From Clerks to Corpora Editors:

From Clerks to Corpora Editors:
Clerks to Corpora
essays on the english language yesterday and today
Philip Shaw
Britt Erman
Gunnel Melchers
Peter Sundkvist
From Clerks to Corpora: essays on the
English language yesterday and today
Philip Shaw, Britt Erman,
Gunnel Melchers & Peter Sundkvist
Essays in honour of Nils-Lennart Johannesson
Stockholm English Studies 2
Editorial Board
Claudia Egerer, Associate Professor, Department of English, Stockholm University
Stefan Helgesson, Professor, Department of English, Stockholm University
Nils-Lennart Johannesson, Professor, Department of English, Stockholm University
Maria Kuteeva, Professor, Department of English, Stockholm University
Nils-Lennart Johannesson is normally a member of the Editorial Board for the
Stockholm English Studies book series. He has however not been involved in the
editorial process of this publication.
Published by
Stockholm University Press
Stockholm University
SE-106 91 Stockholm
Text © The Authors
License CC-BY
Supporting Agencies (funding): Department of English, Stockholm University
First published 2015
Cover Illustration: MS Wellcome 537, f. 15r.
Reproduced by permission of © Wellcome Library, London.
Cover designed by Karl Edqvist, SUP
Stockholm English Studies (Online) ISSN: 2002-0163
ISBN (Hardback): 978-91-7635-004-1
ISBN (EPUB): 978-91-7635-006-5
ISBN (PDF): 978-91-7635-005-8
ISBN (Kindle): 978-91-7635-007-2
Erratum: 2 weeks after publication the following text was added to the title page
of chapter 20 (p367): ’Based on an unfinished manuscript, posthumously edited by
Cecilia Ovesdotter Alm.’
This work is licensed under the Creative Commons Attribution 3.0 Unported License.
To view a copy of this license, visit or send a
letter to Creative Commons, 444 Castro Street, Suite 900, Mountain View, California,
94041, USA. This license allows for copying any part of the work for personal and
commercial use, providing author attribution is clearly stated.
Suggested citation:
Shaw, P., Erman, B., Melchers, G. and Sundkvist, P. (eds) 2015. From Clerks to
Corpora: essays on the English language yesterday and today. Stockholm: Stockholm
University Press. DOI:
To read the free, open access version of this book online,
visit or scan this QR code with
your mobile device.
Stockholm English Studies
Stockholm English Studies (SES) is a peer-reviewed series of monographs and
edited volumes published by Stockholm University Press. SES strives to provide a
broad forum for research on English language and literature from all periods. In
terms of subjects and methods, the orientation is also wide: language structure,
variation, and meaning, both spoken and written language in all genres, as well as
literary scholarship in a broad sense. It is the ambition of SES to place equally high
demands on the academic quality of the manuscripts it accepts as those applied by
refereed international journals and academic publishers of a similar orientation.
Titles in the series
1. Begam, R. and Soderholm, J. 2015. Platonic Occasions: Dialogues on
Literature, Art and Culture. Stockholm: Stockholm University Press. DOI:
2. Shaw, P., Erman, B., Melchers, G. and Sundkvist, P. (eds) 2015. From Clerks
to Corpora: essays on the English language yesterday and today. Stockholm:
Stockholm University Press. DOI:
Acknowledgments vii
Introduction ix
A Personal Tribute to Nils-Lennart Johannesson xiii
Sydney Lamb
1The Middle English Development of Old English y and
Lengthened y: Spelling Evidence 1
Gjertrud F. Stenbrenden
2Linguistic Mysteries Around the North Sea 17
Östen Dahl
3The Late Middle English Version of Practica Urinarum in London,
Wellcome Library, MS 537 (ff. 15r-40v) 35
Javier Calle-Martín
4Is Plant Species Identification Possible
in Middle English Herbals? 53
David Moreno Olalla
5The Periphrastic Subjunctive in the Old English Multiple Glosses to
the Lindisfarne Gospels 71
Marcelle Cole
6On the Place-Name Isle of Dogs 87
Laura Wright
7English Genres in Diachronic Corpus Linguistics 117
Erik Smitterberg & Merja Kytö
Here is an Old Mastiffe Bitch Ø Stands Barking at Mee: Zero Subject
Relativizers in Early Modern English (T)here-Constructions 135
Gunnel Tottie & Christine Johansson
9“Norfolk People Know Best”: On the Written Representation
of Accents as Performed and Perceived by ‘Insiders’ and
‘Outsiders’ 155
Gunnel Melchers
10Sublime Caledonia: Description, Narration and Evaluation in
Nineteenth-Century Texts on Scotland 177
Marina Dossena
11The Development of Attitudes to Foreign Languages as Shown in
the English Novel 193
Philip Shaw
12“Mythonomer”: Tolkien on Myth in His Scholarly Work 215
Maria Kuteeva
13Reflections on Tolkien’s Use of Beowulf 229
Arne Zettersten
14Commentators and Corpora: Evidence about Markers of
Formality 239
David Minugh
15Recent Changes in the Modal Area of Necessity
and Obligation – A Contrastive Perspective 267
Karin Aijmer
16Motion to and Motion through: Evidence from a Multilingual
Corpus 285
Thomas Egan
17Using the World Wide Web to Research Spoken Varieties of
English: The Case of Pulmonic Ingressive Speech 303
Peter Sundkvist
18Another Look at Preposition Stranding: English and Swedish
Discourse Patterns 323
Francesco-Alessio Ursini
19There is Nothing Like Native Speech: A Comparison of Native and
Very Advanced Non-Native Speech 349
Britt Erman & Margareta Lewis
20“Bachelor Means Nothing Without Husband and Father”: What
Collocations Reveal about a Cognitive Category 367
Christina Alm-Arvius (1945–2013)
About the Authors 387
We are grateful, first of all, to the Department of English, Stockholm
University for generously sponsoring this publication. We are also
grateful to many people for help and advice: to Christina Lenz for
prompt and friendly advice throughout the publication process; to
David Minugh for help with proof-reading; and to Ingrid Westin for
knowledge and help. We are also grateful to all those who attended and
presented at the symposium for Nils-Lennart in February 2013 and to
all our contributors.
This volume reports studies of English past and present, all based on
empirical work and illumined by a variety of theoretical and methodical
approaches. Its range and eclecticism makes it a fitting tribute to NilsLennart Johannesson, Professor of English Linguistics, Department of
English, Stockholm University, on the occasion of his 67th birthday. In
addition to knowing him as a wise and supportive colleague, we have
all experienced him as a dedicated teacher and some of us even as a
brilliant fellow-student.
Nils-Lennart first came to our department in the late 1960s as an
undergraduate student. After proceeding quickly through the undergraduate as well as the graduate courses, and having spent a fruitful period of study at Yale University under the wings of Professor
Sydney Lamb, he was ready to present his doctoral dissertation, The
English Modal Auxiliaries: A Stratificational Account, in 1976. After
a brief period as Research Fellow at Lund University, he returned to
our department in 1978, first as Research Fellow, then as Docent and
Senior Lecturer. In 1991 he left the department for the University of
Trondheim, where he served as Professor of English Linguistics for
almost a decade. In the year 2000 he returned to Stockholm, having
been appointed to the Chair at our department in succession to Magnus
It is a daunting task to describe Nils-Lennart’s multi-faceted contributions to the department: as a distinguished and productive scholar,
committed teacher, organizer of symposia, editor, administrator, and
mentor to students as well as colleagues. The following text does not
presume to do justice to his complete oeuvre and achievement.
His scholarly work has above all been concerned with historical linguistics, culminating in a passionate, long-standing commitment to the
Middle English twelfth century homily collection Orrmulum, written
by the Augustinian canon Orm from Lincolnshire. The ultimate aim of
Nils-Lennart’s research on Orrmulum is the production of a new text
edition, based on a new transcription of the existing manuscripts. Since
From Clerks to Corpora
nobody can improve on the creator’s own account of the project, the
reader of this preface is referred to the excellent website Over the years Nils-Lennart has delighted the language seminar at our department with glimpses of his ongoing research within the
Orrmulum project, characterized by linguistic stringency, wide-ranging
use of all technical support, and rich cultural and historical depth.
It should not be forgotten, however, that this is just one of many
interests. Above all, he is a leading expert on Old English syntax, regularly presenting papers at international conferences. He has contributed
greatly to studies on dialect in fiction, with special reference to J.R.R.
Tolkien, and he takes a great interest in linguistic as well as literary
aspects of figurative language, which is reflected in his active role as
co-organizer of our department’s annual ‘Metaphor Festival’ and an
editor of the festival’s proceedings.
He has been a model for us not only as a researcher but also as
a teacher and supervisor. His students are always impressed by his
conscientiousness and creativity, marked by the production and constant improvement of challenging text material for his courses. He is a
painstaking and selfless editor and the co-producer of most handbooks,
festschrifts and conference volumes published by our department (with
the obvious exception of the present volume).
We all know him to be unusually knowledgeable in a number of
other fields: he is a fine musician, known to play the bagpipe as well
as the flute; he is extremely well read, often quoting Galsworthy and
Dorothy Sayers by heart, and has published beautiful translations of
works by the Orcadian writer George Mackay Brown. To at least one
of us he is even known to be very knowledgeable about the intricacies
of traditional knitting.
For a number of years Nils-Lennart served as Head of Department –
a complicated and strenuous function which he carried out in a democratic spirit with fairness, care and great commitment. Many of his
colleagues and students are deeply grateful for his help and advice in
difficult situations, personal as well as academic.
The common thread of the volume is empirical work on English
based on actual data, often from corpora and often diachronic. The first
five chapters deal with Old and Middle English phonology, syntax, lexis
and text editing, forming an overview of current issues in the study of
older stages of the language. Cole shows, very much in the spirit of
Johannesson’s close examination of the Orrmulum manuscript, that the
glosses in the Lindisfarne gospels provide evidence of linguistic change
in progress. Stenbrenden examines a corpus and challenges aspects of
the established phonological history of English. Dahl suggests that the
unexpected similarities between standard Scandinavian languages and
English are evidence for ‘reverse’ influence of English on Old Norse
under Canute’s empire. Calle Martin provides an edition of a previously unedited text. Moreno questions the established interpretation of
a plant name in a Middle English text.
The next six contributions look at Early Modern and Modern
English from a historical point of view, using the variety of methods
associated with Nils-Lennart’s work. Wright looks through everyday
texts for early references (going back to the sixteenth century) that illuminate the etymology of the place-name Isle of Dogs. Smitterberg and
Kytö discuss the complexity of the notion of genre, which they show
is central to historical corpus design. The others all adopt a corpus
approach to samples of more or less literary texts. Tottie and Johansson
examine the development of a non-standard structure in the drama section of the Corpus of English Dialogues, 1560–1760. Melchers looks
at representations of dialect in nineteenth century novels while Dossena
and Shaw both use corpora and closer reading to examine attitudes
to respectively, the Scottish landscape, and foreign languages in nineteenth-century writing.
Nils-Lennart has for many years been interested in the work of J.R.R.
Tolkien and the next two chapters describe aspects of Tolkien’s scholarly work. Kuteeva looks at Tolkien’s concepts of myth and myth study
while Zettersten draws on his own personal friendship with Tolkien to
illuminate the work.
The remaining chapters discuss aspects of Modern English, the first
four once again using corpora, one of the tools which Nils-Lennart
pioneered in the 1980s. Minugh examines aspects of register as revealed
by corpora of different genres. Aijmer and Egan both make use of parallel multilingual corpora. Aijmer looks at developing meanings of
must, and Egan examines a central topic in comparative language studies, the expression of the components of motion, here in English and
French. Peter Sundkvist shows how a phonological topic which cannot
be examined via available corpora can nevertheless be elucidated using
the affordances of the internet.
The last three chapters also cover topics related to Nils-Lennart’s
many-faceted work. He began his career working with Sydney Lamb
on the non-generative modelling of syntax, and this theme is taken
up by Ursini, who shows how preposition-stranding can be handled
From Clerks to Corpora
in type-logical syntax. Later he became interested in pragmatics, and
Erman and Lewis take this up in their paper on the vocabulary use of
advanced second-language users of English, including their use of pragmatic markers. Finally, along with the late Christina Alm-Arvius, NilsLennart was a founder and organizer of Stockholm University’s annual
Metaphor Festival, and it is fitting that the book closes with a chapter
based on her unpublished work.
In view of the richness and quality of the contributions, inspired
by Nils-Lennart’s interest and erudition in various fields of linguistic
and literary studies, this volume will be of interest to a wide academic
audience. Many – if not most – of the contributions offer new and challenging approaches to English studies and linguistics in general, such
as reassessments of the dramatic sound change known as the Great
Vowel Shift and new theories concerning the complicated English–
Scandinavian language contact situation. The importance of English
literature as data for linguistic studies is brought to the fore in many
articles, most of which demonstrate the value of new and ambitious
corpora. The internet, another topical source of linguistic data, is used
in innovative and sophisticated ways by some of the contributors.
It is our hope that this collection of texts – in addition to serving as
a tribute to a much–esteemed colleague – may contribute to inspiration, discussion and research in a number of linguistic fields, such as
historical linguistics, variationist studies, sociolinguistics/dialectology,
syntax, phonetics/phonology, pragmatics, corpus design and studies,
and language typology.
Stockholm, February 2015
Philip Shaw
Britt Erman
Gunnel Melchers
Peter Sundkvist
A Personal Tribute to Nils-Lennart Johannesson
What a pleasure it is to recall Nils-Lennart Johannesson and to learn
that he and his admirable scholarly career are being honored by this
Festschrift. It has been many years now since I have seen him, but I
remember him well from the time he spent with me at Yale working
on his dissertation during the year 1975–76. I recall being repeatedly
impressed by his insight and his diligence as he delved into the rich
complexities of the English auxiliaries, and by his cogent portrayal of
their behavior, enlivened by his well-chosen textual examples, whose
humor often lightened the reader’s load. Although I was the mentor and
he the mentee, he taught me many things about the English auxiliaries,
and he was in fact one of the best students I ever had the pleasure of
working with in my half-century teaching career.
Sydney Lamb
1 The Middle English Development of Old
English y and Lengthened y: Spelling
Gjertrud F. Stenbrenden
University of Oslo
1. Introduction1
The ‘Great Vowel Shift’ is the term used about a set of changes in the
phonetic realisation of Middle English (ME) long vowels, which took
place around 1400–1750 according to the handbooks. In this shift, the
non-close vowels /e:/, /ɛ:/, /a:/, /o:/, /ɔ:/ were raised one step in the vowel
space, and the close vowels /i:/ and /u:/ were diphthongised (Jespersen
1909: 231 ff.; Luick 1914–40: §§479–488; Dobson 1957 passim).
In the late Old English (OE) and early ME periods, changes happened
to the long vowels /y:/, /ɑ:/, and /o:/, as described by e.g. Luick (1914–40:
§§287, 369–370, 406) and Jordan (1968: §§39–42, 44–46, 53–54).
However, these changes are not regarded as part of the ‘GVS’, because
(i) they are said to have been completed before the earliest stages of
the ‘GVS’ took place (the changes to /y:/ and /ɑ:/), and/or (ii) did not
take place in those dialects which later contributed to the phonology
of StE (the fronting of /o:/ in dialects north of the Humber). Critical
voices have been raised, suggesting that the ‘GVS’ started earlier than
textbooks suggest, most notably by Stockwell & Minkova (1988a,
1988b). This paper treats the ME development of OE y– and lengthened
y, for convenience called ‘eME y–’, seeking to establish (a) its phonetic
1 A very early and unfinished version of this paper was read at the conference Historical
Language and Literacy in the North Sea Area, Stavanger, 26–28 August 2009. I am
grateful for valuable comments by Meg Laing, Roger Lass and Merja Stenroos, and
for suggestions from an anonymous reviewer. Any remaining shortcomings remain
my responsibility.
How to cite this book chapter:
Stenbrenden, G. F. 2015. The Middle English Development of Old English y
and Lengthened y: Spelling Evidence. In: Shaw, P., Erman, B., Melchers, G. and Sundkvist,
P. (eds) From Clerks to Corpora: essays on the English language yesterday and today. Pp.
1–16. Stockholm: Stockholm University Press. DOI:
License: CC-BY.
From Clerks to Corpora
developments in the dialects of ME, (b) the approximate dates at which
its various developments started, and (c) whether the said changes were
in fact completed before the ‘GVS’ set in. The answers to these questions may have far-reaching consequences for our interpretation of the
2. Handbooks on the development of OE y
According to standard handbooks, the reflexes of OE /y:/ and /y/ in
lengthening contexts were unrounded to [i:] in late OE or early ME
in “all northern counties”, in parts of the East Midlands, “including
Lincolnshire, Norfolk, and the districts bordering on these counties”,
and in parts of the South-West, “especially Devonshire, Dorsetshire,
and Wiltshire” (Wright & Wright 1928: §57 1; cf. Jordan 1968: §41).
They became [e:] “in Kent and parts of Middlesex, Sussex, Essex, and
Suffolk during the OE period” (Wright and Wright 1928: §57 2; cf.
Jordan 1968: §40). In the remaining areas, i.e. parts of the South and the
West Midlands, the y– remained until the late fourteenth century, when
it was unrounded to [i:] (Jordan 1968: §§39, 42–43; Luick 1914–40:
§§287–288; Wright & Wright 1928: §57 3). Thus, the changes to the
reflexes of eME /y:/ and lengthened /y/ are believed to span a period of
at least three hundred years, even by conventional accounts.
In those dialects where late OE /y:/ was unrounded to [i:], this [i:]
later participated in the ‘GVS’, yielding PDE /aɪ/; an example is OE
hwy– WHY, RP /waɪ/. In those dialects where the /y:/ was unrounded
and lowered to [e:], this [e:] also participated in the ‘GVS’. For instance,
OE my–s MICE became me–s in Kentish, and, after the ‘GVS’, is reflected
as [mi:s] in the modern dialect (Wright & Wright 1928: §57). It should
therefore be possible to infer something about the probable ME reflexes
of eME y– from its modern dialectal pronunciations.
3. Middle English spellings and dialect material
Dialect material in the form of spellings has been extracted from the
Linguistic Atlas of Early Middle English (LAEME), which covers the
period c. 1150–1325 for all of England, as well as from the Survey
of Middle English Dialects 1290–1350 (SMED), and the Linguistic
Atlas of Late Mediaeval English (LALME), which covers the period c.
1350–1450. All tokens for the lexical items listed in the Appendix were
abstracted from all LAEME source texts; from SMED and LALME,
material was extracted for all relevant lexical items.
The Middle English Development of Old English y and Lengthened y
However, ME spelling is not phonetic transcript, so the implied sound
value can only be inferred. Traditionally, <i> and <y> for eME y– are
taken to indicate unrounded [i:]; <u> and <ui/uy> are believed to correspond to a retained front rounded [y:], whereas <e> and <ee> imply
lowering and unrounding to [e:].2 When OE y–-words are spelt with <i>
in late OE or early ME, it seems safe to assume that such spellings do
indeed indicate unrounding, especially if the modern dialect shows /aɪ/,
which is the ‘GVS’ output of ME ı . This assumption is strengthened if
spellings with <y> for etymological ı also appear in the same ME dialects. However, it would be a mistake to view the continued use of <y>
to simply represent [i:] a priori in dialects where the rounded vowel is
believed (in hindsight) to have been retained. In such cases, <y> could
correspond to [y:], although such an interpretation would be highly
improbable if <y> also appears for etymological ı .3 In other words, the
scribe’s entire orthographical system must be taken into account before
his likely pronunciation is inferred, since occasional spellings are by
definition deviations from the scribe’s norm, and may reveal something
about his spoken system.
In those dialects where the OE y– remained front and rounded, i.e. in
parts of the South, and in the West (and Central) Midlands, this vowel
is usually represented by <u>, <ui/uy> in ME – and not by <y> alone –
from around 1100 onwards (Wright & Wright 1928: §57). The use
of <u> for this purpose was made possible when OE u , traditionally
spelt <u>, started to be spelt <ou/ow> during the ME period, due to
French spelling practice (Stenbrenden 2013).
Gradon (1962) cites spellings indicating late OE unrounding of the
reflex of OE y, as well as conditioned rounding of the reflexes of OE i
and ı , in the SW Midlands. Forms with <y> for etymological i in a set
of Exeter documents “are probably to be regarded merely as back-spellings” (1962: 66), based on the merger between OE y and i at [i], but
a number of other such spellings in ten Winchester texts cannot be so
dismissed. More specifically, Gradon claims that OE ı after w seems
2 Anderson (1988) argues convincingly that in Kentish, the reflexes of OE y–/y must
have lowered to [ø(:)] first, before unrounding to [e(:)].
3 For instance, the latter part of the account of Ohthere’s voyage in the OE Orosius,
which is found only in the later, eleventh-century MS (MS Cotton Tiberius B.1),
shows numerous back spellings with <y> for etymological OE ı /i, which suggests
that in late WS, etymological y /y had already been unrounded. Examples of back
spellings are <swyþe> for OE swı þe ‘very’, <scypa> ship, <swyna> swine Such back spellings are absent from the earlier Lauderdale MS, which has
been dated to the first half of the tenth century.
From Clerks to Corpora
to have undergone rounding. Besides, there is evidence that OE y was
unrounded before palatals even in the SW Midlands, whereas it was
retained in other phonetic contexts (1962: 72).
4. Discussion
The extracted LAEME material shows a variety of spellings for eME
y–: <i>, <y>, <e>, <ee>, <eo>, <ey>, <u>, <ui>, <uy>, <yu>, <ou>. Again,
it must be stressed that spellings cannot simply be interpreted as transcriptions of sounds. However, interaction between written norms and
spoken systems must be assumed, resulting in hyper-adaptations, back
spellings, and the like, and when the material is systematised, patterns
emerge. Most LAEME sources show a mixture of spellings for etymological y– which seem to contradict each other in terms of their implied
sound value. A case in point is the text with index number 1300, whose
language has been localised to Suffolk and dated to the second half of the
twelfth century: it has dominant <i> (indicating unrounding), a secondary variant <u> (implying a retained front rounded vowel), and minor
variants <ui> (implying retained [y:]) and <e>, <eo> (implying lowered
and unrounded [e:]). Thus, it is difficult to draw any definite conclusions
from the material. Nevertheless, the following observations can be made.
Unrounding of OE y–/y to [i:] started in late OE and is indicated
in source texts whose language has been localised to Essex, Suffolk
and perhaps Hampshire from the late twelfth century; in sources localised to Oxfordshire, Kent, Northamptonshire and Worcestershire
from the early thirteenth century; in texts localised to Cumberland,
Cheshire, Somerset and Surrey from the mid-thirteenth century; in
sources localised to Lincolnshire, Norfolk, Devon, Gloucestershire and
Herefordshire from the late thirteenth century; and in texts localised
to Ely, Huntingdonshire and the North Riding of Yorkshire from the
early fourteenth century. Thus, unrounding seems to have started in the
South-East and South-West, and to have spread northwards, which goes
against the assumption that the unrounding originated in the North
(Jordan 1968: §41). However, the paucity of ME texts from northern
England from the early ME period precludes any definite conclusion
regarding the locus of this change.
Lowering and unrounding to [e:] is indicated in sources whose language has been localised to Essex and Suffolk from the late twelfth
century; in texts localised to Kent from the early thirteenth century; in texts localised to Somerset and dated to 1240; in sources
The Middle English Development of Old English y and Lengthened y
localised to Gloucestershire and Wiltshire from the second half of the
­thirteenth century; and in a text whose language has been localised to
Lincolnshire from the early fourteenth century. Hence, eME y– > [e:]
seems to have started in the South-East (Kent, Essex, Suffolk), but also
to have taken place independently barely a half-century later in the
South-West. Forms with <e> are dominant in sources whose ­language
has been localised to Kent (the texts with index nos. 8, with a secondary variant <i>, and 142, with minor variants <éé> and <ie>), Essex
(no. 160), Gloucestershire (no. 161), Somerset (no. 156, with <y>
co-varying with <e>), and Lincolnshire (no. 169, also with <y>
co-varying with <e>).
Retained [y:] is indicated in sources whose language has been localised to Berkshire, Essex, Suffolk and Worcestershire from the latter half
of the twelfth century; in texts from Northamptonshire, Herefordshire
and Shropshire from the early thirteenth century; in texts localised
to Cheshire, Gloucestershire, Wiltshire and Surrey from the midto-late thirteenth century; and in sources from Oxfordshire, Ely and
Huntingdonshire from the late thirteenth or early fourteenth century.
The <u> spellings from Berkshire, Essex, Suffolk and Surrey are early,
but they seem to suggest that Wright & Wright (1928: §57) may be wrong
in stating that the reflex of OE y– had become [e:] in Essex and Suffolk in
the late OE period; <e> forms do indeed occur in Essex (text nos. 4, 64,
1200) and Suffolk (text no. 1300), but they are not dominant. Sussex is
poorly represented in the early ME material, but text no. 67 (1200–50),
shows <i>, not <e>, for eME y–. Surprisingly, <u>-type spellings also linger on in the East (Ely, Huntingdonshire) as late as the early fourteenth
century, although the <u> forms here are minor variants.
Regarding retained [y:], the LAEME material seems to also run counter to Wright & Wright’s explicit claim concerning the development of
y– in Wiltshire: dominant <u> in text no. 280 (1250–74) suggests that y–
had not been unrounded in Wiltshire in late OE, but remained rounded.
The same text shows dominant <ou> and <u> for the reflex of eME u ,
and interestingly shows one <ou> for the reflex of eME y– as well, which
indicates a rounded vowel.
Lass & Laing (2005) suggest that, despite what is traditionally
claimed, western ME did not have front rounded vowels, i.e. [y(:)] and
[ø(:)] from OE y–/y and e–o/eo respectively.4 Instead, they maintain that y–
4 Lass & Laing’s claims regarding the reflexes of OE e–/eo will not be addressed here.
From Clerks to Corpora
became [i:] or [e:] or merged with the reflex of eME u in different areas.
That y– changed to [i:] and [e:] is no more than the traditional account,
but Lass & Laing’s claim that it merged with the reflex of eME u
in the SW Midlands certainly needs closer examination. Lass & Laing
use material from LAEME texts 277 (Worcestershire), 272 (Shropshire),
and 280 (Wiltshire) to back up their claims, which is why spellings for
eME y– in these three texts must be investigated in some detail.
The extracted LAEME material for text 272 shows dominant <u>,
and a secondary variant <v> for eME u ; and dominant <u>, and secondary <i>, <e> for eME y–. Likewise, text 277 shows dominant <u>
and secondary variants <ou>, <v>, <o> for eME u ; and dominant <u>,
and minor variants <i>, <eo> for eME y–. Text 280 shows dominant
<ou>, <u>, and minor variants <v>, <o>, <ow> for eME u ; and dominant <u>, and minor variants <i>, <ou>, <eo> for eME y–. Lass & Laing
also claim that there are no instances of <y> for eME y– in the SW
Midlands. Close inspection of all LAEME source texts localised to the
W Midlands reveals that there are, but only for WHY, in text nos. 246
and 1100 from Herefordshire, 2002 from Gloucestershire, and 1600
from Oxfordshire.
Table 1 provides a complete list of all LAEME texts whose language
has been localised to the W Midlands, and their spellings for eME y–
and u . It seems to be true that many W Midlands texts show <u> for
both eME u and y–, but most of them also show different secondary
and minor spellings co-varying for each reflex. For instance, <ou/ow>,
<o>, <uu>, <v>, <ov>, <w> are not infrequent as non-dominant vari—
ants for eME u , whereas such spellings are rare for eME y–. For eME y–,
non-dominant spellings such as <ui/uy>, <e>, <eo>, <i> are more frequent. In some W Midlands texts, the two reflexes appear to be kept
apart; in these, the spellings suggest unrounding (and sometimes
­lowering) of eME y–. Such sources are no. 232 (Oxfordshire, 1175–1224),
no. 189 (Herefordshire, 1200–24), no. 273 (Herefordshire, 1225–49),
as well as no. 161 (Gloucestershire), no. 248 (Herefordshire), and no. 3
(Worcestershire; all 1275–99). Again, most of the W Midlands sources
show a mixture of spellings which often contradict each other in terms
of their implied sound value.
LAEME spellings for lengthened OE y may prove helpful. Most of
the source texts whose language has been localised to the W Midlands
show dominant <u> for the reflexes of OE y in lengthening contexts,
though quite a few show minor <i>, particularly for OE yht, and particularly towards the later period. In other words, lengthened OE y seems
to have remained rounded in most of the W Midlands in the ME period.
The Middle English Development of Old English y and Lengthened y
Table 1. LAEME spellings for eME y– compared to spellings for eME u ;
W Midlands texts only.
Text no. Date
OE ӯ
OE y; -yht
OE ū
u; -uht/uct
ey; -uht/uh
u; -uht
i, u
u; -urht/iht
u; -uht
u; -uht
u, ou
u ((o))†
u ((o, ow, v))
u ((o))
u ((o))
u ((v, e))
u ((v))
u ((v))
u ((v, o, ?eo))
u ((v, o, uu))
u ((v, o, uu))
u (v) ((o, uu))
u ((v))
u ((w))
u ((i*))
u (i*)
i* ((u))
u (i*)
u ((i*, ui))
u ((i*))
u ((i*))
u (i)
u ((i*))
u ((i, ui))
u (i*) ((ui))
u (i*)
i* (u)
((uu, ei))
u ((i, ui))
u, ui
u, i*
u ((i*))
u ((ou, i*, eo))
u, ui, i*
u (i*)
u (ui, y*, i*)
u; -uht/iht
e, u
u; -ucht/ugh
u, i; -iþ(t)/iht
ey; -eyt/eʒt
u, i; -ist
u ((v, o))
ou, u
ou ((u, o, ow))
u ((v, i, o))
ou, u ((v, o, ow))
ou (u) ((o, ow))
u ((o, v, w))
u ((ou, o, uu))
u, ou, ow, ey
ou ((o, u))
u ((o))
u ((ou, o, v, oi,
uu, w, eo, eþ))
From Clerks to Corpora
Table 1. Continued
Text no. Date
OE ӯ
OE y; -yht
OE ū
u, ui, i*
i (u)
u, uy; -it
ou ((ov, o, u))
u ((ou, o))
u ((o, ou, e))
u ((i*))
u ((i, v))
u; -iht/uht/(ih)
u, i; -ih(t/uht)
u (i*) ((y*))
ui (i*)
((y*, u))
u, i
u, uy
((i, y, yu))
u ((v, ou, o))
u ((ou, o, v, eo,
eu, eou, uo, e,
u; -iht/yht/uht u ((v, w, ow/ou,
o, uo, uw, e, ey))
ui, u, ey; -iʒt/
ou ((o, ow, u, e,
u, ei
u ((o))
u, uy (ui); -iʒht ou (o) ((u, ov,
ow, v, uy))
ou, o
ov ((u, v, o))
‡ Suggested dates for source texts follow LAEME: the first number refers to century; ‘a’ and ‘b’
refer to the first half and the second half of the century, respectively; and ‘1’ and ‘2’ refer to the
first and second quarter of each half-century.
† Use of round brackets to enclose non-dominant spellings follows LALME practice: single
brackets enclose secondary variants, and double brackets enclose minor variants.
* An asterisk indicates that the text has <i> or <y> for why and/or OE by–sen only.
Exceptions are found in the following sources: no. 260 (Shropshire,
1200–24) has dominant <ey> for lengthened OE y other than yht (for
which <uht>, <uh> are found); no. 158 (Gloucestershire, 1250–74) has
<e> as well as <u>; no. 161 (Gloucestershire, 1275–99) shows only
<ey> for lengthened OE y; no. 10 (Gloucestershire, 1275–1324) has
<u> and <ei>. All of the preceding forms may point to lowering and
unrounding of y– to [e:]. However, generally speaking, the reflex of OE
yht appears to be in the process of merging with the reflex of OE iht.
Thus, examination of spellings from all LAEME sources localised to
the W Midlands corroborates at least two of Lass & Laing’s claims: (a)
there is no “neat geographical tri-partition for /y/”, and (b) not “only
are the symbol-to-sound mappings more multiplex than is suggested,
The Middle English Development of Old English y and Lengthened y
but there is a strong element of lexical specificity in the set of reflexes”
(Lass & Laing 2005: 281). In other words, certain lexical items seem to
be spelt in certain ways, which supports the theory of change by lexical
diffusion (Phillips 2006a, 2006b): sound changes seem to start in certain
phonetic contexts, and/or in very frequent words, whence they spread
from context to context, and from frequent to less frequent words. Their
third claim, that there are no “particular spellings uniquely associated
with OE /y(:)/” (ibid.), is not entirely correct, as <ui/uy> seem to be used
for the reflex of eME y– only.5 It is their assumption that the reflex of
eME y– fell in with the reflex of eME u in the SW Midlands which is
the most difficult to prove. ME spellings seem to lend support to their
view: Table 1 shows clearly that <u> remains dominant for the reflexes
of both eME y– and u in the SW Midlands up until the last quarter of the
thirteenth century, although <ou/ow> slowly take over as the dominant
spellings for eME u from c. 1250. A very few examples of <ou> for the
reflex of eME y– appear to support the merger hypothesis also.
Evidence from the modern dialects may settle the issue: if presentday SW Midland accents show a merger between the reflexes of eME
u and y–, the ME spellings may be taken to indicate just that; if not,
the same spellings simply show a lack of distinct representations on
the orthographic level, but not phonetic or phonological merger. The
Survey of English Dialects (Orton & Dieth 1962) yields material for
the lexical items listed in the Appendix for the W Midlands, and for
Wiltshire and Berkshire. This phonological material from the modern
dialects implies that there was no merger between the reflexes of eME
y– and u in the W Midlands, since generally eME y– is reflected as a num—
ber of diphthongs with unrounded off-glides, and eME u is reflected as
a series of diphthongs with rounded off-glides. However, exceptions are
found in some localities in Cheshire and Staffordshire:
Locality 1: merger in [aɪ] or [ɑɪ];
Locality 2: merger in [æɪ] for some words;
Locality 3: merger in [ɛɪ] for some words;
Locality 4: merger in [aɪ];
Locality 5: merger in [ɛɪ] or [aɪ].
Locality 2: merger in [ɛɪ];
5 Only text no. 1600 (Oxfordshire, 1275–1324) has <uy> for the reflex of eME u .
From Clerks to Corpora
Localities 7–9: merger in [ɒɪ];
Locality 10: minimal distinction between [ɑ:ɪ] and [a:ɪ];
Locality 11: merger in [aɪ].
That is, in these localities, there is sometimes a diphthong with an
unrounded off-glide for the stressed vowel of cows/kyes (OE cy–), mouse,
house, mouth, drought, thousand, clouds (all with eME u ), as well as
for dry, hide, fire, wright, why (all with eME y–). Interestingly, these
have converged on the ‘normal’ reflex for eME y– and not for eME u .
Thus, the ME spellings from the SW Midlands likely indicate merger of
spellings, but not of sounds, as otherwise it would have been impossible
for the reflexes of the two sounds to be distinguished again later on an
etymologically correct basis. The only way in which merged sounds
could unmerge would be if two different systems co-existed, one of
which kept the reflexes apart, and the distinction was re-introduced
into the system in which merger had taken place. But such a scenario
remains speculative and unlikely.
Interestingly, even in many LAEME source texts with dominant <u>
for eME y–, there is only <i> for the lexemes WHY and OE by–sen ‘example’. This seems to suggest that if the unrounding of y– started in any one
word, that word is most likely WHY, although it should be noted that
there was an OE variant whie.6
5. Summary of early ME material
The material from LAEME suggests the following developments for
the reflex of eME y–:
1. It was unrounded to [i:] across the country, and this process
seems to have started in the (South-)East and the W Midlands.
2. It was lowered and unrounded to [e:] in Essex, Suffolk and
Kent in the late twelfth or early thirteenth century; in Somerset,
Gloucestershire and Wiltshire in the latter half of the thirteenth
6 Likewise, if the lowering and unrounding of y– to [e:] started in a specific word, it may
have been in the OE word þy–ster ‘dark, gloomy’ and cognates, since there are frequent
<e> and <eo> for these lexemes even in texts with dominant or exclusive <u>, <ui/uy>
for all other words with eME y–. However, OE þy–ster and cognates also appear with
OE ie, eo, so ME <e>, <ie> etc. for these probably go back to forms which did not have
OE y–, since they show a very distinct pattern in the extracted material. For this reason,
ME spellings for OE þy–ster and cognates have not been included in Table 1.
The Middle English Development of Old English y and Lengthened y
century; and perhaps in Lincolnshire in the early fourteenth
3. It remained as [y:] in the W Midlands and parts of the SouthWest, as well as in Ely and Huntingdonshire (at least as a minor
variant) in the late thirteenth or early fourteenth century.
With regard to Lass & Laing’s hypothesis concerning the phonetic
nature of the ME reflex of OE y–, the modern material does not support
their claim that the reflex of eME y– had merged with the reflex of eME u
in the ME dialects of the W Midlands, even though <u> is the dominant
spelling for both in ME in the area in question.
6. The later ME material
The LAEME material for eME y– needs to be tied up with the later
material from SMED and LALME. Conclusions based on analysis of
this material are briefly summarised below.
Kristensson (SMED1–5) finds that, in the fourteenth century, all the
Northern counties except two were [i:]-areas. The case for the West
Riding of Yorkshire and Lancashire is less straightforward: Kristensson
concludes that they most likely had [i:], but that [y:] was also used in
Lancashire south of the Ribble. As for the W Midlands, [y:] was found
in Cheshire, Staffordshire, Shropshire, Herefordshire, Worcestershire,
Warwickshire, Gloucestershire and Oxfordshire. Derbyshire also had [y:],
except in the easternmost tip, which had [i:]. Leicestershire seems to have
had [y:] in the west and [i:] in the east and south. Nottinghamshire had [i:].
In the E Midlands, Rutland, Huntingdonshire and Norfolk had [i:],
Bedfordshire, Hertfordshire, Middlesex and Buckinghamshire had [y:],
Suffolk and Essex had [e:]. Northamptonshire had [i:] in the northern
half, [y:] in the southern half. Kristensson thinks Cambridgeshire had [i:]
north of the city of Cambridge (including Ely); south of it, it had [e:] to
the east and [y:] to the west. In the South, Devon seems to have had [i:]
and [y:]; Somerset, Dorset, Wiltshire, Hampshire, Berkshire and Surrey
had [y:] (although Surrey may have had [e:] in the easternmost tip).
Sussex had [e:] in the east, [y:] in the west; Kent had [e:].
The development of eME y– before palatals appears to have been different from that in other contexts: in this environment, y– was unrounded
7 Perhaps more likely, lowered and unrounded [e:] spread occasionally to Lincolnshire
from Suffolk through Norfolk, although in that case, one would expect <e>-type
spellings in Norfolk too. That is, however, not the case.
From Clerks to Corpora
to [i:] in Hampshire and Dorset, and also in Berkshire, Wiltshire and
Somerset. There are traces of such a development in Surrey also, but
not in Sussex or Kent. Finally, Devon shows a tendency to unrounding
regardless of phonetic context.
Ek (1972) investigates the ME development of OE e–o and y– in the
South-East, using onomastic material which partly overlaps with that
of SMED, although much of Ek’s material is earlier. His conclusions
differ somewhat from Kristensson’s, particularly regarding the extent
of the [e:]-area. However, Kitson (1998: 170) concludes that since Ek’s
material is earlier, “what the two investigations show between them is a
retreat of the e-reflex in favour of the u-reflex as well as, further north,
the i-reflex”.8 In other words, Ek’s and Kristensson’s data demonstrate
change in progress.
The LALME material suggests that the reflexes of eME y– had been
unrounded to [i:] in the East and the North in the late ME period.
Retained [y:] is implied by spellings from the West and the South, and
from the West Riding of Yorkshire, whereas <e>-type spellings linger on in most of the East, and in parts of the West (Gloucestershire,
Worcestershire, Oxfordshire), the South-West (Devon, Dorset,
Hampshire, Wiltshire) and the South-East (Kent, Sussex, Surrey). That
is, <i>-type and <e>-type forms co-vary in the East, but whether these
systematically correspond to [i:] and [e:] is an open question.
7. Conclusions and implications
The three corpora exploited all contain spellings which seem to indicate
that the ‘GVS’ started much earlier than 1400, perhaps around 1250 or
1300 (Stenbrenden 2010). For eME y–, for instance, LAEME text 142
(1275–99), whose language has been localised to Kent, shows dominant
<e>, but has one <i-hierde> for the 3.pres.sing. of OE hy–ran HIRE v.
(beside three tokens with <e> for the same), which may indicate raising of
e– to [i:]. Source text no. 2002 (1275–99), whose language has been localised to Gloucestershire, has dominant <ui> and a secondary variant <i>
as well as minor variants <y> and <u>. It has <flei> for FLY (noun); this
may however go back to OE fle–oge or fly–ge. For OE y in lengthening contexts, text no. 137 (Cambridgeshire, 1275–99) has <bein> for OE bycgan
8 Wyld’s two articles (Wyld 1913–14a, 1913–14b) on the dialectal development of
eME y– are classics, but have not been included in the discussion in the present
paper. For a full account, see Stenbrenden (2010).
The Middle English Development of Old English y and Lengthened y
BUY (verb); this text has <i> for the long vowel, so vowel-shift may be
inferred. Text no. 269 (Norfolk, 1275–1324) likewise shows <beyn> for
OE bycgan, but there are no tokens for the long vowel in this source text,
so it is difficult to assess whether this form indicates vowel shift or not.
Generally, there are more irregular spellings which may indicate vowel shift for the lengthened OE y than for the etymologically
long vowel, which may suggest that the vocalisation of post-vocalic
–h, –g in late OE produced a minimal diphthong [ɪi] (rather than a
long monophthong). This diphthong would have been an allophone
of /i:/ (since words with OE –iht, –ig and –yht subsequently had the
same development as OE ı -words), and may have triggered the vowel
shift, as suggested repeatedly by Stockwell (1964, 1972, 1978), and by
Stockwell & Minkova (1988a, 1988b). LAEME, SMED and LALME
also contain irregular spellings indicating early vowel shift of eME u ,
e– and o– (Stenbrenden 2010), lending support to this ‘Early Vowel Shift
It must be concluded, therefore, that there is a long temporal overlap between the constituent ‘GVS’ changes and the assumed earlier set of changes. Consequently, the two sets of changes cannot be
treated separately: the changes to early ME y– must be seen as part
of the Shift, as must the changes to OE a– (south of the Humber)
and OE o– (north of the Humber). This points to a very lengthy
period of long-vowel shifting, from c. 1100 (or earlier) to c. 1750,
which raises the question, Is it possible for one unitary and coherent
‘Great Vowel Shift’ to take place over 650 years or more? Clearly
not. Rather, Stockwell seems right when he states that “the series of
changes of which the GVS is a part have been going on at a remarkably steady rate for more than 1500 years” (Stockwell 1969: 93),
a claim which has not received much support until now.
A. LAEME: lexical items with OE y–/y searched for in all source texts
For OE y–:
4scy–te, a–ly–fedlı c, a–ly–man, a–ly–sedness, a–ly–send, a–ly–sendness, behide,
bride, bridegift, bridegroom, by–sen, by–snian, cy–þan, dry (adj., v.),
dry– (n.), dryfoot, fire, fireburning, hide (n. and v.), hire (n. and v.),
hireman, ly–t, ly–þerlı ce, ly–þerness, ly–þre, ly–tlian, ly–tlum (adv.), pride,
þwy–rian, þy–fel, þy–ster, þy–sterness, þy–strian, þy–strig, why
From Clerks to Corpora
For lengthened OE y:
bitight, buy, buyer, crockwright, flight, forbuy, fright, frighten,
frighty, frightyhood, frightyly, hyht, hyhtan, hyhtlı c, kind (n.), mankind,
offrighten, wright
B. The Survey of English Dialects: questionnaire items examined for the
W Midlands, and for Wiltshire and Berkshire
For OE u and lengthened u:
plough, cow(s), sow, snout, mouse, boughs, house, mouth thousand,
clouds, drought
For OE y– and lengthened y:
dry (III.1.9), hide (noun), fire, dry (VII.6.19), wright, why
Anderson, J. (1988). The great kEntish collapse. Kastovsky & Bauer (eds)
1988: 97–107.
Dobson, E. (1957). English Pronunciation 1500–1700, Vol. I. Oxford: Clarendon.
Ek, K-G. (1972). The Development of OE y– and e–o in South-Eastern Middle
English. Lund: Gleerup.
Gradon, P. (1962). Studies in Late West-Saxon Labialization and Delabialization.
Davis, N. & Wrenn, C.L. (eds), English and Medieval Studies Presented to
J.R.R. Tolkien on the Occasion of his Seventieth Birthday: 63–76. London:
George Allen & Unwin.
Jespersen, O. (1928). A Modern English Grammar, Part I. (4th ed.) Heidelberg:
Carl Winters Universitätsbuchhandlung.
Jordan, R. (1968). Handbuch der mittelenglischen Grammatik: Lautlehre.
(3rd ed.) The Hague: Mouton.
Kastovsky, D. & Bauer, G. (eds). (1988). Luick Revisited. Tübingen: Gunter Narr.
Kitson, P. (1998). Review of Kristensson 1995. NOMINA 21: 169–178.
Kristensson, G. (1967). A Survey of Middle English Dialects 1290–1350. The
Six Northern Counties and Lincolnshire. Lund: CWK Gleerup.
———. (1987). A Survey of Middle English Dialects 1290–1350: the West
Midland Counties. Lund: Lund University Press.
———. (1995). A Survey of Middle English Dialects 1290–1350: the East
Midland Counties. Lund: Lund University Press.
The Middle English Development of Old English y and Lengthened y
———. (2001). A Survey of Middle English Dialects 1290–1350: the Southern
Counties I. Vowels. Lund: Lund University Press.
———. (2002). A Survey of Middle English Dialects 1290–1350: the Southern
Counties II. Diphthongs and Consonants. Lund: Lund University Press.
LAEME: Laing, M. (2008). A Linguistic Atlas of Early Middle English.
University of Edinburgh. URL:
LALME: McIntosh, A., Samuels, M., Benskin, M. et al. (1986). A Linguistic
Atlas of Late Mediaeval English, Vols. I–IV. Aberdeen: Aberdeen University
Press. URL:
Lass, R. & Laing, M. (2005). Are front rounded vowels retained in West
Midland Middle English? In Ritt, N. & Schendl, H. (eds) Rethinking Middle
English: Linguistic and literary approaches: 280–290. Frankfurt am Main:
Peter Lang.
Luick, K. (1914–40). Historische Grammatik der englischen Sprache, Vol. I.
Oxford: Basil Blackwell.
Orton, H. & Dieth, E. (1962). The Survey of English Dialects. Leeds: Arnold.
Phillips, B. (2006a). Word Frequency and Lexical Diffusion. Basingstoke:
Palgrave Macmillan.
———. (2006b). Word Frequency Effects in the Great Vowel Shift. Paper read
at the 14th International Conference on English Historical Linguistics,
Bergamo, Italy, 21–25 August 2006.
SMED: see Kristensson 1967, 1987, 1995, 2001, 2002.
Stenbrenden, G.F. (2010). The Chronology and Regional Spread of LongVowel Changes in English, c.1150–1500. PhD dissertation, University of
———. (2013). The diphthongisation of ME ū: the spelling evidence. Andersen,
G. & Bech, K. (eds) English Corpus Linguistics: Variation in Time, Space
and Genre: Selected Papers from Icame 32 (Language & Computers): 53–
67. Rodopi.
Stockwell, R.P. (1964). On the Utility of an Overall Pattern in Historical
English Phonology. Proceedings of the Ninth International Conference of
Lingustics: 663–671. The Hague: Mouton.
———. (1972). Problems in the Interpretation of the Great English Vowel
Shift. Smith, M.E. (ed.) Studies in Linguistics in Honor of George L. Trager:
344–362. The Hague: Mouton.
From Clerks to Corpora
———. (1978). Perseverance in the English vowel shift. Fisiak, J. (ed.) Recent
Developments in Historical Phonology: 337–348. The Hague: Mouton.
Stockwell, R.P. & Minkova, D. (1988a). The English Vowel Shift: problems of
coherence and explanation. Kastovsky & Bauer (eds) 1988: 355–394.
———. (1988b). A rejoinder to Lass. Kastovsky & Bauer (eds) 1988: 411–417.
Wright, J. & Wright, E.M. (1928). An Elementary Middle English Grammar.
Oxford: Oxford University Press.
Wyld, H.C. (1913–14a). The Treatment of OE. y– in the Dialects of the Midland,
and SE. Counties in ME. Englische Studien 47: 1–58.
———. (1913–14b). Old English y– in the Dialects of the South, and South
Western Counties in Middle English. Englische Studien 47: 145–166.
2 Linguistic Mysteries Around the North Sea
Östen Dahl
Stockholm University
Everyone who knows anything about the history of the Germanic languages knows that English was influenced by Scandinavian around
the Viking Age. The question I will raise in this paper, without giving
any definite answer, is: Did the contacts across the North Sea have any
effect on the Scandinavian languages? As far back as ten years ago,
Maria Koptjevskaja Tamm and I asked the same question in a conference presentation that never resulted in a published paper (but see
Dahl 2010a, 2010b for some discussion). Our point of departure was
the striking similarities between the possessive constructions called
“­s-­genitives’ in English and the Scandinavian languages, and the observation that in Scandinavia, s-genitives are by and large found only in
standard Danish, standard Swedish, and some Danish-influenced varieties of Norwegian. Moreover, it seemed that the origin of s-genitives
was in western Denmark, next to the North Sea.
In this connection, two papers by the Danish Scandinavianist Kristian
Ringgaard seemed relevant (Ringgaard 1986, 1989). Ringgaard argues
against the common view that the simplification processes that took
place in the Danish inflectional system in the Middle Ages were due
to the intensive contacts with Low German in the Hanseatic period.
Referring to earlier work by Anders Bjerrum, he argues that these processes started much too early to be triggered by the influx of German
merchants, whose culmen was in the second half of the 14th century.
He also argues that there were significant differences within the Danishspeaking area as to the strength and chronology of these processes, with
a cline Jutland>Island Danish>Scanian. In the second paper, Ringgaard
focuses on Jyske Lov ‘The Jutlandic Law’, one of the provincial Danish
laws, signed by King Valdemar II in 1241. Ringgaard’s textual source is
How to cite this book chapter:
Dahl, Ö. 2015. Linguistic Mysteries Around the North Sea. In: Shaw, P., Erman, B.,
Melchers, G. and Sundkvist, P. (eds) From Clerks to Corpora: essays on the English language yesterday and today. Pp. 17–34. Stockholm: Stockholm University Press. DOI: License: CC-BY.
From Clerks to Corpora
the Flensburg manuscript, which was at the time considered the oldest
version, dated to 1300; later, another manuscript, Codex Holmiensis,
has been reliably dated to 1271. Ringgaard describes the language in
Jyske Lov as having among other things the following features – all
innovations relative to the language found in Runic inscriptions:
two genders for nouns (residual feminine gender)
no nominative-accusative distinction
generalized s-genitive
dative only in lexicalized expressions
definite articles are rare and always postposed
no case inflections in adjectives
participles tend to lack a neuter form
examples of singular verb forms with plural subjects occur
No exact date for when the law was formulated can be given;
Ringgaard notes that the version that existed in 1241 would have been
written by people born around 1200, even if they may have relied on
earlier legal rules. He notes that modern dialects in western Jutland
show features that mark it off as a “radical innovation area”. What
we see then is that a language with a surprisingly “modern” grammar
must have been spoken somewhere in western Denmark well before
Hanseatic contacts could have had an effect on the language.
Ringgaard is himself slightly baffled by his findings; he wonders why
innovations would spread from western Jutland, which was not the
centre of the realm in historical times, and speculates that in fact the
changes started much earlier, perhaps as early as the period 100–500 CE.
But perhaps part of the explanation could rather be found on the other
side of the North Sea.
The story of the Scandinavian presence in Britain has been told many
times, but much of what really happened is still in dispute. Viking raids
began already in the 8th century; they were undoubtedly spectacular
but had hardly any bearing on linguistic developments. Towards the
end of the 9th century, more ambitious attempts of military takeover
began, and in 886 the Danes and the kingdom of Wessex made an
accord in which the Danes were given control over a large part of
northern and eastern England (the “Danelaw”). During the period
that followed, there was considerable migration from Scandinavia to
the Danelaw areas, the extent and nature of which have been much
debated, as we shall see later. What is not being questioned, however, is
Linguistic Mysteries Around the North Sea
that the Scandinavian settlements in the Danelaw were the basis for the
impact that Scandinavian languages have had on English. Around the
middle of the 10th century, the Danes had lost their political control of
the Danelaw, and the Scandinavian parts of the population were presumably being integrated with the Anglo-Saxon elements, although it is
not possible to judge how fast this process was.
Towards the end of the century, the Danes renewed their attacks
on England, as a result of which the English were forced to pay large
tributes (“Danegeld”). On November 13, 1002 (“St. Brice’s Day”), the
English king Æthelred “the Unready”, in his own words, sent out a
decree “to the effect that all the Danes who had sprung up in this island,
sprouting like cockle amongst the wheat, were to be destroyed by a most
just extermination” (as translated in Whitelock 1996: 545). From the
modern point of view, this seems like a surprisingly candid admission of
ethnic cleansing. It is unlikely that it was directed against the population
of the Danelaw; rather, it concerned recently arrived Danes, but it indicates that they must have been present in significant numbers. Æthelred
did not attain the goal of ridding England of Danes; on the contrary,
after a decade of continued raiding, the Danish king Sweyn Forkbeard
managed to secure the English throne in 1013, but died shortly thereafter. After three years of political unrest, his son Cnut became King
of England and ruled it together with Denmark and Norway until his
death in 1035, when he was succeeded by his son Harthacnut whose
death in 1042 marked the end of Danish rule in Britain. The epilogue
came 25 years later, after the Norman invasion, with “the Harrying
of the North”, another campaign, the character and extent of which
are controversial, but which is claimed by contemporary sources to
have laid large parts of the Danelaw waste, with possibly as many as
a ­hundred thousand people killed.
Turning now to the main issue of the paper, the first observation to
be made here is that there is a similarity between the histories of English
and East Nordic (Danish and Swedish) in that there is a hiatus in the
beginning of the second millennium CE, during which there is very
little written documentation of the languages. Furthermore, when the
languages started being used in writing after the hiatus, they seem to
have undergone, or started to undergo, quite similar simplifications in
their morphology. It is also during and around this period that contacts
across the North Sea are at their peak.
The extent to which Scandinavian was spoken in England, as well as
its impact on English, has been much disputed. To start with, there are
From Clerks to Corpora
different views on the length of the migration period. Thomason and
Kaufman (1988: 267) say that “Norse speakers settled in the North
and East of England during the period from 865 to 955 (though not
in the East after 920)”. Baugh and Cable (1993: 96), quoted approvingly by Emonds and Faarlund (ms.), say that “[u]p until the time of
the Norman Conquest the Scandinavian language in England was constantly being renewed by the steady stream of trade and conquest…”
and consequently, “many of the newcomers … continued to speak their
language at least as late as 1100”. Emonds and Faarlund agree that
“we can say with some certainty that at the outset of Norman rule, the
Danelaw contained many speakers of two distinguishable languages,
one of them being Norse”. Thomason and Kaufman (1988: 288), on
the other hand, think that Norse had disappeared from large parts of
the Danelaw before the end of the first millennium although it was
spoken in present-day Yorkshire until about 1015. Similarly, they say
(1988: 302) that the influence of Norse on English “was pervasive, in
the sense that its results are found in all parts of the language; but it
was not deep, except in the lexicon”. For Emonds and Faarlund (ms.),
on the other hand, “influence” is the wrong word, since Old Norse is in
their view the mother language of Middle and Modern English.
For the understanding of what happened to English and the North
Germanic languages during and around the Viking age, it is important
to keep apart two major types of contact-induced change: one is transfer of features, leading to increased similarity between languages; the
other is simplification due to imperfect learning by second-language
speakers. The latter type is of particular interest since both English and
Mainland Scandinavian underwent quite significant simplification processes in their morphology. As noted above, the breakdown of the old
inflectional system in Mainland Scandinavian has been attributed to
contact with Low German, but Ringgaard argued that this would be
the wrong time and the wrong place. For English, three possible culprits
have been proposed: Norman French, Celtic, and Old Norse. Trudgill
(2010) invokes what he calls “sociolinguistic typology” and the suggestion that the structure of a language may be at least partly determined by
the type of social environment and social structure in which it is spoken.
Thus, he claims, long-term contact situations influencing child language
acquisition will tend to lead to complexification through the addition
of features from other languages, whereas short-term contact involving adult language learning tends to lead to simplification, due to the
problems adults have in coping with irregularity and nontransparency.
Linguistic Mysteries Around the North Sea
Norman French, like Low German in the Scandinavian case, can be
acquitted since it would be the wrong time – the changes begin earlier –
and the wrong place – the changes start in the north, where there were
few Normans. As for the choice between Celtic and Old Norse, several
scholars have recently suggested that Brittonic Celtic may have survived
much longer than was earlier thought (Tristram 2002, 2004; Schrijver
2006; Laker 2008; Trudgill 2010). Furthermore, the sociolinguistic
situation makes simplification much more probable in the case of the
contact between the Anglo-Saxons and the Celts than in the contact
between the Anglo-Saxons and the Vikings. In the former case, a large
population submitted to a relatively small group of intruders; in the
latter case, the situation would be the reverse. Tristram (2004) hypothesizes that there was a long period of diglossia, in which the majority of
the population spoke “Brittonic English” and only the aristocracy the
“purer” variety of Old English. The variety that the Norse immigrants
mainly met could well be Brittonic English, in which categories such
as gender and case would already have disappeared due to imperfect
learning. The role of the contact with Norse would then be at most to
strengthen the processes that had already been initiated earlier.
Against this background, let us now consider what happened at
roughly the same time or a bit later in Scandinavia. As already noted,
many linguistic changes in the Middle Ages seem to have originated
in western Denmark (Jutland) and progressed east. In fact, some of
them continued beyond what was then considered as Denmark (that
is, including the southernmost Swedish provinces). However, as we get
further away from the point of origin, changes tend to be later and
weaker. The regions in Mainland Scandinavia that show most resistance are Norway, northern Sweden and the trans-Baltic areas. Island
Scandinavian (Icelandic and Faroese) were often not reached at all. The
striking observation is now that in a number of cases, these developments have close parallels in Britain. Let us look at the most important
of them.
Reduction of unstressed vowels and apocope
This is a type of phonological change that took place in large parts of
the Germanic-speaking area and also has parallels in some Romance
languages. It has been blamed on earlier prosodic changes (initial wordstress) and has also been used to explain the general simplification of
the inflectional systems in West European languages. There are ­basically
From Clerks to Corpora
two possible outcomes of this process: (1) unstressed vowels (particularly final ones) are reduced to e or schwa; (2) these vowels disappear
altogether (apocope or “schwa loss”). During the Middle Ages, the first
possibility was realized generally in West Germanic and in western and
central Danish – but only to a limited extent in Peninsular Scandinavian;
the second outcome is found in a smaller area, according to Thomason
and Kaufman (1988: 319): “English, Dutch, some Frisian, some Low
German, and some High German”. Crucially to our discussion, “some
Danish” should be added to this list, more specifically West Jutlandic.
In English, the reduction process started already in the OE period. The
apocope started in the north and was already spreading to the south
in the period 1100–1250 (Minkova 1991: 30). According to Perridon
(2005: 1023), all unstressed final vowels were reduced to æ or e in
western and central Danish as early as “in the 12th century, or even earlier”. Loss of final vowels is attested in early manuscripts such as Jyske
lov – that is, more or less simultaneously with the spread to southern
England – and has been generalized in modern Jutlandic dialects. The
timing of the process in Danish would allow for influence from the
south for the vowel reduction but hardly for the apocope, which seems
to have taken place later in the neighbouring languages, to the extent
that it took place at all (Ringgaard 1986: 182).
The case system
As noted above, the breakdown of the old noun case system had already
gone quite far in the language of 13th century western Denmark as represented in Jyske Lov. The distinction between nominative and accusative had disappeared almost wholly, with some remnants in adjectives,
and the dative was used almost exclusively in lexicalized prepositional
phrases (Bjerrum 1966: 58). (The genitive will be discussed below.) In
Peninsular Scandinavian, the process was much later, and has not yet
been brought to completion in some dialects in the north.
Looking at Middle English, we find that the situation in Jyske Lov
comes very close to the way the mid-12th century text in the Peterborough
chronicles is presented in Burrow & Turville-Petre (1991: 29), where
the nominative and accusative have no endings, and the dative in -e of
strong nouns is said to sometimes occur after prepositions, “but is as
often uninflected” and “…in later texts the inflexion is dropped altogether except in a few phrases”. Likewise, case inflections in adjectives
were preserved only in the South “to a limited extent”.
Linguistic Mysteries Around the North Sea
It is interesting to contrast this with the Old Danish letter from
late 14th century Halland (then part of Denmark, now a province of
Sweden) quoted in Ringgaard (1986: 181), where “there is not much to
object to the inflectional system”, that is, the old system is more or less
fully preserved – as it also was in Middle Low German (a distinction
nominative-oblique is still alive in present-day Low German). In other
words, 13th century western Danish aligns more closely with 12th
century Midland English than with 14th century eastern Danish and
with its neighbours to the south.
The genitive
In many Germanic languages, the genitive case has undergone developments that set it off from the rest of the case paradigm. Thus, one
single case suffix, ‑(e)s, which originally marked the genitive singular of
non-feminine nouns, has survived even in the languages where the case
system has otherwise broken down but is used in innovative possessive constructions and has been generalized to feminine and sometimes
plural nouns. As noted above, the s-genitives of English, Danish and
Swedish share a number of properties, in particular: a uniform suffixed
marker -s which (i) can be used with all types of noun phrases; (ii) is
added to the last word of the possessor NP (rather than to the head
noun); (iii) always precedes the head noun of the possessee NP; (iv) is
used with a possessee NP which lacks definiteness marking. Properties
(i–ii) make these s-genitives different from similar constructions in West
Germanic, which tend to be restricted to proper and kinship nouns and
mostly do not appear in complex NPs. Properties (iii–iv) distinguish
them from constructions found in some Swedish traditional dialects,
where the possessee NP takes a definite suffix and the genitive phrase
may follow the head noun of the possessee NP. It should be noted
that s-genitives are virtually non-existent in the traditional dialects of
Northern Scandinavia, where a plethora of other possessive constructions are used instead (Delsing 2003; Dahl 2010a).
Ringgaard’s statement about Jyske Lov, “the s-genitive was generalized”, has to be modified – it was rather “generalizing”. Perridon
(2013) made a thorough investigation of a number of manuscripts and
found a rather complex pattern of variation. In particular, the genitive
is sometimes zero-marked, as in sankte knut sun ‘St. Cnut’s son’. In
definite nouns, forms such as landæns ‘the land’s’ competed with forms
with double genitive marking like landzæns. In a later development, a
From Clerks to Corpora
“linking pronoun construction” – parallel to the much debated English
type John his house – gained ground in western Denmark and is preserved in modern Jutlandic, while the s-genitive was further generalized
in more eastern varieties.
Middle English is usually said to have a single genitive ending -es
but zero marking is also found. A particularly striking attestation is
found on the Kirkdale sundial, dated to 1060, which contains the
phrases in Eadward dagum ‘in Edward’s days’ and [i]n Tosti dagum
‘in Tosti’s days’. Thomason and Kaufman (1988: 289) note that this is
one trait found in the sundial text that characterizes Northern Middle
English although “the writers of the text were evidently trying to write
Standard West Saxon”. Klemola (1997) argues on the basis of modern dialect data and quotations from earlier scholars that there was
a loss (apparently general) of the genitive endings in a northern area
including Yorkshire. Zero marked genitives are one of the traits that are
mentioned as characteristic of Yorkshire speech. As the s-genitive has
been generalized in Standard English, this means that the similarities
across the North Sea are today greater in the standard languages than
in the dialects in the areas where the original contact took place. As for
the development of the -s marker from an affix to a clitic (or “phrasal
affix”) in English and Scandinavian, it is hard to say if they have a common origin, since unequivocal attestations of “group genitives”, that is,
s-genitives marked on the last word of the NP rather than on the head
noun, are relatively late. Allen (2008: 153) provides an example from
1387 of an -s attached to a postmodifier: þe kyng of Fraunces men.
Perridon (2013: 142) cites as the first comparable example from Danish
pana hans allar rigens aff Danmarks wegna ‘on behalf of him or the
kingdom of Denmark’ from 1410.
Summing up, the story of the s-genitive is a complex one: on both sides
of the North Sea we see a competition between generalized ­s-genitives,
zero-marked genitives and linking pronoun constructions, but and even
if the parallels are striking, it is not possible to construct a coherent
narrative that would show how the developments are linked together.
In standard Danish and Swedish and conservative Bokmål Norwegian
the original Germanic three-gender system – which is still preserved in
most non-standard varieties in Peninsular Scandinavian and even in some
Danish traditional dialects – has been reduced to a distinction between
Linguistic Mysteries Around the North Sea
common and neuter gender. In general, these varieties use the erstwhile
demonstrative den to refer to non-neuter inanimates. According to
Ringgaard, (1989: 163), the two-gender system is found already in Jyske
Lov, meaning that the simplification process must have started no later
than the 13th century and probably earlier. In large parts of Jutland, the
common:neuter distinction has also been abolished or (mainly for pronominal reference) been transformed into a semantically-based count:
mass distinction. Even this more radical state-of-affairs may be reflected
in the earliest texts to some extent (Perridon 2005: 1021).
The three-gender system was more or less intact in Old English but
had already gone in the earliest texts in Early Northern Middle English
and was also on its way out in other varieties of Early Middle English. In
the choice of pronouns, however, there was a certain tendency towards
“natural gender” already in Old English (Curzan 2003).
Gender, especially of the “non-natural” type not motivated directly
by semantics, is often thought of as a quirk that languages will tend to
get rid of as soon as possible. However, gender systems are astonishingly stable overall. (For a more detailed discussion, see Dahl (2004:
196–202), and for a survey of earlier views on gender, see Kilarski
(2007)). Only a small number of Indo-European languages have lost
their gender systems altogether – outside Germanic it has happened in
some Iranian languages and Armenian. According to McWhorter (2002:
230), the loss of NP-internal gender in English makes it unique “among
all the languages of Europe” except for “a few nonstandard dialects of
particular languages”. He mentions two non-Germanic examples, both
said to be due to language contact, and two Germanic ones, “Western
Danish” and “Ostrobothnian Swedish” – which he labels as “cases of
internal loss”. The references to the varieties in question are not quite
exact – “Western Danish” is the area in Jutland referred to above, and
“Ostrobothnian Swedish” is not all of the Swedish speaking area of
Ostrobothnia but only the northernmost part of the Swedish speaking
area of Finland. Given the proximity to Finnish, a genderless language,
an account in terms of language contact would seem natural here too,
which leaves English and Jutlandic as the only cases to be explained –
which makes one wonder if they are really independent of each other.
“Direction of case levelling”
As an argument for their position that English is a North Germanic
language, Emonds and Faarlund (ms.) mention that both English and
From Clerks to Corpora
Danish/Norwegian (but not Swedish) share the tendency to use the
object forms of pronouns in all positions other than uncoordinated
subjects of overt finite verbs, as in It’s me or Danish Det er mig. Prima
facie it looks as if English could have been influenced by French on this
point, since in French the pronouns used in those positions are derived
from Latin accusative forms. However, French makes a distinction that
English lacks, namely between the clitic pronouns used with uncoordinated direct objects, e.g. me ‘me’ or te ‘you’, and the “full’ pronouns used
e.g. in C’est moi ‘It’s me’. Basing themselves on this difference, Emonds
and Faarlund (ms.) reject the possibility of French influence.
Timing is critical here. Emonds and Faarlund’s formulation “some kind
of extension of either subject or object forms took place in all Germanic
languages in which case distinctions are restricted to pronouns” suggests
that they assume a connection between the loss of case distinctions on
nouns and the extension of subject or object forms. But this seems to
imply that either (i) the nominative-accusative distinction had already
disappeared in the last common ancestor of English and Scandinavian
or (ii) the similarity is due to later contact between Scandinavia and
England. In fact, Emonds and Faarlund say that they do not exclude this
possibility even if their preferred hypothesis seems to be that “ME pronominal case patterns simply continued those of Old Danish”. But if the
change is due to contact, there is not really any reason why it could not
equally well go from west to east.
The definite article
Most Scandinavian varieties have one postposed and one suffixed definite article whose distribution varies according to slightly complex
patterns (Dahl, 2003). West and South Jutlandic varieties also have two
definite articles, with basically the same distribution as the standard
Danish ones, but the one that corresponds to the suffixed article, that is,
the one that is used when there is no modifier before the head noun – is
an invariable marker æ which always precedes the noun, as in æ by ‘the
town’. This has been claimed to be due to influence from Low German.
Perridon (2005: 1019) argues that the isomorphism between the Danish
and the Jutlandic systems speaks against this. He also argues against
the possibility that the Jutlandic dialects have replaced an earlier suffixed article by a prefixed one, and hypothesizes that Jutlandic æ and
the definite suffix arose at the same time, in the 11th or 12th century, in
spite of the fact that æ is only attested from the 16th century onwards.
Linguistic Mysteries Around the North Sea
He attributes the difference in position to the timing of the change from
postposition to preposition attribute-noun order, which was earlier in
South and West Jutland. Yet, if Perridon is right about the time of the
change, it coincides with the period when influence from English could
be expected. Furthermore, the Jutlandic and the English articles share a
feature not found elsewhere in Germanic languages at this time – they
are invariable, not only in gender and case but also in number. The
invariable article þe is another of the innovations in Middle English
that were established early in the North and then spread to the south.
White (2002) and Tristram (2004) invoke influence from Brittonic
Celtic, where there was also an invariable definite article which, as they
argue, was older than the English one.
Verb morphology
In the present indicative, Old English distinguished all persons in the
singular and none in the plural. By contrast, Old Norse distinguished
all persons in both the singular and the plural except for the second
and third singular. Both Northern Middle English and Old Danish had
virtually the same simplified system in which there were no person distinctions except that the first person singular was optionally different
from the second and third persons (although there seems to have been
a difference in the extent to which this distinction was made).
It thus seems that the systems have converged in that distinctions
that were not made in both Old English and Old Norse were abandoned. However, this convergence was at least initially restricted to
NME and Old Danish; other varieties of Middle English and Medieval
Table 1. Verb endings in Old English, Northern Middle English, Old Danish,
and Old Norse.
Old Danish
From Clerks to Corpora
Scandinavian kept the old systems to a much larger extent. The system
of Modern Icelandic is identical to that of Old Norse and in Swedish,
three persons were distinguished in the plural at least in the written
language for many centuries, and Elfdalian still keeps this system.
Possessive reflexives
Older forms of Germanic made a distinction in the third person between
(i) the reflexive pronoun sı n which agreed with its head noun and
(ii) the non-reflexive, non-agreeing genitive pronouns. In the ­modern
languages, this distinction is retained only in North Germanic. In
West Germanic, two different developments have taken place, both
leading to the disappearance of the distinction. In Continental West
Germanic, the process seems to have been one of fusion between the
two types of pronouns, resulting in a generalization of the agreeing
type. In English, on the other hand, the non-reflexive forms were generalized and the reflexive pronoun disappeared without trace already in
the course of the Old English period.
As an exception to the general retention of the reflexive possessive
in North Germanic, Danish uses the possessive reflexive sin only in the
singular, and in West and South Jutlandic the distinction between reflexive and non-reflexive pronouns has generally been lost; sin is retained
as a general non-human possessive, for human referents the ordinary,
non-agreeing possessive pronouns are used. Perridon (1999: 185) rejects
the possibility of Low German influence, as Low German does not have
non-agreeing possessives. I am not sure how strong this objection is, but
it can be noted that it could not be used against the alternative hypothesis that we are dealing with influence from English, where, as mentioned above, the non-agreeing pronouns were generalized.
Other phenomena
There are a few other points where there is a split between Danish and
some or all Peninsular Scandinavian varieties, and the Danish pattern is
also found in English, but where I have not found sufficient information
about the historical development. These will just be briefly mentioned
here with Swedish as representing the other side of the split:
• In presentational sentences (‘There is beer in the fridge’), Swedish
uses a neuter pronoun (det) as an expletive but English and Danish
use an original adverb (there and der respectively);
Linguistic Mysteries Around the North Sea
• In Swedish, pronominal objects normally follow a verb particle, in
English and Danish they normally precede it;
• In Danish possessive pronouns consistently precede the head noun;
in Swedish they optionally follow kin terms, and in many Swedish
and Norwegian varieties postposed possessives are the norm.
Thus, we have seen that the grammatical changes that have shaped
the modern standard Scandinavian languages, in particular standard
Danish and Swedish, largely originated in western Denmark, and that
many of them have been more strongly implemented in Denmark,
­especially the western parts. We have also seen that a significant part of
the changes had already made their way into Old Danish as it was written in western Denmark in the mid or late 13th century. Furthermore,
many of them closely parallel what happened in the transition from
Old English to Middle English, while resemblances to Continental
West Germanic are considerably weaker, and parallel changes in those
­languages are often later.
These observations fairly strongly suggest that Old Danish, and later
on and more indirectly, other Scandinavian varieties were influenced by
English or Anglicized Norse, or both, during the transition period from
Old to Middle English. There are some major stumbling blocks here,
One is that the influence appears to be essentially restricted to grammar and phonology. There are a number of words in the Scandinavian
languages that are usually regarded as loans from Old or Middle
English (although usually originally from Latin or Greek), but they tend
to be connected with religion (such as Swedish ängel ‘angel’ and kyrka
‘church’) and are assumed to have arrived with Anglo-Saxon missionaries. It should be added that the number may be a bit larger than is usually thought, since it is not always possible to see which West Germanic
language a word comes from, and there may be a tradition to routinely
ascribe a continental origin to such cases. Quak (2005: 569) notes a
number of presumed loans from Old Frisian which could equally well
be from Old English. However, the absence of a larger number of clear
loanwords from Old or Middle English in Scandinavian can be seen
as a serious problem for the hypothesis that Scandinavian was under
influence from the British Isles, especially if we assume that England
was the culturally and economically more advanced part. (Compare the
large number of Low German loanwords in Scandinavian and French
loanwords in English.)
From Clerks to Corpora
The second stumbling block is how to find a credible account of
how the influence took place. The result of the changes that have
been discussed here was a major restructuring of the grammar of the
Scandinavian languages, in particular of the inflectional system. An
explanation in terms of contact-induced change demands that the
contacts were more than superficial. We know that there were intimate contacts between Scandinavian and English in Britain, even if
the intensity and length of the contact is controversial. It is a common
assumption that this led to a mutual convergence of the languages, possibly including morphological simplification due to imperfect learning.
For instance, Braunmüller (2005: 1033) speaks of a “transition from
focused bidialectalism to the use of diffuse, jargon-like interdialectal
variants in the next generation(s)”. Thomason and Kaufman (1988)
think that “there must have been heavy borrowing between the two
languages before the Norse speakers in the end switched to English”
and that “[if] the Norse had survived we would have seen a Norse
equally riddled with English traits”. But maybe the Norse survived
after all, in some sense. If Danelaw Scandinavian was still around when
the Danes took control of England after the millennium shift, it might
have been strong enough to become some kind of prestige dialect in
Cnut’s empire, including Denmark.
Not much seems to have been said about language at Cnut’s court.
However, Frank (1994: 108) says in her paper on the poetry of the
Scandinavian “skalds” there that there was a “casual use in the verse of
Old English words, idioms, and syntax”, as if
the skalds were composing for a Norse-speaking community enisled
in a sea of Anglophones; at times the language seems almost AngloDanish, a dialect as distinct from Old Norse as Québecois is from
metropolitan French.
She bases this statement on Hofmann (1955), one of the very few
works that look seriously at the possible influence of English on
Viking Age Scandinavian. Hofmann analyzed a number of poems by
Norse “skalds” and indeed finds quite a few Anglicisms in them; it
has to be admitted, however, that as examples of code-mixing they
are not so impressive in comparison to what has been documented in
the rich literature on such phenomena (see e.g. Muyskens 2000). Still,
Frank’s picture does not sound too unrealistic. In the enumeration of
Cnut’s earls in Keynes (1994) we find 12 Scandinavian earls and 9
English ones, giving the impression of a relatively even distribution of
Linguistic Mysteries Around the North Sea
both groups in the ruling élite. There would have been a considerable
linguistic diversity in these circles. On the other hand, the period of
Danish control was probably too short for the linguistic situation to
It is clear that being part of Cnut’s empire also had an impact
on Denmark. Two areas where Cnut had clear ambitions are often
­mentioned in the literature: religion and coinage (Lund 1994; Jonsson
1994). There were mints in all major regions of Denmark during Cnut’s
reign, and the Anglo-Saxon influence on the Danish church “was also
more than traceable, it was massive” (Lund 1994: 39). Already Cnut’s
predecessor Sweyn Forkbeard replaced the German bishops with
English and Norwegian ones, and Cnut continued this policy. What this
means from a linguistic point of view is that there will have been significant ­numbers of speakers of various forms of English and/or “AngloDanish” in Denmark – although it may be questioned how much this
would influence the language spoken in Denmark in general.
One important but obscure point is to what extent there was remigration of Scandinavians from England. It seems likely that part of the
Danes in England, particularly those who had arrived recently, would
move back to Denmark when the Danes lost the political control. It
may also be speculated that the harsh treatment of the population in
the north after the Norman invasion may have driven some of them to
seek refuge on the other side of the North Sea.
The ruling élite seems to have been quite mobile during this period.
Among royals, marriages across ethnic and linguistic borders seem to
have been the rule rather than exceptions, although this may of course
have been less pronounced further down in the hierarchy. Cnut himself
may have had a Slavic mother and possibly a Slavic paternal grandmother; his wives Ælfgifu and Emma were Anglo-Saxon and Norman,
respectively. His nephew Sweyn Estridson, who was King of Denmark
between 1047 and 1074, had a Norwegian father and was born in
England; he married two Swedish women and one Norwegian woman.
All this may still seem a bit thin as a basis for the assumption that
Old Danish was shaped in a decisive fashion by contacts across the
North Sea. On the other hand, the alternative hypothesis is that the
similarities we see between Early Middle English and Old Danish are
due to parallel but independent developments – and that it is a pure
coincidence that these parallels are strongest between the Danelaw area
and western Jutland. For the time being, I think the linguistic mysteries
around the North Sea will remain.
From Clerks to Corpora
Allen, C.L. (2008). Genitives in Early English: Typology and Evidence: Oxford/
New York: Oxford University Press.
Baugh, A.C. & Cable, T. (1993). A History of the English Language. London:
Bjerrum, A. (1966). Grammatik over Skånske lov: efter B 74. Copenhagen:
Københavns Univ. Fond til Tilverjebringelse af Læremidler.
Braunmüller, K. (2005). Language contact during the Old Nordic period I: with
the British Isles, Frisia and the Hanseatic League. O. Bandle, L. Elmevik &
G. Widmark (eds) The Nordic Languages: An International Handbook of
the History of the North Germanic Languages (Vol. I). Berlin: Mouton de
Gruyter, 1028–1039.
Burrow, J.A. & Turville-Petre, T. (1991). A Book of Middle English. Oxford:
Curzan, A. (2003). Gender Shifts in the History of English. Cambridge:
Cambridge University Press.
Dahl, Ö. (2003). Competing definite articles in Scandinavian. B. Kortmann
(ed.) Dialectology Meets Typology. Berlin: Mouton de Gruyter, 147–180.
———. (2004). The Growth and Maintenance of Linguistic Complexity.
Amsterdam: Benjamins.
———. (2010a). Grammaticalization in the North: Noun Phrase Morphosyntax
in Scandinavian Vernaculars. Stockholm: Department of Linguistics,
Stockholm University.
———. Review of C. Allen: Genitives in early English: Typology and evidence.
Diachronica 27:3, 489–496.
Delsing, L.-O. (2003). Syntaktisk variation i nordiska nominalfraser. L.-O.
Delsing, Ø.A. Vangsnes & A. Holmberg (eds) Dialektsyntaktiska studier av
den nordiska nominalfrasen. Oslo: Novus.
Emonds, J. & Faarlund, J.T. (ms.). English as North Germanic.
Frank, R. (1994). King Cnut in the verse of his skalds. A.R. Rumble (ed.) The
Reign of Cnut, King of England, Denmark and Norway. London: Leicester
University Press, 106–124.
Hofmann, D. (1955). Nordisch-englische Lehnbeziehungen der Wikingerzeit.
Copenhagen: Munksgaard.
Jonsson, K. (1994). The coinage of Cnut. A.R. Rumble (ed.) The Reign of Cnut:
King of England, Denmark and Norway. London: Leicester University
Press, 193–230.
Linguistic Mysteries Around the North Sea
Keynes, S. (1994). Cnut’s Earls. A.R. Rumble (ed.) The Reign of Cnut: King
of England, Denmark and Norway. London: Leicester University Press,
Kilarski, M. (2007). On grammatical gender as an arbitrary and redundant
category. D.A. Kibbee (ed.) History of Linguistics 2005: Selected papers
from the Tenth International Conference on the History of the Language
Sciences (ICHOLS X), 1–5 September 2005, Urbana-Champaign, Illinois.
Amsterdam: Benjamins, 24–36.
Klemola, J. (1997). Dialect evidence for the loss of genitive inflection in English.
English Language and Linguistics 1:2, 349–353.
Laker, S. (2008). Changing views about Anglo-Saxons and Britons. H. Aertsen
& B. Veldhoen (eds) Six Papers from the 28th Symposium on Medieval
Studies. Leiden: Leiden University Press, 1–38.
Lund, N. (1994). Cnut’s Danish kingdom. A.R. Rumble (ed.) The Reign of
Cnut. King of England, Denmark and Norway, Leicester. London: Leicester
University Press, 27–42.
McWhorter, J.H. (2002). What happened to English? Diachronica 19:2,
Minkova, D. (1991). The History of Final Vowels in English: The Sound of
Muting. Berlin: Walter de Gruyter.
Muysken, P. (2000). Bilingual Speech: A Typology of Code-Mixing. Cambridge:
Cambridge University Press
Perridon, H. (1999). Review of A. Torp. 1998. Nordiske språk i nordisk og
germansk perspektiv. Tijdschrift voor Skandinavistiek 20:2, 182–186.
———. (2005). Dialects and written language in Old Nordic II: Old Danish
and Old Swedish. O. Bandle, L. Elmevik & G. Widmark (eds) The Nordic
Languages: An International Handbook of the History of the North
Germanic Languages (Vol. I). Berlin: Mouton de Gruyter, 1018–1027.
———. (2013). The emergence of the s-genitive in Danish. Language Sciences
36, 134–146.
Quak, A. (2005). Nordic and North Sea Germanic relations. O. Bandle,
L. Elmevik & G. Widmark (eds) The Nordic Languages: An International
Handbook of the History of the North Germanic Languages (Vol. I). Berlin:
Mouton de Gruyter, 568–572.
Ringgaard, K. (1986). Flektionssystemets forenkling og middelnedertysk.
Arkiv för Nordisk Filologi 101, 173–183.
———. (1989). Fleksionssystemets forenkling i dansk. Arkiv för Nordisk
Filologi 104, 160–165.
From Clerks to Corpora
Schrijver, P. (2006). What Britons spoke around 400 A.D. N.J. Higham (ed.)
Britons in Anglo-Saxon England. Woolbridge: Boydell, 165–71.
Thomason, S.G. & Kaufman, T. (1988). Language Contact, Creolization, and
Genetic Linguistics. Berkeley: University of California Press.
Tristram, H.L. (2002). Attrition of inflections in English and Welsh. M.
Filppula, J. Klemola & H. Pitkänen (eds) The Celtic Roots of English.
Joensuu: Joensuu University Press, 111–49.
———. (2004). Diglossia in Anglo-Saxon England, or what was spoken Old
English like? Studia Anglica Posnaniensia 40, 87–110.
Trudgill, P. (2010). Investigations in Sociohistorical Linguistics. Cambridge:
Cambridge University Press.
White, D.L. (2002). Explaining the innovations of Middle English: What,
where, and why. M. Filppula, J. Klemola & H. Pitkänen (eds) The Celtic
Roots of English. Joensuu: Joensuu University Press, 153– 174.
Whitelock, D. (1996). English Historical Documents: Volume 1. London:
3 The Late Middle English Version of Practica
Urinarum in London, Wellcome Library,
MS 537 (ff. 15r-40v)1
Javier Calle-Martín
University of Málaga
1. Introduction to the edition
Mediaeval uroscopic treatises have been traditionally underestimated
in the last decades, to such an extent that they can be safely deemed
to be the Cinderella of the subject categories of Medicine, especially
if compared with Herbals, Remedybooks and Leechbooks (Keiser
1998: 3661; Calle-Martín 2012: 243). Using Voigts and Kurtz’s electronic database of Scientific and Medical Writings in Old and Middle
English (2000), Tavormina has found that the subject field ‘Urine and
Uroscopy’ contains 408 texts, representing “the fourth largest in the
entire database, preceded only by ‘Recipes’, ‘Alchemy’, and ‘Herbs and
Herbal Medicine’” (Tavormina 2005: 40; 61). Unfortunately, however, many of these uroscopic treatises are hitherto unedited and, more
importantly, they are still waiting for a patient scholar to investigate the
textual tradition of the extant witnesses (Tavormina 2009: S33–S41).
According to Tavormina, there is a scholarly gap in the field and further
research is needed in the following three areas: a) the publication of
more uroscopic treatises in the vernacular; b) the complete taxonomy
of the texts in Middle English; and c) the relations of these texts to their
Latin/English sources (2005: 41).
1The present research has been funded by the Spanish Ministry of Science and
Innovation (grant number FFI2011–26492) and by the Autonomous Government
of Andalusia (grant number P11-HUM7597). These grants are hereby gratefully
acknowledged. This Festschrift article is dedicated to Prof. Nils-Lennart Johannesson,
a keen attender of Selim conferences in Spain (The Spanish Society for Medieval
English Language and Literature), on the occasion of his retirement.
How to cite this book chapter:
Calle-Martín, J. 2015. The Late Middle English Version of Practica Urinarum in
London, Wellcome Library, MS 537 (ff. 15r-40v). In: Shaw, P., Erman, B., Melchers, G.
and Sundkvist, P. (eds) From Clerks to Corpora: essays on the English language yesterday and today. Pp. 35–52. Stockholm: Stockholm University Press. DOI: http://dx.doi.
org/10.16993/bab.c License: CC-BY.
From Clerks to Corpora
Our analysis then falls within Tavormina’s first area and contains
the semi-diplomatic edition of the Middle English version of Practica
Urinarum, a small uroscopic treatise instructing the mediaeval ­practitioner
on the examination of urine in the light of the following three elements:
substance, colour and sediment. The text also reports what must be considered for the treatment of the patient, i.e. his/her age, duration of illness,
the time of greatest suffering together with “further questions regarding
the time of the specimen and the diet of the patient” (Keiser 1998: 3662).
As far as we have been able to investigate, the text has been preserved
(completely or partially) in seven different witnesses (Keiser 1998: 3852):
• London, Wellcome Library, MS Wellcome 537, ff. 15r-40v (c. 1460).
• Cambridge University Library, Gonville and Caius College, MS
336/825, ff. 132v-136r (1480–1500).
• Cambridge University Library, Trinity College, MS O.1.77,
ff. 21r-29v (c. 1460).
• British Library, MS Sloane 2320, ff. 4r-9v (c. 1458).
• British Library, MS Sloane 3566, ff. 24r-33v (1450–1475).
• Boston Medical Library, MS Countway 19, ff. 14r-18v (c. 1460).
• Yale University Library, Takamiya 33, ff. 24v-37r (1480–1500).
MS Wellcome 537 has been used as the source text for the present
edition, which is hitherto unedited. With the title Miscellanea Medica
VII, MS Wellcome 537 is a relatively small volume written in octavo
size, measuring approximately 14.5 x 11 cm. The volume houses a
collection of eight anonymous treatises, both in English and Latin, on
medical astrology, weather prediction, uroscopy together with a collection of medical receipts. The following collation relies on Moorat’s
accurate description of the volume (1962: 394–395):
• [Anon.] Canon pro medicinis dandies (in Latin), ff. 6r-12r.
• [Anon.] Menses per circulum anni (in Latin), ff. 12v-14v.
• [Anon.] Seeing of Urines (in English), ff. 15r-40v.
• [Anon.] Medical receipts (in English), ff. 40v-46v.
• [Anon.] Practica medicinae (in English), ff. 48r-310v.
• [Anon.]
Table of Golden Numbers and Dominican Letters,
ff. 315v-318v.
• [Anon.] De cursu lunae (in Latin), ff. 318v-325r.
• [Anon.] On the Changes of the Moon (in English), ff. 326r-333r.
Even though the volume had been previously dated by Francis Douce
(1757–1834) to the first half of the 15th century, Moorat reports that
The Late Middle English Version of Practica Urinarum in London
there is an autograph letter to Dr Joseph Frank Payne (1840–1910)
from Thomas Wright (1810–1877), proposing to date the composition
of the volume to the year 1462 because “the whole of it is written in
the hand which is common enough throughout the reigns of Henry VI
and Edward IV” (Moorat 1962: 395). The palaeographic analysis of the
scribe’s handwriting confirms it to be a hybrid script composed in the
second half of the 15th century in the light of the overwhelming preference for the Secretary script, with just sporadic tinges of the Anglicana
(Petti 1977: 15). Fig. 1 below reproduces the inventory of letterforms
used in MS Wellcome 537 where the letters have been numbered for
reference purposes. As shown, the Secretary script is witnessed in the use
of distinctive letterforms, which sharply differ from the conventional cursive hand of the Anglicana. Among others, the following stand out: the
single-lobed <a> with a pointed head (1); two-lobed <b> (2); the letter
<d> with a looped stem (4); the letter <k> with its characteristic rightarced headstroke (10); the letter <l> with a lobed arm (11) (Clemens and
Graham 2007: 167–168); the right-shouldered <r>, footed and sitting
on the script line (17); the heavy ascender of the letter <v> (24); and the
letter <w> already resembling a double v (25).
Already in the second half of the 15th century, “the two competing Gothic cursiva scripts often seem so intermingled as to have produced a new cursive book script, typically Secretary in ductus, but using
Anglicana round e and sigma-shaped s as cursive features and liable to
adopt other Anglicana letterforms” (Roberts 2005: 212–213; Derolez
2003: 162; also Denholm-Young 1954: 28–29; Hector 1966: 58). In
this fashion, the Anglicana script is limited to the use of the f­ollowing
four distinctive letters in MS Wellcome 537: the round <e> (5);
Figure 1. Inventory
of letterforms in MS Wellcome 537.
From Clerks to Corpora
the two-compartment figure 8 form of <g> (7); the sigma-like <s> (20);
and the letter <w> consisting of a double <l> together with a 3 (26).
The principles of a semi-diplomatic edition have been adopted in the
present edition so as to render an accurate reproduction of the source
text, complying with the following guidelines, partially adapted from
Clemens and Graham (2007: 75–81; also Calle-Martín and MirandaGarcía 2012: 67–68): a) the spelling, capitalization and punctuation
of the original have been retained; b) abbreviations have been expanded with the supplied letter(s) italicized;2 c) insertions, both of single
letters and words, have been inserted in their proper sequence in the
body of the text;3 d) word separation has been normalized, regardless
of whether words are run together without separation or whether a
gap is left within a word; and e) lineation and paragraphing have been
disregarded for reasons of space.
2. The Middle English text4
{f. 15r} . ihc .
¶ here begynneþ þe practise of the sighte of vrynes . HIt is to vndurstonde þat who so wille loke an vryn him behoueþ to considere þre
þinges / Firste he muste aske þe age of þe pacient or seke body . and
how longe his sekenes haþ holde him . and what tyme hit toke him
furste . and whiche tyme of þe day or of þe nyȝt hit greueþ him moste
. Also if þe vryn were made and brouȝte in a clene vessel . and þat
hit were þe furste water after mydnyȝt Also þe dyete of þe seke body
{f. 15v} wheþer he haþ vsed highe metis and drinkes or ellis lowe . Also
hit behoueþ to consider þe substaunce of þe vryn wheþer hit be þicke
or þynne . and þe qualite wheþer it be highe or lowe of coloure . and þe
quantite wheþer it be litil or moche . and þe contentis . and þe cloudes
þat ben þeryn . And in whiche regioun þey ben yn and if þey be oueral
in liche colourid . or if hit be more remise in oon regioun þan in anoþer
. or if hit be of twoo colouris or if hit be oueral in liche colourrid .
boþe intencioun and remissioun . {f. 16r} Than hit behoueþ principally
to wyten by consideracioun of þe vryne and by askyng also what þe
maledy is . wheþer hit be a feuer or anoþer maledy . and wheþer hit be
2Superior letters are lowered to the line, as in wt > with or þe > þe. Superfluous
­brevigraphs and otiose flourishes, in turn, have been disregarded.
3 Insertions above the line and in the margin have been noted with the use of slashes
(/ \) and double slashes (// \\), respectively.
4 (c) Wellcome Library (London) as the owner of the manuscript.
The Late Middle English Version of Practica Urinarum in London
in þe begynnyng . or ellis in state . or in þe ending . Also when þu hast
vndirstonde . al þis . and þe vryne of þe pacient . is brouȝte to þe þu
muste considere þre þingis in euery vryn . þat is to sey . þe substaunce .
þe coloure and þe sedymen . Oon is cause of colour . anoþer is cause of
substaunce . þe þirde is cause of sedymen . Now in euery man is body
is foure qualitees . hete and colde . moyste {f. 16v} and drye hete and
colde þey ben causers of colours . Drynes and moystenes þey ben
cause of substaunce . /hete\ is cause of rede coloure . drynes is cause
of þyn substaunce . moystenes is cause of þycke substaunce . As þus .
if þe vryn of þe pacient be rede and þicke it signifieþ þat blode haþ
more dominacioun þan an oþer . Why For blode is hote and moyste .
If it be rede and þynne hit sheweþ þat colere haþ dominacioun . For
why . colere is hote and drye . If þe vryn appere white and þicke hit
betokeneþ fleume . For fleume is colde and {f. 17r} moyste . If þe vryn
shewe white and þynne it sygnifieþ malencoly . For malencoly is colde
and drye . When þu hast considerid wel as þis þen beholde þe diuersite
of colours of þe vryns and þe cercles of hem . whiche now folowen .
Karapos . is an vryn in coloure like to kamellis flesshe . whiche is a
colour white douny . and in substaunce hit is þyn . The whiche sweweþ5
and signifieþ þe droppesy . or a wynde vndur þe syde . þe stone . þe hede
ache . a posteme in þe longis . or a {f. 17v} or elles a fleume . ¶ Plumbea is
an vryn in colour lyke to lede . whiche signifieþ þe falling yuel . or membris broken . or a palsy . or a feuer tertian . or difficulte of water making
. or womman is floures ¶ Pallida is an vryn /in colour\ like to flesshe
halfe soden þe iuse not putte oute . which signifieþ a feble hede . or a
feuer . or a colde stomake . or a brennyng vndur þe lyuer . or wastyng
of longes . ¶ Plumbea is an vryn in coloure {f. 18r} like to lede . as
hit writen tofore of colour pallida ¶ Viridis is an vryn þat is grene in
colour like to þe caule leaf . whiche betokeneþ wasting of kynde/li\ hete .
þe morfu . or þe hede ache . or þe iaundyse . or chasing of þe luyer and
wasting of kyndeli hete . ¶ Inopos is an vryn in colour like to a man is
lyuer . or to olde rede wyne yturned into blackenes . whiche betokeneþ
þat þe galle brenneþ þe lyuer . and þerof commeþ þe iaundise . or hit
signifieþ a veyne roten /or broken\ of the /r\eynes . or to myche delynge
wiþ a womman . {f. 18v} ¶ Subpalida is an vryn in coloure /like\ to iuse
of flesshe halfe soden . and þe iuse yputte oute . The whiche betokeneþ
a posteme in þe riȝt syde . or a tysike . or hardenes to make water ¶
5Probably an erroneous form for <sheweþ>. The e-MED notes a similar form (ie.
swewe) which is identified as an error.
From Clerks to Corpora
Rubicunda is an vryn in colour like to a flaume of fyre sentte oute .
The whiche betokeneþ a feuer acute and sharpe . and vndursette at þe
herte . defaute of wynde and a drie coughe . ¶ Subcitrina is an vryn in
colour . like to þe iuse of an appel orynge . whiche betokeneþ a feuer
agu . and þe same as doþe cytryne {f. 19r} coloure but not so myche .
ne so perfite . ¶ Subrufa is an vryn in colour like to golde and siluer
medelid togedir . whiche betokeneþ a feuer tertian . or a posteme vndur
þe ribbes . or þe goute . or a tisike . /or\ shortenes of wynde . and simple
bitternes in taste and grete dryness ¶ Subricunda is an vryn in colour
like to a flaume of fyre not sente oute . This betokeneþ as doþe subrufa
. but þat þe yuel is more greuous in þis þan hit is in þat other . {f. 19v}
¶ Rubea Rede vryn is in coloure like to safron in dorte . whiche betokeneþ a feuer þoroughe chafyng of þe lyuer . or a feuer quarteyn . or
ellis a posteme of þe longis . begynnyng of a dropesy . or of a morfu . ¶
Citrina is an vryn like to þe coloure of an appelle of an orynge . whiche
betokeneþ wasting of þe splene and of þe lyuer . commyng of þe frenesy
. and hete aslaked in a feuer tercian . ¶ Subrubea is an vryn in colour
like to safron belynge . whi{f. 20r}che betokeneþ a feuer of corrupte
blode /in þe veynes\ . or a veyne broken in þe backe . or ellis þe floures
of wommen . ¶ Rufa is an vryn in colour like to pured golde . The
whiche betokeneþ þe emoyraudis . or þe dropesy . or passing of floures
. ¶ kaynos is an vryn in colour like to blacke wyne or to blacke water
. And hit betokeneþ a crampe in þe veyne of þe herte . or a posteme in
þe stomake . or ellis þe goute .
¶ Lactea is an vryn in coloure {f. 20v} like to whey of coughe mylke
and hit betokeneþ deþe if þe paciente slepe not . or ellis a frenesy . or
a longe /during\ feuer . ¶ Subuiridis is an vryn in colour clere as þe
welle water . And hit betokeneþ þe same as doþe þe rede coloure. but
not so perfitly . ¶ Alba is an vryn in colour as clere welle water . and
hit betokeneþ wasting of þe splene . or þe dropesy . or þe stone and
greuaunce in pissyng . or pissing ageyne wille . or þe emoyraudis . or
stopping of floures . cha{f. 21r}fyng of þe lyuer . wodenes . and in axces
and feuer hit is deþe . ¶ Nigra is an vryn in colour like to blacke ynke
or to a brent horne . And hit betokeneþ a feuer quarteyn . or a blacke
iaundise . peryl of deþe in shorte tyme . ¶ Glauca is an vryn like in
colour to a shenyng horne . And hit betokeneþ a posteme in þe riȝt side
. or a feuer etike . or /a\ bladder on the longis .
¶ Her endeþ þe significacioun of /þe coloures\ of vryns Now we wille
{f. 21v} declare and telle of þe Cercles of vryns and begynne at þe hede
of man A Cercle þat is grete shevweþ þe feblenes of þe hede and moche
The Late Middle English Version of Practica Urinarum in London
akynge þeryn . ¶ A . cercle þat is white and þycke . signifieþ þat fleume
haþ duracioun in þe hynder parte of þe hede . for þat is his resting place
. ¶ A . cercle white and clere . signifieþ þat malencoly haþ duracioun
on þe lifte syde of þe hede . for þat is his sete . ¶ A . cercle þat is rede
and þicke . signifieþ þat blode haþ duracioun in þe forehede . for þat is
his sete . ¶ A . cercle rede and clere signifieþ þat colere haþ duracioun
. {f. 22r} in þe riȝte side of þe hede . for þat is his restyng place . ¶ If a
cercle shewe or appere grene hit betokeneþ grete ache in þe hede . and
vnder þe riȝte side . ¶ If a cercle appere white hit signifieþ grete ache
in þe hynder party of þe hede . ¶ If a cercle appere rede hit betokeneþ
ache in þe forehede . And if hit appere blacke hit signifieþ þe falling
yuelle . ¶ If þer be in a cercle smale bollis as þey were of reyne water
hit betokeneþ a wynde in þe {f. 22v} hede . risyng oute of þe stomake .
¶ If þer be in a cercle mottis as ben in þe sunne hit betokeneþ a reume
of þe hede . or þe pose . or deffenes . or heuynes . ¶ If þer appere in
a cercle a tree hit betokeneþ a posteme in þe bladder . or ellis in þe
lyuer . ¶ If þer appere in a cercle fattenes hit signifieþ a feuer . or ellis
streittenes at /þe\ breste . ¶ If blode appere in a cercle hit betokeneþ
breking of a veyne in þe backe . If þe /vryn\ appere blac{f. 23r}ke hit
betokeneþ deþe . which is causid of þe blacke iaundise . or ellis of þe
feuere quarteyn . ¶ Water ful of smale þredis betokeneþ dry complexioun . or ellis a posteme in þe riȝte side . ¶ White grauel appering in a
cercle signifieþ þe stone in þe bladder . Rede grauel betokeneþ ache in
þe reynes . and also a spice of þe stone . Blacke grauel betokeneþ ache
in þe rigge bitwene þe kidneys . Shales as hit were of fysshes betokeneþ
stopping of floures . and a scabbe in þe matrice .
¶ here endiþ þe cercles {f. 23v} of vryns with her significacions and
now foloweþ þe contentes of hem . and þe passing excesse of hem .
and þe oþer declaryng of which ben gode and holsum and commendable . Eche vryn is clensing of blode . as hit is propurly proued of two
þinges significatife . or forsoþe hit signifieþ of þe lyuer and of þe veynes
. or of þe bladder and of þe reynes . Of oþer þinges hit is yppirliche
significatyfe . ¶ But in vryn lokeþ and beholdeþ diuerse þinges . þat is
to wite . substaunce . coloures . regions . and þe contentes . Anoþer is
because {f. 24r} of substaunce . Anoþer is because of coloure . Anoþer is
becau6 of residens . Now in euery man is body is foure qualitees . hete
and colde . moystenes and drynes . hete and colde þey ben cause of
6 An erroneous form instead of <because>. The scribe does not provide the full form
of this conjunction probably because the word is broken at the end of the line.
From Clerks to Corpora
coloures . Drynes and moystenes þey ben cause of substaunce . As þus
hete is cause of rede coloure . Colde is cause of white coloure . Drines
is cause of þyn substaunce . moystenes is cause of þicke substaunce
. ¶ An vryn is deuided in foure parties . þe one party is þe cer{f. 24v}
cle . þe secunde is þe body of þe eyre . þe þridde is perforacioun . þe
fourþe is þe grounde . ¶ By þe cercle is seen þe sekenes of þe hede
and of þe brayne . By þe body alle yuelis and sekenes of þe spirituel
membris and of þe stomake . By perforacioun sekenes of þe lyuer and
of þe splene . By þe grounde we shewen accidentes of þe reynes and
of þe matrice . and of þe priue membris / ¶ An vryn haþ þre regions .
þat is . þe neþereste bygynneþ fro þe grounde of þe vrynal and dureþ
{f. 25r} by þe space of two fyngers / þe myddel regioun begynneþ þer
as þe neþer endeþ and dureþ to þe cercle . þe whiche cercle is þe ouermeste regioun . And when þer is in þe ouermeste regioun as fome hit
signifieþ ventosite bulnyng vppe in þe veynes of þe vryn . or ellis inflacioun . or oþer vices or sekenes of þe lungis . ¶ If /so be þat\ þe cercle be
þicke hit signifieþ to moche replecioun in þe hede and doloure or ache
. A . white cloude appering in þe myddel regioun in hole men is an yuel
signe . but in men þat haue þe feue{f. 25v}res . hit signifieþ digestioun of
þe mater and of þe yuel . In þe neþermeste regioun if þer appere grauel
or litil stones . hit sheweþ /þat\ þe paciente is ful of litil stones . and is
greued with þe vice of þe stone . Oþerwhile þe residence is blacke if þat
falle in a feuer quarteyn . or for kyndely hete yquenchid hit is a signe
of deþe . If it be for expulsioun of venomous mater whiche is putte oute
by þe vryn hit is a token of helþe / ¶ If an vryn be white at morwe and
brovne after mete hit {f. 26r} signifieþ gode helþe and welle . ¶ If an
vryn be fatte and trobely hit betokeneþ water in þe bowwelis / ¶ If an
vryn be riȝt rede and clere . hit signifieþ a brennyng feuer / ¶ If an vryn
be oueral rede and þicke with a blacke cercle hit signifieþ grete sekenes
and feruent . and but if he swete hit is deþe / ¶ If an vryn be white
and þynne and a litil dymmy hit signifieþ a coldenes in þe body . ¶ If
an vryn be dymme and sumdel rede . with a blacke cercle /it\ signi{f.
26v}fieþ a badde stomake . do medisyn . ¶ If an vryn be clere and haþ
a blacke cercle hit is deþe . ¶ If an vryn be þynne . and grene abouen
hit signifieþ a colde complexioun . and if in þe casting be blode abouen
it is a tisike . ¶ If an vryn be dymme and blacke abouen . or vndir þe
cercle in þe casting hit betokeneþ luste of kynde . and he wolde haue
a woman to pley / ¶ If an vryn be þicke and trobely like horse pisse .
hit signifieþ grete hede ache . do to him medisyne . {f. 27r} ¶ If an vryn
be fatte in þe bottym . and white in þe myddes . and rede abouen . hit
The Late Middle English Version of Practica Urinarum in London
is a feuer quarteyn . ¶ If an vryn appere flessheli and haue drestis in
þe bottum as shauyng hit is a dropesy . Also if hit be as whey aboue .
and clere in þe myddis . and shadweþ beneþe hit is þe colde dropesy
. ¶ If an vryn haue drestes in þe bottum like gobettis of coles and ben
departid asounder and not to grete hit signifieþ a worme in þe body .
do medisine / ¶ If an vryn haue blacke contentes and smale as mottes
stonding in {f. 27v} þe myddes of þe water . hit signifieþ a posteme in
þe side . ¶ If an vryn haue in þe bottum like gobettis of flesshe of smale
shawyng þat is a token of stopping of þe reynes and also of þe pipis
of þe lyuer . do to him medisine ¶ If an vryn /be\ blacke and grene
with longe contentes like to þe sparme of man . hit is þe palsy . and
hit be frotthi and like lede in þe myddes it is þe same / ¶ If an vryn
haue greynes vndir þe sides of þe cercle hit signifieþ a stomake ful of
{f. 28r} humours . and brennyng at þe breste . do to him medisyne . ¶ If
an vryn haue askes aboue hit /is\ harme in alle þe spiritual parties of
al þe body . do medisine / ¶ If an vryn haue contentes white and blacke
in þe bottum hit is costifnes / ¶ If an vryn abide longe lefting þinne
hit signifieþ swelling in þe body . and moche reume and humours þeryn
. ¶ If an vryn be þycke and litil in quantite hit is token of þe stone /
¶ If an vryn be froþthi and clere and litil in quantite and also rede
{f. 28v} hit signifieþ peyne in þe riȝt side ./ ¶ If an vryn haue grete froþe
after þe casting . hit signifieþ wynde in þe bowel/l\is and peyne vndir þe
sides and bad stomake / ¶ If an vryn be rede as brenning golde hit is þe
dropesi of deþe / ¶ If an vryn be as brennyng cole hit is a posteme on
þe lyuer . do to him medisyne .
URyn as flaume of fyre sente oute / Vryn as safron belynge / Vryn as
safron dorte / Vryn as flavme of fyre not sente oute / These foure signifien excesse of digestion / {f. 29r} ¶ Vryn of þe coloure of a man is lyuer
/ Vryn of þe coloure of blacke water / Vryn þat is grene as a caule lefe
/ These thre . signifien moche adustioun and brennynge . ¶ Vryn þat is
pale of colour as is lede / Vryn þat is blacke in colour of adustion . as is
ynke / Vryn þat is blacke in colour as blacke water . or as a blacke horne
/ These þre betoken mortificacioun /and\ sleinge / ¶ Vryn þat is clere as
welle water / Vryn þat is shenyng as a horne / {f. 29v} ¶ Vryn in colour
as whey of mylke / Vryn þat is white douny as is þe flesshe of a camel
/ These foure signifien indigestioun / ¶ Vryn like to iuse of flesshe halfe
soden . þe iuse yput oute / Vryn like to flesshe soden þe iuse not yputte
oute / These ijº signifien þe begynnyng of digestioun / ¶ Vryn þat is as
þe colour of þe iuse of an orynge appel / Vryn þat is like to þe appel of
orynge /rinde\ / These ijº signifien myddel digestioun / {f. 30r} ¶ Vryn
From Clerks to Corpora
þat is as golde and siluer meddelid togeder / Vryn þat is as golde purid
. þese ijº signifien parfite digestioun / And þis vryn rufa betokeneþ
helþe and gode disposicioun of man is body / Vryn subrufa signifieþ
gode helþe but not so perfitely in ony maner as doþe rufa . ¶ Vryn þat
is of þe colour of an orynge appel . Whan hit is white and a myddel
substaunce . and when þe cercle is of þe same coloure hit is praisable /
Vryn þat is of colour as is þe iuse of an appel orynge is not so {f. 30v}
perfite as is colour citrine for citrine is best colour / ¶ Vryn þat is of coloure as is þe rede rose betokeneþ a feuer þat is callid effimera . and if a
man pisshe contynuelly hit signifieþ a feuer contynuel . ¶ Vryn þat is of
colour as blode in a glasse betokeneþ a feuer of to moche blode . and
þan anoon a man muste be letid blode . but if þe mone be in þe signe of
geminorum . ¶ Vryn þat is grene as is þe caule lefe when hit is pissid
and afterwarde appereþ rede {f. 31r} hit betokeneþ adustioun and sore
barnyng . and hit is mortal and deþe / ¶ Vryn þat is rede in al maner .
remeued fro cleernes betokeneþ declynyng of þe yuel . and amendyng
of þe sekenes . ¶ Vryn þat is rede meddelid sumwhat with blackenes
betokeneþ þe yuel of þe lyuer . and of þe herte . ¶ Vryn þat is pale as
flesshe halfe soden betokeneþ defeccioun or feblenes of þe stomake .
and letting of þe secounde digestioun . ¶ Vryn þat is white as is welle
{f. 31v} water in hole men hit betokeneþ rawenes of humours . and in
sharpe feuers hit is deþe / ¶ Vryn in colour as whey of mylke with þicke
substaunce . if hit falle in a womman hit is not so perilouse as hit is in
a man / for þe inordenaunce of þe matrice . but neuer þe late in sharpe
feueres hit is deþe / ¶ Vryn in coloure aboue as whey of mylke . and
byneþe shadueþ abouten þe myddel regioun clere . hit betokeneþ þe
dropesy / Also if þe vryn of him {f. 32r} þat haþ þe dropesy be rufa . or
subrufa . hit betokeneþ deþe / ¶ Vryn þat is white douny . as is camelle
flesshe betokeneþ multitude of corrupte humours . which befalleþ in
fleumatikes . ydropekes . podagres . and so of oþer . ¶ Vryn þat is blacke
may be of kyndely hete yquenchid . and þan hit is mortale and deþe /
Or hit may be for expulsioun of venomous mater . whiche is putte oute
by þe vryn . and þan hit betokeneþ helþe . In a feuer quarteyn hit is
euermore deþe / {f. 32v} ¶ Vryn in colour as /a\ shenyng horne betokeneþ
yuel disposicioun of þe splene . and disposicion of þe feuer quarteyn / ¶
Vryn in colour as safron in dorte with þicke substaunce . and stinkyng
and fomy . betokeneþ þe iaundise . ¶ Vryn in colour as purid golde .
and vryn as golde and siluer meddelid togedir . hauyng byneþe rounde
resolucioun and white aboue and sumwhat fatty . betokeneþ þe feuer
etike . ¶ Vryn in þe grounde of þe vessel vnto þe myddil clere {f. 33r}
The Late Middle English Version of Practica Urinarum in London
after þicke and þynne betokeneþ harme in þe breest ¶ Vryn þat is fomy
and clere and in colour as safron belinge betokeneþ more harme in
þe riȝt syde þen in þe lifte side / ¶ Vryn if hit be white and fomeþ  hit
betokeneþ more harme in þe lifte side þen in þe riȝt side . for þe lifte
side is more colde þen þe riȝte side . ¶ Vryn þat is þinne and pale and
also cleer betokeneþ acetosum fleuma / If þe cercle of þe vryn be not
resting but tremblyng hit betokeneþ rennyng of fleume {f. 33v} and of
oþer humoures fro þe hede . by þe necke to neþer parties / ¶ Vryn þicke
and mylky and litil in quantite and fatte byneþe with shalis betokeneþ
þe stone / and if þe vryn be withoute shalis and be þicke and mylky
and litil in quantite hit signifieþ þe fluxe in þe wombe / Vryn þicke
and mylky and haþ meny dropes in þe ouer parties signifieþ þe goute
in membris of þe body . ¶ Vryn þat is pale bineþe in men betokeneþ
harme in þe reynes Vrin in whiche litil gobettis apperen . if hit be litil
and trobelid {f. 34r} hit signifieþ brennyng of a veyne aboute þe reynes
in þe bladder / ¶ Vryn in whiche quitoure appereþ in þe grounde of
þe vessel betokeneþ roting of þe bladder . or ellis a posteme / And if
þer appere quitoure þoroughe alle þe vryn hit betokeneþ roting of al
þe body / Vryn in whiche frustis apperen litil and brode hit signifieþ
excoriacioun of þe bladder / Vryn motty signifieþ by longe tyme þe
stone in þe reynes / ¶ Vryn þat is white wiþouten feuer boþe in man and
womman oþer{f. 34v}while hit signifieþ harme in þe reynes . and oþerwhile in wommen hit signifieþ impregnacioun . and cesyng of childe
berynge / ¶ Vryn of hem þat ben with childe / If þey oon moneþe or
two or iije florisshe so þeyr vryn shal be ful cleer . and hit shal haue a
white ypostasy in þe grounde / And if þey haue foure moneþes þe vryn
shal be clere . and haue a white ypostasi and a þicke in þe grounde . Hit
was wonte an ymage to appere in þe glasen vessel / If þilke vrin be of
a womman hit signifieþ þe concepcioun to be made . and {f. 35r} if þe
ymage of him þat shewiþ appereþ in þe vrin of a seke man hit signifieþ
feuers interpolatas . or ellis sekenes of þe lyuer . or drawyng of lengþe
of þe sekenes / ¶ Vryn fomy signifieþ ventosite of þe stomake in wommen . or hete from þe nauyl to þe þrote and grete dryness and þurste /
Vryn of a meyde shal be as þe iuse of an orynge appel . where vrin liuida
and ful clere declareþ þe meyde stedfaste / Vrin trouble in þe whiche
sede appereþ in þe grounde hit signifieþ þe womman to be with a man
{f. 35v} late / Vryn of a womman þicke declareþ hir to be corrupte . þe
foreseide vrin namely trobul in whiche sede appereþ in þe grounde of
þe vessel / For if suche vrin be in a man hit declareþ him to haue delite
with a womman / Vryn of a womman þat is menstruate makeþ a blody
From Clerks to Corpora
vryn / If þis blode be cruddid in a womman she shal seme with childe
til hit be dissolu/e\d . as it is shewid in Antidoda .
Here endeþ þe tretise of þe cercles of vryns with her significacions
. and also also þe propirtes {f. 36r} of vryns . and /þe\ ouer passing
excesse of hem . and nowe foloweþ remedy and medysyn for meny of
hem þat ben rehersid .
MEdisyn for colour as camelle flesshe / Take þe croppe of sauge . and
drinke trisandala . and auente þe in þe leggis / Or ellis take þe ynner
barke of eldryn tree and grynde hit with ale and clense hit . and drinke
hit seuene sponeful at ones and vse hit / al liuida ¶ Plumbea vrina .
For þis . take scamonie and do hit in a pere and roste hit and vse hit /
Also make a poudre of þe rotis of pyony and persil rotis and þe barke
of an {f. 36v} oke and vse hit daily til þu be hole / ¶ Pallida vrina . For
þis . take mirabolane . indi . aloes . epatice . cene . sugur . of alle þes a
dramme and make a poudre of hem and vse in þy potage or in drinke /
¶ Viridis vrina . For þis . take gynger . canel or synamon and sene . and
tempere hem with clarified hony . and vse þis at morwe and at euen .
¶ Inopos vrina . For þis . take osemounde . saueyn . and gromel . and
boyle hem in white wyne and a litil hony and streine it {f. 37r} and vse
þis drinke erly and late / ¶ Subpallida vrina . For þis . take poudre of
grete toonecrasses sede . and vse hem in þy potage . ¶ Rubicunda vrina .
For þis . take vynegre . mustarde and boyle hem togedir . þen take talwe
as moche as a notte and do þerto . and ete þis when þe sekenes takeþ þe
/ vse also dyacameron and trisandala . ¶ Subcitrina vrina . For þis . take
polipodium . anes . alloe . epatice . meddele hem with sugre . and make
a poudre of þes and vse in þy potage . {f. 37v} ¶ Subrufa vrina . For þis
. take poudre of elenacampana . turmentille medle hem with hony clarified and boyle hem togedre and make þerof a confeccioun and use hit
day by dai / al subrubia . ¶ Subrubicunda vrina . For þis . / vse þe same
medisyn þat longeþe to subrufa . as it is nexte afore . ¶ Rubea vrina .
For þis . take triacle and boyle hit in þe iuse or water of moleyne and
vse hit daily / Also take cassiafistula . tamarindes and triasandali and
medle hem togedir and vse þe medisyn daily / {f. 39r} ¶ Citrina vrina .
For þis . take fyges . licoris . ysope . elenacampana . horehonde . yliche
moche . boyle hem in a galoun of water into þe halfe be wastid . and
vse þis drinke . at euen and at morwe vij sponful warme . ¶ Subrubea
vrina . For þis take water crasses and grynde hem smale and boile hem
in wyne . and vse it / ¶ Rufa vrina . For þis take hony and seþe hit til hit
be harde and do þerto poudir of sene and gynger . and make a plaster
of dewete and of wyne / If it be þe flores ley it on þe nauyl . {f. 39v} ¶
The Late Middle English Version of Practica Urinarum in London
Kaynos vrina . For þis . take þe bone of an hertis herbe and make þerof
/a\ poudir and medle hit with diacameron . and vse þis poudir /in\ þy
potage or ellis in þy drinke . ¶ Lactea vrina . For þis . take þe sede of
white popy and bray hem smale and putte hem in letuse water or in þe
iuse þerof . and medle hem togeder and drinke þat licoure . and vse hit
. And anoynte þy templis with þis onyment made of henbane . letues
. popi and mandrake . and vse hit daily . ¶ Subuiridis vrina . For þis .
{f. 40r} lete þe seke blede vnder þe ancle . or in þe veyne of þe arme .
and lete him take a pynte of white wyne . and þeryn boyle a dramme of
alloes and anoþer of epatice and drinke of þat a sponeful /at\ morowe
and anoþer at euen . first and last / ¶ Alba vrina . For þis . take saxifrage
. gromel . persile . and sage . yliche quantite and boyle hem in gode stale
ale . and vse þat drinke on þe morowe and at euen . and also make of
þes herbes a poudir and vse hit daily in þy potage . ¶ Nigra vrina . For
þis . drinke {f. 40v} þe iuse of celidon and of moleine and triacle with
wyne . and drinke hit at morwe and at euen . ¶ Glauca vrina . For þis
take þe iuse of tansy . þe /ribbe\ worte and boyle hem in /white\ wyne
with a litil hony and vse hit And þus þis practif of fisike endeþ /
¶ Of euery sekenes thre tymes þer be þat muste be obserued and
tendrid . þat is to wite . þe begynnyng . þe state and þe ending / If a
man sekele of þe axes and be in þe begynnyng {f. 41r} þerof take for
him percelly and fenel . an handful of eche . and seþe hem in a galoun
of welle water vntil halfe be almoste wateri . þen take halfe an handful
of violettis and putte þerto . and let/t\e al /þis\ seþe til hit be incorporate and of oon strengþe . þen streyne hit þoroughe a lynen cloþe and
close hit in a close vessel . and vse þis drinke fyue daies contynuelli vij
spoonful at morwe and at euen as moche luke warme . and sanaberis
∴ or sumwhat more of almondis . and a gode handful of violettes ∴
{f. 41v} ¶ For him þat is in þe axes in state . Take sage . meyntis .
foþeþistil . and lady þistil and sorel yliche moche . and wasshe hem
clene and bray hem in a morter and take þe iuse of hem and putte in a
potel of mylke . þen sette hit ouer þe fire and lete hit seþe til þe mylke
may take þe vertu of herbes . þen take a quarte of gode stale ale and
put þer to . or þre peyntes if hit so nede . and make a possette and vse
þe drinke ix daies in al dryness and hetis nyȝt and day as ofte as þu
wilt . and þu shalt be hole by þe grace of god . {f. 42r} ¶ For him þat
is trauelid with þe feuers daily or oþer . eiþer day . Take camylmelle
and stampe hit . þen streyne oute þe iuse þorough a canues . /þen\ take
þat and sette /it\ ouer þe fire and lette hit seþe . and alwey scome awey
þe fome til hit be clere . and if þu wilt þu may put a quantite of sugur
From Clerks to Corpora
þerto to make hit þe more liking and when /it\ is clere . take hit doune
and lette hit kele . þen put hit in a close vessel . and vse þe drinke ix
daies . vj or vij sponful at ones in þe mornyng and in þe euenyng luke
warme . And þe þenke camelmyl ouerbitter þu may take sorel . in þe
selfe wise . {f. 42v} ¶ For to swage and abate þe malice of þe hote axes /
Take mandrake . sorel . syngrene and dauntlyoun yliche moche . and
seþe hem in barly mele and vynegre and make of þis a plaster . and ley
hit to þe sides of þe seke and to his breste and to þe stomake and to þe
powes of þe handes and of þe fete and he shal be be hole in short tyme
¶ For him þat pisseþ blode / Take wilde sage . saxifrage . percely yliche
moche and stampe hem . and put þe iuse in a potel of mylke and lette hit
seþe a while . þen streyne hit . and vse þe drinke ¶ For dymnes of yen .
take cen{f. 43r} tuary and planteyn yliche moche . seþe hem in white
wyne and put a litil quantite of piper þerto . þen streyne al þorough a
clene cloþe and drinke þis medisyn luke warme last when þu goest to þy
bedde / ¶ For þe coughe / Take elemini fenel sede . longe piper . commyn
. and hony as sufficeþ . Of elemini take . dragme . 1 . 3 . handful of fenel
sede and longe piper ana . dragme vj . of commyn halfe an vnce . and
seþe al in a quantite of hony as nedeþ and vse þe drinke . first and laste /
¶ Electuary for þe coughe / Take cynamun . spice . gynger . {f. 43v}
galangal . aneys . longe piper . dra/ga\gante . ana . dragme semis . þe
kernel of pyne clene clensid . dragme . drie figes . dragme ij . and clarified hony as nedeþ and make electuari and vse hit ¶ Sargarismus . for
purgyng of þe hede / Take mustard staffeacre . pelleter . cokil . comfery
spiconarde . hony . vinegre . or white wyne menge al togeder and take
ijº sponful þerof and holde in þy mouþe and rense hit wel þerwith . vntil
þe fleume descende . þen spet oute þat and take more . so ofte . til þu be
esid ¶ For a collerike man þat is dissesid with þe axes þat is caused of
{f. 44r} brennyng coler and hete . and also for to ease þe grete dryness
þat he haþ by þis sekenes / Take borage . violettes . sorel . letwis . endif
. souþeþistil . stra/u\bery leues . mary goldes . þe tendrons of vynes .
dauntlyon . pety morel . malowis . sengryn . persely . langedebef . an
handful of eche with þre or iiije stickes of likeris ybrusid . to pouder .
and al yput in a galoun of welle water and yconsumed and soden vntil
halfe be wastid . þen take hit doune and streyne hit and put hit in a glas
. or in a clene erþen pot wel closid . and vse þerof at morwe and at euen
vj or {f. 44v} vij sponful luke warme and sanab/is\ Gargarismus . for þe
hede þat is stoppid with reugme and viscous fleume .
{f. 45r} ¶ Ther ben iiij ages in whiche euery humoure haþ his
duracioun to her likenes . þat is childehode . bachelerhode . manhode .
The Late Middle English Version of Practica Urinarum in London
and fadirhode / Childehode is fro þe begynnyng of birþe into xv ȝere /
bachelerhode . fro xv /ȝere\ vnto xxx . ȝere / manhode fro xxx ȝere vnto
lx ȝere / faderhode and olde age fro lx ȝere vnto þe ende of man is lyfe
/ ¶ And þer ben iiij cesons in þe ȝere apriped to þe foure complexions
of men / ver . sommer . haruest and wynter / Ver . is hote and moyst /
sommer is hote and drye / haruest is colde and drye / wynter is colde
and moist / ¶ Collerike men ben wel in vere . and yuel in sommer / and
wel in haruest and in wynter beste at ease / Sanguyne folke ben ywel
in ver . and better in sommer and haruest and wynter best at ease /
Fleumatike folke ben wel in ver in haruest best and sommer best . and
worste in wynter / Malencolious folke ben best in ver and somer .
{f. 45v} and worste in haruest . and wynter is not gode / Also þer ben
iiij wyndes . eest wynde is moyst and hote / souþe wynde is hote and
drye / west wynde is colde and moyst / Norþe wynde is colde and drie /
And þes ben þe iiij parties of þe worlde . eest . west . souþe and norþe /
¶ Collerike folke ben wel while þe sonne and þe wynde ben eest . west
. or norþe / and þey ben yuel . while þe sonne and þe wynde ben souþe
/ for þat is her quarter / Sanguyne folke ben yuel when þe sonne and
þe wynde ben eest . for þat is þeir quarter . and þey ben wel when þe
sonne and þe wynde ben souþe . west or norþe / Fleumatike folke ben
wel when þe sonne and þe wynde ben eest . souþe or norþe / and þei
ben werste when þe sonne and þe wynde ben west /for þat is þeir quarter\ / Malencolious folke ben werst when þe sonne and þe wynde ben
norþe {f. 46r} and þey ben wel when sonne and þe wynde ben souþe .
est . and west / souþe is þeir quarter / Nowe folweþ þe reignyng of þe
iiije humours in euery man or womman . ¶ Fro þe iij of þe clocke after
mydde nyȝt vntil ix of þe clocke before no/u\ne . regneþ sanguyne / and
from ix of þe clocke afore none vntil þre after no/u\ne . regneþ colerike / and from þre of /þe\ clocke after noune vntil ix before mydnyȝt .
regneþ fleumatike / and from ix of þe clocke before mydnyȝt vntil iij
. after mydnyȝt . regneþ malencoly / And euery man in þe tyme of þat
humoure þat is most regnyng in him / and in þe tyme . þat his humour
regneþ . is most greued ∴ ¶ For þe morfu . wasshe al þy body with
water þat stampid garleke is soden yn . and þy body is dried and clene
. rubbe al þe morfu with garlike and affadile stampid togeder . and vse
it so a while and drinke federfoy . eldren fumetory and rede docke rotes
made in {f. 46v} tysane . and euery morwe erly drinke a gode draugte .
and vse þis medisyn ix daies and þu shalt be clene of þe morfu with yn
and wiþoute . if þe medisin be wel vsid ¶ For þe iavndise take greynes
clowes mases galenga notemygges of eueryche of þes an halfe peny
From Clerks to Corpora
worþe / of ysope of hertistonge of celidon ylichemoche and seþe al in
þre quartis of white wyne vnto þre pyntes . and vse þis medisyn warme
at morwe /and\ at euen . vj or vij sponeful at ones.
Appendix 1
Figure 2. MS Wellcome 537, f. 15r.
© Wellcome Library, London, as the owner of the manuscript.
Calle Martín, J. & Miranda García, A. (2011). From the manuscript to the
screen: Implementing electronic editions of mediaeval handwritten material. Studia Anglica Posnaniensia 46:3, 3–20.
The Late Middle English Version of Practica Urinarum in London
Calle Martín, J. & Miranda García, A. (2012). The Middle English Version of
De Viribus Herbarum (GUL MS Hunter 497, ff. 1r-92r). Bern – Berlin –
Bruxelles – Franfurt – New York – Oxford – Wien: Peter Lang.
Calle Martín, J. (2012). A late middle English version of The Doom of Urines
in Oxford, MS Rawlinson C. 81, ff. 6r-12v. Analecta Malacitana 35:1–2,
Clemens, R. & Graham, T. (2007). Introduction to Manuscript Studies. Ithaca
and London: Cornell University Press.
Denholm-Young, N. (1954). Handwriting in England and Wales. Cardiff:
University of Wales Press.
Derolez, A. (2003). The Palaeography of Gothic Manuscript Books. From the
Twelfth to the Early Sixteenth Century. Cambridge: Cambridge University
Hector, L.C. (1966). The Handwriting of English Documents. London: Edward
Keiser, G.R. (1998). A Manual of the Writings in Middle English 1050–1500.
Volume X. Works of Science and Information. New Haven, Connecticut:
The Connecticut Academy of Arts and Sciences.
Lewis, Robert E., Kurath, H., Sherman, M.K. & Reidy, J. (eds) (1952–2001).
Middle English Dictionary. Ann Arbour: University of Michigan Press.
(Electronic dictionary: <>).
Moorat, Samuel A.J. (1962). Catalogue of Western Manuscripts on Medicine
and Science in the Wellcome Historical Medical Library. Vol. I. Manuscripts
Written before 1650 AD. London: Publications of the Wellcome Historical
Medical Library.
Moreno-Olalla, D. and Miranda-García, A. (2009). An annotated corpus of
middle English scientific prose: aims and features. J.E. Díaz Vera and R.
Caballero (eds) Textual Healing: Studies in Medieval English Medical,
Scientific and Technical Texts (123–140). Bern – Berlin – Bruxelles – Frankfurt –
New York – Oxford – Wien: Peter Lang.
Petti, A.G. (1977). English Literary Hands from Chaucer to Dryden.
Cambridge, Mass: Harvard University Press.
Roberts, J. (2005). Guide to Scripts Used in English Writings up to 1500.
London: The British Library.
———. (2005). The Twenty Jordan Series: an illustrated middle English uroscopy text. ANQ: A Quarterly Journal of Short Articles, Notes and Reviews
18:3, 40–64.
From Clerks to Corpora
Tavormina, M.T. (2009). Practice, theory and authority in a middle English
medical text: ‘Barton’s urines which he treated at Tilney’. Origins of
Nephrology 22, S33–S41.
Voigts, L.E. & Kurtz, P.D. (2000). Scientific and Medical Writings in Old and
Middle English. An Electronic Reference. Ann Arbor: Michigan University
4 Is Plant Species Identification Possible
in Middle English Herbals?
David Moreno Olalla
Universidad de Málaga
Brother Cadfael tucked up his habit and ran for the shelter of the
cloister, there to shake off the water from his sleeves and cowl, and
make himself comfortable to continue his reading in the scriptorium. Within minutes he was absorbed in the problem of whether the
‘dittanders’ of Aelfric was, or was not, the same as his own ‘dittany’.
Ellis Peters, The Devil’s Novice (1983), Chapter 1.
The problem
The transmission of the classical textual corpus during the Middle Ages
was complicated because the oldest volumes that served as exemplars
for the medieval manuscripts were copied and recopied at monasteries, universities and private scriptoria. The sad result was that many
MSS present gibberish fragments due to the many scribal mistakes and
hypercorrections that accumulated in them during the several copying
processes. This is of course an academic truism, which we may call
the universal cause of error as it was at work in virtually every medieval textual tradition regardless of its actual subject. But in the case of
texts dealing with the Natural Sciences, and very acutely in pharmacological treatises, a second, particular cause should also be taken into
account. This has to do with the fact that Western phytography, i.e.
plant description, was still in its infancy, having receded rather than
proceeded since Graeco-Roman times. In fact the opinions held by
the likes of Dioscorides or Pliny the Elder in their fundamental treatises
How to cite this book chapter:
Moreno Olalla, D. 2015. Is Plant Species Identification Possible in Middle English
Herbals? In: Shaw, P., Erman, B., Melchers, G. and Sundkvist, P. (eds) From Clerks to
Corpora: essays on the English language yesterday and today. Pp. 53–70. Stockholm:
Stockholm University Press. DOI: License: CC-BY.
From Clerks to Corpora
(De materia medica and Naturalis historia)1 were still taken as law without much ado fifteen centuries after they were first put down in writing.
This argumentum ab auctoritate moved in two directions. On the one
hand, it seems to have hampered any real development in botanical fieldwork, and so the descriptions appearing in medieval treatises remained
vague. On the other hand, the desire to interpret the Classical texts correctly led the medieval scribes to collect all possible synonyms from all
accessible sources, forgetting that the same species may/will be known
under different names in different places and oppositely that different
species may be designated with the same name,2 which in practice meant
the sloppy application of syllogisms (‘A is B and B is C, therefore A is C;’
in relation to this, see Moreno Olalla 2013a: 398–399). This (mis)treatment of plant-names is behind the well-known problem that it is very
difficult sometimes to be certain as to the actual species being mentioned
in a medieval text: more often than not, at least two or three different
species are theoretically possible. Of course the problem is most acutely
felt with laconical texts such as synonyma, which seldom provide a physical descriptions of the plants, but it can also usually be detected in more
verbose texts such as receptaria and even medical herbals.
In the particular case of medical and botanical treatises composed in
Atlantic countries such as England, the above confusion became worse,
since those species that did not grow in the Mediterranean milieu were
ascribed as a matter of course to plant-names already used by the
Greeks and Romans. The Middle English plant-name and the Latin synonym therefore do not necessarily tally. This last hindrance is still very
well alive in Contemporary English and examples abound: think for
instance of the acacia, which according to the Classical evidence (MM i
101; NH xxiv 109) probably meant a species of the genus Acacia Willd.
(so García Valdés 1998: 201, fn. 201), or perhaps some Mimosa spp.
(André 1956: s. v.; Simpson & Weiner 1989: s. v. acacia1), but which is
1These will be indicated respectively as MM and NH henceforward. I ignore the
information from Theophrastus’s Περὶ φυτικῶν ἱστοριῶν as this work was virtually
unavailable to Western scholars until its 1483 translation into Latin (De historia et
causis plantarum). Note that Dioscorides, just like Galen, was in fact best known in
Western Europe through Latin translations and the works of epigones, but for the
sake of convenience his work will be here quoted in the original Greek.
2Prior 1863: xx mentions the well-known case of the Bluebell, which refers to
Hyacinthoides non-scripta L. in England and to Campanula rotundifolia L. in
Scotland—and, one may add, to sundry species in the US (Mertensia virginica (L.) Pers.
ex Link, Eustoma russellianum Salisb., Phacelia campanularia A. Gray, inter alia), the
Caribbean (Clitoria ternatea L.) or Australia (Wahlenbergia gloriosa Lothian).
Is Plant Species Identification Possible in Middle English Herbals?
commonly understood in English today as the Robinia pseudacacia L.,
a species brought from the New World (Prior 1863: 1).
The outcome of all these factors is that we cannot always know for
a fact which plants are really being treated by the writer of a Middle
English treatise: frequently a very broad identification by genus is the
nearest we can get, while in some cases we must rest content if we
can identify the family to which some obscure plant-name refers. But
a more precise identification by species is occasionally possible. In the
following pages I intend to (a) comment on problems, both medieval
and contemporary, that are often encountered in connection with plant
identification, (b) show how the meanings usually given to plant-names
in dictionaries, even in the most scholarly ones such as the Middle
English Dictionary (MED), are sometimes vitiated because of those
very problems and do not withstand a careful textual analysis, and
hence the species provided by general lexicographical works should
never be accepted at face value, and (c) suggest that despite all odds, it
is sometimes possible to identify the species meant by the writer—or, at
least, to provide very educated guesses.
I will draw my examples from Lelamour’s Herbal (LH henceforward). This is a medicinal collection of 214 plant species,3 alphabetically
arranged by their English names and kept in London, British Library,
Sloane MS 5, ff. 13ra–57ra (S for short). Although in its explicit the
piece purports to be a Middle English translation of Macer Floridus’s
De Viribus Herbarum made in 1373 by an otherwise unknown Hereford
schoolmaster called John Lelamour, the version preserved in S was actually composed in the 1460s near or in London and it is best described as
an assortment of entries drawn from different ME traditions.4 It would
be a moot question even to decide whether Lelamour translated some
entries himself: the Rue Herbal and Agnus Castus, the two main detected sources for LH, were already in English in MSS older than S.5 Apart
from the substantial number of entries, which makes it one of the most
3 It would be more accurate to say that there are 214 entries but no more than 211
species, for at least three of them were treated twice (Moreno Olalla 2007: 120).
4 The only 14th-century piece in the MS is a brief botanical trilinguale (ff. 4r–12v),
which was bound together with the rest of the volume at a later date: the wear and
tear and the dirt on ff. 4r and 13r suggest that these pages were left uncovered for
quite a long time. On f. 3r–v there is a late 13th-century medical fragment in Latin
on urines.
5 See Moreno Olalla 2007: 122; 2013b: 948 about the Rue Herbal. Agnus Castus
was edited in Brodin 1950.
From Clerks to Corpora
important herbals in Middle English (Hunt 1989: ix), LH is actually quite
an average work; as is usually the case with vernacular herbals, there are
no illustrations, and whenever the text provides a physical description of
the species, this is by no means detailed. Broad remarks on the size and
colour of leaves and flowers, and sometimes the presence of bulbs and
seeds, is about all one can wish to find there. A detailed comment on the
usual habitat of the species is the exception and not the rule.
I will reference my quotations from this source using a folio/line system, since LH remains unpublished even though it has been known since
the late 17th century and used as a source text since the 1840s at least.6 It
has been the subject of two theses as well. The first of them was an M.A.
completed in the late thirties (Whytlaw-Gray 1938), which was used by
the editors of the MED. Whytlaw-Gray’s editorial and lexicographical
approach to the text was considered dated and inaccurate at times, and
so a fresh edition, following more modern criteria, was included as part
of a Ph.D. dissertation a few years ago (Moreno Olalla 2002).
Scribes behaving badly: Carthnote
I would like to begin with perhaps the most simple form of distortion:
scribal mistakes and overzealousness through the ages and their modern consequences. On f. 20va/6–19, in the section of the herbal for
plant-names beginning with the letter ‹C›, we read the following:
Carthnote, that is an erbe that haþe levis like to fenell and wiþ flouris
and smale stalkys. He growiþ in wodis, also in medis. The vertu of
this erbe is þis: that, and he be stampid and laid to a sore, he will
feche a-wey all dede fleshe and helpiþe renewe the quyck fleshe. Also
stampe this erbe and put him to þat place þat lackiþ here: he shall
restore hit a-gayne with-in schorte tyme of plaster layeng.
According to MED (Kurath et al. 1954–2001: s. v.), carthnote is a
hapax legomenon vaguely defined there as ‘[s]ome medicinal plant’. It
is certainly possible to go further than that just by reading the whole
entry: the text itself provides us with several clear clues, beginning with
6 Sloane 5 was catalogued already in Bernard 1697: ii.251. To my knowledge, LH
was first quoted in Halliwell-Phillipps 1889: i.xx. The excerpt chosen was the entry
Mowsere, i.e. Mouse-ear (Hieracium pilosella L.). The MS was also perused by contributors to the NED/OED, as witnessed by several quotations contained therein
(see Moreno Olalla 2007: 119 for a list of entries).
Is Plant Species Identification Possible in Middle English Herbals?
the accompanying Latin heading. Cidanum is not recorded in any text
or glossary I know of, but Cidamum is (the faulty reading in S can be
easily explained as a scribal misreading/mispelling of the original cluster of minims). This word appears in another important medical herbal,
Agnus Castus (AC henceforward), which offers almost a twin entry of
S, although it also provides a collection of synonyms that is missing in
LH: ‘dilnote or slyte or haylwourth’ (Brodin 1950: 216).
The lists of synonyms provided in AC and other sources such as
Alphita (see next paragraph) make it clear that cidamum must be taken
as a mistake for *Ciclamen, while carthnote is obviously another error,
for *earthnote this time (cf. AC corhnote < OE eorðnut) due to the
Lombardic initial ‹E› in the exemplar, which probably had a round
shape. Cyclamen was used to refer to the name-sake genus and especially to the Sow-bread (C. purpurascens Mill.), and indeed both virtues
of carthnote mentioned in S (against wounds and alopecia) are also
reflected in Dioscorides’s account of the Sow-bread (MM ii 164).
Still, there is an important detail in the English text that seems to
gainsay this identification: according to Dioscorides ‘the cyclamen has
leaves similar to those of ivy’ (κυκλάμινος φύλλα ἔχει ὄνομα κισσῷ)7
while the author of LH wrote that ‘[e]arthnote […] haþe levis like to
fenell’. The connection between both texts can, nevertheless, be maintained. Cf. the following definition from Alphita: ‘[c]iclamen uel ciclamum, sive citeranum, panis porcinus, malum terre idem. angl. dilnote’
(Mowat 1887: 39). The synonym panis porcinus restitutes the lost link
between both texts again, cf. ‘[p]anis porcinus, ciclamen, malum terre
idem. a[nglic]e dilnote uel erthenote’ (Mowat 1887: 134). The identification of erthenote with cyclamen is assured as well by the translation of the original L. succum ciclaminis as ‘þe Iuse of erþenote’ in the
Cyrurgie of Guy de Chauliac (taken from Kurath et al. 1954–2001: s.
v. ērthe).
Yet another problem remains, not only in LH but in many medieval
texts including Alphita or AC. OE eorþnut did not refer to the Sowbread, but was an umbelliferous species, identified as some Bunium sp.,
especially the Earth-chestnut (B. ferulaceum Sibth. & Sm.), or else the
Pig-nut (Conopodium majus (Gouan) Loret = B. flexuosum Stokes).
This is confirmed by the explicit mention of the similarity of this species
with dill and the resemblance of its leaves to those of fennel. I think
that the confusion between both plants ultimately lies in a mistaken
7 Translations from Greek are my own.
From Clerks to Corpora
reading in NH xx 21, where Pliny confused the Greek plant-name
βουνιάς (L. napus = OE næp) with βούνιον (L. bunium = OE eorþnut;
see André 1956: s. v. nāpus). Napus, on the other hand, was also called
rapum in some Post-Classical sources; and finally, rapum was also a
name for cyclamen (vid. André 1956: s. vv. cyclamı–nos, nāpus, rāpum).
The tiresome scribal thirst for synonyms, which is so frequently encountered in medieval glossaries, equated both plants even though they were
clearly distinguished in the Classical literature. The misidentification
seems to have passed unnoticed to scribes perhaps because all those
plants present a big edible bulb which swine craved after.8
As a short excursus, I would like to highlight here that misreadings are
not peculiar to medieval scribes only, but are shared by modern researchers and unsuspectingly transmitted by serious scholarly works sometimes.
Drawing from LH, MED includes an entry ara-wŏrt, which is laconically
defined as ‘[a] Flowering plant of some kind’, and even given a tentative
etymology: ‘[c]p. wŏrt plant, & ?arwe arrow’. The word is presented as a
hapax taken from the entry Pes columbe (probably, the Soft Cranesbill,
Geranium molle L.) in LH: ‘Coluyr-fote is an herbe, his levis beþ like
to araworte’. This should in fact be put down as a ghost entry, for it is
due to a faulty reading by Whytlaw-Gray that crept into the dictionary.
The manuscript actually reads ‹Maworte›, i.e. some Mallows (Malva sp.).
Comparison of the flowers of the Cranesbill with those of the Mallows
was traditional: it is used already in MM iii, 116: καλεῖται δὲ ὑπ’ ἐνίων καὶ
ἕτερον γεράνιον, ἔχον […] φύλλα μολόχῃ ἐμφερῆ ‘Some call another species
“Cranesbill”, one that has […] leaves like those of the Mallows’.
8This also explains why several fungi of the genus Tuber L. (i.e. the truffles), the
arachis, etc. were also called ‘earth-nut’. Already in NH xxv 114, rāpum, tu ber and
cyclamınos are said to be similar plants, providing with an early instance of quasi-confusion between them. This mistake was transmitted to the late OE Durham
Glossary of the names of worts, which seems to be the first instance in English where
cyclaminos is equated with some plant different from the Sow-bread: ‘Cyclaminos,
Eortheppel, slite, attorlathe’ (Cockayne 1864–1866: iii-301). Judging from the
English names (see Bosworth & Toller 1898: s. v. átorláþe), the Cyclamen was
apparently confused here with the Mandrake. Note that the Mandrake was called
malum terrae (André 1956: s. v. mandragoras), but this name was also applied to
the Cyclamen sp. (André 1956: s. v. cyclamı nos), thus providing a bridge between
both plants: cyclamen <---> malum terrae <---> mandragoras.
Is Plant Species Identification Possible in Middle English Herbals?
Habits die hard: Horse-þistill
Another common mistake in modern works is to accept uncritically
the identification of a ME plant-name suggested traditionally or on the
basis of a single scholarly work, even though the context in another
treatise may be against that equation. On f. 22vb/28–23ra/18, for
example, we can read the following:
Endyue ys an erbe that som men calliþ horse-þistill. Þis erbe haþe
prykkys with-oute. Þe lefe ar longe and when he is brokyn he dropiþ
mylke, and he haþe a litell yelow flour and his sede blawiþ a-waye
with þe wynde as doþe dent de lyon. The vertu of him is þis: take
þe juis þere-of and medill hit with hote water and drynke hit, and
þat heliþ þe stoppinge of þe mylte and þe lyuer. Also þis erbe is gode
y-dronke for þe jaundys and for þe feuer tercian and for þe hote
postem. Also þis erbe a-swagith þe grete hete of þe lyuer and of þe
stomake, for he is colde and moiste.
In Present-day English, Endive is a common name for two species of
Chicory, a family of Compositae (Cichorium spp.): C. intybus L., indigenous to Europe, and C. endivia L., which was imported from China as
early as the 16th century. Obviously only the former could fit in here;
but the description does not fit at all: C. intybus bears bright blue flowers,9 and its leaves do not have ‘pricks’, as stated in the text.
The editor of AC (Brodin 1950: 221), a text which shares this
entry with LH, was evidently aware of the impossibility of accepting
a Cichorium species, so in the accompanying glossary he suggested,
perhaps on account of the yellow flowers, that the plant intended here
might be some wild lettuce, Lactuca virosa L. or else L. scariola L. This
identification was accepted by MED, where the text of LH is actually
quoted (Kurath et al. 1954–2001: s. v. thistel n., sense b). Still, this identification should be rejected since the general look of these species is not
reminiscent of a thistle at all, see Figure 1 for details.10
Again, reading closely the physical description provided in the actual
ME text—which is uncharacteristically detailed—will not go unrewarded. The comparison of this species with a thistle, implicit in the ME
designation and explicit in the mention of a dandelion, and the colour
9 ‘Blue daisy’, ‘blue dandelion’, ‘blue-sailors’ are modern synonyms for this species.
The engravings used in this chapter were taken from Bauhin 1598, and kindly
provided by the Missouri Botanical Garden.
From Clerks to Corpora
Figure 1. Θρίδαξ ἀγρία–Lactuca sylvestris (after Bauhin 1598: 400).
of the flowers make me think that the plant actually intended here may
well be some Helminthotheca sp., a genus akin to Cichorium that normally displays yellow flowers. There is one species in particular of this
genus, the Bristly Ox-tongue (Helminthotheca echioides (L.) Holub)
that fulfills the physical description in LH quite well. Its leaves and
bracts are noticeably covered with white bristles that very much resemble small ‘prikkys’, just like any other Liguliflorate Composite this species yields a white latex when the stem is broken (‘when he is brokyn
he dropiþ mylke’), and its pappused achenes are easily blown away
with the wind (‘his sede blawiþ a‑waye with þe wynde’). It is interesting
also to note here that the distribution of this plant in Britain seems to
be the south-west counties (Martin 1965: 50), and this is exactly the
same area where, to judge from the linguistic evidence, the text may
have been originally composed (Moreno Olalla 2007: 126–132). Since
the description of the species seems not to have been recorded in the
Classical literature, could this support the hypothesis that the author of
LH did some field work after all?
Is Plant Species Identification Possible in Middle English Herbals?
One size fits all: Affodil
I would like to stress further the pitfalls of accepting uncritically the
meanings provided by modern lexicographical works through the case
posed by the species called Centum capta in the Latin heading and
­affodill in the ME text. The fragment where this plant is described in
LH (f. 14va/1–29) runs as follows:
Centum capta
Affodill is an herbe þat beriþe a faire yolewe floure and at þe toppe
he haþe ronde coddys, in þe whiche he berith sede and his levis beth
smale and longe. The vertu of him is that þe branchis of this erbe ben
gode to hele þe dropesy. Also drynke þe juis of þe flourys of the[s]11
erbe in wyne and that will sle byting of venymous wormys. Also
take þe more of this erbe and þe juis of his leuys and a litell safar, lat
this boyle to-gadrys with swete wyne streyned fayr: hit is gode for
renynge eyen. Also þe more j-brent and made to pouder temper þat
with a litell oyle, a-noynte that place wher that lackyth here and hit
shall make hit to growe a-yene. Also, and a harde sharpe cloþe be
wette in þe juis of this erbe, let rubbe the morfue with that cloþe and
hit shall fall a-way, for this erbe is hote and drye in the secund degre.
At least three different species from three different genuses can be
proposed as the plant referred to here, depending on which authority we accept. MED (Kurath et al. 1954–2001: s. v. affodil, sense a)
states that this plant-name normally stands for the Ramsons (Allium
ursinum L., also called Wild, Bear or Wood Garlic; an image is given
below as Figure 2), and provides thirteen illustrative quotations. The
meanings ‘asphodel’ (sense b, three quotations, two of them taken from
synonyma) and ‘rhododaphne’ (sense c, a single quotation) are preceded
by a question mark, which denotes that these senses are uncertain—
all in all a sensible editorial policy: we have just seen how compilers
were not too particular when it came to the gathering of synonyms and
so, their equivalences should be taken with a big pinch of salt. In the
same vein, Tony Hunt identified three Allium spp. as the species behind
Latin “Centum Capita”: A. ursinus, A. vineale L. (Crow Garlic) and
A. schoenoprasum L. (Chives). It is worth mentioning that, even though
other ME synonyms are given in Hunt’s glossary (ramese, crowgarlyk,
wilde garlek, civys and maudefeloun), affodill stands out as the most
usual ME plant-name to refer to this Latin species (Hunt 1989: s. v.).
11 ‹s› added over the line.
From Clerks to Corpora
Figure 2. Allium ursinum (after Bauhin 1598: 422).
It is only natural therefore that one would assume that the entry in LH
deals with Ramsons too.
Still, the physical description provided in the entry may point to a second candidate: the species referred to by the medieval author could also
be the daffodil (Narcissus pseudo-narcissus L.; see Figure 3). Indeed,
the confusion between the Asphodelus and the Narcissus species in
England can be traced at least to the 1550s, as recorded by William
Turner in his New Herball: ‘I could neuer se thys herb [ie. asphodelos,
‘ryght affodill’] in England but ones, for the herbe that the people calleth here Affodill or daffodill is a kynd of narcissus’ (Turner 1551: B.iij
verso); in fact, the very word ‘daffodil’ is etymologically connected with
‘affodill’ (see Simpson & Weiner 1989: s. v. daffodil for details). This
possibility is not contemplated in Kurath et al. 1954–2001: s. v. affodil,
but in view of the description and the post-medieval evidence, it may
be worth speculating whether the entry in LH may be after all devoted
to some Narcissus.
So far we have paid attention to lexicography and phytography. Yet if we turn to the list of healing properties proposed in the
Is Plant Species Identification Possible in Middle English Herbals?
Figure 3. Νάρκισσος–Narcissus (after Bauhin 1598: 858).
text, we will notice that the entry in LH fits extremely well with the
Dioscoridean entry ἀσφόδελος (MM ii 169). Compare the following Greek excepts with the virtues given in LH: ὁ δὲ χυλὸς τῆς ῥίζης
προσλαβὼν οἴνου παλαιοῦ γλυκέος καὶ σμύρνης καὶ κρόκου συνεψηθέντων
ἐπὶ τὸ αὐτὸ ἔγχριστον γίνεται ὀφθαλμοῖς φάρμακον ‘Adding old sweet
wine, myrrh and saffron to the juice of the root and boiling all together,
one makes a healing ointment for the eyes’, καεῖσα δὲ ἡ ῥίζα, τῆς τέφρας
ἐπιπλασσομένης, ἀλωπεκίας δασύνει ‘having burnt the root, and anointing [the bald place] with the ash, it brings back the hair’, and ἀλφόν τε
λευκὸν προανατριφθέντα [ἐν] ὀθονίῳ ἐν ἡλίῳ καταχρισθεῖσα ἡ ῥίζα σμήχει
‘The root, used as a salve, wipes off white leprosy if rubbed with a linen
cloth in the sun’. The virtue of affodill against poison mentioned in LH
also appears in MM, although the wording is somewhat different.
The Dioscoridean species has been tentatively identified as
some Asphodelus spp., perhaps A. aestivus Brot. or A. fistulosus L.
(García Valdés 1998: 346, fn. 208), but I think that we can attempt a
From Clerks to Corpora
different species identification. There is indeed a yellow-flowered species (Asphodeline lutea L.),12 but this was at the time a native of the
Eastern Mediterranean only and, as such, apparently unknown to the
Classical and medieval authors. If the Latin heading in LH, Centum
capta (for *capita, meaning ‘a hundred heads’) and the mentioning of
‘ronde coddys’ at the top of the stalk is taken into account, Asphodelus
aestivus Brot. is a very attractive candidate (cf. the capsules at the end
of the stalk of Figure 4). This is the third, and last, possibility.
Identification of the species treated in LH with the Ramsons presents two almost insurmountable problems. To begin with, the general
description of the plant in the ME entry is very much against this idea:
the leaves of the Ramsons are long indeed, but also rather broad, as seen
in Figure 2 above, while their flowers are conspicuously white. Second,
the pharmacological virtues proposed for this species, which are apparently the same as those of the common garlic (Allium sativum L.; see
12 See
Figure 4. Ἀσφόδελος–Hastula regia (after Bauhin 1598: 450).
Is Plant Species Identification Possible in Middle English Herbals?
NH xix 116), do not match at all with those appearing in LH. Since
neither the physical nor the medical description favour this identification, we can, I think, safely reject this candidate MED notwithstanding.
Note moreover that A. ursinum is not a Mediterranean species, and
this runs counter to the other entries in LH, which ultimately derive
from Classical sources.13 We have then to decide between the daffodil
(as suggested by the general physical description), and the asphodel
(as suggested by the matching virtues with Dioscorides’s account and,
partially, also by its physical description). Two factors should be taken
into account before deciding which is likely to be the actual species in
the ME text.
As mentioned above, LH is a compilation of several Latin sources:
this means that as a rule those works by Dioscorides, Pliny or Galen
were the ultimate sources of information, rather than the reflection of
any personal fieldwork undertaken by Lelamour himself—or indeed
any other English contemporary botanist. Therefore, that particular
entry in Pliny’s Natural History or Dioscorides’s Materia Medica that
fits the Middle English text best will probably be the plant that the
original ME translator intended to describe and extol in his text. And
it happens in this case that the virtues offered in the Middle English
match perfectly those of the asphodel, having nothing to do with those
of the daffodil, as they are given in MM iv 158 or NH xxi 128.
The second factor to bear in mind is purely pharmaceutical, and has
to do with the chemical constituents present in those species. There
is only one property of the daffodil worth mentioning: it contains
an alkaloid, called narcissine after the plant, which is emetocathartic
and phlogogenetic (i.e., induces vomit and causes inflammation; see
Felter & Lloyd 1905: s. v. narcissus). Moreover, narcissine has strong
stupefacient properties as well: cows avoid the plant, for eating it could
paralyse them for some minutes (Font Quer 1987: 911). This feature
was already known to the Greeks—although, contrary to NH xxi 158,
the plant-name is apparently unrelated to the root meaning ‘numbness’
Ramsons, in fact, is not recorded in the Dioscoridean entry for Garlic (MM ii 152),
nor appears in Mattioli, 1558; the image used for Figure 2 is actually an addition
by Caspar Bauhin (that is why there is no Greek name in the caption), who included
it in his edition of Mattioli together with the description and an image of ‘Allium
Anguinum’ (i.e. A. victorialis L.), another Allium sp. that he found in the Sudetes
(‘montes qui Bohemiam à Silesia disterminant’) in 1573 (Bauhin 1598: 423).
From Clerks to Corpora
that we can also find in narcotic.14 These two virtues were of course
recorded in MM and would not fail to appear in any medieval treatise,
but they are missing in LH. A lacuna in the text is highly unlikely, since
AC, which again runs parallel to LH here, does not record such properties in any of the several extant MSS either.
These reasons seem to support that the asphodel (Asphodelus aestivus Brot.) is the plant treated here, but the question remains as to
why the general description of the plant is that of a daffodil. Is is an
addition in LH, or else was something taken from his exemplar? The
latter is a more sensible option, if only because this is also stated in AC.
The description of the plant in this textual tradition (161/26·30; Brodin
1950: 124–125) is in any case fuller and different in some minor details:
Affadilla is an herbe þat men clepe affadille or belle blome. It is lyke
to lek and it hath a ʒelwʒ flour, and in þe crop a round codde quanne
þe flour is falle. In quyche is seed lyke to onyoun seed.
Be that as it may, the description of the plant should not be used as
compelling evidence. We should never forget that the plants described
by Greeks and Romans did not always grow on an Atlantic island.
This is in fact a case in point: the asphodel does not grow naturally in
Britain, but the daffodil does. What we have here seems to be a case of
name shift: the original name is used to refer to another species, and
in fact daffodil is but a by-form of an original affodil (see Simpson &
Weiner 1989: s. v.). The second phase of this identification is that of
virtue shift: the medicinal properties of the former are also assumed for
the latter, to such a degree that the early English botanists came to think
that the plant that they were then holding was the same one described
by Dioscorides many centuries before.
We can logically deduce from here that the ME author’s purpose
was to translate the uses of the plant Centum capta, which cannot but
be the Asphodel and, probably, the Asphodelus aestivus Brot., and he
described physically the English plant that he honestly thought was the
asphodel. He actually never intended to talk about a daffodil; had he
known that the plant that he had in front of him as he was writing this
entry was the Narcissus of the Greeks and Romans, not the asphodelus
14‘A narce [i.e. νάρκη] narcissum dictum, non a fabuloso puero.’ Similar remarks
can be found in Plutarch’s Questiones conviviales 647b. Cf. Chantraine 1999: s. v.
νάρκισσος: ‘il ne peut s’agir que d’une étymologie populaire’.
Is Plant Species Identification Possible in Middle English Herbals?
or centum capita, it stands to reason that he would have written a very
different entry.
Quis custodiet ipsos custodes?: Conclusions
Is plant species identification possible in Middle English herbals? I think
that for many cases the slightly disappointing answer is: ‘yes, but…’.
Twisted textual transmissions, recurrent scribal crazes for synonyms
(at least since Greek times), and misapplication of the same names and
virtues to species growing in separate areas of the world but sharing
some feature are conspicuous dangers in this journey. I have provided
a few examples that will hopefully demonstrate that close reading can
and must be instrumental for the job of defusing (albeit just partially)
such minefields.
Perhaps the main problem here is that trustworthy literature on this
topic is scarce in comparison with the attention paid to the same matter
among Classicists (see André 1956, 1958, 1985; Fortes Fortes 1984a,
1984b inter alios). Tony Hunt’s stab at a solution (Hunt 1989) must be
taken as a thoroughly scholarly yet preliminary work, since he provides
no discussion on the whys and wherefores that motivated his decision
to equate Middle and Contemporary English plant-names. Carole
Biggam’s initiative, the ASPNS (Anglo-Saxon Plant Name Survey),15
and the articles by her collaborators and herself (for example Biggam
1994, 2003) do hit the mark fully, but they deal by definition with Old
English names only.
Historically, moreover, there has been a perceptible scholarly habit
towards passing the buck and—if one is allowed to continue with
clichéd idioms—leaving the proverbially drowsy dogs safely tucked
inside their kennels when it comes to equating ME plant-names and
the modern binomial nomenclature. While this is perhaps a bit of a
foregone conclusion (see McCarren 1998 for similar caveats but with
a more general scope), the examples analyzed above, which are in no
way unique to LH or any other ME herbal that I know of, teach us that
when it comes to plant-identification we simply should never trust the
information fed to us by dictionaries.
It would, however, be very unfair to put the blame on dictionaries and glossaries. We are dealing here with a lexicological, not a
15 The project seems sadly
From Clerks to Corpora
lexicographical, problem. Therefore it is neither Tony Hunt nor the
editors of the MED, but their colleagues working on manuscripts, who
are ultimately responsible for the current situation. In the particular
case of LH, for instance, MED was misled by the work of WhytlawGray, which was poor by any scholarly standard. Reliable editions
must be done of the many treatises, big and small alike, that still await
publication on the shelves of libraries scattered over the world, while
critical revisitations of those treatises already edited and discussed will
be much appreciated. Text glossaries remain an absolute need in any
edition indeed, but no more than ample textual and linguistic notes
that should accompany them, discussing at length why the editor
deems that this or that particular species is the one referred to in that
particular text.
André, J. (1956). Lexique de termes de botanique en latin. Paris: Klincksieck.
———. (1958). Notes de lexicographie botanique grecque. Paris: Honoré
———. (1985). Les noms de plants dans la Rome antique. Paris: Les Belles
Bauhin, C. (ed.) (1598). Petri Andreæ Matthioli medici Cæsarei et Ferdinandi
Archiducis Austriæ, Opera quæ extant omnia. Francofurti: Officina Nicolai
Bernard, E. (1697). Catalogi librorum manuscriptorum Angliæ et Hiberniæ in
unum collecti, cum indice alphabetico. Oxoniæ: Theatro Sheldoniano.
Biggam, C.P. (1994). Hæwenhnydele: an Anglo-Saxon medicinal plant.
Botanical Journal of Scotland, 46:4, 617–22.
Biggam, C.P. (ed.) (2003). From earth to art: the many aspects of the plantworld in Anglo-Saxon England. Amsterdam-New York: Rodopi.
Bosworth, J., & Toller, T.N. (1898). An Anglo-Saxon dictionary based on
the manuscript collection of the late Joseph Bosworth. Oxford: Oxford
University Press.
Brodin, G. (ed.) (1950). Agnus Castus. A Middle English herbal, reconstructed from various manuscripts. Upsala-Copenhagen-Cambridge (Mass.):
A.-B. Lundequistska Bokhandeln-Ejnar Munksgaard-Harvard University
Cockayne, O. (ed.) (1864–1866). Leechdoms, wortcunning, and starcraft of
early England. London: Longman.
Is Plant Species Identification Possible in Middle English Herbals?
Chantraine, P. (1999). Dictionnaire étymologique de la langue grecque.
Histoire des mots. (2nd ed.) Paris: Klinksieck.
Felter, H.W., & Lloyd, J.U. (1905). King’s American dispensatory. (19th ed.)
Cincinnati: The Ohio Valley Company.
Font Quer, P. (1987). Plantas medicinales. El Dioscórides renovado. (10th ed.)
Barcelona: Labor.
Fortes Fortes, J. (1984a). Fitonimia griega I. La identificación de las plantas
por los fitónimos griegos. Faventia, 6:1, 7–29.
———. (1984b). Fitonimia griega II. Las fuentes del vocabulario fitonímico
griego. Faventia, 6:2, 7–15.
García Valdés, M. (ed.) (1998). Plantas y remedios medicinales (De materia
medica). Madrid: Gredos.
Halliwell-Phillipps, J.O. (1889). A dictionary of archaic and provincial words,
obsolete phrases, proverbs and ancient customs from the fourteenth century.
(11th ed.) London: Reeves and Turner.
Hunt, T. (1989). Plant names of medieval England. Cambridge: D. S. Brewer.
Kurath, H., Kuhn, S.M., Reidy, J., & Lewis, R.E. (eds) (1954–2001). Middle
English Dictionary. Ann Arbor: University of Michigan Press.
Martin, W.K. (1965). The concise British flora in colour. London: Ebury Press
and Michael Joseph.
Mattioli, P.A. (1558). Commentarii secundo aucti, in libros sex Pedacii
Dioscoridis Anazarbei De medica materia. Venetiis: Ex Officina Erasmiana,
Vincentij Valgrisij.
McCarren, V.P. (1998). Editing glossographical texts: to marrow and to marrow and to marrow. V. P. McCarren & D. Moffat (eds) A guide to editing
Middle English. Ann Arbor: University of Michigan Press, 141–55.
Moreno Olalla, D. (2002). Lelamour’s Herbal (MS Sloane 5, ff. 13–57). Critical
edition and philological study. (Ph.D. University of Málaga PhD dissertation), Universidad de Málaga, Málaga.
———. (2007). The fautys to amende. On the interpretation of the explicit of
Sloane 5, ff. 13–57, and related matters. English Studies, 88:2, 119–142.
———. (2013a). A plea for ME botanical synonyma. V. Gillespie & A. Hudson
(eds) Probable truth. Editing medieval texts from Britain in the twenty-first
century. Brussels: Brepols, 387–404.
———. (2013b). The textual transmission of the Northern Macer tradition.
English Studies, 94:8, 931–957.
Mowat, J.L.G. (1887). Alphita. A medico-botanical glossary from the
From Clerks to Corpora
Bodleian Manuscript, Selden B.35. London-Oxford: Henry Frowde-Oxford
University Press.
Prior, R.C.A. (1863). On the popular names of British plants, being an explanation of the origin and meaning of the name of our indigenous and most
commonly cultivated species. London: Williams and Norgate.
Simpson, J.A., & Weiner, E.S.C. (eds) (1989). The Oxford English dictionary.
(2nd ed.) Oxford: Oxford University Press.
Turner, W. (1551). A new herball, wherin are conteyned the names of herbes in
Greke, Latin, Englysh, Duch Frenche, and in the potecaries and herbaries
Latin, with the properties degrees and naturall places of the same. London:
Steven Mierdman.
Whytlaw-Gray, A. (1938). John Lelamour’s translation of Macer’s Herbal.
(M.A. University of Leeds MA dissertation), University of Leeds, Leeds.
5 The Periphrastic Subjunctive in the Old
English Multiple Glosses to the Lindisfarne
Marcelle Cole
Utrecht University
The recessive nature of the subjunctive as a formal category in Old
English is witnessed in the use of alternative grammatical structures
other than inflectional subjunctives in contexts of non-fact modality.
An increasing analytic reliance on grammatical devices signalling nonfact modality (e.g. gif ‘if’, sua hua ‘whoever’, etc.) both fostered and
facilitated the occurrence of the indicative in such contexts. The modal
verbs magan, *sculan and willan served as fully independent verbs in
Old English, but even during the Old English period there appears to
have been a ‘modern’ tendency to use modal constructions involving a
(subjunctive) modal + infinitive construction, instead of an inflectional
subjunctive, with little (or no) underlying sense of non-modal notional
In the Old Northumbrian (ONbr) interlinear gloss to the Latin text
of the Lindisfarne Gospels (British Library, MS Cotton Nero D.iv;
henceforth Li), the increasing lack of direct correspondence between
contexts of non-fact modality and the subjunctive in Old English is
attested in the widespread tendency for present-indicative forms in -s
and -ð to supplant subjunctive forms, e.g. 7 swiðe bebead him þætte hia
ne æwades ł mersades hine L et uehementer comminabatur eis ne manifestarent illum ‘And he very much commanded them that they should
not make him known’ MkGl (Li) 3.12, or for indicative forms to alternate with subjunctive forms, which in ONbr ended in -a/-e/-o in both
1 For detailed discussion of the semantics of the Old English modal verbs see Standop
(1957: 18–66, 94–132, 133–155), Visser (1963–1973, iii §1483, §1562, §1653),
Mitchell (1985, §§1012–1015, §§1019–1020, §§1021–1022), and the dictionaries.
How to cite this book chapter:
Cole, M. 2015. The Periphrastic Subjunctive in the Old English Multiple Glosses to
the Lindisfarne Gospels. In: Shaw, P., Erman, B., Melchers, G. and Sundkvist, P. (eds)
From Clerks to Corpora: essays on the English language yesterday and today. Pp.
71–85. Stockholm: Stockholm University Press. DOI:
License: CC-BY.
From Clerks to Corpora
the singular and plural, e.g. gif gie habbas ł hæbbe leafo L si habueritis
fidem ‘If you have faith’ MtGl (Li) 21.21 and þætte gie eta 7 drincga
[...] 7 gie sittað ofer heh sedlo L ut edatis et bibatis [...] et sedeatis super
thronos ‘That you may eat and drink [...] and sit on high thrones’ LkGl
(Li) 22.30 (see Cole, 2014 for extended discussion).2 Periphrastic modal
verb + infinitive constructions also occur in Li to translate the Latin
subjunctive mood, as in 7 sohton ða hehsacerdas 7 ða uðuuto huu hine
mið facne gehealdon ł mæhton hia gehalda 7 ofslogon ł hia mæhton
ofslaa L et quaerebant summi sacerdotes et scribae quomodo eum dolo
tenerent et occiderent ‘and the chief priests and the scribes sought how
they might with wile lay hold of him and they might kill him’ MkGl (Li)
14.01.3 The glossator’s reliance on structures other than the inflectional
subjunctive is possibly all the more surprising given the requirements of
the glossing process to render the Latin as atomistically and faithfully
as possible in the target language.
The present paper examines the glossator’s use of the modal + infinitive construction in Li in relation to that of the inflectional subjunctive and indicative simple verb forms. The strategy of using modal +
infinitive structures in the translation of Latin subjunctives in Li has
long been noted (Bosworth & Toller 1898, magan), yet data drawn
from the ONbr glosses are not included in any of the major studies
on Old English modal verbs (cf. Gorrell 1895; Standop 1957; Ogawa
1989). Modal + infinitive constructions in Li occur frequently (though
not exclusively) in multiple glosses whereby a single Latin lemma is
rendered using at least two Old English glosses, separated by Latin vel
‘or’ (abbreviated to ł in the manuscript). Multiple glosses conveniently
facilitate the comparison of the forms that occur in identical contexts of
non-fact modality in ONbr. Certain properties of the multiple glosses
2 The abbreviations used in this paper to refer to the Lindisfarne Gospels (Li) and
the West Saxon Gospels (WSCp) are those employed by the Dictionary of Old
English Web Corpus (DOEC) and identify gospel, chapter and verse. The DOEC
relies on Skeat’s (1871–1887) edition of the Gospels. Citations are taken from the
DOEC, checked against the online facsimile of Li available at <
manuscripts/FullDisplay.aspx?ref=Cotton_MS_Nero_D_IV>. Biblical translations
translate the Old English text as opposed to the Latin and are my own. Multiple
glosses in the Old English text are provided with one Modern English translation.
3 Northern variants of uton + infinitive also occur in Li to translate the Latin hortative subjunctive, e.g. gæ we ł wutum geonga L eamus ‘Let’s go!’ but this usage will
not be dealt with in this paper.
The Periphrastic Subjunctive in the Old English Multiple Glosses
may also function as a diagnostic for evaluating the status of modal +
infinitive constructions in Old English.
The periphrastic subjunctive in Old English
The extent to which modal + infinitive constructions functioned interchangeably with inflectional forms as periphrastic expressions of mood
is a vexed question that has received a great deal of attention in the literature (Gorrell 1895; Standop 1957; Krzyszpién 1980; Mitchell 1985;
Goossens 1987; Ogawa 1989). The need for caution in too readily interpreting modal verbs as grammatical circumlocutions for the inflectional
subjunctive has been reiterated in the literature: modal verbs often present little loss of primary meaning and occur in the subjunctive under
the same circumstances that trigger its use with other verbs (Mitchell
1985, §§2971–2980). The terms ‘modal’ and ‘modal auxiliary’ are in
themselves problematic, given that these verbs functioned as independent lexical items in Old English and the issue of whether magan, *sculan and willan expressed mood at this early stage is controversial. I
follow Mitchell (1985, §991) in using the label ‘modal auxiliary’ for
want of a better term, but in full awareness of the potential prolepsis
involved in employing the terms ‘modal’ and ‘auxiliary’ with regard to
the function of these verbs in Old English.
From a historical point of view, it has been argued that in the initial
(Old English) stage of the history of the subjunctive and the ‘modal auxiliaries’ in English, modal verbs with an infinitive were treated as grammatical equivalents to inflectional subjunctive forms (Gorrell 1895).
The growing tendency in the language to use auxiliary constructions
was triggered and propagated by the breakdown in the formal distinction between the indicative and subjunctive inflectional forms of verbs.
Gorrell’s examination of the frequency of Old English ‘modal auxiliaries’ in relation to that of inflectional subjunctives in indirect discourse
indicates a striking increase in the use of the periphrastic construction
with a distinct tendency to occur when the corresponding inflectional
forms of verbs would prove ambiguous (1895: 458).
Observations in the literature that the use of a periphrastic subjunctive
was fostered by the breakdown in the formal distinction between the
indicative and subjunctive mood is particularly pertinent to the ONbr
gloss. One of the main characteristics of the ONbr texts is the advanced
state of morphological simplification across the verbal system caused by
various processes of reduction and levelling, including the proliferation
From Clerks to Corpora
of the northern present-tense marker -s at the expense of -ð, and the early
loss of final -n, most notably in the infinitive and present-plural subjunctive, but also in the preterite-present plural verbs and preterite indicative
and subjunctive (see Cole 2014). The preterite subjunctive and indicative
are more often than not indistinguishable from each other as the preterite plural subjunctive shows preterite-indicative -on endings rather than
-en and indicative forms occur with subjunctive -en, which suggests the
coalescence of the endings in [-ən].
The view of the history of the modal auxiliaries as a simple chronological development whereby modal + infinitive constructions increasingly
occur as a subjunctive equivalent as the Old English period progresses
is belied, however, by the textual variation that the incidence of modal
verbs exhibits in the Old English period. Ogawa’s (1989) survey of Old
English texts indicates a varying incidence in the use of the periphrastic
construction across different text types and time periods that challenges
what he terms the ‘substitution theory’, the notion of the history of the
English subjunctive as characterised by the use of modal auxiliaries to
compensate for the loss of inflectional morphology. His analysis of the
meanings of the modal verbs in subjunctive contexts does not show the
degree of loss of semantic meaning which would be expected if the verbs
were being used as auxiliaries. Instead modal verbs are employed to
convey a particular shade of meaning as required by the context: their
use adds a nuance that is not explicit in the inflectional subjunctive.
The suggestion that the periphrastic construction is not a mere grammatical alternative to the inflectional subjunctive but expresses instead
a semantic nuance not found in the simple subjunctive is in line with
Standop (1957: 169) and Krzyszpién (1980). The latter argues that periphrastic magan + infinitive and the inflectional subjunctive were not
wholly semantically interchangeable. The use of one form or the other
was determined by a difference in meaning: inflectional subjunctives
denoted general non-fact modality whereas the periphrastic magan +
infinitive expressed a particular aspect of non-fact modality, objective
possibility. When magan itself occurred in the subjunctive inflectional
form “general non-fact modality was ‘superimposed’ on the narrower
meaning of objective possibility” (Krzyszpién, 1980: 51).
Ogawa’s findings, in particular, emphasise the relevance of text type
and stylistic factors in determining the occurrence of the modal verbs;
there is a preference for poetry as opposed to prose to favour the use
of modal verbs across the entire Old English period, probably as a stylistic poetic device that added emphatic detail (Ogawa 1989: 231–232;
The Periphrastic Subjunctive in the Old English Multiple Glosses
Gorrell 1895: 458). In prose, modal verbs are favoured in argumentative religious and philosophical writings and homiletic literature rather
than in narrative prose. Text type and subject matter are also found
to affect the individual incidence of each modal verb, e.g. magan and
*sculan are common in didactic and religious writings where the latter commonly serves to emphasise commands, whereas narrative prose
works show a preference for willan (Ogawa 1989: 235). Interestingly,
the West Saxon Gospels (Corpus Christi College Cambridge MS, 140;
henceforth WSCp) exhibit an “almost entire neglect of the periphrastic
[modal] forms” (Gorrell 1895: 458). The low incidence of modal verbs
in WSCp is attributed by both authors to the translator holding “slavishly” to the Latin original (Gorrell 1895: 458; see also Ogawa 1989:
225, 236). Verse translations in the Paris Psalter and the Kentish Psalm
reflect a similar reluctance to employ periphrastic forms (Ogawa 1989:
236), which tells in favour of a close degree of dependence on the Latin
source inhibiting the use of periphrastic [modal] forms. The avoidance
of periphrastic forms involving modal verbs found in WSCp does not,
however, find a parallel in the interlinear ONbr glosses in Li; despite
the glossarial nature of the text type, the glossator’s language is not as
subjugated to the demands of atomistic glossing as might be expected.4
Modal + infinitive and the subjunctive mood in Li
Li is consistent with general Old English usage in its employment of
the modal + infinitive compound instead of an inflectional subjunctive form of the verb in indirect discourse after verbs of thinking and
believing and in the employment of willan with expressions of promise (Gorrell 1895: 449–455), exemplified in (1a) and (1b), respectively.
Such usage overlaps broadly with the future-in-the-past employment of
4 It should be borne in mind, nevertheless, that neither the Old English translation in
WSCp nor Li is consistent in its attitude towards the Latin original. The effect of
Latin influence is at times blatantly obvious in Li; for instance, the Latin negative
imperative construction nolite (plural) / noli (singular) + infinitive is categorically
translated using a contracted negative form of the verb willan followed by an infinitive in an attempt, no doubt, to render the Latin construction as atomistically as
possible, e.g. nallaðgie g[e]wyrce L nolite facere ‘make not!’ JnGl (Li) 2.16 and
nælle gie gedoema L nolite iudicare ‘judge not!’ JnGl (Li) 7.24. Contrastively, the
continuous prose translation in WSCp diverges from the Latin in its rendering of
negative imperatives and follows a more native OE ne + V + Spro structure compared with the literal counterparts found in Li, as in ne wyrce ge Jn (WSCp) 2.16 or
ne deme ge Jn (WSCp) 7.24.
From Clerks to Corpora
the preterite subjunctive (Mitchell 1985: §646) found in (1c). It should
be noted that (1d) is the only instance in which *sculan is used periphrastically to translate a Latin subjunctive in Li (Kotake 2006: 44).
Ogawa’s (1989: 235) observation that *sculan commonly serves to
emphasise commands would seem to find exemplification here.5
(1)a. wiste forðon huoelce uoere seðe salde hine ł ualde hine sealla
‘for he knew which one it was who would betray him’
sciebat enim quisnam esset qui traderet eum
JnGl (Li) 13.11
ðona ł forðon mið að gehatend wæs hir þæt sealla walde suæ
huæt wælde giwiga ł giuiade from him
‘therefore with an oath it was promised to her that he would
give whatsoever she would ask of him’
unde cum iuramento pollicitus est ei dare quodcumque postulasset ab eo
MtGl (Li) 14.7
7 forhuon ne saldes ðu feh meh to wege ł to disc 7 ic miððy
cuome mið agnettum ł uutedlice ic giude ł walde giuge þæt
‘and why did you not give my money to the bank, and when I
came, with usury indeed I would have exacted it?’
et quare non dedisti pecuniam meam ad mensam et ego ueniens
cum usuris utique exigissem illud
LkGl (Li) 19.23
7 huu auritten is on sunu monnes þætte feolo geðolas ł scile
ðoliga 7 gehened ł geniðrad ł geteled
‘and how it is written about the Son of man that he must suffer
many things and be despised’
5*sculan is more widely used in Li to translate the Latin future tense where it approximates a modern periphrastic future on occasions, e.g. from hernise gie geheras 7
ne oncnæugie ł ne cuðon ge 7 gesegende ge sciolon gesea ł ge geseas 7 ne geseað
ł ne sciolon gesea L auditu audietis et non intelligitis et uidentes uidebitis et non
uidebitis ‘By hearing you shall hear, and shall not understand: and seeing you shall
see, and shall not perceive’ MtGl (Li) 13.14; in caelo geong sua huæt ðu hæbbe
bebyg 7 sel ðorfendum 7 hæfis ł ðu scealt habba gestrion in heofne L uade cumque
habes uende et da pauperibus et habebis thesaurum ‘Go, sell whatsoever you have
and give to the poor and you shall have treasure in heaven’ MkGl (Li) 10.21; 7 gie
geseað ł scilon gesea sunu monnes to suiðrom sittende ðæs mæhtes L et uidebitis
filium hominis a dextris sedentem uirtutis ‘And you shall see the Son of man sitting
to the right of the power of God’ MkGl (Li) 14.62
The Periphrastic Subjunctive in the Old English Multiple Glosses
et quomodo scribtum est in filium hominis ut multa patiatur et
MkGl (Li) 9.12
The construction magan + infinitive is used in Li in both the present and preterite in clauses of purpose, a usage that is not peculiar to
Li, but which differentiates it from WSCp where magan is never used
in purpose clauses: there is a preference for willan instead (Ogawa
1989: 236). In the case of (2a) and (2b), the inflectional subjunctive
gesii alternates with (subjunctive) magan + infinitive. Behre (1934:
92, fn. 1) states that the subjunctive form of magan + infinitive is
the only way of unambiguously expressing subjunctive mood in the
first-person singular given the lack of formal distinction between the
first-person present-indicative and subjunctive. In ONbr gesii is a subjunctive form (Ross 1937: 133), which effectively eliminates morphological ambiguity as a motive for including mæge + infinitive alongside gesii in the double gloss. Standop (1957: 60–61) suggests that
magan expresses a different kind of uncertainty from the subjunctive of a simple verb. Similarly, as previously mentioned, Krzyszpién
(1980: 51) argues that subjunctive magan + infinitive is not a mere
circumlocution for the subjunctive but contributes semantic precision
by expressing a particular aspect of non-fact modality, that of objective possibility. The inclusion in Li of both an inflectional subjunctive
and mæge + infinitive would seem to corroborate this view. Behre’s
hypothesis may hold nonetheless in preterite contexts where the simple verb form is formally indicative but indistinguishable from the
subjunctive in speech (example 2c).
huæd wilt ðu ðe þæt ic gedoe se blinde uutedlice cuoeð him
laruu god þætte ic gesii ł þæt ic mæge sea
‘what (do) you want that I do to you? The blind man indeed
said to him: master, that I may see’
quid uis tibi faciam caecus autem dixit ei rabboni ut uideam
MkGl (Li) 10.51
cuoeð huæd ðe wilt ðu þæt ic doam ł gedoe soð he cuoeð la
drihten þætte ic gesii ł gesea mæge
‘he said: what (do) you want that I do to you? Indeed he said:
Lord, that I may see’
dicens quid tibi uis faciam at ille dixit domine ut uideam
LkGl (Li) 18.41
From Clerks to Corpora
monigo forðon he gehælde ðus þætte hia raesdon on him þætte
hine hie gehrindon ł hrina mæhtæs
‘for he healed many, thus that they pressed upon him that they
might touch him’
multos enim sanabat ita ut inruerent in eum ut illum tangerent
MkGl (Li) 3.10
Clauses of indefiniteness involving sede and sua hua, sua huelc ‘whoever’, suæhuælc ‘whatever’, etc. generally required a subjunctive verb
form in Old English, although the indicative became increasingly common as the period progressed (Visser 1963–1973, i. §886). Indicative
and inflectional subjunctive forms both occur in indefinite contexts in Li,
e.g. in suahuelcum hus gie inngae L et in quamcumque domum intraueritis ‘in whatever house you enter’ LkGl (Li) 9.4; on sua huelcne hus gie
ingæeð L in quamcumque domum intraueritis ‘in whatever house you
enter’ LkGl (Li) 10.5. There is also one instance of a double gloss consisting of both an inflectional subjunctive and an indicative, e.g. sua huelc
iuer hæbbe ł hæfeð friond L quis uestrum habebit amicum ‘whoever of
you has a friend’ LkGl (Li) 11.5. The Li glossator also avails himself of a
further strategy in double glosses and renders the Latin subjunctive using
a (subjunctive) modal + infinitive construction. In double glosses present
indicative ł welle + infinitive is the particular combination of grammatical forms that is employed in Li to translate the Latin subjunctive mood
in clauses of indefiniteness. This strategy is illustrated in (3):6
seðe soðlice ðerhwunes ł ðerhwunia węlla wið ł in ende ðes hal
‘he that truly perseveres until the end, he will be saved’
qui autem perseuerauerit in finem hic saluus erit
MtGl (Li) 10.22
7 sua hua dringe selles ł sealla węlle anum of lytlum ðassum
cælc ł scenc wætres caldes [...] ne loseð meard his
‘and whosoever gives drink to one of these little ones a cup of
cold water [...] he loses his reward’
quicumque potum dederit uni ex minimis istis calicem aquae
frigide [...] non perdet mercedem suam
MtGl (Li) 10.42
6 See 1b in the present paper for the same combination of grammatical forms used in
a multiple gloss in an indefinite clause, but in the preterite: suæ huæt wælde giwiga
ł giuiade. The occurrence of unambiguous indicative forms in the present tense,
illustrated in (3), suggests that giuiade, though formally indistinguishable from the
subjunctive preterite, was intended as an indicative form.
The Periphrastic Subjunctive in the Old English Multiple Glosses
7 sua hua cueðes ł cueða wele word wið sunu monnes forgefen
bið him
‘and whosoever speaks a word against the Son of man, it shall
be forgiven him’
et quicumque dixerit uerbum contra filium hominis remittetur
MtGl (Li) 12.32
7 seðe suæ huælc wælla suerige ł seðe suerias on wigbed noht
is seðe sua huelc uutedlice wælla sueria in gefo þæt is ofer ðæt
is rehtlic
‘and whosoever swears on the altar, it is nothing; indeed whosoever swears by the gift that is on it is a debtor’
et quicumque iurauerit in altari nihil est quicumque autem
iurauerit in dono quod est super illud debet
MtGl (Li) 23.18
7 cuoeð to him sua hwælc forletas ł forleta welle wif his 7 oðer
laede derneleger efnesende ł geendade ofer hia ł bi hir
‘and he said to them: whosoever puts away his wife and marries another, commits adultery against her’
et dicit illis quicumque dimiserit uxorem suam et aliam duxerit
adulterium committit super eam
MkGl (Li) 10.11
f. seðe welle losige ł loses ł fordoes ł forfæras sawel his fore meh
onfindes hia ł ða ilco
‘he that loses his life for me, shall find it’
qui perdiderit animam suam propter me inueniet eam
Mt (Li) 10.39
In double glosses, therefore, there is a preference for the modal periphrastic construction, as opposed to an inflectional subjunctive, to
occur alongside a present-indicative form in clauses of indefiniteness.
Clearly, both forms were acceptable grammatical alternatives in this
context, although it should be noted that the subjunctive modal form
welle is used and willan here retains an implied element of volition,
intention or acquiescence. Given that the indicative also occurs in this
context, the use of welle + infinitive may be an attempt to vivify the
sense of modality inherent in clauses of indefiniteness as to the person,
place, time referred to in the clause. The modal’s primary meaning has
not been eliminated, but such usage provides an insight into an intermediate stage in the development of the English periphrastic subjunctive.
From Clerks to Corpora
The apodosis of a hypothetical condition in ModE requires a preterite modal: in fact its occurrence in this environment serves as a test of
modal status (Denison 1993: 313). In Old English the preterite inflectional subjunctive generally occurred in this context, e rally ‘if God
were your father indeed you loved me’ Jn (WSCp) 8.47, where lufedon,
though formally preterite indicative, is to be understood as a subjunctive. In Li, however, preterite forms of willan + infinitive occur in the
apodosis of a hypothetical proposition, compare, gif god faeder iuer
woere gie ualde lufiga uutedlice mec ‘if God were your Father, you
would indeed love me’ L si deus pater uester esset diligeretis utique me
Jn (Li) 8.42. Visser (1963–1973, §1532, §1607, §1672) records such
usage in Old English with should and might, but cites no examples
of would for Old English, apparently dating the emergence of would
in the apodosis of a hypothetical proposition to the early ME period.
Instances of wolde in the apodosis of a conditional would appear to
exist in Old English, however: Ogawa (1989: 131) identifies an example of wolde used with an inanimate subject in Ælfric’s Lives of Saints
(ÆLS 31.672) that “may well serve as a pure expression of an imaginary event in the past”. The examples in (4) illustrate the occurrence of
walde in the apodosis of a hypothetical condition in Li.
gif nere ðes yfeldoend ne ðe ue gesaldon ł nalde ue gesealla
hine ðe
‘if he were not this evildoer, we would not have delivered him
up to you’
‘si non esset hic malefactor non tibi tradidissemus eum
JnGl (Li) 18.30
gif ðu gegiuuedes from ðæm 7 gesalde ðe ł æc ualde gesealla
ðe uæter cuic ł lifwelle uæter
‘if you had asked of him, and he would have given you living
petisses ab eo et dedisset tibi aquam uiuam
JnGl (Li) 4.10
Here willan serves to express an intermediate idea between simple
volition and an imaginary past result: it cannot be said that walde is
entirely non-volitional in nature, but it does approximate ‘modern’
usage in conveying the probable past results of an unreal past condition. The examples also succinctly exemplify the lack of formal opposition between the indicative and the subjunctive in the preterite that
may well have contributed to the glossator’s decision to signal non-fact
modality more explicitly using a periphrastic form.
The Periphrastic Subjunctive in the Old English Multiple Glosses
As previously noted, periphrastic subjunctives occur frequently in
double glosses in Li, which, given the preference for multiple glosses in
Li is not in itself significant.7 Nonetheless, the double-glossing translation technique provides a unique insight into what type of grammatical
units co-occurred in identical contexts. Multiple glosses in Li generally
involve two items, though triple and even quadruple glosses also occur
(see examples 1d and 3f in the present paper, in which the triple glosses
gehened ł geniðrad ł geteled and loses ł fordoes ł forfæras translate
Latin contempnatur ‘scorn, despise’ [3sg pres.subj.pass.] and perdiderit
‘lose, destroy’ [3sg perf.subj.act.], respectively). The multiple glosses
vary with regard to the type of information that they supply; some
provide lexical alternatives for a single Latin lemma, involving the use
of synonyms or near-synonyms, e.g. berað ł bringeð L adferte ‘bring!’
JnGl (Li) 21.10. Other double glosses provide grammatical alternatives
for a single Latin lemma, e.g. geseað ł geseas L uideritis ‘you see’ LkGl
(Li) 21.20, ne habbas ł nabbas L non habent ‘they do not have’ MtGl
(Li) 14.16, or supply both a grammatical and a lexical alternative, e.g.
gæ we ł wutu[m] geonga L eamus ‘Let’s go!’ MkGl (Li) 1.38.
In double glosses translating the subjunctive mood the preferred strategy in Li is to juxtapose an indicative with a (subjunctive) modal + infinitive, rather than an inflectional subjunctive. It would be an oversimplification to infer that the occurrence of the periphrastic construction
alongside simple forms in double glosses is in itself proof that the modal +
infinitive structure is a mechanical substitute for the inflectional subjunctive. Double glosses in Li clearly provide alternatives that are equally
acceptable in a given context but they convey differing nuances. Double
lexical glosses generally involve an item that introduces a semantic
nuance, such as berað ł bringeð L adferte ‘bring!’ JnGl (Li) 21.10, cited
above, or a contextualised nuance. Pons-Sanz (forthc.) cites the rendering of L puella as dohter ł mægden at MkGl (Li) 5.41. The Latin lemma
puella would normally be translated using OE mægden ‘girl, maiden’,
but here, Pons-Sanz argues, OE dohtor ‘daughter’ is included because
the girl referred to is the daughter of the leader of the synagogue mentioned a few lines previously. Even alternative grammatical glosses, such
as geseað ł geseas L uideritis LkGl (Li) 21.20 and ne habbas ł nabbas L
non habent MtGl (Li) 14.16, which appear to present simply morphological variants, provide an additional form that is more idiomatic or
colloquial, such as the dialectal ONbr second-person plural form geseas
7Kotake (2006: 37, fn. 4) gives the total occurrences of multiple glosses in Li at
3159, of which 466 are grammatical glosses.
From Clerks to Corpora
alongside geseað ‘you see’ at LkGl (Li) 21.20, or nabbas instead of ne
habbas at MtGl (Li) 14.16. Similarly, the double glosses under scrutiny
involving ‘modal auxiliary’ constructions are a complex case that do not
necessarily involve grammatically interchangeable forms with no difference in meaning. Careful analysis suggests that the periphrastic modal
subjunctive is a grammatically acceptable – if somewhat stylistically
different – alternative to the inflectional subjunctive.
With regard to the ordering of the items in double grammatical
glosses in Li, i.e. whether the glossator chooses to place a term in first
or second position, Kotake (2006) identifies a considerably consistent
pattern in the ordering of the alternative grammatical glosses, including those that translate Latin verbs forms conveying future tense and
subjunctive mood. In the majority of cases, the double glosses consist
of a simple form in first position followed by a periphrastic construction. He attributes this ordering preference to the glossator placing the
Old English grammatical category that is morphologically closest to
the Latin first, followed by the “more morphologically marked” periphrastic construction (2006: 44–46). I interpret the grammatical unit
that most deviates from the Latin original as the more idiomatic or colloquial item in the doublet. The translation technique of double glossing provides the glossator with the scope to include one Old English
translation that is atomistic and closely parallels the Latin morphologically and another that distances itself from the original Latin text and
provides a more concrete or colloquial rendering. Such an interpretation both corroborates and finds support in Ogawa’s contention that
the modal verb phrase probably reflects colloquial Old English usage.
Ogawa notes the tendency for modal verbs to occur more readily at the
colloquial rather than literary level of Old English prose and in direct
speech and suggests that “the colloquial level of style favours the use
of modal verbs in its endeavour to make description vivid and concrete” (1989: 237–238). The colloquial sphere of usage is also a locus
of grammatical change where the ‘modern’ periphrastic future and subjunctive forms would be expected to make their first appearance.
The data in Li provide further insight into an intermediate stage in the
history of the periphrastic subjunctive whereby modal verbs were used
initially to emphasise a particular aspect of non-fact modality that an
inflectional subjunctive could only do more generally and which indic-
The Periphrastic Subjunctive in the Old English Multiple Glosses
ative forms left unexpressed. The notion of the history of the English
subjunctive as characterised by the use of modal auxiliaries to compensate for the loss of inflectional morphology only tells part of the story
of the English periphrastic subjunctive: stylistic factors were of considerable importance. In a succinct summary of the Old English state of
affairs, Ogawa (1989. 223–234) highlights that the varying distribution
of modal verbs across text types
points to no clear tendency for them to form the ‘periphrastic subjunctive’ when the corresponding simple verb form would be ambiguous with respect to mood. […] Although the distinction between
the modal verb construction and the simple verb form is not always
easy to explain, the former can be generally shown to stand for
clearer, more concrete expressions, emphasizing and specifying, by
the appropriate choice of a relevant modal verb, a particular nuance
of desired relationship as the context requires it
The status of the modal verb construction as a clearer, more concrete
expression that emphasises and specifies is also in line with the double
glossing technique found in Li. Double grammatical glosses generally
provide one translation that is morphologically closer to the Latin and
an additional gloss that is more idiomatic or colloquial, in this case, the
modal verb construction.
Nevertheless, the type of grammatical structures that alternate in
double glosses alongside modal verb constructions also suggest that
the breakdown in formal opposition between the indicative and subjunctive played a role in the history of the subjunctive and the ‘modal
auxiliaries’ in English. Double glosses such as þætte ic gesii ł þæt ic
mæge sea L ut uideam MkGl (Li) 10.51 indicate that mæge sea is not a
mechanical grammatical circumlocution for the subjunctive form gesii.
But in cases such as ne ðe ue gesaldon ł nalde ue gesealla L tradidissemus JnGl (Li) 18.30, the lack of formal opposition between the indicative and the subjunctive in the preterite may well have fostered the use
of the periphrastic structure as a more distinctive means of signalling
non-fact modality; after all, walde + infinitive is the only way of unambiguously expressing non-fact modality given the coalescence of the
preterite indicative and subjunctive. It is also important to bear in mind
that even in the present where unambiguous subjunctive forms were
retained for longer, modal + infinitive constructions co-occur mainly
with indicative forms in double glosses rather than the inflectional
subjunctive. The lack of direct correspondence between contexts of
From Clerks to Corpora
non-fact modality and the subjunctive (inflectional or periphrastic) in
Old English and the increasing occurrence of present-indicative (or of
morphologically ambiguous preterite) forms would also have fostered
the use of periphrastic subjunctive forms. The glossator’s preferred
strategy in Li of juxtaposing an indicative with a (subjunctive) modal +
infinitive, rather than an inflectional subjunctive militates in favour of
Krzyszpién’s (1980: 62) view that one of the functions of modal verb
compounds was to serve as “a more distinctive and productive means
of signalling non-fact modality” [be it a particular aspect of non-fact
modality] left unexpressed by indicative forms.
Behre, F. (1934). The Subjunctive in Old English Poetry. (Göteborgs Högskolas
Årsskrift 40). Göteborg: Elanders.
Bosworth, J., & Toller, T.N. (1898). An Anglo-Saxon Dictionary Based on
the Manuscript Collection of the Late Joseph Bosworth. Oxford: Oxford
University Press.
Cole, M. (2014). Verbal Morphosyntax in Old Northumbrian and the
(Northern) Subject Rule. (NOWELE Supplement Series). Amsterdam &
Philadelphia: John Benjamins.
Denison, D. (1993). English Historical Syntax. London & New York: Longman.
[DOEC =] Dictionary of Old English Web Corpus. (2007). Antonette diPaolo
Healey et al. (eds) Toronto, ON: University of Toronto. <http://www.doe.>.
Goossens, L. (1987). The auxiliarization of the English modals: a functional
grammar view. In M. Harris & P. Ramat (eds) Historical Development of
Auxiliaries. Berlin & New York: Mouton de Gruyter, 111–143.
Gorrell, J.H. (1895). Indirect discourse in Anglo-Saxon. PMLAA, 10, 342–485.
Kotake, T. (2006). Aldred’s multiple glosses: is the order significant? In
M. Ogura (ed.), Textual and Contextual Studies in Medieval English:
Towards the Reunion of Linguistics and Philology. Bern: Peter Lang, 35–51.
Krzyszpién, J. (1980). The periphrastic subjunctive with magan in Old English.
Studia Anglica Posnanienia 11, 49–64.
Mitchell, B. (1985). Old English Syntax, 2 vols. Oxford: Clarendon.
Ogawa, H. (1989). Old English Modal Verbs: A Syntactical Study. (Anglistica,
26). Copenhagen: Rosenkilde & Bagger.
The Periphrastic Subjunctive in the Old English Multiple Glosses
Pons-Sanz, S. (Forthcoming). A study of Aldred’s multiple glosses to the
Lindisfarne Gospels. In J. Fernández-Cuesta & S.M. Pons-Sanz (eds) The
Old English Glosses to the Lindisfarne Gospels: Language, Author and
Context. Berlin: Mouton de Gruyter.
Ross, A.S.C. (1937). Studies in the accidence of the Lindisfarne Gospels. Leeds
School of English Language Texts and Monographs 2, Kendal.
Skeat, W. W. (ed.) (1871–1887). The Holy Gospels in Anglo-Saxon, Northumbrian,
and Old Mercian versions. Cambridge: Cambridge University Press.
Standop, E. (1957). Syntax und Semantik des modalen Hilfsverben im
Altenglischen: ‘magan’, ‘motan’, ‘sculan’, ‘willan’. (Beiträge zur Englischen
Philologie, 38). Pöppinghaus, Bochum-Langendreer.
Visser, F. T. (1963–1973). An Historical Syntax of the English Language, 3 parts,
4 vols. Leiden: E. J. Brill.
6 On the Place-Name Isle of Dogs
Laura Wright
University of Cambridge
1. Introduction
In a contribution to the study of the Ormulum (Johannesson 2012),
Professor Nils-Lennart Johannesson pinpointed, by means of pictorial
as well as written evidence, the social nuance inherent in Orm’s metaphor of hunting with nets and dogs to represent the disciples’ catching
of men’s souls for their Lord. Professor Johannesson showed how trapping deer in nets for use as food was the workaday task of servants, as
opposed to the lordly pursuit of hunting as a pastime by giving chase.
Orm’s net metaphor (spelless nett ‘net of preaching’), therefore, aligned
his rendition of the Gospels with the servant class: the disciples serve
their master. Professor Johannesson notes that Orm’s hunting dogs were
not present in his Latin source texts but were his own invention, and
presumably reflect twelfth-century Lincolnshire reality, where his audience would have expected dogs to accompany hunting (Johannesson
2012: 237–238). In what follows I continue with the theme of dogs, the
dogs in question being not literal but to do with word-play.
2. Explanations for the place-name Isle of Dogs
The place-name Isle of Dogs refers in Present-Day English to the land
within a meander of the River Thames in East London. In high Victorian
style, B.H. Cowper tells us that the Isle of Dogs is “embosomed, by
our noble river, which describes a magnificent curve in the form of a
horseshoe from Limehouse to Blackwall” (Cowper 1853: 1). The land
embosomed by this magnificent curve was not, historically, an island.
It was artificially made into an island when the West India Docks were
How to cite this book chapter:
Wright, L. 2015. On the Place-Name Isle of Dogs. In: Shaw, P., Erman, B., Melchers, G.
and Sundkvist, P. (eds) From Clerks to Corpora: essays on the English language yesterday and today. Pp. 87–116. Stockholm: Stockholm University Press. DOI: http://dx.doi.
org/10.16993/bab.f License: CC-BY.
From Clerks to Corpora
created in 1802, by means of the narrow channels of the West India
Dock Canal at Limehouse and Blackwall, but this was almost three
hundred years after the place-name Isle of Dogs came into use. Strype
(1720), Cunningham (1849: vol 2, 417) and Mills (2004: 121) list the
following explanations for the name Isle of Dogs:
Next is the Isle of Dogs; being a low Marshy Ground, so called, as
is reported, for that a Waterman carried a Man into this Marsh,
and there murthered him. The Man having a Dog with him, he
would not leave his Master; but Hunger forced him many times
to swim over the Thames to Greenwich; which the Waterman
who plied at the Bridge observing, followed the Dog over; and
by that means the murthered Man was discovered. Soon after
the Dog swimming over to Greenwich Bridge, where there was a
Waterman seated, at him the Dog snarled, and would not be beat
off, which the other Watermen perceiving, (and knowing of the
Murther) apprehended this strange Waterman; who confessed
the Fact, and was condemned and executed.
(Strype (ed.) 1720: Vol 1, Book 1, 43)
My objection to this explanation is that there is no supporting
evidence, and the date of 1720 is two hundred years after the
date of the first attestation of the name Isle of Dogs. Plus, only
one dog is mentioned.1
The fertile Soil of the Marsh here is much admired, usually
known by the Name of The Isle of Dogs: So called, because,
when our former Princes made Greenwich their Country Seat,
and used it for Hunting, (they say), the Kennels for their Dogs
were kept on this Marsh; which usually making a great Noise,
the Seamen and others thereupon called the Place The Isle of
Dogs: Though it is not an Isle, indeed, scarce a Peninsula, the
Neck being about a Mile in length.”
(Strype (ed.) 1720: Vol 2, Book 6, 102)
This explanation was given to John Strype by the Reverend
1 My grateful thanks to Prof Richard Coates, who asked me about the Isle of Dogs
in the first place – and who noticed the singularity of dogs in the first explanation,
which I had overlooked. I am also grateful to Prof Ian Donaldson, Peter Guillery,
Prof Derek Keene and Prof Nicholas Rodger for criticising and commenting on
earlier drafts, and to Steve Roberts for supplying the fruiterers’ number-system
discussed in Section 5.
On the Place-Name Isle of Dogs
Dr Josiah Woodward, Minister of the Chapel and Hospital of
Poplar. Again, there is no evidence, and the dating is two hundred years after the first attested usage.
That Isle of Dogs was originally Isle of Ducks, or possibly Isle
of Docks.
My objection here is that there are no written occasions
on which the area is referred to as either of these, so far as
I know. Voiced word-final stops do not usually become devoiced
in London English, and there is no obvious motivation for the
replacement of a semantically-transparent and contextually-­
relevant word (whether ducks or docks) with a less relevant one.
That Isle of Dogs was originally Isle of Doggers, from the
­fourteenth-century fishing vessels known as doggers.
The difficulty with this explanation is that there were no
ports, hithes or landing-stages on the land embosomed by the
curve of the river.2 There was nowhere for a fishing vessel to call
in, or for fishing vessels to congregate, unlike on the southern
side where there were hithes at Greenwich and Deptford. Also,
there is no evidence that the land was ever called Isle of Doggers
by anybody.
That there were either wild dogs there, or dead dogs washed up
by the tide there.
We have no way of knowing whether either of these were true,
or more true of this part of the Thames than any other.
That Isle of Dogs alludes to the Canary Islands, because Latin
Insulae Canariae means ‘island of dogs’. The historian Pliny says
that these islands were so-named because there were large dogs
This hypothesis lacks evidence or circumstance that would
make this likely. Canary Wharf appears not to have become
2I thank an anonymous reviewer for pointing out that the name Stepney, earlier
Stebbenhithe (and spelling variants) implies a hithe or landing-stage. However it
is not thought that the hithe in question was situated on the marsh, but at Ratcliff
Cross, in present-day Limehouse: “Evidence for Saxon settlement is etymological.
The first reference to Stepney is to men of the bishop of London’s estate (vill)
of Stybbanhythe c. 1000, recording a hithe or landing-place either on the Thames
or the Lea. Since the place-names Old Ford and Stratford are associated with the
Lea, while the name Stepney has always been linked with the southwest quarter of
the parish, the hithe was probably on the gravel at Ratcliff Cross, one of the few
sites below London Bridge suitable for landing before the marshes were embanked
and wharfed.” (Baker (ed.) 1998: 13–19).
From Clerks to Corpora
so-named until 1936, when the shipping company Fred Dessen &
Co., which unloaded fruit from the Mediterranean and Canary
Islands, was granted permission to rename the wharf known previously as West Wood Wharf. 3
I conclude that none of these explanations really holds water, with
no supporting evidence for any of them.
3. Early attestations of the place-name Isle of Dogs
The name Isle of Dogs is first attested in 1520. Before 1520, the place
we know as the Isle of Dogs was called Stepney Marsh, for which
Mills (2001: 121) has a first attestation date of 1365. Stepney Marsh
is relatively well documented: there exists a field survey of the marsh
from around 1400, and also wills of several landlords who bequeathed
land in Stepney Marsh to their heirs in the late 1300s and early 1400s.
There was a settlement in the marsh from at least the second half of the
twelfth century; the manor house at the southern end, somewhat inland
from the sea-wall, was owned by William de Pontefract, and his manor
house, chapel and the hamlet around became known as Pomfrets,4
which manor was ultimately owned by the Bishop of London (Cowper
1853: 16). However this place-name did not last. The manor house was
in ruins by the 1360s, and the hamlet was abandoned in 1448 when
the river burst through the sea-wall (Dugdale 1662: 72). It seems that
the manor of Pomfret became abandoned because of partible inheritance, as none of the descendents of William de Pontefract actually
lived at Pomfrets or maintained it or the sea-walls (Currie (ed.) 1998:
1–7, 52–63).
Further place-names of Old English etymology in the marsh are as
follows. At the northern perimeter of Stepney Marsh the long street
village of Poplar5 runs from foreshore to foreshore, with the lime kilns
3 ‘The West India Docks: The buildings: warehouses’, Survey of London: volumes 43
and 44: Poplar, Blackwall and Isle of Dogs (1994), pp. 284–300. URL: http:// india docks
buildings war. Date accessed: 06 December 2012.
4 ‘The Isle of Dogs: Introduction’, Survey of London: volumes 43 and 44: Poplar,
Blackwall and Isle of Dogs (1994), pp. 375–387. URL: http://www.british-history. Date accessed: 06 December 2012.
5AN popler ‘at the poplar tree’, Gover, Mawer, Stenton 1942: 133–134.
On the Place-Name Isle of Dogs
of Limehouse at the north-western point of the curve6 and Blackwall7
at the north-eastern point. Moving south around the curve, the western
embankment of Stepney Marsh was called Westwall;8 the north-eastern
embankment was called Blakewalle, present-day Blackwall (after which
the hamlet was named)9 and some, or perhaps all, of the embankment
round Stepney Marsh was called Themsewall and Longewall.10 Further
dwellings in the marsh were situated at Westwall and at Newebygynge.11
From various wills we learn that the marshland was called
Stebbenhith mersche; Stebbenhithmerssh;12 that fields in the marsh were
called Margarusagre atte gate, ‘Margaret atte Gate’s acre’,13 Potterisfeld,
‘Potter’s Field’ and Chafcroft, ‘Calves’ Croft’.14 We learn that there was
a bridge or jetty called ffisshbregge; and that the chapel at Pomfrets
was dedicated to St Mary and All Saints in the marsh: Beate Marie &
Omnibus Sanctis,15 Beate Marie de Capelle in le Merssh,16 Capelle Beate
Marie in Marischo.17 From the Stepney Field Survey of c.140018 we
learn that ffysshysbregge, ffysshysfeld and ffysshyslond were owned by
Thomas ffyssh; that Thomas Edewyne had three rods of land lying by
the wall called Thameswall; and that John Hamme had one rod lying
above le Throwedych.19 The names of hills, fields, walls, enclosures,
6 Gover, Mawer, Stenton 1942: 150; OE līm ‘lime’ + OE āst ‘oast, kiln’.
7 Gover, Mawer, Stenton 1942: 135; presumably OE blǣc ‘black’ + OE weall ‘wall,
rampart of earth or stone’.
8 Kew, TNA, SC 12/11/31 fo 12v.
9 Kew, TNA, SC 12/11/31 fo 16v.
Kew, TNA, SC 12/11/31 fos 17v and 19. Cowper (1853: 4), discussing process of
building these medieval embankments, writes “With might and main they toiled,
and by might and main they overcame”.
Kew, TNA, SC 12/11/31 fo 12. “wall vocata Newebygynge” – which might indicate
ongoing embanking work as bigging meant both ‘dwelling’ and ‘building’ (see OED
bigging, n.).
12 E.g. 1404; London Metropolitan Archives DL/C/B/004/MS09171/002, fo 54v; Will
of Roger Grummote.
13 1376, LMA DL/C/B/004/MS09171/002, fo 36v. Roman type indicates expansion of
abbreviations and place-names have been given initial capital letters.
14 1380, LMA DL/C/B/004/MS09171/002, fo 71; Will of Williamo pottere de maresco
de Stebunhith.
15 1380, LMA DL/C/B/004/MS09171/002, fo 71.
16 1402, LMA DL/C/B/004/MS09171/002, fo 26; Will of John Broun of Stebbenhith.
17 1405, LMA DL/C/B/004/MS09171/002, fo 51v.
18 Kew, TNA SC/12/11/31. I have tried to restrict this list to features in the marsh
alone, but it is possible that a few may have lain further north. See Croot (1997) for
a description and discussion of the Stepney Field Survey.
19 Kew, TNA SC/12/11/31, fos 17v, 18.
From Clerks to Corpora
ditches, weirs, hedges, bushes and meadows in the marsh recorded in
the Field Survey were:
I have counted 74 place-names in Stepney Marsh in the late 1300s
and 1400s, but no mention at all of the name Isle of Dogs. If the name
had been in use then, one might expect it to show up in these detailed
sources. But it does not.
Table 1. Names of hills, fields, walls, enclosures, ditches, weirs, hedges,
bushes and meadows in Stepney Marsh (Field Survey of Stepney, c.1400,
TNA SC/12/11/31).
fo 11 Sandhell, Battysfeld
fo 11v Westwall, Westyerde, Kalfstokkysfeld, Pylyslond
fo 12 Sonderesffeld, Sondereslond, Sonderesthrowe, Wall vocata
Newebygynge, Amystonerffeld, le Chekyr
fo 12v Cochysgate, Worlycheslond, Sandhell in South Newelond,
Codyneshawe, Kalstokkyslond, Kalstokkyswere, Chapell
Lond, Rodeberdeslonde, Coughdesyerde, Long Acre, Karles
Acre, Breweresyerde, Schypmade, Gattyswere, Goldyngeslond,
ffanneresmade (belonging to Henry Vannere), Shyftylcroft
fo 14 le ffanneresmade apud le Wylde
fo 15 Byllokysland, Northbroke, Elderbussh, Scottysacre,
Longemade, le Hyedoune, le Netherdoune iuxta le Hye doune,
Crepyneslond in le Hyedoune, Wowehegge
fo 17v Rancesfeld, Rysshcroft par le Rance, Themsewall, Deleswey,
Admondeslond in Estnessh
fo 18 Le Throwedych in Gabelond, Admondeswere, Madehawys,
Deleslond, Boleyffeld
fo 19 Nethyr Somerlase, Nokysfeld, Byllokyswere, Southawys,
Wereye, Madeye, ffleecroft, Longewalle iuxta Madeye
fo 19v Southmadeye, Schatffletebregge, Chynham
fo 20 Brademade, Sandhell in Brademade, le Longeforlond, Est
Bryȝt Onelond, Colmansyerde
fo 20v Wyseffletdych, Rowynglond, Thornhegge
fo 21 Chalfcroft, Grandylhope, Bradecroft in le Hooke
fo 21v Pareshawe apud Pomffret
fo 22 Buntynglond, Smythlond
On the Place-Name Isle of Dogs
We next turn to some information provided by Sir William Dugdale
in his History of Imbanking and drayning of divers Fenns and Marshes
of 1662:
Upon an inquisition taken in 27 H. 6. the Jurors presented, that by
the violence of the tides upon the banks of Stebenhithe marsh, a
great part of the said banks, adjoining to that marsh, was then ruinous and broken through the neglect of the Land-holders there: And
that through the default of one Iohn Harpour Gentleman, in not
repairing his bank, opposite to Deptford Strond, there was, on the
Monday, being the Feast of the Annunciation of the blessed Virgin,
in the 26 year of the reign of the said King H. 6. a breach made in the
said bank of the before-specified Iohn Harpour, for the length of xx
Rods, unto the land of Iohn Fyloll, in so much as a thousand Acres
of land, lying within the said marsh, were drowned.
(Dugdale 1662: 72)
It is significant that the 26 Henry 6 (1448) breach of the banks
occurred opposite Deptford Strand, because in 1520 the place-name
Isle of Dogs occurs, also referring to a specific area of Stepney Marsh
opposite Deptford Strand; specifically, opposite the Royal Dockyard at
Deptford. It is in The Boke of thaccoumptte of Costys Ande Charges
don & made on the Kinges Schyppes for the transportynge of the kinge &
the Qwen to Calyce to the metynge of the frensche kinge & from
thence in to Inglond Ageyn.20 This is the accounts-book of Henry VIII’s
Dockyard at Deptford, where warships were built and repaired, and
which had been in existence for seven years in 1520. The place-name
Isle of Dogs is mentioned as a berth:
“& to John holmes for a hose for the mary George lynge in at doke
at theille of dogges afor depford xd”
(October 1520; Kew, TNA MS E 36/11, fo 117v, calendared in
Letters and Papers of Henry VIII, volume 3, pp 369–381)
The location of theille of dogges is ‘afor depford’ (the word opposite
was not available, as it did not carry its present meaning at this date,
see Wright 2006). The same ship, the Mary George, was again berthed
in the same place five years later:
“vppon the Este side of the ile of doggis”(the leaf is damaged here so there
is no more context) … “Item the Mary george beinge of portage / ijC &
20 Kew, TNA E 36/11, fo 104; ‘Calyce’ is Calais.
From Clerks to Corpora
L tonne / lythe / vppon the sowthe syde / of the Ile of doggis / and
muste be Calkyd / wtine the borde & wtoughte / also she must be
seerchyd for wurmehoolys because she hath ben in leevaunte /”
(October 1526; London, British Library Cotton MS Otho E IX, fo
6821 calendared in Letters and Papers of Henry VIII, volume 4, pp
4. Early map evidence
The first map to show the place-name Isle of Dogs is Robert Adam’s
map Thamesis Descriptio, of 1588, which shows the lower reaches
of the River Thames below the Pool of London.22 Robert Adam was
Surveyor of Works to Queen Elizabeth I. His map was made as part
of the response to the Spanish Armada, the Secretary of State for War
having asked what arrangements had been made for the defence of the
kingdom. It shows the river as seen from on board a vessel, plus the
army’s camp further inland at West Tilbury. The map demonstrates,
by means of arcs of sweeping lines, the rakes of gun-fire possible from
the defences positioned on the riverbank. There are tiny horses and
riders depicted on the causeway from Tilbury Fort to the camp “beset
wth twentye & seven ensigns”, and rowbarges and small boats at the
barrage or boom between Gravesend and Tilbury Fort. Two batteries
are shown upriver on either side at Lee Ness on the south bank and
Saunders Ness opposite on the north (Saunders Ness lies on the eastern
side of Stepney Marsh). These are the first marked defences downriver
from London, and there is another boom across the river at this point.
The river seems well defended, but the effort was not, in the event,
necessary, as the Armada was blown off course and never attempted to
attack London.
The careful labelling of this map deserves a fuller treatment.23 There
are three types: labels written perpendicular to the river, labels written
horizontally to the river, and labels written in the river. The first group
contains natural features of the riverbank which enable that part of the
bank to be identified visually from onboard ship, with two non-natural
21See State Papers Online database:|MC4301001776&v=2.1&
London, British Library, Adds MS 44839.
I am particularly grateful to Prof Derek Keene for drawing my attention to the
significance of the orientation of the labelling.
On the Place-Name Isle of Dogs
Thamesis Descriptio Anno 1588, surveyed by Robert Adams, 1738. Crace
Collection, British Library. Reproduced with permission. (Note: this is
a different map to the one described, the same in all essentials but with
shipping omitted and slightly different spellings).
exceptions (Gallion and Tripcotts, the names of riverside buildings in
what was otherwise featureless marsh). Head-nouns are breache, creeke,
elmes, haven, ile, mouthe, nesse, pointe, tree. The second group, written
horizontally to the river, contains place-names. The third group, written
in the river, contains the names of stretches of the river that could be
viewed from bend to bend. Head-nouns are checke, hope, poole, reache.
Group One, written perpendicular to the river, grouped alphabetically according to head-noun, contains:
Ye Litle Breache
Ye greate Breache
Stackie Breache
From Clerks to Corpora
Barkinge Creeke
Daignâ Creeke
Dartfoorde Creeke
East Ham Creeke
Rainam Creeke
Rauensborne Creeke
Podds elmes
Saunders elmes
Ile of Dogges
Cuckolds haven
Brode Nesse
Crosse Nesse
Erithe Nesse
Gallion Nesse
Greenhith Nesse
Hooke Nesse
Lee Nesse
Magott Nesse
Northfleete Nesse
Saunders Nesse
Staffleete Nesse
Stone Nesse
Theeves Nesse
Tilberie Nesse
Gilian tree pointe
Middway tree
Saunders Elmes and Podds Elmes are marked by trees. Gallion and
Tripcotts are marked by small buildings. The breaches are marked by
lesser or greater pools, and the creeks are marked by tributaries entering the river. The features labelled with Group One names all act as
On the Place-Name Isle of Dogs
checkpoints in otherwise featureless marsh, either distinguishing one
reach of the river from the next, or helping to identify that particular
reach. They are navigational aids, especially necessary in darkness or
foggy weather.
Group Two, written horizontally to the river, and looking downstream from Westminster, contains place-names:
Westminster, Lambeth, London, Southwarke, Raderife, Ratcliffe,
Limehouse, Depthfoorde, Greenewiche, Woolwiche, Erithe, Rainam
Marshe, Pourfleete, Stone, St Clementes, Greenehithe, Graies,
Northfleete, Grauesende, Tilberie forte, The Campe, Grauesende
forte, the olde Blockhouse (marked thus on both sides of the river),
East Tilberie, Cliffe.
Group Three labels are written in the river, looking from downstream from London:
the Poole
Ratcliffe Checke
Limehouse Reache
Greenewiche Reache
Blackwalle Reache
Cockpull Reache
Podds elmes Reache
Woolwiche Reache
Gallion Reache
Tripcott Reache
Crosse Nesse Reache
Erithe Reache
Longe Reache
St Clements Reache
Northfleete Hope
Grauesende Reache
Tilberie Hope
These short stretches of river from bend to bend enable the sailor to
identify his position at any one point.
There are two labels written on Stepney Marsh: Saunders Nesse at
the south-eastern point of the curve and Ile of Dogges at the south-­
western point. The I of the label Isle of Dogges is located right next
From Clerks to Corpora
to two small islands in the breach opposite Deptford. Both labels
are written perpendicular to the river, aligning them with features of
the riverbank that could be identified from on board ship as aids to
­navigation. They are both identifying features, a ness and an isle, in
otherwise uninterrupted marshland. It might be argued that there is
little room on Stepney Marsh for a label to be written horizontally, and
that therefore the certain identification of the label Ile of Dogges with
the two small islands, rather than the mainland, is not secure. However
this is also the case with what appears to be an anomaly in Group Two,
Rainam Marshe, which, being a natural feature of the riverbank, might
be thought to sit better in Group One. However, Rainam Marshe is
not an identifying feature – all the foreshore from the estuary up was
lined with indistinguishable marshland. This particular marsh is named
because it is depicted as an island, created by a small channel slicing
the tip of Erith Ness from east to west. The perpendicular label Erithe
Nesse (the identifying feature) intersects the horizontal label Rainam
Marshe (the place-name) at right angles. Similarly, the perpendicular
label Greenehithe Nesse (the identifying feature) intersects the horizontal label St Clementes (the place-name) at right angles. The feature
salient to shipping – in this case, the ness or bend of the land – is clearly
differentiated in each case from the place-name. Had Isle of Dogs been
a marsh place-name, it could have been positioned horizontally so as to
intersect the label Saunders Nesse, just as the horizontal label Rainam
Marshe has been positioned. But it was not, and therefore has to be
interpreted as marking the two small islands.
Lastly, let us consider the shipping in the river. There are four separate groups of ships depicted, at the Tilbury Fort/Gravesend boom
(4 or 5 rowbarges and several small boats), Ratcliffe Checke (9 ships),
the Pool (5 ships), and just upstream of London Bridge (3 large vessels,
6 small). The only other vessel to be depicted is a three-masted ship in
the river between Deptford and the Isle of Dogs islands.
Recall the Field Survey of c.1400, where two island names were
mentioned on folio 19, Wereye and Madeye. These are presumably
derived from Old English wer + īeg, ‘Weir Island’ and mǣd + īeg, ‘Mead
Island’. There are only two island names mentioned in the Field Survey
and I speculate that they are the pre-1520 names of the two eyots afore
Deptford, bestowed by those who worked in the marsh, catching fish
in weirs and farming sheep on meadows. Eyots in the Thames are not
permanently fixed entities but shift in shape and position over time, and
can be unified at low water yet divided into two or more at high water.
On the Place-Name Isle of Dogs
Therefore a name of Isle of Dogs for two eyots is not inappropriate: the
Thames eyots upriver at Brentford now collectively known as Brentford
Eyot currently fit this description.24 The foreshore opposite Deptford at
this point in the curve is particularly shifting and unstable,25 and the
two Stepney eyots have since disappeared, but B. H. Cowper records
that the two eyots were positioned where Messrs Ferguson’s mast-pond
was situated in 1853, on the south side of the pond adjoining Tindall’s
dock and the mast house (Cowper 1853: 17) “which pond is an indentation of the river bank, and called Drunken Dock”. Apparently one
of the eyots was still there at that time. This mast pond was directly
opposite the King’s Yard at Deptford (Cowper 1853: 19).
To recap so far: the place-name Isle of Dogs referred originally
not to Stepney Marsh, but to two small islands lying in the river in
a breach on the Stepney side opposite Deptford. The eyots may not
predate 1448, which was when the river flooded that part of the marsh
(they may have resulted from that inundation), and are likely to have
been known to marsh-dwellers as Wereye ‘Weir Island’ and Madeye
‘Mead Island’ between 1448 and 1520. The dating and the context
make it likely that the name or nickname Isle of Dogs was bestowed
by workers in Deptford Dockyard, as they were the ones who created a
dock out of the Stepney Marsh eyots. In which case, what do we know
about the ships that were berthed at this place at this point in time?
Can the Mary George and her ilk tell us anything about the Stepney
Marsh dock?
5. Early Ships at Deptford
We return to the Book of Accounts of the King’s Ships where the name
Isle of Dogs is first mentioned:
“Here Aftyre Ensuythe All Suche Costys Chargys & sundry expenses
hade made & done by the Commaundement of the kinge owre moste
drade soueren lorde henry the viij from the xijth daye of marche in
the xjth yere of his moste nobull Reyngn of fore and apon the kinge
is grette barke the lesse barke the newe barke namyd the Kateryn
24 The western end of Brentford Eyot, which at high water becomes a separate islet, is
known nowadays as Smith’s Eyot or Lot’s Eyot.
25Information from Julian Kingston. Prof Derek Keene suggests that the name Isle
of Dogs denoted the larger island, on later maps marked ‘Osier Hope’, rather than
both of them.
100 From Clerks to Corpora
plesaunce The mary & John wt the ij Rowbarges whiche schippes
were prepared Ryggyd & sett forthe ffor the transportyng of owre
seid soueren lord to Calice and from Calice in to Inglond ageyn”
(October 1520, Kew, TNA MS E 36/11 fo 105)
Table 2. The King’s ships in 1525 (Sources: Loades (2002); “Names of the
King’s ships at Portsmouth and Thames, 22 Oct 17 Hen 8”, Letters and
Papers, Foreign and Domestic, Henry VIII, volume 4: 1524–1530. Tonnage
given where specified. Where tonnages vary it is because more than one figure
is given in the source documents. (bark) signifies that a vessel was specified as
such in the source documents.)
Henry Grace A Dieu
Great Galley
Gabriel Royal
Katherine Fortileza
Mary Rose
Great Barbara
Great Nicholas
John Baptist
Peter Pomegranate
Mary James
Mary George
Great Bark
Mary and John
Lesse Bark
Mary Gylforde
Henry of Hampton
Maudelen of Deptford
Mary Imperial
Katherine Bark
(1,500 tons)
(800 tons)
(800 tons)
(650–700–750 tons)
(700 tons)
(600 tons)
(400 tons)
(400 tons)
(400 tons)
(340 tons)
(260 tons)
(240–250 tons)
(200–250 tons)
(200 tons)
(180 tons)
(160–180 tons)
(160 tons)
(160 tons)
(160 tons)
(120 tons)
(120 tons)
(120 tons)
(100 tons)
On the Place-Name Isle of Dogs 101
Table 2. Continued
Bark of Bullen
Trinity Henry
Bark of Murlesse (Morlaix)
Great Sabra
John of Greenwich
Lesser Sabra
Carvel of Eu
Jennett of Purwyn
Katherine Plesaunce
Mary Fortune
(80 tons)
(80 tons)
(80 tons)
(bark, 65 tons)
(60 tons)
(60 tons)
(50 tons)
(50 tons)
(40 tons)
The king’s lesser bark, his great bark, the little bark and his new bark
were being repaired or built on the Thames in 1520.26 In 1495 the
peacetime navy had no more than three or four ships. By 1523 it had
about thirty, of varying sizes (Loades 2002: 24), as Henry VIII built
up the navy. By 1525, the following 40 of the King’s ships were listed
as being in the Thames, or in the Thames and at Portsmouth:
A bark was a smaller, masted, sailing vessel. If we assume that anything greater than 250 tons (the Great Bark) was not called a bark,
and only include of those without specific tonnage the vessels that were
explicitly named as barks, then even at this conservative estimate, more
than half of Henry’s fleet were known, at the time, as barks.
At this point, let us leave ships and consider other islands in the
Thames. Upstream, small islands are known as eyots (Brentford Eyot,
Chiswick Eyot, Isleworth Eyot, also spelt ait), from Old English īeg
26 Letters and Papers, Foreign and Domestic, Henry VIII, volume 3, October 1520.
102 From Clerks to Corpora
‘island’ with diminutive suffix –eth (OED ait, n., eyot, n., with fifteenth
and sixteenth century attestations spelt le Eyte, hayte). Downstream,
the far bigger islands came to be known as isles (Isle of Grain, Isle of
Sheppey, Isle of Harty, Isle of Thanet). Isle is from Anglo-Norman, ultimately from Latin insula (OED isle, n. 1. a.); ilde of Wi3t is attested from
c1320; jlde off Tenett (Thanet) from 1473.27 I argue in Wright (2010)
that the Thames term for a tidal eyot found at tributaries’ deltas was
a horse (e.g. Fobbing Horse, Upper Horse, Lower Horse, Wodeham
Horse, Sea Horse, Wyllyspitt Horse and perhaps Horse End and
Horsleydown) from OE horsc ‘mud’. I have speculated that the traditional names for the islands opposite Deptford were Wereye and
Madeye, and we might also reasonably expect these islands to have
been known collectively as Stepney Eyot. Instead, the name Isle of
Dogs occurs, in the context of ship-building and dock-working. To
return to the barks, I wonder whether this is Tudor dockers’ word-play.
The grandiose downstream Isle of – has been applied bathetically to
very small islets, and the name Dogs bestowed as a pun, the eyot at
Stepney harbouring barks.
Does this hypothesis fit what we know of the word bark? OED bark
| barque, n. 2 derives bark ‘small ship with sails’ as being “possibly from
Celtic”; “a small vessel with sails; the latter was the sense with which
the word was taken from French into English”. The first attestation in
English is by Caxton in 1477. OED bark, n. 3 derives bark ‘the sharp
explosive cry uttered by dogs’ from the Old English verb beorcan ‘to
bark’. The first written attestation of the noun is not, according to the
dictionary, until 1562, but this need not detain us as the attestation is
poetical (literature being the text-type most thoroughly scrutinised). As
the verb is Old English, the noun could have been derived at any point
in the language’s history. The present-day quasi-archaic spelling barque
is a red herring: a search of the Early English Books Online database
reveals that in the sixteenth century the spelling was always bark(e,
with the –que spelling an innovation of later centuries.
That the pun was possible does not mean that it was certainly the
origin of the place-name Isle of Dogs, of course; it merely raises the possibility. We next hear of the place-name Isle of Dogs on John Norden’s
map of Middlesex of 1593. This is a map of the whole of Middlesex
and so the area is necessarily small and lacking in detail, but two parallel horizontal streams are drawn east-west on Stepney Marsh so as
27 Note that the ‘island’ element in Canvey Island seems to postdate the period under
discussion here. It is attested as Canwaie Iles in 1586 (Reaney 1935: 148).
On the Place-Name Isle of Dogs 103
to cut the marsh into two horizontal bands, with a meridional stream
flowing north-south between the lower horizontal stream and the
Thames, and another smaller north-south stream at Saunders Ness (neither Stepney Marsh nor Saunders Ness is so labelled on the map). Four
labels appear in the marsh, with three symbols identified in the key. The
labels are Stepney at the far north, identified with the symbol for a parish; Limehouse in the north-west and Blackwall in the north-east, both
identified with the symbol for “Hamletes or villages”; and Isle of doges
ferm, the lettering printed in the middle of the marsh but the identifying
symbol placed precisely where Pomfrets was previously situated. This
symbol, the key informs us, indicates “Howses of Knightes, Gent. &c.”,
and indicates that by 1593 the place-name Isle of Dogs had superseded
the place-name Pomfrets for the gentleman’s house and hamlet situated
on the inland part of Saunders Ness. It does not follow that the whole of
Stepney Marsh had become known by that name at that date. In fact all
the evidence accrued so far indicates that Stepney Marsh was still known
as Stepney Marsh until at least 1600 (and indeed considerably later).
6. On London Workers’ Word-Play
If the punning explanation of the place-name Isle of Dogs suggested in
Section 4 is correct – and we will never know, but it does fit the facts,
which no explanation has done hitherto – then dockers’ word-play goes
back at least to the days of the Tudors. That there is ample evidence
that writers of Tudor literature enjoyed word-play perhaps needs no
elaboration here.28 But what about Tudor dock-workers? We cannot
28I leave it to the reader to decide whether both senses of bark are operative in
Shakespeare’s Sonnet 80 (1609):
O how I faint when I of you do write,
Knowing a better spirit doth vse your name,
And in the praise thereof spends all his might,
To make me toung-tide speaking of your fame.
But since your worth (wide as the Ocean is)
The humble as the proudest saile doth beare,
My sawsie barke (inferior farre to his)
On your broad maine doth wilfully appeare.
Your shallowest helpe will hold me vp a floate,
Whilst he vpon your soundlesse deepe doth ride,
Or (being wrackt) I am a worthlesse bote,
He of tall building, and of goodly pride.
Then If he thriue and I be cast away,
The worst was this, my loue was my decay.
104 From Clerks to Corpora
know whether they too created puns. However, we do know that there
has in more recent centuries been a tradition of London workers using
word-play as an integral part of their working day. I give by way of
illustration here a counting system29 long in use between (although not
limited to) importers, wholesalers and retailers in the fruit trade:
Table 3. Fruiterers’ counting terms, London, 2012.
Some of these terms are demonstrably over a hundred years old and
must speak to continuity of usage from father to son. Henry Mayhew
attests to numbers four and seven pronounced backwards: “‘I’ll try
you a “gen”’ (shilling), said a coster; ‘And a “rouf yenap”’ (fourpence),
added the other.” (Mayhew 1851: I 17/2, cited in OED rouf, adj. and
n.; see also neves adj. and n.). OED marks headwords rouf and neves
as ‘now rare’ and ‘obs. rare’, although they have subsequently been
common in the spoken functional variety of market traders.30 Carpet,
from rhyming carpet bag with drag (OED carpet n. II. 6.; carpet bag
n.2; drag n.8 b.), presumably postdates 1830 when carpet bags were in
vogue, but is in reference to drag in the sense ‘stretch of imprisonment’,
attested from 1781. The sense development is from drag ‘handcart’
29Kindly provided by Mr Steve Roberts of Westminster Produce, who is a third-­
generation Covent Garden fruiterer.
30 Sugar must have come later than rouf as sugar loaf rhymes with [rəʊf].
On the Place-Name Isle of Dogs 105
to “The drag, is the game of robbing carts, waggons, or carriages.of
trunks, bale-goods, or any other property. Done for a drag, signifies
convicted for a robbery of the before-mentioned nature” (James Hardy
Vaux A Vocabulary of the Flash Language, 1819) to Henry Mayhew’s
“Sometimes they are detected, and get a ‘drag’ (1851: I. 219/2), in
reference to the length of the term of imprisonment, which seems to
have been at first six, then three months. Monkey, ‘500’, first occurs in
St James’s: A Satirical Poem, in Six Epistles to Mr. Crockford, by
someone writing under the pseudonym of ‘Westminster St James’, first
published in London in 1827. The subject is aristocrats gambling away
the family inheritance at Crockfords gambling club, so that their children, the heirs, are ruined by the losses of the father. The poem mentions an Earl losing ‘ten or twenty ponies’ and a footnote explains: “It
is not every reader that is aware of the modern title by which, in the
Clubs, certain sums of money are recognised. A pony is £25, a rouleau
£50, and a monkey £500. The noble Earl in question, who is gifted with
two sons, partaking very largely of the latter quality, was accosted by
a friend at Crockford’s one evening, (when His Lordship had been a
loser,) who thought him looking very much out of humour, and asked
the cause. ‘I have lost a monkey,’ replied the Earl.” (OED monkey, n.
IV. 23; St James 1827: 134). Pony, ‘25’ (pounds, guineas or sovereigns),
is first attested in a novel of 1797 by Mary Robinson, and again the
meaning is made clear in a footnote: “There is no touching her even for
a poney. [Note. Half a rouleau or twenty-five guineas].” (OED pony,
n.1 and adj. 3.; Robinson 1797: II, 97). Although I can find no specific
supporting evidence or previous discussion, ching ‘five’ would seem to
be derived from Anglo-Norman cinq, surviving along with ace, deuce,
trey, quatre, sise, still in use (or in use until very recently) in the context
of card-play (OED cinque | cinq, n.),31 and in Channel Islands French.
The relevance for our present purposes is that the counting system
(or parts thereof) must have been used over several generations in an
entirely non-literary, working context. It does not constitute proof that
Tudor dock-workers did the same, of course, but it gives a measure of
plausibility to the suggestion that they might have done so.
31 For the palatalisation of the first phoneme, cf. ME chiche ‘chickpea’ < AN chiche < L
cicer (OED chich, n.). Jerriais chîn, chînq ‘five’.
106 From Clerks to Corpora
7. Jonson and Nashe’s play Isle of Dogs
In the summer of 1597 the place-name Isle of Dogs achieved some kind
of notoriety, but it is unclear why. The playwrights Ben Jonson and
Thomas Nashe (and perhaps members of Pembroke’s company) wrote
and performed, but did not publish, a satirical comedy called The Isle
of Dogs (; Donaldson 2012: 101). As a direct result,
Jonson and two members of the company were imprisoned, and Nashe
fled into hiding. Donaldson (2012: 103–107) reconstructs events: in late
July or early August, the Queen’s inquisitor, interrogator and torturer
Robert Topcliffe was instructed by the Privy Council to discover the
instigators and perpetrators of the play, and to seek out all copies and
their owners. Nashe could not be found, but Jonson and his two fellow
players were arrested and imprisoned. They were examined on 15th
August by the Privy Council Court at Greenwich, and accused of “lewd
and mutinous behaviour,”32 the play containing “very seditious and
slanderous matter”. It is likely that they were tortured, but seven weeks
later, on 8th October 1597, the three men were set free. Simultaneously,
although it is not clear whether the two events were related, the Lord
Mayor and Court of Common Council requested on 28th July 1597
that the Privy Council ban all theatrical activity, it being to the detriment of the well-being of the citizens (the Corporation of London
had made similar requests on previous occasions). The same day, the
Privy Council banned all plays in London throughout the summer, and
ordered that theatres be dismantled – although it seems that the second
part of this edict was not carried out. Was this prohibition a response
to recent productions of The Isle of Dogs at the Swan playhouse? The
impresario Henslowe recorded a memorandum on 10th August that the
current restraint was “by the meanes of playinge the Jeylle of Dooges”,33
so he seems to have thought that it was. Donaldson (2012: 106) summarises the speculations of literary historians about who, exactly, might
have found the play offensive. Individuals have been suggested,34 but
Donaldson suggests that the severity of the response seems too great
to have been triggered by an attack on a mere individual, and suggests
that it may have touched upon matters of national defence (2012: 107).
32 lewd meant ‘evil, wicked, unprincipled’ in this context; see OED lewd, adj. 5.
33,_The. Jeylle of Dooges = Isle of
34 The King of Poland, the late Lord Chamberlain (Donaldson 2012: 106, and references therein).
On the Place-Name Isle of Dogs 107
Following Donaldson’s suggestion, if we take the phrase ‘Isle of
Dogs’ to have conveyed a meaning of something along the lines of
‘place where warships were fitted out’, then the question follows, what
was happening to shipping that summer? Wernham (1994: 143–190)
describes in detail the movements of the fleet. The Queen’s main priority that year was to prevent the Spanish armada35 from attacking
Britain. At the beginning of 1597, relations with France were “about
at their lowest” (Wernham 1994: 146), and her two counsellors Sir
Robert Cecil and the Earl of Essex were quarrelling with each other.
There was rebellion in Ireland, which rebels (it was feared) might seek
help from Spain, and the wheat harvest was failing. National security
seemed greatly threatened. Elizabeth was persuaded that her best form
of defence was attack, and on 10th March, she made Essex Master of
the Ordnance. However it was not clear during the spring what his
commission would be – whether to attack the armada at Ferrol, to
put down the rebellion in Ireland or to attack Calais – and it was not
until 9th May that the Privy Council wrote to the Lords Lieutentant
and Commisioners of Musters instructing them to mobilise troops for
service abroad. In early April, a plan was made to fit out two fleets.
The first would lie off the Spanish coast and intercept the armada.
The second fleet would follow, and be the striking force. But Elizabeth
delayed, and withheld her permission until the 20th May. Immediately
thereafter, during the end of May, 28 ships of war were fitted out,
plus around 30 flyboats to transport troops.36 The two-fleet plan was
dropped and consolidated into one ‘powerful strike force’, with Essex
in command.
Essex’s commission had been sealed on 4th June, that is, 4th June
is the date from when it could have been known that a war fleet was
being assembled. He received his detailed instructions privately on 15th
June: he was to destroy the armada then lying at Ferrol, and afterwards sail to the Azores to intercept the Spanish treasure ships where
they revictualled on returning from the East and West Indies, and to
return home by winter. He was forbidden to attack either Spain or
Portugal. Essex’s fleet finally set out on 10th July, but a week of storms
caused half the fleet to return, leaking and damaged. The other half
Known as the Ferrol armada, from where it spent the summer of 1597.
36 Cadwallader (1923: 5) specifies 38 vessels and troop-carriers, plus 5/6 small vessels
with each of the 4 squadrons, plus 20 voluntary barks on the look-out for plunder.
The discrepancy arises, I think, due to differences in classifying the smaller ships.
108 From Clerks to Corpora
Table 4. Ships of the 1597 Islands Voyage fleet. (Sources: Wernham (1994),
Gorges (1625), Cadwallader (1923). Dating and construction information
taken from Colledge and Warlow eds (2006),
Where built
When built
Ark Royal
Due Repulse
Elizabeth Jonas
Golden Lion
Mary Rose
St Andrew
St Mathew
Warspite/Wast Spite
White Bear
Deptford 1597
On the Place-Name Isle of Dogs 109
sailed close to Ferrol, close enough that the armada could see them,
arriving there on 25th July. After this show of bravado, the half-fleet
sailed back to Plymouth and regrouped on 31st July. Frustrated by bad
weather, Essex and Sir Walter Raleigh planned to quit Spain altogether
and sail instead to the West Indies to capture treasure ships. But the
Queen forbade it: defeat of the armada was the nation’s first priority.
On 14th August the wind changed, and the fleet duly set out for Ferrol,
only to become becalmed. This is the point at which Jonson and his
two collaborators were being interrogated. Looking back, on 28th July
when the edict against plays and playhouses went out, half the fleet had
been lying storm-damaged at Plymouth and Falmouth, and the other
half had been taunting – but not destroying – the Spanish armada. In
June, Elizabeth had been vacillating, and at the end of May, the dockyards had been busy fitting ships for war. We cannot know precisely
when the play Isle of Dogs was written, but if it was topical, it was
created and performed during sensitive times for national security.
There might be evidence that it predates June 11th: an entry in the
Calendar of State Papers for that date reads:
John Chamberlain to Dudley Carleton, attendant on the ambassador
at Paris. … There are great preparations for a voyage, some say for
Calais, some the islands of Jersey, some the King of Spain’s navy, or
the Indian fleet. There are 15 of the Queen’s ships, two Spanish ships
taken last year and re-fashioned, 22 Holland men-of-war, and 24
fly-boats for carriage of men and victuals. … There is a new play of
humours in great request, but it is great cry for little wool.”
(Green (ed.) 1869: 437–8)
Humours at this date meant ‘an excited state of public feeling’.37
If this is our play, then it was composed before June 11th, during the
weeks when the Deptford Dockyard would have been fitting out the
fleet for war. If national defence was the target, or perceived target, of
the play’s satire, as Donaldson (2012: 107) hypothesises, then against
a background of such threats to national security, any criticism of
37 OED humour, n. II. 5. c., first attestation 1600: “It was not fitte to stirre up humours
in Spaine.”
110 From Clerks to Corpora
strategy, policy or capability of the fleet could have been interpreted as
mutinous and seditious.38
Subsequent events are not material to the play, but if The Isle of
Dogs had criticised national defence, its authors would indeed have
been percipient. Never had England been allowed to be so off-guard.
Essex’s fleet had left Plymouth on 17th August, and again been assailed
by poor weather, causing damage to some of the ships. By 27th August
the fleet had been scattered and partly blown past Ferrol. Unable to sail
back due to an adverse wind, Essex decided to press on and intercept
the Spanish treasure fleet returning from the Indies. On 30th August
he was misinformed that the armada had set sail for the Azores, and
so the fleet arrived there in early September. Raleigh’s troops caused
damage to one of the islands, and Essex did manage to capture three
Spanish cargoes, but on 9th October the fleet set sail for home, arriving
separately in the last week of October. They did not know that on 9th
October the armada had also set sail, leaving Ferrol bound not for the
Azores but for England. The Spaniards had intelligence that Essex was
in the Azores and had left England undefended, and their plan was to
attack Falmouth and Plymouth, then lie in wait to intercept Essex’s fleet
as it returned. In the event, the armada got to within 30 to 10 miles
off the Lizard when adverse weather set in, storms damaged the ships,
and the armada returned to Coruña and Ferrol. There was no attack
(Wernham 1994: 183–190; Green (ed.) 1869: 520).
There were two further dramatic uses of the place name Isle of Dogs
published sufficiently close to 1597 to constitute a possible reference to
the events of that summer. The first occurs in Middleton and Dekker’s
play The Roaring Girl, or Moll Cutpurse, published in 1611. As well
as a pun on dogs, there is a context of knavery and fighting. The placename occurs in a conceit of sea-faring, which has nothing to do with
the rest of the play, which is set in Holborn.
Souldier? thou deseru’st to bee hang’d vp by that tongue
which dishonours so noble a profession, souldier you
38 A trivial possibility, for example, arises from the detail of Essex’s instructions on his
appointment to the office of Master of the Ordinance. “We would prevent your falling into the errors of your predecessors” … “You shall keep the quantity of stores
a secret from all but our sworn servants, not using your own clerks or strangers,
which has heretofore done disservice” (Green (ed.) 1869: 381–3). Apparently the
previous Master of the Ordinance, the Earl of Warwick, had pilfered the stores and
stolen munitions.
On the Place-Name Isle of Dogs 111
skeldering varlet? hold, stand, there should be a trapdore
here abouts.
Pull off his patch
The balles of these glasiers of mine (mine eyes) shall be
shot vp and downe in any hot peece of seruice for my inuincible Mistresse.
Iac. Dap. I did not thinke there had bene such knauery in blacke
patches as now I see.
Oh sir he hath bene brought vp in the Ile of dogges, and
can both fawne like a Spaniell, and bite like a Mastiue, as
hee finds occasion.
L. Nol. What are you sirra? a bird of this feather too.
T. Cat. A man beaten from the wars sir.
T. Long. I thinke so, for you neuer stood to fight.
Iac. Dap. What’s thy name fellow souldier?
T. Cat. I am cal’d by those that haue seen my valour, Tear-Cat.
Omnes. Teare-Cat?
A meere whip-Iacke, and that is in the Common-wealth of
rogues, a slaue, that can talke of sea-fight, name all your
chiefe Pirats, discouer more countries to you, then either
the Dutch, Spanish, French, or English euer found out,
yet indeed all his seruice is by land, and that is to rob a
Faire, or some such venturous exploit; Teare-Cat, foot sirra
I haue your name now I remember me in my booke of
horners, hornes for the thumbe, you know how.
T. Cat. No indeed Captaine Mol (for I know you by sight) I am
no such nipping Christian, but a maunderer vpon the pad
I confesse, and meeting with honest Trapdore here, whom
you had cashierd from bearing armes, out at elbowes
vnder your colours, I instructed him in the rudements of
roguery, and by my map made him saile ouer any Country
you can name, so that now he can maunder better then my
(Middleton & Dekker, 1611,
The Roaring Girl, or Moll Cutpurse)
The place-name Isle of Dogs also occurs in Beaumont and Fletcher’s
Thierry and Theodoret, first published in 1621. Again, there is a pun –
this time on cats – and a context of quarrelling and fighting:
Enter Thierry, Theodoret, Brunhalt, Ordella, Memberge, Martell.
Thier. What villain dares this outrage?
112 From Clerks to Corpora
Devitry. Hear me, Sir, this creature hir’d me with fifty crowns in
hand, to let Protaldye have the better of me at single Rapier
on a made quarrel; he mistaking the weapon, laies me over
the chops with his club fist, for which I was bold to teach
him the Art of memory.
Omnes. Ha, ha, ha, ha.
Theo. Your General, Mother, will display himself.
‘Spight of our Peace I see.
Thier. Forbear these civil jars, fie Protaldy,
So open in your projects, avoid our presence, sirrah.
Devi. Willingly, if you have any more wages to earn,
You see I can take pains.
Theo. There’s somewhat for thy labour,
More than was promis’d, ha, ha, ha.
Bawdb. Where could I wish my self now? in the Isle of Dogs.
So I might scape scratching, for I see by her Cats eyes
I shall be claw’d fearfully.
Thier. We’ll hear no more on’t,
Soft Musick.
(Beaumont and Fletcher, 1621,
Thierry and Theodoret, Act II Scene 2)
The place-name Isle of Dogs may have been chosen for no other
reason than it allows wit about dogs and cats and was part of the zeitgeist.39 The quarrelsome contexts could be due to coincidence (quarrels
being dramatic staples), or there may be some allusion here to the contents of the 1597 play. Duplicity would seem to be relevant to fawning
like a spaniel and biting like a mastiff. And it must be borne in mind
that the 1597 play may have had nothing to do with Donaldson’s suggestion of the Islands Voyage at all, as there were plenty of other sensitivities to probe.
39 Names which catch the zeitgeist can move well outside their original sphere of reference; see Wright (2011) for waterloo blue (battle 1815, dye 1823), magenta red
(battle 1859, dye 1860), solferino (battle 1859, dye c.1865), waterloo bang-up (a
type of cracker, 1826), waterloo cracker (1833), trafalgar chair (battle 1805, chair
1822), trafalgar cotton (1826), trafalgar coach (1848). Gibraltar rock (1831) and
Wellington pillars (1851) were types of sweets. In the 1860s the name Garibaldi
was marketed as the name of a type of blouse, in the 1880s as both a hat and a type
of fish, and then from the 1890s to the present as a type of biscuit.
On the Place-Name Isle of Dogs 113
8. Conclusion
It now remains to sort hypothesis from fact. I have hypothesised that
the name Isle of Dogs may be Deptford dockworkers’ word-play on the
barks berthed at Stepney Eyot. This hypothesis hinges on the two meanings of the word bark on the one hand, and the fact that the place-name
is first known to us in the context of the Deptford Dockyard in 1520,
just seven years after that dockyard opened, on the other. Were further
research to reveal an earlier, pre-1513 attestation, this hypothesis would
no longer stand. It is also worth emphasising that just because the pun
was possible, it does not follow that it is necessarily the correct explanation. What is certain, however, is that sixteenth-­century dockworkers
used the name Isle of Dogs for the eyots opposite their yard, rather
than for the whole of Stepney Marsh as it is known today. Therefore,
whatever the contents of the lost 1597 play The Isle of Dogs, the reference may have been concerned with that dockyard in some way, if that
reference were indeed topical rather than allegorical or fantastical.
Kew, The National Archives MS E 36/11. Accounts Book of the King’s Ships.
1520. Calendared in Letters and Papers of Henry VIII, volume 3, pp
Kew, The National Archives, MS SC 12/11/31. Field Survey of Stepney, c.1400.
London, British Library, Adds MS 44839. Robert Adam, Thamesis Descriptio,
London, British Library, Cotton MS Otho E IX, 1525. Calendared in Letters
and Papers of Henry VIII, volume 4, pp 757–772.
London, London Metropolitan Archives, DL/C/B/004/MS09171/002, wills,
Printed Sources
Baker, T.F.T. (ed.). (1998). Stepney: Settlement and Building to c.1700. A
History of the County of Middlesex: Volume 11: Stepney, Bethnal Green.
British History Online. Web. 05 November 2014. http://www.british-­
Cadwallader, L. H. (1923). Career of the Earl of Essex 1597–1601. Philadelphia:
University of Pennsylvania.
114 From Clerks to Corpora
Colledge, J.J. & Warlow, B. (eds) (2006). Ships of the Royal Navy: the Complete
Record of all Fighting Ships of the Royal Navy from the 15th Century to
the Present. London: Chatham.
Cowper, B.H. (1853). A Descriptive Historical and Statistical Account of
Millwall, Commonly Called the Isle of Dogs; Including Notices of the
Founding, Opening, Etc., of the West India Docks and City Canal, and
Notes Relating to Limehouse, Poplar, Blackwall, and Stepney. London:
Robert Gladding.
Croot, P. (1997). Settlement, tenure and land use in medieval Stepney: evidenceof a field survey c.1400. London Journal 22:1, 1–15.
Cunningham, P. (1849 edition). A Handbook for London, Past and Present.
Volume 2. London: John Murray.
Currie, C. R. J. (ed.). (1998). A history of Middlesex. Vol.11, Early Stepney with
Bethnal Green. The Victoria history of the counties of England. Oxford:
Published for the Institute of Historical Research by Oxford University Press.
Donaldson, I. (2012). Note on The Isle of Dogs, lost play (1597). D. Bevington,
M. Butler and I. Donaldson (eds) The Cambridge Edition of the Works
of Ben Jonson. Volume 1, 1597–1601. Cambridge: Cambridge University
Press, 101–109.
Dugdale, W., Sir. (1662). The History of Imbanking and drayning of divers
Fenns and Marshes, both in forein parts, and in this Kingdom. And of the
improvements thereby. London: Alice Warren.
Fletcher, J. & Massinger, P. (1679). The Tragedy of Thierry and Theodoret.
London: John Martyn, Henry Herringman.
Gorges, A., Sir. (1625). A Larger Relation of the said Iland Voyage. Purchas,
Samuel. Purchas his Pilgrimes. Book 4. (1938–1969). London: Henrie
Green, M.A.E. (ed.). (1869). Calendar of State Papers, Domestic Series, of the
reign of Elizabeth, 1595–1597. London: Longmans, Green, Reader, and
Johannesson, N-L. (2012). “Rihht alls an hunnte takeþþ der. /Wiþþ hise 3æpe
racchess”: Hunting as a metaphor for proselytizing in the Ormulum. R.
Dance and L. Wright (eds) The Use and Development of Middle English:
Proceedings of the Sixth International Conference on Middle English,
Cambridge 2008. Frankfurt am Main: Peter Lang, 229–240.
Loades, D. ([1995] 2002). From the King’s Ships to the Royal Navy 1500–
1642. J.R. Hill & B. McL. Ranft (eds) The Oxford Illustrated History of the
Royal Navy. Oxford: Oxford University Press, 24–55.
On the Place-Name Isle of Dogs 115
Middleton, T. & Dekker, T. (1611). The Roaring Girle. Or Moll Cut-Purse.
London: Thomas Archer.
Mills, A. D. ([2001] 2004). Oxford Dictionary of London Place Names.
Oxford: Oxford University Press.
Norden, J. (1593). Speculum Britannicae. The First Parte. An Historicall, &
Chorographicall Discription of Middlesex. Wherin Are Also Alphabeticallie
sett Downe, the Names of the Cyties, Townes, Parishes Hamletes, Howses
of Name &c. Wth Direction Spedelie to Finde Anie Place Desired in the
Mappe & the Distance betwene Place and Place without Compasses. No
Reaney, P. H. (1935). The Place-Names of Essex. English Place-Name Society,
volume 12. Cambridge: Cambridge University Press.
Robinson, M. (1797). Walsingham: or, the Pupil of Nature: a Domestic Story.
2 volumes. London: T. N. Longman.
St James, Westminster. (1827). St James’s: A Satirical Poem, in Six Epistles to
Mr. Crockford. London: Ibotson and Palmer.
Strype, J. (ed.). (1720). A survey of the Cities of London and Westminster:
Containing the Original, Antiquity, Increase, Modern Estate and Government
of those Cities. Written at first in the year MDXCVIII by John Stow. London:
A. Churchill, J. Knapton, R. Knaplock, J. Walthoe, E. Horne, B. Tooke,
D. Midwinter, B. Cowse, R. Robinson and T. Ward. Volume 1.
Wernham, R.B. (1994). The Return of the Armadas: The Last Years of the
Elizabethan War against Spain 1595–1603. Oxford: Clarendon Press.
Wright, L. (2006). Street addresses and directions in mid-eighteenth-­century
London newspaper advertisements. N. Brownlees (ed.) News Discourse
in Early Modern Britain. Selected papers of the Conference on Historical
News Discourse 2004. Linguistic Insights Studies in Language and
Communication, Volume 30. Bern: Peter Lang, 199–215.
———. (2010). A pilot study on the singular definite articles le and la in
­fifteenth-century London mixed-language business writing. R. Ingham (ed.)
The Anglo-Norman Languge and its Contexts. York: York Medieval Press
and The Boydell Press, 130–142.
———. (2011). The Nomenclature of some French and Italian Fireworks in
Eighteenth-Century London. The London Journal 36:2, 109–39.
Online References
Early English Books Online:
Letters and Papers, Foreign and Domestic, Henry VIII, Volume 3: 1519–1523
116 From Clerks to Corpora
(1867), pp. 369–381. URL:
compid=91055. Date accessed: 16 November 2012.
Literature Online:
OED Online.
‘The Isle of Dogs: Introduction’, Survey of London: volumes 43 and 44: Poplar,
Blackwall and Isle of Dogs (1994), pp. 375–387. URL: Date accessed: 15 November 2012.
‘The West India Docks: The buildings: warehouses’, Survey of London: volumes 43 and 44: Poplar, Blackwall and Isle of Dogs (1994), pp. 284–300.
URL: Date
accessed: 16 November 2012.
7 English Genres in Diachronic Corpus
Erik Smitterberg & Merja Kytö
Uppsala University
1. Introduction
This chapter is about problems and possibilities associated with using
genre as a parameter in corpus-based historical linguistics. We will begin
by discussing why genre has become an increasingly central concept
in historical linguistics and by defining genre and related terms. Next,
we will discuss a number of challenges that corpus linguists need to
address when they use genre as a parameter in their research. To begin
with, we will discuss potential conflicts between two key desiderata,
namely representativity and comparability. We will also take up the
problem that not all genres are attested for the whole history of English
and that even genres which have a long history may have changed over
time. Thirdly, we will discuss how historical linguists have used genre
comparisons to access an approximation of past speech. We will then
devote a section to two case studies where genre plays a central role. We
first look at how additional information can be gained by considering
genre differences within the framework of multi-feature approaches to
genre variation in the past; this account is followed by an analysis of a
single linguistic feature, viz. the units co-ordinated by and. The chapter
ends with a summary of our main points and some desiderata for future
2. The Centrality of the Genre Concept
Even though linguistic variation according to genre was recognized as an
important variable before the advent of corpus linguistics, it has become
even more central in corpus-based approaches (see Lange 2012: 401).
How to cite this book chapter:
Smitterberg, E. and Kytö, M. 2015. English Genres in Diachronic Corpus Linguistics.
In: Shaw, P., Erman, B., Melchers, G. and Sundkvist, P. (eds) From Clerks to Corpora: essays on the English language yesterday and today. Pp. 117–133. Stockholm: Stockholm
University Press. DOI: License: CC-BY.
118 From Clerks to Corpora
This is so partly because every corpus-compilation project has to take
the genre parameter into account. If the researcher is compiling a singlegenre corpus, delimiting the genre sampled is crucial in order to reach
reliable results. And if the corpus project includes several genres, considering genres in relation to one another is a key issue when the compiler
decides what research questions studies based on the corpus can hope to
According to Kohnen (2001: 115), genres can serve as vehicles for
spreading language change. That is, while genres themselves may not
bring about change, they can certainly affect whether or not a change
will spread through a language. For instance, if an incoming informal
feature such as the contracted form can’t instead of cannot becomes
accepted in informal writing, colloquial written genres can function as
“bridgeheads” from which the innovative feature can colonize other
forms of writing. Hundt & Mair (1999: 236) note that a new form
which arises in speech often “then spreads at differential speeds through
various genres until at a very remote point it can be said to have been
established in ‘the language’ ”. At the same time, genres may also retard
a change that is spreading through a language by preserving conservative
and/or fossilized usage; for instance, many legal texts preserve an obligation use of shall that is no longer current in everyday communication.
There is a certain amount of terminological confusion in linguistic
research that considers the genre parameter. Terms such as genre,
register, and text type are sometimes used by different linguists to mean
more or less the same thing. In this chapter, we use the term “genre” to
refer to categories of texts that are defined on extralinguistic or text-external grounds (the term “register” has also been used to cover such categories in previous work); in contrast, we reserve the term “text type”
for categories that are defined based on their linguistic characteristics.
Within this framework, the linguistic make-up of the text itself thus
does not determine what genre it belongs to; for instance, a novel may
be written in the form of a series of letters or diary entries and still
remain a novel. This means that genres are “fuzzy sets”: central members of a genre category will have a form that is close to the genre prototype, while more peripheral members of the set will deviate from the
prototypical pattern. For instance, a prototypical member of a genre
such as academic writing will contain a relatively large number of linking adverbials, prepositional phrases, and passive clauses. However,
text types do not have to correlate with genres; it would be fully possible, for instance, to write a novel where the text type was Scientific
English Genres in Diachronic Corpus Linguistics 119
Exposition, although it would not be a prototypical member of the
genre; Susanna Clarke’s 2004 novel Jonathan Strange & Mr Norrell,
for instance, remains a historical fantasy novel even though it contains
numerous footnotes, a feature not usually associated with that genre.
Other criteria for genre membership include the function of the texts in
the society in which they are used (a novel, for example, may be read
primarily for amusement) and audience expectations (members of the
audience at a linguistic symposium, for instance, will typically expect a
paper to inform them with regard to some aspect of linguistics). What
we will mainly focus on in the remainder of this section is a set of problems that historical linguists typically encounter when including genre
as a parameter in their research.
We will begin by addressing the issue of representativity. If a corpus
is representative, the study of that corpus (or combination of corpora)
“can stand proxy for the study of some entire language or variety of
a language” (Leech 2007: 135). However, while this notion is simple
enough to define, the practical application is problematic, especially
with regard to historical texts (see, for instance, Biber 1993 and Leech
2007 for different suggestions on how to operationalize this parameter).
Several genres – e.g. everyday conversation – are absent from the historical record even though they were important components of the
language variety that researchers wish to represent, and even attested
genres contain only a few of the textual witnesses that once existed.
Another desideratum of corpus-based research is comparability,
viz. the extent to which sets of material are equivalent except for one
single variable (Leech 2007: 141): in the case of historical linguistics,
that variable is usually time. A typical example of comparability concerns the “clones” of the LOB corpus with texts from 1961; roughly
equivalent publications have been sampled from later years (FLOB) as
well as earlier decades (e.g. BLOB-1931) at intervals of approximately
30 years.
However, one important problem in diachronic corpus design is that
representativity and comparability may clash. One reason why this
happens is that genres develop and change through time, as shown in
Figure 1. In Figure 1, the two ellipses represent the textual universes of
newspaper English in two different periods. The difference in horizontal position indicates genre development: the textual universe in period
2 is greater, as new genres have been added (e.g. interviews), but a few
genres such as shipping news have also all but disappeared during the
time that separates the periods. If precedence is given to comparability,
120 From Clerks to Corpora
Period 2
News report
Shipping news
Period 1
Figure 1. Textual universes of newspaper language in two periods.
only the genres that are present in both periods should be sampled
(as illustrated by the two circles); this would make the period samples maximally comparable, but each sample would become less representative of the language of its period. The other strategy would be
to make each period sample representative (as illustrated by the two
narrow ellipses), which would instead decrease the comparability of
the period samples.
Whole genres may even change across time to serve language users’
needs. Drama texts included in the Helsinki Corpus (1420–1500) may
serve as an example. In the Late Middle English section of the corpus, the Drama texts included are religious mystery plays, while the
Early Modern English section contains Drama comedies (this difference
between the samples is also recognized by the corpus compilers in that
“Drama” has been qualified with different subtitles in these two cases).
Both mystery plays and comedies are Drama texts in the sense that
they are scripted dialogue texts used for stage performances. However,
differences in the intended main functions of these two forms of drama
(religious instruction vs. entertainment) may decrease the comparability of such samples from a corpus-linguistic standpoint. On the other
hand, including such heterogeneous genres will make the corpus more
representative of late Middle and early Modern English.
English Genres in Diachronic Corpus Linguistics 121
Genres represented in period 2
Genre P
Genre Q
Genre P
Genre S
Genre R
Genre S
Genres represented in period 1
Figure 2. Variation in genre representation between two periods.
In addition, even if genres do not change over time, they may emerge
late in the period studied or die out, leading to other representativity
and comparability issues. This is illustrated in Figure 2. If the aim is to
produce two comparable period corpora, only genres P and S should be
included in corpus compilation, as they are represented in both periods.
However, this would lead to each period sample being less representative, as genre Q in period 2 and genre R in period 1 would be ignored.
As Figure 2 implies, not all genres have been attested throughout the
recorded history of English; for instance, the emergence of some genres
is tied to technological developments (e.g. e-mail). The patchy picture
that some genres afford researchers is illustrated in Figure 3.
Figure 3. Genre representation in English across time.
122 From Clerks to Corpora
Figure 3 illustrates one genre and two genre groups. While informal
conversation has of course existed since the beginnings of the English
language, it is only recorded from the 20th century on; the line is thus
dashed for most of the history of the language. Genres that belong to
newspaper language can be said to have existed in English since 1665,
when the Oxford Gazette was first published, although there were precursors of newspapers before then, such as corantos (see Dahl 1953
and Brownlees 2012 for accounts of the beginnings of periodical news).
Finally, genres may disappear at a certain point in the history of a given
language only to re-emerge later on. This is the case with law texts in
English. They are attested in the Old English period, but are absent
from most of the Middle English period, as Latin and, later, French
took over in official use in England. However, after a gap of several
centuries, law texts written in English re-appeared in the late Middle
English period (Claridge 2012: 239–240).
Yet another scenario that may introduce problems for corpus compilers is one in which two or more genres exist in all periods studied,
but their relative importance has changed over time. The question then
becomes whether this change should be reflected in corpus compilation.
Figure 4 illustrates this situation. The news report did exist in seventeenth-century England: as mentioned above, the Oxford Gazette, later
the London Gazette, appeared in 1665, and there were precursors such
as corantos. But it was not a central genre compared with the Bible,
The late twentieth century
The Bible
The news report
The late seventeenth century
Figure 4. The relative importance of two genres in late seventeenth-century
and late twentieth-century English.
English Genres in Diachronic Corpus Linguistics 123
which was part of most language users’ daily lives. The situation had
arguably been reversed by the late twentieth century: the secularized
nature of modern Britain means that most speakers have limited contact with biblical texts, while news in some form – newspapers, television, the Internet, etc. – reaches nearly all language users. In addition,
multiculturalism has led to the establishment of religions that do not
make use of the Bible in Britain, which further decreases the special
standing of that text.
There are thus clear difficulties involved both in achieving representativity and in balancing comparability and representativity. Depending
on the underlying research questions that the study of a given corpus is
intended to help answer, corpus compilers have approached these difficulties from different angles. For instance, the compilers of the BLOB1931 clone of the LOB corpus gave precedence to comparability over
representativity in the selection of newspaper texts (Leech 2007: 143):
priority was given to including the same newspapers in BLOB that had
been included in LOB, even though a given newspaper may not have been
equally representative of its genre around 1931 and in 1961. However,
as Leech (2007: 143) notes, in investigations with greater time-depth,
complications relating to genre evolution would have made comparability very difficult to achieve. In such cases, varying the granularity of
the genre parameter may be useful. When possible, corpus texts may
be coded not only for genre, but also for subgenre and/or subfunction.
Biber and Gray (2013) argue that keeping subgenres such as newspaper
articles and news magazine articles constant may be essential in order to
allow researchers to identify and describe language-change phenomena
with a high degree of reliability. As regards subfunctions, Kohnen (2007)
suggests that coding parts of corpus texts according to the subfunction
they fill can help to make texts more comparable; for instance, a genre
like religious instruction can be divided into subfunctions such as narration and exegesis, and texts belonging to the same genre can then be
compared to see whether they also emphasize the same subfunctions.
Alternatively, grouping several genres together into hyperonymic
entities can enable scholars to collect roughly comparable text categories if not all genres are attested in all periods sampled. The prototypical text categories in the Helsinki Corpus can be used as an example
of this; for example, the category Imaginative Narration covers genres
such as Fiction, Romance, and Travelogue, which suggests that these
genres share some features that may make them roughly comparable if
one or several of them are missing from some period samples.
124 From Clerks to Corpora
In contrast, if priority is given to representativity, some steps have
been taken to improve a corpus in this regard. These include covering
a wide range of genres, giving precedence to texts that are considered
good representatives of their genres, aiming at a proportional representation of the genres included, and simply enlarging the size of the
corpus (see Kytö & Smitterberg forthcoming for a fuller discussion). In
historical linguistics, the most serious obstacle to achieving representativity is of course the lack of spoken texts. The available material has
been preserved in writing, while a great deal of the actuation of change
is likely to have taken place in speech and notably in speech used in
dialogue situations (see, however, Biber & Gray 2011 for an account
of change that has spread mainly in non-speech-related writing). One
solution has been to turn to comparisons of texts and genres that stand
at different distances from past speech. For instance, written records
of spoken language can be assumed to come closer to the actual spoken language of the time than written language that was not based on
language taken down in speech situations or created to imitate speech.
Even though it is practically impossible to take down speech in writing in all its aspects, previous research indicates that the essence of
what was said was relayed in the recorded version. Consequently, it is
necessary to look for what have been referred to as “spoken”, “oral”
or “colloquial” genres. To this end, scholars have compiled genre-specific corpora that focus on speech-related language. For example, in
the Corpus of English Dialogues 1560–1760 (CED), past speech can
be studied from the perspective of speech-purposed (e.g. Drama) and
speech‑based (e.g. Depositions) genres (see Culpeper & Kytö 2010 for
more detailed information on CED, including case studies). The results
of such analyses can then be compared with those based on speech‑like
genres such as Private Correspondence in order to shed further light on
the spoken language of the period.
3. Case Studies: Drama and Science in Focus
We will now consider two actual studies to show how the impact of the
genre perspective has affected historical corpus linguistics. Our examples are taken from the Modern English period. One study considers
the co-variation of a large number of features, while the other focuses
on a single linguistic feature.
One way of looking at linguistic variation is to consider how a large
number of linguistic features co-vary in texts. Within this framework,
English Genres in Diachronic Corpus Linguistics 125
different features load on a number of dimensions of variation. Features
that co-occur in texts end up on the same pole on a dimension of variation; features that tend not to co-occur also belong to the same dimension, but will be placed on opposite ends. When genres are positioned
on these dimensions, they can be shown to be more or less “involved”,
“informational”, etc., depending on the co-occurrence patterns of the
linguistic features that are included in the analysis. Among the pioneers
in extending this methodology to historical texts are Douglas Biber and
Edward Finegan; we will consider one of their studies here, viz. Biber
& Finegan (1997). This study is based on the original version of A
Representative Corpus of Historical English Registers (ARCHER), a
1.7-million-word corpus of British and American English which covers
the period 1650–1990. The features studied include colloquial features
such as contractions as well as features characteristic of impersonal
styles, e.g. passives (Biber & Finegan 1997: 258–259).
In their study, Biber and Finegan (1997) demonstrate that not all
English genres have followed the same trajectory through Modern
English. Whereas specialist expository writing has consistently tended
towards more “literate” styles, popular non-expository texts show a
reversal of this trend towards more “oral” styles during the Late Modern
English period (Biber & Finegan 1997: 272–273). As a result, the linguistic differentiation between different kinds of writing increases during the period studied. For instance, while Drama leads the way towards
a more involved style of communication, science writing exhibits an
equally clear trend towards the informational end of the same dimension (Dimension 1, “Involved versus Informational Production”) (Biber
& Finegan 1997: 266).
Another way of considering linguistic variation is from the perspective of a single linguistic feature. The second case study, which concerns
the co-ordinator and and the linguistic units that and can co-ordinate,
exemplifies this perspective. Following Quirk et al. (1985), we will refer
to these units as “conjoins”. And can link conjoins on different levels of
syntactic structure, from individual morphemes to whole sentences. In
the literature, a basic distinction is often made between clausal co-ordination and phrasal co-ordination (see, for instance, Culpeper & Kytö
2010). In example (1) below, the co-ordination is clausal because the
two conjoins are main clauses. In example (3), in contrast, the two
conjoins are adjective phrases, so we are dealing with phrasal co-ordination. We apply a similar, but slightly modified classification to the
data in this study.
126 From Clerks to Corpora
The main reason why co-ordination by and is of interest is that “oral”
and “literate” genres display different preferences as regards phrasal and
clausal co-ordination. Biber et al. (1999) show that, while clausal co-ordination predominates in conversation in Present-day English, phrasal
co-ordination is frequent in academic writing. Moreover, Biber (2003)
demonstrates that clausal co-ordination and phrasal co-ordination are
characteristic of “oral” and “literate” discourse, respectively. We might
thus expect expository genres to display more phrasal co-ordination than
non-expository genres in Late Modern English.
To investigate this, Smitterberg (forthcoming) retrieved a random
subset of and in A Corpus of Nineteenth-century English (CONCE),
a one-million-word multi-genre corpus of British English from the
1800s. We focus on results for Drama and Science, one stereotypically
“oral” and one stereotypically “literate” genre. The analysis is based
on 400 randomly selected instances per genre from each of the two
periods included, viz. 1800–1830 and 1870–1900. The conjoins of
each instance of and retrieved were classified according to their syntactic make-up. Three categories are recognized in this classification.
Smitterberg refers to the first category, which is exemplified in (1) and
(2), as “super-phrasal”. (Conjoins are given in bold face in numbered
examples; speaker identifications and stage directions are enclosed in
square brackets and dollar signs in the corpus.)
(1) [$Blunt.$] No. I came too late, and I am sorry for it: […]
(CONCE, Drama, Holcroft, 1800–1830, p. 25)
(2) Soon after the application of the heat, a dark line, thin and
delicate as a spider’s thread, was observed to be slowly creeping down each of the bright sodium lines and exactly occupying the centre of each.
(CONCE, Science, Lockyer, 1870–1900, p. 128)
In order to be included in this category, the conjoins have to meet
two criteria. First, both have to contain more material than one syntactic phrase. Second, both conjoins have to contain at least part of a verb
phrase. For the second category, the traditional term “phrasal” is used;
it is illustrated in (3) and (4).
(3) [$MISS T.$] Have those base and servile things called settlements been satisfactorily adjusted? [$eating$]
(CONCE, Drama, Gilbert, 1870–1900, p. 25)
(4) It would undoubtedly be advantageous to the capitalists of
England, and to the consumers in both countries, that under
English Genres in Diachronic Corpus Linguistics 127
such circumstances, the wine and the cloth should both be
made in Portugal, […]
(CONCE, Science, Ricardo, 1800–1830, pp. 160–161)
In phrasal co-ordination, the conjoins are either on or below the level
of a syntactic phrase and do not consist of full verb phrases. Finally,
Smitterberg (forthcoming) recognizes an indeterminate category, to
which examples (5) and (6) belong. This category contains examples
whose conjoins did not meet all criteria for either of the other categories.
(5) [$Mait.$] A vindictive temper is the master passion that
degrades and ruins the peace of Mr. Anson: […]
(CONCE, Drama, Holcroft, 1800–1830, p. 34)
(6) [$Admiral. [More sandwich.]$] If ever there was a jewel
of a wife it’s Lady Darby. God bless her! Here’s her health.
[$Drinks.$] I don’t deserve her. She’s too good for me. When
I remember what an unfaithful rascal I’ve been, and the lies
I’ve had to tell – the awful lies – [$Is overcome with painful
reminiscences and weeps.$]
(CONCE, Drama, Jones, 1870–1900, p. 50)
In (5), both conjoins contain one verb phrase but no other material;
and in (6), the first conjoin consists of a clause and the second of a noun
The manual analysis of 400 instances from each period/genre
sample led to the exclusion of 116 instances, or c. 7% of the data.
These include instances of and in stage directions, chapter headings,
and numerical expressions such as four and a half; a small number of
Figure 5. Conjoins of and in Drama.
128 From Clerks to Corpora
Figure 6. Conjoins of and in Science.
instances that resisted classification were also excluded. The total number of instances of and included is 347 from Drama, period 1, 378 from
Drama, period 3, 376 from Science, period 1, and 383 from Science,
period 3. The period/genre distribution is given in Figures 5 and 6.
As Figures 5 and 6 show, the proportion of super-phrasal co-ordination
is higher in Drama than in Science; this difference is statistically significant in both periods (for period 1, d.f = 2; χ2 = 28.9; p < 0.001; for
period 3, d.f. = 2; χ2 = 22.1; p < 0.001). This genre difference tallies well
with previous research on spoken and written communication. As mentioned above, Biber’s (2003) factor analysis of present-day academic
English demonstrates that clausal co-ordination and phrasal co-ordination are characteristic of spoken and written English, respectively.
As Drama comes out as a stereotypically “oral” genre in Biber and
Finegan’s (1997) diachronic factor score analysis of Modern English,
the predominance of super-phrasal co-ordination in this genre is to be
expected. Likewise, the high percentage of phrasal co-ordination in the
“literate” Science genre is in accordance with what would be expected
against the background of Biber’s (2003) results.
In contrast to the cross-genre differences, there are no clear indications of change across time in either genre. While there are tendencies
towards, for instance, more phrasal co-ordination in Drama and less
indeterminate co-ordination in Science, these period differences do not
reach statistical significance (for Drama, d.f. = 2; χ2 = 4.72; p = 0.095;
for Science, d.f. = 2; χ2 = 5.06; p = 0.080). Neither the consistent trend
towards more literate styles in specialist expository writing nor the
English Genres in Diachronic Corpus Linguistics 129
reversal towards more oral styles in a popular written genre such as
Drama noted by Biber and Finegan (1997: 272–273) is thus mirrored
in our results. Regarding the results for Science, Geisler (2002) found
that this genre did not change significantly on any dimension in his
factor score analysis of CONCE; it is possible that differences in textual selection and/or the time span covered underlie these differences
between results found for scientific English in CONCE and ARCHER
(cf. Biber & Finegan 1997).
The genre perspective comes across as crucial in two respects in our
case study on co-ordination. First, including both “oral” and “literate” writing is necessary to obtain a full picture of co-ordination in
nineteenth-century English as a whole; neither genre included could
have stood proxy for the entire language variety. Secondly, it is well
known that, while language change presupposes language variation, not
all variation leads to change; a genre perspective can help to uncover
such cases of stable variation in language.
The two approaches exemplified here both yield important information on the development of the English language and on how this
development is connected to the genre parameter. For instance, detailed
studies of single features can show what features are worth including
in multi-feature analyses; at the same time, the overall view afforded
by multi-dimensional studies provide single-feature analyses with
an overall theoretical framework (see Biber 1988: 62–63 for further
4. Concluding Remarks
As we hope to have shown in this chapter, genre is an indispensable
parameter in historical linguistics. If the language of a period is treated
as a monolithic phenomenon, patterns such as genre drift and genre
differentiation may go unnoticed. Similarly, factors underlying the distribution of individual linguistic features may escape notice if genre
differences in their occurrence are not taken into account.
However, as we have also demonstrated, the limitations imposed
by historical material require that attention be paid to methodological
issues. Above all, it is crucial that researchers account for their definitions of the genre concept and their criteria for identifying and classifying genres (e.g. the relative importance of linguistic and extralinguistic
criteria). Attention must also be paid to the socio-historical contexts of
the genres sampled; for instance, factors such as literacy and level of
130 From Clerks to Corpora
education will have affected the size of the audience of a written genre
at any point in the history of English. Conversely, the author perspective is important especially when dealing with genres that have few textual witnesses. A text such as the Orrmulum, for instance, can be seen
as representing the genre of homily collections in the twelfth century.
However, it can also be seen as a concrete representation of the author’s
own idiolect as against the more abstract genre level.
As we have discussed above, the desiderata of comparability and
representativity may clash when corpora are compiled from a genre
perspective. Genres have emerged and died out through the recorded
history of English, causing gaps in genre representation across time.
There are of course also plenty of examples of genre continuity; but
even in such cases, attention must be paid to genre evolution and the
extent to which what is nominally the “same” genre can be said to
occupy the same communicative space in an ever-changing society.
The two case studies we have reported on demonstrate that the genre
concept is of central importance in diachronic corpus linguistics. This
is of course true of multi-feature studies, which are typically based on
contrasting linguistic co-occurrence patterns in several different genres.
However, even studies of single linguistic features such as and and its
conjoins often require a genre perspective, as different patterns may
manifest themselves in different genres and genre groups.
As regards future developments, the genre concept is likely to continue to grow in significance in historical corpus linguistics. Above all,
genres are likely to become central parameters in a wide range of subdisciplines, from pragmatics and discourse studies to grammar. (See, for
instance, Walker 2007 for a study where the genre concept is crucial for
an understanding of pragmatic variation and change in pronoun usage.)
Regarding methodology, there is a need for new approaches to studying
past language forms on their own terms. In terms of resources, there
is plenty to do in, for instance, providing faithful linguistic editions of
early manuscripts that can be used as the basis for new historical corpora. We also need to pay better attention to poorly represented speaker
groups, for instance the language of women and lower socio-economic
strata. In addition to studying British English, we should also consider
the development of genres in overseas varieties of the language.
English Genres in Diachronic Corpus Linguistics 131
Corpora Referred to
ARCHER = A Representative Corpus of Historical English Registers, originally
compiled under the supervision of D. Biber and E. Finegan (modified and expanded by subsequent members of a consortium of universities). For more
details, see
BLOB-1931 = The BLOB-1931 Corpus, compiled by G. Leech, P. Rayson &
N. Smith. For more details, see
CED = A Corpus of English Dialogues 1560–1760 (2006), compiled by
M. Kytö & J. Culpeper. For more details, see
CONCE = A Corpus of Nineteenth-century English, compiled by M. Kytö &
J. Rudanko. For more details, see Smitterberg (2005).
FLOB = The Freiburg–LOB Corpus of British English (1999), compiled by C.
Mair. For more details, see
Helsinki Corpus = The Helsinki Corpus of English Texts (1991), compiled
by M. Rissanen (Project leader) & M. Kytö (Project secretary); L. KahlasTarkka & M. Kilpiö (Old English); S. Nevanlinna & I. Taavitsainen (Middle
English); T. Nevalainen & H. Raumolin-Brunberg (Early Modern English).
For more details, see
LOB = The Lancaster-Oslo/Bergen Corpus (1976), compiled by G. Leech,
S. Johansson & K. Hofland. For more details, see
Biber, D. (1988). Variation across Speech and Writing. Cambridge: Cambridge
University Press.
———. (1993). Representativeness in corpus design. Literary and Linguistic
Computing 8:4, 243–257.
———. (2003). Variation among university spoken and written registers:
A new multi-dimensional analysis. P. Leistyna & C. Meyer (eds) Corpus
Analysis: Language Structure and Language Use. Amsterdam & New York:
Rodopi, 47–70.
132 From Clerks to Corpora
Biber, D. & Finegan, E. (1997). Diachronic relations among speech-based and
written registers in English. T. Nevalainen & L. Kahlas-Tarkka (eds) To
Explain the Present: Studies in the Changing English Language in Honour
of Matti Rissanen. Helsinki: Société Néophilologique, 253–275.
Biber, D. & Gray, B. (2011). Grammatical change in the noun phrase: The influence of written language use. English Language and Linguistics 15:2, 223–250.
———. (2013). Being specific about historical change: The influence of
sub-register. Journal of English Linguistics 41:2, 104–134.
Biber, D., Johansson, S., Leech, G., Conrad, S. & Finegan, E. (1999). Longman
Grammar of Spoken and Written English. Harlow: Pearson.
Brownlees, N. (2012). The beginnings of periodical news (1620–1665).
R. Facchinetti, N. Brownlees, B. Bös & U. Fries News as Changing Texts:
Corpora, Methodologies and Analysis. Newcastle upon Tyne: Cambridge
Scholars Publishing, 5–48.
Claridge, G. (2012). Linguistic levels: Styles, registers, genres, text types.
A. Bergs & L.J. Brinton (eds) English Historical Linguistics: An International
Handbook. Vol. 1. Berlin & Boston: Walter de Gruyter, 237–253.
Clarke, S. (2004). Jonathan Strange & Mr Norrell. New York & London:
Culpeper, J. & Kytö, M. (2010). Early Modern English Dialogues. Spoken
Interaction as Writing. Cambridge: Cambridge University Press.
Dahl, F. (1953). A Bibliography of English Corantos and Periodical Newsbooks
1620–1642. Stockholm: Almqvist & Wiksell.
Geisler, G. (2002). Investigating register variation in nineteenth-century
English: A multi-dimensional comparison. R. Reppen, S.M. Fitzmaurice &
D. Biber (eds) Using Corpora to Explore Linguistic Variation. Amsterdam &
Philadelphia: John Benjamins, 249–271.
Hundt, M. & Mair, C. (1999). “Agile” and “uptight” genres: The corpus-based
approach to language change in progress. International Journal of Corpus
Linguistics 4:2, 221–242.
Kohnen, T. (2001). Text types as catalysts for language change: The example
of the adverbial first participle construction. H-J. Diller & M. Görlach
(eds) Towards a History of English as a History of Genres. Heidelberg:
Universitätsverlag C. Winter, 111–124.
———. (2007). From Helsinki through the centuries: The design and development of English diachronic corpora. P. Pahta, I. Taavitsainen, T. Nevalainen &
J. Tyrkkö (eds) Towards Multimedia in Corpus Studies. Helsinki: Research
Unit for Variation, Contacts and Change in English (VARIENG), University
English Genres in Diachronic Corpus Linguistics 133
of Helsinki. Available at
kohnen/; last accessed 13 February 2013.
Kytö, M. & Smitterberg, E. (Forthcoming). Diachronic registers. D. Biber &
R. Reppen (eds) The Cambridge Handbook of English Corpus Linguistics.
Cambridge: Cambridge University Press.
Lange, C. (2012). Text types, language change, and historical corpus linguistics. C. Lange, B. Weber & G. Wolf (eds) Communicative Spaces: Variation,
Contact, and Change. Papers in Honour of Ursula Schaefer. Frankfurt am
Main etc.: Peter Lang, 401–416.
Leech, G. (2007). New resources, or just better old ones? The holy grail of
representativeness. M. Hundt, N. Nesselhauf & C. Biewer (eds) Corpus
Linguistics and the Web. Amsterdam: Rodopi, 133–149.
Quirk, R., Greenbaum, S., Leech, G. & Svartvik, J. (1985). A Comprehensive
Grammar of the English Language. London & New York: Longman.
Smitterberg, E. (2005). The Progressive in 19th-century English: A Process of
Integration. Amsterdam & New York: Rodopi.
———. (Forthcoming). Colloquialization in Nineteenth-century English.
Walker, T. (2007). Thou and You in Early Modern English Dialogues: Trials,
Depositions, and Drama Comedy. Amsterdam & Philadelphia: John
8 Here is an Old Mastiffe Bitch Ø Stands
Barking at Mee: Zero Subject Relativizers in
Early Modern English (T)here-Constructions1
Gunnel Tottie
The University of Zurich
Christine Johansson
Uppsala University
1. Introduction
There are two kinds of zero relativizers – those where the gap functions
as direct object, as in (1), and those where the gap functions as subject
in the relative clause, as in (2), henceforth ZSR. 2
(1) I have nothing Ø I can call my own…. (Thomas Killigrew,
Chit-Chat, 1719)
(2) There were Seven Horses Ø came in. (The Tryal of Ambrose
Rookwood, 1696)
Our purpose here is to throw light on the use of one of the major
types, viz. ZSRs in presentative constructions introduced by there or
here, as in (2), in Early Modern English, and to attempt to establish
factors that influence the choice of surface or zero variant. The gap will
be marked Ø throughout.
1We are indebted to participants at the symposium in honor of Nils-Lennart
Johannesson in February 2013 and at the ICAME 34 conference, Santiago de
Compostela, 22–26 May, 2013, for constructive comments, especially Holger Diessel,
Uwe Vosberg, Lilo Moessner and Yoko Iyeiri. We are also indebted to Hans-Martin
Lehmann and Sebastian Hoffmann for helpful comments and technical support.
2 Terminology varies here; the term bare relatives is used by Huddleston & Pullum
(2002: 134, 155). There are also scholars who challenge the classification of
zero subject relativizers as relativizers, (e.g. Erdmann 1980, Nagucka 1980 and
Lambrecht 1988). We will not enter that discussion here but will use traditional
terminology and consider zero relativizers as variants of one relativizer variable.
How to cite this book chapter:
Tottie, G. and Johansson, C. 2015. Here is an Old Mastiffe Bitch Ø Stands Barking
at Mee: Zero Subject Relativizers in Early Modern English (T)here-Constructions. In:
Shaw, P., Erman, B., Melchers, G. and Sundkvist, P. and Sundkvist, P. (eds) From Clerks to
Corpora: essays on the English language yesterday and today. Pp. 135–153. Stockholm:
Stockholm University Press. DOI: License: CC-BY.
136 From Clerks to Corpora
2. ZSRs in Present-Day English
ZSRs get scant attention in the major grammars of contemporary
English. Quirk et al. (1985: 1250) consider them to be “of doubtful
acceptability” or “slovenly,” Biber et al. (1999: 619) assert that they
occur in “conversational varieties” or “marginally non-standard usage”
and Huddleston & Pullum (2002:1055) declare that “they fall at the
boundary between very informal and non-standard.” However, ZSRs are
characteristic of many British and American English dialects, as is clear
from e.g. Ihalainen (1980), Hackenberg (1981), van den Eynden (1993:
160) and Kortmann & Schneider (2004).
Moreover, according to a large-scale quantitative and accountable
study by Lehmann (2002), based on the spoken demographic component of the British National Corpus (4.2 million words) and on the
Longman Spoken American Corpus (5 million words), ZSRs account
for 13% of all subject relativizers in British Present-Day English (PDE)
but for less, 2.5%, in American PDE.3 Following Shnukal (1981),
Lehmann lists the major types of constructions in PDE as those shown
in (3)–(6), adding one residual mixed category of “other” types, listed
as (7) (the examples given below are from various sources):
(3) Existential there-constructions : There was a farmer Ø had a
dog (Lambrecht 1988)
(4) Clefts: It was he Ø took you out. (Erdmann 1980)
(5) Be-constructions: …they were people Ø got in there for the
summer…(Lehmann 2002: 171)
(6) Have-constructions: We had a client Ø came in about two
weeks ago (Erdmann 1980)
(7) Other types: I knew a girl Ø worked in an office…(Shnukal
Lehmann found the distribution of the five ZSR types shown in Table 1.
In both British and American English, existential there-constructions
3Accountable studies account for all the constructions where alternative variants
could have been used, in this case the relativizers that or which/who as in There is
a man who/that/Ø wants to see you.
Here is an Old Mastiffe Bitch Ø Stands Barking at Mee 137
Table 1. The use of ZSR constructions in PDE. Based on Lehmann (2002:172,
Table 2).
British English
(1581 subject rel.
American English
(3741) subject rel.
Existential there-constructions 126
Other types
Totals zero subject relatives
3. ZSRs in some previous studies
Zero relatives have received a fair amount of coverage in diachronic
studies of English, but ZSRs have been less well described, although
they may be the older type. Both existed in OE, according to several
writers (see e.g. van der Auwera 1986) but zero objects increased in
frequency and are now very common (see e.g. Johansson 2012; Tottie
Previous studies have either concentrated on the origin of the ZSR
and how it decreased from the Old English period and onwards (see
e.g. Erdmann 1980; Nagucka 1980; van der Auwera 1984), or have
described its use in written data and in other time periods – see e.g.
Dekeyser (1984) and Rydén (1966).
Ukaji (2003) is based on all of Shakespeare’s plays, three plays by
Ben Jonson, and Nashe’s The Unfortunate Traveller. In his material,
consisting of 303 examples, 40% of all ZSR instances are either hereor there-constructions, 14% are it-clefts, 10% are have-sentences, and
23% various other types, thus a distribution not unlike that found by
Lehmann for Present-Day English.
Although some earlier historical grammarians have quantified their
material, none seem to have carried out accountable studies. Our study
appears to be the first diachronic study of ZSRs that is both quantified
and accountable. We also aim to pinpoint factors that trigger or constrain
the choice of ZSRs, something that has not previously been attempted
138 From Clerks to Corpora
beyond the basic establishment of the major contexts favoring ZSR, i.e.
existential (t/here-), cleft, have- and to some extent be-constructions.
4. Our speech-related data
As the zero relativizer has been described as colloquial and “spoken”
in character (see Erdmann 1980; Nagucka 1980; Romaine 1981:94;
Dekeyser 1986), we will focus on the use of ZSRs in speech-related
texts: Trials and Comedy, which have been deemed the most speech-like
(cf. Culpeper and Kytö 2000:186–193).
We studied transcripts of trials and plays from the category Drama/
Comedy in the computerized Corpus of English Dialogues, 1560–1760
(CED) (for a full description of this corpus, see Kytö & Walker 2006).
We will use the term Comedy for simplicity. The material is presented
in Table 2:
Table 2. Our CED sub-corpus consisting of Trials and Comedy.
CED periods
Totals N words
1. 1560–1599
2. 1600–1639
3. 1640–1679
4. 1680–1719
5. 1720–1760
As appears from Table 2, our sub-corpus amounted to just over half
a million words from five successive 40-year time periods between 1560
and 1760. It is skewed in several respects: The first two periods contain
fewer words than the others because of the scarcity of trial transcripts
from the 16th and early 17th centuries, but in the following three time
periods, Trials account for more material than Comedy, in fact twice as
many words in period 4, 1680–1719. This must be kept in mind when
the results are presented.
5. Method and first results
We first did a simple lexical search for there, here, it is, and it was
(it-clefts); we included here because it had been bundled with there-ex-
Here is an Old Mastiffe Bitch Ø Stands Barking at Mee 139
amples by earlier writers (e.g. Jespersen 1927: 147; Rydén 1966: 267–
268). We then read the whole corpus to weed out irrelevant material
and to find additional types less amenable to lexical searches. We ended
up with 210 relevant examples, i.e. relativizers in subject function in
either (t)here-constructions or clefts, with explicit or zero realizations.
We also found 17 miscellaneous examples where only zero realizations
were collected because of problems of finding alternants with surface
subject relativizers, as in example (8).
180 159 160 140 120 100 51 80 60 17 40 20 0 (t)here Cle3 Other Figure 1. The distribution of constructions capable of taking zero relative
­constructions in Trials and Comedy in CED.
(8) [Daffodil] Not do it! [hops] Why, I’ll get a Chelsea Pensioner
Ø shall do it in an Hour, with his wooden Leg.
(David Garrick, The Male-Coquette, 1757)
The different types of constructions were distributed as shown in
Figure 1:
The largest category was thus (t)here-constructions – 159/210
or 76% of the instances where an accountable analysis was possible, and thus the type most amenable to a quantitative analysis.
(T)here-constructions will therefore be the focus of our investigation.
Henceforth we will only distinguish between zero or non-zero
constructions; the use of that- versus wh-relatives has been adequately treated in earlier work, such as e.g Johansson (2012), and will
therefore not be discussed here. The distribution of zero relatives in
(t)here-constructions is shown in Table 3.
140 From Clerks to Corpora
Table 3. The distribution of zero subject relatives in (t)here-constructions in Trials
and Comedy in CED.
subj rels
subj rels
subj rels
Table 3 shows the total number and distribution of relevant (t)here
constructions in the entire sub-corpus. Trials are accounted for in columns II–IV, Comedy in columns V–VII, and the entire corpus in columns
VIII–X. The total proportion of ZSRs was 50/159 or 31%, but notice
the difference between the two genres: ZSR-constructions amount to
20% in Trials and 50% in Comedy, as appears from columns IV and
VII. This discrepancy is itself an interesting finding that will be discussed in section 7.4 below.
The frequency of 50 ZSRs in there-constructions in our 524,430-word
sample corresponds to 95 instances per million words (pmw). If we
exclude 21 instances of here and consider only there-constructions,
we get 29 ZSRs per 524,430 words, or 55 pmw. Lehmann’s figure of
126 ZSRs in there-constructions in the 4.2 million words of British
PDE corresponds to 30 instances pmw, and his figure for American
English of 27 ZSRs in there-constructions in 5 million words corresponds to 5.4 per million words – we can thus see a sharp decline in the
use of ZSRs in these constructions over time. This corroborates findings
reported in earlier works (e.g. Erdmann 1980; Nagucka 1980).
6. Factors determining the choice of ZSRs
After gathering the relevant data, we checked a large number of factors
for their possible influence on the choice of ZSR. We first cast our net
wide, including all those listed below, extra- and intra-linguistic. Only
Here is an Old Mastiffe Bitch Ø Stands Barking at Mee 141
those marked in boldface yielded interesting results and will be discussed in detail below.
• Sociolinguistic factors: socio-economic class and gender of
• Time period
• Text type/genre – Trial or Comedy
Here vs. there – constructions
Distance between head of antecedent and relativizer slot
Antecedent number (There is someone…/there are two men…)
Antecedent head: Indefinite noun, definite noun, pronoun, name
Animacy of antecedent head
Tense in matrix clause (There is, there was…)
Verb form in relative clause – Finite, Modal, or other Auxiliary
Polarity – positive or negative matrix clause
Question or statement in matrix clause
Prose or meter – convenience of zero in blank verse could be a
For coding, we used Goldvarb X, which is useful for establishing
basic statistics even when the material inspected is too small to permit
a regression analysis, as in the present case.
7. Extralinguistic factors
We begin with socio-economic class as a possible factor favoring ZSR,
as the construction has been labeled “non-standard” by grammarians.
7.1 Socioeconomic class
A fine-grained analysis was not possible, but we were able to categorize
most speakers as belonging to either a higher class, which we called
U(pper class), consisting of nobility, clergymen, judges and attorneys,
and a class comprising other speakers, Non-U(pper class), made up of
servants, cooks, nurses and most defendants. Three individuals could
not be classified, but that still left a total of 156 speakers. As shown in
Figure 2, there was no difference between U- and Non-U speakers as
measured by the percentages of ZSRs – the proportion of ZSR use was
142 From Clerks to Corpora
80% 72% 68% 70% 60% 50% 40% 32% 30% 28% ZERO that, wh-­‐ 20% 10% 0% U N=99 Non-­‐U N=57 Figure 2. The distribution of ZSR and surface relativizers in the CED subsample.
roughly the same in the two samples, 32% for U speakers, and 28% for
Non-U speakers.
The small difference between U and Non-U speakers was not significant. It thus appears that in Early Modern English, the use of ZSRs
was not a marker of socio-economic class, but that more educated
speakers were as likely to use these constructions as less educated ones.
Examples (9) and (10) are from Trials and illustrate there + ZSR with
U and Non-U speakers, respectively.
(9) [Parson] There was a Gentleman Ø fetch’d Me and the Clerk
from the Fleet.
(Tryals of Haagen Swendsen, 1702)
(10) [Mr. Baley] My Lord, there has been at least 500 people Ø
have viewed her.
(Tryal of Mary Moders, 1663)
7.2 Speaker gender
Speaker gender could not be analyzed in the same way as socio-economic
class, as there were only 11 observations based on women speakers in
the whole sub-corpus, i.e. 7% of the data, as appears from Table 4.
But this in itself is an interesting fact that prompted us to investigate
whether speech by women actually accounted for only 7% of the total
CED sub-corpus.
Here is an Old Mastiffe Bitch Ø Stands Barking at Mee 143
Table 4. The distribution of zero subject relatives in (t)here-constructions in
Trials and Comedy in CED.
Male speakers
Female speakers
All speakers
148 (93%)
11 (7%)
159 (100%)
A representative sample consisting of 20% of our sub-corpus shows
that women actually participated very little – the overall proportion
of women’s speech is about 16% in the whole material.4 There is a big
difference between Trials and Comedy, however: in Comedy, women
account for about 28% of all speech, and in Trials for only 4%. Three
texts are exceptional in having high proportions of female speech,
Farquhar’s The Beaux Stratagem, the Tryal of Stephen Colledge and the
Tryals of Haagen Swendsen, all from Period 4. A couple of illustrative
examples are (11) and (12) from the Haagen Swendsen trials, where
two women, Mrs. Rawlins and Mrs. Busby, are very talkative:
(11) [Mrs. Busby] I’ll tell you Sir, if you please, I did not know what
I was Arrested for, it might be Murder or Treason for ought
I knew, there was a little Boy by, Ø said Madam I know Mr
Unkle. (Tryals of Haagen Swendsen, 1702)
(12) [Mrs. Rawlins] It was Saturday morning before I was releas’d,
there was some of my Friends Ø came to the place where I was.
(Tryals of Haagen Swendsen, 1702)
Finally, because of the paucity of examples, the only conclusion to
be drawn considering the importance of gender is that women seem to
have used fewer subject relativizers than men overall, or 7% of the total
number of relativizers produced in 16% of the total number of words,
leaving plenty of room for speculation.
4 A random sample consisting of 20% of each text was studied.
144 From Clerks to Corpora
70% 61% 60% 50% 40% 30% 20% 10% 0% 20% 46% 14% 21% ZSR 1560-­‐1599 1600-­‐1639 1640-­‐1679 1680-­‐1719 1720-­‐1760 n=10 n=23 n=41 n=43 n=42 Figure 3. The distribution of ZSRs in there-constructions: Aggregate data for
Trials and Comedy in Periods 1–5.
7.3 Change over time in Early Modern English
We have seen that ZSRs have become less frequent over time, and we
therefore investigated the distribution of ZSR and surface relativizers
across the time period covered by CED in Figure 3.
The data presented in Figure 3 must be analyzed with caution. Recall
that the samples from the five different time periods were of very different sizes, as shown in Table 1. The low number of examples from
Period 1, only 10, makes that data highly unreliable. Period 2 yields 23
examples even though the sample size is smaller, but only in Periods 3–5
do we have fairly large numbers of examples, 41, 43 and 42, respectively. Starting with period 3 (1640–1679), we see a downward trend
for ZSRs; the difference between the proportions of ZSRs in Period
2 and Period 5 is significant at 0.0036 (chi-square 8.46, 1 d.f.), and
between Period 3 and 5 at 0.0302 (chi-square 4.7, 1 d.f.).
7.4 Genre: Trials and Comedy
Moreover, the skewness of our sub-corpus makes it necessary to account
for the distribution of ZSRs separately in Trials and Comedy. This is
shown in Figure 4. For Period 2 the proportion 17% ZSR in Trials is
based on a single instance out of six there-constructions and is thus
highly uncertain. The 76% ZSR from Comedy is slightly better, but it
is based on only 13 out of 17 observations. But with more data from
Here is an Old Mastiffe Bitch Ø Stands Barking at Mee 145
80% 76% 70% 60% 50% 40% 30% 20% 10% 0% 60% 42% 40% 17% 29% 11% Trials 33% Comedy 10% 0% Per. 1 Per. 2 Per. 3 Per. 4 Per. 5 Figure 4. The distribution of ZSR over time in (t)here-constructions in Trials
and Comedy.
periods 3–5, we see that the downward trends in Trials and Comedy
parallel each other, and that the general tendency for ZSRs to decrease
in frequency over time holds for both genres in Periods 3–5.
8. Intralinguistic factors
As also shown in Table 2 above, Comedy has consistently higher frequencies of ZSRs than Trials. The question must now be why there is a
difference in ZSR frequency between the two genres.
8.1 Here and there
We first checked the possibility of influence of meter, as blank verse
might have required either a surface or a zero realization to make lines
scan. That hypothesis could quickly be discarded, as there was little
blank verse (or indeed other meter) in our sample, only five instances,
with two zero and three surface relatives, respectively.5 Instead, we need
a two-step explanation:
5 Interestingly, the vast majority of Ukaji’s ZSR examples are from blank verse: only
60/303, or about 20%, are from prose. An accountable study would be necessary
to find out what the proportions are in prose and poetry, respectively.
146 From Clerks to Corpora
90% 77% 80% 70% 60% 60% 50% 40% 40% Zero that, wh-­‐ 23% 30% 20% 10% 0% here, N=35 there, N=124 Figure 5. The use of ZSR in here- and there-constructions.
90% 80% 70% 81% 67% 64% 62% 60% 50% 40% 33% 36% 38% ZERO 30% 19% 20% that-­‐ wh-­‐ 10% 0% here n=27 there n=33 COMEDY n=60 here n=8 there n=91 TRIALS n=99 Figure 6. Here- and there-constructions in Comedy and Trials.
Firstly, here-constructions are more apt to take ZSR than there-constructions, as shown in Figure 5 and secondly, here-constructions are
more frequent in Comedy than in Trials, as shown in Figure 6.
Figure 5 shows that here-constructions have 60% ZSRs, but
there-constructions only 23%; the difference is highly significant
(p<0.0001, chi-square 15.32, 1 d.f.)
Furthermore, most here-constructions occur in Comedy, as shown
in Figure 6: Here accounts for 27/33 instances, or 45%, in Comedy,
but only 8/99 or 8% in Trials. This difference is highly significant
(p<0.0001, chi-square 27, 1 d.f.).
Here is an Old Mastiffe Bitch Ø Stands Barking at Mee 147
The fact that here-constructions, with their high proportion of ZSR,
abound in Comedy thus goes a long way to explain why the ratio of
ZSRs is higher in that text type.
The next question must be why here-constructions are so frequent in
Comedy. The answer is supplied by the instances presented in (13) – (16):6
(13) [Lemot]…here is one Ø had hanged himselfe for loue …
(George Chapman, An Humorous Dayes Myrth, 1599)
(14) [Medley] Dorimant! you are luckily come to justify Your self
— here’s Lady — [Bellinda] Ø Has a word or two to say to
you from a Disconsolate person.
(George Etheredge, The Man of Mode, 1676)
(15) [Dash (servant)] Here are Gentlemen in hast Ø would speake
with you.
(Lording Barrey, Ram-Alley, 1611)
(16) [Daffodil] My Lord Marquis, here is a Letter Ø has started
Game for you already…
(David Garrick, The Male-Coquette, 1757)
Examples (13)–(16) show that here-constructions fulfill an important function in plays, introducing new participants or objects appearing on stage. They have mostly a locative meaning, or very occasionally,
a temporal one, as in (17):
(17) [Galleypot] Here’s a whole Morning Ø has been thrown
(James Miller, The Mother-in-Law, 1734)
It is clear that here and there, although they have been lumped
together by earlier writers, and although they are both presentatives, have
different discourse functions, with here almost invariably having a locative meaning, and there an existential one. Checking the there-instances
closely, we only found five with a locative meaning, as in (18), but (19)
is typical:
(18) [Winwife] Will you see sport? looke, there’s a fellow Ø gathers vp to him, marke.
(Ben Jonson, Bartholomew Fayre, 1631)
6 There were no instances in Period 4.
148 From Clerks to Corpora
100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 60% 40% 82% 18% 74% 26% 100% 0% Zero that, wh-­‐ Figure 7. Distance between antecedent head and relativizer slot.
(19) [William Small-shanke] There be so many rascals, and tall
yeomen Ø VVould hang vpon me for their maintenance…
(Lording Barrey, Ram-Alley, 1611)
8.2 Distance between antecedent head and relativizer
The distance between the antecedent head and the relativizer slot was
also a factor influencing the choice of ZSR.7 Compare e.g. (20), where
there are no intervening words between Day and goes and (21), where
there are five words between Gentlemen and that:
(20) [Scrub] There’s not a Day Ø goes over his Head without
Dinner or Supper in this House.
(George Farquar, The Beaux Stratagem, 1707)
(21) [Cary] …there are some Gentlemen at the Queens-Head at
Bow that have sent me with a Letter to you…
(Tryals of Robert Green, 1678)
7 Ukaji (2003:257f) discusses the adjacency requirement, i.e. the fact that a “relative
pronoun is as a rule placed immediately after its antecedent NP,” and he also discusses exceptions from it in his material.
Here is an Old Mastiffe Bitch Ø Stands Barking at Mee 149
We coded instances into four categories based on the number of
words intervening between the antecedent head and the relativizer slot,
No distance:
Short distance: 1–2 words
Medium distance: 3–4 words
Long distance: ≥ 5 words
Table 5. Proportions of ZSR according to distance ­antecedent head – relativizer slot.
No distance (N = 100)
1–4 words distance (N = 45)
≥ 5 words distance (N = 14)
(N = 159)
The distribution of the categories is shown in Figure 7.
We see that no distance is by far the most frequent option, and that
in those cases, the proportion of ZSR is 40%. With ≥5 words or more
between the antecedent head and the relativizer slot, ZSR goes down
to 0% in our material.8 There was little difference between the results
for examples with Short and Medium distance between antecedent
head and the relativizer slot (4/22 or 18% ZSR vs. 6/23 or 26% ZSR);
if those categories are conflated to one, we arrive at the distribution
shown in Table 5. The distance between antecedent Head and relativizer slot is thus an important factor for the choice of the zero option.
9. Summary and discussion
The purpose of our paper was to describe the use of zero subject relativizers (ZSRs) in Early Modern English speech-related trials and
8 ZSR is not impossible in such constructions: Cf. Shakespeare, A Winter’s Tale (I,
2, 190):
And many a man there is, even at this present,
Now while I speak this, Ø holds his wife by the arm,
That little thinks she has been sluiced in’s absence,
And his pond fish’d by his next neighbour…
150 From Clerks to Corpora
drama (comedy) in 524,430 words taken from the Corpus of English
Dialogues, CED, covering the period from 1560 to 1760. Because of the
paucity of other types (it-clefts and a few others), we focused on presentative constructions with there and here. We performed an accountable
variationist study, i.e. we examined all contexts where subject relatives
occurred, not just those showing the zero variant. We found that 50 out
of a total of 159 cases – 31% – were ZSRs, thus a much higher proportion than those found by Lehmann (2002) for Present-Day British and
American English (13% and 2.5%, respectively).
We also checked our material for factors that might favor or disfavor
the choice of the zero variant. In contrast to Present-Day English, we
found that socio-economic class did not influence the choice of variant
as in Present-Day English, where the use of ZSRs has been described
as informal or non-standard (Huddleston & Pullum 2002: 1055; Biber
et al 1999: 619). In the EModE period, however, ZSRs are used as often
by Upper as by Non-Upper class speakers. Influence of gender could
not be determined because of the low proportion of female speakers.
We had also suspected that meter would play a part in favoring ZSR,
but we found only a few examples of verse in our material.9
We found four factors with a bearing on the choice of relativizer, viz.
Time period
Text type/genre – Trial or Comedy
Here vs. there-constructions
The distance between antecedent Head and relativizer
Our data substantiated earlier findings about the downward trend of
ZSRs over time, from 61% in Period 2 (1600-1639) to 21% in Period 5
(1720–1760), with some fluctuations and reservations for low data in
earlier periods.
Text type was another decisive factor: Comedy has consistently
higher frequencies of ZSRs than Trials, 50% vs 20% overall, with some
fluctuations over time. The reason for this must be that it is in this text
category that most here-constructions are found, and they are more
apt to take ZSRs than there-constructions. Here-constructions serve a
particular function in plays, to introduce new characters or objects, as
in Here is an old mastiffe bitch Ø stands barking at mee. They thus usually have a locative meaning, as opposed to there-constructions, which
9However, as already pointed out above, among the 303 examples cited by Ukaji
(2003) 253 were in verse; the factor obviously merits further study.
Here is an Old Mastiffe Bitch Ø Stands Barking at Mee 151
are almost invariably existential. This in turn makes the previous joint
categorization of here and there-constructions as one type questionable.
The distance between antecedent Head and relativizer slot turned
out to be an important factor. First of all, most subject relative clauses
with adjacent antecedent heads and relativizer slot, i.e. no distance, is
by far the most frequent type: 100/159 instances, or 63% are of this
type. Moreover, among those sentences the proportion of ZSR is 40%.
With a 1–4 word distance, the ZSR frequency goes down to 22%, and
with ≥5 words or more between the antecedent head and the relativizer slot, ZSRs are non-existent in our data. This is in accordance with
the complexity principle postulated by Rohdenburg (1996: 151): “In
the case of more or less explicit grammatical options the more explicit
one(s) will tend to be favored in cognitively more complex environments.” In relative clauses, the more explicit wh-forms or that are thus
preferred when elements intervene between antecedent head and relative clause.
As stated earlier, we have focused on presentative constructions
with (t)here in our paper since this is where ZSRs are most frequent in
EModE. In PDE, they occur almost exclusively in these constructions
(see e.g. Quirk et al 1985: 1406–1407). According to van der Auwera
(1984), ZSRs found in OE and ME began to disappear because English
at later stages required explicit subjects, which had not been necessary
as long as verbs were marked for person. ZSRs survived in pragmatically focused constructions, i.e. the presentative constructions, since
there is less need for an explicit subject in these constructions (see van
der Auwera 1984; Dekeyser 1986).
Scholars have offered different explanations why presentative
there-constructions can occur without a subject. Lambrecht (1988)
claims that they have to be without a subject in order to express the
information as one grammatical unit, which is probably how they
occur in discourse (presentative amalgam constructions, Lambrecht
1988: 336). Diessel (2004) sees ZSRs in there-constructions and in
other contexts as precursors of relative clauses with surface relativizers
in L1 acquisition, which appear with increasing frequency as children
are exposed to more adult speech and become literate. Literacy also
seems to have a great deal to do with the fact that ZSRs are found in
Present-day English dialects without written records and in substandard varieties used by speakers with little knowledge or use of written language. In the standard language, it is possible that ZSRs and
there are learned as one construction or chunk and that ZSRs survive
152 From Clerks to Corpora
in impromptu speech because of the frequency of there-constructions of
which they are a part (cf. Bybee 2010: 156, 159).
The explanations sketched above as to why ZSRs have survived in
there-constructions are of course not mutually exclusive; rather they
support each other. We will explore these explanations and others in
more detail in further work.
Biber, D., Johansson, S., Leech, G., Conrad, S., & Finegan, E. (1999). Longman
Grammar of Spoken and Written English. Harlow: Longman.
Bybee, J. (2010). Language, Usage and Cognition. Cambridge: Cambridge
University Press.
Culpeper, J., & Kytö, M. (2010). Early Modern English Dialogues. Cambridge:
Cambridge University Press.
Dekeyser, X. (1984). Relativizers in Early Modern English. A dynamic quantitative Study. J. Fisiak (ed.), Historical Syntax. Berlin: Mouton, 61–87.
Dekeyser, X. (1986). English contact clauses revisited. Folia Linguistica
Historica 7, 107–120.
Diessel, H. (2004). The Acquisition of Complex Sentences. Cambridge:
Cambridge University Press.
Erdmann. P. (1980). On the history of subject contact-clauses in the history of
English. Folia Linguistica Historica 1, 139–170.
Hackenberg, R.H. (1972). Appalachian English: a sociolinguistic study.
Unpublished PhD thesis. Georgetown University.
Huddleston, R. & Pullum, G. K. (2002). The Cambridge Grammar of the
English Language. Cambridge: Cambridge University Press.
Ihalainen, O. (1980). Relative clauses in the dialect of Somerset. Neuphilologische Mitteilungen 81, 187–196.
Jespersen, O. (1927). A Modern English Grammar on Historical Principles.
Part 3. Copenhagen: Munksgaard.
Johansson, C. (2012). Relativization in Early Modern English: Written versus speech-related genres. L. Brinton & A. Bergs (eds) English Historical
Linguistics: An International Handbook, Vol. I. Berlin: Mouton, 776 –790.
Kortmann, B. & Schneider, E. (eds) (2004). A Handbook of Varieties of
English: A Multimedia Reference Tool. Berlin: Mouton.
Kytö, M. & Walker, T. (2006). Guide to A Corpus of English Dialogues
1560–1760. Acta Universitatis Upsaliensis. Studia Anglistica Upsaliensia 130.
Here is an Old Mastiffe Bitch Ø Stands Barking at Mee 153
Lambrecht, K. (1988). There was a farmer had a dog: Syntactic amalgams
revisited. Paper presented at the Berkeley Linguistics Society. Proceedings
of the Fourteenth Annual Meeting, 319–339.
Lehmann, H.-M. (2002). Zero subject relative constructions in American
and British English. P. Peters (ed.), New Frontiers of Corpus Research.
Amsterdam and New York: Rodopi, 163–177.
Nagucka, R. (1980). Grammatical pecularities of the contact-clause in EModE.
Folia Linguistica Historica 1, 171–184.
Quirk, R., Greenbaum, S., Leech, G. & Svartvik, J. (1985). A Comprehensive
Grammar of the English Language. London: Longman.
Rohdenburg, G. (1996). Cognitive complexity and increased grammatical
explicitness in English. Cognitive Linguistics 7:149–182.
Romaine, S. (1981). The relative clause marker in Scots English: Diffusion,
complexity and style as dimensions of syntactic change. Language in Society
9: 221–247.
Rydén, M. (1966). Relative Constructions in Early Sixteenth Century English.
With Special Reference to Sir Thomas Elyot. Acta Universitatis Upsaliensis.
Studia Anglistica Upsaliensia 3.
Shnukal, A. (1981). There’s a lot mightn’t believe this...variable subject relative
pronoun absence in Australian English. D. Sankoff & H. Cedergren (eds)
Variation Omnibus. Edmonton, Alberta: Linguistic Research, 321–328.
Tottie, G. (1997). Relatively speaking. Relative marker usage in the British
National Corpus. T. Nevalainen & L. Kahlas-Tarkka (eds) To Explain the
Present: Studies in the Changing English Language in Honour of Matti
Rissanen. Helsinki: Société Néophilologique, 465–481.
Ukaji, M. (2003). Subject zero relatives in Early Modern English. M. Ike-uchi,
M. Ukaji, & Y. Nishimura (eds) Current Issues in English Linguistics Tokyo:
Kaitakusha, 248–277.
van den Eynden, N. (1993). Syntactic Variation and Unconscious Linguistic
Change. A Study of Adjectival Relative Clauses in the Dialect of Dorset.
Frankfurt a. M.: Peter Lang.
van der Auwera, J. (1984). More on the history of subject contact clauses in
English. Folia Linguistica Historica 5, 170–183.
9 “Norfolk People Know Best”: On the Written
Representation of Accents as Performed and
Perceived by ‘Insiders’ and ‘Outsiders’
Gunnel Melchers
Stockholm University
1. Introduction
The quote in the title of this presentation is taken from the conclusion of Peter Trudgill’s article ‘Dedialectalisation and Norfolk Dialect
Orthography’ (Trudgill 1999b), in which he explains why
… the non-traditional, outsiders’ spelling <bootiful> is objected to
so strongly by the local community. Native dialect-speaking insiders interpret the <oo> spelling as indicating the utterly nonexistent
pronunciation */bu:təfəl/ rather than the correct /bʉ:təfəl/. As usual,
Norfolk people know best. (Trudgill 1999b: 329)
The concept of ‘insider’ vs. ‘outsider’ plays a significant role in the
assessment and discussion throughout my paper, which – based mainly
on nineteenth-century fiction – is concerned with the representation
in writing of regional and social features of language in England with
special reference to accents.
Writers and readers can be insiders as well as outsiders, here simply
defined as members or non-members of the speech community they
represent in writing or interpret in reading. From the writer’s perspective, the language of a written text can be described as ‘intrinsic’ or
‘extrinsic’ to the author (cf. Hickey (2010: 9), who states that “where
the language being represented is extrinsic to the author it may well be
unreliable”, characterized among other things by vacuous re-spellings).
A thorny issue in the analysis of dialect representation in fiction is
the significance of authenticity, which is highlighted in Jane Hodson’s
pioneering publication Dialect in Film and Literature (2014), among
How to cite this book chapter:
Melchers, G. 2015. “Norfolk People Know Best”: On the Written Representation of
Accents as Performed and Perceived by ‘Insiders’ and ‘Outsiders’. In: Shaw, P., Erman,
B., Melchers, G. and Sundkvist, P. (eds) From Clerks to Corpora: essays on the English
language yesterday and today. Pp. 155–176. Stockholm: Stockholm University Press.
DOI: License: CC-BY.
156 From Clerks to Corpora
other things offering critical views on one-sided, detailed linguistic
assessments but also on the generally impressionistic appraisals by literary reviewers of the accuracy of dialect representation, “approving of
those representations which they felt to be authentic, and condemning
those they felt to be inauthentic” (p. 220). While recognizing the importance of Hodson’s subtle critique, the general view taken in this paper is
that authenticity remains an important factor.
The paper is structured as follows: a brief introduction to nonstandard language in writing as a field of study is given, followed by
some observations on the concepts of ‘insider’ and ‘outsider’. Section 4
focuses on the general problem of representing accents by means of the
orthographic system (‘semi-phonetic spelling’) and section 5 provides
a more detailed account of the representation of regional accents in
19th century England. Section 6 summarizes and discusses the efforts
by insiders and outsiders and presents four case studies of 19th century writers who can be said to represent both categories. The paper is
concluded by an attempt to draw some general conclusions from the
somewhat conflicting data and viewpoints accounted for.
2. Some notes on the study of nonstandard language in
Kirk (1999: 45) distinguishes two approaches to the study of nonstandard language in literary texts:
… the dialectological, which uses literary texts as evidence of the
spoken language and considers the significance provided by the use
of the dialect and the nonstandard within the literary work as evidence for the dialect, often historical …. The second approach is
stylistic, which considers how effective or realistic of speech the language in a particular text is and considers the role and effectiveness
of the dialect and nonstandard within the literary work as a whole.
My background may be that of a dialectologist but I recognize the
importance of both the above approaches. Scholars who argue that
authenticity is not a major issue in analysing dialect representation in
fiction emphasize that focus should be on the function of the dialect
within the text. While recognizing the importance of function, the view
taken in this paper is that the importance of function does not exclude
the value of authenticity. In 19th-century texts as well as their adaptations for television, for example, authenticity is generally expected
“Norfolk People Know Best” 157
with regard to physical environment, dress codes, polite behaviour etc.,
and language should be no exception in providing an authentic setting.
Hence the paper is indeed concerned with authenticity, in comparing
features found in the selected texts to authentic data, above all drawn
from Wright’s English Dialect Dictionary (EDD) and the Survey of
English Dialects (SED). There is, admittedly, a certain amount of circularity in the case of EDD in that a great many entries are taken from
literary works but these were carefully selected by its eminent editor,
Joseph Wright. An important token of the value of dialect in writing
as a source of information in historical linguistics is the chapter The
Dialects of England since 1776, (Ihalainen 1994), which is based on
data contained in works by Wright and other 19th century dialectologists as well as fictional writers.
Sweeping statements are often found in literary criticism such as
“Her men and women have characteristic modes of speech. Sometimes
they are easy to recognize, as, for instance, by their dialect, which, incidentally, she used well” (Pollard 1965: 254, writing about Elizabeth
Gaskell). In my view, such observations are of no value, let alone completely misleading and faulty ad hoc views such as Q.D. Leavis’ claim
that in Dickens’ representations of East Anglian accents the sounds
are not represented (“…a matter of vocabulary and grammar only”)
(cf. Poussa 1999: 34). Opinions vary on this issue, however, as briefly
discussed in the conclusion of this paper.
A distinction is usually made between ‘dialect literature’ and ‘literary
dialect’, the former term referring to works composed wholly, or at least
partly, in non-standard dialect, produced for a local readership, whereas
the latter refers to the representation of non-standard speech (almost
exclusively in the dialogue) in literature otherwise written in Standard
English. The novels by Elizabeth Gaskell, who was “well aware of the
need to balance authenticity and accuracy against accessibility” (Beal
2006: 534) are said to exemplify the latter category. This distinction is
somewhat fuzzy, in that writers of dialect literature more often than
not produce themselves in Standard English as well, and novelists like
Gaskell may have ‘insider’ knowledge about the variety they represent.
It is not completely true that “dialect is a variable dependent on the
demands of fictional situation rather than on the probable behaviour
of an actual speaker” (Page 1988: 59).
Schneider (2002: 71f), who is exclusively concerned with written
texts as data for linguistic, notably variationist, studies, provides a
useful and widely quoted classification of text types according to their
158 From Clerks to Corpora
proximity to speech, considering category (recorded, recalled, imagined,
observed, invented), reality of speech event (ranging from real to hypothetical), speaker-writer relationship, and temporal distance. According
to Schneider, texts should be as close to speech as possible to be of
value; in addition they must fulfil certain size requirements. Literary
writing is dismissed, since it normally displays ‘categorical invariant
usage’. Unfortunately, there has been a shortage of substantial corpora
of 19th-century fiction including non-standard dialogue, in spite of the
increase in dialect literature as well as literary dialect in the 19th ­century,
related to the rise of the novel. A close reading of, for example, Gaskell’s
fictional texts, however, makes it obvious that they cannot be characterized as displaying categorical invariant usage. It is my opinion that her
fictional representation of spoken language could be varyingly classified
as any of Schneider’s categories, i.e. ‘recorded’, ‘recalled’ etc.
The focus on factual knowledge above together with the plea for
authenticity may have given the reader of this text the impression that
authentic representation as propagated here should be characterized by
minute phonetic detail. This is by no means the case. The concept of
‘enregisterment’, as recently developed by Joan Beal (2009) and others,
whereby specific, often somewhat ‘levelled’, linguistic features become
associated with a particular variety and ‘reified’, is not in opposition to
authenticity as viewed in this contribution.
3. ‘Insider’ versus ‘outsider’ writers and readers
In his frustration after trying to include some phonetic symbols in rendering Cockney accents in Pygmalion, G.B. Shaw referred to “… this desperate attempt to represent her dialect without a phonetic alphabet” which
“must be abandoned as unintelligible outside London” (Pygmalion,
Act I). It would appear that Shaw believed only ‘insider readers’, i.e.
Londoners, would be able to understand his Cockney representations.
This may seem plausible enough, but the ability to read and interpret
dialect in writing is more complicated and requires more insight than
simply being a native speaker. The roles of ‘outsiders’ and ‘insiders’ can
be discussed from a variety of perspectives, such as the phonological level
of representation, language awareness and attitudes, and the ideology
underpinning the wish to write and read dialect texts. This is illustrated
in the following section by examples representing different time periods
(occasionally beyond the 19th century for clarification of some particular
issue), genres, and social/regional speech communities.
“Norfolk People Know Best” 159
A number of distinguished writers have represented varieties of language of which they have insider knowledge as well as varieties of which
they have acquired knowledge or limited experience. Elizabeth Gaskell,
for example, wrote two novels (Mary Barton and North and South) set
in Manchester, her hometown, but also a novel set in Whitby (Sylvia’s
Lovers), where she had spent only a fortnight. Another example of a
writer who wrote as an insider as well as outsider is Fanny Burney,
whose novel Camilla contains representations of a number of different
regional accents. Charles Dickens, as we all know, included speakers of
accents from various parts of Britain as well as abroad in his novels. He
has often been accused of lacking linguistic insight and reliability and
was, unlike Gaskell and Eliot, not accepted as an ‘informant’ by Joseph
Wright for the English Dialect Dictionary. Such a severe assessment
is somewhat unjustified, as shown by Poussa (1999) and in the sadly
neglected monumental work Sound and Symbol in the Dialogue of the
Works of Charles Dickens (Gerson 1967).
Even though writers producing non-standard spellings may be quite
knowledgeable about the variety they want to represent, they often fail
miserably due to the inadequacy of the orthographic system. Tennyson’s
elaborate use of ‘outlandish’ spellings in his Lincolnshire poems, for
example, is known to have made them largely inaccessible to the general reader. This is an important issue and the limitations of the orthographic system are therefore discussed in some detail in the following
4. The problem of representing phonetics/phonology
by means of the orthographic system
Whereas a near-authentic use of dialectal morphosyntax and ­vocabulary
is – at least superficially – fairly easy, the representation of sounds,
by contrast, is fraught with innumerable problems, for the writer as
well as the reader. In his 1809 work on the dialect of Bedfordshire, for
example, Batchelor describes “the Deficiencies of the English Alphabet,
when applied in the Explanation of provincial Errors of Pronunciation”
(Zettersten 1974: 157). It is an indisputable fact that the only way to
truthfully represent the actual pronunciation of vowels and consonants
in writing is by using a phonetic transcription (since the late 19th century preferably the International Phonetic Alphabet (IPA)), as exemplified in a simple but correct way in the title of Tony Harrison’s famous
poem Them and [uz]. Understandably, this is not a generally conceiva-
160 From Clerks to Corpora
ble alternative for publishers, writers and readers (see the Shaw quotation in the previous section).
The following examples will illustrate some aspects of the problem
through various employed strategies:
a)Aj häv to tjildren (‘I have two children’; from a textbook in
English for Swedish emigrants in the 1890’s)
b)<foot> vs. <strut>; <arrm> vs. <ahm> ; <the rang spee-oon>; <a
stee-an hoose> (Trudgill 1990)
c)laugh (laaf||laf (regional variants)); one (wun)||Northern England
also won) (Wells, Reader’s Digest Illustrated Dictionary 1984)
d)Ah, Apollo jars. Arcane standard, Hannah More. Armageddon
pier staff. (‘I apologize. I can’t stand it anymore. I’m a-gettin’
pissed off’) (Kingsley Amis 1968, I want it now)
e)To a Londoner, the strawbreez at Wimbledon ah veddy good with
clotted cream … (English Today 6, 1986)
f)Eh? good daäy! good daäy! thaw it bean’t not mooch of a daäy.
Nasty casselty weather! An’ mea haäfe down wi’ my haäy!
(Tennyson, The Church-Warden and the Curate)
In spite of their ‘outlandish’ appearance, all the above examples are –
with a varying degree of sophistication – exponents of some knowledge
of the represented accent; yet they cannot be characterized as very successful. Whereas the ‘transcriptions’ no doubt make perfect sense to
their creators, they are bound to be misleading, if not impenetrable, at
the receiver end. Although they all are meant to represent varieties of
English, some – notably a), d) and e) – clearly presuppose knowledge
of the creator’s variety, in the case of a) even another language. In a
manner of speaking they can – with the exception of a) – be said to
have been produced by ‘insiders’ with little or no consideration of the
‘outsider’ audience. More often than not, however, the transcriptions
tend not to be transparent to the insider audience either.
As for d), which from a British perspective is a very amusing and
adequate representation of an American accent, including rhythmical
features, it “would be rather impenetrable for an American audience”
(Wells 1982: 529). Conversely, e) includes apt observations of characteristic realizations and non-realizations of r in Received Pronunciation,
which would be below awareness of the speakers of this accent.
Examples b) and c), both designed by linguists for pedagogical reasons, certainly demonstrate more insightful attempts at ‘semi-phonetic
spellings’ (cf. Beal 2006: 531), which implies serious attempts at suggesting alternative pronunciations. Of a somewhat different nature is
“Norfolk People Know Best” 161
so-called ‘eye-dialect’, here used in the sense of ‘respellings which reflect
no phonetic facts’, such as sez for ‘says’, wimmin for ‘women’, including
representations of allegro speech such as ‘cause, ‘bout, showing natural
phonetic processes (richly demonstrated in local glossaries, for example).
A reasonably successful system of using near-exclusively letters of
the alphabet to indicate pronunciation was created by J.C. Wells for
the Reader’s Digest Great Illustrated Dictionary (1984) (cf. example c),
where [ə] (for ‘schwa’, the most common vowel in unstressed position)
is the only symbol taken from outside the alphabet. Some regional features, “considered standard in a particular region” are also catered for:
laugh, for example, is transcribed as (laaf||laf) and one is presented as
(wun||Northern England also won), in an unsuccessful attempt to indicate the ‘FOOT-STRUT split’ as featured in Tony Harrison’s [uz], i.e.
the lack of a phonemic opposition between the vowels of these words
and others belonging to the same sets (Wells 1982: 350f).
Similarly, for the first edition of his introductory textbook The
Dialects of England (1990), Trudgill designed a system consisting of
alphabetic letters, exemplified in b), in which the FOOT-STRUT split
is shown as (oo) vs (u), for example. In the second edition of his book,
however, “at the request of many readers”, he felt the need to complement these transcriptions with IPA versions, no doubt a justified step in
the case of a linguistic textbook. Having taught dialect courses based
on this textbook for a number of years, I can testify that students kept
begging for phonetic transcriptions (not a very common experience).
Trudgill’s fairly detailed orthographic system is of special interest here,
however, since he exemplifies most of his presentations of regional dialects by means of literary texts, as in the following extract from ‘The
Lincolnshire Poacher’ (Mabel Peacock, 1890s):
But I’d rather be doon wheare th’fire
An’ brimstun foriver bo’ns,
An’ just goä roond wi’ a bucket
An’ give fook drink by to’ns –
Then sit I’ yon stright made heaven,
Wheare saints an’ aängels sing …
Here the semi-phonetic spellings are the poet’s own, but my general
impression is that Trudgill, at least to some extent, has been inspired by
genuine dialect writing in designing his own system.
Tennyson’s dialect poetry, with its elaborate use of ‘outlandish’
spellings of the same character as ‘The Lincolnshire Poacher’, has been
162 From Clerks to Corpora
characterized as “largely inaccessible to the general reader because of
an unsuccessful attempt by the poet to indicate the precise nature of the
sounds of his native dialect” (Tilling 1972: 108). This is probably due
to lack of phonetic insight as well as the inadequacy of the orthographic
system, but above all to the poet’s overenthusiastic, unrealistic attitude
as a committed, ‘evangelizing’ insider.
Semi-phonetic spelling is also a long-standing concern of the BBC
pronunciation unit (cf. In
his assessment, dated 9 December, 2011, of its recommendations, Wells
draws attention to some particular problems in finding satisfactory
symbols in respelling systems for English, including
–– the PRICE vowel, for which neither y nor igh is unambiguous,
while ı has a diacritic
–– the MOUTH vowel, for which both ou and ow are ambiguous
(cf. soul, show)
–– the GOAT vowel, for which oh may wrongly suggest a short
vowel and oa, ou, ow are ambiguous (cf. broad, loud, now)
–– schwa. If oh represents a long vowel, how can we make it clear
that uh represents a short weak one?
In the following section the limitations of semi-phonetic spelling are
further discussed in some detail in connection with an ‘inventory’ of
the representation of accent in 19th century fiction, the main purpose
of which is to provide a background to understanding the case studies
and general discussion presented towards the end of the paper.
5. Regional accents in 19th century England – factual
knowledge and fictional representation
In comparison with earlier periods, the factual knowledge of English
accents and dialects as they were spoken in the nineteenth century is
more than significant. The first truly ambitious as well as insightful
attempt at mapping English dialect areas was that of Alexander Ellis
(Ellis 1889), but of even greater importance within the context of the
present paper is Joseph Wright’s English Dialect Dictionary (EDD)
(Wright 1898–1905), which contains more information, is more accessible (especially after its recent digitization), and is largely based on
examples drawn from fiction, generally representing insiders. The fact
that Wright, with his monumental knowledge of English dialects, has
included a word form constitutes a guarantee of its real-life existence
but there is, admittedly, a danger of circularity here.
“Norfolk People Know Best” 163
An important factual source of another kind is the Survey of English
Dialects (SED) (Orton et al. 1962–71), based on fieldwork in the
mid–20th century but in its focus on old speakers reflecting regional/­
nonstandard usage not too distant in time from the fictional representations featuring in this paper. Indeed, according to Ihalainen (1994: 205),
“no radical changes took place in English dialects in the post-1776 period
until the second half of the twentieth century”. The fairly recent, widely
quoted mapping of ‘Traditional Dialect areas’ (Trudgill 1999) largely
draws on SED data and is also very relevant for this presentation. His
maps confirm, among other things, that the major division is a north–
south one, e.g. demonstrating the FOOT–STRUT split as mentioned
above. As for fictional representations of regional dialects, however, it
should be pointed out that those referring to the south and middle of
England are neither as easily found nor as well researched as those of the
north. This may be due to a perception of general southern features as
connected with the standard, whereas northern speech is ‘marked’.
Trudgill’s mapping based on phonological criteria defines a staggering
number of dialects and subdialects, such as “Southern Eastern Central
East”. It is not the purpose here, nor would it be possible, to describe
and exemplify all these varieties through fiction. Rather, from a selection of characteristic features, I will demonstrate how writers have –
successfully as well as unsuccessfully – tried to represent different kinds
of phonological features, considering the limits of semi-phonetic spellings. Interestingly, but perhaps not surprisingly, the general north–south
distinctions, such as the FOOT–STRUT split, do not appear to be represented (cf. Wales 2010: 70, however). Searching for examples is not
an easy task, since text corpora have, until recently, not included dialect
literature and deliberately shunned literary dialect. Thanks to the launch
of the Salamanca Corpus (
html), searches will hopefully be more successful in the future. For the
present study, most of the examples derive from the quotations found in
EDD. Unfortunately, however, regional areas are very unevenly represented in the dictionary, as recently demonstrated by Praxmarer (2010).
In sociolinguistics and dialectology, vowels have generally attracted
more attention than consonants. This also appears to be characteristic
of literary representations; the reason may be that it is relatively easier to create semi-phonetic spellings by modifying vowel symbols than
by drastically exchanging one consonant symbol for another. The following listing exemplifies fairly successful renderings of regional vowel
features by 19th century writers in various genres (the words written
in capital letters are key words, representing a category characterized
164 From Clerks to Corpora
by the same vowel and the words in bold exemplify a semi-phonetic
spelling). By ‘fairly successful’ I mean that any reader (outsider as well
as insider) conversant with English could be expected to perceive the
intended sound quality represented by the semi-phonetic spelling. This
must, unfortunately, be characterized as a qualified guess; I am not
aware of any substantial study investigating the reading aloud of dialect texts.
Most of the following features are described as traditional dialect
features in Trudgill 1999.
a)Realization of LAND ([land] generally in the north; [lænd] in the
south; [lɒnd] in the West Midlands: E’s gotten a bwile in ‘is lonk,
poor bwoy (Herefordshire 19th-century anonymous text). Other
examples from the area include mon, hond, ony and such representations are also found in parts of the North (Wales 2010:
70). This spelling works well. Distinguishing between [land] and
[lænd] is obviously more problematic; nor have any examples
been found. That general north-south distinction may well be disregarded for the same reasons as the FOOT-STRUT split: it is
difficult to represent and taken for granted at least if the writers
are insiders.
b)Monophthongal realization of DOWN: doon, roond (Peacock,
Th’Lincolnsheer Poächer), demonstrating Lincolnshire’s ‘northern affiliation’, coo, thoosand (north of the Humber, cf. Wales
2010: 71).
Insertion of <w> before <o> (involving a semi-vowel + a
vowel) (Dorset, Somerset): primrwose, hwome (William Barnes,
Blackmwore Maidens).
d)Yod-dropping (East Anglia), also involving a semi-vowel: solitoode, gratitoode (Dickens, Great Expectations).
e)[y:], i.e. [i:] with lip-rounding, in GOOSE words (Devon): güzechick, güze vlesh (EDD, unidentified source). This will presumably work well if the reader has some knowledge of German.
f)short /a/ in the verbs make and take in the North, up to the
Durham–Northumberland border: mak/mek; tak (Wales 2010:
It is symptomatic that out of the four vowels posing particular problems in respelling according to Wells (see above) three are realised as
diphthongs. This is well illustrated in Tennyson’s Lincolnshire poetry,
with its excess of unusual letter combinations including diacritics
and diaeresis (the pronunciation of vowels in a diphthong separately)
“Norfolk People Know Best” 165
(cf. Tilling 1972). Similarly, the bulk of Gerson’s (1967) inventory of
Dickens’ respellings relates to diphthongs.
Some examples of felicitous representations of consonants are:
• H insertion (the hypercorrect counterpart to H dropping): hodd
‘odd’, hany ‘any’ (George Eliot, Adam Bede), suggesting a desire
to identify with the gentry on the part of one of the characters (cf.
Dickens in David Copperfield, who characterizes Uriah Heep’s
speech by exaggerated H-dropping).
• voiced initial fricatives (the West Country): vorzeaken ‘forsaken’,
zot ‘sat’ (William Barnes’ poem The Broken Heart; the protagonist’s name is Fanny, however).
• interchange of /v/ and /w/ (Eastern counties): “A man your vorship, may call out ‘boots’ and not wiolate any hact vatsomdever”
(The Times, Jan. 26, 1835). Wells (1982: 333), assessing ‘literary
Cockney’, dismisses Dickens’ elaborate representation of this feature as a literary stereotype, seriously out-of-date at the time of
writing. [w] for /v/, however is reported by Skeat from London in
the latter half of the nineteenth century (Skeat’s personal observation, cf. Gerson 1967: XIX).
• realizations of consonant /k/ clusters (Yorkshire, Cumberland):
tnit ‘knit’, tnee ‘knee’; tlay ‘clay’, dlass ‘glass’ (Wales 2010: 71).
• West Midland ‘g-ful’ endings: playingk for ‘playing’ in an 1886
Cheshire text (cf. Skeat 1911: 122).
A characteristic problem in the representation of consonants, i.e.
rhoticity vs. non-rhoticity, was already touched upon in the previous
section (cf. examples d) and e)). In 19th-century fiction, due to the relative lateness of the emergence of non-rhoticity in conjunction with
the spelling convention, ‘r-fulness’ does not appear to be indicated. In
The Mayor of Casterbridge, set in the West Country, Hardy occasionally marks the speech of the Scotsman Donald Farfrae, e.g. in warrld
‘world’ (presumably signalling focus), but never in the representation
of his local speakers (whose speech was presumably also ‘r-ful’). This
may well be an ‘insider effect’. The interpretation of r-fulness/r-lessness
based on spelling must be made with caution, however, since the social
significance of r-dropping remains unsettled even in the early twentieth
century (Ihalainen 1994: 215) and it is the occasional presence of rhoticity that attracts attention (Wells 1982: 30).
With the exception of representations of allegro speech, such as
‘cause, ‘bout, usually showing natural phonetic processes but richly
166 From Clerks to Corpora
demonstrated in local glossaries, examples are scarce when it comes to
indicating suprasegmental features such as syllabic structures or prosody. Sentence stress is occasionally indicated by capital letters, as in
Mr. Podsnap’s famous didactic conversation with a French gentleman:
You find it Very Large? And very Rich? (Dickens, Our Mutual Friend).
No examples are given here of representations of prosodic patterns such
as rhythm or intonation, although 20th-century examples show that it
is possible (cf. the Kingsley Amis representation of an accent from the
American South ((d), above) and the trendy rising intonation in ‘uptalk’
as featured in Jane Smiley’s Moo (1995): “No, sir? You gave me an
A?, See, that was the only A I’ve gotten here?”. For an i­nteresting and
original analysis of significant pitch span variation expressed in fi
­ ction,
including works by 19th and 18th century writers, see the chapter
‘Paralinguistic features’ in Gillian Brown’s Listening to Spoken English
(1977). Hodson (2014: 85) draws attention to Dickens’ representation
of the Artful Dodger’s speech in Oliver Twist, in which there is
… some attempt to indicate his intonation through punctuation,
such as the exclamation marks which indicate emphasis, the question marks which indicate a rising intonation, and the dashes in the
word ‘com-pan-ion’ which presumably indicate that each syllable of
the word is sounded out in full.
Interestingly, and clearly related to the insider/outsider thrust of this
paper, Hodson continues:
All of this can be contrasted with the way in which Oliver’s speech
is reported in this passage, which is given the form of indirect speech
as the narrator summarizes what Oliver said, without giving any
flavour of how he said it … the dialect speech is ‘other’ and its peculiarities are highlighted, while the main narrative work is conducted
in Standard English.
6. Insiders and outsiders in the perception, interpretation
and representation of accents – a brief summary and four
illustrative case studies
In studying the use of nonstandard varieties in English literature, it is
important to remember that English spelling does not represent any
existing dialect phonetically. By convention, therefore, when a writer
uses normal English spellings in dialogue, for example, we infer
that the pronunciation intended is the standard of the audience for
which the work is written, while special deviant spellings indicate the
“Norfolk People Know Best” 167
pronunciation of a dialect that is not the audience’s standard. This
can lead to some rather unusual variations. For example, a writer
representing an Irishman to a predominantly English audience might
be inclined to use spelling to indicate Irish pronunciation, while the
same writer might not do so when presenting an Irishman to a predominantly Irish audience.
(Traugott & Pratt 1980: 339)
In this final section of the paper the insider–outsider aspect, highlighted in the main title, takes centre stage. This aspect has already
surfaced in various contexts; hence it makes sense to begin by summarizing what has emerged so far. This summary is followed by a few brief
‘case studies’ of certain works by 19th-century writers who have produced representations of accents both intrinsic and extrinsic to them:
Fanny Burney, Elizabeth Gaskell, George Eliot, and Alfred Tennyson.
By way of conclusion, various general aspects of the topic are briefly
considered, such as the phonological level of representation, linguistic
awareness and attitudes, and the ideology underpinning the wish to
write and read dialect texts, including conscious effort-raising measures
in, for example, the education system.
A brief summary
• Doubts and fears have been voiced as to the reliability of outsiders representing accents as well as being able to interpret/read
representations of other accents than their own, intrinsic variety
(Hickey 2010: 9; Shaw (Pygmalion));
• In an extended sense, the outsider/insider factor also applies to
linguists as well as perceptive writers of fiction who do not envisage the problems facing the readers of their efforts (cf. examples a)–f) in section 2 above). If, for example, students exposed to
the semi-phonetic spellings in Trudgill’s The Dialects of England
have a Yorkshire accent or have Swedish as their first language,
they tend not to be able to read out the semi-phonetic spellings in
the desired manner; hence they demand phonetic transcriptions.
It has been pointed out that “any attempt at indicating accent
through orthographical manipulation will only work if writer and
reader share an understanding of the variety being so represented” (Hodson 2014: 92);
• Whereas the concept of ‘semi-phonetic spellings’ implies serious attempts at suggesting alternative pronunciations, the only purpose
168 From Clerks to Corpora
of ‘eye-dialect’ (respellings which reflect no phonetic facts, such as
sez, wimmin, ennything) appears to be a signal to the reader that a
character uses vulgar or nonstandard language. Consider, for example, the use of nonstandard spellings in Gone with the Wind for
the speech of blacks while using standard spelling for whites, even
though the speech of both groups is phonetically very similar. In this
case, writers as well as readers might well be extrinsic to the variety.
• As noted, for example, in connection with Tennyson’s elaborate
spellings, and also in the general analysis of regional features, representations are often inaccessible to the reader because of inability to indicate the precise sounds of an accent. This inability
could be due to phonetic ignorance but above all to the sheer
impossibility of representing phonetic detail in an orthographic
transcription, which will suggest different realizations to outsider speakers of different accents. Representations like Tennyson’s
would presumably cause problems to insider readers as well, unless they were given special training.
Four case studies:
Fanny Burney:
In her widely popular novel Camilla (1796) Fanny Burney included
a chapter containing a lively account of a performance of Othello, in
which all the parts except Iago were played by actors speaking their
own local dialect. Cassio, for example, who is presented as hailing from
Norfolk, says dewk for ‘duke’ (a representation of Yod-dropping, not
quite as felicitous as Dickens’ <oo> exemplified in section 3 above).
Othello himself is said to be ‘a true Londoner’, as exemplified by wery,
avay, and Desdemona’s father, a West Country man from Somerset,
­produces zpeak, confez (the second example not adequately exemplifying ‘voicing of initial fricatives’). The actress playing Desdemona is
said to come from Worcestershire, but her speech is – confusingly –
­represented mainly through excessive H dropping as well as H insertion.
Fanny Burney was known to have ‘a good ear for dialect’ which is,
on the whole, apparent from her representations. She appears to be particularly successful in representing her home dialect (Norfolk), i.e. as an
insider. It seems likely, however, that due to her exposure to other local
accents through a wide circle of acquaintances in her London life, her
awareness of the home dialect had been heightened. It is interesting to
note – as pointed out by Blank (1996: 3) – that for Renaissance writers
“Norfolk People Know Best” 169
dialects appeared to have nothing to do with ‘home’; Shakespeare, for
example, never represented Warwickshire speech.
George Eliot:
Warwickshire speech as perceived by an insider some 300 years later,
however, occasionally features in George Eliot’s novels, for example
Silas Marner. An experienced translator and editor, she had given much
thought to the problem of representing local speech, as seen in the following quotation from a letter to Skeat, published in the Transactions
of the English Dialect Society:
It must be borne in mind that my inclination to be as close as I could
to the rendering of dialect, both in words and spelling, was constantly checked by the artistic duty of being generally intelligible. But
for that check I should have given a stronger colour to the dialogue
in Adam Bede, which is modelled on the talk of North Staffordshire
and the neighbouring part of Derbyshire. The spelling, being determined by my own ear alone, was necessarily a matter of anxiety, for
it would be as possible to quarrel about it as about the spelling of
Oriental names. The district imagined as the scene of Silas Marner
is North Warwickshire; but here, and in all my other presentations
of English life except Adam Bede, it has been my intention to give
the general physiognomy rather than a close portraiture of the provincial speech as I have heard it in the Midland or Mercian region.
It is a just demand that art should keep clear of such specialties as
would make it a puzzle for the larger part of its public; still, one is
not bound to respect the lazy obtuseness or snobbish ignorance of
people who do not care to know more of their native tongue than
the vocabulary of the drawing-room and the newspaper. (cf. Cooke
1883: 293)
George Eliot, indeed, practises as she preaches, i.e. indicating ‘the
g­eneral physiognomy’; hence not much of interest with regard to
phonetic detail is found in her work. Instead, she provides rich and
consistent details representing local syntax and morphology, including allegro features (i’ ‘in’, ha’ ‘have’, wi’ ‘with’). Her representations,
incidentally, are strikingly in accordance with Tolkien’s use of dialectal Warwickshire/Oxfordshire forms in The Lord of the Rings, characterized by Johannesson (1994: 55) as “selective rather than wholesale”. A comparison between the dialogues in Silas Marner and Adam
Bede suggests clear but subtle distinctions between the two represented
regional varieties. In Adam Bede, especially in Lisbeth’s speech, there
170 From Clerks to Corpora
are examples of generalized definite article reduction (th’) and, as stated
by the author herself, there is generally a closer ‘portraiture’. Hence
Eliot as an outsider has produced a more detailed representation than
Eliot as an insider.
Elizabeth Gaskell:
Elizabeth Gaskell was – like George Eliot – “well aware of the need
to balance authenticity and accuracy against accessibility” (Beal 2006:
534). In contrast with Eliot, however, she was keen to include phonetic
detail in her renderings of social and regional dialects, but an i­ nteresting
shift in the character of these representations can be observed. As her
writing and creative power of representing human speech and behaviour matured, she appears to have shifted from an essentially itembased approach to a more discourse-based, psychologically motivated
one, yet without foregoing her linguistic intuition and knowledge.
Her linguistic reliability was recognized by Joseph Wright in that
he used two of her novels (Mary Barton and Sylvia’s Lovers), exemplifying two different dialect areas, as data for specimens in his dictionary. In Mary Barton, her first novel, which is set in Manchester, her
home town for the last seventeen years, Gaskell somewhat didactically
made a point of using certain tokens of Lancashire dialect vocabulary,
for which she could provide explanations or etymologies in footnotes,
often exemplifying the use of the same words by renowned writers such
as Chaucer and Shakespeare. For the verb form getten, for example,
a reference is made to Chaucer’s Canterbury Tales: ‘For he had geten
him yet no benefice’. Such references seem to signal the author’s explicit wish to raise the status of the dialect. As the story proceeds, she
appears to have given up such elaborate references and they are not
found in her later works. It has always been assumed that Elizabeth
Gaskell drew a great deal of her knowledge of regional dialect from her
husband, the Rev. William Gaskell, who was known to have lectured on
Lancashire dialect. A close study of his lectures, however, reveals that
his interest in dialect seemed to be of a more traditional character than
what is signalled in the works of his wife. His articles deal exclusively
with vocabulary, presented in item-based listings, including some wild
A close examination of the manuscript of her ‘Whitby novel’ Sylvia’s
Lovers reveals that she made a number of changes suggesting linguistic
awareness, for example with regard to the use and form of the definite
article, and for the second edition she ‘corrected’ the dialect, changing
“Norfolk People Know Best” 171
some Lancashire forms into East Yorkshire ones. It should be noted that
she largely based her familiarity with Whitby speech on two weeks’ holiday in the area and some later, more ‘academic’ consultations. In Mary
Barton, published in 1848, as well as Sylvia’s Lovers (1863) dialect is
used extensively, including some detailed representation of segmental
phonology, which is in accordance with SED findings (Melchers 1978:
116–18). If anything, there is more detail in the Whitby novel, i.e. the
representation of the more ‘extrinsic’ variety.
The detailed representation in Sylvia’s Lovers, her last novel but one,
is not quite in accordance with her development as outlined above. In
North and South (1855), set in Manchester, and – in particular – Wives
and Daughters (1865), probably representing her most ‘intrinsic’ area,
she develops fine nuances of social/regional differences in discourse,
and the item-based features tend to be more generalized. Interestingly,
the recent TV adaptations of Gaskell’s novels demonstrate an awareness of ‘enregisterment’ in that regional and social differences are not
explicit but hinted at in subtle ways and through few but recurrent
linguistic features.
Alfred Tennyson:
Of the four writers featured in the case studies, Alfred, Lord Tennyson,
is by far the one who has made the greatest effort to create a ‘genuine’
local dialect representation, resulting in a staggeringly complex array
of outlandish vowel symbols. Consider, for example, the beginning of
his well-known poem Northern Farmer, Old Style, the first of his poems
written in Lincolnshire dialect:
“Wheer ‘asta beän saw long and meä liggin’ ‘ere aloän?”
Alfred Tennyson (1809–92) was born in Somersby, Lincolnshire, but
left the county for the south of England a good twenty years before
he began writing poetry in dialect, paid only occasional visits to the
area after that and had little contact with the speech he was trying to
convey. He did, however, publish seven long poems in dialect, the last
of them published posthumously. Tennyson is known to have taken
great trouble in consulting experts in the field, including A.J. Ellis, who
‘proofread’ his poetry carefully and offered a great deal of criticism but
also profited on his many discussions with the poet for the compilation
of his monumental On Early English Pronunciation (1889). It is also
worth mentioning that most of Tennyson’s dialect poems were used as
source material in another monumental publication, Joseph Wright’s
EDD (1898–1905).
172 From Clerks to Corpora
In his pioneering study Local Dialect and the Poet (1972), Philip Tilling,
editor of the SED volumes on East Midland dialects, scrutinizes and discusses Tennyson’s representation of dialect in detail as collated with the
findings at the Lincolnshire localities. This penetrating and knowledgeable
investigation demonstrates how a substantial number of Tennyson’s representations correspond to SED data from other parts of England, at best
from other parts of Lincolnshire (incidentally, a most complex area which
probably explains some mistaken advice from the experts consulted) but
also the North and the West Country. Tilling’s general conclusion is that
“the poems, though they contain much that seems to be genuine, cannot
really be said to give an entirely reliable impression of the Lincolnshire
dialect heard by Tennyson in his youth” (p. 107).
7. Concluding remarks
In concluding this attempt to discuss some aspects of the representation
of English accents in 19th-century fiction, it seems justified, if not inevitable, to ask the following question: Who does it best – the outsider or
the insider? The text so far has not been conclusive: on the one hand
profound knowledge of a variety is required, but it can also lead to
exaggerated narrowness in the representation with a frustrated readership, including insiders, as a result. “It is to be acknowledged that even
texts by native or local writers, however informal, have potential problems as accurate or reliable sources of linguistic data, particularly phonological” (Wales 2010: 68). A close study of some appreciated fictional
writers, in fact, reveals that they may often be more competent in representing dialects other than their native tongue. According to Hickey
(2010: 9), who talks about ‘scalar insiderness’, the status as complete
outsider nearly always goes together with a satirical approach. This is,
however, hardly borne out by the data considered here. Consider also
how Dickens, who skilfully represented an amazing number of different accents, as carefully documented in Gerson 1967, reacted when
accused by the Spectator of using dialect as a means of mockery:
I believe that virtue shows quite as well in rags and patches as she
does in purple and fine linen, … even if Gargery and Boffin did not
speak like gentlemen, they were gentlemen.
(Gerson 1967: 371–2)
From the readers’ point of view, the more accurate the phonetic
spelling, the more frustrating it will be to read. Most adults read word
“Norfolk People Know Best” 173
by word, not sounding words out letter by letter, so forcing adults to
sound out nonstandard phonetic spellings would slow readers down,
potentially irritating them, and thus distract them from the actual
story. A native of Lincolnshire comments on Tennyson’s elaborate use
of dialect and ‘semiphonetic’ spelling in the following way: “ploughing
through line after careful line, I found them as thick as porridge”. On
the same note: in his lecture Local Speech in Writing: Surely Nobody
Reads It!, Stanley Ellis (1989: 20)) questions the value of elaborate
representations of local accents even by ‘insider star performers’ such
as Emily Brontë, wondering “whether Joseph really matters so much;
if people find Emily so well worth while that they are prepared to read
Wuthering Heights without even seeing the dialect bits”.
In an important paper, Trudgill (1999b) shows how Norfolk Yod
dropping is seen as a very salient feature by outsiders and thus represented in writing (Dickens, for example, writes dootiful for ‘dutiful’),
whereas insiders do not bother to change the spelling or, confusingly,
produce spellings such as bewtiful (in fact, representing a traditional-dialect, closer and more centralized vowel, which is undergoing
dedialectalization to [u:]) in another set of words such as boat, road,
fool. Trudgill’s conclusion is: “As usual, Norfolk people know best”.
This is not an unqualified truth, however, since insiders are known to
have a tendency to seek to confirm their own preconceived notions
and stereotypes, whereas outsiders may have ‘fresh ears’ (cf. Melchers
1996: 164f). Ideally, linguistic investigations of local and social dialects
should be carried out by insiders and outsiders working together, as
recommended, for example, by Lesley Milroy (
my-esrc/grants/R000221074/read). The same approach would not be
amiss in the representation of dialect in fiction.
Beal, J. (2006). Dialect representations in texts. K. Brown (ed.) Encyclopedia
of Language and Linguistics (2nd ed.) Amsterdam; Oxford: Elsevier Ltd.,
———. (2009). Enregisterment, commodification, and historical context:
“Geordie” versus “Sheffieldish”. American Speech 84:2, 138–56.
Blank, P. (1996). Broken English. Dialects and the Poetics of Language in
Renaissance Writings. London: Routledge.
Brown, G. (1977). Listening to Spoken English. London: Longman.
174 From Clerks to Corpora
Burney, F. (1796). Camilla. E.A. & L.D. Bloom (eds) 1983. Oxford: Oxford
University Press.
Cooke, G.W. (1883). George Eliot: A Critical Study of her Life, Writings and
Philosophy. Cambridge: Cambridge Library Collection.
Ellis, A.J. (1889). On Early English Pronunciation, Part V, the Existing
Phonology of English Dialects Compared with That of West Saxon (Early
English Text Society, Extra Series 56). London: The Philological Society.
Ellis, S. (1989). Local speech in writing: Surely nobody reads it! The Hilda
Hulme Memorial Lecture. University of London.
Gerson, S. (1967). Sound and Symbol in the Dialogue of the Works of
Charles Dickens. Stockholm Studies in English XIX. Acta Universitatis
Stockholmiensis. Stockholm: Almqvist & Wiksell.
Harrison, T. (1986). Selected Poems. Harmondsworth: Penguin.
Hickey, R. (2010). Linguistic evaluation of earlier texts. R. Hickey (ed.)
Varieties of English in Writing. Amsterdam: Benjamins, 1–14.
Hodson, J. (2014). Dialect in Film and Literature. Basingstoke: Palgrave
Ihalainen, O. (1994). The dialects of England since 1776. R. Burchfield (ed.)
The Cambridge History of the English Language. Vol. V. Cambridge:
Cambridge University Press, 197–274.
Johannesson, N-L. (1994). Subcreating a stratified community – On J.R.R.
Tolkien’s use of non-standard forms in The Lord of the Rings. G. Melchers
& N-L. Johannesson (eds) Nonstandard Varieties of Language. Stockholm
Studies in English LXXXIV. Acta Universitatis Stockholmiensis. Stockholm:
Almqvist & Wiksell International, 53–63.
Kirk, J.M. (1999). Contemporary Irish Writing and a Model of Speech Realism.
I. Taavitsainen, G. Melchers & P. Pahta (eds) Writing in Nonstandard
English. Amsterdam: Benjamins, 45–62.
G. Melchers (1978). Mrs. Gaskell and Dialect. M. Rydén & L.A. Björk (eds)
Studies in English Philology, Linguistics and Literature. Stockholm Studies
in English XLVI. Acta Universitatis Stockholmiensis. Stockholm: Almqvist
& Wiksell International, 112–124.
———.(1996). On the value of the SED recordings of spontaneous speech. J.
Klemola, M. Kytö & M. Rissanen (eds) Speech Past and Present. Studies
in English Dialectology in Memory of Ossi Ihalainen. Bamberger Beiträge
zur Englischen Sprachwissenschaft 38. Frankfurt am Main: Peter Lang,
“Norfolk People Know Best” 175
Orton, H. et al. (1962–71) Survey of English Dialects. The Basic Material Vols.
I–IV. Leeds: E.J. Arnold.
Page, N. (1988). Speech in the English Novel. London: Macmillan.
Peacock, M. (c. 1890). ‘Th’ Lincolnsheer Poächer’ in G.E. Campion,
Lincolnshire Dialects. Boston, Lincolnshire: Richard Kay 58–59.
Pollard, A. (1965). Mrs. Gaskell. Novelist & Biographer. Manchester:
Manchester University Press.
Poussa, P. (1999). Dickens as sociolinguist: Dialect in David Copperfield.
Taavitsainen, I., G. Melchers & P. Pahta (eds) Writing in Nonstandard
English. Amsterdam: Benjamins, 27–44.
Praxmarer, C. (2010). Joseph Wright’s EDD and the geographical distribution of dialects: A visual approach. M. Markus, C. Upton & R. Heuberger
(eds) Joseph Wright’s English Dialect Dictionary, and Beyond. Frankfurt
am Main: Peter Lang, 61–73.
Reader’s Digest Great Illustrated Dictionary (1984). 2 vols. London: The
Reader’s Digest Association Ltd.
Schneider, E. (2002). Investigating variation and change in written documents.
J.K. Chambers, P. Trudgill & N. Schilling-Estes (eds) The Handbook of
Language Variation and Change. Oxford: Blackwell, 67–96.
Skeat, W. (1911). English Dialects from the Eighth Century to the Present
Day. London: Cambridge University Press.
Tilling, P. (1972). Local dialect and the poet: A comparison of the findings
in the Survey of English Dialects with dialect in Tennyson’s Lincolnshire
poems. M. Wakelin (ed.) Patterns in the folk speech of the British Isles.
London: Athlone Press, 88–108.
Traugott, E.C. & M.L. Pratt (1980). Linguistics for Students of Literature.
New York: Harcourt Brace Jovanovitch.
Trudgill, P. (1990). The Dialects of England. Oxford: Blackwell.
———. (1999a). The Dialects of England. Second edition. Oxford: Blackwell.
———. (1999b). Dedialectisation and Norfolk dialect orthography.
Taavitsainen, I., G. Melchers and P. Pahta (eds) Writing in Nonstandard
English. Amsterdam: Benjamins, 323–9.
Wales, K. (2010). Northern English in writing. R. Hickey (ed.) Varieties of
English in Writing. Amsterdam: Benjamins, 61–80.
Wells, J.C. (1982). Accents of English. Vols. I-III. Cambridge: Cambridge
University Press.
176 From Clerks to Corpora
Wright, J. (ed.) (1898–1905). The English Dialect Dictionary. 6 vols. Oxford:
Oxford University Press.
Zettersten, A. (ed.) (1974). A Critical Facsimile Edition of Thomas Batchelor,
An Orthoëpical Analysis of the English Language and An Orthoëpical
Analysis of the Dialect of Bedfordshire (1809). Part I. Lund Studies in
English 45. Lund: Gleerup.
10 Sublime Caledonia: Description, Narration and Evaluation in
Nineteenth-century Texts on Scotland
Marina Dossena
Università degli Studi di Bergamo (I)
1. Introduction
The nineteenth century proved crucial for the establishment of a romanticized image of Scotland. Despite, or, more correctly, as a result of,
the impact of the Highland Clearances, which left many areas virtually
deserted, the country came to be perceived in the Lowlands, in England,
and even abroad, as a picturesque wilderness, a totally appropriate setting for ballads, stories and legends. Indeed, even dramatic episodes of
(often forced) emigration became the object of artistic representation,
as – perhaps most famously – in the painting The Last of the Clan
(1865, by Thomas Faed, currently at the Kelvingrove Art Gallery and
Museum in Glasgow).
In this contribution I intend to highlight the main features employed
in a sample of nineteenth-century texts relating to Scottish history and
landscape, in an attempt to identify what linguistic choices played a key
role in the construction of a romanticized environment. More specifically, I intend to discuss two case studies, Nattes’s Scotia Depicta (1804)
and Robert Louis Stevenson’s Edinburgh: Picturesque Notes (1879),
while placing them in the framework of other materials available in the
Corpus of Modern Scottish Writing, travelogues, and other well-known
publications, both literary and non-literary texts. Special attention will
be paid to evaluative language and stylistic moves that enabled authors
to signify their appreciation of their topics. Such strategies of description and narration clearly aimed to persuade readers of the validity of
the authors’ views, and were often accompanied by illustrations meant
How to cite this book chapter:
Dossena, M. 2015. Sublime Caledonia: Description, Narration and Evaluation in
Nineteenth-century Texts on Scotland. In: Shaw, P., Erman, B., Melchers, G. and Sundkvist,
P. (eds) From Clerks to Corpora: essays on the English language yesterday and today.
Pp. 177–191. Stockholm: Stockholm University Press. DOI:
bab.j License: CC-BY.
178 From Clerks to Corpora
to provide a visual counterpart of the textual statements. The role of
intertextual references will thus be taken into consideration, in order to
outline the textual networks that appear to be in place.
After an overview of the perception of Scotland’s antiquity in Late
Modern times, my analysis will focus on the two texts mentioned above
(Nattes’s and Stevenson’s), both published with a manifestly descriptive
aim, in order to assess their persuasive quality. Close readings of the
texts will be supplemented with corpus-based investigations of specific
lexical items. Finally, the concluding section will summarize the main
strategies that appear to be at work in both texts.
2. Scotland in Late Modern times: views of language and landscape
Many eighteenth- and nineteenth-century texts discussed Scotland’s
unique landscape together with the specificity of its language. While
Scots was stigmatized in everyday usage, its occurrence in literary texts
was perceived to be both ‘pithy’ and appropriate. Comments on Burns’s
poetry were typically accompanied by remarks on the difficulty of his
language (Dossena 2012a), which stressed the ‘exotic’ character of
Scots, but praises of its antiquity were also frequent, on account of its
supposedly greater proximity to ‘pure Saxon’.
This attention to linguistic roots appears to have close cultural
connections with the fashionable search for antiquity, the picturesque
and the sublime, which persisted through the times of the Napoleonic
wars and reached a turning point during the Victorian age. In particular, for the Romantics the chief attraction in Scotland was possibly
the isle of Staffa, first discovered by Joseph Banks in 1772. Banks
reported that “There is a cave […] which the natives call the Cave of
Fingal”, and although this was probably a misunderstanding, the place
inspired countless creative artists, not least J.M.W. Turner and Felix
It is beyond the scope of this contribution to discuss the impact
of royal visits, starting from the momentous event when, in 1822,
George IV visited Scotland for the first time since James VII and II’s
stay in 1681/82, long before the Union of Parliaments. The role played
by Walter Scott in the organization of the visit is also well-known, and
1 On Turner in Scotland see Grenier (2005: 56, 94, 100) and Mitchell (2013: 197–
206, 219). On Mendelssohn-Bartholdy see Grenier (2005: 161).
Sublime Caledonia: Description, Narration and Evaluation 179
does not need to be summarized here.2 However, the new attitude of
the monarchy after the repeal of the Act of Proscription3 in 1782 contributed to the success of travelogues published both north and south
of the border.
As for the interest in Scotland’s antiquities, this had begun in the
late seventeenth-early eighteenth century: Martin Martin’s A Late
Voyage to St. Kilda (1698) and A Description of the Western Islands
of Scotland (1703) report on journeys that appear to have been made
mainly at the request of an antiquary, Sir Robert Sibbald. Although
earlier journeys had been made into the Highlands, it was only in the
eighteenth century that their accounts became popular, as in the case
of Sir Donald Monro’s Description of the Western Isles of Scotland
(1549), a few copies of which were printed in 1774. While Thomas
Pennant’s accounts had a more naturalistic focus (Youngson 1973 and
1974), an antiquarian focus was a common denominator in Bishop
Pococke’s letters of 1760 (Youngson 1973: 2).
The raging fashion for antiquity also gave rise to satire (Brown 1980:
10); nonetheless, this interest was made explicit in the 1829 reprint
of Petruccio Ubaldini’s Descrittione del Regno di Scotia, originally
published in 1588. Moreover, the success of Scott’s novels was a great
boost to the recognition of Scotland as a tourist destination. Theodor
Fontane’s accounts (1860/1989: 129, 159) explicitly refer to Rob Roy
and The Fair Maid of Perth; a quotation from Burns’s Drumossie Moor
introduces the section on Culloden (1860/1989: 187), while Fontane’s
notes on the Old Town of Edinburgh and its Spukhäuser (1860/1989:
22, 83) seem to anticipate Robert Louis Stevenson’s Edinburgh:
Picturesque Notes (1879). Finally, the Ossian quest had also been at
the centre of Louis-Albert Necker de Saussure’s Voyage en Ecosse et
aux Iles Hébrides (1821). Towards the end of the century, an anonymous author referred to the same texts by Sir Walter Scott (Anon. 1894:
16, 37–38), but also to works by James Hogg (Anon. 1894: 63) and
indeed to Robert Louis Stevenson’s Kidnapped (Anon. 1894: 103–106,
111), outlining a trail around Rannoch Moor and ‘Cluny’s cage’ that
would then be followed by numerous readers and enthusiasts – see for
instance Nimmo (2005).
On this topic see, most recently, Kelly (2010: 187–211) and Mitchell (2013:
3 This is the Act of Parliament which in 1746, after the last Jacobite rebellion, forbade the use of Highland garb (19 Geo. II, ch. 39, sec. 17, 1746), thus reinforcing
previous ‘Disarming Acts’.
180 From Clerks to Corpora
In addition to antiquity and literature, travelogues typically stressed
the scenery’s most romantic traits, and though they often repeated
cultural stereotypes, such as the fact that “Some of the poor in Skye
have scarcely a notion of any food but oatmeal” (Sinclair 1859: 181),4
they contributed to the creation of an idealised picture of Scotland.
Travelogues and geographical narratives thus appear to have had both
an informative and a promotional function, not least in terms of cultural perception. In what follows I will investigate two texts currently
available in electronic format, in order to assess the linguistic choices
that appear to be most significant in this respect. As I mentioned above,
these are case studies: other texts could be selected, so as to give greater
generic depth to the study; however, space constraints suggest a more
focused approach. Apart from their intrinsic interest, both Nattes’s and
Stevenson’s texts were selected because they place themselves at significant points in the history of Late Modern English: the nineteenth century was particularly innovative both from a lexicological point of view
and in relation to knowledge dissemination strategies, towards which
journals and travelogues made an important contribution (Dossena
2012b and in preparation).
3. Two milestones
The value of travelogues for the dissemination of knowledge concerning specific areas and cultures is well-known: already in Elizabethan
and Jacobean times the collections published by Richard Hakluyt and
Samuel Purchas5 had provided the reading public with intriguing material concerning distant, exotic places and peoples, encouraging the
development of an adventurous approach to discovery, exploration,
and in fact colonization (Carey & Jowitt 2012). By the beginning of
the nineteenth century travel accounts were a well-established genre, in
which the authors’ comments and observations were not disregarded
as subjective assessments, but were perceived as valuable sources of
reliable information (Dossena 2013).
4 This evokes one of the most notorious entries in Johnson’s Dictionary, i.e. the one
on oats – see Dossena (2014).
5 Several collections were published in a relatively short time span: first of all, Richard
Hakluyt’s Divers Voyages Touching the Discoverie of America, of 1582, and The
Principal Navigations, Voiages, Traffiques and Discoueries of: the English Nation
of 1598–1600. These would then be followed by the works of Samuel Purchas
(Purchas, his Pilgrimage, of 1613; Purchas, his Pilgrim, of 1619; and Hakluytus
Posthumus or Purchas his Pilgrimes, of 1625).
Sublime Caledonia: Description, Narration and Evaluation 181
What is particularly interesting within this framework, then, is the
way in which description and evaluation appear to interact, in order
to make the text both convincing and reliable. The two texts selected
for this analysis, in spite of apparent similarities, place themselves at
opposite ends of a chronological and generic cline. One, Scotia Depicta,
published in London in 1804, is a prototypically illustrated narrative in
which a sequence of 48 sights is presented to the reader with a clearly
defined agenda; the subtitle provides a detailed list of what will be the
object of representation, both in words and in pictures:
With Descriptions, antiquarian, historical, and picturesque.
Robert Louis Stevenson’s Edinburgh: Picturesque Notes, on the
other hand, is a series of essays published in 1879, in which the author
presents his own views and comments on selected traits of the Scottish
capital and its suburbs. As Stevenson was born in Edinburgh, his work
could not be defined as a travelogue strictu senso; however, the author
offers his own subjective views in order to guide readers through a
maze of city lamps, sights and legends with an insider’s knowledge that
intrigues while guaranteeing reliability. Indeed, Stevenson’s own travel
writings take the form of essayistic memoirs in which the journeys provide the framework for the author’s thoughts and reflections.6
While Nattes’s readers are expected to take an interest in what is
majestic, antique and sublime, Stevenson highlights what is appealing
in potentially familiar neighbourhoods. With more than seven decades
separating them and with this different approach to narration, the
two texts (Scotia Depicta, henceforth SD, and Edinburgh: Picturesque
Notes, henceforth PN) may thus provide useful benchmarks for the
identification of informative and persuasive strategies in their linguistic
choices and textual organization.
The books are obviously quite different in many ways: SD discusses
48 images, while Stevenson never refers to the illustrations, which
6 This concerns both his travels around Europe and his experiences crossing first the
Atlantic and then the USA, prior to settling down permanently in Samoa. See www. (accessed July 2014).
182 From Clerks to Corpora
are added on.7 Text length also differs: PN includes 25,612 words,
SD includes 19,628; although this difference might not seem particularly important, it becomes much more considerable when the type/
token ratio is compared: PN has 20.31 vs 16.47 in SD. This finding is
somewhat unsurprising in the light of the different literary skills of the
authors under discussion; nonetheless, it may also be indicative of the
more or less sophisticated approach taken by the two texts. In the next
section a more fine-grained analysis will be offered on a few relevant
4. Findings
Table 1 below presents the absolute and relative frequency with which
selected lexical items occur in SD, PN, and in the nineteenth-century
section of the Corpus of Modern Scottish Writing (henceforth CMSW),
employed as a reference corpus. In the case of adjectives, such items
were selected on account of their evaluative quality, while nouns were
selected on the basis of their relative keyness.
While percentages are too low in PN and CMSW to enable statistical
generalizations, it may be interesting to compare these with the ones in
Table 1. Selected lexical items in the text under investigation and in CMSW.
CMSW (19C only)
6 (0.03%)
11 (0.06%)
10 (0.05%)
12 (0.06%)
39 (0.20%)
18 (0.09%)
3 (0.01%)
29 (0.15%)
3 (0.02%)
47 (0.24%)
6 (0.03%)
14 (0.07%)
2 (0.00%)
3 (0.01%)
2 (0.00%)
4 (0.01%)
6 (0.02%)
5 (0.02%)
6 (0.02%)
3 (0.01%)
1 (0.00%)
2 (0.00%)
85 (0.00%)
31 (0.00%)
337 (0.01%)
396 (0.01%)
97 (0.00%)
108 (0.00%)
48 (0.00%)
82 (0.00%)
92 (0.00%)
166 (0.00%)
120 (0.00%)
667 (0.01%)
7 In the 1879 edition there are 6 etchings and 12 vignettes, while there are 27 illustrations in the 1889 edition (see and
stream/edinburghpicture00stev, accessed July 2014).
Sublime Caledonia: Description, Narration and Evaluation 183
SD, in which scenery and picturesque emerge as recurring items, in line
with the centres of interest indicated in the subtitle. Also ruins appear
to elicit significant interest, which might have been predicted of a text
published in indisputably romantic times.
In what follows a few instances are provided from both SD and PN,
in which descriptive and evaluative elements are seen to co-occur; the
former are italicized, while the latter are in boldface:
(1) The top is surrounded with battlements, which project a foot
beyond the walls, and from the broad shadows formed by a
­declining sun, frequently produce the most picturesque effect.
(SD, Balgonie Castle)
(2) No country is more diversified, adorned, and benefited, by
the different lochs, that are scattered over its surface, than
Scotland; in almost every part of which they produce great
variety of scenery, form a beautiful and picturesque series of
views, and afford a plentiful and cheap article of food. (SD,
(3) Chartered tourists, they make free with historic localities,
and rear their young among the most picturesque sites with a
grand human indifference (PN, ch. 1)
(4) the place is full of theatre tricks in the way of scenery (PN,
ch. 6)
The idea of what is beautiful, picturesque and sublime is supported
in both texts with intertextual references which may be literary, artistic,
or historical. In the Introduction to SD the author reassures readers
that ‘works of authority’ have been consulted for the acquisition of
antiquarian details, and states that “Grose, Pennant, Cordiner, and that
valuable mass of materials comprehended in the Statistical Account
of Scotland, have been carefully examined, as well as numerous other
records”.8 Such sources are intended to provide credibility and make the
descriptions reliable – an important detail meant to increase the book’s
appeal to the reader. In addition, first-hand experience is highlighted:
8 Though SD does not list these sources, apart from John Sinclair’s Statistical Account
of Scotland (1791–99), they are presumably works published in the previous
three decades: Thomas Pennant’s A Tour in Scotland (1771), Charles Cordiner’s
Antiquities & Scenery of the North of Scotland, in a series of letters, to Thomas
Pennant, Esqr. (1780) and Remarkable Ruins, and Romantic Prospects, of North
Britain (1788); and Francis Grose, The Antiquities of Scotland (1789 and 1791).
184 From Clerks to Corpora
Nattes’s drawings are said to have been made while travelling with
Dr. John Stoddart, author of Remarks on Local Scenery and Manners
in Scotland (1801), which is described as “a work of very considerable
merit, in which the author has united great depth of research with a
correct and enlightened taste for the picturesque, the beautiful, and the
sublime” (SD, Introduction).
Stevenson does not appeal to authority, but adds credibility with
personal anecdotes, whether referring to himself or his own family; two
examples are given below:
(5) I look back with delight on many an escalade of garden
walls; many a ramble among lilacs full of piping birds; many
an exploration in obscure quarters that were neither town
nor country; and I think that both for my companions and
myself, there was a special interest, a point of romance, and
a sentiment as of ­foreign travel, when we hit in our excursions on the butt-end of some former hamlet, and found a
few rustic cottages embedded among streets and squares.
(PN, ch. 6)
(6) My father has often been told in the nursery how the devil’s
coach, drawn by six coal-black horses with fiery eyes, would
drive at night into the West Bow, and belated people might see
the dead Major through the glasses. (PN, ch. 4)
Another interesting difference is in the use of geographical labels: it
is tempting to perceive a unionist attitude in SD’s use of ‘North Britain’
as opposed to ‘Scotland’, choosing the term that had come into use after
the Union of Parliaments of 1707 and was sometimes abbreviated as
‘N.B.’. Stevenson, on the other hand, disliked it, as seen in a letter dated
(7) Don’t put ‘N.B.’ on your paper: put Scotland, and be done with
it. Alas, that I should be thus stabbed in the home of my friends!
The name of my native land is not North Britain, whatever may
be the name of yours.
(RLS to S. R. Crockett, c. 10 April 1888,
in Booth & Mehew 1995: 156, original emphasis)
9 I am indebted to Richard Dury, former colleague and current co-editor of the New
Edinburgh Edition of the Collected Works of Robert Louis Stevenson, for the observation of this detail, and for numerous pleasant exchanges on Stevenson’s life and
works over the years.
Sublime Caledonia: Description, Narration and Evaluation 185
Nevertheless, in Stevenson’s text aspects of Scottish history are sometimes treated with much less rhetorical reverence than in SD. Among
these, places associated with the House of Stuart, and particularly with
Mary Queen of Scots, elicit vocabulary that leaves little doubt about
the author’s stance in relation to the people and events at hand; in the
quotation below, for instance, ‘mariolaters’ merges ‘Mary’ and ‘idolaters’, suggesting the irrationality of uncritical appreciation of a controversial historical figure:
(8) On the opposite side of the loch, the ground rises to Craigmillar
Castle, a place friendly to Stuart Mariolaters. (PN, ch. 9)
Nor is this the only instance in which Stevenson’s linguistic choices
express his evaluations; the Reformation and the Covenanters are discussed in approving terms conveying the author’s point of view in ways
as effective as they are unobtrusive:
(9) Down in the palace John Knox reproved his queen in the
accents of modern democracy. […] There, in the Grassmarket, stiff-necked, covenanting heroes, offered up the often
unnecessary, but not less honourable, sacrifice of their lives
(PN, ch.1)
(10) The martyrs’ monument is a wholesome, heartsome spot in
the field of the dead; and as we look upon it, a brave influence comes to us from the land of those who have won their
discharge and, in another phrase of Patrick Walker’s, got
‘cleanly off the stage.’
(PN, ch. 5)
On the other hand, sectarianism is condemned in equally clear terms:
(11) We are wonderful patient haters for conscience sake up here
in the North. […]. Indeed, there are not many uproars in
this world more dismal than that of the Sabbath bells in
Edinburgh: a harsh ecclesiastical tocsin; the outcry of incongruous orthodoxies, calling on every separate conventicler to
put up a protest, each in his own synagogue, against ‘righthand extremes and left-hand defections.’[…]. Shakespeare
wrote a comedy of ‘Much Ado about Nothing.’ The Scottish
nation made a fantastic tragedy on the same subject.
(PN, ch. 4)
186 From Clerks to Corpora
The paragraph closes with an intertextual reference to Shakespeare
and, as a matter of fact, both PN and SD elicit or maintain the reader’s
interest with frequent literary and cultural references. These may be
more or less elliptical, depending on the degree of background knowledge readers may be expected to share. References to Scottish literature,
for instance, may be assumed to be fairly transparent: when Stevenson
writes that Robert Burns “came [to Edinburgh] from the plough-tail, as
to an academy of gilt unbelief and artificial letters” (PN, ch. 1) readers
are assumed to be familiar with the sudden success story of the ‘heaven-taught ploughman’, as Henry Mackenzie would dub him. Similarly,
Robert Fergusson’s unfortunate fate is evoked in the brief statement
that “Burns’s master in his art, […] died insane while yet a stripling.”10
SD also refers to Burns assuming that readers will recognize his
poem The Brigs of Ayr:
(12) On entering this place the most striking objects are the new
and the auld bridges, which Burns has personified with so
much successful humour. The former is handsome and convenient, and was built from a plan of Adams’s, while the
auld brig, if we may believe the poet, is so narrow and bad,
that “twa wheel-barrows tremble, when they meet.” (SD,
The town of Ayr)
Ossian is mentioned as many as four times in SD, and Plate no. 26 is
devoted to ‘Fingal’s cave’, for which Stoddart’s description is quoted in
full, and where reference is made to Banks’s visit of 1772 (see above),
expressing appreciation for the resulting contribution to geological
knowledge and ‘taste’; again, descriptive and evaluative tones merge in
the text:
(13) The entrance is an irregular arch fifty-three feet broad and
one hundred and seventeen high; the interior is two hundred
and fifty in length, and appears still longer from the diminishing perspective. The sides, which are straight, are divided
into pillars; some of those on the east, having been broken off
10The paragraph then continues with Stevenson’s comments on the connection
between Burns and Fergusson, two poets to whom he felt very close, particularly the latter. In a letter to Charles Baxter, dated 18th May 1894, he announced
he wanted to repair the gravestone that Burns had set up for Fergusson in the
Canongate Kirkyard, and wrote: “I had always a great sense of kinship with poor
Robert Fergusson [...]. It is very odd, it really looks like transmigration of souls”
(Booth & Mehew 1995: 8/290).
Sublime Caledonia: Description, Narration and Evaluation 187
near the base, form a passage along that side, by which, with
some difficulty, I reached the farthest end, and seated myself
in a kind of natural throne, formed in the rock. From this
seat, the general effect of the cave appears truly magnificent,
and well calculated to form the eye and taste of a picturesque
architect. The broken, irregular, basaltic roof resembled the
rich ornaments of some grand gothic building; (SD, Fingal’s
From “a kind of natural throne” which the commentator has
reached, the basaltic rock formations give the impression of being “the
rich ornaments of some grand gothic building”. The objectivity with
which size and shapes are described gives way to subjective perception
of something “truly magnificent”.
Personal experience thus proves crucial for the presentation of striking sights and memorable venues. SD and PN, however, appear to take
different approaches to personalization strategies: Table 2 below presents the absolute and relative frequency of first- and second-person
subject pronouns, in order to highlight what subjects appear to take or
be given responsibility for the predication.
These data show that in SD the author appears to prefer an inclusive
use of we, allowing readers to participate in the enjoyment of what
is represented on the page and, consequently, in their imagination.
Stevenson, instead, like the authors in CMSW, stresses the subjectivity
of his representations, using I twice as often as we, but he also appeals
to the reader much more directly, you being the most frequently occurring pronoun in his text. See the examples below:
(14) The harbour was formerly called Slochk Ichopper, meaning
the inlet, where vessels came to barter and sell their fish; and
Table 2. First- and second-person subject pronouns in SD, PN and CMSW.
1PS + 2PS
(19C only)*
17 (0.09%)
24 (0.12%)
6 (0.03%)
98 (0.38%)
49 (0.19%)
126 (0.49%)
12241 (0.23%)
5031 (0.09%)
4008 (0.07%)
* Excluding Verse, Drama, Imaginative prose, and the works of orthoepists.
188 From Clerks to Corpora
we find in the arms of the town a net with a herring and this
motto, Semper tibi pendeat halec. (SD, the Port of Inverary
(15) Into no other city does the sight of the country enter so far;
if you do not meet a butterfly, you shall certainly catch a
glimpse of far-away trees upon your walk; […]. You peep
under an arch, you descend stairs that look as if they would
land you in a cellar, you turn to the back-window of a grimy
tenement in a lane:—and behold! you are face-to-face with
distant and bright prospects. You turn a corner, and there is
the sun going down into the Highland hills. (PN, ch. 6)
As regards language, nineteenth-century travelogues did occasionally
include remarks on differences between English and Scots, but these
were typically seen as a source of puzzlement for English visitors – see
for instance Sinclair (1859: 98–99):
(16) If a Scotch person says, “will you speak a word to me,” he
means, will you listen; but if he says to a servant, “I am
about to give you a good hearing,” that means a severe
scold. The Highland expression for two gentlemen bowing
to each other, amused us extremely on a late occasion, when
a Scotchman said to his friend, “I saw your brother last week
exchange hats with Lord Melbourne in Bond Street!”
Stevenson’s metalinguistic comments, instead, focus more on actual
usage and semantic variety. The section in which he discusses winter
weather shows how culture and environment influence lexical distinctions, to which an interesting touch of perceptual dialectology is added:
the author imagines a cold wind blowing onto his face from distant
hills even as the words are written on the page, thus emphasizing their
evocative quality:
(17) The Scotch dialect is singularly rich in terms of reproach
against the winter wind. Snell, blae, nirly, and scowthering,
are four of these significant vocables; they are all words that
carry a shiver with them; and for my part, as I see them
aligned before me on the page, I am persuaded that a big
wind comes tearing over the Firth from Burntisland and the
northern hills; I think I can hear it howl in the chimney, and
as I set my face northwards, feel its smarting kisses on my
cheek (PN, ch. 9)
Sublime Caledonia: Description, Narration and Evaluation 189
Scots also occurs in snatches of conversation and the names of traditions associated with Hogmanay, i.e. the celebration of New Year’s Day:
(18) For weeks before the great morning, confectioners display
stacks of Scotch bun […] and full moons of shortbread
adorned with mottoes of peel or sugar-plum, in honour of
the season and the family affections. ‘Frae Auld Reekie,’ ‘A
guid New Year to ye a’,’ ‘For the Auld Folk at Hame,’ are
among the most favoured of these devices. (PN, ch. 9)
Stevenson’s knowledge of popular culture is also evident in his
r­eferences to folk lore (Thomas the Rhymer) and ballads (Johnnie
Faa and Sir Patrick Spens are mentioned explicitly). SD, instead, does
not seem to go beyond the Ossian myth – whether this may be indexical of the envisaged readership’s expectations can only be a matter
of conjecture.
5. Concluding remarks
This overview, albeit brief and restricted to a few features, has shown a
greater variety of involvement strategies in PN. While both texts examined here describe places, narrate events, and express authorial stance by
means of recurrent lexical items, in Stevenson’s text literary references
and popular culture appear to mix more freely than in SD, which places
greater emphasis on the ‘romantic’ and ‘sublime’ traits of the landscape
and of the buildings represented in the tables. PN also appeals to readers more directly, using second-person pronouns more frequently and
thus encouraging direct participation in the virtual journey presented
in the text. In SD the constant interaction of words and images conveys
meaning and maintains the readers’ interest; in PN, instead, greater lexical richness, witnessed by a higher type/token ratio, stresses the value
of language as a powerful communicative tool. Though both texts rely
on literary and cultural references, Stevenson’s linguistic skill appears
to encourage readers to move from what is virtual to what is real, from
a “somewhere-else of the imagination” to an experience of place and
identity where they can actually ‘see for themselves’.
190 From Clerks to Corpora
Primary sources
Anon. (1894 [2002]). Victorian Travel on the West Highland Line: By
Mountain, Moor and Loch. Colonsay: House of Lochar.
Booth, B.A. & Mehew, E. (eds) (1995). The Letters of Robert Louis Stevenson,
vol. 6. Yale: Yale University Press.
CMSW, Corpus of Modern Scottish Writing. At
cmsw/, accessed July 2014.
Fontane, T. (1860/1989). Jenseit des Tweed. Bilder und Briefe aus Schottland.
Frankfurt a. M.: Insel.
Nattes, J.C. (1804). Scotia Depicta; or, The Antiquities, Castles, [...] and
Picturesque Scenery of Scotland. […] With Descriptions, Antiquarian,
Historical, and Picturesque. London: Printed by T. Bensley and published
by W. Miller et al. At, accessed July 2014.
Sinclair, C. (1859). Sketches and Stories of Scotland and the Scotch, and
Shetland and the Shetlanders. London: Simpkin, Marshall, and Co.
Stevenson, R.L. (1879). Edinburgh: Picturesque Notes. London: Seely, Jackson
& Halliday. At, accessed July 2014.
Secondary sources
Brown, I.G. (1980). The Hobby-Horsical Antiquary. A Scottish Character
1640–1830. Edinburgh: National Library of Scotland.
Carey, D. & Jowitt, C. (eds) (2012). Richard Hakluyt and Travel Writing in
Early Modern Europe. Farnham: Ashgate.
Dossena, M. (2012a). “A highly poetical language”? Scots, Burns, Patriotism
and Evaluative Language in Nineteenth-century Literary Reviews and
Articles. C. Percy & M. C. Davidson (eds) The Languages of Nation:
Attitudes and Norms. Bristol: Multilingual Matters, 99–119.
———. (2012b). Late Modern English – Semantics and Lexicon. A. Bergs &
L. Brinton (eds) HSK 34.1 – English Historical Linguistics – An International
Handbook. Berlin: De Gruyter, 887–900.
———. (2013). “John is a good Indian”: Reflections on Native American
Culture in Scottish Popular Writing of the Nineteenth Century. C. Sassi &
T. van Heijnsbergen (eds) Within and Without Empire: Scotland Across the
(Post)colonial Borderline. Newcastle u.T.: Cambridge Scholars, 185–199.
———. (2014). The thistle and the words: Scotland in Late Modern English
Lexicography. Scottish Language 31–32 (2012–13), 64–85.
Sublime Caledonia: Description, Narration and Evaluation 191
———. (in preparation). “I tell you this, because I come from your country.”
The Popularization of Science and the Linguistic Construction of Reliability
in Nineteenth-century Travelogues and Ego Documents. Subplenary talk
presented at the 14th ESSE Conference, Košice, 29.08-02.09.2014.
Grenier, K.H. (2005). Tourism and Identity in Scotland, 1770–1914: Creating
Caledonia. Farnham: Ashgate.
Kelly, S. (2010). Scott-land: The Man who Invented a Nation. Edinburgh:
Mitchell, S. (2013). Visions of Britain 1730–1830: Anglo-Scottish Writing and
Representation. Basingstoke: Palgrave Macmillan.
Nimmo, I. (2005). Walking with Murder. On the Kidnapped Trail. Edinburgh:
Youngson, A.J. (1973). After the Forty-five: The Economic Impact on the
Scottish Highlands. Edinburgh: Edinburgh University Press.
———. (1974). Beyond the Highland Line – Three Journals of Travel in
Eighteenth-Century Scotland: Burt, Pennant, Thornton. London: Collins.
11 The Development of Attitudes to Foreign
Languages as Shown in the English Novel
Philip Shaw
Stockholm University
It was in Sweden that his career was finally doomed. For some time
past he had been noticeably silent at the dinner table when foreign
languages were being spoken; now the shocking truth became apparent that he was losing his mastery even of French; many ageing diplomats, at a loss for a word, could twist the conversation and suit
their opinions to their vocabulary; Sir Samson recklessly improvised
or lapsed into a kind of pidgin English.
Evelyn Waugh Black Mischief (1932) chapter 2
1. Introduction
Societies are characterized by the patterns of language knowledge and
language use which are studied in the “sociolinguistics of society”
(Fasold 1984). One aspect of this is the languages people know and use
at a given time and place, part of the local ‘language ecology’ (Haugen
1971). Haugen’s term has been used of all the languages used in an
environment and their mutual relations. Thus Smalley (1994) discusses
all the Tai, Mon-Khmer, Malay and Chinese languages of Thailand and
their status and uses in different environments in the country. In the era
of globalization it can appear that what is missing from his study is the
omnipresence of English (or at least the Latin alphabet) and the place of
that language high up in the hierarchy. Just as alien species find a place
in biological ecology, so do foreign languages play a part in language
Thus discussions of language ecology need to take account of the
languages that are learned in an environment as well as those that
How to cite this book chapter:
Shaw, P. 2015. The Development of Attitudes to Foreign Languages as Shown in the
English Novel. In: Shaw, P., Erman, B., Melchers, G. and Sundkvist, P. (eds) From Clerks
to Corpora: essays on the English language yesterday and today. Pp. 193–214. Stockholm:
Stockholm University Press. DOI: License: CC-BY.
194 From Clerks to Corpora
are spoken as first language. There is quite a large literature on the
history of language teaching and learning in Britain and the rest of
Europe (Hüllen 2006; McLelland 2005). Recently, for example Nicola
McLelland and Richard Smith have launched a co-ordinated research
project exemplified by the papers they have collected in a recent issue
of Language and History (Glück 2014; Besse 2014; Sanchez 2014).
But this type of institutional and method-oriented history tells us little
about the attitudes and assumptions of English-speakers in Britain and
elsewhere to the foreign languages they were taught. These attitudes
and assumptions can, however be inferred from the surviving documents in a straightforward way. The twelfth-century writer Orm (e.g.
Johannesson 2008) decided to write the Ormulum in English and this
implies that he intended to address an audience for whom English was
the primary language. The fact that his sources are in Latin and not
Greek or Hebrew implies that he could read Latin but not the other two
languages. Johannesson (personal communication) reports that when
Orm noticed that he had used a Romance or Latin loanword in his
writing he often replaced it with a Germanic equivalent. This tells us
both that monks like him were familiar with words of this type and that
ordinary lay people could not be assumed to know all of them.
Language attitudes can also be inferred from fiction, since writers
can use the reader’s assumed knowledge of them to characterize their
creations. A well-known example is Chaucer’s Prioress, who spoke
French “After the scole of Stratford-atte-Bowe,/ For Frenssh of Parys
was to hire unknowe” (Robinson 1957: 18). The contemporary reader
had a clear idea of what was meant by Stratford French, and could infer
something about the Prioress’s personality from it. The modern reader
has to infer the language ecology that Chaucer is referring to and might
assume that her variety of French was a learner variety like modern
school French or some kind of vernacular Anglo-Norman patois, or
even an East London dialect. At any rate the reference is often taken
as showing that her French was somehow inferior – if she had known
Paris French she would have used it in preference. By looking at many
documents and their language forms, Rothwell (1985) is able to reconstruct the place of ‘insular French’ in the language ecology of the time.
Interestingly, it was an autonomous second-language variety like Indian
English nowadays – acquired at school but needed in everyday life.
The French used in England from the early thirteenth century to the
end of the fourteenth is the only variety to be on a par with francien
The Development of Attitudes to Foreign Languages as Shown in the English Novel 195
in the sense of being an official language of record widely used by
the dominant classes in a vigorously developing nation. (Rothwell
1985: 47)
Even so, one kind of insular French seems not have had a high status
for Chaucer. Thus references in fiction can illuminate what we know
from other sources and in turn are illuminated by that knowledge.
In this note I attempt to retrieve attitudes and assumptions about
foreign languages from incidental observations made while r­eading
­nineteenth-century novels and then attempt a more systematic investigation of a corpus of eighteenth and nineteenth-century novels. The aim is
to try to infer the wider (foreign) language ecology of these centuries in
Britain as it appeared to contemporaries and to see how the changes that
took place in the nineteenth century are reflected in fiction.
Language knowledge and attitudes are local and class-based. Sailors
know a different set of languages from monks and prioresses – the
oceans have their own language ecology. In the seventeenth century,
alongside French, a poet like Milton knew the classical languages and
Italian but in the repertoire of an adventurer and seaman like Edward
Coxere French was accompanied by Dutch and Spanish (Meierstein
1946). Literary works from the past may be more likely to tell us about
the language repertoires of writers and high-status individuals than
about those languages of the majority on which language ecology studies have focused. In particular, the choice in this chapter of mainstream,
predominantly English, sources means that the focus is on references
to foreign-language knowledge, rather than local low-status ones. I do
not consider references to Greek and Latin, although they are fairly
frequent, of course, and typically gendered as male accomplishments.
As noted, a framework for the investigation is first built up by
describing anecdotal observations from the work of several novelists,
and then the value added by a corpus approach is assessed.
2. Pre-systematic observations
Jane Austen
References to modern language knowledge in Austen’s novels seem to be
confined to French and Italian, although translated German literature is
mentioned (notoriously Lovers’ Vows in Mansfield Park). French seems
to be common knowledge. In Emma (1816) Mr Knightley can comment
196 From Clerks to Corpora
to Emma on the supposed difference in meaning between French aimable and English amiable in words that suggest both are equally familiar
with the language. The impression that girls could be expected to know
the language is strengthened by the rich young Miss Bertrams’ contempt
for their poor cousin Fanny’s ignorance of French (Mansfield Park 1814)
and Austen’s assurance that Fanny learned the language once she had a
governess to teach her (“Miss Lee taught her French1.”).
Italian is different. Vulgar Mrs Elton (Emma) irritates Emma by her
references to her caro sposo, flaunting her knowledge of at least some of
the language, while the thoughtful Anne Elliot (Persuasion 1818) attracts
admiration for the quality of her translation of song texts despite her
modest denial of proficiency. Knowledge of French seems unmarked and
it is ignorance of it that is commented on in Austen, whereas knowledge
of Italian is an accomplishment to be commented on and, by the vulgar,
flaunted, and perhaps particularly a typically female accomplishment.
Charlotte Brontë
Thirty years on or so, it is French and German that figure in Charlotte
Brontë’s novels. Interestingly, there is a good deal of untranslated French
(from Adèle in Jane Eyre (1847) and from various Belgian characters in
Villette (1853)), which presupposes that the reader understands the language rather as Mr Knightley presupposes a knowledge in Emma. Teaching
French is clearly an essential requirement for a governess. Villette suggests
some degree of mutuality in the English-French relation, for the francophone little girls in Mme Beck’s school (in a thinly disguised Brussels) are
learning English from Lucy Snowe, the heroine. Furthermore, German
has apparently replaced Italian as the desirable extra accomplishment. In
Brussels Lucy has a colleague who teaches German, and when Jane Eyre
finds her cousins they start learning this language. This is a hint of the situation around 1900 suggested in Shaw (2005) where everyone educated
read all three of French, German, and English.
One other language is mentioned in Villette, Dutch/Flemish, but it
does not share the same status. One character is so uneducated that she
can only speak Dutch: (“the aboriginal tongue”)
This was no more than a sort of native bonne, in a common-place
bonne’s cap and print-dress. She spoke neither French nor English,
1Given the variety of editions of the classic novels referred to, I do not give page
numbers. All examples can be located by searching the Gutenberg Project electronic
editions available at
The Development of Attitudes to Foreign Languages as Shown in the English Novel 197
and I could get no intelligence from her, not understanding her
phrases of dialect.
Still, it is notable that a local middle-class friend who up to now has
only spoken French (and English) could in fact speak Dutch:
Addressing the aged bonne, not in French, but in the aboriginal
tongue of Labassecour [Belgium], he….
Finally, in Jane Eyre (1847), Hindustani (what is now called Hindi/
Urdu) figures as the language St John Rivers wants Jane to learn rather
than German so that she can join him as a missionary.
Anthony Trollope
Writing perhaps with a wider audience in mind, Trollope never breaks
into French, but the language is often called upon for characterization.
In Can you forgive her? (1865), it is part of Burgo Fitzgerald’s fecklessness that he cannot speak the lingua franca of Baden well enough to
inhibit a switch to English. Plantagenet Palliser has been established as
an exemplary figure, and his French is in keeping.
Burgo, … walked up to him, and, speaking in bad French, desired
him to leave them. “Don’t you see that I have a friend with me?”
“Oh! a friend,” said the man, answering in bad English. “Perhaps de
friend can advance moneys?”
“Misther, Misther!” said the man in a whisper.
“What do you want of me?” asked Mr Palliser, in French.
Then the man spoke in French, also. “Has he got any money? Have
you given him any money?” (Trollope 1938 [1865]: 454, 457)
So a superior person like Plantagenet knows French well. Not to
know French indicates social inadequacy like Burgo’s or lack of education like that of Dorothy Stanbury (He Knew He Was Right 1869),
who says ‘I can’t play, or talk French, or do things that men like their
wives to do.2’
By contrast, by the 1860s and 70s knowing German indexes a kind
of (male) cleverness associated with outsiderhood. Figures who are
described as knowing both French and German well are often young
men who have been educated at the superior German universities.
These are people whose place in England is insecure like Lucius Mason,
2 See below for this association of music and French.
198 From Clerks to Corpora
denouncer of British legal unreason and heir of ill-gotten wealth in
Orley Farm (1862), or Ralph Newton in Ralph the Heir (1871) who is
illegitimate (and “spoke German and French as if they were English”).
This dialogue from The Prime Minister (1876) between the outsider
Lopez and his crustily conventional father-in-law illustrates the point
[Lopez] had been at a good English private school........Thence at the
age of seventeen he had been sent to a German University,
[Lopez:] ‘I was sent to a German university with the idea that the
languages of the continent are not generally well learned in this
[Father-in-law:] ‘I dare say French and German are very useful. I
have a prejudice of my own in favour of Greek and Latin’
[Lopez :] ‘But I rather fancy I picked up more Greek and Latin at
Bonn than I should have got here....’
The alienness of outright scoundrels like Melmotte in The Way We
Live Now (1875) is marked by their not merely knowing French and
German but actually speaking German to people in England. Madame
Max, in several novels including Phineas Redux (1874), is a positive character although she has the same central European capitalist
background, and she is never shown committing this solecism, despite
­having an apparently German maid.
In Trollope’s later novels knowledge of German seems to have
become a normal product of education for girls. As early as 1873 the
Fawn sisters are planning to carry on conversations in French and
(with difficulty) German among themselves (The Eustace Diamonds).
In Ayala’s Angel (1881) two sisters are characterized by their language
knowledge: pretty Ayala “had been once for three months in Paris
and French had come naturally to her”, whereas sensible Lucy knows
“something of French and German, though as yet not very fluent with
her tongue”. An intriguing detail in this connection is that we are told
in passing that Alice Vavassor in the earlier Can you Forgive Her (1865)
was educated in “Aix-la-Chapelle” – Aachen. She must have been sent
to this relatively fashionable stage on the grand tour to learn German.
In the 1860s this aligns her with the young outsider men and relates to
her interest in politics and resistance to conventional female roles.
Knowledge and use of Italian, on the other hand, usually indexes the
deep depravity and often the suspected bigamy of aristocratic insiders –
earls and marquises, mostly. The Stanhope family in Barchester Towers
(1857) are merely the least unpleasant and wealthy of this group. There
The Development of Attitudes to Foreign Languages as Shown in the English Novel 199
is little trace of the status of the language as an accomplishment that
can be seen in Jane Austen.
Only French really worked as an international language of science.
In Orley Farm (1862) there is a big international conference on the
law in Birmingham. Lectures are given in English, French, German and
Italian (and not, for example, Spanish or Dutch or Russian). The characters and the narrator comment on the ineffectiveness of German and
Italian in this context, but do not mention French. Students and practitioners of the law treat French as the unmarked option for international
communication, but baulk at German and Italian.
Two occluded languages are mentioned, and the attitudes expressed
are similar to those of Brontë to Flemish/Dutch. Trollope’s remarkable
first novel The Macdermots of Ballycloran (1847) is set in the west of
Ireland. When Macdermot, a Catholic small landowner, accidentally
kills a policeman in the middle of a panic about terrorism, he makes
the mistake of running away to the wild boys in the mountains. There
he meets an old man who can only speak Irish – but it turns out that
like Brontë’s Belgian character, Macdermot himself spoke the language
in his youth and can communicate effectively. In Phineas Redux (1874)
Madame Max finds a key witness in Prague and brings him over to
London. The narrator observes that he “naturally did not speak English
and unfortunately did not speak German either”. Not to speak English
is normal for a Central European, but not to speak one of the major
languages is a marker of lack of education.
George Eliot
The most famous person not to know German in nineteenth-century
literature is Mr Casaubon in Middlemarch (1872), whose academic
efforts were pointless because he did not read the cutting-edge publications in that language. We are told this by his nephew Ladislaw – an
angry young outsider, like Trollope’s German-speaking Englishmen of
the period.
Matthew Arnold
Arnold is not of course a novelist but his comments fit into the pattern.
In 1864 he observes
How much of current English literature comes into this ‘best that is
known and thought in the world’? Not very much I fear, certainly
less, at this moment, than of the current literature of France or
Germany (“The Function of Literature at the Present Time”)
200 From Clerks to Corpora
As a young-ish intellectual he has access to French and German, but
does not mention Italian. In 1887 he makes it clear that he reads French
as easily as English, but does not make the same claim for German, and
shows that no one can be expected to know Russian or Swedish.
I take Anna Karénine as the novel best representing Count Tolstoi.
I use the French translation.....Anna Karénine is perhaps .. a novel
which goes better into French than into English, just as Frederika
Bremer’s Home goes into English better than into French. (“Count
Leo Tolstoi”)
Mention of language knowledge or learning very often has the purpose
of placing a fictional character, and is rather rarely merely a plot device.
By this I mean, for example, that we are told that St John Rivers is
studying Hindustani because it serves to define his dedication and thoroughness, not because later on a mysterious Indian will appear and the
plot requires someone to understand him. Possible exceptions are the
francophone Belgian and anglophone Irishman who happen to recall
the occluded languages, and thus enable the ignorant figures to play
their part.
The incidental observations give a picture of a nineteenth-century
Britain in which knowledge of French was to be expected of every educated person and it was rather lack of it that was to be commented on.
German became more widely known and developed into a pre-­requisite
for scholarship, but was never a pre-requisite for being considered
­educated. Nor did knowledge of the language correlate with high social
status, somewhat the opposite in fact. Italian seems to have lost status,
so that German rather than Italian is the extra accomplishment. Not to
know one of English, French, or German is a mark of total lack of education. As in Shaw (2005), it is unclear whether or not Italian belongs
to this group of required international languages.
3. Corpus investigation
Incidental observations may be influenced by priming for what one
expects to notice, but corpus investigations only turn up what the
search items reveal. The incidental observations showed the range of
lexical items that occurred when the investigator noted references to
languages by eye and knowledge of this range made it possible to create a list of items for a more extensive search. I compiled a corpus of
The Development of Attitudes to Foreign Languages as Shown in the English Novel 201
English novels by downloading plain-text versions from the Gutenberg
Project (n.d.). 64 novels were taken from 34 novelists (mostly two each)
(see Appendix 1). In addition five less fictional works were included:
Borrow’s autobiographical Lavengro (1851) and Romany Rye (1857),
George Moore’s Confessions of a Young Man (1888), Swift’s satirical
Tale of a Tub (1704) and Defoe’s (or Nathaniel Mist’s) General History
of the Pirates (1724). The aim was to include a wide range of novels but
focus on those which draw on the contemporary reader’s own knowledge of the contemporary language ecology, rather than informing
about the ecology of another time (historical novels) or place (novels
set in exotic locations). The notion of a historical novel proved rather
difficult to operationalize. Is Vanity Fair (1847) historical when it deals
with Waterloo (1815)? And if not is Redgauntlet (1822) historical
when it deals with a fictional uprising of the 1760s? Four novels (those
by Scott and Stevenson) deal with other times and non-metropolitan
spaces, and four (Haggard’s King Solomon’s Mines 1885, Kipling’s Kim
1901, Conrad’s Nostromo 1904, and Doyle’s The Lost World 1912)
are clearly “exotic”. Many others deal at least in part with British
characters travelling in Europe, but these generally focus on the cultural-insider characters rather than the non-British environment. The
most recent writer examined was Virginia Woolf, represented in the
Gutenberg Project by the early works Night and Day (1919) and Jacob’s
Room (1922). There were about 1, 480, 000 words in the sample.
Once the corpus was assembled the program AntConc (Anthony
2007) was used to search for instances of the words listed in Table 1,
which are mainly nationality adjectives/language names. The aim was
however not to observe forms but to use the corpus to find instances
Table 1. Items searched for via Antconc.
(low) Dutch (cf. German)
German/High Dutch
202 From Clerks to Corpora
of reference to foreign languages. The method has obvious weaknesses.
Instances of actual use of French (which seemed quite frequent in
Charlotte Bronte at least) or potentially German or Italian (cf. caro
sposo) could not be captured in this way, nor was any attempt made
to search, for example, for instances of the numerous Indian languages
probably mentioned in Kim. The anecdotal survey just described suggested, however, that a high proportion of relevant text-segments could
be identified in this way. To simplify the task, the numerous references
to Greek (all ancient as far as I could judge), and Latin, were neglected
in the analysis.
All the instances of these words found were then examined by eye and
those which did not refer to language use (the majority) were excluded.
Although the search was for words, the aim was to find instances of
reference to languages, and it is the instances that are classified not
the actual use of the word. So the example “Again, there was the little
French chevalier opposite, who gave lessons in his native tongue” is
classified as a reference to the teaching of French, and the example “She
had a French master, who complimented her upon the purity of her language” is classified as a reference to language quality, even though the
word French does not refer to the language in either case.
Individual speaking or knowing LX
Reference to text/literature in LX?
Reference to quality of or ignorance of LX
Ignoranceof LX?
actually speaking LX?
Meta comment or text ?
Individual teaching or learning LX?
Figure 1. Flow diagram of criteria and categories for classifying instances of
language reference.
The Development of Attitudes to Foreign Languages as Shown in the English Novel 203
The incidental survey suggested several hypotheses. First, French,
German and Italian would predominate among the languages mentioned. Second, there would be rather frequent reference to the quality of
the language used as a means of characterizing fictional figures, and this
would be disproportionately focused on French, since knowledge of this
language was an index of general education. Third, diachronically, where
there was reference to knowledge of or learning a language, instances
would mainly refer to French, German, and Italian, with French predominant throughout the period, and Italian gradually giving way to
German. Finally, other languages would mainly be mentioned in other
contexts than as being spoken or known by characters in the novels.
In order to produce some quantitative support for these hypotheses,
the collected instances of reference to language were examined by eye
and an attempt was made to develop categories likely to illuminate
these hypotheses (Figure 1). While the categorization does indeed support the incidental observations, as I show below, its main value is to
add more examples for qualitative analysis3. As Figure 1 shows, in some
areas the categorization is rather fine, in others quite coarse, reflecting
areas judged to be interesting and others judged to be less so. There
is no suggestion that the categories are mutually exclusive. Thus all
categories are ordered with features judged to be most interesting for
the present study coming first, such that if the relevant feature is found,
the instance falls into that category and potential membership of other
categories is not considered. Five primary criteria were used: the first
was whether or not the reference was to an individual actually knowing or speaking a foreign language. Where this was met cases were first
isolated in which there was reference to someone’s ignorance of the language in question and then those where there was reference to the quality of knowledge or production, and the remaining cases were subdivided into references to production and knowledge. Thus “ ‘X’ she said
in bad French” and “Her French was bad” would be in the QUALITY
category while “ ‘X’ she said in French” and “She knew French” would
be respectively SPEAK and KNOW. The same procedure was followed
for subsequent criteria and categories (denoted by capitalized labels in
Figure 1). Examples are shown in Table 2.
3Thus it turned up the first recorded backpacker English teacher: the Vicar of
Wakefield’s son set off to Holland to teach English to the Dutch with only a ‘satchel’
as his luggage, but was frustrated when he realized that he would need to know
Dutch first; the Direct Method did not occur to him.
204 From Clerks to Corpora
Table 2. Examples from the analytical categories.
IGNORANCE 1 They asked me what I was, in Portuguese, and in Spanish, and
in French, but I understood none of them. Robinson Crusoe
2 “If Mr. Casaubon read German he would save himself a great
deal of trouble.” Middlemarch
LANGUAGE 1 peecoly Rosiny,” says James, in a fine Scotch Italian, “e la piu
bella, la piu cara, ragazza ma la mawd.” The Newcombes
2 was genteel and extremely polite; spoke French well, and danced
to a miracle; Amelia
3 my bungling half-English horrid French, Harry Richmond
1 chattering to each other in the Gaelic. Kidnapped
3 He promptly replied in French, “No. Not this one.” Little Dorrit
3 QUE VOULEZ VOUS? as the French valet said to me on the
occasion. The Absentee
1 And have you learnt French?” “Yes, Bessie, I can both read it
and speak it.” Jane Eyre
2 “Of course I know French,” says the other; “but what’s the
meaning of this?” The Newcombes
1 Percy sings a Spanish seguidilla, or a German lied, or a French
romance The Newcombes
2 Diana sang alone for the credit of the country, Italian and
French songs, Irish also. Diana of the Crossways
1 if he had to sacrifice one it would be the French literature or the
Russian? Jacob’s Room
2 they are nothing more nor less than Chinese writing expressing
something, though what I can’t say Romany Rye
1 Tibby put a marker in the leaves of his Chinese Grammar and
helped them. Howards End
2 kept a school in town, where he taught the Latin, French, and
Italian languages; Roderick Random
3 passed third in algebra, and got a French prize-book at the
­public Midsummer examination Vanity Fair
1 those sensations which the French call the mauvaise honte Amelia
2 might be well called den wild zee, as the Dutch call the sea in a
storm. Robinson Crusoe
3 some gibberish, which by the sound seemed to be Irish Roderick
1 “Do you think French useful in a military education, sir?”
Harry Richmond
2 I declare I’d as soon teach my parrot to talk Welsh Evelina
The Development of Attitudes to Foreign Languages as Shown in the English Novel 205
The example under TEACH/LEARN illustrates the principle: the
reference is actually to a META text, but the higher-ranked category
TEACH/LEARN is chosen because the quotation shows that Tibby is
learning Chinese. The OTHER/META distinction is the least reliable.
Generally, because the categories have not been intersubjectively verified, they are best regarded merely as giving a general idea of the range
of uses. As noted above, the most useful data from the concordance
lines are the added examples for qualitative analysis.
4. Results
Frequency of language references overall
The number of works referring to a particular language is a better
measure than the number of references, since for example, the sixteen
references to Danish all refer to a single item, Borrow’s decision to
learn Danish to study Danish folk-songs (1851). Borrow’s enthusiasm
also accounts for many of the references to Welsh. Similarly there are
Table 3. Overall numbers of references to languages, by occurrences and works.
German/High Dutch
Dutch/Low Dutch
Hindustani etc.
Indian (Amerindian)
% of all works
206 From Clerks to Corpora
s­ everal references to Tibby’s study of Chinese in Howard’s End (1910)
but what is significant is that Forster makes him study that language
(and not Amharic or Hittite, for example), not how many times it is
referred to.
Table 3 gives the numbers of references to languages and the number
and percentage of works referring to languages. Since Erse may refer
to Irish or Scottish Gaelic, the numbers for Irish, Erse, and Gaelic are
merged. High Dutch and German are assumed to be synonyms4.
As expected, references to French predominate. Four out of five
of the works consulted mention the French language, and only half
the next most frequently mentioned language, Italian. As expected,
German and Italian are close to one another in frequency and the rest
are far behind. The four languages that occur in as much as one-sixth
of the works examined are mentioned in very different contexts. Irish/
Erse/Gaelic (both in Scotland and in Ireland) occurs surprisingly frequently but this is often period colour (Kidnapped, Redgauntlet, The
Antiquary) or expression of personal antiquarian interest (Romany Rye,
Lavengro). Otherwise Dutch and Spanish are mentioned mainly in seafaring contexts and Arabic and Chinese in a wide variety. The Swedish,
Norwegian, Serbian, Romanian, Turkish, Siamese and Japanese languages happened not to be mentioned, although there were occurrences
of the words Swedish, etc. with other reference.
Categories of use of various languages
Table 4 shows the number of references to various languages classified by the content of the reference. It is moderately surprising that
languages show relatively similar profiles across the various uses. That
is, for example, the largest numbers are found in the LANGUAGE
QUALITY, SPEAK, TEXT and META columns for French, and also for
German, Italian, and Irish/Erse/Gaelic.
French predominates in all categories, most in references to texts
(French novel, French saying), confirming, perhaps, that texts in French
are regarded as accessible and valuable. It predominates greatly in
judgements about quality of knowledge or performance of language,
but relatively less in statements of outright ignorance. This confirms the
view that some knowledge of French is conceived of as widespread, and
that (as for the Prioress) a social judgement is often made on the basis
of the quality of one’s French.
4 Although by 1873 the meaning of High Dutch was lost and the Fawn sisters in The
Eustace Diamonds use it to mean ‘double-dutch’.
The Development of Attitudes to Foreign Languages as Shown in the English Novel 207
Table 4. Numbers of works containing specific types of reference to languages.
High Dutch/German
Irish Erse/Gaelic
Arabic, Chinese,
Hindustani (etc.),
Danish, Dutch, Russian,
This social significance of French, or the widespread extent of
knowledge and therefore of ascribed ability to discriminate, is shown
by the types of adjective applied to it. In the category QUALITY terms
like pure, excellent, faultless, exquisite, to perfect oneself, are frequently
applied to French but not to German or Italian. On the other hand
tolerable, bad, broken, have a smattering, clumsy apply to all three.
Not only is French more frequently referred to than other languages,
knowing it well is highlighted in a different way.
A suggestive exposure of similar values appears in a few of the
instances turned up by the concordance but unclassified above, where
the reference is to a foreign accent in English. Thackeray refers to “the
sweetest French accent,” “that charming French English” (both The
Newcombes 1855) and “Her pretty French accent” (Vanity Fair 1847),
Gissing to “fluent French-English, anything but disagreeable” (Born
in Exile 1892) and James to “a queer little dialect of French-English”
208 From Clerks to Corpora
Table 5. Percentage of works containing reference to selected languages,
by period.
1 (up to 1830)
2 (after 1830)
(Portrait of a Lady 1881). It sounds as though a French accent can be
judged positively, as further evidence of the social status of the language. Attitudes of this kind to German seem more mixed, by contrast.
Although Thackeray also allows (ironically?) that one can have a good
German-accented French (“said the courier in a fine German French”,
Vanity Fair), a German accent in English annoys George Eliot’s characters (“suddenly speaking in an odious German fashion” (Daniel
Deronda), “some disgust at the artist’s German accent” (Middlemarch)).
Attitudes to accent reflect the high status of French and the somewhat
ambiguous position of German.
Change over time
Chaucer and Spenser read Italian, and Milton wrote in that language. But, as we have seen, by Matthew Arnold’s time, German had
become the literary Englishman’s second foreign language. To show
that this is reflected in the novels, I divided the works in the corpus
into two p
­ eriods – up to 1830 and after 1830. The date was chosen
to put Jane Eyre (1847) learning German clearly in Period 2, and
Anne Elliot (Persuasion 1818) using her Italian in Period 1. Table 4
shows percentages of works referring to French, Italian, and German.
There was indeed a large increase in references to German, which did
indeed become part of general education. From Period 1 (Lady Susan
1793) comes the interesting transitional observation “It is throwing
time away to be mistress of French, Italian, and German, music, singing, and drawing.” References to Italian in the novels did not decrease
correspondingly to the rise in German, but Italian became exotic in a
new way. Characters in novels visit Italy much more than in Period 1
(Portrait of a Lady, Room with a View, Daniel Deronda) and Italian
characters are frequent (Nostromo). In Period 2 references to Italian
arise mainly in references to this travel, while those to German make it
part of a general education.
The Development of Attitudes to Foreign Languages as Shown in the English Novel 209
5. Further Observations
In this section I note, qualitatively, two usages where the special position of French is very noticeable, both of which seem to be commonplaces in earlier novels but less frequent in later ones, and one that
might show the Celtic languages becoming even more occluded.
From “French and dancing” to “French and German”
In texts from Period 1 it is common for French to be mentioned along
with dancing and music as a polite accomplishment, but such references are much less frequent in Period 2.
Moll Flanders (1722) is made to say “in short, I learned to dance
and speak French as well as any of them, and to sing much better”. In Amelia (1751) we can read “the gentleman was genteel and
extremely polite; spoke French well, and danced to a miracle; ……”.
In Humphrey Clinker (1771) we read: “the girl’s parts are not despicable, and her education has not been neglected; that is to say, she can
write and spell, and speak French, and play upon the harpsichord; then
she dances finely, ……” In Evelina (1778) the ill-bred Captain Mirvan
rejects both dancing and French as possibilities for himself: “What, I
suppose you’d have me to learn to cut capers? and dress like a monkey?
and palaver in French gibberish? hey, would you?” French and dancing are part of a non-academic training for polite life suitable perhaps
especially, though not exclusively, for women. Italian does not occur in
these contexts; it is not basic in the same way.
Similar quotations, again not always referring to female education,
can be found from the later nineteenth century. Thackeray writes: “She
could not play on the piano; she could not speak French well; she could
not tell you when gunpowder was invented:…” (The Newcombes
1855), and also “Tom was absent taking his French and drawing
­lesson of M. de Blois.” Quite a bit later George Eliot gives Gwendolen
Harleth’s qualifications to be a governess: “your French, and music,
and dancing – and then your manners and habits as a lady, are exactly
what is wanted” (Daniel Deronda 1876). But in the course of the nineteenth century modern languages became part of an academic syllabus
(Archer 1921) and examples like those above become less frequent. Two
things happen. First, knowledge of German appears as an accomplishment. Ann Brontë describes an unsatisfactory pupil: “everything was
neglected but French, German, music, singing, dancing, fancy-work,
and a little drawing”, …(Agnes Grey 1847). And Charlotte lets a similar character describe herself: “I know nothing--nothing in the world – I
210 From Clerks to Corpora
assure you; except that I play and dance beautifully, – and French and
German of course I know, to speak; but I can’t read or write them very
well.” (Villette 1853). Second, mention of learning French in connection
with artistic and aesthetic skills seems to become less frequent, and the
associations of French seem to be more those suggested by Moore’s
“Neither Latin, nor Greek, nor French, nor History, nor English composition could I learn” (Confessions of a Young Man 1886) – a school
subject like others. It is possible that when Trollope makes Dorothy
Stanbury say “I can’t play, or talk French, or do things that men like
their wives to do.” (He Knew He Was Right 1869) he is characterizing
her as someone whose idea of a sophisticated education itself is a bit
In general, references to learning French decrease and those to the
quality of French knowledge certainly do not increase5. Portrait of a
Lady (1881) has a reference to perfecting one’s French and one to a
character’s imperfect French while the otherwise rather similar The
Good Soldier (1915), also set mainly in continental Europe and very
concerned with social distinction, never refers to the French language.
“As the French have it”
In the corpus it is not uncommon to show knowledge of a language by
citing a word or phrase and naming the source language as in “T, on the
left arm of the dead man, signified the Italian word ‘Traditore,’” (The
Woman in White 1860) or “ ... must be my amende honorable, as the
French have it” (Clarissa 1748). Borrow uses phrases or literal translations from Welsh, Irish, and Chinese like this; otherwise Dutch occurs in
Robinson Crusoe (1719) and Humphrey Clinker (1771), Italian in The
Woman in White (1860) and A Room with a View (1908), and German
in Confessions of a Young Man (1886) and The Adventures of Sherlock
Holmes (1892). French is much more common: no less than fourteen
of the works examined use it in this way, from “those sensations which
the French call the mauvaise honte” in Amelia (1751) to “Bon voyage,
my dear sir – bon voyage, as the French say.” in The Woman in White
(1860). Again the references seem clustered in the earlier works, and to
be less frequent later in the nineteenth century.
5 The social and geographical range covered by the novel as a form also changed, so
that we are not always comparing like with like.
The Development of Attitudes to Foreign Languages as Shown in the English Novel 211
“All Greek to me”
Some languages are marked as incomprehensible, in occasional usages
like “Most of the matter might have been written in Chinese for any definite meaning that it conveyed” (The Lost World 1912) or “as intelligible
to me as if he had spoken in Arabic or Irish” (Roderick Random 1748) or
“I declare I’d as soon teach my parrot to talk Welsh.” (Evelina 1778). These
few examples suggest a greater presence of the Celtic languages in Period
1, but also their status as strange and difficult. No one says “It could have
been French/ German/ Italian/ Spanish for all I understood.”
6. Conclusion
Use of a corpus does indeed confirm the special status of French and the
rise of German. It also confirms that Italian was the only other language
that was widely known or referred to. It confirms that some knowledge
of French was presupposed, perhaps especially later in the period, and
that the quality of one’s spoken French was a significant index of one’s
status as a cultured person, for both sexes, perhaps especially earlier.
We can probably see a shift in the status of modern languages from
polite accomplishments along with dancing and music to school subjects along with History and Geography, and this may be associated
with increasing identification of all education (and not just the classics)
with the institution of the school.
Use of a corpus implies quantification and led me to attempt to
classify references to languages in terms of their content. This proved
difficult to do reliably, but nonetheless produced the unexpected finding that similar proportions of references to most languages fell into
each category. Apart from particular usages (such as the identification of Chinese, Arabic, Welsh and Irish as incomprehensible, or the
­co-occurrence of French with dancing) similar sorts of thing seem to
be said about all the languages mentioned; it is frequency of reference
that differs dramatically, not content. The “linguicist” attitudes towards
minor European languages noted from Brontë and Trollope, the condemnation of those who cannot manage English, French German, or
Italian, are not prominent in the corpus, but on the other hand the
minor languages are not mentioned much at all.
Though knowledge of German is represented as having spread quite
rapidly after, say, 1830, the language never acquired the position of
French, and speaking German could be used for negative categorization in a way that speaking French could not. While French was clearly
212 From Clerks to Corpora
the main lingua franca in Europe in the nineteenth century, the corpus
data perhaps suggest a weakening in its position as a status symbol
from the second half of the nineteenth century. Nonetheless the fate of
Sir Samson in the passage used as an epigraph to this chapter is worth
noting: humiliated by his inability to speak French in Stockholm, he
was exiled to the remotest of outposts. From Chaucer to Evelyn Waugh,
characters’ French defines their status.
Period 1
21 works
Daniel Defoe
Jonathan Swift
Henry Fielding
General History of the Pirates 1724, Moll Flanders
1722, Robinson Crusoe 1719
Gulliver’s Travels 1726, Tale of a Tub 1704
Tom Jones 1749, Amelia 1751
Period 1
21 works
Oliver Goldsmith
Samuel Richardson
Tobias Smollett
Maria Edgworth
Walter Scott
The Vicar of Wakefield 1766
Clarissa 1748
Peregrine Pickle 1751, Humphrey Clinker 1771,
Roderick Random 1748
Camilla 1796, Evelina 1778
Lady Susan 1793, Mansfield Park 1814, Persuasion
Castle Rackrent 1800, The Absentee 1812
Redgauntlet 1824, The Antiquary 1816
Period 2
48 works
Ann Brontë
Charlotte Brontë
Emily Brontë
George Borrow
Anthony Trollope
Agnes Grey 1847
Jane Eyre 1847, Villette 1853
Wuthering Heights 1847
The Newcombes 1855, Vanity Fair 1847
Lavengro 1851, Romany Rye 1857
Framley Parsonage 1861, The Way We Live Now
Sybil 1845, Tancred 1847
Little Dorrit 1857, Our Mutual Friend 1865
Cranford 1851, North and South 1855
Fanny Burney
Jane Austen
Benjamin Disraeli
Charles Dickens
Elizabeth Gaskell
The Development of Attitudes to Foreign Languages as Shown in the English Novel 213
George Eliot
George Gissing
George Meredith
George Moore
Henry James
Oscar Wilde
Rider Haggard
Rudyard Kipling
Thomas Hardy
Wilkie Collins
Robert Louis
Conan Doyle
Middlemarch 1872, Daniel Deronda 1876
Born in Exile 1892, Henry Ryecroft 1903
Diana of the Crossways 1885, Harry Richmond 1871
Esther Waters 1894, Confessions of a Young Man
Portrait of a Lady 1881, What Maisie Knew 1897
The Portrait of Dorian Gray 1890
King Solomon’s Mines 1885
Kim 1901, The Light that Failed 1890
Jude the Obscure 1895, Tess of the Durbervilles 1891
The Moonstone 1868, The Woman in White 1860
Kidnapped 1886, Treasure Island 1881
Adventures of Sherlock Holmes 1892, The Lost World
Period 2
E.M. Forster
Ford Madox Ford
H.G. Wells
Joseph Conrad
A Room with a View 1908, Howards End 1910
The Good Soldier 1915
Ann Veronica 1909, Tono Bungay 1909
Nostromo 1904, The Secret Agent 1902, Under
Western Eyes 1911
Jacob’s Room 1922, Night and Day 1919
Virginia Woolf
Anthony, L. (2007). AntConc 3.2.1w (Windows) [Computer Software]. Tokyo,
Japan: Waseda University. Available from http://www.antlab.sci.waseda.
Archer, R.L. (1921). Secondary Education in the Nineteenth Century. London:
Besse, H. (2014). La SIHFLES, ou vingt-cinq ans d’investigations historiographiques sur l’enseignement/apprentissage du français langue étrangère
ou seconde. Language & History 57:1, 26–43.
Fasold, R. (1984). The Sociolinguistics of Society (Vol. 1). Oxford: Basil
Glück, H. (2014). The history of teaching German as a foreign language in
Europe. Language & History 57:1, 44–58.
214 From Clerks to Corpora
Haugen, E. (1971). The ecology of language. Linguistic Reporter. Supplement
25, 19–26.
Hüllen, W. (2006). Foreign language teaching – a modern building on historical foundations. International Journal of Applied Linguistics 16:1, 2–15.
Johannesson, N.-L. (2008). Icc hafe don swa summ þu badd: An Anatomy of
the Preface to the Ormulum. SELIM: Journal of the Spanish Society for
Medieval English Language and Literature 14, 107–140.
Mclelland, N. (2005). Dialogue and German language learning in the
Renaissance. D.B. Heitsch & J.-F. Vallée (eds) Printed Voices: The
Renaissance Culture of Dialogue. Toronto: University of Toronto Press,
———. (2012). Walter Rippmann and Otto Siepmann as Reform Movement
textbook authors: a contribution to the history of teaching and learning
German in the United Kingdom. Language & History 55:2, 123–143.
Meyerstein, E. (ed.) (1946). Adventures by Sea of Edward Coxere. Oxford:
The Clarendon Press.
Project Gutenberg. (no date)
Robinson, F.N. (ed.) (1957). The Works of Geoffrey Chaucer. Oxford: The
Clarendon Press.
Rothwell, W. (1985). Stratford atte Bowe and Paris. The Modern Language
Review 80, 39–54.
Sanchez, A. (2014). The teaching of Spanish as a foreign language in Europe.
Language & History 57:1, 59–74.
Shaw, P. (2005). The languages of international publication in economics in
1900. U. Melander-Marttala, H. Näslund, I. Bäcklund, & U. Brestam (eds)
Text i Arbete/Text At Work. Uppsala: Institutionen för Nordiska Språk,
Smalley, W.A. (1994). Linguistic Diversity and National Unity: Language
Ecology in Thailand. Chicago: University of Chicago Press.
12 “ Mythonomer”: Tolkien on Myth in His
Scholarly Work
Maria Kuteeva
Stockholm University
By the bye, we now need a new word for the ‘science of the nature of
myths’ since ‘mythology’ has been appropriated to the myths themselves. Would ‘mythonomy’ do? I am quite serious. If your views
are not a complete error this subject will become more important
and it’s worth while trying to get a good word before they invent a
beastly one.
(C. S. Lewis to J.R.R. Tolkien, in Lewis 1988: 255)
1. Introduction
This article is dedicated to Nils-Lennart Johannesson’s life-long interest
in Tolkien. Nils-Lennart shares Tolkien’s scholarly interests in AngloSaxon and Middle English and continues the tradition of the philological school that developed at Oxford in the late nineteenth century
(e.g. Palmer 1965, Shippey 1982). This approach to the study of language in connection to literary and historical texts made a paramount
contribution to the study of historical development of English and the
compilation of what is known today as the Oxford English Dictionary.
For example, Tolkien’s essay on ‘Beowulf: the Monsters and the Critics’
marked a new epoch in the study and appreciation of the poem. For
Nils-Lennart, the study and publication of the Ormulum manuscript
has been the major project of his academic career.
In this article I examine the ideas of myth as expressed in Tolkien’s
scholarly work. What is ‘myth’? The meaning of this word has changed
throughout human history, causing a certain degree of semantic confusion. As a result, today myth is thought of as ‘fiction’ or ‘illusion’.
People generally distinguish between myth as a story about ancient
How to cite this book chapter:
Kuteeva, M. 2015. “Mythonomer”: Tolkien on Myth in His Scholarly Work. In:
Shaw, P., Erman, B., Melchers, G. and Sundkvist, P. (eds) From Clerks to Corpora: essays on the English language yesterday and today. Pp. 215–228. Stockholm: Stockholm
University Press. DOI: License: CC-BY.
216 From Clerks to Corpora
gods and myth as any fictitious narrative. However there is more to
find in ‘myth’. In the OED the word is given the following definitions:
“A purely fictitious narrative usually involving supernatural persons,
actions or events, and embodying some popular idea concerning natural or historical phenomena. Properly distinguished from allegory
and from legend.”; and in generalized use: “an untrue or popular tale,
a rumour”. The second meaning is identified as concerned with “a fictitious or imaginary person or object.” Whether regarding a tale or an
object, one quality seems to be particularly underlined in the above
definitions of ‘myth’, that of falsehood.
It is not only in English but in all European languages that the word
‘myth’ denotes a ‘fiction’, and this connotation goes back some twentyfive centuries. It was in ancient Greece that myth became crucial for the
development of epic poetry, tragedy, comedy and the plastic arts. But, on
the other hand, due to the highly advanced philosophy, ‘myth’ was subjected to a subtle and critical analysis, which contributed to the process
of its demystification. The Sophists interpreted myths as allegories revealing naturalistic and moral truths. Socrates, along with his disciple, Plato,
was rather sceptical about such attempts at explaining mythical narratives (Phaedrus, Republic E378). Nevertheless, they both also pointed
to the existence of “the under-meaning of ancient mythology” (Müller
1881: 580). For Plato, ‘myth’ was a form of knowledge itself, one of
the human ways of knowing the world. Even earlier Ionian ­rationalism
intensely criticised the classic mythology expounded by Homer and
Hesiod. Xenophanes of Colophon (circa 565–470) was the first to reject
mythological concepts of divinity employed by the two poets as “the
fables of men of old” (Spence 1921: 41). From that point, the Greeks
continued to empty muthos of its religious and metaphysical value. In
short, the debate about the origins and validity of myths began in ancient
Greece, the country with which today many people associate the word
mythology. The following sections will focus on Tolkien’s views on myth.
2. J.R.R. Tolkien on Myth
The study of myth was not Tolkien’s chief occupation as a scholar, and in
his writings he largely expresses his personal views and beliefs. Two main
concerns for him are myth as story and the relationship between myth
and history. He touches upon the matter of myth in his famous essay on
‘Beowulf: the Monsters and the Critics’ (1936) and chooses to speak on
a related subject in his Andrew Lang lecture ‘On Fairy-Stories’ (1939).
“Mythonomer”: Tolkien on Myth in His Scholarly Work 217
In ‘Beowulf: The Monsters and the Critics’ Tolkien argues against
definitions of the poem as “a wild folk-tale” or “mythical allegory”. He
asserts that the account of myth as mythical allegory of nature – the
sun, the season, and so forth – is to be discredited. He also sees the term
‘folk-tale’ as misleading. In Tolkien’s opinion, there should not be any
special distinction between myth and folk-tale:
Folk-tales in being, as told – for the ‘typical folk-tale’, of course,
is merely an abstract conception of research nowhere existing – do
often contain elements that are thin and cheap, with little even potential virtue; but they also contain much that is far more powerful, and
that cannot be sharply separated from myth, being derived from it, or
capable in poetic hands of turning into it: that is of becoming largely
significant – as a whole, accepted unanalysed. The significance of
myth is not easily to be pinned on paper by analytical reasoning. It is
at its best when it is presented by a poet who feels rather than makes
explicit what his theme portends; … myth is alive at once and in all
its parts, and dies before it can be dissected. (Tolkien 1983: 15)
Another interesting point is made by Tolkien with regard to the features
of northern and southern mythologies. Tolkien deeply regrets that practically nothing has survived from pre-Christian English mythology. His
general assumption is that it was essentially similar to Norse mythology. In connection with Beowulf, he suggests similarities between
pagan English and Norse mythic imagination in their “vision of the
final defeat of the humane (and of the divine made in its image), and
in the essential hostility of the gods and heroes on the one hand and
the monsters on the other” (Tolkien 1983: 21). Since the status and
significance of monsters are of particular importance in his essay, the
differences between southern and northern mythologies are discussed
in connection with this theme:
We may with some truth contrast the ‘inhumanness’ of the Greek
gods, however anthropomorphic, with the ‘humanness’ of the
Northern, however titanic. In the southern myths there is also
rumour of wars with giants and great powers not Olympian … But
this war is differently conceived. It lies in a chaotic past. The ruling
gods are not besieged, not in ever-present peril or under future doom
… The gods are not allies of men in their war against these or other
monsters … In Norse, at any rate, the gods are within Time, doomed
with their allies to death. Their battle is with the monsters and the
outer darkness. (Tolkien 1983: 25)
218 From Clerks to Corpora
Thus the northern gods degenerated in mythic imagination into the
mighty ancestors of northern kings: “When Baldr is slain and goes to
Hel he cannot escape thence any more than mortal man” (ibid.).
Tolkien admits that, in the case of southern mythology, gods appear
to be more godlike. They are timeless and have no fear of death. Such
a mythology is more likely to contain profound thought behind it. Yet
one of the characteristics of classical mythology was its continuous
development and change. Thus, he claims, it gradually evolved into
philosophy (Greece) or regressed into anarchy (Rome). According to
Tolkien, the northern mythology escaped this destiny by putting monsters in the centre:
It is the strength of the northern mythological imagination that it faced
this problem, put the monsters in the centre, gave them victory but no
honour, and found a potent but terrible solution in naked will and
courage … So potent is it, that while the older southern imagination
has faded for ever into literary ornament, the northern has power, as
it were, to revive its spirit even in our own times. (Tolkien 1983: 25–6)
The power ascribed by Tolkien to northern mythology is also significant for his own mythopoeic work.
‘On Fairy-Stories’ was an Andrew Lang lecture delivered at the
University of St Andrews on 8 March 1939. Lang’s twelve-volume collection of fairy-stories is referred to as having no rivals in “the popularity, or the inclusiveness, or the general merits” in English (Tolkien 1983:
114).1 However, Tolkien entirely disagrees with Lang’s presentation of
the material as intended specifically for children.
‘On Fairy-Stories’ is Tolkien’s most well-known piece of academic
writing. As Richard Purtill remarks it is even “too well known”, although
“the more obvious points Tolkien makes in this essay have been repeated
over and over again, and the subtler points have often been neglected”
(Purtill 1984: 13). One of the latter points is the relationship between
fairy stories and myth. As Tolkien argued in his lecture on Beowulf,
folk-tales are nothing else but derivatives of myth. Folk-tales, in turn,
belong to what he generally defines as fairy-stories which he considers to
be ‘lower mythology,’ the humbler part of mythology. Therefore a great
deal of what Tolkien says about fairy-stories can be applicable to myth.
1 Tolkien is referring to the twelve volumes of different colours published between
1889 and 1910 by Longmans and Green.
“Mythonomer”: Tolkien on Myth in His Scholarly Work 219
Myth in Tolkien’s understanding is, above all, a story. Discussing
the question of its origins, he emphasises that when enquiring into the
origins of ‘fairy-stories’, one should enquire into the origins of the fairy
elements: “To ask what is the origin of stories (however qualified) is
to ask what is the origin of language and of the mind” (Tolkien 1983:
119). The parallel he draws between story (or myth), language, and the
mind gives us clues for understanding Tolkien’s own account of myth.
Before considering the question of origins, Tolkien argues against
the application of certain approaches to the study of fairy-stories. In
fact the fundamental question of the origin of “the fairy element” ultimately leads the enquirer to the same mystery of the origin of story,
language and the mind. There are many elements in fairy-stories that
can be studied regardless of this main question. These kinds of studies,
Tolkien asserts, are usually scientific by intention: “they are the pursuit of folklorists or anthropologists: that is of people using the stories
not as they were meant to be used, but as a quarry from which to dig
evidence, or information, about matters in which they are interested”
(Tolkien 1983: 119). Tolkien expresses particular distaste with scholarly interest in recurring similarities, typical of certain schools: “We
read that Beowulf “is only a version of Dat Erdmänneken”; then “The
Black Bull of Norroway is Beauty and the Beast”, or “is the same story
as Eros and Psyche”, and the Norse Mastermaid . . . is “the same story
as the Greek tale of Jason and Medea”” (ibid.). For him, such comparisons appear to be pointless whether in art or literature. As he remarks
in his essay on Beowulf: “myth is alive at once and in all its parts, and
dies before it can be dissected” (Tolkien 1983: 15).
Tolkien admits that the fascination of the desire to explain the history of the evolution of stories is very strong in himself. He expresses
certain reservations regarding this kind of investigation of the Tree of
It is closely connected with the philologists’ study of the tangled
skein of Language, of which I know some small pieces. But even with
regard to language it seems to me that the essential quality and aptitudes of a given language in a living moment is both more important
to seize and far more difficult to make explicit than its linear history.
I feel that it is more interesting, and also in its way more difficult,
to consider what they are, what they have become for us, and what
values the long alchemic processes of time have produced in them.
In Dasent’s words I would say: “We must be satisfied with the soup
220 From Clerks to Corpora
that is set before us, and not desire to see the bones of the ox out of
which it has been boiled.”2 (Tolkien 1983: 120)
Tolkien believes that to find out how the whole picture is formed is
often too challenging a task. What can usually be done is to explain one
particular element or detail (like a word in language).
Thus Tolkien believes it unworthy to go into inquiries concerning
the history of stories. Fairy-stories are ancient indeed, and are found
wherever there is a language. As in archaeology or comparative philology, the debate arises “between independent evolution (or rather
invention) of the similar; inheritance from a common ancestry; and
diffusion at various times from one or more centres” (Tolkien 1983:
121). Invention is considered by Tolkien to be the most important and
fundamental, and, at the same time, the most mysterious factor of these
three: “To an inventor, that is to a storymaker, the other two must in
the end lead back” (ibid.). Thus Tolkien shows that both diffusion (borrowing in space) and inheritance (borrowing in time) are ultimately
dependent on invention.3
Tolkien dismisses Max Müller’s account of mythology as a “disease of language”,4 although, he admits, language “may like all human
things become diseased” (Tolkien 1983: 121). He emphasises how intimately linked the origin of language, the mind, and mythology are:
You might as well say that thinking is a disease of the human mind. It
would be more near the truth to say that languages, especially modern European languages, are a disease of mythology. But Language
cannot, all the same, be dismissed. The incarnate mind, the tongue,
and the tale are in our world coeval. (Tolkien 1983: 122)
What creates the fairy element, typical of fairy-story and myth, is the
human art of ‘Sub-creation’. The latter, in turn, is based upon a unique
admixture of human capacity for abstract thought and generalisation,
for distinguishing objects from their qualities (e.g. green-grass). Indeed,
2 Tolkien refers to Sir George Dasent’s translation of Popular Tales from the Norse,
collected by P. C. Asbjörsen and J. I. Moe (Edinburgh, 1859), p. xviii.
3On Tolkien’s understanding of invention as ‘discovery’ see Shippey, The Road to
Middle-Earth, p. 22.
4 In the second series of his Lectures on the Science of Language (1864) Müller dedicates five chapters to the discussion of myth. Here he states: “Mythology, which
was the bane of the ancient world, is in truth a disease of language. A mythe means
a word, but a word which, from being a name or an attribute, has been allowed to
assume a more substantial existence’’(p. 12, emphasis added).
“Mythonomer”: Tolkien on Myth in His Scholarly Work 221
the invention of the adjective is seen by Tolkien as a great step in the
evolution of mythical grammar: “When we can take green from grass,
blue from heaven, and red from blood, we have already an enchanter’s
power - upon one plane; and the desire to wield that power in the world
external to our minds awakes” (Tolkien 1983: 122). Tolkien believes
that this is how by ‘fantasy’ a new form is made, and a human being
becomes a sub-creator.
Tolkien points out that, as an aspect of mythology, sub-creation has
been given too little consideration.5 Rather, a good deal of attention
has been spent on examining representation or symbolic interpretation
of various phenomena of the real world. Thus certain opinions divide
mythology into ‘lower’ and ‘higher’, distinguishing ‘Faërie’, “the realm
or state in which fairies have their being” (Tolkien 1983: 113) from
Tolkien rejects the allegorical explanation of myths which claims that
‘nature-myths’ are based upon personification of natural phenomena
(the sun, the dawn, and so forth) and that fairy tales are simply a sort
of debased mythology. This seems to Tolkien “the truth almost upside
down” (Tolkien 1983: 123). Indeed, the closer the so-called ‘naturemyth’ is to its prototype, the less interesting it is as a myth offering an
explanation of the real world. Any ‘personality’ can only be obtained
from a person. Thus Tolkien concludes:
The gods may derive their colour and beauty from the high splendours
of nature, but it was Man who obtained these for them, abstracted
them from sun and moon and cloud; their personality they get direct
from him; the shadow or flicker of divinity that is upon them they
receive through him from the invisible world, the Supernatural.
There is no fundamental distinction between the higher and lower
mythologies. Their peoples live, if they live at all, by the same life,
just as in the mortal world do kings and peasants. (ibid.)
Tolkien illustrates this point by the example of the Norse god Thórr.
His name means ‘thunder’, but his character and appearance cannot
originate in thunder or lightening. He looks more like a not particularly
clever, red-bearded, and remarkably strong Northern farmer, the kind
by whom Thórr was beloved. So, what came first, Tolkien asks, the
thunder or the farmer? It seems to him that “the farmer popped up in
5 Zettersten (2011) devotes an entire chapter ”Tolkien’s double worlds” to the subject of sub-creation.
222 From Clerks to Corpora
the very moment when Thunder got a voice and face; there was a distant growl of thunder in the hills every time a story teller heard a farmer
in a rage” (Tolkien 1983: 124).
Another important point in Tolkien’s discussion of mythology is
its relation to religion. This is where its ‘higher’ element comes from.
Tolkien refers to the view, previously expressed by Andrew Lang, that
mythology and religion are two different things. Although they have
become confused, mythology as such does not have any religious significance. Tolkien largely agrees with this distinction, although he admits
that in fact religion and mythology have become entangled. He does
not deny a possibility that they could be separated long ago, but since
then, through error and confusion, came back to re-fusion.
In Tolkien’s view, confusion itself appears to be a significant factor in
the process of making mythical or fairy-story characters. Using Dasent’s
image of the ‘Soup’, he talks about the continuously boiling ‘Cauldron
of Story’ (Tolkien 1983: 125). Thus many famous historical ­characters
find themselves “thrown” into the Cauldron, from which they re-­appear
changed into mythical characters. The most obvious example of this
kind is king Arthur, but Tolkien also mentions Charlemagne’s mother,
the Archbishop of Canterbury, Froda, King of the Heathobards, and
others. His conclusion is this: “History often resembles ‘Myth’, because
they are both ultimately of the same stuff” (Tolkien 1983: 127). This
notion is anticipated in ‘The Monsters and the Critics’.
Above all, Tolkien is concerned with the question: What effect is produced by fairy-stories now? Myths and fairy-stories are old and therefore appealing but there is more to them:
Such stories have now a mythical or total (unanalysable) effect, an
effect quite independent of the findings of Comparative Folk-lore,
and one which it cannot spoil or explain; they open a door on Other
Time, and if we pass through, though only for a moment, we stand
outside our own time, outside Time itself, maybe. (Tolkien 1983:
It is because of this ‘timeless’ effect that old elements in myths and
fairy-stories have been preserved. Even though some ancient elements
in such stories are often dropped out, replaced, or changed by their oral
narrators, the essential mythical elements survive because of their felt
literary significance. For example, when a myth has an explanation of
some ritual or taboo which no longer means anything, the value of the
story itself still depends upon this ritual or taboo.
“Mythonomer”: Tolkien on Myth in His Scholarly Work 223
To sum up Tolkien’s argument, “There is no fundamental distinction
between higher and lower mythologies” (Tolkien 1983: 123), regardless of whether myth and folk-tale speak of real persons and things,
or whether they are merely a product of human imagination. Since he
speaks of fairy-stories and myths as stories enjoyed by people today,
the main criterion for defining a good story is how well it temporarily
convinces the reader of the imaginary world of the story, with its own
standards of truth. Thus the secondary worlds of fairy-stories, as products of sub-creation, require from the reader a special kind of Secondary
Belief. The ‘primal desire’ for Faërie is often constrained in the modern
reader by religion (or lack of it) and science. This was not the case for
pre-Christian pagans who lived under the ‘spell’ cast by myths which
led them to a more or less permanent state of secondary belief.
As a philologist, Tolkien was aware of the historical relationship
between spell and evangelium. As Shippey notes:
… for the Old English translation of Greek evangelion, ‘good news’,
was gód spell, ‘the good story’, now ‘Gospel’. Spell continued to
mean, however, ‘a story, something said in formal style’, eventually
‘a formula of power’, a magic spell. The word embodies much of
what Tolkien meant by ‘fantasy’, i.e. something unnaturally powerful (magic spell), something literary (a story), something in essence
true (Gospel). At the very end of his essay he asserts that the Gospels
have the ‘supremely convincing tone’ of Primary Art, of truth - a
quality he would also like to assert, but could never hope to prove of
elves and dragons. (Shippey 1982: 47)
Tolkien finishes ‘On Fairy-Stories’ with a discussion of the fairystory element in the Gospels. They contain “a story of a large kind
which embraces all the essence of fairy-stories” (Tolkien 1983: 155).
3. Pretence or Belief?
‘On Fairy-Stories’ was written for a non-specialised audience, and is
therefore the least ‘philological’ of Tolkien’s scholarly works. This lack
of philological core, as well as the comparatively popular style of the
essay, has caused Shippey to remark that there is a sign that Tolkien
tried to “talk down” to his audience, “pretending that fairies are real”
(Shippey 1982: 45). On the other hand, if there had been an element of
‘pretending’, it could have also been intended to keep a form of seriousness about the works of sub-creation. As noted earlier, Tolkien believed
that in order to create secondary belief in the modern reader, there
224 From Clerks to Corpora
must be no laughing at the magic: “That must in that story be taken
seriously, neither laughed at nor explained away” (Tolkien 1983: 114).
It seems that the merits of ‘On Fairy-Stories’, and its consequent
popularity, are rooted in Tolkien’s less formally academic and more
personal approach to the subject. The essay is invaluable for shedding
light upon his views on myth and language. After all, philology was
not the only, although surely the major, source of Tolkien’s inspiration.
‘On Fairy-Stories’ seems to be written more by Tolkien-the artist and
Tolkien-the believer, rather than Tolkien-the philologist or Tolkien-the
scholar. As he remarked in 1951 to Milton Waldman:
I am not ‘learned’* in the matters of myth and fairy-story … for in
such things (as far as known to me) I have always been seeking material, things of a certain tone and air, and not simple knowledge …
Myth and fairy-story must, as all art, reflect and contain in solution
elements of moral and religious truth (or error) but not explicit, not
in the known form of the primary ‘real’ world. (I am speaking, of
course, of our present situation, not of ancient pagan, pre-Christian
days). (Carpenter & Tolkien: 144)
* Though I have thought about them a good deal.
Later in the same letter he again recalls the theme of the Fall which he
touches upon in ‘On Fairy-Stories’:
After all, I believe that legends and myths are largely made of ‘truth’,
and indeed present aspects of it that can only be received in this
mode; and long ago certain truths and modes of this kind were discovered and must always reappear. There cannot be any ‘story’ without a fall – all stories are ultimately about the fall – at least not
for human minds as we know them and have them. (Carpenter &
Tolkien: 147)
For Tolkien, myth is a form of art which inexplicitly transmits elements
of truth. He strongly disagrees with the confusion of myth with allegory and the explanation of pagan gods as ‘personifications’ of natural
phenomena. On the other hand, he also does not draw any clear distinction between mythology and folklore. Myth is interesting to Tolkien
as a story with literary merits. In his opinion, the Tree of Tales should
be left to be enjoyed rather than examined.
4. Myth: thought, language, and story
As mentioned in Section 1, Tolkien disagrees with the account of myth
as an allegory of nature. The debate concerning this problem started in
“Mythonomer”: Tolkien on Myth in His Scholarly Work 225
ancient Greece when Socrates and Plato argued against the Sophists’
explanation of myths as revealing naturalistic and moral truths (Spence
1921: 40–43). During the period of the Enlightenment, myth was interpreted as a lack of rationality, as a “defective understanding of scientific
causes”.6 One of the results of this “defective understanding” is the
account of myth based on the personification of natural forces.
In the nineteenth century, Schelling (e.g. 1856) rejects the principle
of allegory and turns to the problem of symbolic expression in myths.
However, later Müller (e.g. 1881) returned to explaining myths by natural phenomena, although this time his conclusions were supported with
reference to philological studies. Müller’s explanation was attacked by
Andrew Lang and by the time of Tolkien’s scholarly activities was virtually discredited. Tolkien also rejects Müller’s account of myth (see
Section 1).
Let us briefly consider the link between myth, language and thought,
as understood by nineteenth-century scholarship. Max Müller, for example, claims that mythology is “the dark shadow which language throws
on thought, and which can never disappear till language becomes altogether commensurate with thought, which it never will” (Müller 1881:
590). Tolkien, however, does not derive one phenomenon from another.
Objecting to the account of myth as a “disease of l­ anguage”, he asserts:
“It would be more near the truth to say that languages, especially modern European languages, are a disease of mythology” (Tolkien 1983:
122). It is clear that Tolkien is being ironic here but his claim echoes
Schelling’s definition of language as ‘verblichene Mythologie’ (‘faded
mythology’) (Schelling 1856: 52).
In spite of his scepticism regarding attempts to discover the origins of
myth, Tolkien points out three significant factors: invention, inheritance
and diffusion, of which the first is considered as the most fundamental.
The fairy element, typical of myth, is the product of the human art of
sub-creation, or fantasy. Tolkien’s theory of sub-creation and the parallels
found in the ideas of Coleridge, Grundtvig, and other Romantics, have
been discussed by a number of scholars (e.g. Agøy 1995, Seeman 1995).
When Tolkien discusses myth as story in its present form, he equates
it with what is often defined as ‘folk-tale’. He refers to these two kinds
of narrative as ‘higher’ and ‘lower’ mythologies. It is noteworthy that
instead of looking for certain criteria to distinguish myth from other
kinds of stories, Tolkien rather draws a link between mythology and
6J. W. Rogerson, ‘Slippery Words: V. Myth’, The Expository Time 90 (October
1978), 11.
226 From Clerks to Corpora
what is usually regarded as folklore. This type of approach has a few
analogies in both nineteenth and twentieth centuries, the most obvious
of which lie in the ideas of Jacob Grimm.
Grimm’s treatment of myth and mythology implies a very close
connection with folklore, in particular with the tales and legends of
Germany, and surely with the comparative study of languages. Since
much of Germanic mythology is irrecoverably lost, in order to dig out
Germanic divinities and beliefs, Grimm turns to the study of written
and oral testimonials, especially legends, fairy tales, and superstitious
beliefs, to runes and early language. As a result, Grimm is held “responsible for the development by which mythologic study has become the
study mainly of folklore” (Feldman & Richardson 1972: 410). In his
Preface to Teutonic Mythology he asserts:
... these numerous written memorials have only left us sundry bones
and joints, as it were, of our mythology, its living breath still falls
upon us from a vast number of Stories and Customs, handed down
through lengthened periods from father to son … Oral legend is to
written records as the folk-song is to poetic art, or the ruling recited
by schöffen (scabini) to written codes. But folk-tale wants to be
gleaned and plucked with a delicate hand. (Grimm 1883, volume
III: xiii)
In this Preface Grimm analyses fairy-tales as different from folk-tales
(pp. xiv–xv), as Tolkien does in ‘On Fairy-Stories’. He finally comes to the
definition of the fairy-tale in relationship to other forms of common lore.
Although Grimm does not point out the existence of ‘higher’ and ‘lower’
mythologies, he intimately links myths with folk tales and legendarium.
Tolkien is careful in his use of the words ‘myth’ and ‘mythology’, often
preferring the word ‘legend’ and ‘legendarium’, particularly when referring to his own work (Stenström 1995: 310). Both Grimm and Tolkien
are especially enthusiastic about the vivid tales of the North as opposed
to the ones of the South. Whereas Grimm is concerned with the reconstruction of Germanic heritage, Tolkien in particular regrets the loss of
Anglo-Saxon myths (for further details see Kuteeva 1999, chapters 3–5).
Tolkien’s understanding of myth demonstrates strong adherence
to the symbolic interpretations of this phenomenon, which allies him
with the Romantic movement. In other words, Tolkien can be seen as
a ‘re-mythologiser’ of the ideas ‘de-mythologised’ in the second half
of the nineteenth century (e.g. Müller), and his views also appear to
be strongly anti-modernist. This is particularly obvious in Tolkien’s
“Mythonomer”: Tolkien on Myth in His Scholarly Work 227
creative writings, in which he weaves together language and myth in
order to construct his own mythology dedicated to England.
Agøy, N. I. (1995). Quid Hinieldus cum Christo? - New Perspectives on
Tolkien’s Theological Dilemma and his Sub-creation Theory. P. Reynolds
& G. GoodKnight (eds) The Proceedings of the J. R. R. Tolkien Centenary
Conference. Milton Keynes and Altadena: Tolkien Society, 31–38.
Asbörsen, P. C. & Moe, J. I. (1859). Popular Tales from the Norse, trans.
Sir George Dasent. Edinburgh: n/p.
Carpenter, H. & Tolkien, C. (eds) (1981). The Letters of J.R.R. Tolkien.
London: George Allen & Unwin.
Feldman, B. & Richardson, R. D. (eds) (1972). The Rise of Modern Mythology,
1680–1860. Bloomington, London: Indiana University Press.
Grimm, J. (1882–1883). Teutonic Mythology, trans. from the fourth edition,
notes and appendix by James Steven Stallybrass, 4 Vols. London: George
Bell & Sons.
Herder, J. G. (1833). The Spirit of Hebrew Poetry, trans. J. Marsh, 2 vols.
Burlington: n/p.
Kuteeva, M. (1999). Scholarship and Mythopoeia: The Ideas of Language
and Myth in the Works of Owen Barfield, C.S. Lewis and J.R.R Tolkien.
Unpublished PhD thesis. University of Manchester.
Lewis, W. H. (ed.). (1988). Letters of C. S. Lewis (2nd ed.) revised and enlarged
by Walter Hooper. London: Fount Paperbacks.
Mitchell, B. (1995). J. R. R. Tolkien and Old English Studies: An Appreciation.
P. Reynolds & G. GoodKnight (eds) The Proceedings of the J. R. R. Tolkien
Centenary Conference. Milton Keynes and Altadena: Tolkien Society,
Müller, F. M. (1861). Lectures on the Science of Language: First Series. London:
Longman, Green, Longman, and Roberts.
———. (1864). Lectures on the Science of Language: Second Series (London:
Longman, Green, Longman, Roberts, & Green.
———. (1881). Selected Essays on Language, Mythology and Religion.
London: Longmans, Green and Co.
Palmer, D. J. (1965). The Rise of English Studies: An Account of the Study
of English Language and Literature from its Origins to the Making of the
Oxford English School. Oxford: Hull University Press.
228 From Clerks to Corpora
Purtill, R. (1984). J.R.R. Tolkien: Myth, Morality, and Religion. San Francisco:
Harper and Row Publishers.
Rogerson, J. W. (1978). Slippery Words: V. Myth, The Expository Time 90
(October 1978), 10–14.
Schelling, F. W. (1856). Einleitung in die Philosophy der Mythology, in
Sämmliche Werke, I. Stuttgart and Augsburg: J. Verlag.
Seeman, C. (1995). Tolkien’s Revision of the Romantic Tradition. P. Reynolds
& G. GoodKnight (eds) The Proceedings of the J. R. R. Tolkien Centenary
Conference. Milton Keynes and Altadena: Tolkien Society, 73–83.
Shippey, T. A. (1992). The Road to Middle-earth, 2nd enlarged edition.
London: HarperCollins.
Spence, L. (1921). An Introduction to Mythology. London: G. Harrap.
Stenström, A. (1995). A Mythology? For England. P. Reynolds & G. GoodKnight (eds) The Proceedings of the J. R. R. Tolkien Centenary Conference.
Milton Keynes and Altadena: Tolkien Society, 310–314.
Tolkien, J.R.R. (1983). The Monsters and the Critics and Other Essays.
London: Allen & Unwin.
Zettersten, A. (2011). J.R.R. Tolkien’s Double Worlds and Creative Processes:
Language and Life. New York: Palgrave Macmillan.
13 Reflections on Tolkien’s Use of Beowulf
Arne Zettersten
University of Copenhagen
Beowulf, the famous Anglo-Saxon heroic poem, and The Lord of the
Rings by J.R.R. Tolkien, “The Author of the Century”, 1 have been thoroughly analysed and compared by a variety of scholars.2 It seems most
appropriate to discuss similar aspects of The Lord of the Rings in a
Festschrift presented to Nils-Lennart Johannesson with a view to his own
commentaries on the language of Tolkien’s fiction. The immediate purpose of this article is not to present a problem-solving essay but instead
to explain how close I was to Tolkien’s own research and his activities in
Oxford during the last thirteen years of his life. As the article unfolds, we
realise more and more that Beowulf meant a great deal to Tolkien, culminating in Christopher Tolkien’s unexpected edition of the translation
of Beowulf, completed by J.R.R. Tolkien as early as 1926.
Beowulf has always been respected in its position as the oldest
Germanic heroic poem.3 I myself accept the conclusion that the poem
came into existence around 720–730 A.D. in spite of the fact that there
is still considerable debate over the dating. The only preserved copy
(British Library MS. Cotton Vitellius A.15) was most probably completed at the beginning of the eleventh century.
1 See Shippey, J.R.R. Tolkien: Author of the Century, 2000.
2 See Shippey, T.A., The Road to Middle-earth, 1982, Pearce, Joseph, Tolkien. Man
and Myth: A Literary Life, 1998, Drout, Michael D.C., Beowulf and the Critics,
3See, for example, Alexander, Michael, Beowulf: A Glossed Text, 1995, and
McNamara (ed.), Beowulf, 2005.
How to cite this book chapter:
Zettersten, A. 2015. Reflections on Tolkien’s Use of Beowulf. In: Shaw, P., Erman, B.,
Melchers, G. and Sundkvist, P. (eds) From Clerks to Corpora: essays on the English
language yesterday and today. Pp. 229–238. Stockholm: Stockholm University Press.
DOI: License: CC-BY.
230 From Clerks to Corpora
1. Personal memories of Tolkien and biographical background
I met and was in close contact with Professor Tolkien during the last
13 years of his life. I worked and published Ancrene Wisse texts for the
Early English Text Society, Oxford University Press, together with an
international group of Tolkien-inspired scholars. Tolkien had edited the
most important of the Ancrene Wisse texts (MS. Corpus Christi College
Cambridge 402) in 1962, and I later edited three of the other MSS. In
that period I sometimes got first-hand evidence of his views on both the
anonymous Beowulf and his own The Lord of the Rings. I often wondered why Tolkien was so keenly interested in discussing scholarly matters with me as a much younger person. I soon realized that the fact that
I represented the Nordic countries and was able to pronounce the languages and also some of the dialectal variants made a great impression
on him. Once I selected a text from the Gospel according to St. Mark
and read it in Danish, Norwegian and Swedish, as well as in Gothic. The
Gothic text was taken from Wulfila’s translation of the new Testament
into Gothic, probably produced at Ravenna but now kept as the magnificent Silver Bible from c. 550 in Uppsala, written with silver and gold ink
on purple-coloured parchment. I could understand Gothic fairly well but
Tolkien could speak the language, and amazingly enough, he was able to
construct words in Gothic that would have been regarded as real, if the
corresponding texts and contexts could have been expected in written
form. I should like to regard this as a kind of oral emendation. It may
be seen as one of Tolkien’s most remarkable gifts as a linguistic scholar.
Tolkien was extremely fond of telling stories, remembrances and
comparisons between different types of medieval literature. His interest
in all literature after Shakespeare was said to be next to negligible. The
meetings I had with Tolkien in his homes, first in Sandfield Road outside
Oxford and later in his flat in Merton Street, close to Merton College,
were structured according to one and the same pattern. Tolkien did
most of the talking and I was the attentive listener, with the option of
asking questions at irregular intervals. His eloquent talks were similar
to well-structured animated oral essays.
An interesting point about Tolkien’s time as a new professor at
Oxford is the fact that it coincides rather well with C.S. Lewis becoming
a tutor in English and at the same time a fellow of Magdalen College.
The two of them were to mean enormously much to each other as
colleagues and friends but also as critics and competitors. They each
became a springboard for the other in questions that extended from a
dry syllabus to Nordic mythology, from marriage to the existence of a
God, from Tolkien’s The Lord of the Rings to Lewis’s Narnia.
Reflections on Tolkien’s Use of Beowulf 231
Lewis’s note in his diary from 13 May 1926 concerning Tolkien
(pp. 392–93), before they had become good friends, is a good example
of Lewis’s combination of mental agility and wit:
Tolkien managed to get the discussion round to the proposed English
Prelim. I had a talk with him afterwards. He is a smooth, pale, fluent
little chap—can’t read Spenser because of the forms—thinks the language is the real thing in the school—thinks all literature is written
for the amusement of men between thirty and forty—we ought to
vote ourselves out of existence if we were honest—still the soundchanges and the gobbets are great fun for the dons. No harm in him:
only needs a smack or so. His pet abomination is the idea of ‘liberal’
studies. Technical hobbies are more in his line.
When Tolkien had been established as professor and scholar at
Oxford after 1925, he started on a new phase of his career, devoting
much time to the critical essay. He incorporated this type of essay as
a genre in English philological contexts. Several of his essays from the
end of the 1920s to the end of the 1950s, through all his active life as
professor, point back to important lectures that he was invited to give
both in and outside Oxford. This genre in Tolkien’s writing is the very
core of his scholarly achievement. His philological activity with his editions of Sir Gawain and the Ancrene Wisse, his word studies and his
translation of Middle English poetry into Modern English are examples
of his solid learning, but it is within the art of essay-writing that he can
introduce innovative ideas and new results.
Tolkien managed to develop this part of his activities into a kind of
mastership. An example is the essay from 1929, which describes how
he identifies the new literary language from 13th-century England, the
so-called AB-language. There are also, besides “Beowulf: The Monsters
and the Critics”, important essays such as “A Secret Vice” (1931), “On
Fairy-Stories” (1939), “On Translating Beowulf” (1940), “Sir Gawain
and the Green Knight” (1953) and “English and Wales” (1955).
I realized that I had found the origin of this art of essay-writing when
I first set eyes on the series of essays and summarizing notes kept in the
note-books in the Tolkien Collection from 1913 at the Bodleian Library.
Naturally, these mature essays are the result of the tradition of essay-writing characterizing the whole of the English school system. On top of that,
the tradition at the universities was that the students wrote essays for a
tutor every week, which were read, analyzed and criticized in every detail.
The Tolkien material kept at the Bodleian Library is considerably
larger than what one can imagine here from a brief summary. If we add
232 From Clerks to Corpora
the collection at Marquette University, Milwaukee, Wisconsin, we end up
with a great number of texts, which may be used for analyzing the whole
background of Tolkien’s interest in medieval literature and his ability to
write analytical essays within Old and Middle English philology.
Tolkien always felt an irresistible eagerness to throw himself into
new research projects, new languages and new problems. One way
of learning about his intellectual curiosity is to read through his own
preserved letters and notes. In the Tolkien Collection at the Bodleian
Library, under item A 21/1-12, there are among other things a series
of essays or philological annotations from the period May–June, 1913,
and later. Nr A 21/1 contains essays within the following varied areas:
1.Gradation [vowel-changes, for example from ‘i’ or ‘e’ to ‘a’].
2.The origin of the English people.
3.Some sound changes.
4.Chaucer’s language.
5.Deor’s Lament (an Old English poem).
7.Scandinavian influence on English.
8.The Old English poem Waldere.
9.Dialect problems.
10.Lengthening of short vowels in Middle English (1100–1500).
11.Classifying consonants in Old English (700–1100).
The first impression one gets from this fascinating material is
that Tolkien is capable as a student only after a year’s study to draw
advanced conclusions about etymology and philology, which normally
fit the work of an accomplished scholar. At the same time he is ambitious enough to make careful notes about sound-changes and etymologies described in the best known handbooks in Old English. It is also
fascinating to study the attention he pays to the use of a clear handwriting and an unusually elegant calligraphy.
In the part of the manuscript collection from 1913, No A 21/2, there
are notes about lectures that Tolkien attended on historical grammar
with references to well-known philologists, such as Sisam, Emerson,
Napier and Morsbach. Nr 21/5 is particularly interesting with all its
comments on Beowulf, both its contents and language. Tolkien refers
to passages in Beowulf, dealing with legendary persons as well as geographical names related to Denmark. The Danish king Rolf Krake
(=Hrothulf in Beowulf) and place-names such as the old village of Lejre
(=Hleidr; Lat. Lethra) outside Roskilde could be mentioned here. In the
Reflections on Tolkien’s Use of Beowulf 233
historical little village of Lejre, archaeologists have recently excavated
the rest of an impressive hall similar to the great hall Heorot, where
Beowulf fought against the monster Grendel in the poem. This excavated hall at Lejre corresponds closely to the one depicted in the poem.
2. Tolkien and Beowulf
The view of the American scholar Michael D.C. Drout is that Tolkien was
greatly inspired by the poem Beowulf and that Tolkien’s “Beowulf: The
Monsters and the Critics” (edited by Drout in 2002) is a point of origin
of modern Beowulf study. Tolkien may not have been the only scholar to
have helped to change the direction of Beowulf studies and the attitude
to the poem, but he was the driving force with very sound arguments,
which led Michael Drout to call the essay “the single most important critical essay ever written about Beowulf” (Drout, Beowulf and the Critics, 1).
The critics had earlier regarded Beowulf as an important historical or
philological work with uncertain literary qualities. Tolkien´s essay became a
distinct turning-point, even a revolution in the discussion of the poetic value
of the poem. Tolkien could indicate that Beowulf as poetry is more beautiful
and every line more significant than those of other Old English poetry.
Previously scholars had regarded Beowulf as an epic poem consisting of two separate parts which did not hang together very successfully.
Tolkien proves very clearly that the poem forms a unity of two connected
parts, which help to create a poetic wholeness. In contrast to previous critics Tolkien makes clear that the structure is remarkably strong. He calls
the structure inevitable and the design of the poem admirable. Tolkien
also argues convincingly that Beowulf’s various adversaries, i.e. Grendel,
Grendel’s mother and the dragon are the central entities in the poem.
Tolkien was greatly engaged in all aspects of the extensive lost literature within the whole Indo-European language area. At some of our
meetings we discussed that topic most vividly in relation to the Old
English poetic fragments. We also discussed the fascinating book, The
Lost Literature of Medieval England, originally published in 1952 by
R.M. Wilson, but which had appeared in a new and revised edition in
1970. The whole of the great lost treasure of ancient literature is a most
relevant key area for the understanding of Tolkien’s thinking in building up a whole fictional world in an age different from his own.
Already at school, Ronald began to read Beowulf in Old English, Sir
Gawain and the Green Knight in Middle English and several Icelandic
tales in Old Icelandic. Through his deep knowledge of many languages
234 From Clerks to Corpora
he became a precocious philologist, who could combine language and
literature research at an advanced level even at the age of 15–16.
A working philologist needs not only to read, analyze and compare
a large number of literary texts. A medieval philologist must also study
palaeography, meaning the knowledge of manuscripts, characters and
styles, the relations between the manuscripts, and much else regarding
the cultures where the manuscripts belong, and also be well acquainted
with the medieval manuscript collections in England.
It is of course a great advantage for a scholar to have easy access to
the great manuscript collections in, for example, Oxford, Cambridge
and London. For Tolkien it was important to have his favourite texts
Beowulf and Sir Gawain in London and Ancrene Wisse (MS. Corpus
Christi College) in Cambridge. The various college libraries in Oxford
and Cambridge are often well equipped as regards medieval manuscripts. The unique medieval library at Tolkien’s own college, Merton,
owns, for example, a copy of the Latin version of Ancrene Wisse, published by the Early English Text Society in 1944. This library at Merton
creates a most remarkable atmosphere. Once I was guided by Tolkien
himself through the library, where some of the medieval manuscripts
were so valuable that they were chained to the shelves.
The poem Beowulf appeared in a new translation into Modern
English in 1999 by the Irish poet and Nobel Prize winner Seamus
Heaney. In the same year Seamus Heaney was invited by the University
of Copenhagen to Lejre near Roskilde in order to read from his English
translation in the excavated great hall mentioned above. He read with
special focus on the passage in the heroic poem where the huge monster Grendel enters the great hall Heorot and walks towards the sleeping Danes. Since I had taken the initiative to invite Heaney to read
and myself read the original Anglo-Saxon text passage before Heaney
in this reconstruction of the special Beowulf atmosphere of the sixth
century, it was a pleasure for me to link all this to Tolkien, who had
meant so much to Beowulf research in the 1930s.
A great part of the Germanic treasure of heroic legends has disappeared in the course of time and in the relevant countries only a
small part is still extant. The only fully preserved long heroic poem in
England is Beowulf, and furthermore there are only two short poems,
Widsith and Deor, and the two fragments, Waldere and Finnsburg.
Tolkien rewrote and commented on all these texts in various ways.
Both Deor and Waldere belong to the areas he wrote essays about in
his notebooks.
Reflections on Tolkien’s Use of Beowulf 235
The heroic poetry from Anglo-Saxon times seems after this brief presentation to be rather modest in size, but important considering that a great
deal of what once existed is no longer there. The fact that Beowulf in the
first place, but also the other, shorter poems caught Tolkien’s interest was
not only due to their literary qualities but also to their fates as manuscripts
and their connections with other literary traditions within the Germanic
mythological field. All names of heroes in the Germanic traditions and
legends circulating through references to Gothic, Burgundian, Icelandic,
and to other myths indicate that they were current in oral traditions in
Anglo-Saxon England. The absence of written tales of these legends may
possibly indicate that the literary traditions survived in oral form.
As Tolkien mentioned himself in his letters,4 he had been influenced
by Beowulf during his work on The Hobbit and The Lord of the Rings.
For example, some names from Beowulf, such as Eomer, Hama and
Wealhtheow appear in The Lord of the Rings.
It has been known for a long time that Tolkien made a translation of
Beowulf in the 1920s. He finished it in 1926 and put it aside perhaps
without actually wanting to publish it. It could very well have served as a
kind of working copy to be used in connection with lectures or citations
or commentaries for a text edition. Now suddenly, in 2014, Christopher
Tolkien has published an edition of Beowulf, a translation and commentary, together with Sellic Spell (HarperCollins Publishers). Christopher
Tolkien’s edition is a complete prose translation of Beowulf made by his
It is fairly well known that Michael D.C. Drout was involved in editing Tolkien’s translation of Beowulf at the beginning of the 21st century.
It may therefore be of interest to know that Drout’s edition was discontinued before it was ready for publication. Drout had received the permission of The Tolkien Estate to publish a two-volume edition including
some of the comments on textual problems written by Tolkien. Drout
had planned to edit the partial verse translation and a complete prose
translation. The companion volume was supposed to include commentaries made by Tolkien. Before Drout had finished the project, the permission to publish his edition was withdrawn by The Tolkien Estate.
As a great surprise to many, the translation of Beowulf by J.R.R.
Tolkien was suddenly published by HarperCollins in 2014 with
Christopher Tolkien as editor. Christopher has also included his father’s
4See a letter to The Observer: “Beowulf is among my most valued sources”; The
Letters of J.R.R. Tolkien. 31.
236 From Clerks to Corpora
Sellic Spell, a ‘marvellous tale’, which, as Christopher suggests, is in the
form and style of an Old English folk-tale of Beowulf with no connection with the ‘historical legends’ of the Northern kingdoms.
Furthermore, Christopher included two versions of his father’s Lay
of Beowulf, which is a rendering of the story in the form of a ballad
supposed to be sung, now a clear memory after more than eighty years
of Christopher’s first acquaintance with Beowulf and the golden hall of
Heorot. With very little imagination needed, one may quite easily hear
J.R.R. Tolkien’s voice singing the first stanza of the later of the two
poems The Lay of Beowulf, called ‘Beowulf and the Monsters’:
Grendel came forth at dead of night;
the moon in his eyes shone glassy bright,
as over the moors he strode in might
until he came to Heorot.
Dark lay the dale, the window shone;
by the wall he lurked and listened long,
and he cursed their laughter and cursed their song
and the twanging harps of Heorot.
Tolkien’s translation of Beowulf, as presented by Christopher, is all
in prose, as mentioned above, starting in the following way:
Lo! The glory of the kings of the Spear-Danes in days of old we have
heard tell, how these princes did deeds of valour. Oft Scyld Scefing
robbed the hosts of foemen, many peoples, of the seats where they
drank their mead, laid fear upon men, he who first was found forlorn;
comfort for that he lived to know, mighty grew under heaven, throve
in honour, until all that dwelt nigh about, over the sea where the whale
rides, must hearken to him and yield tribute – a good king was he!
The Anglo-Saxon original looks as follows:
/Hwæt we Gar-Dena in geardagum,
þeodcyninga þrym gefrunon,
hu ða æþelingas ellen fremedon.
Oft Scyld Scefing sceaþena þreatum,
monegum mægþum meodosetla ofteah,
egsode eorl[as], syððan ærest wearð
feasceaft funden¸ he þæs frofre gebad,
weox under wolcnum, weorðmyndum þah,
oðþæt him æghwylc þ[ær] ymbsittendra
ofer hronrade hyran scolde
gomban gyldan, þæt wæs god cyning!/
Reflections on Tolkien’s Use of Beowulf 237
In order to make the contents of the beginning of the poem even
more clear, I conclude by presenting Seamus Heaney’s translation of the
same lines into modern English:
/So. The Spear-Danes in days gone by
and the kings who ruled them had courage and greatness.
We have heard of those princes’ heroic campaigns.
There was Shield Sheafson, scourge of many tribes,
a wrecker of mead-benches, rampaging among foes.
This terror of the hall-troops had come far.
A foundling to start with, he would flourish later on
as his powers waxed and his worth was proved.
In the end each clan on the outlying coasts
beyond the whale-road had to yield to him
and begin to pay tribute. That was one good king./
While Christopher Tolkien presented his father’s translation of
the text into modern English prose, Drout, on the other hand, had
announced that he had planned to use both prose and part poetry.
Whether the text of Christopher Tolkien’s edition could have been
expressed more clearly, if his edition had included both prose and part
poetry, is not easy to say. The fact that this edition is now out is, however, an astounding event in itself. In whatever case, the special relations
between Tolkien and Beowulf are clearly made public and are often
well illustrated through Christopher Tolkien’s publication.
Published works by J.R.R. Tolkien or Christopher Tolkien
The Hobbit: Or There and Back Again. (1937). London: George Allen & Unwin.
The Fellowship of the Ring: Being the First Part of The Lord of the Rings.
(1954). London: George Allen & Unwin.
The Two Towers: Being the Second Part of The Lord of the Rings. (1954).
London: George Allen & Unwin.
The Return of the King: Being the Third Part of The Lord of the Rings. (1955).
London: George Allen & Unwin.
Ancrene Wisse: The English Text of the Ancrene Riwle. (1962). Edited by
J.R.R. Tolkien from MS. Corpus Christi College Cambridge 402. Early
English Text Society O.S.249. London: Oxford University Press.
The Silmarillion. (1977). Edited by Christopher Tolkien. London: George
Allen & Unwin.
238 From Clerks to Corpora
The Letters of J.R.R. Tolkien. Edited by Humphrey Carpenter with the
Assistance of Christopher Tolkien. (1995). London: HarperCollins
The Monsters and the Critics and Other Essays. (1983). Edited by Christopher
Tolkien. London: George Allen & Unwin.
Books related to J.R.R. Tolkien
Alexander, M. (ed.) (1995). Beowulf: A Glossed Text. London: Penguin Books.
Carpenter, H. (1977). J.R.R. Tolkien: A Biography. London: George Allen and
———. (1978). The Inklings: C.S. Lewis, J.R.R. Tolkien, Charles Williams and
their Friends. London: George Allen and Unwin.
Drout, M.D.C. (2002). Beowulf and the Critics. Tempe, Arizona Center for
Medieval and Renaissance Studies.
McNamara, J. (ed.) (2005). Beowulf. New York: Barnes & Noble Classics.
Pearce, J. (1998). Tolkien. Man and Myth: A Literary Life. London:
Shippey, T.A. (1982). The Road to Middle-earth. London: George Allen and
———. (2000). J.R.R. Tolkien: Author of the Century. London: HarperCollins.
Zettersten, A. (2008). Tolkien—min vän Ronald och hans världar. Stockholm:
———. (2011). J.R.R. Tolkien’s Double Worlds and Creative Process—
Language and Life. New York: Palgrave Macmillan.
14 Commentators and Corpora: Evidence
about Markers of Formality
David Minugh
Stockholm University
1. Introduction
Dictionary-makers and stylists have long singled out various terms for
special notice, and at times had strong opinions about their use and
abuse. These comments were in many cases essentially a matter of taste
(often masquerading as logic), but until corpus linguistics and powerful
computers arrived on the scene, no tools existed to demonstrate actual
usage, beyond collections of (laudable and reprehensible) examples.
Logical and sentential connectors have not escaped such scrutiny, and
here we shall focus on three fairly formal such terms, all of which have
interesting characteristics from a learner perspective: albeit, notwithstanding and thus. After briefly considering their origins, we will ­examine
some of the comments about them, particularly by grammarians and
style police, and then bring in data from recent corpora to examine their
actual use, which will not always prove to be in formal settings.
2. Origins
The lexical items albeit, notwithstanding and thus are not particularly
obscure in their development, although they do have a reasonably venerable pedigree. The OED Online considers the etymology of albeit as
straightforwardly deriving from all as a conjunction and the present
subjunctive of be, with the first instances surfacing with clauses in the
late 14th century:
[1] “But syn my name is lost thurgh you,” quod she,
“I may wel lese a word on yow or letter,
How to cite this book chapter:
Minugh, D. 2015. Commentators and Corpora: Evidence about markers of formality.
In: Shaw, P., Erman, B., Melchers, G. and Sundkvist, P. (eds) From Clerks to Corpora:
essays on the English language yesterday and today. Pp. 239–265. Stockholm: Stockholm
University Press. DOI: License: CC-BY.
240 From Clerks to Corpora
Al be it that I shal be neuer the better”
[Chaucer, Legend of Good Women, 1361–63]1
Further instances soon show it introducing other constructions, such
as PPs:
[2] We dyd graunte (albeit not for this argumentacyon) that…
[Marshall, 1535]
or as an adverb, as in the OED’s quite recent final citation:
[3] Young skunks begin to spray, albeit inaccurately, at about
one month of age.
[1995, Animals’ Voice Spring 13/1]
Notwithstanding is also a compound form, straightforwardly
derived, as Johnson noted,2 from not + withstand, on the pattern of
Anglo-Norman and Old French non obstant and post-classical Latin
noˉn obstante, with the same sense, appearing shortly after albeit:
[4] Natwith-stondinge his grene mortal wounde, He ros ageyn.
[ c1425, Lydgate Troyyes Bk.]
Its most striking grammatical feature, the ability to function as a
postposition (or adverb, depending on your analysis),3 is also documented from within less than a century later, as in Caxton:
[5] This notwystondyng, alwaye they be in awayte.
[Eneydos, 1490]
a variation which remains its hallmark until the present day.
Our final (and oldest) item, thus, apparently has its roots in the
demonstratives (the OED suggests derivation via either that or this).
Some early examples from the OED:
[6] Sicini [siccine], ac ðus
[c725 Corpus Gloss. 26]
1Much of the detailed OED information has been removed from these citations;
the Chaucer quote follows the text in Fisher 1977:643. The OED also mentions
variants such as al were it, albe (both with further citations from Chaucer), but
the clearest view of the range of this type of construction actually emerges from the
examples cited in Jespersen 1940.
2“[Notwithstanding] is properly a participial adjective, as it is compounded of not
and withstanding, and answers exactly to the Latin non obstante” (1783, Vol. II).
3Cf. Rissanen 2002 and Weber (2010:181–86) for Middle English developments,
Minugh 2002 for modern English use.
Commentators and Corpora: Evidence About Markers of Formality 241
[7] & tuss ȝho seȝȝde inn hire þohht..Þuss hafeþþ drihhtin don
wiþþ me.
[?c1200 Ormulum (Burchfield transcript) l. 235–7]
[8] Here vn-to you þus am I sente.
[c1440 York Myst. vii. 6]
3. Learners’ perspectives
Albeit. From the start, the term albeit stands out for phonetic reasons.
For foreign learners and young native speakers alike, the word is normally first encountered in written form, so that the trisyllabic pronunciation /ɔːlˈbiːɪt/, with a clear be, often comes as a distinct surprise,
particularly if they have previously paid attention to items with reduced
stress, such as the RP pronunciation of secretary. To most speakers of
English, the etymological links to al- (as in although) and subjunctive be
are not at all obvious, particularly since the latter’s primary ­current use,
the mandative subjunctive (e.g. I move that the meeting be adjourned),
is not frequent (Hundt 1998); in addition, the pronunciation of the
final -it as a distinct final syllable is unexpected. Placing the stress
on the first syllable (as in alien, alias) would lead to something like
*/ˈeɪlbɪt/, which has apparently never been current. John Wells (2008:
19) records the frequent but “non-RP” pronunciation /ælˈbiːɪt/ (he also
notes it for AmE, a form that e.g. Elster [1999:13] takes violent exception to). Once learned, its pronunciation is easy enough (in parallel to
although), and as regards usage, it presents no particular difficulties,
functioning as a synonym of even though or although.
Notwithstanding. For learners, the pronunciation of notwithstanding ought to be straightforward (once they grasp that it is a single unit),
and as for usage, its preposed placement predominates in BrE, and
causes no problem. This position allows it to control fairly long (and
relatively complicated) constructions, whereas its postposed use tends
to be limited to controlling short NPs. The postposed use is above all
found in more formal AmE (cf. Minugh 2002 for statistics).
Thus. The voiced initial consonant of thus follows the normal deictic
patterns seen in the, this, those, thy and so on. It has no direct cognates
in Romance or Germanic languages (the sole exception is Dutch dus).4
4 Dus is considerably more frequent than its English counterpart: the Dutch Web
Corpus (via the commercial program called SketchEngine) reports it as having an
occurrence of 1,299 per M words; by comparison, the Oxford English Corpus
(again via SketchEngine) reports thus as having an occurrence of 153 per M words.
242 From Clerks to Corpora
Its simple monosyllabic form and its use parallel those of other logical
connectors such as so. But like therefore and as a result, thus has a distribution heavily slanted towards formal written English; this (and the
lack of cognates) appears to delay its acquisition, at least in Sweden,
where informal English is given priority in the school system.
Of these three items, only thus (which has the widest functional
range) was regarded as sufficiently important to be included in the
classical General Service List (West 1953), the first reasonably modern
word list for learners. When the Academic Word List was developed
(Coxhead 2000), all GSL words were excluded, as already covered, so
that thus was not included in the AWL; the latter does, however, include
both albeit and notwithstanding.5
4. Stylistic comments by dictionaries, grammars
and style manuals
In this section, we will briefly survey what various reference works have
had to say about our three terms, and what claims, if any, they make
about the validity of their comments about the use of albeit, notwithstanding and thus. It should be noted at the outset that there is no
significant disagreement about the semantics of these terms; what is at
issue are matters of register and style.
4.1 Major dictionaries
Johnson’s epoch-making Dictionary (1755) records all three items
without further ado, notably without any comments on their stylistic
level. The reader is reminded that he was by no means above pronouncing judgments about usage: while thus is merely recorded, compare his comment on the very next word, thwack: “A ludicrous word”
(1799, Vol. II).
The first edition of the OED (1933) passes over the stylistic value
of albeit and notwithstanding in silence, but begins the article for
thus with the note “now chiefly literary and formal” (1933:XI, 397);
more interestingly, no changes in this judgment are to be noted even
in the contemporary OED Online. In addition, the one-volume New
5 The increasing impact of the AWL is seen not least at the English Department of
Stockholm University, where an “AWL Vocabulary” test is administered to entering
students early in their first semester. Not until 2013 did its first serious competitor
appear (; cf. Gardner & Davies 2013. Note
also section 4.4, below.
Commentators and Corpora: Evidence About Markers of Formality 243
Oxford Dictionary of English (1998) similarly has a note only for thus:
“poetic/literary or formal”. Webster’s New International Dictionary of
the English Language (1941) records all three without further comment, as do Webster’s Collegiate Dictionary (1988, 1993) the American
Heritage Dictionary of the English Language (1970) and Australia’s
Macquarie Dictionary.6
4.2 Grammars
Turning now to earlier 20th-century grammars, we find that Poutsma
comments “[i]n Present-day English [albeit] is used only in the higher
literary style, mostly without that” (1929:I, ii, 712), with a similar comment on notwithstanding (711). Curme notes that conjunctions used in
concessive clauses include notwithstanding and “[i]n older or archaic
English: albeit (i.e., all be it = be it entirely) that or simple albeit, albe”
(1931:II, 333). Jespersen (1940:51) remarks on the alternate pre-/postposition of notwithstanding and provides numerous examples of albeit
and related subjunctive constructions (1940:364). Interestingly enough,
his volume on pronunciation (1949) does not mention albeit.
For our triad of terms, Swedish-based university grammars of English
have a long tradition of silence as regards form and use, although register is occasionally touched on. Elfstrand & Gabrielsson (1960) only
­mention notwithstanding that as a concessive conjunction. Svartvik &
Sager (1977, 1996) mention albeit functioning to link adjectives
(§353D) in “formal language” and thus as a linking adverbial (§ 439E)
“in formal style”. More recently, Estling Vannestål merely mentions
thus as one of the linking adverbials (2008:269), omitting albeit and
In Quirk et al., the first major grammar with a dawning awareness
of corpus data, albeit is dismissed in a footnote: “the following archaic
subordinators still have a limited currency: albeit, whence, whereat,
wherefore, whither” (1985:998, note [b]). Notwithstanding is mentioned several times, usually with the label “formal”; note particularly:
“Notwithstanding [‘in spite of’] is formal and rather legalistic in style,
particularly when postposed” (1985:706). Together with other prepositional phrases (despite, in spite of, irrespective of, regardless of),
6The New Oxford Dictionary also includes an entry on thusly, which is labeled
“informal” (1998:1935b), while the American Heritage goes further, labelling it
“nonstandard,” noting that it “is termed unacceptable by 97 per cent of the Usage
Panel” (1970:1342); cf. Menken’s comments, in section 4.3, below.
244 From Clerks to Corpora
notwithstanding is “considered stylistically clumsy” (1985:1098). Thus
is consistently labeled “formal”, e.g. “The form thus is largely formal”
(e.g. 1985:557, note [b]).
In their brief discussion of register, Celce-Murcia & Freeman remark:
“In any kind of informal situation, a native speaker of English would
be surprised to hear somebody say notwithstanding the fact that to
express the notion of concession. A connector such as even though
would be much more likely” (1983:323). They nevertheless list albeit
and notwithstanding under “Concession,” without any comments on
register (326).
Turning to modern general learner grammars, we find that A
Communicative Grammar of English (2002) does not include albeit,
but does mention notwithstanding (“very formal” [2002:113]) and thus
(“formal” [2002:110]). The Longman Student Grammar of Spoken and
Written English (Biber, Conrad & Leech 2002) appears to contain no
information on our three terms.7 The Cambridge Grammar of English
(Carter & McCarthy 2006) is silent on albeit and notwithstanding.
They list thus as an option among many, but the only concrete information given is that initial thus can allow inversion:
[9] Thus does Mr Major find himself ever more closely closeted
with Mr Campbell. (2006:782)
4.3 Prescriptive stylists and manuals of style
Like most of the grammars cited above, nearly all of the works cited
below were written in the pre-corpus era. With a single exception to be
discussed below, however, they rarely cite extensive examples to bolster
their claims. Fowler & Fowler (1930:29), for example, using guilt by
association, dismiss albeit as an archaism, listing it with the likes of
bashaw, certes, damsel and quoth(a), terms few would wish to champion as shining examples of modern English. They are silent on notwithstanding, but object strongly to thus in one case:
In this use thus is placed before a present participle (thus enabling &c.),
& its function, when it is not purely otiose, seems to be that of apologizing for the writer’s not being quite sure what noun the participle belongs
to, or whether there is any noun to which it can properly be attached
7 A caveat: since lexical items are not included in the index, the search by subject area
may have missed a minor comment on these words.
Commentators and Corpora: Evidence About Markers of Formality 245
This actually sounds rather like the discourse markers sometimes
referred to as shell nouns, i.e. a way of summing up a form of logical
relationship previously presented in detail in the text (Schmid 2000), a
use which they find to be too vague. In all other respects, thus is passed
over in silence.
The Americanist H.L. Mencken found nothing to comment about on
albeit and nevertheless, but was interested enough in the American use
of ly-less adverbs to comment that: “the use of illy and thusly is confined
to the half educated” (1936:467).8 Copperud (1964) only warns against
the use of for or thus at the beginning of sentences: “…an affectation
by some writers, particularly columnists. This is warranted only when
the sentence draws a conclusion based on what has gone before” (1964:
165). The Longman Guide to English Usage (Greenbaum & Whitcut
1988), silent on notwithstanding, does warn against “the facetious
variant thusly”, and waxes truly eloquent on albeit:
This is often regarded as pretentious when used, unless for humorous effect, as an alternative to (even) though. It is perhaps justified
as a convenient way of linking pairs of adjectives (a small albeit
crucial mistake), although but, yet, and though will also do in this
case (1988: 27).9
Oxford’s Authors’ and Printers’ Dictionary (1956), the Chicago
University Press Manual of Style (1969), Michael Swan’s Practical
English Usage (2005) and Collins COBUILD English Usage (1992) are
among the numerous works silent on these three words. As for student writing manuals, an examination of the popular Writing Academic
English revealed only that thus appears in several lists of “connecting
words and transition signals” (Oshima & Hogue, 2006, Appendix C),
while albeit and notwithstanding are passed over in silence.
However, one work stands out in its detailed comments on albeit,
as well as its extensive use of 20th century citations (almost unique
among style manuals): the Merriam-Webster’s Dictionary of English
Usage (1994). Their opening shot deserves quotation in extenso:
Copperud 1970, 1980 observes that “a generation ago” albeit was
considered archaic but is “now being revived.” The source of the
8 In Supplement Two, he records a congressman using thusly, but adds, “However,
it is often difficult to tell whether a congressman is serious or spoofing” (1948:
390, n. 3).
9 Also noted by Svartvik & Sager 1977 (see section 4.2, above).
246 From Clerks to Corpora
notice of revival is Gowers (in Fowler 1965). This is a most curious
business, since albeit seems never to have gone out of use, though
it may have faded somewhat in the later 19th century. If it did, the
revival began decades before the commentators noticed. (1994:65)10
They go on to trace a lineage of albeit quotes from 1907 to Krapp’s
grammar in the late 1920s, with a last example from the 1980s. As
noted in section 4.2, above, as late as 1985 Quirk et al. labelled albeit
as archaic, despite such evidence.
What we seem to find, then, is a series of fairly random objections
to specific uses or forms (such as thusly), while “allowing” others. This
is hardly surprising, given that these writers were unable to systematically trawl through large amounts of text from many different domains
for matters of interest. To do so, we must turn to recently-compiled
corpora for documentation. In doing so, we will concentrate on these
three terms and their frequencies over the last two centuries. The first
indications of what this can result in may be seen in modern learner
dictionaries, to which we now turn.
4.4 Learner dictionaries
Starting with the first edition of the Collins COBUILD dictionary
(1987), but increasingly in the period after 2005, learner dictionaries
have based their labelling on data drawn from (usually in-house) large
corpora, i.e. corpora now normally in excess of 100 M words. It is nevertheless worth noting the comments from the editors of the Oxford
Advanced Learner’s Dictionary (ALD): when it comes to deciding on
the recently-introduced “Oxford 3000” keywords (“the words which
should receive priority in vocabulary study because of their importance
and usefulness”), they based their decision on corpus frequency and
range of text types—but also as being “very familiar to most users of
English”, as judged by “language experts and experienced teachers”
(2005:R99). In other words, for Oxford, the corpus is definitely not
considered the sole arbiter in adjudicating on such matters.
What, then, do learner dictionaries say about our three terms?
A ­pre-corpus edition of the Oxford ALD (2nd ed., 1963) labels albeit
10Gowers states that “[albeit] has since been picked up and dusted and, though
not to everyone’s taste, is now freely used, e.g. It is undeniable that Hitler was a
genius, a. the most evil one the modern world has known” (1965:16). Note also
that Copperud 1964 was silent on albeit; in later editions he is clearly aware of the
changing perception of albeit’s status.
Commentators and Corpora: Evidence About Markers of Formality 247
as “not colloq[uial]”, but is otherwise silent on this issue. By the 7th
(­corpus-aware) edition of 2005, it labels all three as “formal”, while
­including thus as one of its “Oxford 3000” keywords. In the 8th edition
(2010), albeit and notwithstanding are additionally labeled as aw (i.e.,
part of the Academic Word List, which has now made its entry into the
COBUILD editions (1987, 1995) are relatively consistent, labelling
albeit and notwithstanding as “a formal word”, but thus as “a fairly
formal word”.
By its 4th edition, the Longman Dictionary of Contemporary English
labels all three as “formal”. Thus is noted as w1, i.e. among the 1000
most common words of written English, but with a warning triangle
indicating that when it is used as a sentence adverb, “in spoken English
it is more usual to use so”. In the 5th edition (2009), albeit and notwithstanding receive the additional label ac (i.e., part of the Academic
Word List, which has now made its entry into LDOCE, as well).
The Cambridge Advanced Learner’s Dictionary (3rd ed., 2008) labels
all three as “formal”. Thus is additionally noted as i, for “improver”,
the middle category in its high-frequency words.11
Table 1. Learner dictionary labels for albeit, notwithstanding and thus.
ALD (1963)
ALD (2005)
ALD (2010)
COBUILD (1987)
COBUILD (1995)
LDOCE (2005)
LDOCE (2009)
CALD (2008)
MEDAL (2007)
CDAE (2000)
OADCE (1999)
not colloq.
formal aw
formal ac
formal ★
(no label)
(no label)
formal aw
formal ac
formal I
(no label)
(no label)
(no label)
formal Oxford 3000
formal Oxford 3000
fairly formal
fairly formal
formal w1
formal ★★★
11 More specifically, this applies to thus in the senses ‘in this way’ and ‘with this result’,
with a frequency typically of 200–400 per 10 million words (2008:VIII).
248 From Clerks to Corpora
The Macmillan English Dictionary for Advanced Learners (2nd ed.,
2007) also labels all three as “formal”, but includes albeit among the
7000 most common words of English (one star) and thus among the
2500 most common words (three stars, its highest frequency rating).12
As for these dictionary-makers’ American offshoots (which are invariably smaller, presumably in order to sell better in America), the Oxford
American Dictionary of Current English (1999) notes that thus is formal, but has no labels for albeit or notwithstanding. The Cambridge
Dictionary of American English (2000) omits albeit altogether, but
includes notwithstanding (with only one example—a postposed one!)
and thus (considered “formal”).
Summarizing, we obtain the table below, from which it appears
clear that the dictionary-makers are in agreement on both register and
frequency. This should not lead to conspiracy theories about borrowing from one another, but rather is a consequence of their now having
access to large, proprietary corpora yielding similar results. However,
it has recently become possible for scholars independently to check on
these results, thanks to large-scale publicly-available corpora, to which
we now turn.
Table 2. Frequencies from the BNC (Lancaster interface), including
­comparative data for therefore.
Written %
20, 127
21, 406
Per M wds Spoken %
96.6% 15.13
99.6% 228.97
93.2% 243.52
Per M wds
12This is the BrE version; the AmE edition is Macmillan English Dictionary for
Advanced Learners of American English, which had no second edition, their subscription website having instead taken over all updating.
Commentators and Corpora: Evidence About Markers of Formality 249
5. Corpus data for albeit, notwithstanding and thus
5.1 Contemporary corpora13
The earliest “large” (= 1M word) corpora of contemporary written
English are the BROWN series: BROWN and LOB have matched
AmE/BrEtexts from 1961, and FROWN and FLOB similarly matching
texts from 1991: They produce quite small numbers (only thus yields
results larger than 20 examples per corpus), but they will turn out to be
quite close to the results from the much larger corpora now available.14
The major 90s corpus, the British National Corpus (BNC), at 100M
words (British English only) yields our first solid data on how these
words are distributed along the written/spoken dimension:15
This clear preponderance of written instances suggests that we are
dealing with what is tantamount to words found in written English
only, particularly if one considers that some of the “spoken” data consists of prepared transcripts for radio and TV. In absolute numbers,
thus is once again more common than the other two by more than an
order of magnitude. Using a further analysis from the Brigham Young
(BYU) interface, we can break down the results into different domains
(Figure 1):16
13The corpora in this section all seek to portray modern English from the 1990s
onward. The most purely synchronic of these corpora are of course the BROWN
group, each of which samples only one year. The diachronic corpora in section 5.2,
on the other hand, cover a much larger temporal range, precisely in order to track
changes over time.
14 For descriptions of these earlier corpora, see any standard undergraduate textbook
on corpus linguistics, e.g. McEnery, Xiao & Tono 2006.
15The now much larger Collins COBUILD Bank of English was the first modern
corpus of English, but the open-access policy of the BNC continues to be crucial to
scholarship; meanwhile, the publicly available component of the BOE has evolved
into the 57M Collins Wordbanks Online, currently openly available at http://www.
16A technical note: by comparing instances per 1M words, and selecting for each
word the domain with the largest number of instances as 100%, we can graphically
compare all domains for each word individually.
250 From Clerks to Corpora
albeit notwithstanding em
ic isc
s ad
ic m
r no
e ne
n thus fic
100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Figure 1. BNC distribution frequencies (BYU interface).
As expected, all three connectives occur most frequently in the academic domain. However, the most striking aspect of this comparison is
that albeit is considerably more evenly distributed than the other two;
the discrepancy is so large that it is difficult to ascribe it merely to being
an artefact of the domain definitions.17 The chi-square test returns a
significance of well below p< .001.
Since the BNC is specifically limited to BrE, let us next turn to the
Corpus of Contemporary American English (COCA), which now covers a little over two decades, from 1990 on.18 Containing 450 M words,
including a “spoken” section (largely derived from radio/TV transcripts), it is the largest broadly-based contemporary corpus with free
access, although also limited geographically. Not surprisingly, in raw
numbers, thus again dominates by more than an order of magnitude,
with 62,764, compared to albeit and notwithstanding, with 4,061 and
2,683, respectively. We therefore again choose to display the data as
percentage comparisons to the largest category for each item, again
based on frequency per 1M words (Figure 2):
17The values for the category “miscellaneous”, on the other hand, clearly indicate
that something is escaping this categorization.
18It has been expanding as time passes, now including up to 2012, so that this is
an evolving synchronic record of “contemporary” American English. Like several
other contemporary corpora, it thus will not yield replicable results over time, since
the corpus itself is growing; cf. the “monitor corpus” solution adopted by John
Sinclair and the COBUILD team (for an early description, see e.g. Clear 1998).
Commentators and Corpora: Evidence About Markers of Formality 251
100% 90% 80% 70% 60% albeit 50% notwithstanding 40% thus 30% 20% 10% 0% spoken fic5on magazine newspaper academic Figure 2. COCA distribution frequencies.
From Figure 2 it is clear that for all three items, the academic
domain dominates, having more than twice the frequency found in the
other domains. For thus, the dominance of the academic domain is
overwhelming, while both albeit and notwithstanding have a certain
currency in magazines and newspapers, perhaps due to their feature
Figure 3. COCA (C) and BNC (B) distribution frequencies (comparable
252 From Clerks to Corpora
articles. With the possible exception of spoken notwithstanding, the
three items show similar distribution patterns, with thus having the
widest gap between the other domains and academic English. Via the
chi-square test, both albeit and thus distributions are significant at
p < .001, while notwithstanding has p < .0157.
Naturally, it is interesting to compare the British and American data.
Since both corpora are available with the same (BYU) interface, this
would appear to be simple, but the BNC data has the two extra categories of non-academic and miscellaneous, which in unknown fashion
are redistributed in COCA’s fewer domains. Omitting those two BNC
categories, our comparison looks like this (Figure 3).
The fit between these two geographical domains is quite good
for both albeit and thus, which is not surprising, given that four of
the five are written, the domains where BrE and AmE are traditionally considered to have the smallest differences. The odd man out is
Figure 4. OCE distribution frequencies (SketchEngine interface).
­ otwithstanding, which seems to be more favored in AmE, again with
the reservation for the BNC data loss.
Stepping up to an even larger corpus, the Oxford Corpus of English,
a corpus from the early 2000s based on material from the Web, now
supplemented to reach 1736 M, and including significant input from
English in other parts of the world, we again find that thus is more frequent by an order of magnitude: 152.4 words per million, versus albeit
Commentators and Corpora: Evidence About Markers of Formality 253
60 50 40 30 albeit 20 notwithstanding 10 thus l ni
0 Figure 5. OCE frequencies vis-à-vis formality, per 1M words (SketchEngine
and notwithstanding at 13.7 and 8.4, respectively. Comparing their frequency relative to the largest domain (SketchEngine provides 21 different domains), we see a largely parallel pattern for the three, including
great differences between domain frequencies (Figure 4).19 Albeit weighs
in heavily in the arts, and above all, in news—perhaps an attempt to give
news greater weight? Not surprisingly, notwithstanding’s single largest
component is legal texts. Thus, however, turns out to be relatively evenly
distributed, with the single spectacular exception of science texts, where
it occurs three times more frequently than in any other domain (except
the dubious “unclassified”). All three terms are relatively well represented in “weblogs”, perhaps because these are relatively early blogs,
when they had not yet reached the demotic level of today’s twittering.
Also of interest is that for all three terms, only a few domains reach levels more than 33% of the most frequent domain, again indicating that
the distribution of these words is quite domain-sensitive.20
The SketchEngine software for the OCE also allows us to look at this
data via degrees of formality, ranging from non-standard to formal and
19 The values for the category “unclassified” are uniformly high, again suggesting that
something fairly formal about them is escaping the categorization.
A technical note: to keep the diagram legible, the type of graph has been changed,
but the domains are of course independent of one another.
20In terms of raw numbers, these are quite robust samples, with 16,293 instances
of albeit, 9,024 of notwithstanding, and 189,969 of thus, so that even one of the
smallest, the fiction examples of thus, weighing in at 0.1 per million, still totals 109
separate tokens.
254 From Clerks to Corpora
technical.21 Here, the most striking distribution is that of thus, whose
use peaks in the technical texts (suggesting that this group, rather than
formal, includes most scientific texts), but which is still a clear presence
from the standard level on upward. The other, somewhat surprising
factor is that albeit seems above all to be a marker of standard texts, as
seen in the peaks in the arts and news domains; domains such as science
and law are less entranced with its quasi-literary flavor, it seems.
5.2 Diachronic corpora
Here, we shall consider three very recent diachronic corpora: the TIME
corpus (100 M words, 1923–1996), the Corpus of Historical American
English (COCA; 406 M words, 1810–2009), and the GOOGLE US/
UK corpus (in two parallel parts: AmE 155 B words, BrE 34 B words,
1810–2000), all of them created at Brigham Young University, and with
the same interface.
The TIME corpus is one of the few corpora that chart a single source
over a long period.22 Time Magazine began publication in 1923, and
this corpus includes all the texts in Time (excluding ads, picture captions, etc.) from its inception until 1996. Two factors are of particular importance when using this corpus: first, nearly all of its articles
21 Their category technical is clearly not automatically “more formal” than, say, formal, but presumably much narrower in domain. Since SketchEngine makes this
division of the entire corpus, technical is included in the present discussion.
22 There is a small, but clear overlap between this corpus and both COCA and COHA,
as the latter two corpora could hardly ignore Time when dealing with contemporary and historical American magazine writing.
Figure 6. TIME distribution frequencies.
Commentators and Corpora: Evidence About Markers of Formality 255
Figure 7. COHA distribution frequencies.
are collectively written, with a small staff of writers and editors interacting on many articles; second, it has been read in a large number
of middle-class American homes for generations. As such, it is hardly
­representative of all American writing, but is disproportionately influential (although with a penchant for word play and bons mots). Again,
the most revealing way to look at its statistics is to compare changes in
relative frequencies per decade (Figure 6).
There are two striking changes for our word trio: first, thus has
undergone a steep decline, broken only by a resurgence during the 60s
and 70s, and ending up at less than 20% of its frequency in the 20s,
from 347 per M words to 65 per M words after 2000. This is clearly in
line with the specialization (tantamount to domain loss) we see in the
contemporary corpora, where the vast majority of the modern instances
are in science articles, a domain that does not feature prominently in
Time. The second is the rise of albeit, which, at 0.6 per M words in the
1930s (i.e. less than the frequency of recondite words such as germane,
which in turn is almost 40 times less frequent than relevant), rises uninterruptedly to 10.4 per M words in the 2000s. This is almost double the
peak frequency of notwithstanding, which fluctuates from 3.2 to 5.5
per M words throughout this time period. The fluctuations of notwithstanding suggest that we may not be seeing change that is a trend, but
rather a fairly stable term with a variation of ± 1 per M.
Turning to the COHA corpus, we shift to a more traditional type of
linguist’s corpus, i.e. a sample selected for linguistic purposes. It covers two centuries, and again our comparison is of relative frequencies
(Figure 7). From the perspective of this longer time scale, we see that
256 From Clerks to Corpora
Figure 8. COHA and Time distribution frequencies, 1920s–2000s.
both thus and notwithstanding have declined drastically from the early
19th century to the present, and furthermore, rather consistently. The
odd man out is albeit, which has an extraordinary peak in the 1820s
(probably a result of its sampling,23 and a more genuine higher level in
the late 19th century, but which rises steadily from its low point of 1.22
per 1 M words in the 1930s to 7.41 in the 2000s. The shifts in notwithstanding and thus are both statistically significant (both with p < .001),
but not the variation in albeit.
If we compare these two corpora during the time period 1920–2010
(ignoring for the moment the obvious distortion effects of comparing
an entire range of written language with the language of a small group
of editors and writers working at one publishing house), we find the
following (Figure 8). Both Time and COHA begin with a high level of
thus and notwithstanding, but quite a low level of albeit. They match
quite well for both albeit and thus, the former dipping, then rising
sharply, and the latter dipping, then dropping off sharply (probably
a reflection of the domain loss suffered by thus). Time’s retention of
23 These early decades have far fewer works to draw upon than the rest of the corpus, and are thus more vulnerable to sampling peculiarities. In particular, of the
71 instances of albeit in the 1820s, 55 are from a single work, The Buccaneers:
A Romance of Our Own Count[r]y in Its Ancient Day . . . [by] Yclept Terentius
Phlogobombos [pseud, actually Samuel Judah].
Commentators and Corpora: Evidence About Markers of Formality 257
!!!! !""""" ""
Figure 9. Google US and COHA distribution frequencies.
Figure 10. Google frequency distributions, 1810–2000.
notwithstanding, however, appears to be out of step with its reduced
role in written language in general.24
Further comparison may now be made with material from the newly-released Google corpus from 1810 to 2000 (actually two parallel
corpora of AmE and BrE, respectively). The vast amount of material
available, with text masses in the billions of words, would seem to imply
24 A word of caution about the 1920s issues of Time: a number of other searches indicate that the 20s was a period when the magazine was finding its level of readership,
and is an atypical decade; e.g its use of shall dropped by 50% from the 20s to the
30s; no other decade-to-decade comparison indicates such a major shift for shall.
258 From Clerks to Corpora
Figure 11. C
omparison 1810–2000 for the three major diachronic corpora,
plus Time (from 1923 to 1996).
a major improvement in quality of data, but in his comparison website,
its creator Mark Davies remarks, “All three resources—Google Books
(both versions) and COHA—give nearly the same results for [word
and frequency] searches. The 400 million words in COHA is probably
sufficient for nearly all searches of individual words and phrases.”25 It is
consequently interesting to compare the COHA and Google US material for this period, as seen in Figure 9.
For thus, and even more so for notwithstanding, the fit appears to be
quite good, but much less so for albeit, at least during several periods of
the 19th century, whereas the 20th century appears to be a good fit. The
explanation lies in the data mentioned in note 23, above: a single book
in the 1820s accounts for 77% of the instances in the 1820s COHA
data. This is a useful reminder when there are startling shifts in the data
between adjacent periods.
Within Google, one can compare the relatively massive AmE corpus
with the five times smaller BrE corpus for our three items (Figure 10).
As the graph shows, the fit over two centuries is astonishingly good,
with only a minor blip in the figure for albeit in BrE in the 1840s to
disturb the picture.26 The fit is also relatively good with the COHA
data, and bears out Davies’ prediction for both thus and notwithstanding.
25, accessed March 10, 2014.
26The possibility that albeit was an OCR error for (Prince Consort) Albert was
explored, but the readings are accurate; the Google image was quite clear in all
instances checked—the BrE 1840s texts had a penchant for albeit.
Commentators and Corpora: Evidence About Markers of Formality 259
Figure 12. C
omparison 1810–2000 for the three major diachronic corpora,
plus Time (from 1923 to 1996).
Figure 13. C
omparison 1810–2000 for the three major diachronic corpora,
plus Time (from 1923 to 1996).
The larger-sized Google data on albeit shows far less variation per decade, but from the 1950s onward shows the same sharp, consistent rise
in its use.
Finally, combining these large corpora and looking at the four-­
corpora data for each word, i.e. the words in Google US/GB, COHA
and Time, converted to frequencies per 1 M words, we see the same
patterns in even stronger relief. The numbers are not fully consistent,
particularly in the earlier decades, but what clearly emerges is that the
260 From Clerks to Corpora
British frequencies are consistently a bit higher. This can of course indicate either that these three items are more markedly formal in AmE
(and thus less used), or merely that the distribution of domains (and
hence of formality) differs between BrE and AmE data in the corpora.
Given that all three AmE corpora are in striking agreement for the 20th
century, it seems probable that there is a difference between BrE and
AmE invoved. Furthermore, whether BrE or AmE, it is clear that notwithstanding and thus are dropping in use (except for thus in scientific
texts). As for albeit, the frequency per million is much lower than for
most logical and attitudinal connectors, but rumors of its demise are
clearly exaggerated (it is currently actually more frequent in AmE than
notwithstanding). The key to its revival is to be found in the distributional data from the OCE corpus: the arts and news are the major
domains for its use.
It seems reasonable to argue that earlier style mavens such as Fowler
and Gowers (1965) were particularly aware of domains such as the
arts, and less interested in stylistic uses in e.g. the sciences, so that it
would not be surprising that Gowers should become aware of the
revival of albeit—but the chronology is slightly wrong: the nadir in
its use appears to have been the period 1930–1960, with the real rise
taking place after 1965, the publication date of his revision of Fowler;
moreover, it had always been more in use in BrE (his dialect) than in
AmE, something that is even more clear today.27
6. Some final words
Since this paper has been written with a specific Stockholm Metaphor
Festival scholar in mind, it may be worth mentioning that a review of
the 2006 to 2010 articles from the Festival produced about 170,000
words in English (other languages discounted), and per million statistics of 22/M for albeit, 6/M for notwithstanding and 750/M for thus,
figures which are quite close to the 2000 data for albeit and notwithstanding, but just about double the BrE figure for thus, which is not
surprising, given that the majority of these papers were linguistically
oriented, and follow the scientific pattern seen in e.g. Figure 4.
27 Yet another recently-released web corpus, the Corpus of Global Web-Based English
(also from BYU), reports post-2000 frequencies of 20.24 per M for BrE and 12.61
per M for AmE, which is in agreement with the other corpus data. Released in 2013,
GloWbE contains 1.9 B words from the entire English-speaking world. See http://
Commentators and Corpora: Evidence About Markers of Formality 261
It turns out that by and large, our corpus data from multiple corpora
tends to be in agreement, although since these corpora are constructed
with different metainformation, they will provide us with different types
of information about matters such as style and domain. Even so, they are
clearly of great help in enriching our picture of English, not to mention
their forming the basis for the information found in our modern learner
dictionaries. If there is one specific matter which the present corpus data
suggests, it is that the Academic Word List needs to be re-examined,
based on more extensive corpus data (as Gardner & Davies 2013 does).
Corpus data can of course only produce (massive) descriptive evidence of what people are doing with English at any given time, so that
there will always be room for stylists and language police who wish
to impose prescription upon us—even if they would do well to be far
more heedful of the complexity of language, in particular the different
domains within which language operates. All our data seems to indicate
that while notwithstanding has indeed become relatively infrequent,
thus has found a niche in scientific writing, where it seems to be flourishing. But the most astonishing of our triad is clearly albeit, which has
returned from the moribund, to the joy of those who rejoice at seeing
quirks of syntax live on, albeit in frozen form.
American Heritage Dictionary of the English Language. (1970). New York:
American Heritage Publishing & Houghton Mifflin.
Biber, D., Conrad, S. & Leech, G. (2002). Longman Student Grammar of
Spoken and Written English. Harlow, UK: Pearson Education.
Cambridge Advanced Learner’s Dictionary. (2008). (3rd ed). Cambridge:
Cambridge University Press.
Cambridge Dictionary of American English. (2000). Cambridge: Cambridge
University Press.
Carter, R. & McCarthy, M. (eds) (2006). Cambridge Grammar of English.
Cambridge: Cambridge University Press.
Celce-Murcia, M. & Larsen-Freeman, D. (1983). The Grammar Book: An
ESL/EFL Teacher’s Course. Rowley, Mass: Newbury House.
Clark, S. & Pointon, G. (2009). Words: A User’s Guide. Harlow, UK: Pearson
Clear, J. (1988). Trawling the Language: Monitor Corpora. M. Snell-Hornby
(ed.), ZuriLEX ‘86 Proceedings. Tübingen: Francke.
262 From Clerks to Corpora
Collins COBUILD English Language Dictionary. (1987). (1st ed.) London:
Collins COBUILD English Usage. (1992). Glasgow: HarperCollins.
Collins, F.H. (1956). Authors’ and Printers’ Dictionary. (10th ed.) London:
Oxford University Press.
Copperud, R. (1964). A Dictionary of Usage and Style. New York: Hawthorn
Corpus Of Contemporary American English (COCA), http://www.american​
Corpus Of Historical American English (COHA),
Coxhead, A. (2000). A New Academic Wordlist. TESOL Quarterly, 34(2),
Curme, G.O. (1983 [1931]). A Grammar of the English Language, Vol. II:
Syntax. Essex, Conn: Verbatim.
Davies, M., accessed
Mar. 10, 2014.
Elfstrand, D. & Gabrielsson, A. (1960). Engelsk Grammatik för universitet och
högskolor. (4th ed.) Stockholm: Läromedelsförlagen.
Elster, C.H. (1999). The Big Book of Beastly Mispronunciations. New York:
Houghton Mifflin.
Estling Vannestål, M. (2008). A University Grammar of English with a Swedish
Perspective. Lund: Studentlitteratur.
Fisher, J.H. (ed.). (1977). The Complete Poetry and Prose of Geoffrey Chaucer.
New York: Holt, Rinehart and Winston.
Fowler, H. (1926). A Dictionary of Modern English Usage. Oxford: Oxford
University Press. See also Gowers (1965).
Fowler, H. & Fowler, F. (1930). The King’s English. (3rd ed.) Oxford: Oxford
University Press.
Gardner, D. & Davies, M. (2013). A New Academic Vocabulary List. Applied
Linguistics, doi: 10.1093/applin/amt015, published Aug. 2, 2013.
Gowers, E. (1965). Fowler’s Modern English Usage. (2nd ed.) Oxford: Oxford
University Press. See also Fowler (1926).
Greenbaum, S. & Whitcut, J. (1988). Longman Guide to English Usage.
Harlow: Longman.
Hundt, M. (1998). It is Important that This Study (Should) Be Based on the
Commentators and Corpora: Evidence About Markers of Formality 263
Analysis of Parallel Corpora: On the use of mandative subjunctive in four
major varieties of English. H. Lindquist et al. (eds) The Major Varieties of
English, Papers from MAVEN 97, Växjö: Växjö University,159–175.
Jespersen, O. (1940). A Modern English Grammar on Historical Principles, V:
Syntax (Fourth Volume). Copenhagen: Ejnar Munksgaard.
———. (1949). A Modern English Grammar on Historical Principles, I:
Sounds and spellings. Copenhagen: Ejnar Munksgaard.
Johnson, S. (1783). A Dictionary of the English Language…Abstracted from
the Folio Edition. London.
———. (1799). A Dictionary of the English Language. (8th ed.) London.
Leech, G. & Svartvik, J. (2002). A Communicative Grammar of English,
3rd ed. Harlow, UK: Pearson Education.
Longman Dictionary of Contemporary English. (2005). (4th ed.) Harlow, UK:
Pearson Education.
Longman Dictionary of Contemporary English. (2009). (5th ed.) Harlow, UK:
Pearson Education.
McEnery, T., Xiao, R & Tono, Y. (2006). Corpus-based Language Studies: An
advanced resource book. London: Routledge.
Macmillan English Dictionary for Advanced Learners. (2007). (2nd ed.)
Macmillan Education: Oxford.
Macmillan English Dictionary for Advanced Learners of American English.
(2004). Macmillan Education: Oxford.
Macquarie Dictionary. (1997). (3rd ed.). Sydney: The Macquarie Library.
Manual of Style, A. (1969). (12th ed.) Chicago: University of Chicago Press.
Mencken, H.L. (1936). The American Language. (4th ed.) New York: Alfred
A. Knopf.
———. (1948). The American Language, Supplement Two. New York: Alfred
A. Knopf.
Merriam-Webster’s Collegiate Dictionary. (2003). (11th ed.) Springfield, Mass:
Merriam-Webster’s Dictionary of English Usage. (1994). Springfield, Mass:
Minugh, D. (2002). “Her COLTISH Energy Notwithstanding”: An examination of the adposition notwithstanding. L. E. Breivik & A. Hasselgren
(eds) Language and Computers: From the COLT’s Mouth… and others’.
Amsterdam: Rodopi, 213–29.
264 From Clerks to Corpora
New Oxford Dictionary of English. (1998). Oxford: Oxford University Press.
Oshima, A. & Hogue, A. (2006). Writing Academic English. (4th ed.) Harlow,
UK: Pearson Education.
Oxford Advanced Learner’s Dictionary (ALD). (1963). (2nd ed.) Oxford:
Oxford University Press.
Oxford Advanced Learner’s Dictionary (ALD). (2005). (7th ed.) Oxford:
Oxford University Press.
Oxford Advanced Learner’s Dictionary (ALD). (2010). (8th ed.) Oxford:
Oxford University Press.
Oxford American Dictionary of Current English. (1999). New York: Oxford
University Press.
Oxford English Dictionary. (1933 [1928]). (1st ed.) Oxford: Oxford University
OED Online. Accessed Feb. 25, 2014, at
Poutsma, H. (1929). A Grammar of Late Modern English. Part I: The sentence,
2nd ed. Groningen: P. Noordhoff.
Quirk, R., Greenbaum, S, Leech, G. & Svartvik, J. (1985). A Comprehensive
Grammar of the English Language. London: Longman.
Rissanen, M. (2002). On the Development of Concessive Prepositions in
English. A. G. Fischer, G. Tottie, & H. M. Lehman (eds)Text Types and
Corpora : Studies in honour of Udo Fries. Tübingen: Gunter Narr, 191–203.
Schmid, H.-J. (2000). English Abstract Nouns as Conceptual Shells: From corpus to cognition. Berlin: Mouton de Gruyter.
Svartvik, J. & Sager, O. (1996). Engelsk universitetsgrammatik. (2nd ed.)
Stockholm: Almqvist & Wiksell.
———. (1977). Engelsk universitetsgrammatik. Stockholm: Esselte Studium.
Swan, M. (2005). Practical English Usage. (3rd ed.) Oxford: OUP.
Weber, B. (2010). Sprachlicher Ausbau: konzeptionelle Studien zur spätmittelenglischen Schriftsprache. Frankfurt am Main: Peter Lang.
Webster’s New Collegiate Dictionary. (1993). (10th ed.) Springfield, Mass:
Webster’s New International Dictionary of the English Language. (1941).
(2nd ed.) Springfield, Mass: Merriam.
Webster’s Ninth New Collegiate Dictionary. (1988). Springfield, Mass:
Commentators and Corpora: Evidence About Markers of Formality 265
Wells, J. (ed.). (2008). Longman Dictionary of English Pronunciation. (3rd ed.)
Harlow, UK: Pearson Education.
West, M. (1953). A General Service List of English Words. London: Longman,
British National Corpus,
British National Corpus,
Brown Corpus (Brown University Standard Corpus of Present-Day American
English), available via ICAME:
Corpus of Global Web-Based English.
Dutch Web Corpus. 111 M words, accessed via SketchEngine, https://www.
Frown Corpus (Freiberg-Brown Corpus of American English), available via
FLOB Corpus (Freiburg-LOB Corpus of British English), available via ICAME:
LOB Corpus (Lancaster-Oslo-Bergen Corpus of British English), available via
Oxford English Corpus (with BiWeC), 1736M words, accessed via
Time Corpus.
15 Recent Changes in the Modal Area of
Necessity and Obligation – A Contrastive
Karin Aijmer
University of Gothenburg
Recently we have witnessed a lively discussion about modal changes in
contemporary English. Leech et al. (2009) have for instance shown, on
the basis of corpora from different periods, that the modal auxiliaries
in general and must in particular have declined in frequency over the
relatively short period between 1960 and 1990. The changes affect both
the epistemic and deontic meaning but have been particularly drastic
for deontic must.
There are still many questions in the air; for example what do speakers use to compensate for the loss of the modal auxiliary (if anything).
This question has given rise to the hypothesis that the disappearance of
must is counterbalanced by the emergence and growth in the frequency
of other modal elements in particular semi-modals. However the range
of alternatives which have been studied has been fairly restricted.
The aim of my paper is to contribute to the discussion of the decline
of the modal auxiliary must by a comparison with its Swedish cognate måste which has not undergone the same semantic developments.
We can study both when must is chosen as a translation and when a
different lexical item or construction is preferred. This approach can
provide a rich panorama of expressions of obligation and necessity.
Translations can also confirm hypotheses which have been suggested
on the basis of monolingual corpora.
The structure of my paper is as follows. I will first discuss my methodology and the use of a parallel or translation corpus. I will then
1 With many thanks to Bengt Altenberg for excellent comments on an earlier version
of the article
How to cite this book chapter:
Aijmer, K. 2015. Recent Changes in the Modal Area of Necessity and Obligation – A
Contrastive Perspective. In: Shaw, P., Erman, B., Melchers, G. and Sundkvist, P. (eds) From
Clerks to Corpora: essays on the English language yesterday and today. Pp. 267–284.
Stockholm: Stockholm University Press. DOI: License:
268 From Clerks to Corpora
analyse the translation paradigms with must and alternative realisations of obligation or necessity. The translations will also provide the
raw material for a qualitative analysis contrasting the functions of must
and its most frequent competitors. I will then compare the translation
paradigms in fiction and non-fiction texts in order to look for text-type
specific differences. My paper will end with a summary and a discussion of the advantages of using data from a parallel corpus.
2. Methodology
The data for the present study are taken from the English-Swedish
Parallel Corpus (ESPC) (see Altenberg and Aijmer 2001). The corpus
contains roughly comparable original texts in English and Swedish
with their translations, altogether 2.8 million words. The Swedish and
English texts have the same size and represent the same genres namely
fiction and non-fiction texts. (see Table 1).
The relationship between måste or must in the original texts and
their correspondences in the target texts can be exhibited as a translation paradigm showing how often must and måste correspond to each
other in translation. The translations also show what the alternatives
are when must and måste are not translated into each other.
3. The frequency of must and of måste
I became interested in the on-going restructuring of the English modality system through the observation that English must seemed to be
much less frequent in English than its cognate måste in Swedish. There
were 544 examples of must in the English originals in the corpus. There
were more than twice as many examples of måste in the Swedish original texts (1104 examples). The difference in frequency is found both in
fiction and non-fiction (See Table 2).
The smaller number of examples of must in the English original texts
compared with Swedish måste is interesting against the background of
Table 1. Size of the English-Swedish Parallel Corpus.
Number of words
Recent Changes in the Modal Area of Necessity and Obligation 269
Table 2. Must and måste in the English and Swedish original texts.
Normalized figures to 10,000 words in parentheses.
English originals (‘must’)
Swedish originals (‘måste’)
210 (1.58)
454 (3.42)
334 (2.26)
650 (4.41)
544 (1.94)
1104 (3.4)
what has been claimed about the decline of must in English and the
emergence of competing variants.
The English-Swedish Parallel Corpus is designed as a bidirectional
corpus which can be used to study translations between languages in
two translation directions:
Figure 1. The structure of the English-Swedish Parallel Corpus.
We can therefore use the corpus to test the hypothesis that must
is disappearing in the translation. Must and måste are cognates and
‘favoured’ correspondences in a translation perspective. According
to Gutknecht and Rölle (1996: 237), ‘modals should preferably be
rendered by modals, because they correspond to each other in terms of
various kinds of non-specificity.’ If must is avoided in the translation
this may therefore suggest that it is losing out in the competition with
other grammaticalizing elements.
Correspondences between languages can be established by studying
translations and sources. If must is in the process of disappearing we
would also expect it to be less frequent in English sources of måste
(going from Swedish translations to English originals).
270 From Clerks to Corpora
4. Epistemic and deontic must
At the outset, a distinction needs to be made between deontic and epistemic meanings of must. Epistemic meaning has been defined in terms
of a judgment by the speaker: ‘a proposition is judged to be uncertain or probable in relation to some judgment’ (van der Auwera and
Plungian 1998: 81).
With must the speaker represents a situation as relatively certain:
1. “That’s Davina Flory.” I guessed it must be,” Burden said quietly.
Must can also be deontic. Deontic modality “identifies the enabling
or competing circumstances external to the participant as some person(s), often the speaker, and/or some social or ethical norm(s) permitting or obliging the participant to engage in the state of affairs” (van
der Auwera and Plungian 1998: 81).
2. I put a finger to his mouth: “Don’t bring it up again. You must
allow me this chance in Provence to make up my mind.” (BR1)
4.1 Epistemic meaning
The epistemic meanings of must are generally infrequent. Only 109
examples (25%) of the examples of måste (454 examples) were epistemic and 328 (75%) deontic. Moreover when we compare Swedish
måste in the epistemic meaning and English must we find a fairly high
correspondence (Table 3).
Must dominated in the translations of Swedish epistemic måste
(78.9%) and in the sources (83.5 %). The mutual correspondence
between måste and must (based on the translations in both directions)
is 81.2%.2 Epistemic must seems to be stable. In diachronic terms it
is not threatened by semi-modals which belong to the area of deontic
modality (have got to was only found three times). (See further below.)
In non-fiction texts (not shown in the table) epistemic must was even
more infrequent than in fiction (24 examples) and the translations did
not contain any alternatives.
The translations consist of modal auxiliaries (besides must) such as
could, might, will, would. Other alternatives are adverbs (obviously,
2 Mutual correspondence refers to ‘the frequency with which different (grammatical,
semantic and lexical) expressions are translated into each other (Altenberg 1999:
Recent Changes in the Modal Area of Necessity and Obligation 271
Table 3. The English Translations (SO->ET) and sources of Swedish måste
(ST <-EO). Epistemic meanings. Fiction only.
86 (78.9%)
would (maybe)
have got to
(you) can be sure
could (perhaps)
I suppose
perhaps would
sounds like
must surely
of course
I suppose
it is certain to
91 (83.5%)
177 (81.2%)
presumably, doubtless, maybe, perhaps) or verbs and adjectives (you
can be sure, I suppose, sounds like).
4.2 Deontic modality
Deontic modality (obligation/necessity) is a more complex semantic
field than epistemic modality. As shown by its translations deontic
must has many competitors or what Leech et al. (2009) describe in
diachronic terms as the present-day beneficiaries of must’s decline. For
this reason I will focus on the deontic must and its competitors in the
domain of obligation and necessity (Table 4).
272 From Clerks to Corpora
Table 4. The English translations and sources of Swedish måste (SO ->ET
and ST<-EO). Deontic meanings. Fiction only.
have to
had to
91 (27.7%)
91 (27.7%)
91 (30.5%)
81 (27.2%)
182 (29.1%)
172 (27.5%)
(have) got to
need to
need (main verb)
is to/was to
ought to
be going to
had better
make sb do sth
it meant -ing
be due to
78 (23.8%)
14 (4.3%)
38 (12.6%)
19 (6.4%)
116 (18.5%)
33 (5.3%)
NP modification
can’t wait to
be in need of
be expected to
it was natural for X to
couldn’t possibly
past tense
could not help
it does not necessarily
follow that
emphatic do
it’s time
be required
be forced to
Recent Changes in the Modal Area of Necessity and Obligation 273
Table 4. Continued
know to do something
be obliged to
I don’t mind
could only
The percentages are based on 328 examples in the translations and
298 examples in the sources. In all there were 39 different translation
alternatives. There is a low degree of mutual correspondence between
måste and must. The mutual correspondence of must was 18.5% to be
compared with have to which had a mutual correspondence of 29.1%.
The higher frequency of must in the English translations than in sources
may be induced by the high frequency of ‘måste’ in the Swedish original; cf Johansson (2007: 32–33) a ‘translation effect’. The translator
uses a cognate even when a more ‘idiomatic’ translation is available.
Have to was used in translations (27.7%) and in sources (30.5%). The
frequency of have to would have been even higher if I had included
had to. However had to has been used as an alternative of must for
syntactic reasons. It was used in roughly 23% of the examples (both in
translations and sources) as a past tense form mainly in narrative contexts. Had better is associated with weaker obligation than have to and
it only occurred three times as a correspondence. Need to was more
frequent than should but not as frequent as have got to.
The translations draw attention to the fact that must has a large
number of co-players in the domain of obligation/necessity. Many different grammatical categories are represented in the translations (see
Table 5).
Other modal auxiliaries than must are should, ought to, will/would.
A difference between English and Swedish is that English can express
deontic modality by means of semi-modals. ‘ “ Semi-modals” are not
full modals but are verb constructions (…) which have been moving
along the path of grammaticalization and have gradually acquired an
274 From Clerks to Corpora
Table 5. Grammatical categories of deontic modality.
modal auxiliaries
modal adverbs
modal adjectives
modal nouns
lexical verbs
NP modification
must, should, ought to, will/would
have to, need to, have got to, be to, be
due to, had better
necessarily, inevitably
it was natural for X to
be in need of
be required, be forced to, be obliged to
I cannot wait to do, I don’t mind doing
let us VP
three points to consider
emphatic do, prosodic stress
auxiliary-like function’ (Leech et al. 2009: 91). The semi-modals with
the meaning of obligation/ necessity mostly contain have. Have to, have
got to, had better are all semi-modals. Less frequent semi-modals are be
to, be due to. In addition we find modal adjectives and adverbs (it was
natural for X to, necessarily, inevitably), verbo-nominal expressions of
modality (be in need of), lexical verbs (be obliged, forced, compelled).
More idiomatic translations are for example I cannot wait to do. The
deontic must also shares directional force with the imperative as shown
by the translations. Three points to consider can be paraphrased ‘three
points which need to or must be considered’. The translation is an
example of NP modification. There is also a close association between
deontic modality and emphasis.
4.3 Summing up deontic modality in fiction
Grammaticalization and change are lurking in the background when
we interpret the translation correspondences. The frequency of a certain translation can reflect its status as a ‘substitute’ of the declining
must. In view of the diachronic findings about the decline of must it is
not surprising that have to is more frequent than must in translation.
The increase of have to in fiction is remarkable compared to other studies. Biber et al. (1999: 489) have compared the frequencies of modals
and semi-modals in different registers. Must and have to had the same
frequency in fiction but have to was more frequent in the conversational data.
Recent Changes in the Modal Area of Necessity and Obligation 275
4.4 Competition between must and have to
The translation paradigms only provide the raw material for the semantic analysis. The meanings range from strong obligation or necessity
(represented by must and have to) to weaker elements such as should
or need to. Have to and must differ semantically. Have to (unlike must)
refers to what van der Auwera and Plungian (1998:81) have described
as participant-external necessity. Participant-external necessity makes
reference to the circumstances that are external to the participants and
make a state of affairs necessary.
3. För att hålla mig igång krävs inte mer än ett par 1,5 volts
Jag omsätter samma mängd energi som en
Vattenlösningen, som jag vilar i, måste vara helt
steril. (PCJ1)
To keep me going requires nothing but a couple of 1.5-volt batteries.
I consume no more energy than a 20-watt bulb.
The aqueous
solution in which I rest has to be absolutely sterile.
In this example the source is clearly not the speaker but the properties of the aqueous solution.
If must is declining and have to is increasing in frequency we would
expect the boundaries between them to be drawn up differently. The
translations can be the basis for a deeper and more detailed analysis of
the variation between have to and must. A factor such as the person of
the subject has the advantage that it can be compared in the translations.
In the English Swedish Parallel Corpus have to was more frequent
than must with an impersonal subject (e.g. generic you) than with the
‘direct’ you. It was also frequent with we as the subject (collective we)
Table 6. Variation between have to and must with different subjects.
Type of subject
have to
you generic
you direct
animate subject
we collective
non-animate subject
276 From Clerks to Corpora
or with the passive. Must on the other hand was more frequent with a
first person subject and with the ‘direct’ you (Table 6).
The use of have to with a generic subject is illustrated in the following example:
4. Man måste lära sig ta skydd. (JMY1 )
You have to learn how to take cover. (JM1T)
We can also be used impersonally:
5. Nu när det är krig måste man hjälpa varandra. (JMY1)
“Now that the war is on we have to help each other.” (JMY1T)
Have to is used in an abstract way to make a recommendation (if
something is favourable) or an instruction formulated in general terms.
The following example illustrates that must can have a strong emotional meaning urging the hearer to do something. According to Smith
(2003: 259) such insistence can however sound odd in present-day
English: ‘Even where MUST is used with no obvious hint of speaker-imposed deontic meaning. …, in Present Day English it is liable to be
perceived as off, perhaps because it sounds unduly insistent’.
6. Ja, faster måste förstå mej: det är en ära att få arbeta ihop med ett
sånt snille. (ARP1)
“Yes. You must understand me, Auntie. It’s an honour to be allowed
to work with such a genius.” (ARP1T)
Moreover have to unlike must can be softened and is therefore used
when more politeness is required:
7. Då måste du stanna hemma från skolan och passa henne. (GT1)
“Then you ll have to stay home from school and take care of her.”
8. Då måste jag sätta mig hos ålen. (KE1)
Then I ’ll have to sit with the eel. (KE1T)
Other examples of softening are illustrated by would have to and
might have to (mitigating an inconvenience): The speaker’s staying for
a month may involve an inconvenience for the hearer:
9. Kanske måste jag stanna hos er en hel månad. (KOB1)
“I might have to stay with you for a whole month.” (KOB1T)
Recent Changes in the Modal Area of Necessity and Obligation 277
Checking in with the concierge involves some extra effort for the
10. A visitor would have to check in with the concierge. (FF1)
En besökare måste anmäla sig hos portvakten. (FF1T)
5. Non-fiction translations
By including non-fiction in the study of must and its competitors we
can get a more detailed and richer picture of the expressions of obligation and necessity. The number of examples is higher in non-fiction
than in fiction texts. There were 526 examples in translation of måste
and 412 examples in sources (deontic examples only). There were 41
different competing forms (most of them occurring only once or twice).
See Table 7:
Table 7. English translations and sources of Swedish ‘måste’.
have to
need to
had to
need main verb
mean V-ing
ought to
to be -ed
it is necessary
oblige to
have got to
was made to
will inevitably
(what may happen is) for X to V
it is essential
295 (56%)
63 (12%)
27 (5.1%)
48 (9.1%)
17 (3.2%)
177 (43%)
100 (24.3%)
47 (11.4%)
16 (3.9%)
44 (10.7%)
472 (50.3%)
163 (17.4%)
74 (7.9%)
64 (6.8%)
61 (6.5%)
278 From Clerks to Corpora
Table 7. Continued
will be to
should like to
be bound to
was compelled to
I have no choice but
I cannot help but
agree on the need to
appreciate the need to
be of the need to
to be compelled to
I would note
subject to
was to
entail the need
I regret to say
I’m sorry to tell you
embarrassed to speak to me
forced to
recognize the necessity of
Måste was translated into ‘must’ in 56% of the examples and in the
examples with English sources it was found in 50.3% of the examples.
Have to was chosen less often. Need to and should are also among the
most frequent correspondences. It is interesting to make comparisons
with fiction where the percentages of the most frequent variants are
quite different. The ranking of the most frequent variants in fiction and
non-fiction is shown in Table 8 (percentages only).
Must was more frequent both in translations and in sources in
non-fiction reflecting the fact that must has not declined in frequency to
the same extent as in fiction. Have to, on the other hand is less frequent
in non-fiction where it is ranked below must. Had to is more frequent
Recent Changes in the Modal Area of Necessity and Obligation 279
Table 8. Ranking of the most frequent variants in fiction and non-fiction texts.
have to
need to
had to
(have) got to
17 (2.7%)
8 (1.3%)
in fiction reflecting the fact that it is associated with narrative contexts.
Have got to (gotta) was rare in non-fiction. On the other hand. need
to, should (and ought to) were strikingly more frequent in non-fiction
than in fiction.
5.1 Text-type specific correspondences
Certain correspondences are text-type specific. Have to, need to and
should function as rhetorical devices in non-fiction texts (for example
EU regulations) imposing an obligation also when no specific individual is mentioned. The following example uses have to (and must).
11. Den andra faktorn är att vi måste se till att skaffa en utbildning
som går att använda under lång tid när vi skaffar oss en utbildning.
Det måste vara en bred grundutbildning, eftersom samhället
förändras i allt snabbare takt.
Det går inte att ha snabba utbildningar.
Vidare måste det också vara ett livslångt lärande. (EAND1)
The other factor is that we must ensure that when we obtain an education we obtain one which can be used for a long time.
There has
to be a broad basic education, because society is changing ever more
It is not possible to have a quick education.
there has to also be life-long learning. (EAND1T)
The reference is to a situation in the future when have to is used. The
speaker envisages a broad basic education for everyone. The obligation
is only weak since no individual is under the obligation to do something. Must in the same sentence implies greater imposition (we must
ensure that we obtain a broad discussion even in the face of resistance).
Should is weaker than have to or must. It merely expresses that the
situation referred to is favourable to the speaker, the hearer or to people
in general:
280 From Clerks to Corpora
12. Det viktigaste måste väl ändå vara Sveriges ekonomi och dess
förmåga att kunna ’platsa’ i sällskapet när det gäller inflation, räntevillkor osv. (EAND1)
The most important aspects should still be Sweden’s economy and its
eligibility for a place in the club in terms of inflation, interest rates
and so on. (EAND1T)
When should and stronger deontic forms are used in the same context they overlap semantically. The ordering between must and should
could be changed without any difference in meaning:
13. Särskild vikt måste läggas vid tidig förvarning och tidigt agerande i konfliktlösning.
Förebyggande diplomati måste ytterligare
There must be a particular focus on early warning and early action in
conflict resolution.
Preventive diplomacy should be further strengthened. (LHW1T)
Need to is particularly frequent in non-fiction texts. However it is
not used with its basic meaning of internal necessity or compulsion
but in a more abstract sense imposing an obligation on a non-specific
individual. Because of its basic meaning need to (‘having a need’) is
especially appropriate to express that something is a desirable goal or
in the hearer’s best interest. Like should and have to it was frequent
with the collective we or with a following passive. The combination
we need to was used as a correspondence (as a translation and as a
source) in 34 examples or almost half of the examples of need to (also
when the Swedish original did not contain ‘we’) and need with a following passive verb in 25 examples. When the subject was not we it
was for instance ‘Countries of the European Union’ or ‘Swedes living
and working abroad’. Leech et al. (2009: 111) emphasise the strategic or manipulative function of need to: ‘Here a double mitigation of
imperative force occurs: not only is obligation represented as in the
best interests of ‘us’, but by referring to ‘we’ rather than ‘you’ as the
people with the need, the writer imposes a collective obligation on an
often rather vague community of people including the addresser and
the addressees’. 3
3 Nokkonen (2006: 48) also points out cases where ’we need’ is used in an impersonal, ’strong’ way. She finds examples of this use in informative genres in the FLOB
Recent Changes in the Modal Area of Necessity and Obligation 281
14. Det måste bli en omprövning av de traditionella attityderna gentemot äldre och de roller som man vill ge dem. Speciellt gäller detta
på arbetsmarknadsområdet. (EISC1)
We need to review our traditional attitudes towards senior citizens
and rethink the roles we expect them to play in society. This applies
particularly to the world of work. (EISC1T)
Rather than saying ‘you must’ (which is strongly impositive), the
more polite we need to is used strategically as a way of urging an individual or the community in general to do something.
With a passive following need and a third person subject no direct
reference is made to the speaker and hearer:
15. Flexibiliteten för medlemsstaterna måste matchas av en grupp
indikatorer som skall identifiera behovet. (EMCC1T)
The flexibility for Member States needs to be matched by a range of
indicators to identify need. (EMCC1)
The use of need conveys that the action (matching the flexibility of
EU member states by certain criteria) is judged to be favourable (needs
to be done). The imposition is only expressed weakly since it is not
directed to a special individual.
There is a great deal of overlap between need to and other markers as
indicated by examples where they are used in the same neighbourhood:
16. Alla bidrag för att nå Kyotomålen måste användas, men man
måste också ha deras inbördes relationer klara för sig;.
Maximum efforts should be made to meet the Kyoto targets, but we
still need to keep a sense of proportion.
Should and need are used in a similar way. Their ordering can therefore be exchanged without a difference in meaning.
6. Conclusion
The present study can be seen as a complement to comparative historical corpus studies of must and its changes over time. The developments and changes in the area of modality which have taken place
between 1960 and 1990 are at least to some extent visible in translations. Translations can therefore confirm observations which have been
made on the basis of monolingual corpora about the decline of must
and the emergence of semi-modals and other variants.
• The translations confirm the observation that must has declined
above all in the area of deontic modality
282 From Clerks to Corpora
• The translations also confirm the hypothesis that semi-modals are
becoming more frequent to fill the gap left by must
• The translations also confirm the proposal that other modal
­auxiliaries such as need to and should compete with must and
have to
The translations also showed that:
• must was more frequent than have to in non-fiction suggesting
that the decline of must has made less progress there
• should/ought to and need were more frequent in non-fiction than
in fiction
• have got to was infrequent in non-fiction
Translations also have certain advantages over monolingual corpora.
In a monolingual corpus the range of forms with obligation/necessity
meanings competing with must is not apparent. Monolingual studies
have mainly discussed the rivalry between must and a few selected
Translations on the other hand provide a large number of variants
of must. They may therefore add something to the picture of who the
players are in the semantic domain of obligation and necessity. As an
extra bonus they can also contribute to the discussion of the factors
motivating the choice of a particular form. Leech et al. (2009: 114) used
the term ‘ecology’ to capture the idea that each form [in the same field of
meaning] ‘evolves its own niche in the expression of modality, expanding, contracting or maintaining its “habitat” in relation to other, partially
competing, forms’. The translations show that obligation and necessity
can be expressed in many different ways and that there are conventions
for how the different forms are used. In fiction have to (unlike must) was
generally used with generic or impersonal subjects to make recommendations or to give instructions. The area of semantic overlap between
must and have to is therefore restricted to certain contexts. Have to was
used as a mitigator unlike must which was insistent and emotional.
In non-fiction texts have to, should and need to were typically used
as rhetorical strategies when the speaker addresses a vague community
of individuals. They were for instance used with a similar function in
contexts with the passive or we as the subject. However the high frequency of need to and should may also have to do with their basic
meaning to refer to what is beneficial or the right thing to do.
Recent Changes in the Modal Area of Necessity and Obligation 283
Primary sources:
Fiction –Swedish originals
Pettersson, Allan Rune, Frankenstein’s faster-igen. Stockholm 1989.
Tunström, Göran, Juloratoriet. Stockholm 1983.
Mankell, Henning, Den vita lejoninnan. Stockholm 1993.
Ekman, Kerstin, Händelser vid vatten. Stockholm 1993.
Bornemark, Kjell-Olof, Handgången man. Stockholm 1986.
Scherer, Maria, Kejsarvalsen. Stockholm 1983.
Fiction –English originals
Brink, André, The wall of the plague. London 1984.
Forsyth, Frederick, The fourth protocol. London 1984
Rendell, Ruth, Kissing the gunner’s daughter. London 1992.
Non-fiction-Swedish translator/author
EAND1 Europaparlamentets överläggningar.
EISC1 Europaparlamentets överläggningar.
Secondary sources:
Altenberg. B. (1999). Adverbial connectors in English and Swedish: Semantic
and lexical correspondences. In Hasselgård, H. & S. Oksefjell (eds) Out of
Corpora. Studies in Honour of Stig Johansson. Amsterdam and Atlanta,
GA: Rodopi, 249–268.
Altenberg, B. & K. Aijmer. (2001). The English-Swedish Parallel Corpus: A
resource for contrastive research and translation studies. Mair, C. & M.
Hundt (eds) Corpus Linguistics and Linguistic Theory. Papers from the 20th
International Conference on English Language Research on Computerized
Corpora (ICAME 20), Freiburg im Breisgau 1999. Amsterdam and Atlanta,
GA: Rodopi, 15–33.
Biber, D., S. Johansson & G. Leech. (1999). Longman Grammar of Spoken and
Written English. London: Longman.
Gutknecht, C. & L.J. Rölle. (1996). Translating by Factors. New York: State
University of New York Press.
Johansson, S. (2007). Seeing Through Multilingual Corpora. On the Use of
Corpora in Contrastive Studies. Amsterdam/Philadelphia: John Benjamins.
284 From Clerks to Corpora
Leech, G., M. Hundt, C. Mair,. & N. Smith. (2009). Change in Contemporary
English: a Grammatical Study. Cambridge: Cambridge University Press.
Nokkonen, S. (2006).The semantic variation of NEED TO in four recent
British English corpora. International Journal of Corpus Linguistics 11:1,
Smith, N. (2003). Changes in the modals and semi-modals of strong obligation and epistemic necessity in recent British English. In Facchinetti, R., M.
Krug & F. Palmer (eds) Modality in Contemporary English. Berlin and New
York: Mouton de Gruyter, 240–266.
van der Auwera, J. & V. Plungian. (1998). Modality’s semantic map. Linguistic
Typology 2: 79–124.
16 M
otion to and Motion through: Evidence
from a Multilingual Corpus 1
Thomas Egan
Hedmark University College
1. Introduction
Over the course of the last twenty years, there has been a considerable amount of comparative research into the coding of motion events
in various languages (see, for instance, Filipović & Jaszczolt 2012,
Hickmann & Robert 2006, Viberg 1998, 2003, 2013). This research
has led to a reassessment and subsequent refinement of Talmy’s typology (1991, 2000), according to which languages are said to be either
satellite-framed or path-framed. According to Talmy, satellite-framed
languages, such as English, tend to code manner of motion in the
verb and path of motion in an adverbial (particle) in self-motion constructions, i.e constructions in which it is the syntactic subject which
moves, as in He walked to work. Path-framed languages, such as
Spanish, tend to code path in the verb and manner, if at all, in an
adverbial. In recent years this typology has been expanded to include
so-called equipollent framing, found in various serial verb languages
(Slobin 2006). The clear dichotomy proposed by Talmy has also been
nuanced by scholars who point to the co-existence of several patterns of framing in one and the same language. Kopecka (2006) and
Pourcel and Kopecka (2005), for example, propose such a hybrid situation for French.
In this chapter I take a fresh look at satellite- and path-framing in English
and French in a comparative study of codings of self-motion predications
1 I would like to thank the editors for inviting me to contribute to this volume in honour of Nils-Lennart. I would also like to thank two anonymous referees for helpful
and insightful comments on my chapter.
How to cite this book chapter:
Egan, T. 2015. Motion to and Motion through: Evidence from a Multilingual Corpus.
In: Shaw, P., Erman, B., Melchers, G. and Sundkvist, P. (eds) From Clerks to Corpora:
essays on the English language yesterday and today. Pp. 285–302. Stockholm: Stockholm
University Press. DOI: License: CC-BY.
286 From Clerks to Corpora
in which the path either traverses a space or area (‘through-ness’) or leads
up to a goal (‘to-ness’). The reason for choosing these two path types is
that French in particular is said to avoid the use of manner verbs with telic
actions in general (Aske 1989: 6) and actions involving boundary-crossing in particular (Cappelle 2012: 189). In order to carry out a comparison
of types of predication in two languages one needs a reliable tertium comparationis (see Jaszczolt 2003, Johansson 2007: 39, Krzeszowski 1990:
15). Much earlier research into the ways in which languages code motion
events made use of a tertium comparationis in the form of events in a picture book (such as the Frog story: see Berman & Slobin 1994) or in short
video snippets, which are described by participants in the experiment. The
Oslo Multilingual Corpus (OMC), provides the tertium comparationis for
the present study, where expressions in a source language serve as grounds
for the comparison of their translations into two or more languages (see
Egan 2013, Egan & Rawoens 2013).
I take as my starting point Norwegian predications in the OMC
of self-motion events containing two path prepositions, til (= to) and
gjennom (= through), and compare the English and French translations
of these predications. In section 2 I introduce the corpus and explain
briefly why I consider such a corpus to be suitable for this sort of study.
Sections 3 and 4 compare English and French renderings of the notion
of ‘through-ness’ and ‘to-ness’, as these are coded by the Norwegian
prepositions gjennom and til. The results of the investigation of the two
sorts of path predications are compared in section 5. Finally, section 6
contains a summary and conclusion.
2. Multilingual corpora as sources of tertia comparationis
In a comparative study such as the present one, which is based on
English and French translations of Norwegian predications, the tertium comparationis is given by the original Norwegian texts. This
tertium comparationis can, of course, only be viewed as a guarantor
of semantic equivalence between the English and French expressions
to the extent that the translators have aimed to convey as much as
possible of the meaning of the original texts. My own experience of
working with translation corpora has led me to believe that professional translators try to convey as much as possible of the sense of
the original text most of the time. One will inevitably come across
instances of mistranslation or non-translation, but the former are
very rare in the OMC in my experience, which in itself testifies to the
Motion to and Motion through: Evidence from a Multilingual Corpus 287
quality of the translations in the corpus. More common than mistranslation is the complete omission by a translator of a predication
in the original text. In cases where one of two translators whose texts
are being compared omits to translate a predication in the source
text, one has no grounds for comparing them. Both the translated
and non-translated version of this item must accordingly be excluded
from the comparative study.
The OMC was compiled under the direction of Stig Johansson (see
Johansson 2007). The No-En-Fr-Ge part of the corpus contains texts in
four languages, consisting of long extracts from five Norwegian novels translated into English, French and German. For the present study I
only looked at the English and French translations. A multilingual corpus
has at least two advantages over a bilingual translation corpus. In the
first place it allows for the comparison of identical text types, in that
both texts being compared are translations, whereas in working with
a bilingual translation corpus one is comparing an original text with a
translation. Given that translated texts differ from original texts along
various parameters, it makes obvious sense to compare one translation to
another translation. In the second place, the examination of comparable
translations in a multilingual corpus allows us to estimate the overlap
between equivalent expressions in the two languages being compared.
My tertium comparationis comprises all Norwegian predications of
self-motion events containing the two path prepositions, gjennom and
til in the OMC. I downloaded all tokens containing the two forms in
the corpus, then extracted all tokens coding motion events and finally
discarded tokens coding caused motion, by which I mean tokens containing an explicit causer who/which causes someone/something to
move along a path (such as ‘She drove him to work’). All predications of motion without an explicit causer, in other words all S-V-A
sentences, were categorised as coding self-motion, irrespective of the
degree of agentivity of the mover. A glance at a bilingual dictionary or
contrastive grammar will show that Norwegian gjennom codes relations that may be rendered in English by through and in French by
à travers, among other prepositions. Similarly, Norwegian til codes
relations that may be rendered in English by to and in French by à. My
primary interest, however, is not in the correspondences between the
Norwegian original and its translations into the other two languages,
but in the correspondences between the two sets of translations. To this
end, having extracted all the occurrences in the OMC of Norwegian
gjennom (110 tokens) and til (664 tokens) in self-motion predications,
288 From Clerks to Corpora
the Norwegian originals were set aside and comparisons drawn
between the English and French renderings of these predications.
One point that should be made about the data is that Norwegian is a
satellite-framed language like English but unlike French.2 Moreover, all
the original Norwegian tokens contain a path preposition. Slobin (2006:
70) claims that “in translations […] manner salience follows patterns
of the target, rather than source language”. If he is correct, the fact that
Norwegian is satellite-framed should not affect the results of the comparison, at least as far as coding of manner is concerned. This contention of
Slobin’s has, however, been disputed by Cappelle (2012), who maintains
that translations of motion predications will tend to some extent to borrow the form of the original text, irrespective of typological differences
between the two languages involved. One should bear this argument in
mind in interpreting the data presented in the next two sections.
3. English and French strategies for coding ‘through-ness’
As mentioned in the previous section, the tertium comparationis for my
analysis of ‘through-ness’ consists of codings in Norwegian of this concept by means of the preposition/particle gjennom. The main definition
of gjennom in Norsk Ordbok, the closest Norwegian equivalent to the
OED, is:
[U]sed about a movement or something perceived as motion which
takes place in (within, surrounded by) that which is encoded by the
landmark [i.e. the prepositional object] from one end or side all the
way to the other, containing the whole landmark from start to finish;
(in) from the one side or end and (out) to the other. (Norsk Ordbok
2002, my translation)3
2 In fact Norwegian is rather more satellite-framed than English. As a result of the
Norman conquest English contains path verbs such as enter and descend, where
Norwegian has a combination of a verb and a particle. One reviewer points out
that Old English already contained some path verbs of native origin such as stigan
which can be used in the sense ‘ascend’. This is certainly true, but it is also true of
Old Norse, with the verb stiga being cognate with OE stigan, for instance. The point
is that the number of such verbs in English increased in Middle English.
3 The original definition reads: “Gjennom el igjennom prep, adv 1 a) brukt om rørsle
el noko oppfatta som rørsle for å uttrykkja at ho går føre seg i (innanfor, omslutta
av) det som styringa nemner frå den eine enden, den eine sida heilt ut til (på) den
andre , at ho omfattar det som er uttrykt i styringa frå byrjing til slutt; (inn) frå den
eine og (ut) til den andre sida el enden av”. (Norsk Ordbok 2002)
Motion to and Motion through: Evidence from a Multilingual Corpus 289
This definition of Norwegian gjennom is very similar to standard
definitions of the prototype of through, as described for example by
Egan (2012: 44), Lee (2001: 49), Leech (1969: 181), Lindstromberg
(1998: 31), and Tyler and Evans (2003: 219). The similarity in the prototypical senses of Norwegian gjennom and English through is reflected
in the number of occurrences of through in the English translations, 81
of which (74%) contain the form. This prototypical sense of through is
illustrated here by examples (1) and (2), with the path in italics.4
(1) a.We began to walk slowly through the galleries, and up to the
first floor. (JG3TE)
Nous déambulâmes un moment à travers les salles, puis
montâmes au premier étage. (JG3TF)
(2) a.I suggested I could walk with him through the Retiro Park.
b. Je proposai de l’accompagner à travers le Retiro. (JG3TF)
The French versions of (1) and (2) both contain the preposition
à travers. The prototypical sense of this preposition, according to the definition in Dictionnaire de l’Académie française, differs from its English
counterpart in emphasising the central portion of the landmark, the space
or area through which the path extends (often referred to as a ‘container’
in the literature), backgrounding the elements of entrance and exit.
À TRAVERS, AU TRAVERS DE, locative preposition. Going from
one side to the other, crossing: À travers is mostly used to code an
open or free passage; Au travers de on the other hand is used to
code a passage made between obstacles, or crossing or penetrating
an obstacle; however this distinction is not rigorously observed.
(Dictionnaire de l’Académie française, 8e edition: my translation)5
4 The English and French versions of all tokens are cited in the text itself. The first
part of the code, ‘JG3’ in (1) for example, refers to the text in the OMC from
which the example has been taken. ‘TE’ stands for translated text in English, ‘TF’
translated text in French. The corresponding Norwegian originals are listed in
an appendix.
5The 8th rather than the 9th, and most recent, edition of the dictionary has been
used, as the online version of the 9th edition had not reached the letter ‘t’ at the
time of writing. The original definition reads “À TRAVERS, AU TRAVERS DE, loc.
prép. En allant d’un bord à l’autre, en traversant. À travers se dit principalement
pour désigner un Passage vide, libre; Au travers de se dit plutôt, au contraire, pour
désigner un Passage qu’on se fait entre des obstacles, ou en traversant, en pénétrant
un obstacle; mais cette distinction n’est pas toujours rigoureusement observe”
(Dictionnaire de l’Académie française, 8e edition).
290 From Clerks to Corpora
Note that the French definition of á travers, although not au travers
de, differs from its English and Norwegian counterparts in making no
reference to either entry into nor exit from a container. It is perhaps
therefore not surprising that there are only 22 tokens in the French
translations (20%) in which path is coded by à travers. Indeed French
translators actually prefer to employ prepositions other than à travers
to code motion ‘through-ness’. Moreover, while the basic sense of
‘through-ness’, as coded by Norwegian gjennom, is a path relation
involving subcomponents of ‘entrance to’, ‘crossing of’ and ‘exit from’
an area or container, only the central portion of the path, denoting the
crossing of the area or container, is salient in all tokens in Norwegian.
(Note that in the translation of the definition of gjennom above, the
prepositions in and out are enclosed in brackets.) In addition, many
of the French translations contain a prepositional phrase coding the
area or container within which the motion event takes place (the ‘site’)
rather than the actual path taken through the site by the mover.6 Such
translations either code the path in the verb, as in (3b), or leave it up to
the addressee to infer the extension of the path, as in (4b). There are 20
examples of the preposition dans in the French texts, as opposed to just
two of in in the English ones.
(3) a.During the summer, Dina began wandering about the house.
b. Dina recommença à circuler dans la maison cet été-là. (HW2TF)
(4) a.When he walked through town ... (BHH1TE)
b. Lorsqu’il déambulait dans la ville ... (BHH1TF)
In the English version of (3) the prepositional phrase codes similar
information to the verb in the French version, that the path extends in
a non-linear fashion throughout the space comprised by the dwelling.
The prepositional phrase in the French version just denotes the locus
(site) of movement. Similarly, in (4a) we are given to understand that
the path extends from (near to) one side of the town to the other, while
from (4b) we can merely surmise that the movement took place within
the confines of the town.
6 Note that there are no tokens in which site is encoded by a verb in the present study,
since these would be analysed as predications of location rather than motion and as
such would per definition have been excluded from the data under examination.
Motion to and Motion through: Evidence from a Multilingual Corpus 291
If an area to be crossed is very small, or even two-dimensional, the
translator does not have the option of concentrating on the central portion of the path. Aurnague (2000) dubs such small areas “intrinsically
medial spaces”. In such cases French translators sometimes choose to
code the path in the verb phrase and encode the point of boundary
crossing by the preposition par, as in (5b) and (6b).
(5) a.Then crawled through the open window. (HW2TE)
b.Elle sortait ensuite par la fenêtre. (HW2TF)
(6) a.One spring a duck entered the cookhouse through the open
door (HW2TE)
b.Une année, une mère eider entra par la porte ouverte du fournil
... (HW2TF)
There are 13 instances of par in the French texts, but not a single one
of English by, although the use of the latter in (6a), in which the path
is coded by the verb, would result in an idiomatic English utterance. It
would not, however, be possible to substitute by for through in (5a), in
which the verb codes manner, without changing the direction of motion.
In the French versions of examples (3), (5) and (6) the path is coded by
the verb rather than, or in addition to, an adverbial. This option is chosen by the French translators in just over half of the tokens (57 of 110).
By far the most popular verb is traverser (32 tokens), followed by entrer
(6 tokens), franchir (5 tokens) and passer and sortir (3 tokens each).
Manner, like path, may also be coded by a verb, by an adverbial, or
both. In example (1), for instance, manner is coded by the verb in both
translations. In (7) and (8), on the other hand, in which the form coding manner is underlined and the form coding path in italics, manner is
coded by the verb in English with path being coded in an adverbial, but
in an adverbial in French with path being coded by the verb.
(7) a.Later I slipped through the door of the Grand Café …
b.Après quoi j’ai discrètement franchi la porte du Grand Café ...
(8) a.She ran through the rooms wearing only pantalets … (HW2TE)
b. En pantalon, et en courant, elle traversa la pièce ... (HW2TF)
Figure 1 provides details of how often the two sets of translations
code manner, path and site in verbs, adverbials or both of these.
292 From Clerks to Corpora
100 90 80 70 60 50 40 30 20 10 0 Manner Path Site Just verb Just Adv. Verb + Adv. Just verb English Just Adv. Verb + Adv. French Figure 1. Manner, path and site in English and French codings of self-motion
We can see in Figure 1 that the English texts overwhelmingly code
manner in the verb. Equally clear-cut is the tendency for path in English
to be coded in an adverbial, rather than the verb. In other words, the
evidence of the texts in the present study tends to confirm the view that
English is indeed a satellite-framed language. The picture for French is
more mixed with respect to Talmy’s (2000: 221) typological distinction.
The number of tokens coding path in the verb testify to it being pathframed to a much greater extent than English, but there is a sizable
minority of tokens in which manner is coded in the verb, far more than
one would have expected had French been a pure verb-framed language. Rather it appears to be predominantly verb-framed, but with a
number of alternative possibilities for coding manner and path, as has
been pointed out by Kopecka (2006) and Pourcel and Kopecka (2005).
4. English and French strategies for coding ‘to-ness’
The tertium comparationis for my analysis of ‘to-ness’ consists of codings in Norwegian of self-motion predications by means of the preposition/particle til. The definition of til in Norsk Riksmålsordbok may be
translated as follows7:
7 The definition is taken from Riksmålordbok rather than Den Norske ordboka since
the latter had not reached the letter ‘t’ at the time of writing. The original definition
reads: “brukt for å uttrykke at det styrte ord betegner bestemmelsessted, mål ell.
sluttpunkt for en bevegelse, at det nevnte sted ell. område blir nådd ell. skal nås”
(Norsk Riksmålsordbok: 1983)
Motion to and Motion through: Evidence from a Multilingual Corpus 293
used to express that the word governed codes a target-place, a goal
or the endpoint of a movement, so that the place or area mentioned
is reached or will be reached (Norsk Riksmålsordbok 1983: my
This definition of Norwegian til is very similar to that of English to in
the OED and French à in Dictionnaire de l’Académie.
To: Expressing a spatial or local relation. Expressing motion directed
towards and reaching: governing a n. denoting the place, thing, or
person approached and reached. The opposite of from. (OED)
À introduces a complement denoting a place: 1. The place towards
which there is a movment, in the direction of which one is heading.
(Dictionnaire de l’Académie, neuvième édition, my translation)8
Given the similarity between the definitions of to and à and that
of til, it comes as no surprise that a large number of paths coded in
Norwegian by the latter are coded in a similar fashion in both English
and French. (9) and (10) may serve as typical examples.
(9) a.I ran up to the window… (BHH1TE)
b.Je me suis précipité à la fenêtre… (BHH1TF)
(10) a. “Then they go to church!” Dina commented. (HW2TE)
b. “Et après, elles courent à l’église!” fut le commentaire de Dina.
There are as many as 552 English tokens containing to and 354
French tokens with à. The difference is partly due to a greater tendency
for French to code path in the verb. However, in addition, the French
translators tend to specify the extent to which the landmark has been
actually reached (jusqu’à), as in (11), or whether the mover is still in the
process of approaching the target (vers), as in (12).
(11) a.And calmly strolled across the room to the window! (HW2TE)
Et traversait tranquillement la pièce jusqu’à la fenêtre!
(12) a.Ana raced to the jeep and returned with a small video camera
… (JG3TE)
8 À introduit un complément désignant un lieu : ✩ 1. Le lieu vers lequel il y a mouvement, vers lequel on va. (Dictionnaire de l’Académie, neuvième édition.)
294 From Clerks to Corpora
b.Ana se précipita vers la Jeep et revint avec une caméra de
poing… (JG3TF)
There are 40 French tokens containing the preposition jusqu’à and
90 with vers. In many of the latter there is actually no doubt that the
target has been reached – thus in (12b) Ana could not have got hold of
the camera if she had not reached the jeep – but the French translator,
focussing on the process of the progress towards the jeep, chooses the
more specific (and idiomatic) proposition vers. The English texts differ
markedly in the extent to which they encode the mover’s actual reaching the landmark (there are 24 tokens of up to) and mere progress in its
direction (there are only 5 tokens of towards).
Both sets of translations contain tokens in which the path is coded
by both verb and adverbial, as in (13), or verb alone, as in (14).
(13) a.I want you to go to La Coste, Ramon, and find De Sade.
b.Ramon, je veux que vous alliez à La Coste à la recherche de
Sade (NF1TF1)
(14) a.Then Dagny and the boys arrived. (HW2TE)
b.Ils arrivèrent alors, Dagny et les garçons. (HW2TF)
There are 160 constructions in English and 259 in French that resemble
(13) in containing a double coding of path.9 Moreover, the construction
in (14) containing a single coding of path in the verb is much more
common in French, with 113 tokens, compared to just 19 in English. A
further difference worth noting is that 20 of these French tokens contain a purpose adverbial, the French translator substituting the aim of
the mover in seeking out some goal for the actual goal itself. This sort
of usage may be seen in (15).
(15) a.He returned to Mother Karen. (HW2TE)
b.Il retourna voir Mère Karen. (HW2TF)
As for manner, underlined in the next three examples, this may also
be coded by the verb, as in (16), in an adverbial, as in (17), or in both,
as in (18).
9 In (13) the verbs go and aller are categorised as path verbs since they are used deictically, to encode motion away from the speaker. More often they are categorised as
neutral movement verbs, as are come and venir.
Motion to and Motion through: Evidence from a Multilingual Corpus 295
(16) a.Then she would be running past the horses towards me…
b.Bientôt elle trottinera vers moi entre les croupes des chevaux
… (NF1TF)
(17) a.I’d be late if I went all the way to Majorstua on foot. (KF1TE)
b.S’il fallait que j’aille à pied jusqu’à Majorstuen, j’arriverais
trop tard. (KF1TF)
(18) a.It leaped like a shaggy little animal from person to person.
b.Il sautait de l’un à l’autre comme un petit animal velu.
Double coding of manner as in (18) is much less common in both languages than double coding of path. There are just 16 tokens in English
and six in French. Nor is manner coding by an adverbial alone, as in
(17), frequent in either language, with 17 tokens in English and 27 in
French. More common is coding of manner by the verb alone, as in
(16), with 158 tokens in English and 64 in French.
500 450 400 350 300 250 200 150 100 50 0 Path Manner Just Just Verb + Just Just Verb + Verb Adv. Adv. Verb Adv. Adv. English French Figure 2. Coding of manner and path in English and French translations of
Norwegian ‘to-ness’ predications.
296 From Clerks to Corpora
120 100 80 60 Path 40 Manner 20 0 Just verb Just Adv. Verb + Adv. Just verb Just Adv. Verb + Adv. English French Figure 3. Coding of manner and path in English and French translations of
Norwegian ‘through-ness’ predications.
5. Codings of ‘through-ness’ and ‘to-ness’ compared
This section contains a brief comparison of the English and French
codings of ‘to-ness’ and ‘through-ness’ discussed in sections 3 and 4. In
order to better facilitate the comparison of the two types of predication, the data for ‘through-ness’ in Figure 1 are reproduced in Figure 3,
with the tokens coding site in an adverbial omitted.
Even a cursory glance at Figures 2 and 3 will suffice to reveal that the
codings of the two different sorts of motion predication, one containing a path to a target, the other a path through some sort of container,
resemble one another closely in both languages with respect to the coding of manner. As for the coding of path, there is a significant difference
between the two sorts of predications in both languages. This difference is related to a greater tendency to code path twice, or to split the
denotation of the path between verb and adverbial in predications of
‘to-ness’ compared to predications of ‘through-ness’. This is related to
the fact that predications of ‘through-ness’ just evoke a medial portion
of the path, the ‘route’, whereas predications of ‘to-ness’ presuppose
both a route and an end-point or goal. This double coding of path [to]
is illustrated in (19) and (20) and may be compared to the single coding
of path [through] in (21).
(19) a.But he did not return to the dressing room. (HW2TE)
b.Mais il ne retourna pas dans le cabinet. (HW2TF)
Motion to and Motion through: Evidence from a Multilingual Corpus 297
(20) a.“Won’t you come down to this bad billygoat of a man?”
b.“Ne peux-tu pas descendre vers ce terrible bouc de mari?”
(21) a.I suggested I could walk with him through the Retiro Park.
b.Je proposai de l’accompagner à travers le Retiro. (JG3TF)
In (19) the verbs return and retourna code a path back to a previous
location and the two prepositional phrases the end point of this path
(dans here codes a path [into] rather than a site [in]). In (20a) the verb
come codes path of motion towards the speaker, the particle down a
horizontally descending path and the to phrase the end point of this
path. In the corresponding French sentence it is the verb that codes the
descending path, while the prepositional phrase codes the direction of
the path towards its end point, rather than the end point itself. (21),
on the other hand, contains only one coding of path, referring to the
medial portion, with neither the starting nor end point being specified.
Another striking difference between the tokens coding ‘to-ness’
and those coding ‘through-ness’ is the number of verbs coding neutral
movement, rather than either manner or path. Such verbs include travel
and voyager and non-deictic come, go, aller and venir (the deictic readings of these four verbs code path of motion in the direction of, or away
from, the focused participant). There are 265 neutral movement verbs
in the English translations (43%) and 166 in the French translations
(27%) of til compared to just 20 (17%) for English and 14 (12%) for
French in the translations of gjennom.
In English, manner is far more likely to be coded in a verb than
an adverbial and path more likely to be coded in an adverbial than a
verb. Moreover, when path is coded in a verb, it is likely to be coded
in an adverbial as well, especially in predications of ‘to-ness’. English
thus conforms largely to the prototype of satellite-framed languages.
French is less likely than English to code manner, but if it does do so,
it resembles English in so far as it is more likely to code it in the verb
than an adverbial. Thus while French does conform to some extent to
the prototype of the path-framed language by coding path in the verb,
it diverges from it both in preferring to code manner, if at all, in the verb
and in coding path in an adverbial in addition to the verb, especially in
the case of ‘to-ness’.
298 From Clerks to Corpora
6. Summary and conclusions
In this chapter I compared codings of self-motion predications in which
the path either traverses a container landmark (‘through-ness’) or leads
up to a goal landmark (‘to-ness’). The reason for choosing these two
paths is that the former (potentially) involves the crossing of a boundary,
while the latter does not do so. Given that French is commonly taken
to avoid coding manner in the verb in boundary-crossing predications
(see, for instance, Aske 1989, Cappelle 2012), one might have expected
fewer such verbs in the translations of ‘through-ness’ ­predications. As
we have seen in section 5, this is not in fact the case, there being no
significant difference in the encoding of manner in the verb in French
renderings of the two sorts of predication.
The data for the study were taken from the Oslo Multilingual
Corpus and consist of codings in English and French of the concepts of
‘through-ness’ and ‘to-ness’ as these are instantiated in translations of
the same Norwegian tokens all containing adverbials with the prepositions gjennom or til. With respect to predications of ‘through-ness’ the
two languages differ in the frequency with which they employ the most
frequent preposition in path adverbials. While through occurs in 83%
of the English translations, à travers is only used in 20% of the French
ones. In fact French translators often prefer to encode the landmark as
the site within which an act of motion takes place, without specifying
the nature of the path followed by the mover. Thus 15% of tokens contain the preposition dans used to encode a site (there are three tokens in
which dans encodes a path, with the sense of English into rather than
in). The English translators code manner in the verb and path in an
adverbial in over half of all tokens (55%), whereas the French translators do so just 14% of the time. More common in French is double
coding of path, with the verb coding the general direction of movement
and the adverbial specifying this in greater detail. Also more common
in French than in English are tokens in which the path is denoted by a
verb such as traverser with the ground denoted by a direct object.
The translations of ‘to-ness’ predications differ from those of
‘through-ness’ in that a majority of tokens in both languages code
path in an adverbial in the form of a preposition phrase containing the
default prepositions to and à. There are more such tokens in English
than in French, the difference being due to some extent to a greater
tendency for French to code path in the verb, but also to a tendency
on the part of the French translators to specify the extent to which the
landmark has actually been reached (jusqu’à), or whether the mover is
Motion to and Motion through: Evidence from a Multilingual Corpus 299
still in the process of approaching the target (vers). Both languages contain a large number of constructions with a double coding of path, with
French again outnumbering English. On the other hand while there are
also a large number of tokens in French containing a single coding of
path in the verb, this sort of coding is comparatively rare in English.
Finally one may note that French translators occasionally use a purpose
adverbial in place of a path one, substituting the aim of the mover in
seeking out some goal for the actual goal itself.
If we compare the two types of motion predications to one another,
we see that there is no significant difference between codings of ‘throughness’ and ‘to-ness’ in either language with respect to the coding of manner. On the other hand, there is a significant difference between the two
types of predications in both languages in the coding of path. Another
striking difference is the number of verbs coding neutral movement,
which are much more common in translations of ‘to-ness’ predications
than ‘through-ness’ predications. This difference reflects a difference in
the original Norwegian tokens. It appears that in goal-directed predications, both the original authors and the translators focussed more
narrowly on the target landmark than on the manner of the mover’s
reaching it or the path along which the mover travelled. Having said
that, there are far more Norwegian neutral movement verbs rendered
by path verbs in French than there are in English.
Turning to the question of satellite- and path-framing, we have seen
that in the case of both types of path predication English seems to conform largely to the satellite framed prototype, while the picture for
French is more blurred. The English texts overwhelmingly code manner
in the verb. Equally clear-cut is the tendency for path in English to be
coded in an adverbial, rather than the verb. As for French the number
of tokens coding path in the verb testify to it being path-framed to a
greater extent than English, but there is a sizable minority of tokens
in which manner is coded in the verb, far more than one would have
expected had French been a pure verb-framed language. Rather, French
appears to be predominantly verb-framed, but with a number of alternative possibilities for coding manner and path, as has been pointed out
by Kopecka (2006) and Hickmann et al. (2009).
(1)Vi begynte å gå langsomt gjennom galleriene, og opp i andre
etasje. (JG3).
300 From Clerks to Corpora
Jeg foreslo at jeg kunne følge ham gjennom Retiro-parken.
(3) Dina begynte å gå gjennom stuene denne sommeren. (HW2)
(4) Når han kom gående gjennom byen … (BHH1)
(5) Og så klatre ut gjennom det åpne vinduet. (HW2)
(6)Et år kom ei ærfuglmor seg inn i eldhuset gjennom den åpne
døra … (HW2)
Siden smøg jeg meg inn gjennom døren på Grand Kafé ...
(8) I bare mamelukkene sprang hun gjennom stuene ... (HW2)
(9) Jeg sprang til vinduet … (BHH1)
(10) Siden fer de til kjerka! kommenterte Dina. (HW2)
(11) Og spaserte rolig over golvet og bort til vinduet! (HW2)
(12)Ana styrtet til jeepen og kom tilbake med et lite videokamera …
(13)Jeg vil at de skal reise til La Coste, Ramon, og finne de Sade.
(14) Så kom de til, både Dagny og guttene. (HW2)
(15) Han gikk til Mor Karen enda en gang. (HW2)
(16) Snart småløper hun mellom hesterompene bort til meg. (NF1)
Jeg ville komme for sent om jeg skulle ta meg helt ned til
Majorstuen til fots. (KF1)
(18)Det hoppet som et lite loddent dyr, fra menneske til menneske.
(19)Kan du ikke kom ned til dennan fryktelige bukken av en mann?
(20)Men han gikk ikke til påkledningsværelset. (HW2)
(21)Jeg foreslo at jeg kunne følge ham gjennom Retiro-parken. (JG3)
Dictionnaire de l’Académie, huitième edition. http://artfl-project.uchicago.
Dictionnaire de l’Académie, neuvième édition.
Knudsen, T., Sommerfelt A. & Noreng H. (1983). Norsk Riksmålsordbok:
Bind lV. Oslo: Kunnskapsforlaget.
Oslo Multilingual Corpus:
Motion to and Motion through: Evidence from a Multilingual Corpus 301
Oxford English Dictionary: OED v4.0 on CD-Rom. Oxford: Oxford
University Press.
Vikør, L. S. (ed.). 2002. Norsk ordbok, band IV [vol. 4]. Oslo: Det Norske
Aske, J. (1989). Path predications in English and Spanish: a closer look.
K. Hall, M. Meacham & R. Shapiro (eds) Proceedings of the fifteenth annual meeting of the Berkeley Linguistics Society. Berkeley: Berkeley Linguistics
Society, 1–14.
Aurnague, M. (2000). Entrer par la petite porte, passer par des chemins de
traverse: à propos de la preposition par et la notion de “trajet”. Carnets de
grammaire. Rapport no. 7.
Berman, R. A., & Slobin, D. I. (1994). Relating events in narrative: A crosslinguistic developmental study. Hillsdale, NJ: Lawrence Erlbaum Associates.
Cappelle, B. (2012). English is less rich in manner-of-motion verbs when translated from French. Across Languages and Cultures, 13:2, 173–195.
Egan, T. (2012). Through seen through the looking glass of translation equivalence: a proposed method for determining closeness of word senses.
S. Hoffman, P. Rayson & G. Leech (eds) Corpus linguistics: Looking back
– moving forward. Amsterdam: Rodopi, 41–56.
———. (2013). Tertia Comparationis in Multilingual Corpora. K. Aijmer & B.
Altenberg (eds) Advances in corpus-based contrastive linguistics. Studies in
honour of Stig Johansson. Amsterdam: John Benjamins, 7–24 .
Egan, T. & Rawoens G. (2013). “Moving over in(to) English and French”.
Languages in Contrast 13:2, 193–211.
Filipović, L. & Jaszczolt K. M. (eds) (2012). Space and Time in Languages and
Cultures: Linguistic Diversity. Amsterdam: John Benjamins.
Hickmann, M., & Robert S. (eds) (2006). Space in Languages: Linguistic
Systems and Cognitive Categories. Amsterdam: John Benjamins.
Hickmann, M., Taranne P. & Bonnet P. (2009). Motion in first language acquisition: Manner and Path in French and English child language. Journal
of Child Language 36, 705–741.
Jaszczolt, K. M. (2003). On translating what is said: tertium comparationis in
contrastive semantics and pragmatics. K. M. Jaszczolt & K. Turner (eds)
Meaning Through Language Contrast. Vol. 2. Amsterdam: John Benjamins,
302 From Clerks to Corpora
Johansson, S. (2007). Seeing through Multilingual Corpora : On the use of
corpora in contrastive studies. Amsterdam: John Benjamins.
Kopecka, A. (2006). The semantic structure of motion verbs in French:
Typological perspectives. M. Hickmann & S. Robert (eds) (2006), 59–81.
Krzeszowski, T.P. (1990). Contrasting languages: the scope of contrastive linguistics. Berlin: Mouton de Gruyter.
Lee, D. (2001). Cognitive Linguistics: an Introduction. Oxford: Oxford
University Press.
Leech, G.N. (1969). Towards a semantic description of English. London:
Lindstromberg, S. (1998). English Prepositions Explained. Amsterdam: John
Pourcel, S. & Kopecka A. (2005). Motion expressions in French: Typological
­diversity. Durham and Newcastle Working Papers in Linguistics 11,
Tyler, A. & Evans. V. (2007). The Semantics of English Prepositions: Spatial
scenes, embodied meaning and cognition. Digitally printed (with corr.) ed.
Cambridge: Cambridge University Press.
Slobin, D. (2006). What makes manner of motion salient? Explorations in linguistic typology, discourse and cognition. M. Hickmann & S. Robert (eds)
(2006), 83–101.
Talmy, L. (1991). Path to realization: A typology of event conflation. L.A.
Sutton, C. Johnson & R. Shields (eds) Proceedings of the 17th Annual
Meeting of the Berkeley Linguistics Society. University of California at
Berkeley, CA: Berkeley Linguistics Society, Inc. 480–519
Talmy, L. (2000). Towards a Cognitive Semantics, Volume II: Typology and
Process in Concept Structuring. Cambridge MA: The MIT Press.
Viberg, Å. (1998). Contrasts in polysemy and differentiation: Running and
putting in English and Swedish. S. Johansson & S. Oksefjell (eds) Corpora
and Cross-linguistic Research. Amsterdam: Rodopi, 343–376.
———. (2003). The polysemy of the Swedish verb komma ‘come’: A view from
translation corpora. K.M. Jaszczolt & K. Turner (eds) Meaning through
language contrast. Vol. 2. Amsterdam: John Benjamins, 75–105.
———. (2013). Seeing the lexical profile of Swedish through multilingual corpora. The case of Swedish åka and other vehicle verbs. K. Aijmer & B.
Altenberg (eds) Advances in corpus-based contrastive linguistics. Studies in
honour of Stig Johansson. Amsterdam: John Benjamins, 25–56.
17 Using the World Wide Web to Research
Spoken Varieties of English: The Case of
Pulmonic Ingressive Speech
Peter Sundkvist
Stockholm University
1. Introduction
Undoubtedly, one of the most significant developments for human
interaction and communication within the last thirty years or so is the
internet. Its origins may be traced back to various experimental computer networks of the 1960s (Crystal 2013: 3). It has since grown by a
phenomenal rate: between 2000 and 2012 the number of internet users
increased by an average of 566.4% across the globe, and as of 2012
34.3% of the world’s population was estimated to have access to the
internet (Internet World Stats 2012). Partly as a result of its phenomenal growth the internet has also established itself as an important tool
and source of information within the fields of linguistics and English
studies. While much linguistically-oriented work has focused on features and developments related to the electronic medium of communication itself, inquiry has gradually extended into further domains,
as researchers become more creative in exploring the full potential of
the internet. In particular, the World Wide Web (WWW), defined as
“the full collection of all the computers linked to the internet which
hold documents that are mutually accessible through the use of a
standard protocol (Hypertext transfer protocol, or HTTP)” (Crystal
2013: 13), has opened up a range of new possibilities with regard to
linguistic research. Thus far, however, less attention has been devoted
to its potential for research on spoken varieties of English. The aim of
this paper, therefore, is to explore the World Wide Web as a source of
evidence for a specific feature of spoken language. Drawing upon the
fact that very large amounts of audio and video material have become
available, through a diverse range of online resources, the aim is to
How to cite this book chapter:
Sundkvist, P. 2015. Using the World Wide Web to Research Spoken Varieties of English: The
Case of Pulmonic Ingressive Speech. In: Shaw, P., Erman, B., Melchers, G. and Sundkvist,
P. (eds) From Clerks to Corpora: essays on the English language yesterday and today. Pp.
303–321. Stockholm: Stockholm University Press. DOI:
License: CC-BY.
304 From Clerks to Corpora
gather evidence for and provide documentation of a somewhat elusive
paralinguistic feature within a set of regional Englishes, which has thus
far proved challenging to study by means of more standard methods.
2. Pulmonic ingressive speech
Visitors to the Nordic countries are often struck by a curious feature
in the speech of the locals. In conversation, Norwegians, Swedes and
Danes can frequently be heard drawing their breath inwards, in a
manner vaguely reminiscent of a sudden gasp. Although often erroneously interpreted by the uninitiated listener as a sign of sudden ill
health or astonishment, this is in fact a well-established paralinguistic
feature. The technical term for this phenomenon is ‘pulmonic ingressive speech’. Most commonly of course, human speech occurs on an
outward airstream, directed from the lungs towards the mouth. The
phonetic term for this is pulmonic (‘involving lungs’) egressive (‘outwards’) speech. However, it is also perfectly possible to speak on a
reverse, inward (‘ingressive’) airstream, by drawing air into the lungs
while speaking. In the Nordic languages, this airstream mechanism
is commonly used on short words for ‘yes’ and ‘no’, especially when
provided as backchannel items in a conversation, when the listener
wants to signal to the speaker that he or she is following what is
being said. While the many anecdotes of befuddlement on part of
newcomers may seem to imply that ingressive speech is restricted to
the Nordic languages, it is becoming clear that the phenomenon in
fact occurs in a much wider range of languages and language varieties, including several regional Englishes.
At present there is no consensus regarding the cross-linguistic distribution of pulmonic ingressive speech. According to one view it
may be found in a wide range of languages, on all continents of the
world. Based on an extensive review of various sources, Eklund (2007,
2008) suggests that it has simply gone unnoticed for many parts of
the world, and that insufficient effort has been made to compile and
compare reports from across the globe. In typological terms, pulmonic
ingressive speech is thus seen as constituting a ‘neglected universal’
(Eklund 2007). However, an opposing view suggests that ingressive
speech is geographically restricted. Clarke and Melchers (2005) argue
that ingressive speech occurs primarily within a region stretching in
the east from the Baltic countries (e.g. Estonia) and westwards over
Scandinavia (peninsular as well as insular), northern parts of continental Europe (Germany, Austria), across some regions of the British
Using the World Wide Web to Research Spoken Varieties of English 305
Isles, and finally into Maritime Canada and coastal New England. This
pattern finds a ready historical explanation. Pulmonic ingressive speech,
it is suggested, was first spread by Vikings from Scandinavia westwards
to the British Isles and eastwards to the Baltic countries. Subsequently
it was further transmitted across the North Atlantic by Irish and British
migrants (Clarke & Melchers 2005; Shorrocks 2003). From this perspective, pulmonic ingressive speech is considered to be ‘typologically
highly marked’ (Clarke & Melchers 2005).
While at first sight ingressive speech may merely seem as a quaint
curiosity, it soon becomes clear that further investigation into the
phenomenon is significant for several reasons. From a general viewpoint it is a feature whose typological status remains to be established.
From the present perspective of English studies, its precise distribution
among regional varieties of English remains an open question, in spite
of the fact that it is reported in several areas. Further insight into the
use of ingressive speech among such forms of English may furthermore
reveal historical connections among relevant regions. What is more,
pulmonic ingressive speech is commonly surrounded by popular mythology. For instance, where it occurs, locals often seem to believe that
it is somewhat unique to them and their communities. This, as we will
see, appears to be the case for some regional Englishes also.
3. Pulmonic ingressive speech in regional Englishes:
A test case for the World Wide Web?
One problem which has stalled the study into ingressives is the lack
of objective evidence. For many regions where it reportedly occurs
there is very little documentation in such forms as audio or video
recordings, and the sole evidence consists of informal observation.
The absence of data is in many cases not attributable to a lack of
trying but to various inherent challenges for data collection. While
the most obvious method may seem to be to collect ingressives by
recording interviews, such attempts have had rather mixed success. In
many cases, the interview situation itself seems to militate against the
occurrence of ingressives. It has been argued that pulmonic ingressive speech tends to be used primarily in relaxed, informal situations,
and that a level of interpersonal affiliation between the speaker and
listener is required (Clarke & Melchers 2005). Owing to such challenges facing anyone setting out to record ingressives through interviews, pre-existing data sets and corpora would seem like a suitable
alternative, perhaps especially for such localities where the inquiry
306 From Clerks to Corpora
into ingressives is at an early stage and simply objectively supported
instances would constitute an important step forward. Even for this
approach, however, previous researchers report various difficulties,
as ingressives seem poorly represented in certain types of existing
recordings (cf. e.g. Thom 2005).
A major source of pre-existing recordings is of course audio and
visual media. However, it has been suggested that the media may perhaps not be a very good resource for pulmonic ingressive speech. Thom
remarks that ingressive speech is used especially by such sections of
society which are not well represented in the media – which within
this context means primarily elderly, rural speakers – and therefore is
bound to be underrepresented in media-based corpora: “[…] it was
difficult to find examples of [ingressive speech] in recorded material,
either on radio or on television. This was because [ingressive speech]
tends to be used in informal situations, and (as I discovered) mostly
by groups of people who are rarely seen in the media.” (Thom 2005:
28–9). In Swedish (and Finnish), the word jo/ju (‘yes’) is sometimes pronounced as a voiceless bilabial fricative or approximant, on a pulmonic
ingressive airstream. The frequency of this feature appears to increase
the further north in Sweden you go (Eklund 2008: 262–263), and it is
popularly associated with northern parts. With regard to media, however, Hanell and Salö (2009: 19) claim that “You can actually spend
hours among video clips on YouTube where [well-known people from
the northern part of Sweden], such as Ingemar Stenmark and Maud
Olofsson, go through interview after interview without making a single
[jo/ju, bilabial fricative]” (author’s translation).
In face of the difficulties in documenting ingressives through more
standard methods, it is natural to consider what potential the World
Wide Web may offer. Owing to its massive expansion, a wide range
of audio and visual material is now available online. This includes
not only news broadcasts, documentaries, entertainment etc. but also
unedited material posted at sites such as YouTube. The aim of this paper
is therefore to provide documentation of pulmonic ingressive speech in
regional Englishes, and within the North Atlantic region more generally, based on material accessed through the World Wide Web. As far as
has been possible, this is based on publicly available sources, for which
URLs are provided; in some instances research data has been utilized
as an additional source of information. Secondly, the aim is also to find
illustrations of the popular mythology surrounding ingressives on the
basis of material available through the WWW.
Using the World Wide Web to Research Spoken Varieties of English 307
4. Documentation of pulmonic ingressive speech:
Audio and video
The following section provides illustration of the use of pulmonic
ingressive speech based on audio and visual data. The focus is on
Englishes across the North Atlantic region for which relevant data has
thus far been acquired. In addition, examples are offered from certain
Scandinavian languages, as well as one speaker from western Canada.
Brief extracts of conversations are transcribed and presented; in these,
words spoken ingressively are underlined and in italics. Where available, URLs are stated, as well as the relevant time segment in the files.
4.1 Ireland
Illustration of Irish ingressives may be found in the speech of Paddy
Malone, the piper of the Irish band The Chieftains, who hails from
Dublin. Instances may for instance be heard in the brief documentary
‘San Patricio – Behind the scenes’, which portrays the making of the San
Patricio album (2010), featuring the Chieftains and American guitarist
Ry Cooder. The extract below comes from a conversation between Paddy
Malone (PM) and Ry Cooder (RC), concerning the San Patricio ­battalion –
a group of Irish volunteer soldiers who deserted the US army and fought
on the Mexican side in the Mexican-American war (1846–1848).
(1) RC: It’s a beautiful little song and the lyrics (…) are, if you flip it
around, seems to me that (…) the same could be said for these
poor Irish soldiers, [who] are so far from home
PM: That’s right
RC: Probably never going back
PM: Yeah (ingressive)
RC: Well, [needless] to say
PM: [Yeah]
PM: Yeah (ingressive)
RC: So for them it’s also true
RC: Yeah (ingressive), very much so
RC: You know, Yeah. The diasporic person [is] never gonna get back
PM: [Mmm] [Yeah
Location in link: 7:22–7:41
308 From Clerks to Corpora
4.2 England
It seems to be particularly difficult to find recorded instances of ingressives from England. However, instances occur in the audio recordings
made by the Survey of English Dialects in Patterdale, Westmoreland,
Northern England, in 1974. The recording consists of a conversation on
the topic of gardening, work etc., between the fieldworker, linguist Clive
Upton (CU), and the interviewee Edward Blamire (EB), born in 1903.
(2) CU: You do all sorts of other work, now, around the estate
EB: Well, aye. I mow, I mow the la-, mow the lawns. Do fencing.
And uh
CU: [inaudible]
EB: Anything what, anything what I can do
CU: Hmm
EB: Aye, Aye (ingressive)
CU: Yeah, Hmhm
EB: That’s right
CU: Yeah
Location in link: 4:10–4:30
4.3 Shetland
The Shetland Isles, situated in the North Sea, constitute the northernmost
part of the UK. The presence of ingressives in Shetland has been documented more extensively (Sundkvist 2012a, 2012b). The extract below is
from a conversation between the author (PS) and a Shetland man (LM)
in his 80s, from Shetland’s main town of Lerwick. The interview is part
of a corpus collected by the author 2000–2003 for a study of Lerwick
pronunciation, and the discussion concerns the microphone stand.
(3) LM: Yeah, you can’t take the nut far enough, that’s the thing, yeah
PS: No. And then when you keep
LM: You put a washer in, or something like that with
PS: Yeah, I can put that but I [think] when I come home I need to take it
apart and [do]
LM: [Yeah]
PS: something with it
LM: Yeah (ingressive)
Link: Audio file not made available.
Using the World Wide Web to Research Spoken Varieties of English 309
4.4 Orkney
The Orkney Isles are located just off the north coast of Scotland. The
excerpt below is taken from data collected by Gunnel Melchers and
Arne Kjell Foldvik between 1980 and 1985. The conversation involves
one of the fieldworkers (Arne Kjell Foldvik, AKF) and a farmer from
the isle of Westray (WM), whose age is estimated to be between 45 and
50 by Foldvik. The conversation takes place outdoors, and reflects a
chance encounter with the local.
(4) AKF: But you got to have them inside all winter
WM: [Oh, yes]
AKF: [pass] the winter
WM: Oh yes, they come in in November
AKF: Umhm
WM: and they’re in 'til the first of May
AKF: Silage? Is that […]
WM: [yes, mostly] silage, yes, that’s right, aye (ingressive)
Link: Audio file not made available.
4.5 Fair Isle
Fair Isle is a small island in the North Sea, situated halfway between
the Orkney and Shetland Isles. Although it is officially part of Shetland,
Fair Islanders display a strong local identity, often expressed as being
distinct from that of the rest of Shetland. The presence of ingressives
within the Fair Isle community was established in Sundkvist 2012a,
b. The excerpt below is from an animated discussion between one
local woman (FIF) and two local men (FIM1, FIM2). It was recorded
in 1982 by Gunnel Melchers and Arne Kjell Foldvik. For simplicity,
word forms and spelling have been adjusted to Standard English in
the extract.
(5) FIM1: But someone, for the funeral, someone, someone who
belonged to the body [=person] who was dead, they had to
get around to every house, like
FIM2: Invited every house [to the funeral]
FIM1: [(…) invite to the] funeral
FIF: Yeah
FIM1: Yeah (Ingressive)
FIM2: Yeah
Link: Audio file not made available.
310 From Clerks to Corpora
4.6 Cape Breton
Turning the focus across the North Atlantic, instances of ingressives may
also be found from North America. The example below is from a documentary about the Irish heritage of Cape Breton, Nova Scotia, Canada:
‘The Irish in Cape Breton 1713–1990’ (The Irish In Cape Breton 3 of
6, 1990, produced, directed and narrated by Kenneth Donovan). The
interviewer is Kenneth Donovan (KD), a Parks Canada historian; the
interviewee is Loretta Donovan (LD), an elderly lady.
(6) KD: And how many uh ... children were in your family, in your
father and mother’s family?
LD: Nine, we had six boys and three girls
KD: Six boys [and] three girls?
LD: [Yes]
LD: And I’m the only girl left, but I have two brothers in the States
KD: Right
LD: My older brother
KD: Yeah
LD: Not my older brother [name], Yeah (ingressive), and one is 89
and the other is 84
Location in link: 07.25–07:42
4.7 Newfoundland
As to Maritime Canada, examples of ingressives may be found for
instance in the online archive of Memorial University of Newfoundland.
The archives reveal the presence of ingressive speech in several locations
within Newfoundland. The example presented below is from Lower
Cove. The conversation took place between interviewer Lisa Wilson
(LW) and interviewee Mamiellen Noseworthy (MN) on June 7, 2010,
and concerns local craft.
(7) LW: Doreen just spent 'bout half an hour showing me, just a sample,
how to do a pleat
MN: Yeah
LW: And, she’s going quite fast [laughter], and there’s no way I could
do that without having to watch, a lot
MN: Right
LW: And then practice a lot
MN: Yeah
Using the World Wide Web to Research Spoken Varieties of English 311
LW: So, observe, observe, observe, and then practice, practice,
MN: Yeah (ingressive), yeah
Source: Noseworthy, Mamiellen. Lisa Wilson interviewing Mamiellen
Noseworthy, June 7, 2010, Lower Cove, Newfoundland. Archive ID:
Location in link: 4:30–4:55
4.8 British Columbia
While previous sources have pointed towards the existence of ingressive
speech in Maritime Canada, few have mentioned the possibility for the
western parts of the country. The token below is from a young female from
British Columbia. It occurs in an episode of the crime show ‘48 hours’
titled ‘Highway of Tears’, which deals with a series of disappearances and
homicides along a part of the Trans-Canada highway system supposedly
known by some locals as the ‘highway of tears’. However, only one ingressive token occurs in the material, on reported speech within a conversation between the reporter (R) and the female (Jordanne Bolduc, JB).
(8) R: What did she say to you?
JB: S he was just like shocked, she’s like: really? You’re going? And I
was like: yeah (ingressive), I’m going
JB: And [pause] she kinda begged me, and then I was just well you
can come with us and she said ‘no’, She wanted to stay there with
her tent, for it to be safe
Location in link: 13:30–13:45
4.9 The Nordic languages
The existence of ingressive speech is well documented for Swedish,
Norwegian, and Danish. However, fewer samples appear to be easily available for Insular Scandinavian, such as Icelandic and Faroese.
The following section thus presents examples concerning Icelandic and
Faroese, as well as the Swedish word jo/ju, commented on earlier.
4.9.1 Icelandic
Icelandic ingressives are amply illustrated in the TV show Silfur Egils.
In the excerpt below, the host Egill Helgason is talking to a book keeper,
312 From Clerks to Corpora
Bragi Kristjónsson (aka Bragi bóksali). The conversation concerns a
photographic book.
(9) BK: These are his parents.
EH: Yes.
BK: Haraldur Guðmundsson, master carpenter, and the mother.
EH: Yes (ingressive).
BK: And his brother Leifur who wrote the poem there “The young
poets write verses…”
EH: “…without being able to”.
BK: “…without being able to/ In the Public house1 I get daily meals/
without eating them”.
EH: Yes, yes, yes (ingressive) Yes-yes.
BK: But this brother of his, Magnús Haraldsson
EH: Yes (ingressive).
BK: he died in a car accident.
Location in link: 1:11–1:41
4.9.2 Faroese
Faroese ingressives may similarly be illustrated with material from a
talk show. The interviewer below is journalist and show host Høgni í
Jákupsstovu (speaking with a Tórshavn dialect); the interviewee is artist
Tróndur Patursson (with a Kirkjubø dialect).
(10) HJ: Yes, welcome Tróndur Patursson.
TP: Thank you.
HJ: 4
0 years since you first were at the art museum (= Listasavn
Føroya, National Gallery).
TP: Yes, yes (ingressive).
HJ: You do it like the queen, [hold] a 40 year anniversary?
TP: Well, yes, yes-yes. I suppose it is. I suppose it is. Yes (ingressive)
ow was it to be back at the art museum, after that conflict
[that you were a part of]?
TP: Yes, yes, the art museum is a truly outstanding building (.)
Location in link: 3:05–3:30
1 The word Alþýðuhúsið literally means ‘the house of the public or the public house’.
However this is not the same as a public house as abbreviated into ‘pub’ but rather
a type of communal guesthouse (Friðrik Sólnes, p.c.).
Using the World Wide Web to Research Spoken Varieties of English 313
4.9.3 Swedish ‘jo’/‘ju’
As pointed out previously, the word jo/ju (‘yes’) in Swedish may be
spoken on a pulmonic ingressive airstream; in such instances it may
be generally characterized as a voiceless bilabial fricative or approximant. This pronunciation is often associated with the northern parts
of Sweden, and one individual commonly mentioned with regard to it
is former alpine skier Ingemar Stenmark, whose laconic conversational
style was popularly thought to fit well with the brief reply of a jo/ju.
The discussion below occurs in an episode of the TV show ‘Mästarnas
mästare’, whose participants are former top athletes. It concerns an
annoying interview with a journalist who prematurely made the suggestion that Stenmark’s career had come to an end. Stenmark (IS) is
talking with Peter (“Pekka”) Lindmark (PL), a former ice hockey goal
(11) PL: No, but really, it must be horrible to get a [expression meaning
‘we thank you for the time you have been a skier’] and then
you go on winning another 20
IS: Yeah
PL: Oh, [it] is like kicking the legs out from under you
IS: I really felt like, it would have been better to quit long, long
before, before I had, just when I had reached the top
PL: Yes, because then you were, exactly, then you would have
become immortal, but
IS: [Yes, then I would have]
PL: [But you] really were outstanding, for very many years
IS: Yeah
Figure 1. Documentations of pulmonic ingressive speech in section 4.
314 From Clerks to Corpora
PL: And won and won and won
IS: Yeah (Ingressive; bilabial fricative)
Location in link: 8:57–9:21
4.10 Summary: Audio and video documentation within the North
Atlantic region
The range of localities for which evidence of pulmonic ingressive speech,
based on audio and visual material, has been presented in section 4 is
summarized in Figure 1 below. A (red) dot in the map signifies that an
illustration has been provided for the locality in question. Regarding
access to relevant sound files, URLs are available and provided for
Ireland, England, Cape Breton, Newfoundland, British Columbia; and
also Icelandic, Faroese, and Swedish. For Shetland, Orkney, and Fair
Isle no URLs are available. Further linguistic analysis of the ingressive
tokens presented, including their discourse function, falls outside the
scope of the present study. It should be noted, however, that all tokens
in section 4 constitute expressions for ‘yes’ provided as feedback during
a conversation.
5. Illustration of stereotypes, popular notions
Pulmonic ingressive speech is associated with a range of stereotypes
and popular mythology. A common notion in localities where it occurs
seems to be that it is unique to that particular area; general awareness
of its wider distribution seems fairly limited. This self-perceived exclusivity, or at least typicality, applies not only to Swedes and Norwegians
but also to several English-speaking localities. The following media
clips illustrate the conception that it is unique to or at least characteristic of a particular region, that it may be misunderstood by outsiders,
and that it consequently may need to be explained to them.
The part of North America which seems most commonly discussed
with regard to ingressive speech is coastal Maine. In particular, it tends
to be mentioned in conjunction with the word ayuh or ayup (‘yes’).
Similar anecdotes may be encountered for Maine as from Scandinavia,
in which newcomers are bewildered and confused by the phenomenon.
One such anecdote, illustrated in the link below, concerns a medical
doctor from Kentucky, who was working in Down East Maine, a part
of coastal north-eastern Maine which borders on Canada. The doctor
asks a local patient if he is OK, to which he answers ayuh (pronounced
Using the World Wide Web to Research Spoken Varieties of English 315
as a voiced ingressive). The doctor is puzzled by the patient’s apparent
gasp and becomes concerned about his health, not realizing that it in
fact represents a ‘yes’ to the question if he is OK.
A second case in point concerning coastal Maine is provided by
humourist Tim Sample. This comes from a live stage show, based
partly on an audio record: How to Talk Yankee: A Downeast Foreign
Language Record featuring Bob Bryan and Tim Sample. During the
show, Sample and Bryan discuss several local words and expressions,
including ‘ayuh’. As the conversation is fairly long, and digresses somewhat at several points, only relevant parts are extracted below. Sample
begins by suggesting to the audience that there are several ways of saying ‘ayuh’, including both egressive and ingressive variants. As to the
more exotic ingressive variant, Sample discusses its use, implying that it
forms part of a greeting, and provides explicit instruction and a drill to
the audience on how to perform it. He starts:
(12) So I’ll tell you one thing that will fool even a native occasionally.
You can learn this, you can learn this even if you’re driving a Volvo
(…) (…) This is the ‘Yankee reverse nod with inhalation’. This is
how native Yankees communicate with each other on the road.
Sample goes on to give an instruction:
(13) Try this: you inhale like that, and throw your head back at the
same time.
He finally illustrates its usage, as part of greeting a local:
(14) So you’re driving down the road wishing you was from here. You
see old Bert and his '67 Chevy pickup truck. (…). So, what you
wanna do, you see Bert over the hill. You [Sample inhales and
throws back his head, as instructed] like that. He does it too, see.
And for a moment there, you’re bonded.
Location in link: 0:00–5:00; record also available for purchase (see
list of references).
Reference to ingressive speech is also to be found in Maritime Canada.
Some instances indicate that it is believed to be specific to the region.
316 From Clerks to Corpora
As an example, a host at a local radio station on Prince Edward Island
made the following call for listeners to phone in their ‘islanderisms’:
(15) Now, on to tonight’s topic […]. I have a favour to ask of you. I
need you guys to call me in with your ‘Islanderisms’. What I mean
by that. ‘Islanderism’, definition = something that is specific to
our native tongue. Example: ‘slippy’, as opposed to ‘slippery’. The
sound of an inhalation, as opposed to ‘yes’, and what I mean by
that is: yeah (ingressive). You know what I mean. Stuff like that. I
want you guys to give me a call with your examples, and we can
see just how many islanderisms there are.
Location in link: 0:08–0:44
Turning to Norway, popular commentary on ingressive speech is fairly
frequent. In a self-sarcastic manner, ‘Song about integration’ (as featured in the comedy show ‘Sex og SingelSiv’ broadcast on VGTV) plays
on stereotypes of Norwegian culture and behaviour, with a focus on
aspects that may be unfamiliar and seem peculiar to immigrants. Among
the many stereotypes drawn upon is pulmonic ingressive speech:
(16) (…)
And if you talk to strangers on the bus you are a jerk
And when you say ‘yes’, so say it on ingressive speech
As mentioned, the Swedish word jo/ju (‘yes’) may be uttered as a voiceless bilabial fricative or approximant, on a pulmonic ingressive airstream. This feature appears to be more common in northern Sweden,
and is stereotypically associated with it (Eklund 2008: 262–263).
Popular commentary may be found online. One example is provided
by former ice hockey player Jonas Bergqvist, who was born in southern Sweden (Ängelholm, Scania) but moved to a more northerly part
(Leksand, Dalcarlia) to attend high school. He recalls that he adopted
ingressive jo/ju among other features when moving north, and that it
was overtly commented on when returning south, reflecting its distinct
regional marking:
(17) (…) and that’s how it was, when I moved there, then I had that
real Scanian dialect, and said: “one and two and three”[Scanian
accent] and “are you stupid?” and so on, and everyone said ‘what?’
all the time. So in some way it was only natural that I tried to make
Using the World Wide Web to Research Spoken Varieties of English 317
myself understood, and then it started becoming “one and two and
three.”[clear standard]. Then sometimes a jo/ju (voiceless ingressive
bilabial fricative) would enter [into my speech], and then they nearly
split their sides laughing when I returned home in the summer.
Location in link: 17:40–18:03
6. Conclusion
Although pulmonic ingressive speech may initially admittedly seem
like a fairly peripheral matter, upon further consideration the phenomenon is clearly significant in several ways. Its typological status
remains a topic of debate, with proposals ranging from ‘highly marked’
to ‘neglected universal’. Significantly for our purposes, its pattern of
distribution among regional Englishes has not been sufficiently established; although it has been claimed to occur in a range of regions, there
has been a glaring lack of objective evidence. What is more, it is a feature which may possibly reveal historical connections among Englishes
across the North Atlantic region. Unfortunately, it has proved challenging to elicit and record pulmonic ingressive speech in many localities. In
response, previous researchers have tried several alternatives to standard sociolinguistic methods, including observational methods (Peters
1981) and wide-scope searches of a broad range of written sources
(Eklund 2008). In this paper we explored a further alternative, namely
the World Wide Web and the many sources of information to which it
provides access. Claims were reviewed that ingressives may be hard to
study through the media, partly because relevant speaker groups are
not well represented therein (Thom 2005; Salö & Hanell 2009). While
this may be true to some extent, the massive expansion of the WWW
has opened up new possibilities in this regard and necessitates a re-examination of the potential of the media, broadly defined. In this paper
a collection of audio and video clips was presented; in addition excerpts
from research data sets were provided. This contributed documentation for the existence of pulmonic ingressive speech in an additional set
of Englishes. Despite the moderate total number of tokens, ingressive
speech was illustrated for both genders, in each case on a brief ‘yes’
response. Furthermore, illustrations could also be found of the many
popular notions surrounding ingressive speech.
Owing to the highly transient nature of the medium and technology
involved, it is difficult to predict the future range of applications and
318 From Clerks to Corpora
benefits of the WWW for the study of spoken language (cf. Crystal
2013: 257, 273). At this stage, however, several possibilities as well as
potential limitations are discernible. As to the many possibilities, while
spoken language so far clearly is less well represented than written language on the WWW, there is reason to believe that both the amount
and the proportion of video and sound material will increase as the
WWW continues to grow and technology develops (Crystal 2013: 9;
Ess & Consalvo 2011: 2). Since the expansion of the internet over the
last ten years has been the greatest in Africa, the Middle East and Latin
America, there is also reason to be hopeful that it may provide a source
of evidence for pulmonic ingressive speech in these parts of the world,
for which objective evidence is completely lacking. What is more, linguists are gradually beginning to realize the formidable, largely unexplored pedagogical potentials of the WWW with regard to spoken
language. Squires and Queen (2011: 27) outline some of the benefits
of media collections for the teaching of linguistics, which include the
ability to illustrate particular features of spoken language.
A number of limitations are however apparent. The time factor is
clearly significant. The set of clips presented in this paper was located
through a combination of idle web surfing and purposeful, directed
searches by the present author, sporadically as well as during more sustained intervals stretching over a total period of 3–4 years. Some were
also obtained through kind suggestions by others. It must therefore be
acknowledged that the compilation of even a relatively small collection
of material is time-consuming, and perhaps best approached as a longterm project. A second serious limitation is the transience of the medium.
URLs are notoriously short lived, and there is no guarantee that published
web links will even be alive at the time of publication (Crystal 2013: 257;
Squires & Queen 2011: 229). It is therefore of course preferable to save
material onto more permanent storage, which, however, leads into other
issues such as storage space and copyright (Crystal 2013: 273; Squires
& Queen 2011: 229). In addition, it is also possible that in the future
internet sites providing useful material to a greater extent will require
their users to pay. On the whole, however, echoing Crystal’s remark that
“(…) the sheer scale of the present Internet, let alone its future telecosmic
incarnations, has convinced me that we are on the brink of the biggest
language revolution ever” (2013: 274–275), there is good reason to be
optimistic that the World Wide Web will prove a useful tool for exploring a range of additional features of spoken English, reflecting the value
of the internet in ways far beyond those originally envisaged.
Using the World Wide Web to Research Spoken Varieties of English 319
Appendix: Excerpts in the original (Nordic languages)
Icelandic (example 9)
BK: Sko, þetta eru foreldrar hans.
EH: Já.
BK: Haraldur Guðmundsson trésmíðameistari og móðirin.
EH: Já (ingressive).
BK: Og Leifur bróðir hans sem orti þarna vísuna “Ungu skáldin yrkja
EH: “…án þess að geta það”.
BK: “…án þess að geta það/ í Alþýðuhúsinu er ég í fæði/ án þess að éta
EH: Já, já, já (ingressive), jájá.
BK: En sko þessi bróðir hans, Magnús Haraldsson.
EH: Já (ingressive).
BK: Hann dó í bílslysi.
Faroese (example 10)
Hógni: Ja, vælkomin, Tróndur Patursson
Tróndur: Takk fyri tað
Hógni: 40 ár siðani tú fyrsti ferð vart á Listasavninum.
Tróndur: Ja, ja (ingressive).
Hógni: Tað er [tú gjert?] sum drottningin, 40 árs jubileum.
Tróndur: N
á, ja, jaja, tað er nakað um tað. Tað er nakað um tað. Ja
Hógni: Tú, hvussu tað at vera aftur, aftaná hetta her stíði hjá tykkum?
Tróndur: Ja, ja, Listasavnið er, er ein heilt einastandandi bygningur. (…)
Swedish (example 11)
PL: Nej, men alltså, det måste vara fan ruggigt att få ett tack för den tid
som har varit ungefär och så vinner man 20 till
IS: Jo
PL: Åh, [det] är som att sparka undan bena
IS: [jo]
IS: Egentligen hade man ju lust, det hade varit bättre att sluta långt,
långt innan, innan man hade, precis när man hade kommit upp på
PL: Ja, för då va man, precis, då hade du varit odödlig, men
IS: [Ja, då hade jag]
PL: [men du] var ju hur bra som helst hur många år som helst
IS: Jo
320 From Clerks to Corpora
PL: Och vann och vann och vann
IS: Jo (ingressive, bilabial fricative)
Norwegian (example 16)
Song about Integration: Norwegian lyrics:
Og snakker du med fremmede på bussen er du dust
Og når du sier “ja” så si det på innoverpust.
Orthographic transcription from:
Swedish (example 17)
[…] Och det var ju så när man flyttade dit, då hade jag den här riktigt
skånska dialekten, va, och sa ”itt och två och tre” och ”e du dum i huvudet” och så där, och alla sa ’va’ hela tiden, så då blev det liksom på nåt
sätt att man försökte göra sig förstådd va, och då började det bli ”ett
och två och tre och’. Sen kom det in nåt jo/ju (ingressive) ibland och så
där va, och då skattade de ihjäl sig när man kom hem på sommaren. […]
I would like to thank several people who have provided useful comments and discussions on ingressive speech, translations, and in one
case links to media clips; in particular I am grateful to Edit Bugge,
Daniel Davis, Clelia LaMonica, Gunnel Melchers, and Friðrik Sólnes.
Clarke, S. & Melchers, G. (2005). Ingressive discourse particles across borders:
Gender and discourse parallels across the North Atlantic. M. Filppula, J.
Klemola, M. Palander & E. Pentitilä (eds) Dialects across Borders: Selected
papers from the 11th international conference on methods in dialectology (Methods XI), Joensuu, August 2002. Amsterdam/Philadelphia: John
Benjamins, 51–72.
Consalvo, M. & Ess, C. (eds) (2011). The Handbook of Internet Studies.
Oxford: Blackwell Publishing Ltd.
Crystal, D. (2013). Language and the Internet. Cambridge Books Online.
Cambridge: Cambridge University Press.
Donovan, K. (Producer, Director, Narrator) (1990). The Irish in Cape Breton
Using the World Wide Web to Research Spoken Varieties of English 321
1713–1990 (part 3 of 6). Sydney, Nova Scotia, Canada: Irish Benevolent
Society. Retrieved from:
Eklund, R. (2007). Pulmonic ingressive speech: A neglected universal?
Proceedings of Fonetik 2007, Stockholm, TMH-QPSR, KTH, 50, 21–24.
———. (2008). Pulmonic ingressive phonation: Diachronic and synchronic characteristics, distribution and function in animal and human sound
production and in human speech. Journal of the International Phonetic
Association 38, 235–324.
Ess, C. & Consalvo, M. (2011). Introduction: What is “Internet Studies”?
M. Consalvo & C. Ess (eds) The Handbook of Internet Studies. Oxford:
Blackwell Publishing Ltd, 1–8.
Hanell, L. & Salö, L. (2009, June). Säg jo och andas in – samtidigt.
Språktidningen 3, 14–19.
Peters, F.J. (1981). The Paralinguistic Sympathetic Ingressive Affirmative in
English and the Scandinavian Languages. (Unpublished doctoral dissertation). New York University, New York.
Pitschmann, L.A. (1987). The linguistic use of the ingressive air-stream in
German and Scandinavian languages. General Linguistics 27, 153–161.
Sample, T. (1991). How to Talk Yankee: A Downeast Foreign Language Record
Featuring Bob Bryan and Tim Sample. Down East Publisher.
Shorrocks, G. (2003). Pulmonic ingressive speech in Newfoundland English: A
case of Irish-English influence? H.L.C. Tristram (ed.) The Celtic Englishes
III. Heidelberg: Universitätsverlag Winter, 374–389.
Squires, L. & Queen, R. (2011). Media clips collection: Creation and application for the linguistics classroom. American Speech 86, 220–234.
Steinbergs, R. (1993). The Use of the Paralinguistic Sympathetic Ingressive
Affirmative in Speakers of English in the St. John’s, Newfoundland Area.
(Unpublished master’s dissertation). Memorial University, St. John’s,
Newfoundland, Canada.
Sundkvist, P. (2012a). Pulmonic ingressive speech in Shetland English. World
Englishes 31, 434–448.
———. (2012b). Pulmonic ingressive speech in the Shetland Isles: Some observations on a potential Nordic relic feature. Nordic Journal of English
Studies 11, 188–202.
Thom, E.J. (2005). The Gaelic Gasp and its North Atlantic Cousins: A Study
of Ingressive Pulmonic Speech in Scotland. (Unpublished master’s dissertation). University College London, London, UK.
18 Another Look at Preposition Stranding:
English and Swedish Discourse Patterns
Francesco-Alessio Ursini
Stockholm University
1. Introduction: Basic Aspects of Preposition Stranding
Preposition Stranding (henceforth PS) is a syntactic construction that
can be found across Germanic languages (Bolinger 1977, 1978). PS
involves a non-canonical word order, in which a preposition and its
object noun phrase (henceforth NP) are not adjacent, hence they appear
to be “stranded”. PS includes three sub-types of constructions: pseudo-passives, relative clauses and wh-constructions (Koopman 2000). The
existence of PS in Southern and Western Germanic languages, such as
German and Dutch, is considered a controversial matter (van Riemsdijk
1990, 1998; Maling & Zaenen 1985; Truswell 2009). However, PS is
certainly attested in English and most Scandinavian languages including Swedish, Norwegian, Danish, and Icelandic (Ǻfarli 1992; Takami
1992; van Riemsdijk 1998; Law 2005; Klingvall 2012). Interestingly,
some of these works observe that PS seems to be an uncommon construction in Swedish (Takami 1992; Klingvall 2012). For this reason, in
this paper we concentrate on English and Swedish, in order to address
whether this phenomenon can receive a unified account across both
languages, regardless of its language-related frequency. We start by discussing some preliminary examples in (1)-(8)1:
(1) (This chair)i was sat on (NP)i (by Luigi)
(2) (The room)i that/Ø we went into (NP)i is occupied
(relative clause)
1 The examples in (1)-(8) have been adapted from (corpora-based) examples found
in previous literature, in particular Takami (1992) (Swedish), Truswell (2009)
(English). The discourse-bound examples in (9)-(14) have been also tested with the
help of native speakers of either language, which we thank for their patience.
How to cite this book chapter:
Ursini, F.-A. 2015. Another Look at Preposition Stranding: English and Swedish Discourse
Patterns. In: Shaw, P., Erman, B., Melchers, G. and Sundkvist, P. (eds) From Clerks to
Corpora: essays on the English language yesterday and today. Pp. 323–347. Stockholm:
Stockholm University Press. DOI: License: CC-BY.
324 From Clerks to Corpora
(3) (Which apples)i are you talking about (NP)i?
(4) (De här sängarna)i har sovits i (NP)i
(The here beds)i have been slept in (NP)i
‘These beds have been slept in’
(5) *(De här sängarna)i har sovits i (NP)i av Jon
(The here beds)i have been slept in (NP)i by John
‘These beds have been slept in by John’
(6) Dörr-en är målad av Jon
Door-ART is painted by John
‘The door has been painted by John’
(7) (Det rum)i som/(Ø) vi har betalat 300 kronor för (NP)i är
(The room)i that/(Ø) we have paid 300 crowns for (NP)i is
‘The room that we have paid 300 crowns for is vacant’
(8) (Vilk-a äpple-n)i pratar du om (NP)i?
(Which-PL apple-PL)i talk you about (NP)i?
‘which apples are you talking about?’
The “basic” position of the stranded NP is indicated via brackets; original and stranded NPs share the same index in order to highlight their
structural relation. The English examples in (1)-(3) show that the NPs
this chair, the room and which apple appear stranded from their governing prepositions, respectively on, into and about. Note: that in (2) can be
omitted, a fact we represent via the symbol “Ø” to represent phonologically null/silent heads. The examples in (4)-(8) offer an illustration of PS
in Swedish: pseudo-passives (i.e. (4)-(5)), standard passive constructions
(i.e. (6)), relative clauses (7) and wh-constructions (i.e. (8)). These examples suggest that PS in Swedish follows similar but not identical patterns
of distribution to English, principally centred on two properties.
First, Swedish pseudo-passive sentences cannot include the “passive”
preposition av ‘by’, as examples (4)-(5) show2. Swedish passive sentences,
2 There is an ongoing debate concerning which prepositions can occur in pseudo-passives (Abels 2003: ch.1-2). We gloss over this debate here, since it is orthogonal to
our discussion.
Another Look at Preposition Stranding: English and Swedish Discourse Patterns 325
on the other hand, can normally include av, like their English counterparts, and introduce the deep subject NP (John, in (6)). Second, Swedish
pseudo-passives include the auxiliary verb ha ‘to have’, rather than vara
‘to be’, as their passive counterparts (Holmes & Hinchcliffe 2008: ch.3).
Thus, Swedish pseudo-passive and passive sentences are not as closely
related as their English counterparts. Examples (7)-(8), instead, show
how relative clause and wh-constructions sub-types of PS are realized in
Swedish. The relative pronoun som ‘that’ may be freely omitted, as in
the case of English. In wh-constructions, the wh-pronoun vilka ‘which’
combines with the NP äpplen and agrees in number, as the glosses suggest.
As the examples seem to suggest, the PS patterns in both languages
seem relatively clear. However, theoretical accounts of PS offer fairly
different analyses of this phenomenon. Simplifying matters somewhat,
classic and minimalist (transformational) proposals offer a movement-based analysis (Hornstein & Weinberg 1981; van Riemsdijk
1990, 1998; Maling & Zaenen 1985; Koopman 2000; Truswell 2009).
According to these analyses, PS is an operation that targets NPs and
moves them into sentence-initial position. In our examples, the bracketed and indexed NPs mark the starting position of the NPs that are
moved in sentence-initial position. Within non-transformational theories, one analysis about PS exists within “Head-Driven Phrase Structure
Grammar” (henceforth HPSG). The analysis found in Tseng (2000,
2004, 2005) suggests that PS involves two “copies” of the same NP,
the two indexed NPs in our example. The sentence-initial copy, instead
of the original copy, is the only phonologically realized copy in a sentence. Qua copies, both NPs must be identical in form; if they are not,
PS cannot be licensed. Although other analyses of PS could certainly
be formulated, this analysis represents the most prominent proposal,
within this non-transformational framework.
Given these assumptions and the data in (1)-(8), it seems that both
types of approaches can offer equally plausible, although theoretically
different accounts of PS. Two questions that arise at this point, given
this equilibrium among proposals, can be formulated as follows. A
first more empirical question is whether we can discuss a broader set
of data, in order to better understand the predictive power of these
approaches. A second more theoretical question is whether we can offer
a third, alternative analysis of this broader set of data that can perhaps
improve upon previous analyses.
The goal of this paper is to offer an answer to this question by offering
a “third way” analysis of PS. This analysis is couched in Type-Logical
326 From Clerks to Corpora
Syntax, a formal, non-transformational approach to morpho-syntactic structures (Morryll 2011). Via this approach, we first offer a more
flexible approach to the lexical properties of our NPs. Then, we sketch
an analysis that shows how these properties are related to syntactic
structures, and that a unified account of PS in English and Swedish is
possible and could perhaps be seen as theory-neutral, to some extent.
To offer this solution, we follow this plan. In section 2, we present the
inter-sentential data; in section 3, we introduce our framework; in section 4, we offer our analysis, and compare it with previous analyses; in
section 5, we offer some conclusions to the paper.
2. The Data: PS Licensing in Inter-Sentential Contexts
Standard theoretical analyses of PS tend to focus on intra-sentential
data, as (1)-(8) show. However, descriptive works usually observe that
PS is often licensed in certain inter-sentential contexts, as well. A common pattern is that when the stranded NP denotes a specific, definite
referent in discourse, then PS can be licensed, although this is not a
necessary condition. This often represents the preferred word order
for a sentence (Huddlestone & Pullum 2002: 137-140; Holmes &
Hinchcliffe 2008: 140; Ward & Birner 2012: 1938-1942). This can be
the case, as PS can license the formation of anaphoric relations between
a stranded NP and a possible antecedent, insofar as these NPs select the
same referent in discourse. These anaphoric relations act as ties that
establish the cohesion of a text, and can be established both as relations
between (argument) NPs, but also via other devices (e.g. temporal relations). Since these factors play a key role in our discussion, we discuss
cohesion and anaphoric relations in a compact but accurate manner, in
the remainder of this section.
We start with cohesion, and define it as a syntactic property of sentences in a text to express information about the same topics and referents in discourse (Kehler 2011: 1964). One way to achieve cohesion
is to have NPs referring to the same entities in sentence-initial position
(the same chair, in ex. (9) below), in any sentence after the initial sentence. This is because the sentence-initial position is often reserved for
topical noun phrases, which usually convey “old” information (Zeevat
2011: 956; Ward & Birner 2012: 1945-1948). Another way to achieve
coherence is also by establishing precise temporal and causal relations
among the events described by each sentence (Zeevat 2011: 957-958;
Kehler 2011: 1965). A possibility that arises, then, is that PS may or
Another Look at Preposition Stranding: English and Swedish Discourse Patterns 327
may not uniquely determine whether a mini-discourse is cohesive: when
the stranded NP conveys information about an “old” referent. I show
these patterns in English via (9)-(11), in which I index the anaphorically
related NPs via sub-scripts:
(9) I bought (this chair)i/(a chair)j on the left. (The chair)i/#j has
been sat on
(10) (This room)i is free. (The room)i/(a room)#j that we went into
is not
(11) A: (Which apples)i are you talking about?
B: (The apples)i/(some apples)#j on the left side of the table
I concentrate on cases in which PS seems to be the key construction
that allows the formation of anaphoric relations. In (9), sequence of
tenses “simple past; present perfect” does not allow the formation of
clear temporal relations. Thus, the event of buying the chair on the left
and somebody sitting on it may be not causally “connected”, but just
occur in the past, with respect to the moment of utterance. The same
reasoning can be applied to (10)-(11), although the sequence of tenses
is different. In these cases, I would like to suggest that PS is the key
syntactic construction that allows cohesion to be established. Since precise causal and perhaps temporal relations between the events that the
sentences describe seem to be lacking, PS seems to be the key factor that
licenses a cohesive mini-text3. The principle behind this relation can be
described as follows.
If a first sentence introduces a definite and specific NP, such as the
chair, then PS is licensed in a second sentence, as (9)-(11) suggest. If
an indefinite, non-specific NP such as a chair is introduced, instead,
the whole mini-text becomes incoherent, as the “#” and the lack of
matching indexes display (i on the chair, j on a chair) in each example.
This is because the sentence-initial NP matches in features (the combination of definiteness and specificity, in (9)-(11)) with the previous NP,
otherwise they cannot possibly refer to the same entity in discourse:
that is, be anaphorically related. Thus, it seems that when no other
3 An important aspect is that other parts of speech can license the formation of temporal, anaphoric relations. An anonymous reviewer observes that the presence of
sentence-final adverbs could render a mini-text involving an indefinite-definite NP
sequence cohesive (I bought a chair yesterday. The chair…). In other words, PS may
also involve non-matching NPs, but when other anaphoric relations allow cohesion
to emerge. In these cases, the relation(s) between NPs are semantic in nature, hence
beyond the scope our discussion. See however von Heusinger (2007); Zeevat (2011)
and references therein, for discussion.
328 From Clerks to Corpora
anaphoric relations can be established, an NP should be stranded from
its preposition for a text to achieve cohesion. Furthermore, this stranding procedure seems to involve “feature-matching” as a condition that
licenses anaphoric relations. Note that what is at stake in these cases is
cohesion, not ungrammaticality. Each sentence in a mini-discourse can
be perfectly grammatical, and yet the resulting text can fail to be cohesive, if no anaphoric relations are established. This fact seems to hold for each of the three PS-type constructions, as examples (9)-(11) suggest. These examples also show that the
stranded NPs are not formally identical to their anaphorically related
NP. Consider (10), for instance: the two NPs this room and the room
share the same values of specificity and definiteness, as a deictic phrase
and a definite noun phrase, respectively (Heusinger 2007, 2011; Diessel
2012). Furthermore, the question in (11) contains a form of wh-construction PS, and can be answered only via an answer that also contains
a definite and specific NP, in this case the apple, otherwise the answer is
incohesive (cf. Krifka 2001, 2004). Thus, one type of anaphoric relation
is established when two NPs carry matching morphological features,
even if the NPs do not belong to exactly the same (syntactic) sub-type.
Examples (12)-(14) show that a similar picture holds for Swedish, but
with certain differences:
(12) Jag köpte (stolen)i/(någon stol)j till vänster.
(Den här stolen)i/#j var sutten på
I bought (chair-ART)i/(some chair)j to left-ART
(The here chair)i/#j was sat on
‘I bought the chair/a chair to the left. This chair was sat on’
(13) (Detta rum)i är ledigt. (Det rum)i/(något rum)#j som vi gick
in i
är det inte
(This room)i is vacant. (The room)i/(some room)#j that we
went into is it not
‘This room is vacant. The/a room that we went in is not’
(14) A:(Vilk-a äpple-n)i pratar du om?
(Which-PL apple-PL)i talk you about?
‘Which apples are you talking about?’
B:(Äpple-na)i/(några äpple-n)#j på vänster sida av
(Apple-PL.ART)i/(some apple-PL.ART)#j on left-ART side
of table-ART
‘The apples/some apples on the left side of the table’
Another Look at Preposition Stranding: English and Swedish Discourse Patterns 329
Example (12) shows that when the non-specific någon stol ‘a chair’
(lit. ‘some chair’) occurs in the first sentence, the NP den här stol ‘this
chair’ does not find a suitable antecedent. This is the case, as någon
stol is an antecedent NP that does not match the features that Den
här stolen carries. The net effect is that the speaker appears to identify
via the second sentence one chair that, however, was not mentioned in
the first sentence. Thus, the mini-discourse appears incohesive, since no
other anaphoric relations (e.g. temporal ones) are established. The same
patterns emerge in (13)-(14), modulo the slightly different types of NPs
and PS constructions involved4. Hence, both specificity and definiteness seem to play a role in the distribution of PS in discourse, as these
features allow the establishment of cohesion (or lack thereof) in a text,
when no other anaphoric relations can be established. Hence, an analysis of this phenomenon must include a treatment of the role of these
features, and their relation to word order and cohesion.
One interesting dilemma that these data present consists in assessing
the exact nature of this problem. From a theoretical perspective, specificity and definiteness are features that play a role at a morphological
and semantic level of representation. The morphological values permit an anaphoric relation between two NPs to be established, and the
related mini-text may become cohesive as a result. Both the chair and
this chair are specific, definite NPs that can refer to a given chair. Hence,
they can refer to the same chair in discourse, and license a discourse
that is also semantically coherent (Kehler 2011; Zeevat 2011). As our
data suggest, our problem regarding the nature of PS seems to involve a
complex interplay between morphological, syntactic and semantic levels of representation.
As a consequence of this pin-pointing the nature of our PS dilemma,
our empirical question can now receive a precise formulation. The
extension of either current HPSG or minimalist analyses of PS to
these data does not appear as a simple matter. One problem that
both approaches share pertains to the feature values of anaphorically
related NPs. The approach outlined in e.g. Tseng (2000: ch.4) includes
an inter-sentential treatment of anaphoric phenomena that transformational approaches lack (Klingvall 2012), but the morpho-semantic
4 We note here that this pattern emerges when indefinite and non-specific NPs such
as någon stol are involved. Swedish differs from English in having a second indefinite article: en lit. ‘one’, ambiguous with respect to specificity. However, in PS constructions en can only have specific value (e.g. en stol ‘one (specific) chair’, in (12)).
Our analysis can be extended to en as well, but the use of någon in our example
better highlights the parallels between Swedish and English PS constructions.
330 From Clerks to Corpora
problem of anaphora resolution would remain intact. This is the case,
as this approach crucially relies on copied NPs and their phonologically realized counterparts to be formally identical. Our data in (9)(14) suggest that this assumption is not tenable, at an inter-sentential
level. Two NPs can be anaphorically related, license PS and a cohesive
text, as long as they both carry the definiteness and specificity features.
Therefore, a more flexible approach seems to be called for. Since both
current analyses of PS seem to be problematic, we propose a third
analysis in the next section.
3. The Proposal: An extended TL calculus
In this section our formal analysis is based on a framework known
as Type-Logical calculi (henceforth TL, Jäger 2001, 2005; Moortgat
2010, 2011; Morryll 2011). Our variant of TL implements some
assumptions from certain variants of minimalist syntax, notably
Distributed Morphology (Embick & Noyer 2001; Harbour 2007;
Harley 2012). However, the proposal I wish to make takes an inherently non-transformational perspective, in part closer to HPSG and
other similar frameworks. I will spell out these assumptions, as we proceed in our presentation of the framework. We choose our variant of
TL for two reasons. First, TL is a formally explicit framework that can
treat morphological and syntactic data alike, without any supplementary assumptions specific to each domain. Second, it offers tools that
allow one to easily analyse inter-sentential data, at least in our formulation. Here we offer a compact discussion; a more thorough presentation of this framework is found in previous work of the author (Ursini
2011, 2013 a, b, c, 2015 a, b; Ursini & Akagi 2013 a, b). I present some
key assumptions of our TL analysis, then we move to our innovations.
First, parts of speech are mapped onto or assigned types, which can
be considered as either being “complete” or “incomplete” bits of morpho-syntactic information. Complete types represent constituents that
can stand as distinct, independent constituents (e.g. np for NPs such
as the girl). Incomplete types represent constituents that must combine
with other constituents to form a complete unit. An intransitive verb
such as runs can be assigned type np\s, since it can combine with an np,
the girl. The result is the sentence the girl runs, which is assigned the
type s of sentences. Thus, types can also be used to represent the syntactic valence of lexical items, and possible restrictions on which types
of arguments/phrases heads can take.
Another Look at Preposition Stranding: English and Swedish Discourse Patterns 331
Second, we implement two connectives, “/” and “·”, that are known as
the right division (or just division) and the product connectives (Moortgat
2010: § 2; Morrill 2011: ch. 1). Both connectives are binary and associative, but product is also non-commutative. Products of types (e.g. a·b) are
taken in this order. Third, we follow some TL calculi that take a psycholinguistic model of sentence production, and propose that sentences are
derived in a top-down (“left-to-right”, in linear terms) manner (Morryll
2011). This assumption is also found in psycholinguistics models such as
Levelt (1989); Phillips (2006); Jarema & Libben (2007). More importantly, this assumption will turn out to be germane to our goals, since it
allows us to treat anaphoric phenomena in a straightforward manner, as
shown in section 4. The three assumptions represent innovations that I
introduce with respect to standard TL calculi.
First, I leave aside other standard TL connectives, such as left division
“\” and Jäger’s (2001, 2005) connective “|” for anaphoric relations. In
particular, I will suggest a way to treat anaphoric relations that exclusively rely on the interplay of division and product known as the merge
schema, in part adumbrated in Jäger (2001: 78-81). I aim to show that,
once we offer a formal analysis of the distributional properties of our
lexical items, our basic combinatorial system will suffice to account for
our data. Specifically, I aim to show that our PS data require a simple
but precise analysis of the distribution of prepositions and nouns based
on their morphological features.
Second, I take a more sophisticated view concerning types other
than the one found in standard TL calculi, as we follow recent analyses on the nature of morpho-syntactic categories. Thus, I assume that
lexical and functional categories are not primitive categories, but clusters of morphological features (Hale & Kayser 2002; Harbour 2007;
Adger 2010; Acquaviva & Panagiotidis 2012). For instance, nouns
include features such as gender and number while prepositions lack
these features, but they may include a “spatial” feature or similar other
non-nominal features. Hence, in our system morphemes correspond
to products of features, which can then differ with respect to value
they can carry (e.g. “male” or “female” for the gender feature). Here
I follow proposals that assume the “separation hypothesis”: vocabulary insertion occurs after morphological derivations (Levelt 1989;
Embick & Noyer 2001). Thus, abstract morphological objects may
lack an overt phonological exponent, or are realized by different exponents across languages (here, that vs. som: Embick & Noyer 2001). We
return to this point in section 4.
332 From Clerks to Corpora
Third, I assume the distributional properties of categories can be
represented explicitly, via our type system. If transitive verbs (loves)
and prepositions (e.g. to) act as heads with 2-valence, then their type
should reflect this shared property. Hence, different categories and constituents (verbs, prepositions, phrases, sentences, and discourses) can be
reduced to a handful of types. Depending on the valence of a constituent, one can determine the type assigned to this constituent. I capture
this assumption by using only one basic type p, which is mnemonic for
both “phrase” and “product of features”. The rules for deriving other
types are defined in (15):
(15) 1. p is a morphological type (Lexical type)
2. If x is a type and y is a type, then x/y is a type (Type I.:
3. If x is a type and y is a type, then x·y is a type (Type I.: Product)
4. If x/y is a type and y is a type, then (x/y)·y x, y·(x/y) x (MI:
For. A.)
5. If x/y is a type and y/z is a type, then (x/y)·(y/z) x/z (MI: Cut
6. Nothing else is a type (Closure rule)
In words, rule 1 introduces our basic type p. Rule 2 says that two
basic types combined via division (x, y) form a complex type, e.g. a
head which can take one argument (here, x/y: a definite article taking
an NP as an argument, as in the car). Rule 3 says that two basic types
combined via product form a complex type that bundles information
(i.e. x·y). Rule 3 can also be used to introduce information (i.e. from x
to x·y), in a manner that we will discuss thoroughly in section 4. In this
case, we take our basic type to represent single features, a move that we
also fully motivate in section 4.
Rules 4 and 5 introduce two instances of merge schemas, rules
that govern how (right) division and product types interact. Rule 4 is
known as forward application. It says that the product of a complex
(division) type and a simple type yields a certain output type, provided
that the input type of the complex type (here, y) matches that of the
simple type. Thus, rule 4 governs how a head can combine with an
argument to form a more complex constituent such as a phrase, for
instance. It also determines what the output of this phrase is (here, x),
provided that the two input types “match”. As rule 4 plays a crucial
role in our analysis, we return to the specific details of its application
when we discuss the data.
Another Look at Preposition Stranding: English and Swedish Discourse Patterns 333
Rule 5 is a merge schema known as the cut rule, and says that two
complex types sharing their “internal” type can be combined into a
new type. This rule plays a key role for the analysis of our discourse
data; we delay a more precise explanation to section 4. Rule 6 says that
no other rules are necessary. In our system, the symbol “ ” represents
the merge schemas as ternary relations between the types of two input
constituents and the type of their output, the constituent they form
when merged together. If the NP the girl and the verb runs are assigned
matching types, their merge will form the VP the girl runs. As a consequence, in our system rules 4 and 5 offer two formally precise schemas
to “prove” that larger constituents can be formed, via our basic set of
rules and types.
Overall, our minimal set of derivational rules allow us to generate
complex type sets, intended as types that we can assign to our constituents. For our purposes, the set TYPE={p·p, p·p/p·p, p·p/p·p/p·p} will
suffice to account for all the data at hand. This set respectively includes
arguments taken as bundles of features, 1- and 2-valence heads. The
precise nature of these types will become clear when we discuss how
and why these types are assigned to our constituents, in the next section. However, before we introduce our analysis, we must discuss one
last aspect of our formal apparatus. In order to capture the incremental
nature of our derivations, we define a simple pre-order as the pair of
an interval set I, and an addition operation “+”, i.e. <I,+>. This pre-order represents an index set, which in turn allows the representation
of all the steps in a derivation as sequential elements (e.g. t, t+1, t+2,
etc.). We also implement two labels, Lexical Selection (LS) and Merge
Introduction5 (MI) in order to explicitly mark the introduction of a
new element in a derivation, and the merge of two elements, respectively. With this formal apparatus at our disposal, we move to our data.
4. The Analysis: The Distribution of PS
The goal of this section is to offer our TL analysis of the data. We start
by motivating our type assignment for our constituents, before moving
to the derivations that illustrate how we can account for our data.
5 In TL calculi, merge is an elimination, rather than introduction rule for division: It
removes slashes in a structure. The label “introduction” stresses that morphemes
are combined into more complex structures.
334 From Clerks to Corpora
We start by motivating which categories are assigned the (product)
type p·p, which represents phrases that carry bundled morphological
features. We assign this type to NPs qua phrasal arguments of a head.
As our examples show, NPs carry (at least) the features p+f or p-f, for
specificity, definiteness and similar other features. This simple fact seems
to motivate the use of a product type for NPs. We then assume that
wh-phrases which apples and vilka äpplen can be assigned the same type.
We thus follow standard treatments of this category, although we choose
a more coarse-grained perspective than standard analyses of wh-phrases
(Alexiadou et al. 2000; Bianchi 2002a, b; Vermaat 2005). We represent
this complex type as p±spec, for simplicity, and leave the values for definiteness, (plural) number, pronominal and relative features implicit, in
our analysis. Nothing crucial hinges on this notational simplification.
As our initial type assignment shows, we also assume that features can
have different values, which can in turn determine whether an instance
of merge is successful or not. Analyses of feature systems abound in the
literature, in TL calculi and other frameworks (Johnson & Bayer 1995;
Bernardi & Szabolcsi 2008; Tseng 2005; Adger 2010; Stabler 2013). As
we only discuss cases in which the binary value(s) of features may determine the well-formedness or cohesion of a syntactic (or discourse) derivation, our analysis has an inherently theory-neutral perspective. However,
the core aspects of our analysis are based on Johnson & Bayer (1995)
and their proposal on feature percolation, as it will soon become clear.
We move to spatial prepositions, as we wish to offer an argument
for assigning them the type p·p as well. Several minimalist analyses
treat these prepositions as sequences of functional heads, which may
or may not be phonologically realized. Thus, a non-stranded preposition such as in would be assigned the type p·p/p·p/p·p of heads, with
its complement phrase being an NP. A stranded preposition would be
assigned the same type, but it would also combine with a non-pronounced (silent) copy of the stranded NP (Koopman 2000; Truswell
2009; Svenonius 2006, 2010). Here we follow a different minimalist
analysis of SPs, sometimes known as the “P-within-P” hypothesis (Hale
& Kayser 2002; cf. also Emonds 1985; van Riemsdjik 1990, 1998).
This analysis suggests that spatial prepositions involve a complex structure, in which a silent prepositional head takes another preposition as
its specifier phrase6. Hence, a prepositional phrase such as in the garden
6 TL calculi usually do not employ “silent” categories, although this assumption is
not uniform (cf. Jäger 2005: ch. 2 vs Moortgat 2011 §2). This matter is not crucial,
for our discussion.
Another Look at Preposition Stranding: English and Swedish Discourse Patterns 335
would involve the phrase in (i.e. the “internal” preposition), a silent
head “(P)”, and the complement phrase the garden.
Via this assumption, our prepositions occurring in PS contexts are
assigned the type p·p, which is the same type they would receive in a
non-stranded position. A minimal proviso is that when these prepositions occur in stranded positions, they act as complements of a verb,
rather than specifiers of another preposition. Thus, prepositions such as
in in the room we went in have a particle-like distribution, as complements of the verb (here, went: Åfarli 1992; Abels 2003). Importantly,
although spatial prepositions and NPs are assigned type p·p, they differ
in the values of the features they carry as phrases. For instance, prepositions seem not to carry the feature value p±spec of NPs, a minimal
morphological difference that suffices to distinguish these two categories. Both categories, though, can act as arguments of relational heads,
something we represent via the general use of product types for phrases.
We move to our analysis of heads: verbs, prepositions introducing
pseudo-passive constructions and complementizers. We assign the type
p·p/p·p/p·p of 2-valence heads to these three categories, although with
some minimal differences. First, verbs carry a “voice” feature, thus
either a passive or active value on their output type. They can only
merge with a preposition that matches these features, such as passive
prepositions. For the sake of simplicity, we treat compound verbs as
forming a single lexical unit: both har sovits ‘have been slept’ and has
been sat receive this type, qua verbs. We thus assign type p·p+spec/p·p±/p·p to each verb7, depending on whether it has active or passive
voice. Note that since temporal features do not play a crucial role in
the analysis of our examples, we omit them. However, a more thorough analysis of PS patterns could be offered, by adding a discussion of
these features and their role in forming cohesive texts. We leave such an
extension aside, for the time being.
Second, we also assign the type p·ppass/p·p/p·p to passive prepositions
(by, av) as heads, for a simple reason. Standard analyses of passive sentences suggest that these prepositions take the deep subject NP (Luigi
in (1)) and the passive verb phrase (e.g. this chair has been sat on in
(1)) as arguments (Abels 2003; Gehrke & Grillo 2009; Ishizuka 2010).
Thus, their specifier input type must carry a passive feature value.
Third, we also assign this relational type to complementizers, as heads
7 For simplicity, rather than for necessity, we also assume that our heads always take
a specific (specifier) input type p·p+spec, since all our examples involve specific NPs.
336 From Clerks to Corpora
introducing relative clauses (that, som), in line with standard assumptions (Alexiadou et al. 2000; Bianchi 2002a; Vermaat 2005). We thus
assign the type p·p+spec/p·prel/p·p to this category. Before we show how
these types can be combined together to form our sentences, we summarize our type assignment in (16):
(16) a. p·p={this chair, de här sängarna, up, i, which apples, vilka
b. p·p/p·p/p·p={was sat, by, av, that, som}
From this type assignment and the rules in (15), we can offer an
account of how our examples are derived, starting from pseudo-passives.
We repeat (1) as (17a) to illustrate our account. We use simplified notations for types (e.g. p+ for pspec, p’ for ppass, p’’ for prel) and shortened lexical entries in our derivations (e.g. this for this chair) for reasons of space:
(17) a. This chair was sat on by Luigi
b. t. [ this chairp·p+]
t+1. [ was satp·p+/p·p’/p·p]
t+2. [ this chairp·p+]·[ was satp·p+/p·pp’/p·p]
[p·p’/p·p[ this chairp·p+] was satp·p+/p·p’/p·p]
t+3. [ onp·p]
t+4. [p·p’/p·p[ this chairp·p+] was satp·p+/p·p’/p·p]·[ onp·p]
[p·p’[ this chairp·p+] was satp·p+/p·p’/p·p [ onp·p]]
t+5. [ by p·p’/p·p/p·p]
t+6. [p·p’[this chairp·p+] was satp·p/p·p’/p· p[onp·p]]·[ byp·p’/p·p/p·p]
[p·p/p·p[p·p’[thisp·p+] was satp·p+/p·p’/p·p[ onp·p]] byp·p’/p·p/p·p] (MI)
t+8. [ Luigip·p]
t+9. [p·p/p·p[p·p’[thisp·p+] wasp·p+/p·p’/p·p[onp·p]] byp·p’/p·p/p·p]·[ Luigip·p]
[p·p[p·p’[thisp·p+] wasp·p+/p·p’/p·p[onp·p]] byp·p’/p·p/p·p[ Luigip·p]](MI)
Our derivation reads as follows. The NP denoting the object, this
chair, merges with the passive verb was sat. The result of this operation
merges with the spatial preposition on, thus deriving the VP this chair
was sat on (steps t to t+4). The preposition by carries passive features,
and merges with the passive VP this chair was sat on. This is the case,
as the passive feature values of the two merged constituents match (step
t+6). The deep subject NP Luigi is then merged, and (17a) is derived
accordingly (steps t+7 to t+9). An important result of this derivation,
then, is that the word order in a sentence involving PS can be derived
without assuming silent or copied NPs in the sentence-final position.
Thus, we can explain how English pseudo-passives are derived as complete sentences without resorting to assumptions involving copied/
Another Look at Preposition Stranding: English and Swedish Discourse Patterns 337
moved constituents. Since we have also shown how a grammatical pseudo-passive sentence is derived, we can now show how an ungrammatical sentence is instead blocked. We do so by offering a “compressed”
derivation for our Swedish example (5), repeated as (18a):
(18) a. *De här sängarna har sovits i av Jon
b. t+6. [p·p[De här sängarnap·p+]har sovitsp·p/p·p/p·p[ ip·p]]·[ avp·p’/p·p/p·p]
(Feature mismatch, derivation crashes)
We focus on the derivational step at which the derivation blocks
or “crashes”, t+6. The merge of a clausal phrase that lacks a passive
feature (de här sängarna har sovits i ‘these beds have been slept in’)
and passive preposition results in a mismatch of features. Recall now
that Swedish pseudo-passives differ from “true” passives by having ha
‘to have’ as the main auxiliary verb, hence being closer to active forms.
This fact suggests that har sovits carries the opposite voice feature values of av. As we have established via rule 4, if two merged units do not
match in type/feature value, then a derivation is blocked. Thus, (18b)
suggests that our analysis predicts the ungrammaticality of (18a) as a
feature mismatch case. This basic aspect of merge, together with the
type assignment for our English and Swedish lexical items, seems to
suffice to explain this datum. Via this result, other data can now also
be accounted for, a fact that we show by focusing on relative clause PS
patterns in both languages. Recall that the difference between English
and Swedish pertains to the exponent that realizes the complementizer
head: that and som. If the exponents differ, but the lexical items that are
merged are the same, then the same type of derivation can generate the
structures of both English and Swedish relative PS sentences. We show
this derivational symmetry in (19)-(20):
(19) a. The room that we went into is occupied
b. t. [ the roomp·p+]
t+1. [ thatp·p+/p·’’/p·p]
t+2. [ the roomp·p+]·[ thatp·p+/p·p’’/p·p]
[p·p’’/p·p[ the roomp·p+] thatp·p+/p·p’’/p·p]
t+3. [ we went intop·p]
t+4. [p·p’’/p·p[ the roomp·p+] thatp·p+/p·p/p·p]·[ we went intop·p]
[p·p’’[ the roomp·p+] thatp·p+/p·p/p·p[ we went intop·p]] (MI)
(20) a. Det rum som vi har betalat 300 kronor för är ledigt
b. t. [ det rump·p+]
338 From Clerks to Corpora
t+1. [ somp·p+/p·p’’/p·p]
t+2. [ det rump·p+]·[ somp·p+/p·p’’/p·p]
[p·p/p·p[ det rump·p+] somp·p+/p·p’’/p·p]
t+3. [ vi har betalat förp·p]
t+4. [p·p’’/p·p[ det rump·p+] somp·p+/p·p’’/p·p]·[ vi har betalat förp·p]
[p·p’’[ det rump·p+] somp·p/p·p’’/p·p[ vi har betalat förp·p]] (MI)
These derivations suggest that our subject relative clauses, such as
the room that we went in, can include the stranded NP, in this case the
room. Our type assignment and our derivational rules show that this
subject relative clause is well-formed and can be merged with the rest of
the sentence as a phrase (argument). Thus, our analysis of PS, when also
applied to this type of PS constructions, seems to offer a parsimonious
but overall accurate analysis of how the observed word order can be
Before we move to wh-constructions, we discuss the optional/obligatory realizations of complementizers. Recall now that som and that
may be omitted (cf. (2), (7)), a phenomenon that is “post-syntactic”
in our account (Embick & Noyer 2001). That is, it pertains to which
vocabulary exponents are inserted in a derivation, not to the morpho-syntactic objects that are merged together. Thus, it can be based
on language-specific ellipsis rules, which usually target specific feature
values that elided categories can carry (Merchant 2001: ch. 1-2, 2004).
Since we distinguish between morpho-syntactic derivations and phonological operations, language-specific operations are a natural consequence of our approach, one example being relative clauses in PS
contexts. Once we have this second piece of our PS puzzle in its correct
place, we can offer an analysis of wh-pronoun PS constructions, which
builds on the results obtained so far. We repeat (3) and (8) as (21a) and
(22a), to show this fact:
(21) a. Which apples are you talking about?
b. t. [ which applesp·p+]
t+1. [ are you talkingp·p+/p·p/p·p]
t+2. [ which applesp·p+]·[ are you talkingp·p+/p·p/p·p]
[p·p/p·p [ which applesp·p+] are you talkingp·p+/p·p/p·p]
(22) a. Vilka äpplen pratar du om?
b. t. [ Vilka äpplenp·p+]
t+1. [ pratar dup·p+/p·p/p·p]
t+2. [ Vilka äpplenp·p+]·[ pratar dup·p+/p·p/p·p]
[p·p/p·p[ Vilka äpplenp·p+] pratar dup·p+/p·p/p·p]
Another Look at Preposition Stranding: English and Swedish Discourse Patterns 339
The partial derivations in (21b) and (22b) read as follows. For the
sake of simplicity, we treat both you and du as part of the verbal head,
hence treating these pronouns as clitic-like elements (cf. Koopman 2000;
Jäger 2005). These verbs merge with the relative NPs which apples and
vilka äpplen ‘which apples’, which also carry a specificity feature. Since
we do not explicitly represent other feature values, these derivations
appear equivalent to those we offered in examples (18b)-(20b). Since
our type assignment for relative NPs and our derivational analyses
prove that silent NPs must be merged into our sentences, we can extend
our parsimonious analysis to this PS sub-type, too8.
We now have an account of each of the three sub-types of PS constructions in intra-sentential examples found in (1)-(8). Thanks to this
result, we are in a position to also sketch an account for our inter-sentential examples in (9)-(14). For this purpose, I build on Jäger’s (2001:
84-86) implementation of TL calculi to derive discourse structures.
Differently from his proposal, however, we do not employ a special
type for sentences as part of discourses (his type D). In our system,
we assign type p to sentences qua complete syntactic objects, as our
derivations show. Thus, I sketch an approach in which the same logical
analysis can be applied to different levels of structure, and merge can
act as a schema that combines sentences into discourses.
We must now account for two important problems that arise at this
level of analysis: inter-sentential anaphoric relations and, consequently,
discourse cohesion. Since anaphoric relations can define the cohesion of
discourses (Kehler 2011; Reuland 2011), their analysis can permit us to
sketch a preliminary syntax of discourse that can also account for PS.
For this purpose, I take the fairly standard assumption that a system of
feature percolation is active (Adger 2010: 188-195; Tseng 2005; Stabler
2013). The theory-neutral assumption is that the features of constituents making up sentences can percolate at a sentence level, and constrain
how anaphoric relations can be established. Insofar as at least one set of
features can license the formation of anaphoric relations, then cohesion
is obtained. Thus, PS can potentially permit the formation of a cohesive
text, in cases when no other anaphoric relations can be established (i.e.
8 We must offer one caveat. We partly deviate from standard analyses of questions,
including TL-based ones (Vermaat 2005), since we are concerned with offering an
account of PS. While standard accounts of questions assign a type close to p/p, here
we treat these types of sentences as if they were simple declarative sentences, of type
p. This is a simplification, although a non-problematic one.
340 From Clerks to Corpora
our (9)-(14)). In order to capture these facts, however, we must capture
how feature percolations systems work, in the first place.
For this purpose, I follow feature percolation systems found in TL
calculi (Johnson & Bayer 1995; cf. also Bernardi & Szabolcsi 2008). I
assume that percolation involves the “duplication” of the feature types
of an argument NP to the VP that contains this NP, as per rule 3 (type
formation: product). Since we are discussing PS and its contribution
to cohesion, we can restrict this assumption to the features that NPs
contribute to a sentence. Hence, I assume that the features that allow
the formation of an anaphoric relations are those that the stranded
NP in the second sentence, and its non-stranded counterpart in the
first sentence contribute. Thus, if p represents the type of a VP such as
(17a), p·p+spec can represent9 the type of this sentence, made ready to be
merged with another sentence taking this type as an input10. Once these
features are percolated, an anaphoric relation can be established, and
the two sentences that include these NPs form a cohesive discourse. In
order to show how our analysis works, I repeat (9) as (23a) and offer
its two simplified derivations in (23b-c)11:
(23) a. I bought the chair/a chair on the left. This chair was sat on
by Luigi
b. t. […the chair onp·p+]
t+1. [ This chair…p·p+]
t+2. […the chair on.p·p+]·[ This chair…p·p+]
(MI: cut rule)
[p·p […the chair onp·p+ .This chairp·+p]]
c. t. […a chair onp·p-]
t+1. [ This chair…p·p+]
(Der. crashes)
t+2. […a chair onp·p-]·[ This chair…p·p+] *
In words, the two sentences I bought the chair on the left and this
chair… are both assigned the type p·p+spec. This is the type assigned to
9This type minimally differs from the type assigned to NPs by the specific value
assigned to p, since it may represent sentence types, rather than nominal types.
Again, I do not explicitly represent this distinction, for mere reasons of space.
10 In this case, we assume rather than prove that our mechanism of feature percolation
derives the results we discuss. See Bernardi & Szabolcsi (2008: § 1-2) for discussion.
Note also that we implement product types, although the standard definition of the
cut rule we offer is better suited for functional types.
11 Our system could offer an incremental account of how these mini-discourses are
derived, not unlike other related proposals in the literature (e.g. Asher & Lascarides
2003). Here we only merge two fully formed sentences, in order to sketch how our
apparatus works.
Another Look at Preposition Stranding: English and Swedish Discourse Patterns 341
sentences that receive the relevant features from one of their argument
NPs via feature percolation. Via the cut rule, a merge schema and rule 5
in our set of rules, we can establish an anaphoric relation between NPs
across sentences. In doing so, we also establish one form of cohesion
between the sentences that contain them, as a reflection of the “matching” component of merge. In this way, we can offer a proof on how to
derive our discourse example. Hence, our analysis can derive anaphoric
relations that emerge in PS constructions, as instances of our general
type-matching mechanism that is part of merge.
One important caveat is that our examples highlight the possible
role of PS as one cohesion-building construction, but certainly as not
the only construction to do so. Recall from section 2 that the verbs in
our examples lack temporal features that permit to establish anaphoric
relations between sentences. For instance, the event of buying a chair
and that of someone sitting on this chair, described by the mini-text
in (23), are not necessarily related. In this case, PS seems to reflect the
fact that the only features that can percolate, and allow cohesion to be
established, are those that the stranded NP carries. Consequently, the
derivation in (23c) shows that when not even this anaphoric relation
can be established, a mini-discourse becomes incohesive. The cut rule
cannot merge two sentences that lack matching types, so a cohesive
mini-discourse cannot be formed. Therefore, we can now capture the
fact that PS can allow the formation of cohesive discourses, especially
when no other constructions can do so.
Two other results that we obtain via our analysis are the following.
First, we can now predict a preference for PS constructions in inter-sentential examples. This can be seen as a strategy that allows for an easy
resolution of anaphoric relations, via cross-sentential feature-matching
(Reuland 2011; Ward & Birner 2012). Second, we can indirectly predict
that, when other features can percolate at a sentential level, cohesion
may be established even if a stranded NP does not match the features
of its anaphoric counterpart. If the events described by each sentence
are anaphorically related via e.g. temporal morphology on verbs and/
or adverbs, then we could expect that some chair could occur in (23a),
and the mini-discourse be cohesive. If cohesion emerges insofar as the
types of two sentences match on one feature (value), e.g. p+tense, then PS
does not univocally determine the cohesion of a discourse. Once more,
however, our examples suggest that it may play this role, given the more
general mechanisms that govern cohesion. Thus, an extension of our
analysis could ideally account for these data, as well.
342 From Clerks to Corpora
Given these two results, we have now shown how our non-transformational analysis can now derive all the examples in (1)-(14) in a
principled and parsimonious way. As we have reached our main goal,
we can move to the conclusions.
5. Conclusions
In this paper we have offered an account of Preposition Stranding
(PS) and its three sub-types (pseudo-passives, wh-constructions, relative clauses) in English and Swedish. We have suggested that the three
assumptions on which our account rests upon can offer an empirically
broader and theoretically parsimonious analysis of this phenomenon.
Our first assumption is that, if we treat morphological features as “bits”
of information, then we can successfully account for how (and when)
stranded constituents are inserted (merged) in a sentence. Our second
assumption is that, if we pursue a top-down (“left-to-right”) derivational approach to sentences, then we do not need copying/movement
analyses to account for word order in PS constructions. Our third
assumption is that the feature-matching aspect of merge suffices to
capture the anaphoric relations that arise between sentences when PS
constructions are involved, and that can create discourse cohesion. This
is obtained via a very simple system of feature percolation that simply
copies “old” morphological information at the level of discourse (here,
the specificity and definiteness of NPs).
Thus, our account seems to be successful in explaining our data,
and seems to sketch an alternative analysis to both minimalist proposals (Truswell 2009) and HPSG-based proposals (e.g. Tseng 2000).
However, nothing prevents that feature-based proposals can be offered
within these frameworks, that can cover our data and perhaps broader
sets, in a similarly accurate manner. I would like to suggest that the
current analysis can indeed be seen as complementing and enriching
previous analyses, as well as our understanding of PS as a syntactic
construction and its relation to discourse structure and cohesion (cf.
Ward & Birner 2012).
It is goes without saying that we have not exhausted the discussion
on PS. For one thing, we have focused on a specific sub-set of data,
and left aside the possibility that PS can interact with other syntactic
constructions, and partake in a fairly complex interaction between feature percolation and cohesion phenomena. This has been a necessary
choice, given our limits of space and the complexity of PS as a general
Another Look at Preposition Stranding: English and Swedish Discourse Patterns 343
phenomenon. We also have left aside a discussion of other Scandinavian
languages, such as Norwegian or Icelandic, in which PS seems to be a
more common, but also a more complex phenomenon. Also, we conjecture that this approach can also potentially offer a morpho-syntactic counterpart of the semantic treatments of anaphoric relations
(Elbourne 2005; Kamp, van Genabith & Reyle 2011). Furthermore,
our work sketches the possibility that our Type-Logical analysis can be
extended to a more sophisticated theory of discourse structure, as discussed in the literature (Asher & Lascarides 2003). However, we leave
such theoretical problems for future work.
Abels, K. (2003). Successive cyclicity, anti-locality and adposition stranding.
Ph.D. dissertation, University of Connecticut.
Acquaviva, P. & Panagiotidis, P. (2012). Lexical Decomposition meets conceptual atomism. Lingue e Linguaggio XI:2, 165–180.
Adger, D. (2010). A minimalist theory of feature structure. A. Kibort & C.
Greville (eds) Features: Perspectives on a Key Notion in Linguistics. Oxford:
Oxford University Press, 185–218.
Ǻfarli, T. (1992). The Syntax of Norwegian Passive Constructions. Philadelphia:
John Benjamins.
Alexiadou, A., Law, P., Meinunger, A. & Wilder, C. (eds) (2000). The Syntax of
Relative Clauses. Amsterdam: John Benjamins.
Asher, N. & Lascarides, A. (2003). Logics of Conversation. Cambridge:
Cambridge University Press.
Bernardi, R. & Szabolcsi, A. (2008). Optionality, scope, and licensing: an application of partially ordered categories. Journal of Logic, Language and
Computation 17:3, 237–289.
Bianchi, V. (2002a). Headed relative clauses in generative syntax: Part I. Glot
International 6:7, 118–130.
———. (2002b). Headed relative clauses in generative syntax: Part II. Glot
International 6:8, 235–247.
Bolinger, D. (1977). Transitivity and spatiality: The passive of prepositional
verbs. A. Makkai, V. B. Makkai, & L. Heilmann (eds) Linguistics at the
Crossroads. Lake Bluff, IL: Jupiter Press, 57–78.
———. (1978). Passive and transitivity again. Forum Linguisticum 3, 25–28.
Diessel, H. (2012). Deixis and demonstratives. C. Maienborn, K. von Heusinger
344 From Clerks to Corpora
& P. Portner (eds) An International Handbook of Natural Language
Meaning (Vol. 3). Berlin: Mouton de Gruyter, 2407–2431.
Embick, D. & Noyer, R. (2001). Movement operations after syntax. Linguistic
Inquiry 32:4, 555–595.
Elbourne, P. (2005). Situations and Individuals. Cambridge, MA: The MIT
Emonds, J. (1985). A Unified Theory of Syntactic Categories. Dordrecht: Foris
Gehrke, B. & Grillo, N. (2009). How to become passive. K. Grohmann (ed.)
Exploration of Phase Theory: Features, Arguments, and Interpretation at
the Interfaces. Berlin: De Gruyter, 213–268.
Hale, K. & Keyser, S. J. (2002). Prolegomenona to a Theory of Argument
Structure. Cambridge, MA: MIT Press.
Harbour, D. (2007). Morphosemantic Number: From Kiowa Noun Classes to
UG Number Features. Dordrecht: Springer.
Harley, H. (2012). Semantics in distributed morphology. K. von Heusinger, C.
Maierborn & P. Portner (eds) Semantics: An International Handbook of
Natural Language Meaning. Amsterdam: De Gruyter, 688–709.
Heusinger, K. von (2007). Accessibility and definite noun phrases. M. SchwarzFriesel, M. Consten & M. Knees (eds) Anaphors in Text: Cognitive, Formal
and Applied Approaches to Anaphoric Reference. Amsterdam: Benjamins,
———. (2011). Specificity. K. von Heusinger, C. Maienborn & P. Portner (eds)
Semantics: An International Handbook of Natural Language Meaning,
Volume 2 Berlin: de Gruyter, 1024–1057.
Holmes, P. & Hinchliffe, I. (2008). Swedish: An Essential Grammar. Routledge:
New York.
Hornstein, N., & Weinberg, A. (1981). Case theory and preposition stranding.
Linguistic Inquiry 12:1, 55–91.
Huddleston, R. & Pullum, G.K. (2002). The Cambridge Grammar of the
English Language. Cambridge: Cambridge University Press.
Ishizuka, T. (2010). Towards a Unified Analysis of Passive in Japanese: A
Cartographic Minimalist Approach. Ph.D. Dissertation, UCLA.
Jäger, G. (2001). Anaphora and quantification in Categorial Grammar. M.J.
Moortgat (ed.) Logical Aspects of Computational Linguistics, Springer
Lecture notes in Artiificial Intelligence 2014. Springer:Dordrecht, 70–90.
———. (2005). Anaphora and Type Logical Grammar. Springer: Dordrecht.
Another Look at Preposition Stranding: English and Swedish Discourse Patterns 345
Jarema, G. & Libben, G. (eds) (2007). The Mental Lexicon: Core Perspectives.
Amsterdam: Elsevier.
Johnson, M & Bayer, S. (1995). Features and agreement in Lambek Categorial
Grammar. Formal Grammar. Proceedings of the Conference of the European
Summer School in Logic, Language and Information, Barcelona, 122–127.
Kamp, H., van Genabith, J. & Reyle, U. (2011). Discourse representation theory. Handbook of Philosophical Logic 15. Dordrecht: Kluwer, 125–394.
Kehler, A. (2011). Cohesion and coherence. C. Maienborn, K. von Heusinger
& P. Portner (eds) Semantics: An International Handbook of Natural
Language Meaning (vol. 2). Berlin: Morton de Gruyter, 1963–1987.
Klingvall, E. (2012). Topics in pseudo-passives. C. Platzack (ed.) Working papers in Scandinavian Syntax 90, 53–80.
Koopman, H. (2000). Prepositions, postpositions, circumpositions, and
particles. Koopman, H. (ed.) The Syntax of Specifiers and Heads. London:
Routledge, 204–260.
Krifka, M. (2001). For a structured meaning account of questions and answers. C. Fery, & W. Sternefeld (eds) Audiatur Vox Sapientia. A Festschrift
for Arnim von Stechow. Berlin: Akademie Verlag, 287–319.
———. (2004). The semantics of questions and the focusation of answers. C.
Lee, M. Gordon & D. Büring (eds) Topic and Focus: A Cross-Linguistic
Perspective. Dordrecht: Kluwer Academic Publishers, 139–151.
Law, P. (2005). Preposition stranding. M. Everaert & H. van Riemsdijk (ed.)
The Blackwell Companion to Syntax, Volume III. Malden, MA: Blackwell,
Levelt, W.J.M. (1989). Speaking: From Intention to Articulation. Cambridge
(Mass.): The MIT Press.
Maling, J. & Zaenen, A. (1985). Preposition-stranding and passive. Nordic
Journal of Linguistics 8:2, 197–209.
Merchant, J. (2001). The Syntax of Silence. Oxford: Oxford University Press.
———. (2004). Fragments and ellipsis. Linguistics and Philosophy, 27:6,
Moortgat, M.J. (2010). Typelogical grammar. E.N. Zalta (ed.) The Stanford
Encyclopedia of Philosophy (Winter 2010 Edition). Stanford.
———. (2011). Categorial type logics. J. van Benthem & A. ter Meulen (eds)
Handbook of Logic and Language. (2nd ed.). Amsterdam: Elsevier, 95–179.
Morrill, G. (2011). Categorial Grammar: Logical Syntax, Semantics, and
Processing. Oxford: Oxford University Press.
346 From Clerks to Corpora
Phillips, C. (2006). The real-time status of island phenomena. Language 82:5,
Reuland, E. (2011). Anaphora and Language Design. Cambridge, MA: The
MIT press.
Riemsdijk, H. van (1990). Functional prepositions. H. Pinkster & I. Genee
(ed.) Unity in Diversity. Dordrecht: Foris, 229–241.
Riemsdijk, H. van. (1998). Head movement and adjacency. Natural Language
and Linguistic Theory 16:5, 633–378.
Stabler, E. (2013). Two models of minimalist, incremental syntactic analysis.
Topics in Cognitive Science 5:3, 611–633.
Svenonius, P. (2006). The emergence of axial parts. Nordlyd, Tromsǿ University
Working Papers in Language and Linguistics, 33:1, 49–77.
———. (2010). Spatial P in English. G. Cinque & L. Rizzi (eds) The Cartography
of syntactic Structures: (vol. 6) Oxford: Oxford University Press, 127–160.
Takami, K. (1992). Preposition Stranding. From Syntactic to Functional
Analyses. Berlin and New York: Mouton de Gruyter.
Tseng, J. (2000). The Representation and Selection of Prepositions. Edinburgh:
Ph.D. Dissertation.
———. (2004). Directionality and the complementation of Dutch prepositions. H. Cuyckens, W. De Mulder & T. Mortelmans (eds) Adpositions of
movement. Belgian Journal of Linguistics 18, 167–194.
Tseng, J. (2005). Prepositions and complement selection. A. Villavicencio &
V. Kordoni )(eds) Proceedings of the 2nd ACL-SIGSEM Workshop on the
linguistic dimensions of prepositions and their use in computational linguistics formalisms and applications. Colchester: Essex University, 11–19.
Truswell, R. (2009). Preposition-stranding, passivisation, and extraction from
adjuncts in Germanic. J. van Craenenbroeck & J. Rooryck (ed.) Linguistic
variation yearbook 8. Amsterdam: John Benjamins, 131–177.
Ursini, F.-A. (2011). On the syntax and semantics of “Ser” and “Estar”. Lingue
& Linguaggio 9:1, 57–87.
———. (2013a). On the syntax and semantics of “Tener” and “Haber”. Lingue
& Linguaggio 11:1, 89–120.
———. (2013b). Esse and Sta: Auxiliary selection in the Aquilan dialect.
Dialectologia 10:1, 107–134.
———. (2013b). Another look at spatial prepositions and the modification
problem. Iberia: An International Journal of Theoretical Linguistics 5:2,
Another Look at Preposition Stranding: English and Swedish Discourse Patterns 347
———. (2013c). On the syntax and semantics of spatial Ps in Spanish. Borealis:
An International Journal about Hispanic Linguistics 2:1, 117–166.
———. (2015a). The morphology of spatial P: Is a unitary perspective possible? G. Boyes et al. (eds) Proceedings of Les Decembrettes 8. Toulouse:
Mirail, 1–20.
———. (2015b). On the syntax and semantics of Italian spatial prepositions.
Acta Linguistica Hungarica 63:1, 3–57.
Ursini, F.-A. & Akagi, N. (2013a). On the distributed morphology and semantics of spatial Ps. I.-J. Lee & U. Dolgormaa (eds) Proceedings of the 15th
Seoul International Conference on Generative Grammar (SICOGG 15).
Seoul: Hankuk University Press, 447–468.
Vermaat, W. (2005). The Logic of Variation: A Cross-linguistic Account of
Wh-question Formation. Doctoral Dissertation, Utrecht.
Ward, G. & Birner, M. (2012). Discourse effects of word order variation. C. Maienborn, K. von Heusinger & P. Portner (eds) Semantics: An
International Handbook of Natural Language Meaning (vol. 2). Berlin:
Mouton de Gruyter, 1934–1963.
Zeevat, H. (2011). Discourse relations. C. Maienborn, K. von Heusinger & P.
Portner (eds) Semantics: An International Handbook of Natural Language
Meaning, Volume 1. Berlin: Mouton de Gruyter, 946–972
19 There is Nothing Like Native Speech:
A Comparison of Native and Very
Advanced Non-Native Speech
Britt Erman & Margareta Lewis
Stockholm University
“I’ve been here for 8 ½ years, my English should be more fluent than
this. Yes … sometimes I really stumble on the words...on the words”
1. Introduction
The above quote shows that finding words can be hard even for someone who has lived and worked in the L2 community for a considerable
time. Vocabulary is an area of L2 acquisition that has received increasing attention in the last couple of decades. The present study is part of
the research program “High-level proficiency in L2 use”1. The program
seeks to provide answers to questions pertaining to what characterizes
the very advanced L2 user, and involves several language departments
at a Swedish university. This study compares vocabulary of different
frequencies in the oral production of two groups of speakers of English,
one non-native Swedish group and one native English-speaking group
as a control. The non-native Swedish group has lived and worked in the
UK (London) for an average of 7.3 years. The main aim of the study is
to establish the rate of high-frequency and low-frequency words in the
spoken data of these two groups. The material is made up of a recorded
semi-structured interview. In order to establish lexical variation the
present study, in contrast to several earlier studies, includes results not
only from frequencies of tokens but also frequencies of types and T/T
1 Thanks for generous grants are due to The Bank of Sweden Tercentenary Foundation.
How to cite this book chapter:
Erman, B. and Lewis, M. 2015. There is Nothing Like Native Speech: A Comparison of
Native and Very Advanced Non-Native Speech. In: Shaw, P., Erman, B., Melchers, G.
and Sundkvist, P. (eds) From Clerks to Corpora: essays on the English language yesterday and today. Pp. 349–366. Stockholm: Stockholm University Press. DOI: http://dx.doi.
org/10.16993/bab.s License: CC-BY.
350 From Clerks to Corpora
ratios, (cf. Lindqvist 2010, Lindqvist et al. 2011; Bardel et al. 2012;
Lindqvist et al. 2013; Forsberg Lundell and Lindqvist 2012). Including
types in the study will give indications regarding variation, which is
assumed to distinguish native from non-native speech. Lindqvist (2010)
found in her study of L2 French that the advanced learners used more
general words to refer to key objects in a video film clip compared to a
native control group.
The interview is one of three tasks carried out with the same participants. The results from two earlier studies, one on vocabulary and one
on multiword structures (formulaic language), both involving two other
tasks, a role play (dialogic) and an online retelling task (monologic),
showed that in the role play the results of the London Swedes (LS) were
like the natives in both studies, whereas the retelling task revealed significant differences between the NS and LS groups (Erman & Lewis 2011;
2013; Erman et al. 2014). Some of the questions asked in the interview concerned the Swedish participants’ knowledge of languages and
in particular their knowledge of English. Questions relating to English
included for example the age at which they started learning English at
school (in Sweden), whether they found speaking English difficult when
they arrived in England, and the extent to which they used English also
at home when in the UK. It is worth noting that all the Swedish speakers
used English at work, and most of them had English-speaking partners at
the time of the recording. Reading through the transcribed interviews it
became apparent that the interviewees had rather varied perceptions of
their knowledge of English, as the extracts below show. However, the
general impression from these extracts is that the interviewees believe
that their English is quite good, some even to the extent that English has
taken over at the expense of their mother tongue, Swedish.
• An easy ride when it comes to languages. Watched English TV a
lot when little. Always speak English with my English partner.
• It’s much more natural to use English when speaking about music.
I just can’t find the Swedish word…
• English was one of my worst subjects in Sweden. Wasn’t good at
English at first (was very shy) but then just started speaking to people.
• I was fluent when arriving in England.
• Sometimes I feel when I go back, I become so conscious about my
Swedish. And obviously I can still speak’s no problem,
but ...
• … sometimes I could have difficulty of swinging back into ...into
fluent Swedish. I mean, when it comes to the more advanced
There is Nothing Like Native Speech 351
Swedish, I think. Because, I think, my Swedish stopped developing
when I was 22 and I came here. And...and here I don’t ...I don’t
associate that much with Swedes.
• English is ...what I realize with English is [after living in France]...
it got a lot more words than French. French is, I think, if you’re
good in French, you use grammar to show that you are educated.
In this last extract there is a hint that English is perceived as having a
large vocabulary.
The aim of the present study is not to establish whether the London
Swedes’ own perceptions of their knowledge of English has a bearing
on the results but to find out how the two groups differ in their use of
vocabulary in this task, more specifically across two main frequency
ranges to be explained below.
We start by accounting for earlier research on vocabulary with a
focus on advanced L2 speakers’ spoken production (2). After presentations of aims (2.1), and material and method (3), we discuss the
notion of frequency in relation to L2 acquisition (4). A description of
the 1–2000 frequency range (4.1) is followed by a display of the results
from this range (4.2), a description of the frequency range beyond 2000
words as this is applied in the present study (4.3), and the results of this
frequency range (4.4). Finally, since the 1–2000 frequency range also
includes high-frequency words typical of spoken discourse, we introduce a selection of sequences involving words from the 1000 most frequent words functioning as pragmatic markers (4.5) and present results
from their distribution across the participant groups (4.6). Apart from
offering some general insights to be drawn from the results, section 5
discusses the main contribution of the study. Section 6 winds up by
presenting some more voices from the London Swedes in light of the
2. Earlier research
Establishing methods that relate vocabulary knowledge to different
proficiency levels in L2 production has in the last few years been a
major concern (Daller et al. 2007; Milton 2007; Tidball and TreffersDaller 2007; Lindqvist 2010; Lindqvist et al. 2011; Bardel et al. 2012;
Lindqvist et al. 2013). One method used is the Lexical Frequency Profile
developed by Laufer and Nation (1995). A basic assumption behind
most studies of vocabulary in relation to frequency is that frequency of
352 From Clerks to Corpora
input will affect output, so that the more frequent a word is the more
likely it is to appear in an L2 speaker’s production (Cobb and Horst
2004; Vermeer 2004). There is also evidence to prove that frequency
plays an important role in L2 acquisition, implying that high-frequency
words are shared by more L2 users than low-frequency words (Tidball
and Treffers-Daller 2007). The higher the percentage of words beyond
the 2000 most frequent words is in an L2 user’s production, the more
advanced is this person’s vocabulary (Laufer 1995).The proportion of
low-frequency words is also commonly referred to as lexical richness in
the literature. The results from studies of lexical richness have shown
that the quantity of lemma tokens of different frequencies distinguishes
not only native from non-native speakers but also L2 speakers at different proficiency levels (Bardel et al. 2012). Some advanced non-native
speakers of L2 French have been shown to reach nativelike levels in
their use of low-frequency lemma tokens. But if some of these were
removed from the list containing many low-frequency words (i.e. the
‘Off-list’; see section 3), such as thematic words occurring in teaching
materials and words that are similar in L1 and L2 (and some others),
no non-native speaker of either L2 French or L2 Italian reached nativelike levels (Bardel et al. 2012).
2.1 Aims and research questions
In the aforementioned studies of multiword structures (MWSs) and
vocabulary with the same participants in two tasks (see Introduction)
it was found that the London Swedes behaved like the natives in the
role play, but differed significantly from the native speakers on both
vocabulary and MWSs in the online retelling of a film clip that was
unfamiliar to them. On the basis of these results it is hypothesized that
the London Swedes, being immersed in an English-speaking community, will come close to the native speakers in the interview, since this
task is connected to a situation that is believed to be familiar, notably
answering questions about themselves. As mentioned, two main frequency ranges are examined: the first two thousand words (1–2000 frequency range, i.e. words of high frequency), and those outside the first
two thousand words (the 2000+ frequency range, i.e. low-frequency
words). Our main aim is to compare the LS group with the NS group
with regard to T/T ratios, and quantity of types and tokens in these two
frequency ranges.
Another aspect closely related to vocabulary is the use of pragmatic
markers, which are assumed to vary with text type (Simon-Vandenbergen
There is Nothing Like Native Speech 353
2000). Based on this it is hypothesized that an interaction involving a
description of self such as in an interview will generate a considerable
number of pragmatic markers. Furthermore, pragmatic markers have
been found to distinguish native from non-native speech (Altenberg
1997; Denke 2009; Fant & Hancock 2014). These facts lead to our
second aim, which is to establish how the LS group compares with the
NS group on a selection of frequent pragmatic markers.
3. Material and method
Table 1 provides some more information about the participants.
The method used involves sorting the transcribed texts into frequency ranges by using the Lexical Frequency Profile (LFP), which
is accessible via LexTutor2. By feeding in the transcribed texts in this
program we get not only different frequency lists (see below) but also
the total number of words, which distributes as follows over the two
groups (Table 2).
Lexical frequency profiles are available in LexTutor via the program
Vocabprofile. In Vocabprofile all the words are registered alphabetically in terms of type and token frequency; this makes the data easily
accessible and allows various kinds of analyses. The words have not
been lemmatized, which means that type frequencies are indicated in
terms of ‘word forms’; for example, museum, museums, and call, calls,
Table 1. Participants.
Time with English
Average age
10 Native speakers
10 London Swedes
9 years at school and an average
of 7.3 years’ residency in London
Table 2. Number of words over the native speakers (NS), and London
Swedes (LS).
2 LexTutor is accessible at:
354 From Clerks to Corpora
called, calling are all registered as six separate types, while representing
two lemmas. The LFP program maps the word forms onto their lemmatized forms (i.e. ‘call’ and ‘museum’ for the six word forms above)
in four categories (or lists): the first most frequent 1000 words, the
second most frequent 1000 words, and the Academic Word List (AWL;
Coxhead 2000). The fourth category is a separate list, called the Offlist, comprising any word (or item see 4.2) outside the 2000 most frequent words and the words in the AWL list.
It should be mentioned that although some types, especially in the
high-frequency 1–1000 list, are inflections of one and the same lemma
as in the examples above the majority belong to different lemmas
(see 4.1). The further we move away on a scale from high-frequency
words towards low-frequency words the more likely it is that type
equates lemma type, and is thus unique (see 4.1).
4. Analysis and results
As mentioned, the results are divided into two main groups, the 1−2000
words frequency range and words beyond 2000, the 2000+ frequency
range. The words (i.e. tokens) in the first 2000 word span constitute the
major part of the present material and cover between 88% and 90%
of the texts (see Table 3 below). These figures are above the average for
written text, which is 80% for the first 2000 words;3 this discrepancy
may be explained by the rather informal character of the text type studied here, and by the fact that the present material constitutes spoken
In the present study the 2000+ frequency range is made up of the
words in the AWL list and a pruned version of the words in the Offlist (see 4.3). The words in AWL make up the smallest proportion of
the words for both groups, covering between 1% and 2% of the texts.
It is common in the literature for calculations only to include number of lemma tokens (Bardel and Gudmundson 2012; Lindqvist et al.
2013; Forsberg, Lundell and Lindqvist 2012), but, as mentioned, in
the present study it was relevant also to include the number of types.
For instance, on some measurements the LS group is nativelike on the
number of tokens, whereas they are non-nativelike on the number of
types, which is an indication that this group recycles their types more
often, implying less diversity. Although her own study only includes
3 See
There is Nothing Like Native Speech 355
lemma tokens, Lindqvist (2010: 415) emphasizes the importance of
also including types in studies on vocabulary.
4.1 Description of the 1-2000 frequency range
The 1–2000 frequency range apparently holds the most frequent content words and among the first thousand words we find many grammatical words needed to ensure structure and coherence, such as
determiners, pronouns, conjunctions, etc. Words of high frequency by
necessity come out in different word forms (i.e. types in LexTutor),
some of which are based on the same lemma. In order to provide more
exact relations between LexTutor (LT) type and lemma type we lemmatized all the LT types to find out the proportions over the frequency
ranges. In the 1–2000 frequency range it was found that the proportions of different lemmas to LT types in the two groups are: NS 79.4%
and LS 79.5%, and for the AWL lists: NS 92.6% and LS 93.0%. At the
other end of the scale are the Off-list words where it was found that
for NS 98.1% and LS 97.6% of the LT types belong to different lemmas. In other words, the vast majority of word forms (LT types) in the
interview belong to different lemmas with average percentages for NS
86.7% and LS 85.8%.
Lexical frequency profiles with their focus on words are obviously
independent of syntax and text type. It is not within the scope of the
present study to evaluate the vocabulary produced, i.e. either to establish whether the words are syntactically, semantically or pragmatically
appropriate, or their functions.
We start by accounting for the T/T ratios, and types and tokens per
hundred words pertaining to the 1–2000 frequency range (Table 3)
followed by a corresponding account of the results from the 2000+
frequency range (Table 4). Finally, we present and discuss results from
searches targeting specific sequences (you know, I think, sort of) −
which are among the 50 most frequent collocations according to Shin
and Nation (2008) − and their distribution over the two groups.
The NS group functions as benchmark, and the threshold for significance is set at p < .05.4
4The chi-square test has been used throughout the study. We wish to thank NilsLennart for drawing our attention to this website:
356 From Clerks to Corpora
Table 3. T/T ratios, types and tokens/100 words in 1–2000 range in the
Interview Type/Token
T/T p
Type/100 p
.000 5.4
4.2 Results for the 1−2000 frequency range
Words belonging in the 1–2000 frequency range cover a large part
of the texts as can be seen in the number of tokens per 100 words
(Table 3). We also observe that the results in Table 3 are all based on LT
results, since lemma types and LT types per 100 words yielded the same
result, both showing that the difference between the NS and LS groups
is highly significant (for lemma types per 100 words p <. 000). For this
reason, types refers to LT types throughout the study.
Our hypothesis that the LS group would be nativelike on measurements pertaining to this task given its everyday character is only partly
supported. While the LS group is nativelike on tokens per hundred
words, they use significantly fewer types compared to the NS group.
This result gives support for the inclusion of types in vocabulary studies. The highly significant difference in T/T ratio in the LS group compared to the NS group indicates that they recycle more words in this
frequency range.
4.3 Description of the 2000+ range
In the present study the 2000+ frequency range is composed of a pruned
version of the words in the Off-list combined with the words in the
AWL list. The LexTutor Off-list is a heterogeneous group of items,
low-frequency words as well as very informal high-frequency words and
voiced pausing. In order to avoid a situation where words, because they
are outside the frequency bands of the first 2000 words, would unduly
be considered advanced or low-frequency, the Off-list was scrutinized
and certain items were removed (cf. Lindqvist 2010; Lindqvist et al.
2013). As a consequence, all the items in the LexTutor Off-list that were
deemed as not being part of a language’s vocabulary, such as voiced
pausing and word fragments, were removed. Indeed, equating the Offlist words with lexical richness can be misleading (Lindqvist 2010: 415).
There is Nothing Like Native Speech 357
The following types of items in the Off-lists have also been removed:
names (of people, regions, places, continents, countries (including languages and nationalities, many of which are similar in Swedish and
English, therefore more readily accessible; cf. Horst and Collins 2006;
Milton 2007; Lindqvist et al. 2013)), feedback words (yea, yeah, ok,
huh, mm), foreign words (cher), contractions (wanna, gonna, gotta,
coz), swear words (fucking), slang words (kids, guys, crap, ass), and
voiced pausing (eh, uh/uhm/um(m)), and, finally, fragments of words
(Thur, archi, etc.).
Table 4 below shows the results for T/T ratios and types and tokens
/100 words in the 2000+ word range.
4.4 Results for the 2000+ range
While the LS group is nativelike on T/T ratios, they significantly differ
from the NS group on types and tokens per 100 words in the 2000+
frequency range (Table 4).
Our hypothesis that the LS group would be nativelike also in the
2000+ frequency range in view of the everyday character of this task
was not confirmed by the results. The number of tokens per 100 words
is significantly lower compared to the NS group, and the difference
between the groups in the number of types per 100 words is highly
significant, the p-value being close to zero. One possible explanation
for this result is that the NSs use more specific vocabulary compared to
the NNSs, which is in line with the results from several earlier studies
(Ovtcharov et al. 2006; Lindqvist 2010; Erman & Lewis 2011).
It is worth noting that a comparison of T/T ratios between the three
tasks targeting the LS and NS groups, i.e. the interview in the present
study and the role play and the retelling task in Erman & Lewis (2013),
shows that the interview is the task that demonstrates the highest T/T
Table 4. T/T ratios, types and tokens/100 words in 2000+ range (incl. AWL)
in the Interview.
Interview Type/Token
Token/100 p
358 From Clerks to Corpora
ratio in this frequency range. This is apparently the task where these
speakers display the most diversity.
Summing up, while the non-natives reached nativelike levels in number of tokens in the 1–2000 frequency range, it is in the number of
types in both frequency ranges that differences between natives and
non-natives become visible. In light of the fact that the 1–2000 frequency range covers between 80% and 90% of all spoken texts, and
to judge by the results of the present study, variation in this frequency
range obviously is a nativelike feature, which distinguishes native and
advanced non-native speakers. It is proposed in the present study that
reaching a nativelike level in types in the first 2000 frequency range
should be included in what is considered advanced vocabulary. In other
words, showing variation among the 2000 most common words should
be a skill worth aiming for also for advanced non-native speakers.
On the basis of the results presented in this study it seems reasonable
to suggest that a contributing factor to divergences shown between the
LS and NS groups is the difference in exposure, which has an effect also
on types of high-frequency words as well as in the range of productive
vocabulary at large.
4.5 Combinations of high-frequency words
LexTutor provides not only statistics, and alphabetical lists of words
item per item in the frequency lists, but also the entire texts with each
word marked for frequency and identifiable in the text. Depending on
the query one can do either a search in the texts proper or in the word
lists. If we are interested in specific combinations of words we apply the
search command to the entire texts.
Since our results show that there are significant differences between
the NS and LS groups in both frequency ranges, a sub-study involving
particular, frequent combinations of high-frequency words, the majority functioning as pragmatic markers, was carried out.
The use of pragmatic markers has been shown to distinguish NN
and N speakers of English (Denke 2009), and very advanced NN and N
speakers of French and Spanish (Hancock & Kirchmeyer 2009; Hancock
2012; Fant & Hancock 2014). English pragmatic markers which have
been shown to be used differently by NN and N speakers include you
know (Denke 2009) and sort of (De Cock 2004). Denke (2009) found
that not only is the pragmatic marker you know significantly more frequent in NS than in NNS speech, but the marker is also used differently,
There is Nothing Like Native Speech 359
the NS speakers using the marker to organize discourse, i.e. as a discourse marker, and the NNS group as an editing marker in connection
with stalling and repair. De Cock (2004) found that pragmatic markers
of vagueness (sort of, kind of) are underrepresented in NNS compared
to NS speech. Another English pragmatic marker, which, along with
you know and sort of, belongs to the 50 most common ‘collocations’
in the 10 million word spoken part of the British National Corpus is I
think (Shin and Nation 2008). This pragmatic marker has been shown
to be overused by NN speakers in both speech (Altenberg 1997) and
writing (Aijmer 2001). The results from the study of these collocations
with a potential function as pragmatic markers will be shown below. It
should be noted that this study is purely quantitative.
4.6 Results for pragmatic markers
The results show that in total figures the NS group has twice the
­number of pragmatic markers compared to the LS group (525 vs. 226).
Numerically the LS group comes the closest to the NS group in their use
of you know. Although the difference is statistically significant (p <.03),
it is close to the threshold (p <.05).
The difference between the LS and NS groups for sort of is highly
significant, the LS group using approximately one sixth (1/6) of the
number used by the NS group. The significantly higher figure for I think
in the LS group confirms results from earlier studies showing that there
is a general tendency for non-natives to overuse this marker in both
speech (Altenberg 1997; de Cock et al. 1998) and writing (Granger
1998; Ringbom 1998; Aijmer 2001). I think is a versatile marker and
can signal a tentative attitude as well as authoritative deliberation
Table 5. Collocations (pragmatic markers) over the NS and LS groups in
the interview.
you know
/100 wds
/100 wds
sort of
I think
360 From Clerks to Corpora
(Simon-Vandenbergen 2000; Aijmer 2001), but, as mentioned, the
present study does not take qualitative aspects of these markers into
account. It is worth noting that the p-values for sort of and I think are
close to zero. This result strongly diverges from the LS results for you
know which in comparison differ marginally from the NS group. One
tentative explanation for the overuse of I think is that there are formally
similar phrases in Swedish (‘jag tycker’, ‘jag tror’, ‘jag tänker’) with
partly overlapping meanings and functions with the English phrase.
The formal similarity and shared semantics between the English phrase
and the three Swedish phrases may thus explain an overuse on the part
of the Swedish L2 English users. This contrasts with the underuse of
sort of which has no formal correspondence in Swedish. Swedish uses
other downtoning items.
In sum, results from earlier studies of sort of being significantly
underrepresented and I think significantly overrepresented in non-­
native compared to native speech have been confirmed in the present
study. Nevertheless, it is worth noting that the significant overuse of
I think in the LS group compared to the NS group does not compensate
for a significant underuse by the LS group of all three pragmatic markers when collapsed compared to NS group (p-value < .000).
5. Conclusion and discussion
As is clear from our results, our hypothesis, that the LS group living
and working in the L2 country would be nativelike on both frequency
ranges studied in view of the fact that the participants are invited to
talk about themselves, was in the main contradicted by the results. In
only two out of six measurements (one for each frequency range) did
the LS group score like the NS group. More specifically, they produced
a nativelike number of tokens per 100 words in the high frequency
range (1–2000), and were nativelike on the T/T ratio in the frequency
range beyond 2000 (2000+). The most interesting result cutting across
the two frequency ranges is that the LS group produced significantly
fewer types compared to the NS group. However, the result for high-frequency tokens (the 1–2000 frequency range) for the LS group is in line
with the general assumption that frequency plays an important role in
L2 acquisition (Tidball and Treffers-Daller 2007).
The most important insight gained from the results of the present
study is that when studying vocabulary it is important to analyze tokens
as well as types, since they may yield divergent results. In other words,
it is with regard to types that there is room for further development for
There is Nothing Like Native Speech 361
L2 users in the high-frequency as well as the low-frequency range. The
results of this study suggest that displaying variation in the first 2000
frequency range is as much a native feature as showing variation in the
beyond 2000 frequency range.
Furthermore, results from many earlier studies suggesting that the use
of pragmatic markers is one area that distinguishes NSs and NNSs are
supported in the present study, notably through significantly fewer occurrences of you know and sort of, and significantly more occurrences of I
think in the LS group compared to the NS group. The quantity as well
as proportion of pragmatic markers is thus what distinguishes the two
groups. It is also worth noting that the significant overuse of I think in the
LS group compared to the NS group does not compensate for a significant
underuse by the LS group of all three pragmatic markers when collapsed
compared to NS group (p-value < .000). One plausible explanation for
this result is that although the pragmatic markers are known by the NN
speakers, they have not become routinized. This would be in accordance
with Bialystok (1993) who sees state of knowledge and control of knowledge as two separate processes. In other words, although the LS group
obviously knows these pragmatic markers and may know when to use
them, they might not have automatic control of them, which in turn can
be explained by constraints related to real-time task performance.
The overall results suggest that native speakers have more immediate
access not only to high-frequency and low-frequency words, but also
to productive vocabulary more generally, including pragmatic markers.
This can only be explained by differences in exposure and degree of
more or less immediate access to items relevant for the situation.
6. Winding up
Against the backdrop of the results let us contemplate some more voices
from the London Swedes regarding their beliefs about their knowledge
of English. According to one of them, British English is difficult because
of the rate at which it is spoken, and it is worth noting that this view
persists after several years in the country.
It’s sometimes difficult to actively participate in the social environment.’s uh...British English is difficult, I think. It’s spoken
very, very fast... very quickly and you really have to listen to
understand. Uh...and sometimes you just don’t understand what
...what they’re talking about. I was taken aback by that.
We observe that understanding rapid speech can be an obstacle even
362 From Clerks to Corpora
at high levels of proficiency. From the introduction we recall the words
of another participant concluding that English comes more naturally
when speaking about certain topics (repeated here):
It’s much more natural to use English when speaking about music. I
just can’t find the Swedish word…
And below is one more extract along similar lines:
And I would say that my English now has come to a point…and…
uh… and at work, some topics at work, I feel more confident in
As a linguist it is easy to agree with these two speakers, linguistics being
one of the many domains dominated by English.
I think for certain things my English is better and for other things
my Italian. For the job English is far better, but if I’m talking, I think,
emotionally or generally, then I would be more comfortable in Italian
than...than in English, I think.
Some acknowledge that English is difficult, but also that practice helps,
as in this quote from the introduction, repeated here.
English was one of my worst subjects in Sweden. Wasn’t good at
English at first (was very shy) but then just started speaking to people.
This view is shared by another speaker who, like the former speaker,
eventually realized that participating in conversations is essential in
everyday life.
I always struggled with languages. That was never my strong subject
in school. I’m a physicist. I’m a mathematician and... and I can’t
learn anything by heart. I need to have, you know, I need to understand why it is this way. But I realized that if I don’t say anything,
this is gonna be really, really boring and a bit useless so .. and I just
kind of started speaking.
A couple of speakers comment on their Swedish accent when speaking English. In the second extract the speaker apparently considers her
Swedish accent part of her identity.
When I speak Swedish, it sounds like I’m singing. But when I speak
English, I think my voice sounds really...uh ... monotone, do you see
There is Nothing Like Native Speech 363
that? I heard some people say they thought I was Irish which was for
me very, very strange.
I have a Swedish accent but, yeah, and I don’t think I’m trying to get
away from that.
Whether or not you have an accent is of no importance according to
another speaker.
In England no one cares if you have got an accent.
Finally, two of the ten Swedes comment on English vocabulary and the
limitations they often sense when speaking the language, which provides a clear link to the results of the present study. In fact, these quotes
neatly summarize the overall results from the present study.
Actually, I was surprised. I often find myself using English expressions but with Swedish words. And that’s also funny because if you
migrate away from what you’re used to, you casually speak to someone about something else, you realize how poor your vocabulary
actually is.
I’ve been here for 8 ½ years, my English should be more fluent than
this. Yes, sometimes I really stumble on the words...on the words.
(We recall this quote from the beginning of the article.)
Of the three vocabulary tasks administered to these two groups (the
interview in the present study, a role play and a retelling task (Erman
& Lewis 2013), the interview was the task in which the London Swedes
were the furthest away from the native group. Furthermore, this was
the task in which the natives showed the most diversity in regard to
low-frequency as well as high-frequency words. The fact that the native
speakers distinguish themselves from the non-­native speakers by having
significantly higher numbers of types in both frequency ranges in this
task may be explained by the interviewee being able to talk freely about
anything that comes to mind in answering the questions asked by the
interviewer. All in all, the results have shown that the native speakers
had more immediate access to words across the board. Indeed, the most
important insight gained through this study is that it is frequent word
types, i.e. those within the 1–2000 range, that require practice in order
to approach the quantity of native speakers. Infrequent words are presumably less important for general communication. Finally, the results of
this study should encourage more research involving word types as well
as tokens, and on larger corpora of different types of spoken production.
364 From Clerks to Corpora
Aijmer, K. (ed.). (2001). A Wealth of English: Studies in Honour of
Göran Kjellmer. Gothenburg Studies in English 81. Acta Universitatis
Gothoburgensis, Gothenburg University.
———. (2001). I think as a marker of discourse style in argumentative ­student
writing. K. Aijmer (ed.), 247–257.
Altenberg, B. (1997). Exploring the Swedish component of the International
Corpus of Learner English. B. Lewandowska-Tomaszcyk & P. J. Melia (eds)
Proceedings of PALC’97: Practical Applications in Language Corpora.
Lódz: Lódz University Press, 119–132.
Bardel, C., Gudmundson, A. & Lindqvist, C. (2012). Aspects of lexical sophistication in advanced learners’ oral production: vocabulary acquisition
and use in L2 French and Italian. N. Abrahamsson & K. Hyltenstam (eds)
High-level L2 Acquisition, Learning and Use. Thematic issue of Studies in
Second Language Acquisition, 34:2, 269–290.
Bialystok, E. (1993). Symbolic representation and attentional control in
pragmatic competence. G. Kasper & S. Blum-Kulka (eds) Interlanguage
Pragmatics. Oxford: Oxford University Press, 43–59.
Cobb, H., & Horst M. (2004). Is there room for an academic word list in
French? P. Bogaards, & B. Laufer (eds) Vocabulary in a Second Language:
Selection, Acquisition, and Testing. Amsterdam: Benjamins, 15–38.
Coxhead, A. (2000). A new academic word list. TESOL Quarterly, 34:2,
Daller, H., van Hout, R. & Treffers-Daller, J. (2003). Lexical richness in spontaneous speech of bilinguals. Applied Linguistics, 24:2, 197–222.
Daller, H., Milton, J., & Treffers-Daller, J. (eds) (2007). Modelling and Assessing
Vocabulary Knowledge. Cambridge: Cambridge University Press.
De Cock, S., Granger, S., Leech, G. & McEnery, T. (1998). An automated approach to the phrasicon of EFL learners. S. Granger (ed.) Learner English
on Computer. London & New York: Longman, 67–79.
De Cock, S. (2004). Preferred sequences of words in NS and NNS speech.
Belgian Journal of English language and literatures (BELL) New Series, 2,
Denke, A. (2009). Nativelike Performance: A Corpus Study of Pragmatic
Markers, Repair and Repetition in Native and Non-native English Speech.
Saarbrücken: VDM Verlag.
Erman, B. & Lewis, M. (2011). Multiword structures in the speech of non-native
There is Nothing Like Native Speech 365
and native speakers of English. Paper presented at EUROSLA 21, 21st annual
conference of the European second language association, 8–10 September.
Erman, B. & Lewis, M. (2013). Vocabulary in advanced L2 English speech. N.L. Johannesson, G. Melchers & B. Björkman (eds) Of Butterflies and Birds,
Dialects and Genres. Stockholm Studies in English 104. Acta Universitatis
Stockholmiensis, 93–108.
Erman, B., Denke, A., Fant, L. & Forsberg Lundell, F. (2014). Nativelike
­expression in the speech of long-residency L2 users: A study of multiword
structures in L2 English, French and Spanish. International Journal of
Applied Linguistics 24, doi: 10.1111/ijal.12061 2014.
Fant, L. & Hancock, V. (2014). Marqueurs discursifs connectifs chez des
locuteurs de L2 très avancés: le cas de alors et donc en français et de
entonces en espagnol. M. Borreguero Zuloaga & S. Gómez-Jordana Ferary
(eds) Marqueurs du discours dans les langues romanes: une approche contrastive. Limoges: Lambert Lucas, 317–335.
Forsberg Lundell, F. & Lindqvist, C. (2012). Vocabulary development in advanced L2 French: do formulaic sequences and lexical richness develop at
the same rate? Language, Interaction, Acquisition (LIA), 3:1, 73–92.
Granger, S. (1998). Prefabricated patterns in advanced EFL writing: collocations and formulae. A.P. Cowie (ed.) Phraseology, Theory, Analysis and
Applications, 145–160.
Hancock, V. & Kirchmeyer, N. (2009). Étude du marqueur polyfonctionnel
vraiment. L’information grammaticale, 120, 14–23.
Hancock, V. (2012). Pragmatic use of temporal adverbs in L1 and L2 French:
Functions and syntactic positions of textual markers in a spoken corpus.
C. Lindqvist & C. Bardel (eds) The Acquisition of French as a Second
Language: New Developmental Perspectives. Special issue of Language,
Interaction and Acquisition, 3:1, 29–51.
Laufer, B. (1995). Beyond 2000: A measure of productive lexicon in a second
language. L. Eubank, L. Selinker, & M. Sharwood Smith (eds) The Current
State of Interlanguage: Studies in Honor of William E. Rutherford.
Amsterdam: Benjamins, 265–272.
Laufer, B. & Nation, P. (1995). Vocabulary size and use: Lexical richness in L2
written production. Applied Linguistics, 16, 307–322.
Lindqvist, C. (2010). La richesse lexicale dans la production orale de l’apprenant avancé de français. La revue Canadienne des Langues Vivantes/The
Canadian Modern Language Review, 66:3, 393–420.
Lindqvist, C., Bardel, C., & Gudmundson, A. (2011). Lexical richness in the
366 From Clerks to Corpora
advanced learner’s oral production of French and Italian L2. International
Review of Applied Linguistics (IRAL), 49, 221–240.
Lindqvist, C., Gudmundson, A., & Bardel, C. (2013). A new approach to measuring lexical sophistication in L2 oral production. Eurosla Monographs
Series, 2, 109–126.
McCarthy, P. M., & Jarvis, S. (2007). Vocd: A theoretical and empirical evaluation. Language Testing, 24:4, 459–488.
Milton, J. (2007). Lexical profiles, learning styles and the construct validity of
lexical size tests. H. Daller, J. Milton, & J. Treffers-Daller (eds) Modelling
and Assessing Vocabulary Knowledge. Cambridge: Cambridge University
Press, 133–149.
Ringbom, H. (1998). Vocabulary frequencies in advanced learner English: a
cross-linguistic approach. S. Granger (ed.) Learner English on Computer.
London & New York: Longman.
Shin, D. & Nation, P. (2008). Beyond single words: The most frequent collocations in spoken English. ELT Journal, 62:4, 339–348.
Simon-Vandenbergen, A.-M. (2000). The functions of I think in political discourse. International Journal of Applied Linguistics, 10:1, 41–63.
Tidball, F., & Treffers-Daller, J. (2007). Exploring measures of vocabulary richness in semi-spontaneous French speech. H. Daller, J. Milton, & J. TreffersDaller (eds) Modelling and Assessing Vocabulary Knowledge. Cambridge:
Cambridge University Press, 133–149.
Vermeer, A. (2004). The relation between lexical richness and vocabulary size
in Dutch L1 and L2 children. P. Bogaards, & B. Laufer (eds) Vocabulary
in a Second Language: Selection, Acquisition, and Testing. Absterdam:
Benjamins, 15–38.
20 “Bachelor Means Nothing Without
Husband and Father”1 : What Collocations
Reveal about a Cognitive Category
Christina Alm-Arvius (1945–2013)
Stockholm University. Based on an unfinished manuscript, posthumously
edited by Cecilia Ovesdotter Alm.
1. Introduction
This study combines recent insights into the interpretation of language
in cognitive linguistics with the examination of syntagmatic patterns,
or collocations, in English language use. The focus is on collocations
involving the English lexeme bachelor. To further our science, the aim
must be to try to integrate information from different theoretical perspectives to get as comprehensive and profound a view as possible
of semantic questions (cf. Alm-Arvius 1999). It is hardly possible to
discuss meaning in natural human languages without making use of
such basic notions as reference, denotation, sense, and syntagmatic and
paradigmatic sense relations. Significantly enough, different versions of
the cognitive semantic school have in many respects developed through
critical reactions to these earlier semantic paradigms.
The sense or senses of the lexeme bachelor have been described in
different ways, and often the sense descriptions given are straightforwardly connected with the theoretical standpoint of the linguist(s)
producing them regarding semantic questions in general. In particular
we can note that cognitive semantics has been critical of what it calls
classical categorization. In short, it insists that classical categorization
overlooks certain crucial observations concerning how speakers of
English really understand and use, for instance, the noun bachelor. In
this study I shall look into the semantics of this English lexeme, and
1 From the British National Corpus.
How to cite this book chapter:
Alm-Arvius, C. 2015. “Bachelor Means Nothing Without Husband and Father”: What
Collocations Reveal about a Cognitive Category. In: Shaw, P., Erman, B., Melchers, G.
and Sundkvist, P. (eds) From Clerks to Corpora: essays on the English language yesterday and today. Pp. 367–386. Stockholm: Stockholm University Press. DOI: http://dx.doi.
org/10.16993/bab.t License: CC-BY.
368 From Clerks to Corpora
use it as a touchstone as well as a steppingstone for gaining insight
into general semantic questions and the contributions that different
types of theories can be expected to give to linguistic semantics in
general. The empirical material consists of bachelor occurrences in
The Times and The Sunday Times 1995 and examples in the British
National Corpus. Below, T and ST are abbreviations of The Times
and The Sunday Times respectively, and BNC is short for the British
National Corpus.2
The discussion will consider certain explanatory models within the
cognitive linguistic paradigm, examining their usefulness and descriptive validity when trying to explain the meaning(s) conveyed by bachelor.
Cognitive science is the result of work in a number of related disciplines, including linguistics and philosophy, but in many ways it has been
particularly influenced by findings in cognitive psychology and, more
recently, neurobiology (see e.g. Lakoff 1987; Lakoff & Johnson 1999).
More specifically, the concept of prototype is a central one in cognitive
semantics. It was introduced by Eleanor Rosch and her co-workers in
the 1970s in their studies into the nature of ordinary human categorization (Rosch 1978,1977,1975; Rosch & Mervis 1975).
Fillmore’s notion of frames has proved quite influential in cognitively oriented research on language meaning(s). It is somewhat loose
and open to partly variant understandings. Sometimes it must be taken
to stand for a fairly specific type of situation, not unlike the propositional structures outlined by a predicator and its argument(s) in predicate logic. However, there are also said to be complex frames, a more
comprehensive scenario supposed to constitute a common ground for
the more specific conceptual figures represented by individual words.
This is reminiscent of lexical fields or sense relations in structuralist
semantics, but according to Fillmore and other adherents of the frame
theory it is the conceptual frames which are basic, and words are only
understood and related to each other via their grounding in such experiential schematizations (Fillmore 1985).
A fundamental fact about lexical words in English is that they
are used together in linear sequences according to more general and
2 Even if the bachelor occurrences in these British texts cannot be taken to exemplify
the average way of using bachelor in English at large, they nonetheless show how
native speakers of the language can use the word for different kinds of communicative purposes.
“Bachelor Means Nothing Without Husband and Father” 369
recursive syntactic patterns. Lexical co-occurrence is however restricted
not only by the syntactic properties of words but also by more specific
lexico-semantic selectional tendencies. Some syntagmatic connections
are quite fixed or “frozen” constructions (idioms) whose meanings are
sometimes idiosyncratic and not calculable from the individual words
used in them. On the whole, it seems however more common that
lexical items occur together in more variable combinations or collocations. These are also idiomatic or language specific in a great many
cases and often constitute difficulties both for language learners and
in translation work. Prototype analyses of lexical senses have not paid
much attention to the collocational behaviour of individual lexemes
like bachelor.
2. The lexicon and the grammar
Examining bachelor allows exploration of how cognition, in a wider
sense, and structural conventions in the English language are important for how it can be used and understood in syntagmatic sequences.
The relationship between the lexicon and the grammar of a human
language is a central question in linguistics. It concerns the character
of lexemes, word formation, collocations, and syntactic structures, and
their relationships.
Arguably, syntagmatic sense relations are more basic than paradigmatic or substitutional ones; language is usually realized in compositional linear strings moulded on recurring, and thus conventionalized, constructional patterns. The collocations of lexical items within
syntactic structures are in principle variable, even if they are also
restricted. In fact, the intuitions among proficient speakers concerning
lexical co-occurrence potentials are often intricate, as they involve,
for instance, the possibility for creative figurative extensions of lexical senses. As a result, it has in many cases proved difficult to unravel
them analytically and describe them in a satisfactory way (cf. Cruse
2000: 229–234).
The combination of a premodifier and a nominal head, as in eligible bachelor, poor bachelor and wealthy spinster, is a two-lexeme
collocation and part of a noun phrase, but lexical words – or rather
specific morphosyntactic forms of lexical words – are of course usually strung together in longer and more complex syntagms which
often also include function words. An example of this would be the
370 From Clerks to Corpora
combination of words in the main declarative clause A poor bachelor
wants to marry a wealthy spinster.
As I exemplify below, bachelor can either be treated as a kind of
civil legal term or as a general language word. In the first function it
has a well-defined sense distinguishing it from other terms within the
terminological field dealing with marriage or being single like husband,
wife, spinster, and divorcee. Its use in the general language overlaps
ordinarily with this technical application to a considerable extent, but
it is more variable and open to changes as to what associations are
foregrounded on a particular occasion of use.
3. Collocations: co-occurrence potentials of words
in syntactic strings
The term collocation was introduced by J R Firth (1957: 194–196). It
concerns the semantic significance of the co-occurrence of words within
language sequences. Firth does not provide a more exact description of
this concept, however.
“Meaning by collocation is an abstraction at the syntagmatic level
and is not directly concerned with the conceptual or idea approach
to the meaning of words. One of the meanings of night is its collocability with dark, and of dark, of course, collocation with night.
This kind of mutuality may be paralleled in most languages and has
resulted in similarities of poetic diction in literatures sharing common classical sources.” (1957: 196)
In the passage quoted above, which occurs towards the end of his brief
outline of the notion of collocation, Firth appears, firstly, to distinguish
between the conceptual meaning of a word and its collocability. The
former is presumably the meaning it has also when it is uttered quite on
its own, without the company of other words in a composite syntactic
string. A word’s collocability with another word is said to be a separable kind of meaning, or perhaps we should rather take it to be another
aspect of its meaning. At any rate, Firth can be taken to claim that the
“Bachelor Means Nothing Without Husband and Father” 371
collocability of a word is not directly reflected in, or does not straightforwardly follow from, its conceptual content.3
Quite generally speaking, collocability seems to be connected with
semantic compatibility. More specifically, empirical observations of collocates in syntactic strings suggest that they can occur together
i) because they represent conceptual contents that intersect to a
certain degree
ii) and often also because of established habits within the speech
In other words, words’ collocational behaviour is part of the idiomaticity of the language, and this would appear to account for the perceived
arbitrariness of certain collocational preferences and restrictions – also
in factually oriented descriptions. Why are for instance both a high
building and a tall building possible – and synonymous – while we only
talk of a tall man?
3.1 Collocations involving bachelor
We can note that the extent to which the sense of a possible collocate
is integrated into the sense of a lexeme apparently varies. A potential
collocate is not reflected so directly in the general understanding of
a word, but all the same it can be said to specify the character of a
property that will be found in the members of a given sense category. Be of a certain age is an inalienable characteristic of material
phenomena, including living creatures like human beings. Since the
3Cruse (2000: 221f) suggests two types of co-occurrence preferences in language:
selectional preferences which inescapably follow from the propositional content of
a string, and collocational preferences which do not have the same kind of inescapable logico-factual basis, and that violation of selectional preferences is more serious.
It will result in a paradox or even outright incongruity. Instead I would suggest that
the terms collocability or collocational restrictions and preferences are used in a
broad sense about syntagmatic co-occurrence in general between the words in various actual or conceivable phrases, clauses, and sentences within a given language
system. In English we can, for instance, both ride a horse and ride a bicycle/motorcycle, but the Swedish cognate verb rida is only used in the translation of the first
predication: rida en häst, while another verb must be used in collocation with cykel
and motorcykel, the translation equivalents of bicycle and motorcycle: åka cykel/
motorcykel – in spite of the obvious etymological relation between these Swedish
and English words. Similarly, drivers or passengers can ride in a vehicle like a bus/a
cab/a car/a limousine/a taxi in English, but rida would again not be possible in the
corresponding Swedish expressions: åka bil/buss/limousin/taxi.
372 From Clerks to Corpora
primary sense of bachelor denotes unmarried adult human males, the
adjectives old and young are both natural collocates: an old/a young
bachelor. Similarly, we can talk about, for instance, an aged or ageing
bachelor, a forty-year-old bachelor, or a bachelor in his early thirties
(cf. Cruse 1986: 214ff). Indeed, bachelor is comparatively frequently
found in collocation with words that say something about how long a
man has been a bachelor. This is connected with the fact that a man’s
bachelor days are over as soon as he marries. Comparing the antonyms bachelor and husband, the sense relation between them tends
to involve a temporal aspect: a man is typically a bachelor before he
becomes a husband, although some remain bachelors all their lives.
Consequently, it is hardly surprising that the sentential formulation
X remained a bachelor is often seen in obituaries, sometimes with
minor variations. Instances of this kind of formulaic sentence occur
in The Times and The Sunday. Still a bachelor is another idiomatic
expression that is linked with the temporal aspect of bachelor status in
these newspaper texts. They also contain instances of the collocation
lifelong bachelor.
(1) He remained a bachelor. (T11/10,19), (T1/9,19), (T31/8,17)
(2) Tubby Broomhall remained a bachelor all his life. (T17/1,19)
(3) Having remained a bachelor until he was well into his seventies, Cramrose married in 1986 Princess Joan Aly Khan.
(4) Still a bachelor, he is content with doting on his nephews in
Scotland and has little interest in having children of his own.
(5) Anyone who’s been a bachelor for as long as me, and there
are very few of those around who aren’t gay, can knock up
the odd this and that. (ST22/10,9/15)
To sum up, observations indicate that the collocational potential of a
lexical item typically reflects both the conceptual intersection between the
senses of collocates and established idiomatic practices in the language in
question. This means that collocations tend to be the result of language
specific combinatory relations between word senses, idiosyncratic conventions within a particular speech community. When translated into other
languages, they must commonly be translated as wholes and be calculated
in a more global way, involving idiomatic constructional patterns and
“Bachelor Means Nothing Without Husband and Father” 373
stylistic norms by language (for instance, consider the common English
collocation eligible bachelor vs. a Swedish translation equivalent).4
It is clear that collocations cannot generally be explained by referring
to factual or general conceptual knowledge. From a strictly informative
point of view marriageable bachelor would appear an appropriate collocation, but it is not a normal collocational choice in English. Eligible
bachelor is, on the other hand, an idiomatic collocation conveying the
same sort of meaning. Collocations are thus commonly habitual or idiomatic combinations of words, and there is no strict distinction between
what should be considered a compound or a multi-word expression/
idiom and an idiomatic collocation.
Furthermore, regarding collocation and syntactic structure, lexical
words are necessarily integrated in the grammar of a language and
predisposed to function in particular syntactic slots. As a result, it is
arguable that co-occurrence relations between words – be they lexical
or grammatical – are limited to items which are syntactically related in
a string.
3.2 Types of collocational connections
It is possible – or even necessary in some cases – to distinguish between
different types of collocational connections, depending on what kind
of syntactic relation exists between a ‘collocational focus’ and its collocate(s). In a noun phrase the head appears to be the collocational focus,
while the modifiers are dependent collocates, adding specifying information about it. Similarly, the subject will be the collocational focus in
a clause in relation to predicative items like a verb predicator or a subject complement, and a transitive verb will be collocationally dependent on both its subject and its object(s). In other words, a collocational
4Strings that deal with hypothetical situations – or combinations or blends of situational scenarios that are at least partly hypothetical – as well as strings which
simply negate or question a description of the real world may contain lexical collocations that would appear anomalous in an affirmative assertion describing some
event or circumstance in the real world (cf. Alm-Arvius 1993: 26–28; Fauconnier
1997: 14–18, 99ff, 93f, 156ff).
(1) (2) (3) (4) *Bachelors are married/female.
If bachelors were married/female …
Are bachelors married/female?
Bachelors are not married/female.
374 From Clerks to Corpora
focus typically has a more independent and weightier denotational or
referential status compared to its collocate(s). It tends to be the referential hub in such a combination, a basic link with the extralinguistic
thing described. As a result, its meaning will be more directly influenced by the things referred to, and the reading of a collocate will in
its turn be adjusted to the understanding of the focus. Accordingly, the
distinction between a collocational focus and its collocate(s) is important for the direction of ‘collocational tailoring’, as the interpretation
of a dependent collocate is commonly adjusted to fit the character of a
collocational focus rather than the other way round.
Interestingly enough, the direction of collocational selection seems
largely the opposite. A modifier, for instance, links on to or selects a
phrasal head because, as it were, the contents of this head provide it
with an adequate carrier of the characteristic that it represents. In the
collocation old bachelor the noun head bachelor will stand for a person
that exhibits the quality described by old.
4. Regular vs. occasional features of bachelor
4.1 Regularly incorporated aspects of the meaning of bachelor
So-called analytic sentences tell us something about our experience of
the world and how cognitive categories are represented. Analytic sentences contain a specific, more unusual type of collocation, as they spell
out sense relations between lexical units which are regularly incorporated within the sense of the semantically defined subject constituent, a
bachelor, such as a bachelor being human, male, adult, and unmarried. If
we make the sentences synthetic in (6) through (9) by replacing an initial
indefinite article with the definite article or a demonstrative pronoun,
these strings appear odd – it is difficult to think of a communicative
situation in which they occur. There is usually no need explicitly to add
meaning aspects to an instance of a lexical item by means of a collocate
if these qualities are already regularly incorporated in its sense – unless
one wants to explain the sense in question by using an analytic sentence.
(6) ?The/This/That bachelor is (a) human
(7) ?The/This/That bachelor is (a) male.
(8) ?The/This/That bachelor is (an) adult.
(9) ?The/This/That bachelor is unmarried.
Admittedly, it is often difficult to decide whether a general sentential
description of the contents of word – or part of it – is strictly speaking
“Bachelor Means Nothing Without Husband and Father” 375
an analytic sentence or whether it is instead largely synthetic. Clearly,
the difficulty to draw an absolute boundary between analytic and synthetic sentences is directly related to the often cumbersome distinction
between what can be said to be strictly entailed by a propositional
statement and meaning features that seem merely to be regularly presupposed or even just commonly expected in the use of certain lexical
and grammatical constructs.
Arguably, the following two statements, which essentially express
the same claim, are not unconditional analytic sentences, because even
if they would be true of most men who can be included in the primary
sense of bachelor category, they need not be true of all of them. In
other words, being a bachelor does not entail living in circumstances
that make it possible to enter into the state of matrimony. Instead this
quality is just expected in most cases, although it need not be part of the
characterization of each and every bachelor.
(10) A bachelor can marry.
(11) If a man is a bachelor, he is free to marry.
Such observations directly support a prototype-centred analysis of lexical senses. Indeed, individual bachelors necessarily exhibit a host of
characteristics which are not regularly or even commonly shared by all
the members of this category. This is directly relevant for the impression
that this sense category – like most others – does not have strict conceptual boundaries. Bachelors come in many different shapes or forms,
as it were. They are men that are old enough to marry, but each of
them also has a great many other qualities, and they all participate in a
considerable range of social scenarios, although some of them are more
associated with the status of bachelorhood than others. These unavoidable experiential facts cannot be disregarded when we consider and try
to describe how users of English understand this term within or even
out of specific language contexts.
Cognitive semantics allows us to consider also such merely occasional characteristics of, for instance, bachelors. By comparison, earlier – or classical – types of sense analyses normally aimed at identifying
a skeletal and finite set of supposedly necessary or at least criterial features, but this proved difficult in many cases for a number of reasons.
Our general experience of what can be included in a category like bachelor is many-sided and even partly variable, and our cognitive grasp of
word senses also allows imaginative, unpredictable variations, many of
which are logically and factually impossible. All the same, they occur
376 From Clerks to Corpora
and provide food for speculative thought as well as for figurative extensions of sense categories (cf. Alm-Arvius 1999). In addition, language
habits sometimes impose arbitrary-like preferences or even restrictions
on the use of lexical items or longer expressions, for instance as regards
what words or word forms can naturally collocate in a given language
(cf. Cruse 1986: 281f).
4.2 Occasional collocates of bachelor
Examining the semantic relation between a lexical item like bachelor and its lexical collocates or collocational foci, we can first observe
that generally recognized analytic sentences, used to explain the primary sense of bachelor, are exceptional in containing collocates that
only stand for senses that are just regularly incorporated parts of this
other, typically more specific, lexical sense – or cognitive category, if we
instead use the terminology of cognitive semantics. But if the purpose
of the utterance is not to explain this particular word sense, such a
collocation will appear tautological or pleonastic (cf. Cruse 2000: 45,
223f, 227–229). Importantly enough, this appears to be true of constructions that contain the entailed predicative quality unmarried as
well as those that explicitly attribute the regularly presupposed features
human, male, and adult to the referent of a particular bachelor instance.
The following two examples show that male and unmarried have
been used as premodifying collocates of bachelor also in non-analytic
constructions. These two premodifiers might have been added to these
particular examples of bachelor functioning as noun phrase heads as
semantic emphasis markers.
(12) How, then, can this zeitgeisty little lifestyle development
for the heterosexual male bachelor be on the way out, …
(13) … a revenge drama about an unmarried bachelor who
likes to razor off people’s ears to avenge his deaf brother.
However, normally the collocates of a lexical item with a particular
sense like bachelor add merely occasional features to it. This is in fact
why they are informative. The premodifiers posh, young, kindly, old,
taciturn, painfully shy, American, millionaire5, and middle-aged, as well
5 American millionaire in (18) could well be regarded as one composite modifier consisting of a phrase in which millionaire is the head and American is a premodifier.
“Bachelor Means Nothing Without Husband and Father” 377
as the postmodifiers in his fifties and farming alone … in the examples
below are all possible but by no means regular features of members of
the bachelor category. In other words, these senses are compatible with
the primary sense of bachelor, but not regularly incorporated in it.
(14) … a posh young bachelor … (T31/3,2)
(15) … the kindly old bachelor … (ST26/11,7/7)
(16) Mr Cassidy, a taciturn bachelor, … (T6/11,4)
(17) … the duke, a painfully shy bachelor, … (T24/7,15)
(18) … , American millionaire bachelor, … (BNC,AP7(762))
(19) … a bachelor in his fifties … (BNC,AFC(1644))
(20) … a middle-aged bachelor farming alone after his mother
dies, … (BNC,A36(270))
The observation that the dependent collocates of a particular lexical
word typically add occasional but by no means regular features to
its collocational focus helps us to understand the character of lexical
senses or categories. It shows us that they are semantically flexible, as
they have the potential to take on certain additional meaning aspects
temporarily, in specific language sequences and as a result of the actual
or would-be referents or reference situations described by particular
uses of a lexical sense. Indeed this observation agrees well with prototype analysis of lexical meanings.
Even quite common collocates of, for instance, bachelor which
occur in synthetic sentences rather than in analytic ones – like eligible,
confirmed, or phrases and words that say something about the age of
a bachelor – represent additional, non-core features of their collocational focus. Common collocations like eligible bachelor and confirmed
bachelor ought to have some kind of lexical status, even if they cannot
be considered compounds (cf. ALD 1989: 72; Longman 1987: 63). This
kind of linguistic mental imprint must, however, be quite intricate and
involve possible variations in the realizations of a collocational pattern.
When used in actual syntactic structures in utterances, these collocations must accordingly be realized as parts of whole noun phrases.
Eligible in eligible bachelor is then often, but not necessarily, found in
its superlative form, most eligible, with a preceding geographical term
in the genitive case. In addition, such a noun phrase appears often to be
the complement of the preposition of in a partitive construction: one of
X’s most eligible bachelors.
The regular features of a lexical sense constitute its core, while the
occasional features added by dependent collocates are merely various
378 From Clerks to Corpora
kinds of conceivable but not generally incorporated sense attributes. In
other words, in the case of bachelor collocational meaning additions
will be semantic aspects that are quite compatible with being a bachelor,
even if a man can also very well be a bachelor without being associated
with them either just transiently or more permanently. Most importantly,
it is clear that in addition to shared, regular features of this category (see
above), every bachelor must also have a host of other, either more stable or just incidental, or even just potential characteristics. Speakers of
English are naturally aware of this, and it is debatable whether they ever
think of bachelors as simply having the skeletal set of attributes that I
have called regular, incorporated sense features. Instead it seems likely
that their conceptions of bachelors tend to be both richer and somewhat
variable through the inclusion of merely occasional sense attributes.
More specifically, occasional attributes with bachelor understandings in actual language use can clearly be placed along a continuum
from usually expected ones over those that are just possible but not
ordinarily expected to more unique, individual qualities of particular
bachelors like farming alone … in (20) (cf. Cruse 1986).
Continuum of occasional features of bachelor:
expected -------- possible -------- individual
It seems somewhat problematic to try to sketch a general prototype of
the primary sense of bachelor without considering special types of circumstances or language contexts, but it appears reasonable to suggest
that the regular and typically expected features of bachelors together
form a loose kind of prototypical conception of the sense category we
speak of as bachelor. In particular, it should be noted that this explanatory model of the semantics of this lexical item makes a distinction
between regularly presupposed and usually expected qualities. Only the
former will be felt to be criterial enough to be regularly incorporated in
the primary sense of bachelor.
regular features: entailed – presupposed
occasional features: expected -------- possible -------- individual
Furthermore, it seems as though it would normally be wrong to consider
occasional collocates accidental or selectionally arbitrary, since apparently they can co-occur with, or select, the collocational focus because
the two somehow share a semantic quality. This semantic affinity need
not be the result of the actual character of the kind of phenomena represented by a collocational focus. It could just be a cultural and cognitive
“Bachelor Means Nothing Without Husband and Father” 379
construct, and thus an indication of how speakers of a given language,
or a variety of it, conceive of something part of their life experience.
Actually, also regular sense features can depend on cultural conceptions
and conventions. This applies to the fourth core feature regardless of
whether it can be paraphrased as (a man who) has never been married
or just as (an) unmarried (man). The institution of marriage is a cultural
construct, and the sense of bachelor is directly dependent on it.
We have now discussed how the meaning of a collocational focus,
bachelor, can be made more specific by in particular modifying
­collocates. However, bachelor appears also to be fairly frequently used
itself as a dependent collocate, that is as an adjectivalized premodifier or
as a predicative complement. In turning to such bachelor occurrences,
it is relevant to recall that dependent collocates are often semantically
tailored, that is adjusted, to the meaning of their collocational focus.
In other words, the reading of a collocate can also be influenced by a
collocational relationship, or sometimes perhaps mainly as a result of
the extralinguistic character of the thing(s) described.
All the “bachelor plus noun” combinations exemplified below are
used in my empirical material, and especially bachelor pad seems so
common that it may be appropriate to consider it a compound noun
rather than a collocation.6
(21) The first time Mathias and Christa make love is in a cheesy
bachelor pad, complete with mirrored ceiling. (ST5/2,7/12)
6 Bachelor appears also to be quite frequently used as the head of a noun phrase functioning as a subject complement. I would suggest that such an instance is typically
not as closely integrated semantically with the subject element it adds information
about as many premodifier instances of bachelor and their head nouns. A premodifier and its head are parts of the same noun phrase and will function together
also semantically within such referring or predicative expressions. By comparison,
a subject and its subject complement do not belong to the same syntactic phrase,
just to the same clause. A complement makes up a phrase of its own, which is furthermore syntactically more directly related to the copular predicate verb than to
the subject, as it is first tied to this verb through its complement status. It is then
syntactically related to the subject through it, as an element of the predicate constituent. The syntactic relationship between a dependent collocate and its collocational
focus appears to be relevant for how much the latter can dominate or mould the
reading of the former.
(5)Charles had been a bachelor for thirty-two years. (BNC,A7H(445))
(6)Gary Bond remained a bachelor. (T14/10,23)
(7) Fortunately, too, most of the men are bachelors and so are spared the withering
remarks of bored spouses. (T24/6,SP/2)
380 From Clerks to Corpora
(22) I was pursing a bachelor life with almost no downside.
(23) It was one of the first recipes that my husband attempted in
his bachelor days. (BNC, ABB(864))
(24) “You mean, he had them under the bed in his old bachelor
flat in Wimbledon, …” (ST24/2,5/7) (25) Heuston lived the life of a bachelor don with rooms in college for the next 15 years. (T27/12, 9)
(26) She met Ernest Weekly, a bachelor lecturer from
Nottingham University, when she was 18 and he was 33,
and they embarked on a loveless, dessicated marriage, …
(27) … since Wendy and Tom Witherington had two young
children upon whom their bachelor uncle doted. (BNC,
(28) Or, as he sang back in 1962, “until then I’ll be your bachelor boy and that’s the way I’ll stay, happy to be a bachelor
boy until my dying day.” (ST18/6,3/2)
5. Two types of occasional features or collocates
Within the class of occasional bachelor attributes which occur as
premodifying collocates of this noun we can in fact distinguish two
­different groups depending on what regular features in bachelor they
can be used to comment on. The first group consists of premodifiers that
can also be found with the superordinate man. Such bachelor instances
can be replaced by man without affecting the reading of the premodifier, although such a noun phrase would of course have a less specific
meaning, since bachelor contains the additional information that we
can spell out as has never been married or is unmarried. The strings
from (29) to (31) contain examples of such premodifying c­ollocates
of bachelor.
(29) Starkie, an American bachelor, regularly accom­panies the
duchess on charity trips to eastern Europe, … (ST18/6,1/7)
(30) He is Marc Andreessen, a large, loose-limbed 24-yearold b
­ achelor from the Midwest, who appears to live for
nothing but computers and junk food. (ST13/8,3/8)
(31) He was an old, stooping, emaciated bachelor,aghast at the
facts of life; he never smiled but glowered defensively at
the world from under his eyebrows, … (BNC,ABW(240))
The second group is illustrated below. The collocates eligible, confirmed, incorrigible, entrenched, and lifelong must be taken necessarily
“Bachelor Means Nothing Without Husband and Father” 381
to associate to the entailed unmarried feature of the head bachelor. In
all these cases the combination of the premodifying collocate and bachelor is semantically quite cohesive, and an instance of man could not
be substituted for bachelor in any these phrases, as the choice of these
particular premodifiers would then no longer make sense, at least not in
these specific contexts. Not even unmarried man would seem a possible
alternative, because it is not normally used in collocation with these
premodifying adjectives to convey the same sense as bachelor.
As I have outlined, bachelor seems to be associated with a rich and
variable set of characteristics, and this is no doubt directly connected
with its use in so many different collocations and wider contexts depicting scenarios involving bachelors and aspects of bachelorhood. There
are likely to be differences between individual speakers as regards what
occasional features tend to come to mind when they hear this lexical
item mentioned, especially out of a specifying context, but speakers
are prepared for variation in the use of the noun. This is presumably
a result of their linguistic experience; the many times they have heard
bachelor being used or used it themselves. Their extralinguistic experiences of bachelors and their way of life must, as usual, interact with
their language competence and practices, for instance their recognition
of both frequent and apparently conventionalized collocations and
other collocations considered possible.
(32) John Fitzgerald Kennedy, 34-year-old son of the slain
President, is variously known as “the hunk”, “the sexiest
man alive”, and America’s most eligible bachelor. (T12/9,15)
(33) … he had the slightly panicky look of a confirmed bachelor who has just walked into a maternity ward of bawling
babies, … (ST26/2,10/7)
(34) He had long seemed an incorrigible bachelor but in 1971
he surprised his friends with marriage. (T2/10,23)
(35) … verses aimed at turning the thoughts of entrenched bachelors to the comforts of matrimony … (ST12/2,5/7)
(36) A lifelong bachelor, Paul Hogan leaves no survivors.
(37) Often dubbed Britain’s most eligible bachelor, the late duke
was linked with a series of glamorous women … (T1/11,1)
(38) One Fleet Street columnist after another claimed her wedding to one of the world’s most eligible bachelors was an
appalling mistake. (ST19/11,1/5)
(39) … he was an international champion golfer and one of the
most eligible young bachelors in the London society of the
early 1930s. (BNC,K5J(2175))
382 From Clerks to Corpora
(40) As a youth Richard was red-haired, high, wide and
handsome, and considered the most eligible bachelor in
Christendom. (ST8/1,3/1)
(41) Thus Andrew Davies has Mrs Bennet’s initial announcement to her husband that an eligible bachelor has taken the
tenancy of Netherfield Hall take place most incongruously
as the family hurries home from church. (T6/10,39) (42) But while some women complain that eligible black bachelors are hard to find, others play solitaire by choice.
(43) Housman, a confirmed bachelor, was born in 1859 and had
a stifling Victorian upbringing. (T23/9,5)
(44) Sadly, most seem to be confirmed bachelors. (ST30/4,9/14)
(45) … a retired businessman and confirmed bachelor with bald
head and circular spectacles, dressed in old-fashioned tights
and gaiters. (BNC,Boy(981))
6. Figurative shifts and irregular features
As an interesting contrast, in the next two examples bachelor is used
as a premodifier of the noun phrase heads Mother and queen. These
collocational foci stand for women, and accordingly they force a suppression of the regular sense quality male in the primary sense of the
collocate bachelor. This collocational tailoring of the both syntactically
and semantically dependent item bachelor is necessary if they are to
function together to convey the intended meaning.
(46) Notable among her films during the period were the comedy Bachelor Mother (1939), with David Niven, as a shopgirl who finds an abandoned baby, and the drama Kitty
Foyle (1948), … (T26/4,19)
(47) Queen Christina, directed by Rouben Mamoulian, takes
liberties with history, telling how Sweden’s 17th-century
bachelor queen abjures her love for the Spanish ambassador (John Gilbert) in the higher interests of the state.
In short, such a figurative, or more specifically metaphorical shift in the
reading of bachelor is connected with the introduction of an irregular
feature in the understanding of this lexical item within a specific collocation, and I would suggest that this is commonly the case in figurative
uses. A lexical word like bachelor can be semantically influenced by
its companion. If bachelor is the collocational focus, its collocate can
“Bachelor Means Nothing Without Husband and Father” 383
decisively add a specifying feature to its semantic contents, but if it is
itself a more dependent collocate, it can be collocationally tailored by
the focus. The semantic characteristics that can explicitly be spelt out in
a collocational companion are of three different, general types. Regular
features are only added to a word in specific cases, usually to explain its
meaning in an analytic sentence, while occasional semantic attributes
highlight possible but not regular features of a particular lexicalized
sense. Finally, irregular attributes trigger a figurative shift of some kind
in the reading of a lexical item.
regular -------- occasional -------- irregular
However, bachelor instances like the two in (46) and (47) share other,
both regular and merely occasional attributes of the sense category
bachelor. Actually, more unstable but still common occasional bachelor
features like independent or fending for oneself appear to be important
for the use of bachelor in such cases. Obviously, spinster cannot be used
to convey the complex meanings expressed by presumably incidental
combinations like bachelor mother and bachelor queen, because it is
not typically associated with such positive meaning aspects.7
Adding to the above, in the kind of similarly metaphorical bachelor application exemplified below, it is instead primarily the regular
semantic attribute human in bachelor that is suppressed. And since animals do not marry, although the relations between the sexes among
mammals and birds are similar to what we find in human societies, the
unmarried or has-not-married feature cannot be taken at face value
either in such contexts. Actually, this is a good example of the insistent
impression that our interpretations of words and compositional verbal
strings are typically many-sided, holistic complexes rather than some
kind of simplistic adding-up of discrete sense aspects.
(48) Any sedge warblers that are still singing at midsummer will
be bachelors, just warbling on hopefully. (T29/4,WE/12)
(49) A minority of the males accomplish most of the mating;
and many males die bachelors. (BNC,GUB(2048))
(50) Large groups of dolphins are mixed in age and sex, but
smaller groups generally are of three types: a nuclear group,
comprising a single adult male and female; a nursery group,
with a number of adult females and young; and a bachelor
group, with adult and young males. (BNC,ABC(440))
7 Bachelor girl seems, on the other hand, to be a lexicalised compound (ALD 1989: 72).
384 From Clerks to Corpora
(51) Gelada baboons often move in large herds; individual harems move separately and the ‘bachelor’ males are
found in their own discrete and coherent social units.
Furthermore, there is probably a difference between the systematic
status of the types of figurative bachelor uses exemplified above. The
use of bachelor to denote animal males without a female partner can
be regarded as a lexicalised secondary sense of this lexeme, while it is
questionable whether the use of bachelor in combinations like bachelor
girl and bachelor queen has the same more independent applicatory
status. Both commonly occur as premodifying collocates, but bachelor
as a synonym of male animal without a female partner can clearly also
be used as a noun phrase head that on its own serves to pick out referents in the extralinguistic universe of discourse.
7. Conclusion
This study of the collocational range and the textual environment of
the English lexeme bachelor suggests that its meaning potential is quite
complex and variable. Even if we look just at the primary sense of bachelor, it is clear that its meaning potential is in certain respects far richer,
or more multifaceted and variable, than what was assumed in different
attempts at a classical kind of categorization, which would simply claim
that the noun had a fixed and clearly delimited content that could be
paraphrased as unmarried man, or man who has never been married.
Together with ordinarily expected features like able/free to marry, the
regular, incorporated features summed up in these paraphrases appear
to make up a kind of general sense prototype for speakers of English.
The collocational potential of bachelor makes it quite clear, however,
that the understanding of this lexical item can be associated with a wide
range of merely occasional but still possible meaning features. In other
words, the more peripheral range of this sense category is variable and
rich in possible associations or attributes. There are bachelors of many
different sorts of personalities and characteristics, all with unique life
stories, and speakers of English may differ as to how typical they feel
that a given representative of the category bachelor is. We build up our
understanding of words, including our assessment of their collocational
potential in compositional grammatical string, through active interaction with our environment. In short, the experiences that help form our
“Bachelor Means Nothing Without Husband and Father” 385
conception of how a lexical item like bachelor can be used in English
are both extralinguistic and verbal.
Alm-Arvius, C. (1999). Metaphor and Metonymy as Meaning Generalisations
Comparable to Hyponymy and Meronymy Respectively. S.-K. Tanskanen
& B. Wårvik (eds) Proceedings from the 7th Nordic Conference for English
Studies. Anglicana Turkuensia No 20. Turku, Finland: University of Turku,
Alm-Arvius, C. (1993). (Photocopy version in 1991) The English Verb See: A
Study in Multiple Meaning. Gothenburg Studies in English 64. Göteborg,
Sweden: Acta Universitatis Gothoburgensis.
Cruse, D. A. (1986). Lexical Semantics. Cambridge Textbooks in Linguistics.
Cambridge, UK: Cambridge University Press.
———. (2000). Meaning in Language. Oxford & New York: Oxford University
Fauconnier, G. (1997). Mappings in Thought and Language. Cambridge, UK:
Cambridge University Press.
Fillmore, C. J. (1985). Frames and the Semantics of Understanding. Quaderni
di Semantica, 6:2, 222–254.
———. (1982). Towards a Descriptive Framework for Spatial Deixis.
R. J. Jarvella & W. Klein (eds) Speech, Place and Action. Chichester, New
York: John Wiley & Sons Ltd, 31–59.
Firth, J. R. (1957). Modes of Meaning. Papers in Linguistics 1934–1951.
London, New York & Toronto: Oxford University Press.
Lakoff, G. (1987). Women, Fire, and Dangerous Things. Chicago & London:
The University of Chicago Press.
Lakoff, G. & Johnson, M. (1999). Philosophy in the Flesh. New York: Basic
Longman Dictionary of Contemporary English 1987. (2nd ed.). Burnt Mill,
Harlow, Essex: Longman. (Longman).
Oxford Advanced Learner’s Dictionary 1989. (4th ed.). Oxford, New York,
etc: Oxford University Press. (ALD).
Rosch, E. (1975). Cognitive Reference Points. Cognitive Psychology, 7(4),
386 From Clerks to Corpora
———. (1977). Human Categorisation. N. Warren (ed.) Studies in Crosscultural Psychology Volume 1. London, New York & San Francisco:
Academic Press, 1–49.
———. (1978). Principles of Categorization. E. Rosch & B. B. Lloyd (eds)
Cognition and Categorization. Hillsdale, N.J.: Lawrence Erlbaum, 27–48.
Rosch, E. & Mervis, C. B. (1975). Family Resemblances: Studies in the Internal
Structure of Categories. Cognitive Psychology, 7(4), 573-605.
About the Authors
Philip Shaw is a professor in the English Department at Stockholm
University. He has published within a wide range of areas including historical linguistics and, more recently, World Englishes, applied linguistics, particularly with a focus on academic and business English, and
vocabulary learning among advanced students. With Gunnel Melchers
he has published a textbook on World Englishes (see above), and his
recent articles have appeared in such journals as Journal of Second
Language Writing, TESOL Quarterly, and Journal of Pragmatics.
ORCID: 0000-0002-8301-3960
Britt Erman has retired from a post as associate professor in the English
Department at Stockholm University. Her earlier research focused
on pragmatics and communication. More recently she has published
within the areas of cognitive linguistics and L2 acquisition, the latter focusing on academic writing, formulaic language, vocabulary
and syntactic complexity among high-level learners of English. Her
recent work has appeared in journals such as International Journal of
Corpus Linguistics, Journal of Pragmatics and English Language and
Linguistics (ELL). ORCID: 0000-0002-5009-440x
Gunnel Melchers is Professor Emerita in the English Department at
Stockholm University. Her research has been devoted to regional and
social variation, with special reference to the north of England and
Scotland’s Northern Isles. Among her many publications within these
areas are such books as World Englishes, with Philip Shaw, (Hodder,
2011 2nd ed.), Writing in Nonstandard English, co-edited with Irma
Taavitsainen and Päivi Pahta (Benjamins, 1999), Studies in Anglistics
(1995; co-edited with Beatrice Warren), and Nonstandard Varieties
of English (1994; co-edited with Nils-Lennart Johannesson). ORCID:
388 From Clerks to Corpora
Peter Sundkvist is an associate professor in the English Department at
Stockholm University. His research interests concern phonology, phonetics, and dialectology, relating to varieties of English and Germanic
languages more generally. Most of his empirical work has focused on
the Shetland Islands and such topics as Shetland Scots and Scottish
Standard English. His work has appeared in journals such as English
World-Wide and World Englishes. ORCID: 0000-0001-7870-6351
Karin Aijmer is Professor Emerita in English Linguistics at the University
of Gothenburg. Her research has focused on pragmatics and discourse,
including such topics as modality, pragmatic markers, conversational routines, and fixed expressions. Her recent publications include
Understanding Pragmatic Markers: A Variational Pragmatic Approach
(Edinburgh University Press, 2013), a special issue of Functions of
Language on “Discourse linguistics: Theory and practice” (Benjamins,
2014, co-edited with Anita Fetzer)and the chapter on pragmatics in The
Routledge Companion to English Studies (Routledge, 2014). ORCID:
Christina Alm-Arvius, who passed away in 2013, was an associate professor in the English Department at Stockholm University. Her main area
of research was semantics, in particular lexical semantics. Many of her
studies focused on aspects of polysemy and involved such related fields
as word formation, idiomaticity, the concept of construction, discourse
and text analysis. Her publications include Introduction to Semantics
(Studentlitteratur, 1998) and Figures of Speech (Studentlitteratur, 2003).
Cecilia Ovesdotter Alm is an assistant professor in the Department
of English at the Rochester Institute of Technology. Her specialism
is computational linguistics. Her research has involved interdisciplinary research in this field. Recent publications include the book Affect
in Text and Speech (VDM Verlag, 2009) and ‘The Role of affect in
the computational modelling of natural language’ (Language and
Linguistic Compass: Computational and Mathematical, 2012). ORCID:
Javier Calle-Martin is a tenured senior lecturer in the Department of
English at the University of Málaga (Spain). His research interests
are within historical linguistics and manuscript studies, focusing on
About the authors 389
early English documents. He is also the leading researcher of a project for the electronic edition of hitherto unedited late Middle English
and early Modern English Fachliteratur. He is the editor of The
Middle English Version of De viribus herbarum (2012) and A Late
Middle English Remedybook in MS Wellcome 542 (2013). ORCID:
Marcelle Cole is an assistant professor in the English department at the
University of Utrecht. Her most recent book is Verbal Morphosyntax
in Old Northumbrian and the (Northern) Subject Rule. NOWELE
Supplement Series (Benjamins, 2014). ORCID: 0000-0003-4255-7686
Östen Dahl is Professor Emeritus in General Linguistics at Stockholm
University. His fields of research include linguistic typology and he has
published extensively on tense and aspect, negation, definiteness and
grammaticalization. Among his monographs are Logic in Linguistics,
with J. Allwood and L-G Andersson (Cambridge University Press,
1977), Tense and Aspect Systems (Blackwell, 1985), and The Growth
and Maintenance of Linguistic Complexity (Benjamins, 2004). ORCID:
Marina Dossena is Professor of English Language at the University
of Bergamo, Italy. Her research interests focus on Scots and Scottish
English and the history of specialized discourse. Her many publications include Insights into Late Modern English, co-edited with Charles
Jones (Peter Lang, 2003), Methods and Data in English Historical
Dialectology, co-edited with Roger Lass (Peter Lang, 2005), and the monograph Scotticisms in Grammar and Vocabulary (John Donald, 2005).
ORCID: 0000-0002-8025-6086
Thomas Egan is Professor of English Linguistics at Hedmark University
College in Hamar, Norway. His research interests encompass topics
within the areas of corpus linguistics, contrastive linguistics, cognitive
linguistics and historical linguistics, including grammaticalisation. He
is the author of a monograph on complementation, entitled Non-finite
Complementation: A Usage-based Study of Infinitive and –ing Clauses
in English (Rodopi, 2008). ORCID: 0000-0002-8826-5972
Christine Johansson is a senior lecturer in the English department at
Uppsala University. Her fields of research include syntax, diachronic
syntax, corpus linguistics and learner English. Her publications include
‘Relativizers in 19th-century English’ in Nineteenth-century English:
390 From Clerks to Corpora
Stability and Change (Cambridge, 2006) and ‘Relativization in Early
Modern English: Written versus speech-related genres’ in Historical
Linguistics of English, (Mouton, 2012). ORCID: 0000-0002-9440-9215
Maria Kuteeva is a professor and Director of the Centre for Academic
English in the English Department at Stockholm University. Her main
research interests lie in the field of applied linguistics, with a particular
focus on academic uses of English across disciplines. Her recent work
has appeared in such journals as Applied Linguistics, Journal of English
for Academic Purposes and Journal of Second Language Writing.
ORCID: 0000-0003-2942-1426
Merja Kytö is Professor of English Language at Uppsala University. Her
research focuses on variation in English (synchronic and diachronic),
corpus linguistics, historical pragmatics and early manuscript studies. Among her publications are English Corpus Linguistics: Crossing
Paths (Rodopi, 2012), Testifying to Language and Life in Early Modern
England, with Peter J. Grund and Terry Walker (Benjamins, 2011), and
Early Modern English Dialogues: Spoken Interaction as Writing, with
Jonathan Culpeper (Cambridge University Press, 2010).
Margareta Lewis is affiliated with the English Department at Stockholm
University through the large-scale project “High-Level L2 Proficiency
and Use” involving several language departments. At present Lewis
holds a position as senior lecturer in L2 English within adult education.
Recent publications include ‘Multiword structures in different materials and with different goals and methodologies’, with Britt Erman,
Yearbook of Corpus Linguistics and Pragmatics 2013 (Springer).
ORCID: 0000-0001-9123-116x
David Minugh recently retired from a post as foreign lecturer in the
English Department at Stockholm University. His research has primarily concerned corpus linguistics and idiom research, and American
English, but also includes work on a number of dictionaries. His recent
publications include a chapter in Corpus Linguistics 25 Years on
(Rodopi, 2007), and his compilation PhD thesis Studies in Corpora and
Idioms: Getting the Cat Out of the Bag. 2014. Stockholm: Stockholm
University. ORCID: 0000-0002-6481-1975
David Moreno Olalla is a lecturer at the University of Málaga. Most of
his work deals with Middle English Scientific Fachliteratur and manuscript studies, which he alternates with dialectological research. He has
published of late on the textual transmission of a hitherto unknown
About the authors 391
northern ME translation of Macer Floridus’s De Viribus Herbarum.
ORCID: 0000-0002-9772-8959
Erik Smitterberg is an associate professor and a senior lecturer in
English Linguistics, at Uppsala University. His main fields of research
concern Late Modern English, from a corpus-based perspective.
Among his publications are the books The Progressive in 19th-century English: A Process of Integration (Rodopi, 2005), and Spotting the
Error: A Problem-based Workbook on English Grammar and Usage
(Studentlitteratur, 2007). ORCID: 0000-0001-9596-7406
Gjertrud F. Stenbrenden is an associate professor of English in the
Department of Literature, Area Studies, and European Languages at
the University of Oslo. Her main interests are phonology, dialectology,
diachronic linguistics, linguistic variation and change. Recent publications include ‘The diphthongisation of ME uˉ : The spelling evidence’.
Language and Computers 2013, and Academic Writing in English, with
Per Lysvåg, (Cappelen Damm, 2014). ORCID: 0000-0003-3094-7900
Gunnel Tottie is Professor Emerita at the Department of English,
University of Zurich. She has published widely within such fields as
syntax and pragmatics. Her most recent book is An Introduction to
American English (Blackwell). ORCID: 0000-0001-6994-9829
Francesco-Alessio Ursini is a lecturer in the English Department at
Stockholm University. Recent publications include ‘The interpretation of
spatial At: An Experimental Study’, Journal of Cognitive Science 2013.
ORCID: 0000-0001-7042-3576
Laura Wright is a Reader in English Language at the Faculty of English,
University of Cambridge. Her most recent book is Code-Switching in
Early English, co-edited with Herbert Schendl (Mouton de Gruyter, 2011).
ORCID: 0000-0002-9953-6483
Arne Zettersten is Professor Emeritus of English at the University of
Copenhagen. His scholarly contributions cover a broad range of areas,
including grammar, semantics, lexicography, etymology, and medieval textual studies. He has made important contributions to work on
J.R.R. Tolkien. Among Zettersten’s many works is his monograph on
Tolkien: Tolkien – min vän Ronald och hans världar (Atlantis, 2008)
(later translated into J.R.R. Tolkien’s Double Worlds and Creative
Process (Palgrave McMillan, 2011)). ORCID: 0000-0003-3393-1330
Why is the Isle of Dogs in the Thames called Isle of Dogs? Did King Canute’s
men bring English usage back to Jutland? How can we find out where English
speakers suck their breath in to give a short response? And what did the Brontës
do about dialect and think about foreign languages?
The answers are in this collection of empirical work on English past and present
in honour of Nils-Lennart Johannesson, Professor of English Language at
Stockholm University. The first five chapters report individual studies forming
an overview of current issues in the study of Old and Middle English phonology,
lexis and syntax. The next six look at Early Modern and Modern English from
a historical point of view, using data from corpora, manuscript archives, and
fiction. Two more look at the Old English scholar JRR Tolkien and his work. The
remaining chapters discuss aspects of Modern English. Several use corpora to
look at English usage in itself or in relation to Swedish, French, or Norwegian.
The last three look at grammatical models, the pragmatics of second language
use, and modern English semantics.
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF