University of New Brunswick
Fredericton, New Brunswick, Canada
31 October - 1 November 2014
Université du Nouveau-Brunswick
Fredericton, Nouveau-Brunswick, Canada
31 octobre au 1 novembre 2014
University of New Brunswick
© 2015 by individual authors of the papers
© 2015 Auteures et auteurs des communications
Papers from the 38th Annual Meeting of the Atlantic Provinces Linguistic Association
Actes du 38e Colloque annuel de l’Association de linguistique des provinces atlantiques
v.38 (2014)
Legal deposit / Dépôt légal : 2015
National Library of Canada
Bibliothèque nationale du Canada
ISSN: 2368-7215
About APLA/ALPA 38 ......................................................................................... 1
“The moving adverb moves; and, having moved, moves on.” Adverb
placement in L1 Arabic, L2 English and L3 French................................................ 3
The Dictionary of Cape Breton English: From “after” to “Zombie”...................... 19
Les stratégies verbales de politesse en situation de « double contrainte »
dans le débat télévisé de 2008 au Canada ............................................................... 33
The language of quotations in Hong Kong Chinese newspapers............................ 49
The syntax of null pronouns in Russian.................................................................. 57
Patricia Balcom
Département d’anglais, Université de Moncton
Wladyslaw Cichocki
Department of French, University of New Brunswick
William Davey
Senior Scholar, Cape Breton University
Steeve Ferron
Département d’études françaises, Université de Moncton
Sylvia Kasparian
Département d’études françaises, Université de Moncton
Anthony C. Lister
University of New Brunswick (retired)
Egor Tsedryk
Department of Modern Languages and Classics, Saint Mary’s University
University of New Brunswick
The 38th Annual Meeting of the Atlantic Provinces Linguistic Association (APLA) /
Association de linguistique des provinces atlantiques (ALPA) took place from 31 October
to 1 November 2014 at the Wu Conference Centre on the campus of the University of
New Brunswick in Fredericton NB. The conference was organized by a committee of
linguists from the University of New Brunswick (UNB). The last time that APLA/ALPA
met in Fredericton was in 2007 on the occasion of the Association’s 31st anniversary.
About 30 scholars attended the APLA/ALPA38 conference. In addition to linguists
from the four Atlantic Provinces, the conference attracted researchers from other regions
of Canada, including Alberta, Ontario and Québec, as well as from Japan. About one
quarter of the attendees were students.
The conference theme was “Language and society.” James A. Walker (Department
of Languages, Literatures and Linguistics, York University, Toronto) delivered the
plenary Murray Kinloch Memorial Lecture. His presentation, entitled “Canadian English
in the 21st century: Contact, continuity and change,” discussed ethnolinguistic variation in
Toronto English. He showed that while young Torontonians of all ethnic backgrounds
have adapted to local linguistic norms they convey their ethnicity through differential use
of features.
This volume of PAMAPLA/ACAPLA includes revised versions of five
presentations given at the conference. The following fifteen papers were also presented
but do not appear in these proceedings.
CHARLES BOBERG, McGill University, Variation and change in Gaspé English
To appear as C. Boberg & J. Hotton in English World-Wide
CHERYL ANNE CLELAND, York University, Gendered language in job advertisements and
its effects on job application
PHILIP COMEAU, Université du Québec à Montréal, Les effets de la négation sur deux
variables sociolinguistiques
PAUL COOK, University of New Brunswick, Can we build a corpus of Canadian English
automatically from the Web?
ELIZABETH COWPER & DANIEL CURRIE HALL, University of Toronto & Saint Mary's
University, Morphosyntactic features and the scope of contrast
Newfoundland, How does variation in production interact with perception? A look at
differences in light and dark /l/ in Newfoundland English
PAMAPLA 38 / ACALPA 38 (2014)
BERNARD MULO FARENKIA, Cape Breton University, Lecture pragmatique des comptes
rendus critiques d’ouvrages en linguistique
BRANDON J. FRY, University of Ottawa, Free Merge and minimality
ANNE L. KLINCK, University of New Brunswick, Making a significant difference:
Bilingualism and re-creation in Charles d’Orléans
Published in Neophilologus:
MATTHIEU LEBLANC, Université de Moncton, Politique linguistique = politique de
traduction? : pour une analyse du rôle de la traduction officielle au Nouveau-Brunswick
TAKESHI OGURO, Chiba University of Commerce, Types of WH-questions in Japanese,
with special reference to Speech Act Phrase
MACKENZIE SALT, McMaster University, People, persons, and individuals: Is the DSM
University of Douala, Language policy in education: Minority official language in
technical and vocational programs in Canada and Cameroon
ALEXANDRA TSEDRYK, Mount Saint Vincent University, Pour une approche structurée
d’enseignement de la paraphrase
The conference was made possible by generous support from the following
offices and departments:
Office of the President and Vice-Chancellor, Dr Eddy Campbell, UNB
Office of the Vice President (Academic), Dr Tony Secco, UNB
Office of the Vice President (Research), Dr David Burns, UNB
Office of the Dean of Arts, Dr George A. MacLean, UNB
Second Language Education Centre, Faculty of Education, Dr Joseph Dicks, UNB
Department of English, Chair: Dr Jennifer Andrews, UNB
Département d’études françaises, Chair: Dr Christine Horne, UNB
Department of Psychology, Chair: Dr Sandra Byers, UNB
Photos from the conference are posted on the APLA/ALPA website at
Two student assistants helped to ensure that everything ran smoothly: Tarun
Porter and Natasha Quiring. Thanks also to assistance from Daniel Grant and Natalie
Savoie. The members of the Organizing Committee were Wladyslaw Cichocki (French,
UNB), Paul Cook (Computer Science, UNB), Christine Horne (French, UNB) and
Anthony C. Lister (Retired, UNB).
Université de Moncton
This paper describes a study that explored the acquisition of adverb placement in L3
French by learners whose L1 was Emirati Arabic and whose L2 was English. A
picture-description task was developed to tap what Fassi Fehri calls the
“transportability” of adverbs in Arabic. Participants were asked to write – in all three
languages – up to three sentences describing nine labelled images, using all of the
words on the image. While all four possible word orders were produced in all three
languages, the frequencies differed depending on adverb class and on language. With
imperfective adverbs, in Arabic and English the second and third responses provided
different positions for the adverb (i.e., ASVO, SAVO, SVAO, SVOA); this rarely
occurred in French. With measure and voice adverbs, in Arabic many second and
third responses were with the verb in sentence-initial position (i.e., VSOA and
VOSA); this did not occur in English or French. Transportability in the L1 and freer
adverb placement with imperfective adverbs in the L2 did not transfer to the L3 at the
earliest stages.
According to De Angelis and Selinker (2001), no model of second language
acquisition (henceforth SLA) that incorporates cross-linguistic influence can be
complete if it is based exclusively on two languages. The L1 does not have
“privileged status” (De Angelis, 2005:401), and subsequent languages acquired can
also have an effect on acquisition of the target language. A number of hypotheses
have been put forward concerning "the L3 transfer puzzle" (Rothman and CabrelliAmaro 2010:192): L1 transfer; L2 transfer (the "L2 status factor", Bardel and Falk
2007), which involves factors such as recency and learning environment; and positive
transfer from L1 and/or L2 called the “Cumulative Enhancement Model” (Flynn et al.
2004). Cross-linguistic influence in L2/L3 acquisition can also be affected by
psychotypology, or perceived language distance (Kellerman, 1977).
Arabic and French share rich verbal morphology, which accounts for the
placement of adverbs and negation, but learners in the Arabian Gulf may not perceive
these similarities because English and French are alike in terms of their writing
systems, vocabulary and culture. This study examines adverb placement in L1 Arabic,
L2 English and L3 French by speakers of Emirati Arabic. More specifically, it looks
at whether what Fassi Fehri (1993: 13) called the “transportability” that characterizes
the distribution of adverbs in Arabic is transferred to L3 French, or whether the fixed
position of adverbs from L2 English is transferred.
Basic word order in Emirati Arabic, English and French
Basic declarative word order in Emirati Arabic is SVO, as it is in Gulf Arabic in
general (Holes, 1989). However, VSO and OVS are also common (Aljenaie and
Farghal 2009, Shaalan 2010). Basic declarative word order in English and French is
PAMAPLA 38 / ACALPA 38 (2014)
also SVO. With the exception of topicalization and left-dislocation (OSV), no other
word orders are grammatical in English or French. Topicalization and left-dislocation
are also found in Gulf Arabic (Holes, 1989).
Adverb placement in Emirati Arabic, English and French
For the analysis of adverb position in Arabic, English and French I have adopted the
“feature-based theory” of adverb syntax based on Cinque (1999, 2004). Under this
theory, all adverbs are base-generated as specifiers of semantically related functional
projections between TP and VP/vP. According to Cinque (1999:v), the same
functional projections are found in the same order cross-linguistically, and superficial
differences in adverb placement between languages arise because other syntactic
elements may move up from VP to positions above the adverb. This is perhaps one
of the more controversial aspects of Cinque’s theory. While many researchers –
Alexidou (1997), Fassi Fehri (1997), Abeillé and Godard (2004) and Vainikka (2007)
for example – have argued that adverbs can be base-generated in functional categories
above VP, they also maintained that adverbial complements and other clause-final
adverbials are generated within VP. Cinque (1999, 2004) was mute concerning
possible mechanisms triggering movement. However, Costa (2004:737) proposed that
“remnant VP movement” or scrambling could be involved.
Fassi Fehri (1997) and Vinet (1995) showed that Cinque’s hierarchy was
applicable, with minor modifications, to Arabic and French respectively. Table 1
shows a partial adverb hierarchy for Arabic, English and French, with functional
projections for the adverbs used in the present study.
Table 1
Functional projections for adverbs in English, French and Arabic
a lot
a bit
Traditionally, linguists have not made a distinction between habitual and
frequentative adverbs, referring to both as “frequency” adverbs, but Cinque (1999)
showed that adverbs expressing these meanings can co-occur in English and French,
as shown in (1). According to my informants they can co-occur in Gulf Arabic as
well. In what follows, I will refer to habitual and frequentative adverbs together as
“imperfective adverbs” (Rivero, 1992: 311).
She often usually eats fruit.
Elle mange habituellement souvent des fruits.
Second, as shown in (2a), in English usually (a habitual adverb) cannot appear
in clause-final position while often (a frequentative adverb) can (2b.)
*Mary eats ice-cream usually.
Mary eats ice-cream often.
French and Arabic are somewhat different: the equivalents of both usually and
often can occur in clause-final position as shown in (3) and (4).
regardent la télé habituellement/souvent.
‘They watch the TV usually/often.’
Rachid yektib mesedjat
‘Rachid writes messages
ʔadatan/ marrat
To go in to a bit more detail, as can be seen in (5) and (6), often and usually
occupy different positions in English: SVAO is ungrammatical (5c and 6c), and in
clause-initial and clause-final position grammaticality depends on the adverb,
grammatical with often (5a) and (5d) but ungrammatical with always (6a) and (6d).
Often Meera speaks French.
Meera often speaks French.
*Meera speaks often French.
Meera speaks French often.
?Usually Meera speaks French.
Meera usually speaks French.
*Meera speaks usually French.
*Meera speaks French usually.
The equivalent adverbs in French (7) and (8) show different behaviour with
ASVO; this word order is grammatical with souvent (7a) but not with habituellement
(8a). Otherwise the two adverbs have similar syntactic behaviour: SAVO is
ungrammatical, as shown in (7b) and (8b); SVAO is grammatical in both (7c) and
(8c); and SVOA is grammatical with both.
Souvent Meera parle français.
*Meera souvent parle français.
Meera parle souvent français.
Meera parle français souvent.
?Habituellement Meera parle français.
*Meera habituellement parle français.
Meera parle habituellement français.
Meera parle français habituellement.
In Arabic the equivalents of both often and usually can occur in all four
adverbial positions (9), demonstrating the characteristic of transportability, but SAVO
is preferred.
Marrat/ʔadatan Meera tetkallam faransi.
‘Meera often/usually speaks French.’
Meera marrat/ʔadatan tetkallam faransi.
‘Often /Usually Meera speaks French.’
Meera tetkallam marrat/ʔadatan faransi.
‘Meera speaks often/usually French.’
PAMAPLA 38 / ACALPA 38 (2014)
Meera tetkallam faransi marrat/ʔadatan
Meera speaks French often/usually.
As can be seen in (10) and (11), adverbs of voice (well, badly) and measure (a
lot, a little) can only occur in SVOA position in English.
(10) a.
Meera dances ballet badly.
*Badly Meera dances ballet.
*Meera badly dances ballet.
*Meera dances badly ballet.
(11) a.
Fatima likes Paris a lot.
*A lot Fatima likes Paris.
*Fatima a lot likes Paris.
*Fatima likes a lot Paris.
The French equivalents occur only in SVAO, as shown in (12) and (13).
(12) a.
*Meera danse le ballet mal.
*Mal Meera danse le ballet.
*Meera mal danse le ballet.
Meera danse mal le ballet.
(13) a.
?Fatima aime Paris beaucoup.
*Beaucoup Fatima aime Paris.
*Fatima beaucoup aime Paris.
Fatima aime beaucoup Paris.
In Arabic they are transportable, as shown in (14) and (15), although SVOA is
(14) a.
Meera terges Khalidji muzein.
‘Meera dances the Gulf dance badly.’
Muzein Meera terges Khalidji
‘Badly Meera dances the Gulf dance.’
Meera muzeinterges Khalidji
‘Meera badly dances the Gulf dance.’
Meera terges Khalidji muzein.
‘Meera dances badly the Gulf dance.
Table 2
Positions for measure and voice adverbs3
Table 2 summarizes placement of voice and measure adverbs in the three
languages. In French they have the same position as imperfective adverbs — postverbal — while in English they are clause-final. In Arabic they are transportable,
although SVOA is preferred.
Research questions
Based on the description of adverb placement in Arabic, English and French
presented in Section 1.2, if there is L1 transfer in the L3 acquisition of adverb
placement, Arabophones should demonstrate transportability of adverbs in L3 French.
In that case, we would expect participants to produce sentences exhibiting ASVO,
SAVO, SVAO and SVOA word orders. If there is transfer of adverb placement from
L2 English, adverb placement will be quite rigid in the L3 and will reflect the L2
order. In this case we would expect SAVO for habitual and frequentative adverbs and
SVOA for voice and measure adverbs, but none of the other orders, with the
exception of often, where ASVO and SVOA might be produced.
In sum, my two research questions are: Will Emirati Arabic speakers transfer
the property of transportability of adverbs to L2 English and L3 French? Will they
transfer fixed adverb position from L2 English to L3 French?
As mentioned in Section 1.2, in Arabic most adverbs occur freely in a variety of
positions. In order to tap this property of transportability, a picture-description task
was developed, inspired by Juffs (1996). The tasks were administered over the
semester to the same learners in the three languages, with at least one week between
tasks. Examples in what follows are taken from the English task.
2.1 Materials
Participants were asked to write up to three sentences describing nine labelled
images, using all of the words on the image. If they thought that there were fewer
than three possible sentences, they were asked to write “no other way”. The stimuli
consisted of one adverb of frequency (often) and one habitual adverb (usually), two
voice adverbs (well, badly) and two measure adverbs (a lot, a bit). Three negative
adverbs (not, ever, anymore) were also included in the task, but the results with these
adverbs will not be presented under Results.
Before performing the tasks, participants were given two practice pictures,
one which showed two men looking at each other with the labels Ali, looks at and
Mohamed, and the other with two framed pictures and the words Fatima’s picture,
Meera’s picture and under-above. Possible responses to the first practice image are
shown in (15), and were discussed before participants proceeded to the task.
(15) Ali looks at Mohamed.
Mohamed looks at Ali.
No other way.
Figure 1 gives an example of an image used in the picture-description task. Possible
answers for Figure 1 are given in (16).
PAMAPLA 38 / ACALPA 38 (2014)
(16) Fatima likes a lot Paris.
Fatima likes Paris a lot.
No other way.
Figure 1
Example of an image2 from the picture-description task
There were also grammaticality judgement tasks in the three languages, but only the
results of the picture-description task will be presented in this paper.
2.2 Participants
The tasks in the three languages were administered to intact classes of beginning
French L3 learners who were intermediate to advanced L2 users of English. Students
who were non-Emiratis were excluded from the analysis, as were students whose
mothers were not native speakers of Emirati Arabic, or those who did not complete
the tasks in the three languages. The final set of participants was 16 female students
studying at a post-secondary institution in the United Arab Emirates, ranging in age
from 20 to 26 (mean 21.2, median 21.0).
2.3 Data coding and analysis
All responses were coded according to their structures, i.e. ASVO, SAVO and so
forth. The coding was particularly complex for the Arabic task since, as mentioned in
Section 1.1, basic word order in Gulf Arabic can be SVO, VSO or OVS, while both
English and French have a fixed SVO word order in normal declarative sentences.
For example, when the adverb was in final position, responses included SVOA,
VSOA, and VOSA, as shown in (17).
(17) a.
Meera tetkallam faransi marrat.
‘Meera speaks French often.’
Tetkallam Meera faransi marrat.
‘Speaks Meera French often.’
Tetkallam faransi Meera marrat.
Speaks French Meera often.
The types of responses in (17) were all coded separately, but for ease of
presentation in this paper all were classified SVOA, since the adverb is in clause-final
position. As will be seen in the Results section, in some cases the responses were
SVOA the first, second and sometimes third time: the first response might be SVOA,
the second VSOA and the third VOSA. The same is true of SVAO: any word order
in which the adverb came between the verb and the object is presented as SVAO (for
example VSAO); any pre-verbal adverbs are presented as SAVO (for example
SOAV); and clause-initial adverbs are presented as ASVO (for example AVSO). “No
other way” responses were also coded, and blank answers were coded “No other
Table 3 shows the percentage of “No other way” responses on the picture-description
task. Recall that the possibility of second and third responses was to attempt to tap the
transportability of adverbs. On the Arabic task, only 22.9 % of the second responses
were “no other way”, compared to 40.6% in English and 62.5% in French. In Arabic
70.8% of third responses were “no other way”, compared to 84.4% in English and
97.9% in French.
Table 3
“No other way” second and third responses by language and by adverb class
As can also be seen in Table 3, the response patterns were different with the
various adverb classes. In Arabic and English, there were fewer “no other way” third
responses with imperfective adverbs (15/32 or 46.9% and 19/32 or 59.4%
respectively) than there were with adverbs of measure (24/32 or 75% and 32/32 or
100%) and voice (29/32 or 90.6% and 30/32 or 93.7%). This pattern is not reflected
in the French responses, where third “no other way” responses were similar for the
three verb classes: 93.75% for imperfective adverbs and 100.00% for measure and
voice adverbs.
In what follows, the results will be examined on a class-by-class basis. The
results for the imperfective adverbs often (frequentative) and usually (habitual) will
be presented separately due to the fact that, as was mentioned in Section 1.2, they
belong to different adverb classes and their syntactic behaviour is different.
PAMAPLA 38 / ACALPA 38 (2014)
Table 4
Second and third responses to often and usually
As seen in Table 4, participants treated the two adverbs differently in English
and Arabic: in Arabic often was the adverb with the lowest “No other way” answers
on second and third responses, 0% and 31.2% compared to 31.2% and 81.2% in
English. Conversely, the “No other way” second and third responses for usually were
18.7% and 62.5% in Arabic, and 0% and 37.5% in English. The second responses in
French were more similar to the Arabic results, with fewer “No other way” responses
with often than usually.
Looking at the types of responses more closely, Figure 2 gives the responses
to the frequentative adverb often by language and structure. It is clear that in all three
languages the participants used all four structures, though to varying degrees.
Figure 2
Responses to often by structure and group
For first responses, in Arabic SVAO was preferred (37.5%), in English SAVO
(43.7%) and in French SVOA (57.1%). SVOA was also frequent in Arabic and
English, with a mean of 30.0% in Arabic and 20.9% in English. To avoid long strings
of numbers, in what follows the mean of first, second and third responses was
calculated and will be presented. ASVO occurred frequently in Arabic (mean 27.9%)
and English (mean 16.6%), but very rarely in French (mean 2.4%). SVAO occurred
more frequently in Arabic (mean 16.7%) and French (mean 9.5%) than it did in
English (mean 4.1%), and SAVO occurred relatively frequently in all three languages
(means are 20.9% in English, 15.0% in Arabic and 14.3% in French).
As can be seen in Figure 3, the structures used in responses to the habitual
adverb usually were different from those for the frequentative adverb often, although
again the participants used all four structures in the three languages with differing
Figure 3
Responses to usually by structure and group
In Arabic ASVO was preferred (43.7%) in first responses; in English there were equal
numbers of SAVO and ASVO (37.5% each) and in French SVOA was preferred
(57.1%). SVOA was also frequent in Arabic (mean 31.2%) and English (mean
20.8%). ASVO occurs frequently in both English (mean 29.1%) and Arabic (mean
22.9%), but infrequently in French (mean 7.1%). SVAO occurred more frequently in
French (mean 12.0%) than it did in Arabic (mean 4.2%) and English (mean 4.2%).
SAVO occurred more often in English (mean 33.3%) and Arabic (mean 14.6%) than
it did in French (4.7%).
Figure 4 presents the responses for the voice adverbs badly and well by
language and structure. First of all it can be seen that responses are far less varied and
there are more “No other way” responses than there were for the imperfective
adverbs. In all three languages SVOA is the preferred structure in first responses,
with 100% in Arabic, 78.1% in English and 71.5% in Arabic. SVAO occurred fairly
often in English (mean 10.4%) and French (mean 13.1%) but rarely in Arabic (mean
3.1%). SAVO and ASVO occurred rarely, with means respectively of 0% and 2.1%
in Arabic, 4.2% and 7.3% in English and 7.1% and 0% in French.
PAMAPLA 38 / ACALPA 38 (2014)
Figure 4
Responses to voice adverbs by structure and group
In Figure 5 are given the responses for the measure adverbs a lot and a little
by language and structure. As was the case with the voice adverbs, there are more
“No other way” responses and less variety in structures used.
Figure 5
Responses to measure adverbs by structure and group
SVOA was again the preferred structure in first responses in the three
languages: 87.5% in Arabic (mean 40.6%), 62.5% in English (mean 24.2%) and
64.3% in French (mean 23.8%). SVAO occurred more frequently in English (mean
11.1%) and French (mean 11.9%) than it did in Arabic (mean 1.0%). SAVO was
more frequent in Arabic (mean 10.4%) and French (8.4%) than it was in English
(3.2%). ASVO occurred in no language but Arabic (mean 13.6%).
Table 5 summarizes responses in percentages by adverb class and language.
Shaded cells highlight the most frequent responses. For ease of exposition, often and
usually have been combined under “Imperfective”. Different responses to the two
adverbs were presented above and will be considered in the Discussion.
Generally speaking, in all three languages the four adverb positions were
produced. ASVO is frequent with imperfective adverbs in Arabic (31.19%) and
English (30.00%), but less so in French (10.27%). There are no ASVO responses with
measure adverbs in English and French, or with voice adverbs in Arabic and French.
SAVO is the most frequent response with imperfective adverbs in English (35.72%),
but less frequent in Arabic (18.33%) and French (19.56%). SAVO occurs with the
other adverb classes as well, with the exception of voice adverbs in Arabic. SVAO
also occurred in all three languages. In Arabic, responses ranged from 7.89% to
12.38% (M = 9.84, SD = 2.30, Mdn = 9.26); in English, the range was 5.71% to
28.07% (M = 18.06, SD = 11.36, Mdn 20.40) and, in French, it was 22.68% to
27.03% (M = 24.46, SD = 2.27, Mdn = 23.68). SVOA is the most frequent response
for all adverb classes in the three languages, with the exception of imperfective
adverbs in English. In Arabic, responses with SVOA ranged from 38.09% to 90.74%
(M = 62.94, SD = 26.45, Mdn = 60.00); in English, the range was 28.57% to 55.10
(M = 41.34, SD = 13.29, Mdn= 40.35) and, in French, it was 47.62% to 60.53 (M =
54.02, SD = 6.52, Mdn = 54.05).
Table 5
Mean responses by language and adverb class. Note that for usually in English ASVO
and SAVO were preferred equally (37.5% each).
PAMAPLA 38 / ACALPA 38 (2014)
The research question guiding the current study was whether speakers of Emirati
Arabic would transfer the property of transportability of adverbs in their L1 to L3
French, or whether they would transfer fixed adverb position from L2 English to L3
French. If there is L1 transfer, participants would produce sentences exhibiting
ASVO, SAVO, SVAO and SVOA order with all adverb classes. If there is L2
transfer, adverb order would be quite rigid in the L3 and would reflect the L2 order.
In this case participants would produce SAVO for imperfective adverbs and SVOA
for voice and measure adverbs, but none of the other orders, with the exception of
souvent ‘often’, where ASVO and SVOA might be produced. If there is L2 transfer,
participants would not produce SVAO, which is ungrammatical in English.
4.1 Transportability
As reported in the Results section there were more second and third responses on the
Arabic task than there were on the English and French tasks, which suggests that the
transportability of adverbs was not transferred from the L1 to the L3. That being said,
in all three languages participants provided more second and third responses with
imperfective adverbs than they did with voice and measure adverbs. These additional
responses meant that participants used all four word orders with imperfective adverbs
in all three languages. With measure adverbs, ASVO was not produced in English or
French, and with voice adverbs ASVO was not produced in Arabic or French. SAVO
did not occur with voice adverbs in Arabic.
In Arabic, with imperfective adverbs the second and third responses were
almost always sentences with the adverb in different positions, while with measure
and voice adverbs many second and third responses were due to word order of the
subject and verb, with the adverb typically in clause-final position, i.e. VSOA or
VOSA as well as SVOA. These word orders, with the verb in sentence-initial
position, never occurred in English or French.
4.2 Imperfective adverbs
Participants treated the adverbs often and usually differently in Arabic and English.
The preferred word orders, that is, the largest frequency of first responses, were
respectively SVAO and ASVO in Arabic and SAVO and SAVO/ASVO in English.
(The SAVO/ASVO word orders were produced equally frequently in preferred
responses.) In French both adverbs had the same preferred word order, SVOA. As
noted in Section 1.2, in English usually cannot appear in clause-final position while
often can, but the equivalents of both adverbs can occur clause-finally in Arabic and
French. In all three languages the percentage of SVOA responses was almost the
same for the two adverbs, although it is ungrammatical with usually in English. The
participants had recognized that the two adverbs were different in English,
demonstrated by their different preferred word orders, but they had not learned the
ungrammaticality of SVOA with usually. It is unclear whether SVOA in French was
due to transfer from Arabic and/or English, or whether it was an L3 rule. The latter
possibility is unlikely, however, since White (1990) showed that adverbs were
infrequent in the input in French L2 classrooms.
4.3 Voice and measure adverbs
Participants made a clear distinction between imperfective adverbs on the one hand
and voice and measure adverbs on the other. Responses are much more consistent
with over 60% SVOA for first responses in all three languages. There were fewer
second and third responses and less variety in these responses.
Participants showed sensitivity to adverb classes and did not over-generalize adverb
placement from one class to other classes. This is particularly striking in L3 French,
since the learners had had only 20 hours of exposure to the L3 when the task in that
language was administered. Learners were cautious about transferring “marked”
ASVO position with imperfective adverbs from L2 English to L3 French. This
finding is contra Zobl (1992), who showed that L3 learners were less conservative
than L2 learners. Freer adverb placement with imperfective adverbs in English is
likely due to input and teaching in the L2 rather than L1 transfer, since the responses
to the two adverbs were different in the two languages.
Balcom and Bouffard (2010) showed that Anglophone L2 learners of French
preferred SVOA with imperfective adverbs, as did the French L3 learners in this
study. In the case of the Anglophones, it could not be due to transfer from the L1,
where SAVO is the preferred order. Balcom and Bouffard concluded that SVOA
marked a developmental stage during which learners know that the position of
imperfective adverbs is not pre-verbal in French, but have not been exposed to
enough input to learn that in French the adverb occurs between the verb and object.
The very similar results in L3 French, on a different task, are likely due to the same
process in the development of the Interlanguage grammar. Input with intransitive
sentences might also contribute to this Interlanguage rule, which reflects neither the
L1 nor the L2.
The SVOA responses in L3 French are a problem for Cinque`s theory. Cinque
(1999, 2004) maintained that all adverbs are base-generated in Aspect projections
between VP and Tense, and he claimed that clause-final position is due to movement
of the verb and object to the left of Aspect. Cinque is unclear as to what might trigger
movement of these elements out of VP to a position above the adverb, or indeed what
position the elements move to. Cinque’s analysis is problematic for L2 English and
particularly L3 French, since learners with only 20 hours of exposure would not have
acquired those triggers. The Cumulative Enhancement Model (Flynn et al. 2004)
proposed that in L3 acquisition there can be positive transfer from the L1 or the L2; it
may be that the participants in this study had transferred the full IP tree, including
Aspectual projections and VP from the L1 or the L2. The results of this study do not
allow a determination to be made concerning which language would be the source of
VSOA and VOSA were frequent in L1 Arabic, but these structures were never
produced in English or French. Research with learners at a lower level of proficiency
in English might show an initial transfer of freer word order from L1 Arabic to L2
English, but a later rejection as learners realized these word orders do not occur in
English. Participants in this study, who were intermediate to advanced users of L2
English, had learned that English had fixed word order, and they likely transferred
this knowledge to L3 French. Psychotypology might have played a role, since English
PAMAPLA 38 / ACALPA 38 (2014)
and French are similar in terms of their orthography and culture, and both were
learned in an academic environment.
1. With apologies to Omar Khayyam.
2. Images for the picture-description task were produced by Kerri George of Image en
Action Studio.
3. The format in Table 2 is based on Trahey (1976).
Abeillé, Anne and Danièle Godard. 2004. The syntax of French adverbs without
functional projections. In Martine Coene, Gretel De Cuyper, and Yves D'hulst
(eds.). Current Studies in Comparative Romance Linguistics. Antwerp Papers in
Linguistics 11: 1-42.
Alexiadou, Artemis. 1997. Adverb Placement: A Case Study in Antisymmetric Syntax.
Amsterdam: John Benjamins.
Aljenaie, Khawla and Mohammad Farghal. 2009. Comprehension of three word
orders in Kuwaiti Arabic child language. Language Sciences 31: 494-506.
Balcom, Patricia and Paula Bouffard. 2010. The initial state in L2 and L3 acquisition
of adverbs and complex negation: How long does it last? Poster presented at the
Canadian Association of Applied Linguistics Annual Conference, Montreal.
Bardel, Camilla and Ylva Falk. 2007. The role of the second language in third
language acquisition: The case of Germanic syntax. Second Language
Research 23: 459-84.
Cinque, Guglielmo. 1999. Adverbs and Functional Heads. Oxford: Oxford University
Cinque, Guglielmo. 2004. Issues in adverbial syntax. Lingua 114: 683-710.
Costa, João. 2004. A multifactorial approach to adverb placement: Assumptions,
facts, and problems. Lingua 114: 711-753.
De Angelis, Gessica. 2005. Interlanguage transfer of function words. Language
Learning 55: 379-414.
De Angelis, Gessica and Larry Selinker. 2001. Interlanguage transfer and competing
linguistic systems in the multilingual mind. In Jasone Cenoz, Britta Hufeisen, and
Ulrike Jessner (eds.). Cross-linguistic Influence in Third Language Acquisition:
Psycholinguistic Perspectives, 42-58. Clevedon, England: Multilingual Matters.
Fassi Fehri, Abdelkader. 1997. Layers in the distribution of Arabic adverbs and
adjectives and their licensing. In Elabbas Benmamoun, Mushira Eid, and Niloofar
Haeri (eds.). Perspectives on Arabic Linguistics XI, 9-46. Amsterdam: John
Flynn, Suzanne, Claire Foley, and Inna Vinnitskaya. 2004. The cumulativeenhancement model for language acquisition: Comparing adults' and children's
patterns of development in first, second and third language acquisition of relative
clauses. International Journal of Multiculturalism 1: 3-17.
Holes, Clive. 1989. Colloquial Arabic of the Gulf. Routledge: New York.
Juffs, Alan. 1996. Semantics-syntax correspondences in second language acquisition.
Second Language Research 12: 177-221.
Kellerman Eric. 1977. Towards a characterisation of the strategy of transfer in second
language learning. Interlanguage Studies Bulletin 2: 58-145.
Rivero, Maria-Luisa. 1992. Adverb incorporation and the syntax of adverbs in
Modern Greek. Linguistics and Philosophy 15: 289-331.
Rothman, Jason and Jennifer Cabrelli-Amaro. 2010. What variables condition
syntactic transfer? A look at the L3 initial state. Language Learning 26: 189-218.
Shaalan, Saleh. 2010. Investigating Grammatical Complexity in Gulf Arabic
Speaking Children with Specific Language Impairment (SLI). PhD Dissertation,
University College London.
Trahey, Martha. 1996. Positive evidence in second language acquisition: Some longterm effects. Second Language Research 12: 111-129.
Vainikka, Anne. 2007. Adverb movement in Organic Syntax.
Vinet, Marie-Thérèse. 1995. Adverbial quantifiers, negation, and stress rules effects.
University of Venice Working Papers in Linguistics 5:115-138.
White, Lydia. 1990. The verb-movement parameter in second language acquisition.
Language Acquisition 1: 337-360.
Zobl, Helmut. 1992. Grammaticality intuitions of unilingual and multilingual
nonprimary language learners. In Susan M. Gass and Larry Selinker (eds.).
Language Transfer in Language Learning (rev.ed.), 176-196. Amsterdam: John
PAMAPLA 38 / ACALPA 38 (2014)
Cape Breton University
The Dictionary of Cape Breton English is a regional dictionary that defines words and
senses that are “native” to Cape Breton Island or are “distinctively characteristic” of Cape
Breton English and Atlantic Canada. After a brief consideration of the scope of the
dictionary, this paper examines how selected citations and the largest semantic domains
of the terms defined in the dictionary reflect the social values of various communities on
Cape Breton Island. Like the Dictionary of Newfoundland English and the Dictionary of
Prince Edward Island English, the Dictionary of Cape Breton English gives special
attention to words used in local industries that have become part of the regional lexicon.
Terms from various semantic domains – industry, food, entertainment and social
activities, and religion and beliefs – demonstrate some of the ways that the dictionary
reflects the interests and values of various Cape Breton communities.
Richard MacKinnon and I started work on the Dictionary of Cape Breton English
(DCBE) in 1993, expecting to spend ten years or so, but as often occurs with dictionary
projects, it has taken longer than expected. Like the Dictionary of Newfoundland English
(2nd ed. 1990) and the Dictionary of Prince Edward Island English (1987), ours is a
regional dictionary that seeks to collect and define the words and senses that are either
native or distinctively used in an island community and in the Atlantic region. After a
brief discussion of the scope of the dictionary, this paper examines how selected citations
(dated quotations) and the main semantic domains of the terms defined in the dictionary
reflect social values of various Cape Breton communities.
The DCBE follows Walter Avis’s principles of collection as he articulated them in his
Introduction to the Dictionary of Canadianism (1967, rpt. 1991). Avis (1967:xiii)
identifies two types of words, those that are “native” and those that are “distinctively
characteristic.” By “native” he means those words that originate in the region or have
exclusively survived in the area. In Cape Breton English, suête (or les suêtes) is native as
the word describes the southeast winds that often reach hurricane force along the
northwest coast of Cape Breton Island.1 Occurring in a French-speaking area, suête is a
cursive form of sud-est. Other native words include the card game tarabish, two
meanings of stripper (one designating a crane that removes molds from steel ingots and,
the second, a cow that gives milk during the winter months) and several modified by
Cape Breton, such as Cape Breton fiddling, Cape Breton Silver (moonshine), and others.
In a regional dictionary, however, the more frequent type of term is “distinctively
characteristic.” Characteristic indicates that a term must have sufficient evidence that it
PAMAPLA 38 / ACALPA 38 (2014)
is used in the region represented by the dictionary. Distinctive indicates the word or sense
is limited in usage to a particular region or regions, and not part of “common vocabulary”
(Avis 1967:xiii). These distinctive terms are regional, but do not originate, nor are they
exclusively used, in Cape Breton English. For instance, tittle and tickle both designate a
narrow, dangerous saltwater passage. Although both terms are used in Atlantic Canada,
most North Americans would be unfamiliar with them.
In addition to adopting Avis’s general principles of selection, we have also
accepted a limited number of terms that are known both in and beyond Atlantic Canada.
These terms are frequently represented in our citation files or survey evidence, and have
cultural or historical importance in the development of Cape Breton Island. This type of
term is similar to what Stefan Dollinger, Laurel J. Brinton, and Margery Fee, editors
working on the second edition of the Dictionary of Canadianisms, describe as
“‘culturally’ significant terms.” They explain that this approach is more inclusive
allowing for “terms that are of greater significance to Canadians than to those of other
nationalities” (Dollinger et al. 2012:171-172). They note this broader approach is similar
to the practices of national dictionaries from Australia and South Africa. In the Atlantic
region some fishing terms are shared along the Atlantic coast and beyond, such as flake
or fish flake, a raised wooden frame, covered with spaced boughs and used to dry salted
fish, typically cod. This sense is strongly represented in our files with forty-four citations,
and it is recorded in dictionaries from Canada, Britain, and the USA. Because of its
importance to the economic development of Cape Breton and Atlantic Canada, the word
is defined in our dictionary, but knowledge of this word stretches beyond the DCBE’s
usual regional limits of Cape Breton Island and Atlantic Canada.
Citations in a regional dictionary serve several purposes including attestation, dating the
earliest known record of a term, and illustration of meanings, but they also offer glimpses
into the social values and interests of the region. Like the Dictionary of Newfoundland
English and Dictionary of Prince Edward Island English, our dictionary includes terms
that have migrated from the work place into regional vocabulary. On Cape Breton Island,
this is especially true of mining, fishing, and steel making. As one would expect, the
citations illustrating these terms provide insights into the working lives of Cape
Bretoners. At the headword split, the reader learns that fishers in 1906 received one dollar
for every hundred pounds of fish that they had gutted and split by removing the backbone
so the fish would lie flat, and that a trawl line (a long line with baited hooks used to catch
ground fish) was usually two and a half miles long but could be as long as five miles (at
the headword tub of gear). Several of the citations for coal mining terms mention that the
miners would travel up to five miles under the sea to reach their work place or the wall
face (the vertical coal seam), and that trapper boys (who opened and closed doors in mine
passages to trap and direct the air flow) might begin work underground as young as nine
years. Coal trimmers, who worked to balance (or trim) the coal in the holds of ships,
would occasionally be buried under the coal while it was being loaded (at the headword
nut coal). These details from citations in the dictionary sound like headlines from a
tabloid magazine, but they were the facts of life for those working in these industries.
In addition to these glimpses into the working lives of Cape Bretoners, some of
the citations for industrial terms also reveal other aspects of peoples’ lives. The citations
for the entry pit prop, a vertical support for the mine roof, exemplify aspects of the
island’s economic and social history.
pit prop noun; also prop; compare pit timber, timber
a trimmed log, six to eight feet long and approximately six inches in
diameter, used as a vertical support for the mine roof; later, a steel post
similarly used
1903 VERNON CB Canada 135 The most important uses to
which it [timber] is put are for pit-props for the mines . . .. 1920 Beaton
Institute MG 12 46 A9 n.p. Timber used in Mines year ending Sept. 30,
1920 . . . Props . . . 496, 236. 1974 LOTZ and LOTZ CB Island 97 In
1969, Cape Breton produced a little over a million cubic feet of rough pit
props . . .. 1986 CB’s Magazine 41, 2 See, in the wintertime they cut
timber for the mine–booms and pit props. Pit props were for Glace Bay.
That was part of your farming. 1988 Beaton Institute Tape 2381 . . . pit
props would be piled . . . and Dunlap’s [a general store] would exchange
the wood for groceries or merchandise or money. 1990 Beaton Institute
Tape 3004 They used to ship car loads from River Denys, pit props they
called them. They’re sticks of wood eight feet long about six to eight
inches in diameter, and they ship car loads of that down to the mines in
Sydney. 1996 CB Post 1 June 3A When the lake ice was thick enough the
pit props would be loaded on to a steel shod sleigh and hauled across to
[EDD, OED3 (1794), W3; at prop: EDD, OED3 Mining (1613), W3]2
Along with the factual details about the size of the pit props (6 to 8 feet long with
a diameter of 6 to 8 inches), the citations also refer to the economic spinoffs outside of
the mines. In 1920 the mines used nearly half a million pit props (Beaton Institute MG 12
46 A9 1920). Despite the introduction of iron props, nearly fifty years later, this number
had reached “a little over a million cubic feet of rough pit props” (Lotz and Lotz
1974:97). Social patterns are also illustrated. Many farmers would use the winter months
to cut props as a cash crop: “See, in the wintertime they cut timber for the mine-booms
and pit props. Pit props were for Glace Bay. That was part of farming” (Caplan 1986:2).
The citation at 1988 reveals that in small rural communities they used to barter,
exchanging props for merchandise, or receive money (Beaton Institute Tape 2381, 1988).
The citations for the Boston States reflect social attitudes and migration patterns
in nineteenth- and early twentieth-century Cape Breton.
Boston States, the noun, plural
1. New England, occasionally Boston
1927 ARCHIBALD The Token 11 Well, Angus man, the sealin’
trip may be dangerous, but it’s not so dangerous as the trip to the Boston
States . . .. 1948 MACNEIL Highland Heart 23 . . . Rory had a daughter
who some years before had gone to Boston, often known in Cape Breton as
PAMAPLA 38 / ACALPA 38 (2014)
the Boston States. Boston in those days was the Valhalla of all ambitious
Washabuckt [sic] boys and girls who ventured out into the world to better
themselves. 1951 MACLENNAN Each Man’s Son 161 “. . . once when he
wass buying a suit of clothes in the Boston States the man in the store
asked wass they too loose, so Captain Livingstone swelled out the muscles
of hiss back and shoulders and burst every seam in them just to show a
little of what he could do.” 1975 BIRD NS Historical Quarterly 5 (4), 326
Many of them [visitors] were from the Boston States (as the New England
States used to be called). 1988 MACDONALD Eyestone 34 I left Boston,
left that place your kin fled to over the generations, the Boston States, land
of plenty. 1994 GILLIS Travels 106 She taught school in Port Hood for a
few years and then moved to “The Boston States” in the early 1920s. Aunt
Mary worked in Boston as a domestic (maid), at a salary that was more
than double what she had made as a teacher in Cape Breton. 1994
MACLEOD and ST. CLAIR Pride of Place 59 Work in the “Boston
States” was akin to a finishing school. 2007 ST. CLAIR and LEVERT
Nancy’s Wedding Feast 14 She determined that the best thing for her
fatherless children was for her to leave them with their grandparents and
go to work in Boston, Massachusetts, “The Boston States,” as many still
refer to New England. 2007b BARBER Only in Canada 95 Boston States
(Maritimes & Nfld) New England.
2. United States
1951 MACLENNAN Saturday Night 3 July 11 People no longer
refer to the great republic as “the Boston States.” 1974 LOTZ and LOTZ
CB Island 103 They left for the ‘Boston States’– the Cape Bretoner’s term
for the United States . . .. 2006 MACISAAC Better Life 124 Women who
were good letter writers would sometimes embellish the wonders of their
life in the “Boston States,” a name given to all the American states by the
Highland Scots of eastern Nova Scotia. Survey I Inf. 2 . . . down in the
States. Survey I Inf. 20 The States.
[Sense 1: DPEIE, DC Maritimes (1948), COD Cdn (Maritimes & Nfld),
NCD Maritimes, SSPB. Sense 2: DPEIE (1939); compare “In Nova Scotia,
the United States of America are the Boston states . . ..” ML]
In Atlantic Canada, the Boston States usually meant New England, occasionally
Boston itself, and rarely the United States in general. Beyond this geographic reference,
many Atlantic Canadians saw the Boston area as offering a “better life” with economic
prosperity and social elegance. Like Alberta and Saskatchewan today, the Boston States
were a magnet for young people – women working as domestics and men in factories.
The citation at 2007 tells the story of a widowed mother making the difficult choice to
leave her children with their grandparents so that she could better provide for their futures
(St. Clair et al. 2007:14). A citation from Gillis recalls an aunt who moved to the Boston
States to work as a domestic “ . . . at a salary that was more than double what she had
made as a teacher in Cape Breton” (Gillis 1994:106). In addition to the lure of more
money, MacLeod and St. Clair (1994:59) observe the Boston States were “ . . . akin to a
finishing school,” and Neil MacNeil (1948:23) describes it as “the Valhalla of the
ambitious” young people.
Many of the Cape Breton community histories tell stories of young people visiting
home from the Boston States dressed in the latest fashions and having new fangled ideas
about indoor plumbing. Neil MacNeil (1948:23-24) tells one such story set in
Washabuck, probably occurring in the late 1800s. The daughter of Little Rory Donald
Dhu returned from the Boston States for a visit with her well-to-do husband. She
discovered that her family home still lacked both indoor and outdoor toilets. MacNeil
comments: “The natives of Washabuckt [sic] did not miss sanitary facilities or think them
necessary, for their ancestors had survived in good health and vigor for centuries without
bathrooms and lavatories” (23). At his daughter’s request, Rory built a backhouse (an
outhouse), but after his daughter left, he tore it down as his neighbours would point at
him saying “There goes proud Rory what does it in a box” and nicknamed him “Little
Rory the Backhouse” (23). The story illustrates the possible social friction between urban
and rural practices, and suggests why both social and language change was a conservative
process, especially in rural areas.
The semantic domains of the terms defined in the dictionary also demonstrate how the
entries reflect social values of Cape Breton. Table 1 gives an overview of the main
domains in order of the frequency of entries. The following discussion considers only the
five largest groups and explores some of the thematic patterns within these groups.
Table 1 Main semantic domains of words defined in the Dictionary of Cape Breton
English and their frequency
100-110 entries
mining; fishing/fishery
20-39 entries
food; entertainment and social activities;
religion and beliefs; music, humour and irony; steel
11-19 entries
clothing; ice; alcohol, farming and rural life; historical
reference; terms of address; houses, rooms and related
terms; weather
fewer than 10
lumbering; vehicles; physical actions; topography;
health and disease; trees and other plants;
personal qualities
4.1 Mining terms
With the closure of its mines and steel plants, Cape Breton Island is today a postindustrial society, but 250 years of mining have influenced the general vocabulary of
those living in the former industrial areas, which include not only the mines in Cape
PAMAPLA 38 / ACALPA 38 (2014)
Breton County but also the former mining areas in Inverness, Mabou, Judique, and other
locations in Inverness County.3 The terms pit socks and pit boots are typical examples of
how mining terminology has become part of community lexicon. It is a common
occurrence to hear stories of Cape Bretoners who have asked store clerks in Ontario for
pit socks or pit boots and have received blank stares from the clerks who know them as
work socks and work boots. As the citations for these terms indicate, these terms are
frequently used outside of the mines or pits. These two mining terms are part of the
remembered heritage of this industry, as are some of the early mining methods.
The earliest type of coal extraction, called room and pillar mining or hand pick
mining, has been memorialized in the surviving vocabulary as found in published and
tape recoded accounts, and by institutions like the Glace Bay Miners’ Museum that
recreated an underground mine from the early 1930s. Using this early method, two
partners, or buddies, worked in a room (a tunnel approximately 14 feet wide cut into the
coal seam) to extract the coal at the coal face (the vertical coal seam). The miners would
leave a pillar (a solid block of coal 40 to 60 feet square) between the rooms to support the
roof (the mine’s ceiling), thus giving the name room and pillar to this type of mining.
The miners would first mine or cut the coal (that is, make a narrow channel in the coal
seam with pick axes, or later with specialized machinery) to create a two-foot channel
under or beside the coal face to allow the coal to break when the shotfirer (a miner
trained in explosives and mine safety) would shoot, or blast, the coal. The buddies would
then load the broken coal into coal boxes with pan shovels (short handled shovels with a
heavy, round blade the size of a pizza pan). Pit ponies, which were used in some mines
until 1960 (MacNeil 1990:29), then pulled the coal boxes on rails to the deep (a sloping
haulage tunnel that follows the coal seam) to be pulled to the surface or bankhead (a
multi-storied building near the mine’s entrance where coal boxes were emptied).
Although this work was backbreaking, dirty, and dangerous, it has become part of the
local folklore and vocabulary in the former industrial areas. Later mechanization gave
words like continuous miner, punching machine, and undercutter,4 but it is this early
method that has captured the interest of many writing about Cape Breton mining.
The often turbulent relationship between workers and management has also
generated lasting vocabulary. The company – frequently referred to as a single
monolithic corporation but in fact had several corporate owners during its long history –
owned and operated the famous company stores that offered credit to miners for groceries
and other merchandise until the mid 1920s. Because so many miners were in debt to the
stores, the miners distrusted them and called them the pluck-me stores, believing their
pay was plucked from them by check-offs (or deductions). Frequently, the miners would
receive very little or no cash on pay day, and consequently named their pay slips the
bobtail paysheets or bobtails, which, like the tail of a bobcat, was short.
The degree of distinctiveness of these mining terms, and other work terms
discussed below, varies. Terms like pluckme store and bobtail paysheets are closer to
being Cape Bretonisms than the strictly technical terms. Many of the terms like pit props,
slope, and deep are regional terms known not only in Cape Breton English but also in
other areas where coal mining occurs. However, as Atkins and Rundell (2008:227)
explain, a regional label indicates that “the item is mainly but not exclusively used” in a
particular country or area. Many fishing and steel terms in the dictionary would be known
in regions other than Cape Breton with similar industries. In this same vein, the
Dictionary of Prince Edward Island English defines green as “of a ranch-raised fox,
having a coat that is not yet ready for pelting,” but surely people living in areas where
foxes are raised for pelts would be familiar with this sense of the word, as would trappers
who cure animal pelts. At the same time, however, work-related terms are also regional
as they are not part of “common vocabulary” because their particular meanings would be
unfamiliar in areas without activities such as coal mining, commercial fishing or fox
4.2 Fishing terms
Many fishing terms have also become part of the vocabulary in coastal areas on the
island. The rich fishery along the Atlantic coast drew Europeans to North America first as
seasonal fishers and then later as settlers to exploit the fish stocks during the years that
the island was a French colony and then later an English possession. The longest lasting
technology was the dry fishery for ground fish and is frequently discussed in the
published texts and tape recordings on the fishery. In some areas on Cape Breton Island,
the dry fishery was practiced as late as the 1960’s (Gallant 2002). In the dry fishery, the
fishers unloaded their catch on a stage (a wharf-like platform), often with the stagehead
over the water. The fish were gutted, headed, and split by removing the sound bone (that
part of the backbone near an air bladder called the sound) so the fish could be laid flat.
The fish were then salted, rinsed, and later placed on flakes (raised platforms with spaced
bows) to dry. Salted fish is still available in supermarkets, but it is now a specialty item
rather than a staple during winter.
The lobster fishery has also generated several terms for the various parts of the
lobster trap itself and the different ropes used to haul the traps. As Rose Mary Babitch
(1996) found in her study of fishing vocabulary on the islands of Lamèque and Miscou in
New Brunswick, these terms vary in different fishing areas. The trap typically used in this
area has two sections, the kitchen, where the lobster enters to take the bait, and a second
chamber – called the parlour, jail, or prison – where it becomes trapped. A bridle, a short
rope permanently tied to the kitchen end, is used to hold the rope (often called the snood)
that raises and lowers the trap in the water. As the season begins, the wooden traps need
extra ballast so that they sit still on the ocean floor; these are called dummy rocks in
certain areas.
Specialized boats were used for different types of fishing. Lobster smacks and
well boats transported live lobsters in special holds where water circulates. Others
include the Newfoundland jacks (small schooners built in Newfoundland and used in the
Cape Breton fishery), snapper boats (used in the sword fishery), stemmers (12 to 20 foot
boats used in the inshore fishery and powered by sail or oars), and flats (small boats with
flat bottom).
4.3 Food terms
In addition to these two major categories of mining and fishing terms, several other
semantic domains contributed words. Food terms are among the most productive of the
groups, and several of these are Cape Bretonisms. Cape Breton pork pies are, in fact,
small tarts with date filling, and no longer contain pork, although the citations
accompanying the definition speculate that the pork refers to the pork lard once used in
PAMAPLA 38 / ACALPA 38 (2014)
the pastry. Fat Archies are thick, soft molasses cookies – often buttered like a biscuit –
and Cape Breton steak is a variant of the traditional joke where the steak turns out to be a
less expensive alternative; in Cape Breton it is baloney. The two desserts are highly
prized, and Cape Breton steak illustrates the self-deprecating Cape Breton humour.
Other food terms are known regionally among older speakers or are loanwords.
Baker’s fog derisively refers to light and insubstantial baker’s bread. Krim-ko was once a
brand of chocolate milk available in the Maritimes, Ontario, and some northern US states,
but the word has survived among Cape Bretoners in their fifties as a generic term for
chocolate milk. Loanwords have also contributed to this group, illustrating the cultural
values attached to the ethnic foods. Five come from Gaelic: isbean and marag are homemade sausages, fuarag is a mixture of cream and oatmeal still typically served in some
places at Halloween, and strupag and tea bheag identify tea usually served with a snack.
The Mi’kmaq have a quick bread called lu’sknikn, and fricot (an Acadian stew or hearty
soup made with potatoes and meat, typically chicken) is an Acadian French loanword.5
4.4 Terms for entertainment and social activities
Another category, entertainment and social activities, often reflects the co-operation and
community spirit of the early settlers. For example, the term frolic is a gathering of
volunteers to complete a specific task and is often followed by a social event with food,
music, dancing and - yes - drinking. Although the term frolic probably originated in the
United States, what makes the word distinctive in Cape Breton is the twenty-four
varieties of frolics that are listed as variants (but not defined) at frolic: barn, carding,
chopping, cutting, digging, fulling, haying, hooking, house, milling, mowing, planting,
ploughing, quilting, reaping, rolling, spinning, stumping, tucking, waulking, weaving,
wood; and by analogy, cockfighting and ferret frolic. The reasons for working on these
tasks varied. At times someone might be incapable of working because of injury or
illness; other tasks required many hands like barn raising, haying, or a chopping frolic;
and others were an opportunity to relieve the boredom of a repetitive activity like
spinning wool while chatting with friends. We defined the milling frolic separately as this
activity of fulling/thickening homespun cloth was the most popular of the frolics and has
generated several related words: harrow or milling board, milling song, pairing song,
putting up song, tucking, and waulking. Once the milling was finished, often in the early
morning, pairing songs were sung, as a singer would create an extemporary song to pair
an eligible participant with a potential partner (Dunn 1953:40). Milling frolics are reenacted today as tourist and cultural activities to keep the tradition and the songs alive.
This sense of community is also reflected in various fundraisers like the popular
pie social and box social where males would bid on a dessert or a decorated box with
food enclosed in order to share the food and an evening of dancing with the female
contributor. Box social is an example of a term that is culturally and historically
important, as it is well documented in several American dictionaries, while pie social is
found in the Dictionary of Prince Edward Island, the Gage Canadian Dictionary, and the
Dictionary of American Regional English. Similarly, parcel posts, teas, and pound
parties raised money or food for schools, hospitals, and those in need.6 I recently noticed
parcel post mentioned in a church bulletin for a fundraiser in the town of Inverness.
4.5 Terms associated with religion and beliefs
Religion and beliefs are another productive source of regional terms. One of the most
remarkable examples from formal religion is the Presbyterian open-air communions that
flourished on Cape Breton Island during the second half of the nineteenth century and
early twentieth century (Stanley-Blackwell 2006.). Imported from Scotland, the five-day
communion service was conducted out of doors, as no churches or buildings were large
enough to house the two to five thousand participants. Because of its popularity, several
terms have remained current. On Question Day (or ceist day, a Gaelic loanword), a small
group of the pious and spiritually minded elders, called the men, explored a question to
reveal the spiritual merits of a selected scriptural passage. On other days, the ministers
would speak from a tent - actually, a wooden shed-like structure with a window facing
the congregation seated on a hill overlooking the minister. Because musical instruments
were not used in early Presbyterian services and printed copies of the psalms were not
available for such a large congregation, the precentors would lead the singing of psalms
by singing a verse, which was repeated by the congregation. Many of the sermons on
Saturday were aimed at fencing the tables; that is, defending the communion tables from
unworthy recipients. Those eligible to receive communion would be given a token: “ . . .
only those finally deemed worthy by elders, and by themselves, were issued a
communion token which permitted them to receive the bread and wine” (St. Clair et al.
In addition to terms from formal religion, others were generated by beliefs in
spirits, ghosts, and fairies. The Gaelic loanword bochdan designates a ghost and is often
used in the phrase bochdan stories that tell about ghosts that frighten people and spook
horses travelling over bridges or through wooded areas. MacKenzie (2005:18) writes that
the Irish settlers were more frightened by the bochdans than by bears. Fairies were often
designated by loanwords: lutin from French and the duine beaga (the little people) from
Gaelic. Forerunners, drokes or taibhs7 foretold future events and deaths. For example, a
person hearing (but not seeing) coffin wood being delivered to a neighbour’s home might
consider this as a prediction of death, or a light observed near a person’s home might
foretell that someone would soon die.
The Acadian French communities have also contributed some terms derived from
traditions associated with religion. The feast of Candlemas (2 February) or Chandeleur
became more of a community celebration than a religious feast. Volunteers collected
food for the party and performed the Escaouette (a lively dance and song) in the home of
the potential donors. The word Escaouette also identified the leader of the dance and now
is the name of a summer festival held at Cheticamp. Mi-carême is a similar example of a
community celebration that began as a mid-Lenten tradition where disguised visitors
entered homes of neighbours and family to see if the host family could guess the
identities of the disguised visitors. Now, the displays of masks and mask-making
activities are available during the summer months for interested tourists to learn about the
PAMAPLA 38 / ACALPA 38 (2014)
The citations and the semantic domains of words defined in this or any regional
dictionary illustrate the connection between language and society. Like the Dictionary of
Newfoundland English and the Dictionary of Prince Edward Island English, the
industrial terms have become part of the regional vocabulary in communities near the
work areas, and they give insight into the working conditions and the economy of the
island community. This past summer, for instance, I was delighted to hear a friend use the
term pit socks to describe her “bed socks.” I am sure she has never seen the inside of a
mine but uses the word as a natural part of her speech.
Although this paper has discussed only a few of the semantic groups found in the
dictionary, Table 1 suggests the wide range of connections with various sorts of social
activities. Many of the traditional social gatherings like frolics are certainly well
memorialized in the surviving vocabulary, but many traditions have bridged the gap
between the past and present. The co-operative spirit of the settlers reflected in the frolics
continue today as family members help each other build their homes, and Cape Bretoners
still generously support various fundraisers. The traditional activities of milling frolics,
the Escaouette, the mi-carême, ceilidhs and kitchen parties were once for family and
neighbours, but by both lexical and social extension these activities are now available to
the public as part of cultural tourism. Local Gaelic, Acadian, and historical societies
organize these sorts of events, thereby keeping alive the traditions and terms. Similarly,
many traditional food items are still available. Historian and journalist Jim St. Clair
(2010) explained that fuarag (the mixture of cream and oatmeal mentioned above) is still
served at the Inverness Consolidated Hospital on Halloween.
Finally, the lexemes drawn from the domain of religion and beliefs touch on
another aspect of society. There are many ways of finding comfort and meaning when
bad things happen to good people – or even when good things happen to bad people.
Formal religion and superstitious beliefs are two ways of coming to terms with hardships
and losses. The annual open-air communions were the religious highlight in many
Presbyterian communities and provided spiritual encouragement and religious instruction
(often in the form of notes or memorable statements). Like weather forecasts, the various
types of forerunners often predicted “bad news,” but there is some comfort in having
warnings of death or major change. For the minor challenges of life, like fresh milk going
sour, beliefs may provide an outlet for anger and disappointment by blaming the loss on
the lutin or the duine beaga, those mischievous fairies that like to mess in human affairs.
1. On the west coast of Newfoundland, similar topographic features and winds from the
southeast regularly cause hurricane force winds, and they too have also received a
specific name, the Wreck House Winds. Here and below, the italicized terms indicate
headwords from the DCBE. The accompanying explanations are derived from the
dictionary’s definitions but are abbreviated and informal, and so they do not appear in
quotation marks unless they are taken directly from the dictionary.
2. In many instances, the dictionary entries conclude with “dictionary lists” (enclosed in
square brackets) that list the abbreviations of dictionaries that share a meaning similar to
that found in the DCBE. Eighteen dictionaries have been regularly checked for each
headword, and other foreign language dictionaries have been used for loanwords from
Gaelic, Acadian French and Mi’kmaw. The dictionary list also provides the earliest
attested date among the dictionaries and an indication of the term’s distribution beyond
Cape Breton English. The full list of these abbreviations and their titles is given in the
3. The term company store (discussed below) is one of the mining terms found on the
western side of the island. In a taped interview, Anne Sherrington commented that a Coop store in Judique was known as the company store: “You could have credit for the
whole season there. In fact, you had to because you wouldn’t get paid until you
harvested. So they referred to that as the company store” (DCBE Tape 2007:28).
4. These terms are defined as a continuous miner “a machine that breaks the coal with
tapered metal teeth on a rotating track and transfers it to a conveyor belt in a single
process”; a punching machine “a pneumatic coal cutting machine with a sharpened bar or
bit used to undermine (or cut) the coal face to allow room for the coal to break once
blasted”; and an undercutter “a machine that cuts or punches a narrow channel along the
bottom or sides of the coal being excavated (the coal face), to allow the coal to break
after the explosive charge (the shot) is fired.”
5. Although one might expect that fricot is widely used, of the eighteen dictionaries in the
dictionary list regularly checked for headwords (see note 2 above), only the Canadian
Oxford Dictionary (2nd ed. 2004) records this meaning and has the regional label
6. These terms are defined as parcel post “a fundraising activity where people bid on
wrapped packages without knowing their contents”; tea “a fund-raiser, often in the
afternoon or early evening, with tea, a light meal, and frequently a sale”; and pound party
“a party to which guests bring a pound typically of some food item as a donation, often
followed by a dance; also, the admission fee for a dance.”
7. These terms are defined as forerunner “an omen, heard or seen, foretelling a death and
occasionally other events,” droke “an omen, heard or seen, that foretells a death,” and
taibhs “a spirit, vision, or sign foretelling an event, a forerunner.”
Atkins, B. T. Sue and Michael Rundell. 2008. The Oxford Guide to Practical
Lexicography. New York: Oxford University Press.
Avis, Walter S. 1967, Rpt. 1991. Introduction. In Walter S. Avis, Charles Crate, Patrick
Drysdale, Douglas Leechman, Matthew H. Scargill, and Charles J. Lovell (eds.). A
Dictionary of Canadianisms on Historical Principles, xii-xv. Toronto: Gage.
PAMAPLA 38 / ACALPA 38 (2014)
Babitch, Rose Mary. 1996. Le vocabulaire des pêches aux îsles Lamèque et Miscou.
Moncton: Éditions d’Acadie.
Beaton Institute Tape 2381. 1988. Rita Gillis and Vera Ratchford. Cape Breton Women
Working at the Sydney Steel Plant.
Caplan, Ronald (ed.). 1986. Dan Alex MacLeod: A Working Life. Cape Breton's
Magazine 41: 2.
Dollinger, Stefan, Laurel J. Brinton, and Margery Fee. 2012. Revising “The Dictionary of
Canadianisms on Historical Principles”: A Progress Report, 2006 – (April) 2012.
Dictionaries Journal of the Dictionary Society of North America 33:164–78.
Dunn, Charles W. 1953. Highland Settler: A Portrait of the Scottish Gael in Nova Scotia.
Toronto: University of Toronto Press.
Gallant, Gerald. 2002. DCBE Survey II.
Gillis, Rannie. 1994. Travels in the Celtic World. Halifax: Nimbus.
Lotz, Jim and Pat Lotz. 1974. Cape Breton Island. Vancouver: Douglas, David, and
MacKenzie, A.A. 2005. The Neighbours are Watching: An Extraordinary Entertainment.
Wreck Cove, NS: Breton Books.
MacLeod, Mary K. and James O. St. Clair. 1994. Pride of Place: The Life and Times of
Heritage Homes in Cape Breton. Sydney, NS: University College of Cape Breton
MacNeil, Jack. 1990. History of Sydney Mines: A Centennial Project. Sydney Mines, NS:
Town of Sydney Mines.
MacNeil, Neil. 1948. The Highland Heart in Nova Scotia. New York: Scribner's.
M[anuscript]G[roup] 12, 46. Duncan Seward Morrison. Diaries and Papers, 1847-1953.
Beaton Institute, Cape Breton University, Sydney, NS.
St. Clair, Jim. Email to Bill Davey. 27 Oct. 2010.
St. Clair, James O., Yvonne LeVert, and Peter Rankin. 2007. Nancy's Wedding Feast and
Other Tasty Tales. Sydney, NS: Cape Breton University Press and CBC Radio.
Sherrington, Anne. 2007. DCBE Tape.
Stanley-Blackwell, Laurie C. 2006. Tokens of Grace: Cape Breton's Open-Air
Communion Tradition. Sydney, NS: Cape Breton University Press.
APPENDIX: Abbreviations of Dictionaries Cited in the Dictionary of Cape Breton
Cassell’s Dictionary of Slang. 1998. Jonathon Green (ed.). London: Cassell.
Canadian Oxford Dictionary. 2nd ed. 2004. Katherine Barber (ed.). Toronto: Oxford
University Press.
Concise Oxford English Dictionary. 11th ed. 2009. Catherine Soanes and Angus
Stevenson (eds.). New York: Oxford University Press.
A Dictionary of Americanisms on Historical Principles. 2 vols. 1951. Mitford M.
Mathews (ed.). Chicago: University of Chicago Press.
A Dictionary of American English on Historical Principles. 4 vols. 1938-44. William A.
Craigie and James R. Hulbert (eds.). Chicago: University of Chicago Press.
Dictionary of American Regional English. 6 vols. 1985-2013. Frederic G. Cassidy [vol.
1], Frederic G. Cassidy and Joan Houston Hall [vols. 2-3], Joan Houston Hall [vols.
4-5], and Joan Houston Hall with Luanne von Schneidemesser [vol. 6] (eds.).
Cambridge, MA: Belknap Press of Harvard University Press.
A Dictionary of Canadianisms on Historical Principles. 1967. Reprinted 1991. Walter S.
Avis, Charles Crate, Patrick Drysdale, Douglas Leechman, Matthew H. Scargill, and
Charles J. Lovell (eds.). Toronto: Gage.
Dictionnaire du français acadien. 1999. Yves Cormier (ed.). Saint-Laurent, QC: Éditions
Dictionary of Newfoundland English. Second Edition with Supplement. 1990. G. M.
Story, W. J. Kirwin, and J. D. A. Widdowson (eds.). Toronto: University of Toronto
Press. URL:
Dictionary of Prince Edward Island English. 1988. T. K. Pratt (ed.). Toronto: University
of Toronto Press.
Dictionary of the Scots Language Dictionar o the Scots Leid. 2004. Susan C. Rennie
(ed.). Dundee, Scotland: University of Dundee. URL:
PAMAPLA 38 / ACALPA 38 (2014)
The English Dialect Dictionary. 6 vols. 1898-1905. Reprinted 1986. Joseph Wright (ed.).
Oxford: Oxford University Press.
Funk and Wagnalls Canadian College Dictionary. 1989. Walter S. Avis (editor,
Canadian Edition) and Sidney I. Landau (editor-in-chief). Toronto: Fitzhenry and
Gage Canadian Dictionary. 1998. Gaelan Dodds de Wolf, Robert J. Gregg, Barbara P.
Harris, and Matthew H. Scargill (eds.). Vancouver: Gage.
A Pronouncing and Etymological Dictionary of the Gaelic Language: Gaelic-English,
English-Gaelic. 1925. Reprinted 1979. Malcolm MacLennan (ed.). Aberdeen: Acair
and Aberdeen University Press.
The Gaelic-English Dictionary. 2004. Colin Mark (ed.). London: Routledge.
Mi’kmaw-English Lexicon. 2007. Mi’kmaw Kina’matnewey Board of Directors (eds.).
Eskasoni, NS: Eskasoni Centre of Excellence.
Micmac Dictionary. 1996. Albert D. DeBlois (ed.). Mercury Series Canadian Ethnology
Service, Paper 131. Hull, QC: Canadian Museum of Civilization.
Université de Moncton
À l’aide des théories sur le système de la politesse, cette recherche a pour but de décrire
les stratégies verbales utilisées par les cinq participants lors du débat politique national
télévisé des chefs de 2008 au Canada, alors qu’ils devaient suivre la consigne particulière
de flatter l’un de leurs adversaires. L’analyse démontre l’utilisation de mécanismes
communs et divergents entre eux, au niveau des actes langagiers flatteurs et des actes
langagiers menaçants. Cette contribution originale nous éclaire davantage sur les
stratégies de politesse dans les débats politiques, plus notamment lors d’une situation de
« double contrainte » très prononcée.
Using theories about politeness, this paper aims to describe the verbal strategies used by
five political leaders during the national leaders’ debate televised in Canada in 2008. The
leaders were asked to make positive comments about one of their opponents. The analysis
shows shared and divergent mechanisms, including both Face Flattering Acts and Face
Threatening Acts. This original contribution provides a better understanding of politeness
strategies in political debates and, in particular, in the very explicit situation of “double
La communication occupe une place centrale dans l’univers des politiciens. Avec la
dimension médiatique qui s’est imposée dans cet univers de la communication, la façon
de faire les choses force ces acteurs publiques à ne plus rien laisser au hasard. Par
exemple, bien paraître et se distinguer lors d’un débat télévisé peuvent entraîner des
effets décisifs sur une campagne électorale. Par contre, ce passage en direct peut aussi
s’avérer un terrain glissant. Le travail est exigeant et il n’y a que peu de « damage control
» pouvant être fait quand un imprévu survient devant la caméra.
Lors d’une campagne électorale fédérale récente, on a pu vivre en direct justement
un épisode difficile. Alors que les chefs débattaient et essayaient de marquer des points
depuis près d’une heure, en plein milieu du débat, une question surprise est venue
chambarder le déroulement normal des choses :
Bonjour. Vous êtes les leaders politiques du Canada et votre travail est de vous
assurer du mieux-être et du bien-être des Canadiens, au-delà des chicanes
partisanes. Dans ce contexte, pouvez-vous nommer au moins un bon coup ou une
qualité de l’adversaire qui se trouve à votre gauche ?
PAMAPLA 38 / ACALPA 38 (2014)
« Question intéressante… » se permet de dire l’animateur. Les réactions devant la
caméra nous suggèrent que la nouvelle consigne a pris tous les candidats par surprise.
Visiblement déstabilisés, les chefs autour de la table s’agitent. Certains se redressent sur
leur chaise. L’un se gratte nerveusement, alors qu’un autre prend une gorgée d’eau… Un
candidat signale même à l’animateur son incompréhension de la question. Témoin de ce
malaise général, l’animateur minimise la « gravité » de la question, qu’il qualifie
d’« intéressante » en riant sur un ton léger. Bien qu’ils connaissent les thèmes à l’avance,
les chefs n’avaient visiblement pas prévu une consigne aussi précise. Ils semblent agacés,
voire tiraillés, à l’idée de suivre la consigne en donnant ainsi gratuitement des points à
leur adversaire – et on comprend : cela va à l’encontre même de leur objectif respectif. Ils
ne sont pas là pour donner des points à leur adversaire, mais bien pour leur en rafler le
plus possible (Kerbrat-Orecchioni 2011 : 40).
La « double contrainte », telle que définie par Kerbrat-Orecchioni (1996 : 64) est
un conflit entre la valorisation de soi et celle d’autrui. Dans cette optique,
corrélativement, plus on se valorise au cours d’une interaction, plus on risque de
dévaloriser autrui. En même temps, plus on valorise autrui, plus on se place en position
basse ou de vulnérabilité. Nous expliquons davantage la notion de « double contrainte »
dans la prochaine section, où l’on présente brièvement le cadre théorique du système de
la politesse.
Dans le contexte de « double contrainte » fortement prononcée de cette partie du
débat des chefs de 2008, notre débat à l’étude – coincés entre leurs propres objectifs
personnels (gagner le débat) et celui de devoir se plier à la nouvelle contrainte pour bien
paraître –, quelles stratégies les chefs ont-ils pu employer pour à la fois donner des points
à leur adversaire et éviter d’en perdre ? Comment ont-ils pu tirer leur épingle du jeu et
allier à la fois l’acte d’attaquer leur adversaire et celle de le flatter ? Voilà la question à
l’étude et à laquelle cette recherche vise à répondre. Pour y arriver, nous nous fixons
comme objectif de décrire les rouages au niveau des Face Flattering Acts et des Face
Threatening Acts retrouvés dans le débat à l’étude. Cela nous permettra du même coup de
discerner les stratégies communes et celles divergentes. Notre démarche cherche ainsi à
relever les stratégies ayant permis à chacun des chefs de maintenir sa face et de se
construire malgré tout une place, c'est-à-dire de marquer des points malgré la consigne de
flatter son adversaire. En plus de démystifier un pan de la communication en politique,
nous espérons par nos résultats mettre en lumière certains mécanismes du langage en
interaction concernant la gestion de la « double contrainte » dans le système de la
politesse, aspect très peu étudié dans l’analyse des débats politiques.
Cette analyse est abordée à l’aide d’outils développés dans le cadre théorique de la
politesse, qui est présenté brièvement dans la partie qui suit.
1.1 Cadre théorique du système de la politesse
Intégrée à l’analyse des données naturelles en tant que système régulateur, la politesse est
composée de matériaux du langage – autant verbal que paraverbal et non verbal –
omniprésents au cours d’une interaction, qui à la fois reflètent la relation interpersonnelle
et agissent sur elle. Elle est définie comme suit par Kerbrat-Orecchioni (2005 : 189) :
« […] l’ensemble des procédés conventionnels ayant pour fonction de préserver le
caractère harmonieux de la relation interpersonnelle, en dépit des risques de friction
qu’implique toute rencontre sociale. »
Ferron et Kasparian
Selon Goffman (1967), les participants engagés dans une interaction obéissent
normalement au principe du respect de la face (« face-work »). Ce principe est régi par un
ensemble de micro-rituels conventionnels auxquels les participants ont recours tout au
long d’une interaction, alors qu’ils tentent à travers leurs comportements de maintenir
leur face ainsi que celle de leur partenaire : « la face est donc un objet sacré, et il s’ensuit
que l’ordre expressif nécessaire à sa préservation est un ordre rituel » (Goffman 1967 :
21). Le bon déroulement d’une interaction exige donc normalement un certain travail de
coopération, faute de quoi on risque de perdre la face. On comprend que la face est
indissociable des émotions, qui se manifestent à travers le langage. Comme le dit
Traverso (1999 : 58) : « […] pourquoi en effet consacrer tant d’attention à la préservation
des faces, sinon parce que la perte de face est une blessure, fortement marquée sur le plan
émotionnel ? »
En s’inspirant des notions de Goffman, les linguistes Brown et Levinson (1987)
ont développé un cadre théorique selon lequel tout individu possède deux faces : la face
négative – qui correspond en gros au territoire corporel, biens matériels, savoirs secrets,
et cetera – et la face positive – qui correspond plutôt au narcissisme, aux images de soi
valorisées par l’individu et revendiquées au cours d’une interaction (Kerbrat-Orecchioni
1996 : 51-54). Selon eux, la présence d’un acte de langage peut à tout moment venir
menacer l’une des faces de l’un des participants, ce qu’ils appellent des Face Threatening
Acts (FTA) et qu’ils classent en quatre catégories : (1) actes menaçants pour la face
négative de celui qui les accomplit (promesse, engagement, etc.) ; (2) actes menaçant
pour la face positive de celui qui les accomplit (aveu, excuse, autocritique, etc.) ; (3) actes
menaçants pour la face négative de celui qui les subit (offenses proxémiques, agressions
visuelles, sonores ou olfactives, questions indiscrètes, interdiction, conseil, etc.) et (4)
actes menaçants pour la face positive de celui qui les subit (critique, réfutation, moquerie,
insulte, sarcasme, etc.). Kerbrat-Orecchionni spécifie tout de même que s’il existe des
FTA dans le langage en interaction, il y a aussi des Face Flattering Acts (FFA), comme
le compliment ou le cadeau. Deux types de manifestation linguistique de la politesse
peuvent alors être produites : celle « positive », de nature productionniste de FFA, et celle
« négative » de nature abstentionniste ou compensatoire de FTA.
Il s’avère néanmoins parfois difficile de se sortir d’une impasse. Pour tenter de
nous éclairer davantage, Kerbrat-Orecchioni (1996 :64) fait référence à la notion de
« double contrainte », développée au sein de l’École de Palo Alto. Dans une situation de
« double contrainte », il y a un conflit entre l’altruisme et l’égoïsme. Les conventions de
la politesse servent à contrebalancer les pulsions égocentriques et à neutraliser le pire
pour l’interaction. Le tout réside habituellement dans un état très subtil entre la
préservation de soi et le respect de l’autre. On se voit donc généralement à la recherche
d’un compromis, parfois en adoucissant nos expressions (en prenant des manières
embarrassées, etc.). Il nous arrive ainsi aussi, par exemple, de bémoliser un FFA qui nous
est destiné. C’est le système de la politesse qui nous oblige à ces « contorsions verbales »
si nous voulons maintenir notre rôle en société avec honneur.
Il arrive quand même de retrouver des situations où l’envie de se faire « mousser »
émerge dans une séquence, comme dans un match (Goffman, 1967 : 24). Le débat
politique télévisé représente bien un tel cas, où l’objectif est de marquer des points et de
disqualifier son adversaire. Dans la question portant sur le thème « Leadership et
gouvernance » au milieu du débat des chefs canadiens de 2008, nous retrouvons une
PAMAPLA 38 / ACALPA 38 (2014)
situation de match, mais dans laquelle s’insère une problématique particulièrement
prononcée de la « double contrainte », tel que décrit en introduction. Pour marquer des
points cette fois, la nouvelle contrainte fait en sorte qu’ils doivent en donner à l’autre – ce
qui va à l’encontre de l’objectif visé habituellement dans un débat télévisé. Comment les
chefs arriveront-ils à se valoriser devant la caméra, alors que la règle introduite par le
citoyen leur impose plutôt de donner des qualités à leur adversaire ? C’est cette situation
délicate à gérer qui nous a intéressés et que nous avons étudiée à travers les réactions
verbales et les réponses des chefs, ce qui constitue notre corpus, que nous présentons
avec notre méthodologie dans la partie qui suit.
Le corpus à l’étude est constitué d’une séquence du débat des chefs télévisé (en français)
de 2008 au Canada, soit les réponses des candidats à la question reliée au thème
« Leadership et gouvernance ». Ce débat est animé par le journaliste Stéphan Bureau et
on y retrouve cinq participants : Elizabeth May (Parti vert du Canada), Stephen Harper
(Parti conservateur du Canada), Jack Layton (Nouveau parti démocratique), Stéphane
Dion (Parti libéral du Canada) et Gilles Duceppe (Bloc québécois). Comme de coutume,
la place qui leur a été assignée a été pigée au hasard. On se souviendra que la formule
adoptée était plus conviviale, puisque les participants étaient tous assis autour d’une table
ovale. Cette année-là, le consortium des médias avait demandé aux citoyens d’envoyer
des questions dans une vidéo. Les thèmes étaient connus en avance par tous les leaders,
mais les questions précises n’ont été dévoilées qu’au moment où elles ont été posées en
Pour analyser notre corpus, la démarche transversale (Traverso 1999 : 26-27), qui
consiste à isoler dans le corpus une ou des variables définies à l’avance, est favorisée
dans cette étude afin d'isoler les FFA et les FTA. Afin de manier la lecture à notre guise,
nous nous sommes servis du logiciel de capture d’écran Camtasia Studio et nous avons
enregistré l’intégralité du débat, disponible sur le site web de Radio-Canada. Ensuite,
nous avons procédé à la transcription en nous inspirant des conventions établies en
linguistique par Losier et al. (2002). Nous avons inséré le corpus transcrit en annexe, à la
toute fin de cet article.
Nous procèderons tout d'abord, au point 3.1, à l’analyse de la séquence d’ouverture. Ce
passage illustre bien le tiraillement que la situation de « double contrainte » peut
entraîner. Nous verrons aussi d'ailleurs comment l'ouverture des interventions des chefs
est façonnée par ce contexte difficile au niveau du face-work. Ensuite, aux points 3.2 et
3.3, nous analyserons de façon détaillée la production de FFA et FTA au niveau de la
strate verbale, afin de montrer les stratégies utilisées par les chefs pour répondre à la
« double contrainte », soit d’allier l’acte de se valoriser à celui de valoriser leur
Ferron et Kasparian
3.1 Malaise général en situation de « double contrainte »
La toute première récurrence observable dans le corpus est le malaise, qui est visible chez
les chefs après l’écoute de la question. La présence du rire, ayant servi à minimiser la
menace que représente la consigne, est l’une des réactions significativement notables en
réponse à la nouvelle consigne. On constate que l’animateur lui-même produit un rire
minimisateur quand il présente la question (ligne 5) : « ((SB rit.)) Question intéressante,
pour laquelle vous étiez peut-être pas préparés. » Puis, avant le début des interventions,
Duceppe envoie une blague qui brise la glace pour atténuer l’atmosphère et fait rire tout
le monde (voir Extrait 1 : ligne 13) – du moins visiblement, à défaut de pouvoir se
prononcer sur Harper en raison de l’angle de la caméra :
Extrait 1
GD 1 : [Ah], je croyais que je parlais de vous, Monsieur Bureau : vous êtes à ma
SB 3 : [Alors, c’est–]
GD 2 : Vous posez bien les questions!
EM 2 : ((rire))
Si le rire contribue à atténuer l’atmosphère, le malaise se laisse quand même percevoir
chez chacun à travers d’autres indices, parmi les composantes linguistiques. Au niveau
verbal, on a des hésitations, entre autres aux lignes 49, 59, 85, 91, 96 et 124. En effet, par
exemple, Harper (voir Extrait 2 : ligne 68) et May (voir Extrait 3 : ligne 124) ont des
départs un peu plus laborieux, pouvant nous indiquer le malaise qu’ils ressentaient, tout
en pouvant gagner quelques secondes :
Extrait 2
((SH regarde JL et sourit.)) Monsieur Layton, Jack, eu… Je peux–  <Je
peux dire> des bonnes choses sur Jack
Extrait 3
EM 9 : Oui, je pense que Monsieur Harper et Monsieur Layton ont travaillé d–
125 fort ((EM regarde et pointe JL.)) pour  quelque chose ensemble, mais je laissais
ça pour le moment ((EM balaye l’air de la main.)) parce que j’arriveras ici. Je suis
très heureuse avec ça. ((EM regarde SH.)) Monsieur Harper, j’ai trouvé fort pour,
((EM gesticule.)) comme vous avez peut-être souviens, ((SH se frotte le poignet
nerveusement.)) pour trouver les choses que nous avons ensemble partagées.
Layton semble également ressentir une certaine gêne au départ, alors qu’il relâche un
sourire atténuateur de la menace envers sa face, quand l’animateur l’a placé sur la
sellette. À ce moment précis, en tant que bystanders, Dion et May aussi laissent aller un
rire atténuateur (voir Extrait 4 : lignes 46 et 48) :
PAMAPLA 38 / ACALPA 38 (2014)
Extrait 4
SB 6 : Mais c’est pas à vous encore. C’est Monsieur Layton ((EM rit.)) qui doit
dire du bien de son voisin de gauche, Monsieur Dion, en le regardant dans les
yeux. Ça va être beau, ça. ((SB rit.))
JL 1 : Mais voilà– ((SD rit.)) ((JL rit, se rapproche de SD et lui touche le bras.))
Le malaise manifesté au début de la séquence à l’étude vis-à-vis de la consigne ne
s’évacue pas par la suite. En effet, durant la séquence d’ouverture de chaque intervention,
tous les chefs ont utilisé des stratégies d’autodéfense. On relève entre autres le marqueur
d’opposition « mais » utilisé comme ouvreur, affichant d’emblée une opposition avec
l’idée de ne pas être capable d’envoyer un FFA à leur adversaire. On observe chez
Duceppe : « Mais, je pense que Madame May a des préoccupations au niveau de
l’environnement fort importantes. » (ligne 21) ; on note chez Layton : « Mais voilà– ((SD
rit.)) ((JL rit, se rapproche de SD et lui touche le bras.)) Voilà, c’est un professeur comme
moi. » (ligne 49). Selon Maingueneau (cité par Marchand 1998 : 115) : « Le mais de
réfutation récuse la légitimité de ce qu’un destinataire a dit ou pensé, pourrait avoir dit ou
pensé ». En effet, le connecteur « mais » agit comme stratégie de réfutation, ayant l’effet
de nier et de rectifier l'idée présupposée de ne pas être capable de relever le défi de flatter
son adversaire.
D’autres indices révélant une autodéfense ont été perceptibles dans notre analyse.
Harper, par exemple, a adopté un ton de voix plus aigu et hésitant pour dire « Monsieur
Layton, Jack, eu… Je peux– Je peux dire des bonnes choses sur Jack » (ligne 68). May
aussi a éprouvé des difficultés au départ, en raison de ses hésitations, prenant de la ligne
124 jusqu’à 127, avant de commencer à donner des FFA. Elle affirme en plus : « j’ai
trouvé (sic) fort pour, comme vous avez peut-être souviens, pour trouver les choses que
nous avons ensemble partagées » (ligne 127).
Dion, pour sa part, s’est défendu d’emblée de « mettre en cause la sincérité » de
Duceppe (voir Extrait 5 : ligne 94). Néanmoins, comme les autres, il laisse transparaître
des indices d’un malaise à l’idée d’envoyer des FFA à son adversaire : sa gorge se serre
quand il prononce son premier « oui » (Extrait 5 : ligne 91) et les mots « au contraire »
(Extrait 5 : ligne 96). Il serre aussi ses pouces quand il laisse aller son FFA (Extrait 5 :
ligne 98) :
Extrait 5
SB 8 : Monsieur Dion, vous nous avez
parlé récemment de ce que vous étiez aussi, vous, fier nationaliste québécois.
SD 1 : <Oui. >
SB 9 : Vous pouvez dire ça dans le bleu des yeux à Gilles Duceppe ((EM rit.))
((GD sourit.)) en vantant ses mérites. ((SB rit.)
SD 2 : ((SD acquiesce de la tête.))  Oui, certainement, c’est pas moi qui ((SD
jette un regard vers SH.)) va mettre en cause la sincérité ((SD regarde GD et
hoche la tête.)) de Monsieur Duceppe. <Au contraire>, je pense qu’il veut
vraiment tout faire pour aider les Québécois et le Québec, ((SD joue avec ses
Ferron et Kasparian
Des FFA servant surtout ses propres objectifs
Malgré le malaise des candidats à l’idée de donner des qualités de leur adversaire, les
chefs doivent répondre à la consigne. Ils finissent donc par produire quelques FFA.
Néanmoins, on relève plusieurs formes de FFA : les directs, les inclusifs (du locuteur), les
détournés vers soi-même, ceux valorisant explicitement sa propre face, les bémolisés ou
affaiblis et, finalement, ceux servant à envoyer subtilement de façon détournée un FTA
au premier ministre, la réelle cible.
Les FFA directs destinés à leur adversaire sont rares. On en retrouve à quelques
reprises, par exemple : quand Duceppe dit que May a « des préoccupations
environnementales fort importantes » (ligne 21), quand Layton dit à Dion qu’il est
« honnête » et « intelligent » (ligne 51), lorsque Harper dit apprécier que Layton est
« honnête envers le débat » (ligne 85), ou encore lorsque May dit à Harper qu’il est un
« bon père » (ligne 130). On remarque toutefois que les FFA directs produits par Harper
et Dion ne sont arrivés qu’à la clôture de leur intervention, ce qui pourrait être interprété
comme des actes réparateurs de leur déviance de la consigne. Concernant les FFA directs,
il est intéressant de constater qu’ils concernent généralement les qualités personnelles
plutôt que politiques.
Les FFA bémolisés sont ceux qui sont suivis d’atténuateurs qui ont pour effet
d’affaiblir l’acte de langage flatteur. Ce procédé est surtout utilisé par Dion, par exemple
quand il dit que Duceppe a un sens de l’État « quand il est à son meilleur » (ligne 114).
May également se servira de cette stratégie, en disant à Harper qu’il est un homme de
principe, tout en ajoutant que les dits principes vont conduire le pays dans une « direction
dangereux (sic) » (ligne 135).
Les FFA inclusifs, c’est-à-dire ceux utilisés pour s’inclure parmi les destinataires
du FFA, sont plus nombreux que les FFA directs. Par exemple c’est le cas de Duceppe
quand il remercie May pour avoir eu des « bonnes paroles pour le Bloc » (ligne 37) ; le
cas de Layton quand il dit à Dion qu’il est un « professeur » comme lui (ligne 50) ; le cas
aussi de Harper quand il remercie Layton d’avoir coopéré « avec le gouvernement »
(ligne 78) ; et le cas de Dion quand il affirme viser la « même chose » que Duceppe, soit
de vouloir aider le Québec (ligne 100). Il est intéressant de souligner qu’il n’y a en fait
que May qui ne détourne jamais ainsi subtilement un seul FFA vers elle-même.
Les FFA produits envers soi-même ne sont pas toujours aussi subtils. En effet, les
chefs réussissent à insérer certains FFA qui sont bien explicites dans lesquels une certaine
valorisation de soi est très clairement énoncée. Voici des exemples : « il faut élire des
députés du Bloc » (Duceppe, ligne 41) ; « comme premier ministre, je vais le faire »
(Layton, ligne 61) ; « C’est ce que le Parti libéral pourra offrir » (Dion, ligne 111), etc. Si
Harper n’a pas été aussi explicite durant la première partie, il l’a été néanmoins durant la
deuxième, par exemple quand il fait son auto-éloge à la clôture du débat, en disant avoir
dirigé le plus long mandat d’un gouvernement minoritaire dans l’histoire du pays (ligne
317). Le premier ministre en a quand même profité souvent pour rappeler son rôle de
« gouvernement » dans la première partie, ce qui constitue un taxème qui lui est
particulier, comparativement aux autres chefs. Enfin, encore une fois, May est la seule à
ne pas s’auto-décerner de FFA direct ou valoriser son propre parti, durant toute la
séquence du débat.
Enfin, certains FFA donnés à leur adversaire ont pu être transformés en FTA pour
être détournés vers le premier ministre Harper, leur véritable adversaire. Deux chefs ont
PAMAPLA 38 / ACALPA 38 (2014)
surtout utilisé cette stratégie, soit Duceppe et Dion. Duceppe, par exemple, rappelle qu’il
a participé avec May à des événements importants où les experts « dénonçaient tous la
politique du gouvernement Harper » (ligne 27). Dion, lui, affirme que Duceppe a raison
« quand il dit qu’il faut pas que Monsieur Harper devienne majoritaire » (ligne 105). De
plus, au début de son intervention, le chef libéral soutient Duceppe en déclarant,
contrairement à Harper, ne pas remettre en cause sa sincérité, tout en jetant un regard
déictique accusateur envers Harper (ligne 95) : « Oui, certainement, c’est pas moi qui va
mettre en cause la sincérité de Monsieur Duceppe. »
Finalement, même s’ils ont produits des FFA, on peut voir que les chefs s’en sont
généralement servis à leurs propres fins. Ils ne se sont pas gênés non plus pour envoyer
des FTA, malgré la consigne.
3.3 Des FTA, malgré tout
Certains chefs ont profité de la consigne pour produire des FTA explicites, parfois envers
leur adversaire ou encore pour attaquer le premier ministre. Au niveau des FTA envers
leur adversaire, par exemple, Layton a clairement critiqué le bilan du Parti libéral de Dion
(ligne 52) : « Mais malheureusement, son parti a… pendant treize ans… pas un bilan,
eu… des étoiles, si vous voulez. » Pour sa part, Dion a également dit à Duceppe qu’il
manquait un peu d’ambition (ligne 107) : « il manque un peu d’ambition. » Il faut noter
deux choses importantes néanmoins. D’abord, les FTA relevés dans le corpus concernent
strictement un aspect politique ou idéologique de leur adversaire concerné. De plus, ils
sont toujours accompagnés de minimisateurs ou coussins atténuateurs (ex. : « un peu »,
une hésitation, etc.) et/ou même un rire complice (ex. : celui entre Dion et Duceppe,
quand il est dit de son adversaire qu’il manque d’ambition, à la ligne 107). Il y a bien
d’autres atténuateurs au niveau non verbal et paraverbal. Nous n’abordons ici que ceux au
niveau verbal. Aucun FTA n’est émis directement à la face de son adversaire sans qu’on
y décèle des indices d’un certain malaise ou un enrobage de politesse négative.
Certains chefs ont souvent dérivé délibérément de la consigne et se sont permis de
concentrer leurs attaques sur le premier ministre, la réelle cible. On peut ainsi retrouver
cette affirmation de Duceppe : « pour vaincre et empêcher une majorité de Monsieur
Harper, il faut élire des députés du Bloc québécois » (ligne 41). Il y a également celle-ci,
de la part de Dion : « C’est ce que le Parti libéral pourra offrir, eu… plutôt que les
conservateurs » (ligne 112). Il est toutefois intéressant de noter que ces FTA sont produits
alors que les chefs parlent de Harper, et non à Harper. Le message est quand même assez
clair pour tout le monde, y compris les téléspectateurs (qui sont en fait toujours le réel
destinataire visé par les candidats).
Dans cette recherche, nous avons voulu savoir comment les candidats aux élections
fédérales de 2008 au Canada, en plein milieu du débat télévisé des chefs en français, ont
pu utiliser des stratégies verbales relevant du système de la (im)politesse pour tirer leur
épingle du jeu, dans le contexte d’une question venant d’un téléspectateur qui les a mis
dans une situation de « double contrainte » fortement prononcée, les poussant à devoir
donner des points positifs à leur adversaire pour répondre à la consigne s'ils veulent
Ferron et Kasparian
gagner le débat, ce qui va à l'encontre des règles habituelles d'un débat politique télévisé,
soit de marquer des points pour soi et se faire valoir.
Tel que démontré à l’aide de notre analyse, les chefs produisent des actes flatteurs
(FFA) envers leur adversaire – ils doivent répondre au citoyen, à défaut d’être
disqualifiés en direct. Toutefois, il a été démontré clairement dans notre analyse que la
plupart du temps, ces FFA ont servi à atteindre leurs objectifs personnels. En effet, de
façon générale, ils s’en sont servis – quelquefois par le biais de la reprise dialogique –
pour s’inclure à l’occasion parmi les destinataires de leur FFA, ou encore pour s’allier
avec leur adversaire et attaquer finalement le premier ministre, l’ultime adversaire.
D’ailleurs, les chefs ne se sont pas gênés pour lancer des FTA malgré tout, y compris à
l’adversaire de leur gauche, contrairement à ce qu’exigeait la consigne. Il a néanmoins été
intéressant de constater que les FTA concernaient essentiellement l’aspect idéologique ou
partisan de leur adversaire, alors que les FFA utilisés pour répondre à la consigne
touchaient généralement leur côté plus intime ou personnel.
Enfin, les résultats obtenus révèlent des stratégies divergentes chez Elizabeth
May, qui avait sans doute le plus grand défi d’entre tous : flatter le premier ministre, celui
qui a été la cible principale d’attaques de tous côtés depuis le début du débat. Malgré tout,
elle s’est distinguée au niveau des FFA au cours du débat, que ce soit dans la première ou
dans la deuxième partie de notre corpus. En effet, elle est à peu près la seule à avoir
utilisé des FFA directs et francs (voir Extrait 6 : lignes 130 et 132), sans les détourner en
sa faveur :
Extrait 6
130 pense que vous êtes un… un bon piè– père pour les en– pour– avec vos enfants.
Les enfants : très charmants, très engagés. ((EM sourit à SH.)) Et je pense que vos
efforts pour le pays du Canada sont basés sur vos principes.
Si May n’a pas pu utiliser des alliances pour attaquer le premier ministre,
contrairement aux autres, elle a misé sur l’intégrité en lui lançant directement un FTA à la
ligne 135 (« je pense que vos– votre– vos principes va changer Canada dans une direction
dangereux »), qui disqualifie du même coup l’un des FFA qu’elle lui avait destiné, à la
ligne 132, au sujet de ses valeurs idéologiques : les efforts de Harper pour le pays sont
basés sur ses principes, mais ses principes sont « dangereux ». Elle s’est permis tout de
même une petite gêne et a atténué ses FTA à plusieurs reprises. Elle a, par exemple,
utilisé un euphémisme (« autocrate »), à la ligne 143, au lieu d’un mot plus dur (ex. :
« dictateur », « tyran », etc.). De plus, elle n’a pas hésité à la ligne 131 à enrober de
procédés de politesse positive les FFA qu’elle envoie à Harper au sujet de ses enfants et
de ses efforts pour le pays, avant de lui lancer les FTA.
Ce n’est pas seulement avec Harper que les stratégies de May ont été singulières.
Plus loin dans le débat (voir Extrait 7), on a pu la voir dans d’autres interventions
renforcer des FFA que Jack Layton s’administrait, en qualifiant d’« éloquent » le discours
de leur ancien chef de parti, et venir même carrément au secours de Gilles Duceppe en
difficulté – sans qu’elle ne soit sollicitée de le faire, alors même que l’animateur avait
placé le chef du Bloc québécois sur la sellette – en appuyant son idée pour le soutenir.
Personne d’autre durant le débat ne produit de FFA de façon aussi gratuite et altruiste,
surtout pas pour sauver la face de qui que ce soit.
PAMAPLA 38 / ACALPA 38 (2014)
Extrait 7
EM 15 : C’est le dernier discours de Monsieur Broadbent. Je le souviens très bien.
JL 5 : [Oui.]
EM 16 : C’est très, eu… éléqu– eu… éloquent ?
JL 6 : Oui.
EM 17 : Et il a rét– eu… souligné les droits humains dans le niveau de Nations
Unies. Il– Il y a le respect l’un à l’autre ici, dans cette Chambre des communes…
Nous manque le respect là. Je pense pour le- le leadership, c’est très important de
faire ce– comme Monsieur Duceppe a dit, devoir servir le public. Il faut d’être à
l’écoute des citoyens.
Elizabeth May a donc bien pu tirer son épingle du jeu, malgré l’ampleur du défi
qui l’attendait pourtant. En effet, son comportement honnête et poli dénote par rapport au
comportement des autres candidats. Se pose alors la question à savoir s’il y a un lien avec
le fait que May soit une femme. Son comportement est-il représentatif de stratégies
typiquement « féminines » ? De plus, y a-t-il une corrélation possible avec le fait que la
Parti vert soit le 5e parti à la Chambre des communes, donc un parti marginal, ayant ainsi
une plus grande marge de manoeuvre discursive ? Si l'on se réfère à ce qui est démontré
par plusieurs chercheurs, plus le discours se rapproche du pouvoir, plus il se « vide » de
contenu, devient « ramasse-tout », « neutre », etc. Ainsi, les petits partis auraient
tendance à être plus affirmatifs de leurs valeurs et n’hésiteraient pas à afficher plus
clairement leur position idéologique. En fonction des résultats obtenus, nous ne pouvons
que formuler des hypothèses à cet égard. Un regard dans la littérature autour de la théorie
du genre, ou encore du discours en marginalité, pourraient nous donner des pistes
intéressantes à suivre.
En plus des pistes mentionnées ci-haut, il serait intéressant d’élargir les horizons
de la recherche à d’autres situations de « double contrainte », un phénomène
particulièrement présent dans notre vie quotidienne en société. De telles analyses peuvent
s’avérer précieuses pour nous aider à mieux pénétrer les rouages du langage dans la
gestion de l’alter et de l’ego en interaction.
Brown, Penelope et Steven Levinson. 1987. Politeness: Some Universals in Language
Usage. Cambridge, R-U : Cambridge University Press.
Caillat, Domitille. 2012. La notion de temporalité au coeur de l’analyse des discours
représentés à l’oral : essai de typologie, procédés rhétoriques et portées
argumentatives des DR dans le débat présidentiel français de 2007. In F. Nevue, V.
Muni Toke, P. Blumenthal, T. Klingler, P. Ligas, S. Prevost et S. Teston-Bonnard
(éds.). SHS Web of Conferences, 3e Congrès Mondial de Linguistique Française 1 :
427-439. doi :
Ferron et Kasparian
Cosnier, Jacques et Catherine Kerbrat-Orecchioni. (éds.) 1991. Décrire la conversation,
2e éd. Lyon : Presses Universitaires de Lyon.
Goffman, Erving. 1967. Interaction Ritual: Essays in Face-to-Face Behavior. Chicago:
Aldine Publishing.
Jacquin, Jérôme. 2011. Orientation interactionnelle et multimodale vers l’activité de
« débattre ». Analyse d’un extrait de débat public. Mots. Les langages du politique
96. URL :
Jefferson, Gail. 1979. A technique for inviting laughter and its subsequent
acceptance/declination. In G. Psathas (ed.). Everyday Language: Studies in
Ethnomethodology, 79-96. New York: Irvington.
Kerbrat-Orecchioni, Catherine. 1990-1994. Les interactions verbales. Paris : Armand
Colin. Tome I, 1990 ; tome II, 1992 ; tome III, 1994.
Kerbrat-Orecchioni, Catherine. 1996. La conversation. Paris : Seuil.
Kerbrat-Orecchioni, Catherine. 2005. Le discours en interaction. Paris : Armand
Kerbrat-Orecchioni, Catherine. 2011. L’impolitesse en interaction : Aperçus théoriques et
étude de cas. Lexis Special 2 : « Impoliteness / Impolitesse », Journal: Studii si
Cercetari Filologice : Seria Limbi Romanice, Editura Universitatii din Pitesti 1 : 9.
Losier, Line, Kasparian, Sylvia et al. 2002. Guide de présentation de mémoires et de
thèses en linguistique et conventions pour la transcription de conversations.
Université de Moncton.
Marchand, Pascal. 1998. L'analyse du discours assistée par ordinateur. Paris : Armand
Radio-Canada. URL:
Sandré, Marion. 2011. Mimiques et politique. Analyse des rires et sourires dans le débat
télévisé. Mots. Les langages du politique 96. URL : ;
doi : 10.4000/mots.20203
Traverso, Véronique. 2007. L’analyse des conversations. Paris : Armand Colin.
Turbide, Olivier. 2009. La performance médiatique des chefs politiques lors de la
campagne électorale de 2003 au Québec. Thèse de doctorat, Université Laval.
PAMAPLA 38 / ACALPA 38 (2014)
APPENDICE: Transcription du corpus
[citoyen] : Bonjour. Vous êtes les leaders politiques du Canada et votre travail est
de vous assurer du mieux-être et du bien-être des Canadiens, au-delà des chicanes
partisanes. Dans ce contexte, pouvez-vous nommer au moins un bon coup ou une
qualité de l’adversaire qui se trouve à votre gauche ?
SB 1 : ((SB rit.)) Question intéressante, pour laquelle vous étiez peut-être pas
préparés. ((SD s’ajuste sur sa chaise et bouge ses papiers.)) ((JL se gratte le nez et
se frotte les doigts.)) ((GD bouge les doigts et se rajuste sur sa chaise.)) Alors,
vous êtes pas certaine d’avoir bien compris ?
EM 1 : Non, je n’ai pas bien compris. Pardon.
SB 2 : [Eu… Alors, eu… ] Vous allez voir, c’est assez simple. Il suffit de voir ce
que va faire Monsieur Duceppe, puisqu’il va parler de vous, madame May, et
GD 1 : [Ah], je croyais que je parlais de vous, Monsieur Bureau : vous êtes à ma
SB 3 : [Alors, c’est–]
GD 2 : Vous posez bien les questions !
EM 2 : ((rire))
SB 4 : Oui, j’aurais beaucoup aimé ça, mais malheureusement vous parlez de
madame May, qui est à votre gauche dans ce cas-ci. ((JL retouche ses manches.))
((SD joue avec son stylo.))
GD 3 : Mais, je pense que Madame May a des préoccupations au niveau de
l’environnement fort importantes. ((GD tourne son regard vers EM et penche son
buste dans sa direction.)) On a participé d’ailleurs ensemble, hein, à des
rencontres à Ottawa où on a honoré les scientifiques, hein, qui ont mérité ce prix,
avec Al Gore=
EM 3 : [Oui, oui.] ((EM acquiesce de la tête.))
GD 4 : =et qui dénonçaient tous la politique du gouvernement Harper. ((GD fait
un geste accusateur avec les index et penche son buste vers l’avant.))
EM 4 : Oui.
GD 5 : Et ça, j’en suis redevable à madame May qui a bien fait ce travail-là. ↑ Je
lui dis cependant – ↓ je pense qu’elle est d’accord avec moi ((GD regarde EM et
hoche de la tête de façon affirmative.)) ((EM acquiesce de la tête.)) –, ↑ si on veut
empêcher Monsieur Harper d’avoir une majorité, hein, bien, ((GD penche son
buste vers l’avant, en ouvrant les bras et les paumes de main tournées vers le
haut.)) il faut faire en sorte qu’on le batte. Et puis, je pense que notre programme
– ((GD tourne son regard vers EM, penche son buste vers elle et tend les bras vers
elle.)) → et vous l’avez déjà reconnu en disant que Bernard Bigras était le
meilleur critique en environnement puis le Bloc avait été le parti le plus solide en
environnement au parlement. Je pense qu’au Québec, ((GD commence à relâcher
un sourire en coin.)) pour vaincre et empêcher une majorité de Monsieur Harper,
il faut élire des députés du Bloc Québécois. Je vous remercie d’avoir dit et eu ces
bonnes paroles pour le Bloc. ((GD rit.))
EM 5 : ((EM rit.)) Oui. ((EM rit.)
SB 5 : Alors, vous avez compris l’exercice ?
Ferron et Kasparian
EM 6 : Oui, oui, oui.
SB 6 : Mais c’est pas à vous encore. C’est Monsieur Layton ((EM rit.)) qui doit
dire du bien de son voisin de gauche, Monsieur Dion, en le regardant dans les
yeux. Ça va être beau, ça. ((SB rit.))
JL 1 : Mais voilà– ((SD rit.)) ((JL rit, se rapproche de SD et lui touche le bras.))
Voilà, c’est un professeur comme moi. ((JL se vise avec ses mains, puis rapproche
son buste vers SD.)) C’est un homme honnête, intelligent. Mais malheureusement,
↓ son parti a… ((JL s’abaisse la tête.)) [ǂ] pendant treize ans… ((JL se rabaisse la
tête.)) pas un bilan, eu… ((JL mime des guillemets.)) des étoiles, si vous voulez.
↑ ((JL se redresse droit et produit des battements rythmiques de la tête.)) Mais ce
que j’ai apprécié, c’est que / on peut travailler ensemble. ((JL regarde Dion.)) On–
Et avec ↑ tous les partis, et ↑ tous les chefs ((JL ouvre grand les bras et regarde
tout le monde.)) ((SD se retire contre le dos de sa chaise.)), on a travaillé dans la
Chambre des communes d’une façon respectueuse. Et moi, je peux– ↑ je pense
que c’était le… l’essentiel de cette question. ((SD avance son buste et regarde.
JL.)) ((JL jette un bref regard régulateur vers SD.)) C’est comme (sic) le citoyen
nous demande de travailler pour les meilleurs résultats. ↑ Et comme premier
ministre, je vais le faire. Je vais rassembler les leaders pour avoir les discussions,
pour assurer de faire avancer les préoccupations ((JL produit un regard
synchronisateur envers SD.)) des familles. ((SD répond au synchronisateur de JL
en acquiesçant.))
SB 7 : Merci, eu… beaucoup. Notre prochain à se prononcer, c’est Monsieur
Harper sur Monsieur Layton.
SH 1 : ((SH regarde JL et sourit.)) Monsieur Layton, Jack, eu… Je peux– ↑ <Je
peux dire> des bonnes choses sur Jack ((SH lève la main droite et la rabaisse en
balayant.)), malgré nos grandes différences ((SH sourit et effectue des « va-etvients » kinémimiques avec ses mains.)) en… en philosophie, nous avons travaillé
ensemble sur des questions où nous sommes– ((SH produit des battements
rythmiques de la main droite et de la tête, en plus de pencher la tête vers JL, en le
regardant et en haussant les sourcils.)) où nous étions d’accord. ((JL hoche la
tête.)) Par exemple, la reconnaissance de la nation québécoise, ((JL hoche la
tête.)) les excuses pour les Aut– les écoles autochtones ((SH bat de la main
droite.)) où vous avez fait un très… ((SH tend la gauche bien ouverte vers JL.))
((JL hoche la tête.)) un très bon travail avec le gouvernement et je l’ai mentionné
à la Chambre des communes. ↑ Même sur la Loi sur la– l’imputabilité à la
responsabilité, ((SH tend la main vers JL.)) nous avons eu des différences, ((SH
effectue des « va-et-vients » kinémimiques avec ses mains.)) mais nous avons eu
beaucoup de… beaucoup de points communs. Et moi, je trouve en général que,
malgré nos différences, que=
JL 2 : Énormes.
SH 2 : =vous– ((JL rit.)) Vous êtes… Vous êtes honnête envers le débat et je
JL 3 : ((JL acquiesce de la tête.)) Merci beaucoup.
SB 8 : [Merci
], Monsieur Harper. Monsieur
Layton qui souligne ici les différences énormes. Monsieur Dion, vous nous avez
parlé récemment de ce que vous étiez aussi, vous, fier nationaliste québécois.
PAMAPLA 38 / ACALPA 38 (2014)
SD 1 : <Oui. >
SB 9 : Vous pouvez dire ça dans le bleu des yeux à Gilles Duceppe ((EM rit.))
((GD sourit.)) en vantant ses mérites. ((SB rit.)
SD 2 : ((SD acquiesce de la tête.)) → Oui, certainement, c’est pas moi qui ((SD
jette un regard vers SH.)) va mettre en cause la sincérité ((SD regarde GD et
hoche la tête.)) de Monsieur Duceppe. <Au contraire>, je pense qu’il veut
vraiment tout faire pour aider les Québécois et le Québec, ((SD joue avec ses
pouces.)) et– On a un désaccord sur la façon dont on doit s’y prendre ((SD bat des
mains et hoche de la tête, en regardant brièvement GD.)) ((GD acquiesce de la tête
et pointe ses mains vers SD.)), mais on vise la même chose ((SD détourne le
regard.)) ((GD se redresse sur sa chaise, incline la tête en direction de SD et en
maintenant le regard vers lui.)) En tant que chefs d’opposition, on a coopéré très
souvent. Ça a été une belle expérience pour moi. ((SD hoche de la tête.)) Eu…
Bien sûr, je trouve que Monsieur Duceppe a raison ((SD regarde Duceppe.))
quand il dit qu’il faut pas que Monsieur Harper devienne majoritaire. C’est une
chose– ((SD regarde GD en hochant la tête.)) ((GD acquiesce de la tête, sourit et
pointe SD avec ses pouces.)) Puis je trouve que là, il manque un peu d’ambition.
((SD agite les mains vers Duceppe, le regarde, penche son buste vers lui et
sourit.)) ((GD rit, mais se redresse de suite après sur sa chaise.)) En fait, on
devrait avoir un gouvernement ↑ progressiste qui va travailler pour ↑ tous les
Canadiens. C’est ce que le Parti libéral pourra offrir, eu… ((SD désigne SH avec
sa main.)) plutôt que les conservateurs. Donc, on peut faire un pas de plus ((SD
bat la main et la tête, en regardant GD.)) dans cette direction. Je voudrais ajouter
que Monsieur Duceppe, quand il est à son meilleur, ((SD serre les pouces, puis
gesticule des mains.)) a un sens de l’État et ça, c’est très apprécié <de ma part>.
((SD se serre les lèvres.)) ((GD incline la tête légèrement, horizontalement, puis
détourne le regard.))
SB 10 : Merci beaucoup, Monsieur Dion. Alors, vous terminerez le bal, madame
May. Vous avez compris les règles du jeu ?
EM 7 : [Oui.]
SB 11 : Et il reste une personne dont on n’a pas encore parlé.
EM 8 : [Oui.]
SB 12 : C’est Monsieur Harper, qui est à votre gauche.
EM 9 : Oui, je pense que Monsieur Harper et Monsieur Layton ont travaillé d–
fort ((EM regarde et pointe JL.)) pour ↑ quelque chose ensemble, mais je laissais
ça pour le moment ((EM balaye l’air de la main.)) parce que j’arriveras ici. Je suis
très heureuse avec ça. ((EM regarde SH.)) Monsieur Harper, j’ai trouvé fort pour,
((EM gesticule.)) comme vous avez peut-être souviens, ((SH se frotte le poignet
nerveusement.)) pour trouver les choses que nous avons ensemble partagées. Je
pense que vous êtes un… un bon piè– père pour les en– pour– avec vos enfants.
Les enfants : très charmants, très engagés. ((EM sourit à SH.)) Et je pense que vos
efforts pour le pays du Canada sont basés sur vos principes. Mais, évidemment,
((EM lève ses bras de chaque côté, en ouvrant la paume des mains vers le haut.))
((SH ferme les yeux brièvement et on entend quelqu’un relâcher un léger soupir.))
je– je pense que vos– votre– vos principes va changer Canada dans une direction
dangereux. Alors, aussi, aussi, c’est
Ferron et Kasparian
SB 13 : Mer…
EM 10 : C’est o– C’est– Malheureusement, je pense que votre sorte de
SB 14 : [Merci…]
SD 3 : [xxx] débat, là.
SB 15 : Alors– Alors–
EM 11 : =c’est plus autocrate. ((EM affiche une mimique faciale exprimant le
SB 16 : Merci, madame May. ((SD rit.))
EM 12 : Oui. Oui.
SD 4 : C’est pas le débat, là.
EM 13 : Oui. Oui.
SB 17 : Bien, le débat, on parle de leadership et moi je vais vous poser une
EM 14 : ((EM se tourne vers SH, le regarde et se rapproche de lui, en lui touchant
le bras.)) [xxx]
SD 5 : Bon. ((SD détourne son regard vers le stylo avec lequel il commence à
SB 18 : Cette question a été posée, Stéphane Dion. C’est qu’il y en a plusieurs qui,
quand ils voient les campagnes électorales, se demandent c’est quoi le ton
SD 6 : Oui. ((SD hoche la tête plusieurs fois.))
PAMAPLA 38 / ACALPA 38 (2014)
University of New Brunswick (retired)
In Hong Kong standard written Chinese is Mandarin while Cantonese, which is the
language of the vast majority of the territory’s population and is entirely acceptable when
spoken, has very low status in its written form. This poses a problem for newspaper
editors when quoting Cantonese speech. In the past, it was translated into Mandarin, but
in recent years there has been an increasing use of Cantonese. This article examines the
reporting of a speech in two different newspapers, the quality Ming Pao and the mass
circulation Apple Daily, to study whether there was variation in the amount of Cantonese
in the quotations. It was found that there was slightly less in Ming Pao than in Apple
Daily, though not as much as might have been expected. There was also a greater use of
quotation in Ming Pao, which again was not expected based on earlier research.
However, comparison with the actual words as recorded in a YouTube video revealed that
both newspapers still reduced the amount of Cantonese.
The present article discusses quotations in Chinese newspapers and, in particular,
examines an account, in two different newspapers, of remarks made by Li Ka-shing at the
annual shareholders meeting of Hutchison Whampoa Limited and Cheung Kong
Holdings. According to the business magazine Forbes, Li, who is the Chairman of the
two companies concerned, is the eighth richest person in the world as of March 2013,
with an estimated wealth of US $31 billion. This is of relevance since his status validates
Cantonese, which he often uses and which he also used in this meeting rather than
Mandarin or English.
Quoted speech in Hong Kong newspapers is of particular interest due to the
language situation in the territory. Cantonese is one of the three official spoken languages,
the others being Mandarin and English. However, there are only two official written
languages, Mandarin and English. Although Cantonese can be written, the written form
is looked down on as being low class, uneducated and crude, and has traditionally been
reserved for working class material, popular songs and light entertainment. It is not used
for serious matters. Spoken Cantonese does not suffer from this discrimination, and it is
widely used in Hong Kong in all areas, from the lowest to the highest ranks of society. It
is also the language most commonly used in the Hong Kong Legislative Council.
Nevertheless, it is interesting to note that Snow (2004) reports that in recent years there
has been an increasing use of Cantonese in Hong Kong newspapers. For a review of
Snow’s study, see Lister (2009).
The problem for newspaper editors is how to quote Cantonese speech, since
reproducing the speech verbatim could be seen as denigrating the speaker. An alternative
is to translate the quotation into Mandarin, but this risks losing nuances of meaning. A
PAMAPLA 38 / ACALPA 38 (2014)
third solution is to employ indirect speech. This allows much more leeway since the
speech can be summarized, and for that purpose the use of Mandarin is entirely
The problem does not exist across the border in the province of Guangzhou. Even
though Cantonese is widely spoken there, it is not an official language. All newspapers
are written in Mandarin, and speeches and talks are given mainly in Mandarin. Cantonese
speech, when it does occur, is translated into Mandarin. This is in line with the official
policy of the Chinese government, which is to promote Mandarin and to discourage the
use of dialects. In laissez-faire Hong Kong, in comparison, editors are free to publish in
whatever dialect they please, and they encounter no obstacles in using Cantonese. The
contrast between newspaper quotations in Hong Kong and in Guangzhou is quite
dramatic, the former appearing more colourful and authentic, the latter more official and
formal. It does appear also, from a cursory examination, that the frequency of quotations
is higher in Hong Kong newspapers than in the mainland press.
One of the problems encountered for this study was determining the original
language of a particular quotation. However, if there are any Cantonese characters in the
direct quotation, it is virtually certain that the original was Cantonese. If the quote is
entirely in Mandarin, the original may have been in English, Mandarin or Cantonese.
A further problem is defining exactly what is Cantonese and what is Mandarin.
Some terms are clearly Cantonese, and are represented by a series of characters that are
specific to Cantonese. However, some words exist in Mandarin also but with different
meanings, or are used in a different way, or occur far less frequently. Cantonese speakers
are also not always certain whether particular terms or words are also used in Mandarin,
and sometimes they will err on the side of caution. The best way to establish whether a
particular word is Cantonese is to ask native speakers of Mandarin whether they would
understand and/or use it, though this method is not completely foolproof. Another related
problem is that formal Cantonese is very similar to Mandarin, and there may be little or
no difference between the two in a given quotation.
In this paper, it is assumed that the speech studied was in Cantonese, since
Cantonese terms are scattered throughout the quotes. If it had been in English or
Mandarin, this would certainly not have been the case. Moreover, part of the speech was
broadcast on YouTube, including many of the quotations mentioned in this paper.
The amount of Cantonese in Hong Kong newspapers varies. Quality newspapers,
such as Ming Pao, have traditionally used less Cantonese than popular mass circulation
newspapers such as Apple Daily, and the pro People’s Republic of China newspaper Wen
Wei Po contains little or no Cantonese, as is the case with newspapers in Guangzhou. The
choice of language is both political and cultural. Ming Pao tends to use less Cantonese
than Apple Daily for cultural reasons, and it is considered to be among the best
newspapers for its use of Chinese.
According to a Media Credibility Survey conducted by The Chinese University of
Hong Kong in 2006, Ming Pao was selected as the most credible Chinese language
newspaper. It aims at providing comprehensive and accurate reports on political and
economic issues in Hong Kong and mainland China. Well known for its accuracy in
language, many secondary schools in Hong Kong encourage their students to subscribe to
Ming Pao to improve their Chinese language.
Wen Wei Po, on the other hand, avoids Cantonese in accordance with the Chinese
government's policy of promoting Mandarin and of discouraging the use of other
varieties of Chinese. The mass circulation Apple Daily uses Cantonese to identify more
closely with its readership and to demonstrate its opposition to China's language policy of
promoting Mandarin at the expense of regional dialects.
This study examines reporting of the speech made by Li Ka-shing to shareholders
on 5 August 2011. It compares the reports in two different newspapers, the quality Ming
Pao and the mass circulation Apple Daily.
2.1 Two examples of quotations in Ming Pao
There were 14 quotations in the Ming Pao article. One was repeated three times; one was
repeated twice; and there were 12 indirect quotations. No distinction is made between
different types of quotations, such as the three categories – direct quotations, iconic
quotations and iconic expressions - identified by So Fung Ming (1998)1. The first
quotation in the passage, given in (1), is the one that is repeated three times. Cantonese
words and expressions are in bold type.
(1) 「 要 識 得
做 嘢, 唔 係 剩
係 識
得 口 講」
‘Jiu3 sik1 dak1 zou6 je5,
m4 hai6 zing6 hai6 sik1 dak1 hau2 gong2.’
must know
do things, not is
mouth speak
‘He must know how to do things, and not only talk.’
The quotation is almost completely in Cantonese, probably because it was an
important statement by Li concerning one of the candidates in the election for Hong
Kong's Chief Executive. It was probably safer to report it word for word in the original
language rather than translate it into Mandarin. In addition, the sentence would also lose
some of its flavour and impact in translation.
In the quotation given in (2), there are only a few Cantonese words or
expressions. It is certainly not accurate reporting since the mixture of Mandarin and
Cantonese would be most unlikely in normal Cantonese speech. For example, the
possessive particle 的 (Mandarin de, Cantonese dik1), the verb to be 是 (Mandarin shì ,
Cantonese si6) , and the negative 不 (Mandarin bù, Cantonese bat1) are Mandarin, and in
Cantonese would be respectively 嘅 ge3, 係 hai6 and 唔 m4, while 點 解 dim2 gaai2
‘why’, is Cantonese for Mandarin 為什麼 wèishéme and Cantonese 成 日, seng4 jat6 ‘all
day’, in Mandarin would be 整天 zhěng tiān. 叻 lek1 ‘clever’ is a Cantonese adjective. In
Mandarin the character exists but is only used in place names.
(2) 「司 長
的 意思
是,你 點 解
成 日 complain(抱 怨),
‘Si1 zoeng2 dik1 ji3 si1 si6, nei5 dim2 gaai2 seng4 jat6 complain ( pou5 jyun3),
the chief
LP meaning is, you why
day complain (complain),
點 解
不 努 力 去
麼。 我
14 歲 打 工
dim2 gaai2 bat1 nou5 lik6 heoi3 hok6 se1 sam6 mo1 . Ngo5 14 seoi3 daa2 gung1
not strive
learn some things.
14 years work
PAMAPLA 38 / ACALPA 38 (2014)
負 擔
起 家 庭, 一 路 走
來, 我 不 是 最 叻
那 個,
fu6 daam1 hei2 gaa1 ting4 , jat1 lou6 zau2 loi4 , ngo5 bat1 si6 zeoi3 lek1 naa5 go3,
one way go come, I not am most clever that one,
我 走 的 路 是 可 以 有
就 的。」
daan6 ngo5 zau2 dik1 lou6 si6 ho2 ji5 jau5 sing4 zau6 dik1.’
go LP way is can
have success
‘The principal secretary's meaning was, why do you always complain, why do you
not energetically go and learn something. When I was fourteen, I worked to
support my family. From then until now, I was not the smartest, but the road I took
was one which could lead to success.’ [PRT=particle]
2.2 Two examples of quotations in Apple Daily
In the example from Apple Daily given in (3), half of the words or expressions are in
(3) 「 將
來 2017 年 你
哋 都
有 得
‘Zoeng1 loi4 2017 nin4 nei5 dei6 dou1 jau5 dak1 syun2,
2017 year you
also can
而 家
ji4 gaa1 gong2 zou2 se1 siu2
speak early somewhat little PTR
‘In the future in 2017, you can also vote, I only spoke now a little early.’
The second example, given in (4), contains no Cantonese at all and is interesting since Li
Ka-shing was quoting from a speech that he delivered to graduating students in Shantou,
China, and that would have been in Mandarin.
(4)「逆 境
和 挑 戰
只 要
激 發 起 生
‘Jik6 ging2 wo4 tiu1 zin3
zi2 jiu3 nang4 gik1 faat3 hei2 sang1 ming6 dik1
hardship and challenges only must can
力 度,
我 們
的 成
是 可 以 超 乎
lik6 dou6, ngo5 mun4 dik1 sing4 zau6
si6 ho2 ji5 ciu1 fu4
LP accomplishments is can
soeng2 zoeng6
zi6 gei2
‘As long as hardship and challenges can arouse life's vigour, then our achievements
can surpass our own imagination.’ [LP=linking particle]
However, it is likely that he quoted his words with Cantonese pronunciation with a few
minor changes, such as replacing Mandarin shi4 “to be” with its Cantonese equivalent
hai6. The whole passage would then have been translated back to Mandarin by the
2.3 Comparison of a quotation in Ming Pao, Apple Daily and YouTube
The following quotation appeared in the two newspapers, and Li's actual words were also
recorded in a YouTube video clip (no longer available). In Ming Pao, the quotation is
reported as in (5).
(5) 他 表 示, 為 免
對 「有 心
和 有 可 能
出 來 選
Taa1 biu2 si6, wai6 min5 deoi3 ‘jau5 sam1
wo4 yau5 ho2 nang4 ceot1 loi4 syun2
he said
, to avoid
have intention and have probability participate elect
首 的 人 士
道」 ,不 會
評 論 個 別
dak6 sau2 dik1 yan4 si6 bat1 gung1 dou6’ , bat1 wui5 ping4 leon6 go3 bit6
LP people
not fair
, not will comment individual
人 選.
jan4 syun2
‘He said, in order to avoid ‘being unjust to people who intend to and have the
probability of participating in the election for the leader’, he would not
comment on individual contestants.’
In the Apple Daily reporting, given in (6), Li's words are reported more directly.
(6) 李 嘉
評 論 個 別
人 選,
Lei5 gaa1 sing4 keoi5 zyut6 ping4 leon6 go3 bit6 jan4 syun2.
Ka shing refused
comment individual contestants
「我 講
嘢, 對
任 何 想
特 首
‘ngo5 gong2 mat1 je5 deoi3 jam6 ho4 soeng2 zou6 dak6 sau2
say anything about any
wishes be leader
嘅 人
ge3 jan4
LP person entirely not
gung1 ping4.’
Li Ka-shing refused to comment on individual contestants, ‘It would not be fair
if I said anything about any person who wishes to be the leader.’
What Li actually said, transcribed from the YouTube clip, is given in (7).
(7)「我 今
日 講
任 何 邊
說 話
對 有 心
‘Ngo5 gam1 jat6 gong2 jam6 ho4 bin1 geoi3 syut3 waa6 deoi3 jau5 sam1
this day say
about have intention
PAMAPLA 38 / ACALPA 38 (2014)
出 嚟
嘅 可 能
嘅 特 首
喺 唔 公
ceot1 lei4 ging6 syun2
ge3 ho2 nang4 ge3 dak6 sau2 hai2 m4 gung1 dou6 ge3’
out come compete election LP probable
LP leader
is not fair
‘I say today that it would not be fair if I said anything about those probable
contestants in the leadership election.’ [Note that 邊 bin1 is probably a mistake
for jat1 ‘one’]
It can be seen that neither newspaper accurately reported the actual words spoken
by Li, in spite of the quotation marks, although the general meaning of his remarks was
accurately reflected. The quality newspaper Ming Pao diverged from the original words
by translating them into Mandarin, and it also embedded them in the sentence. The
quotation marks could have been omitted, since in this case direct speech and reported
speech would have been the same. The only reason the quotation marks were added was
to give the appearance of authenticity. Overall, accuracy was sacrificed to language
The mass circulation Apple Daily retained the original Cantonese and, in addition,
used the first person pronoun ngo5. This version certainly appears more authentic than
the translated report in Ming Pao, but again it is a summary of what was said rather than
a true representation of the actual words. Moreover, while in the audio clip the number of
Cantonese words was 12, including the grammatical particle ge3 that was repeated three
times, in Apple Daily there were just four Cantonese words, including two occurrences of
ge3. In Ming Pao there were no Cantonese words at all.
Table 1 gives some descriptive statistics about the quotations found in the two newspaper
reports. There was a greater amount of quotation in the quality Ming Pao than in the mass
circulation Apple Daily (31.5% versus 22.5%). However, the quotations in Ming Pao had
a slightly lesser percentage of Cantonese words and expressions than in Apple Daily
(18.4% versus 22.4%).
Table 1
A comparison of the total amount of quotation and the use of Cantonese in the
two newspapers
relative amount of
percentage of
Cantonese in the
total number of
characters in the article
Both results are somewhat unexpected. According to So (1998:76), the overall
percentage of quotations in the two newspapers was almost the same: 2.39% for Apple
Daily and 2.49% for Ming Pao. So also reports that the percentages of quotations in Sing
Pao and the pro People’s Chinese Republic newspaper Wen Wei Po were 0.28% and
0.20% respectively. However, So's results were based on a larger corpus, and the two
passages in this study were chosen specifically for their large number of quotations. So
did not examine the use of Cantonese, but one might have expected that there would have
been a far greater frequency of Cantonese terms in the quotations in the mass circulation
Apple Daily than in the quality Ming Pao. In fact, there were only a slightly larger
number, 22.4% versus 18.4%. However, the fact that one quotation in Ming Pao
containing ten Cantonese words out of thirteen was repeated three times, may have
influenced the results.
Further research would be necessary to determine whether these figures would be
confirmed with a larger corpus. It is possible that the use of Cantonese is growing and is
now used by even the quality newspapers in direct quotations, while in Apple Daily there
may be no increase in the use of Cantonese compared with earlier years. To find out
whether this is so, it would be necessary to compare the present use of Cantonese in both
newspapers with past practice. Both newspapers reduced the number of Cantonese terms,
when the quotations were compared with the actual words spoken by Li, as recorded in
the YouTube video.
1. So Fung Ming (1998:14-15) borrows Yung's concept of iconic quotation (1966) in
which “direct discourse is prepared by indirect discourse and emerges from inside it”
(Voloshinov 1986:132). She defines an iconic expression as one that is in quotation marks
but is in fact not a real quotation, for example, “she seems to know those tricks of his.”
So shows how journalists can reveal their point of view by their use of quotations, but
does not examine the choice of language, Mandarin or Cantonese, or the accuracy of the
quotations, both of which can also reveal the journalist’s and the newspaper’s political
I wish to thank Shirley Leung and Henry Chong for their helpful advice concerning the
translation of the quotations.
Forbes. URL:
Lister, Anthony C. 2011. Written Cantonese in Hong Kong advertisements. Papers from
the 33rd Annual Meeting of the Atlantic Provinces Linguistics Association (2009) 33.
PAMAPLA 38 / ACALPA 38 (2014)
Ming Pao. URL:
So Fung Ming. 1998. Functional Analysis of Quotations in Chinese Newspapers. M.A.
dissertation, Hong Kong Baptist University.
Snow, John. 2004. Cantonese as Written Language. Hong Kong: Hong Kong University
Voloshinov, Valentin Nikolaevich. ([1929]) 1986. Marxism and the Philosophy of
Language. Cambridge: Harvard University Press.
Yung, Vicky. 1996. The Presentation of Voice in Chinese and English Newspapers in
Hong Kong. Perspectives. Working Papers of the Department of English, City
University of Hong Kong 8(1): 64-92.
Saint Mary’s University
This paper focuses on null subjects in Russian root finite clauses, discussing the structural
properties of referential and non-referential NS constructions. The question is: What is
the structural position of a null pronoun (pro)? Hypothetically, the specifier of tense
(Spec,TP) is a position where pro is licensed and interpreted (Rizzi 1986). Under the
current assumptions, the licensing head is a complementizer (C), not an inflectional head
with rich agreement. This is what Sigurðson (2011) calls “context linking,” implying that
C is involved in extra-clausal context scanning and clause-internal C/edge linking. In
Tsedryk (2015), I implement Sigurðson’s idea, proposing that pro is a bare matrix of features valued by C. Pro can also be viewed as a regular pronoun that is void of phonetic
content (Holmberg 2005). Both approaches are discussed in section 3, where I scrutinize
pro from a feature-geometric perspective. Sections 1 and 2 provide empirical
background, introducing different types of NSs (section 1) and presenting intervention
effects (between C and pro) in the context of Sigurðson’s C/edge linking generalization
(section 2). As a theoretical framework, I assume a probe-goal model (e.g., Chomsky
2000 et seq.) and a non-cartographic architecture of the clausal spine with three
functional categories (v, T and C).
Historically, Russian evolved from a canonical pro-drop system towards a non-pro-drop
one, showing a considerable decline of Null Subjects (NSs) after the middle of the 17th
century (Choo 2003:84). From a typological point of view, it is uncontroversial to claim
that, compared to West and South Slavic languages, Modern Russian is not a pro-drop
language (see, e.g., Lindseth 1998). Even though Russian is not a canonical NS language,
it still allows the subject to be dropped. This property of Russian kept it in a somewhat
grey zone of the pro-drop/non-pro-drop dichotomy (see Tsedryk 2015 for an overview).
The Russian type of subject drop is sometimes described in terms of topic-drop or
ellipsis (Franks 1995, McShane 2009). It is usually agreed that subject drop in Russian is
more restricted than it is in canonical NS languages. In fact, a referential subject should
not be left unpronounced unless it can be recovered from the immediate discourse or
context. A subject gap can have an extra-clausal linguistic antecedent, as in (1a), or it can
be associated with a situational referent, as in (1b).
Kto-nibud’ videl Mašui? ____i govorila že, čto ____i pridët
somebody saw.M Maša.ACC
said.F FOC that
‘Did anybody see Maša? (She) did say that (she) would come on time.’
PAMAPLA 38 / ACALPA 38 (2014)
[Pointing toward Maša’s friend.]
____i govorit, čto pjat’ minut nazad Maša
byla ešče doma.
says.3SG that five minutes ago
Maša.NOM was.F still home
‘(She) says that five minutes ago Maša was still home.’
Examples (1a-b) show that referential NSs in Russian are compatible with past, present
and future tenses. Present and future have identical verbal inflection with number and
person agreement; past tense has agreement in gender and number, but not in person.
In (2), the intended interpretation of NS is 1PL, whereas the verbal inflection does
not reflect the person feature in morphology. According to Avrutin and Rohrbacher
(1997), person features in the past tense are by-products of presupposition and they are
supplied by “discourse binding.”
____ včera byli
na rynke, ____ videli tam Mašu.
yesterday were.PL at market
saw.PL there Maša.ACC
‘(We) were at the market yesterday, (we) saw Maša there.’
Note that the first subject gap precedes the adverb (včera). The motivation behind this
particular notation will become clear in section 2.
In addition to referential NSs (available for any tense and person), Russian also
has 3rd person non-referential NSs, as illustrated in (3). These NSs are differentiated by
morphological number: 3PL is comparable to an arbitrary they and 3SG to an impersonal
it. I refer to these instances of NS as “arbitrary” and “impersonal” pro, respectively.
____ govorjat, čto zavtra
say.3PL that tomorrow to.pour rain.NOM
‘(They) say that it will pour with rain tomorrow.’
____ budet
tomorrow cold
‘It will be cold tomorrow.’
Again, for reasons to be explained in section 2, I make a particular choice in the
alignment of the subject gap in (3b). This time it follows the adverb (zavtra), unlike the
NS in (2). From a phonetic point of view, it is unmotivated. From a semantic point of
view, we have a referential pro in (2) and a non-referential one in (3b). From a syntactic
point view, I claim that these NSs are not in the same structural position.
In sum, there are three types of pro: (i) referential pro with any person and
number specification (1SG/PL, 2SG/PL, 3SG/PL), (ii) arbitrary pro (3PL), and (iii)
impersonal pro (3SG). We can entertain the idea of having a null counterpart for each
overt pronoun (following Holmberg 2005) and assuming a special kind of non-referential
pro for each instance in (3). At a more abstract level, we can assume that there is only one
kind of pro and that syntax is able to manipulate features in such a manner that every
single instance of NS (referential or not), as briefly sketched in (1)-(3), is a result of a
syntactic computation. I return to this discussion in section 3.
As mentioned in section 1, Russian is sometimes described as a topic-drop language.
Topic-drop is more restricted than canonical pro-drop found in Romance; e.g., Germanic
topic-drop should have an empty left edge. In this section, Russian is compared to both
Germanic and Romance types of subject drop. While discussing Russian data, I rely on
Sigurðson’s (2011) idea of “context linking.”
According to Sigurðson (2011), the syntactic structure should contain C/edge
linking features (CLn) in the C-domain of a clause. A set of such features includes
different kinds of topics (Top), as well as a logophoric agent (ΛA) and a logophoric
patient (ΛP), also commonly known as the “speaker” and the “hearer.” Their purpose is to
provide a medium between syntax and context, which is formally ensured by the
generalization in (4).
C/Edge-Linking Generalization (Sigurðson 2011:282)
Any definite argument, overt or silent, positively matches at least one CLn in
its local C-domain, CLn  {ΛA, ΛP, Top, ...}.
On the one hand, a CLn scans the context outside the clause; on the other, it is
linked to an argument clause-internally. The umbrella term for both processes is “context
linking” as schematically presented in (5) (based on (35) in Sigurðson 2011:284). In a
non-cartographic setting (assumed here in divergence from Sigurðson 2011), the locus of
{CLn} is C. Slightly simplifying Sigurðson’s proposal, I assume that it is the C head that
ensures context linking (context scanning + C/edge-linking).
context scanning
[CP ... C ... [TP ... pro/agreement ...
What is relevant from a syntactic point of view is C/edge linking, which amounts to
Chomsky’s Agree (2000 et seq.) that relates features, ensuring their interpretability at the
Abstracting away from technical details, it is worth pointing out that C/edge
linking is subject to minimality effects (Rizzi 1990). More precisely, in Germanic (e.g.,
Icelandic), pro is a phrasal category, whereas in Romance (e.g., Italian), it is not:
agreement morphology is an incorporated weak pronoun attached to T, and pro and T
form a phonological unit, making Spec,TP invisible at the interfaces. Thus, an XP
intervenes between C and pro in (6b), but not in (7a) (Sigurðson 2011:286-7).
stundum íslensku.
speak.1SG sometimes Icelandic
‘I sometimes speak Icelandic.’
*Stundum tala
sometimes speak.1SG Icelandic
[CP C ... (*XP) ... pro ... T
PAMAPLA 38 / ACALPA 38 (2014)
Travolta parlo
sometimes speak.1SG Icelandic
‘Sometimes I speak Icelandic.’
[CP C ... (XP) ... pro-T
Let us now consider intervention effects in Russian, beginning with (8), which
features a question-answer pair with two possible answers. A2 in (8) is comparable with
(6b).1 What is interesting is that eë ‘her’ fronting becomes perfectly grammatical if the
verb is plural and the subject has an arbitrary reading, as in (9).
(8) Q:
Kak ty
poznakomilsja s
how you.NOM got.acquainted.M with Maša.INSTR
‘How did you get acquainted with Maša?’
Vstretil u druga na večerinke.
met.M at friend’s on party
‘I met her at a friend’s party.’
vstretil u druga na večerinke.
her.ACC met.M at friend’s on party
vstretil-i na vyhode iz
her.ACC met-PL at exit
from house
‘She was met at the exit from the house.’
As expected, there is no intervention effect in (9), since the arbitrary pro does not require
C/edge linking, whereas the referential pro has to be linked to C. In (10), we have a topic
(Mašu) in a preverbal position. Again, only an arbitrary reading is available. In some
cases, fronting an XP may lead to ambiguity, as shown in (11).
[Maša has disappeared. Everybody talks about her (= given topic), but nobody
knows where she is. Somebody says:]
videl-i včera
v magazine.
Maša.ACC saw-PL yesterday in store
(i) #‘(We) saw Maša yesterday in the store.’ (referential)
(ii) ‘Maša was seen yesterday in the store.’ (arbitrary)
v magazine videl-i Mašu.
yesterday in store
saw-PL Maša.ACC
(i) ‘Yesterday in the store (we) saw Maša.’ (referential)
(ii) ‘Yesterday in the store Maša was seen.’ (arbitrary)
In what follows, I will focus on (10) and (11). Before looking at the structure of these
sentences, let me first spell out some relevant features of Russian syntax.
In (12a), we have a variant of (10)-(11) with an overt pronominal subject. As
shown in (12b), the subject is in its canonical position (Spec,TP), motivated by the
Extended Projection Principle (EPP). Phonetically expressed nodes are in bold.
My videl-i Mašu [PP včera
v magazine].
We saw-PL Maša.ACC yesterday in store
‘We saw Maša yesterday in the store.’
There are two relevant properties of Russian syntax. First, the verb does not move to T.2
Second, the EPP position does not have to be occupied by a nominative DP—any XP can
potentially fill this position (see Bailyn 2012: 317-8).
Returning now to (10), assume that Mašu (being topicalized from its VP-internal
object position via the outer Spec,vP) moves to Spec,TP by-passing the subject pro, as
shown in (13). By virtue of being a topic, Mašu holds the most prominent structural
position, which coincides with a need to satisfy EPP.3
PAMAPLA 38 / ACALPA 38 (2014)
In (13), Mašu intervenes between C and pro, preventing the latter from been contextually
linked. Thus, only a non-referential pro can be used in this configuration.
Consider now the ambiguous sentence in (11). At some point of the derivation, PP
moves to an outer Spec,vP (see (14)). Note that this movement is not the same as
topicalization of Mašu in (10)/(13). In fact, it is common for spatio-temporal topics to be
fronted in all-new-information sentences (see Dyakonova 2009:15), but they do not have
to be as prominent as individual topics. When T is merged, there is an option of satisfying
EPP by either pro or PP. If EPP is satisfied by pro, we obtain (14a). If EPP is satisfied by
PP, we have (14b).
Crucially, since the verb does not move to T, we do not have an overt indication of
whether or not PP has moved to Spec,TP. Both configurations have the same word order
Unlike (14a), the configuration in (14b)—just like the one in (13)—is
incompatible with a referential pro, because there is an intervener (PP). The subsequent
question is whether or not (14a) is compatible with an arbitrary pro. If yes, this structure
alone would be sufficient to account for the ambiguity in (11), and we would have no
empirical motivation behind the structure in (14b). In a more restrictive approach, each
structure in (14) would correspond to a particular reading: referential in (14a) vs.
arbitrary in (14b).
3.1 Approach A: Pro as a P valued by C
In Tsedryk (2015), I consider the arbitrary pro as a by-product of the syntactic
computation. That is, there is no inherent difference between a referential and a nonreferential pro until the latter enters into a probe-goal relation with C. Initially, pro has a
minimal specification in -features, whereas C contains a maximally specified set of features.
For convenience, -features are represented as a hierarchical set:
This feature geometry is adapted specifically to Russian inflectional and pronominal
systems (a slightly modified version of (22) in Tsedryk 2015: 353).4 For all nodes, except
[] (person) and [#] (number), the sisterhood means ‘exclusive disjunction’ (‘either...
or’). First person ([1]) excludes second person ([2]), plural ([PL]) excludes singular
([SG]), and feminine ([F]) excludes masculine ([M]). Dominance is implicational: [1] and
[2] imply [], [F] and [M] imply [SG], and [PL] and [SG] imply [#]. Bare [] is ‘3rd person’
and bare [SG] is ‘neuter’.
Thus, C starts with a feature specification in (17) and ends up with a reduced
specification depending on the context: contextually irrelevant features are dissociated
from the structure as a result of context scanning; see, e.g., (18a). T and pro have only a
bare root [], as in (18b-c).
In addition to (18), we have to stipulate that -features of C and T are [-interpretable],
whereas those of pro are [+interpretable]. That is, -features in C are [-interpretable,
+valued], those in T are [-interpretable, -valued], and for pro they are [+interpretable, valued]. Labeling features as [+/-interpretable] is essential in a “derivational time-bombs
model” (Preminger 2014). Only a [-interpretable] feature can trigger a probe-goal
relation, which ends up copying features from one syntactic node to another.
When T targets pro, nothing happens, since both contain nothing but a bare root
of -features. However, when pro moves to Spec,TP, as in (14a), it can be targeted by C
and -features are copied from C to pro. As a result, pro receives a contextually linked set
of -features, which is possible only if it is probed by C. If, instead of pro, another XP
moves to Spec,TP, as in (14b), C would not be able to probe further than this XP. In this
case, pro cannot receive contextually linked -features from C. This derivation results in
a default 3rd person specification characterizing an arbitary/impersonal pro, as in (3) (see
Tsedryk 2015: 355 for further details).5
This approach clearly separates both structures in (14). It also makes a testable
prediction that there is a correlation between XP fronting and non-referential instances of
PAMAPLA 38 / ACALPA 38 (2014)
pro. Nevertheless, there is a conceptual drawback: it relies on the assumption that
features can be either interpretable or not regardless of their geometric property. That is,
T and pro look the same from the point of view of -features, but we still have to
stipulate that those of pro are [+interpretable] and that those of T are [-interpretable]. This
is hard to motivate besides the need to have T as a probe. In Tsedryk (2015), I simply
assume a four-way feature specification (following Pesetsky and Torrego 2007) and
combine [+/-interpretable] and [+/-valued] in four logical possibilities distributed among
C, T, DP and a bare P.
[-interpretable] [+valued]
[-interpretable] [-valued]
[+interpretable] [+valued]
[+interpretable] [-valued]
3.2 Approach B: Pro as a P/DP with its own values
As an alternative, we can assume that pro has a predetermined set of -features just like
any other pronoun (Holmberg 2005). For example, it can be 1PL (20a) or 3PL (20b).
Note that, if we assume that T has a bare root [], as in (18b), we do not have to
introduce an ad hoc [+/-interpretable] distinction. There is a choice. We can follow
Chomsky (2000) who suggests that -feature interpretability is related to availability of
-feature values (a valued feature is interpretable, while an unvalued one is not).
Alternatively, we can simply dispense with interpretability of features in syntax and
assume a version of agreement developed by Preminger (2014). That is, agreement is an
obligatory operation that is not derivative of some conceptually needed property (i.e.,
uninterpretable features), but it is an obligatory operation on its own. What it does is to
copy one piece of -geometry to another. This route is worth exploring in more detail, but
I leave both options open at this point.
When T (= 18b) targets a fully specified P (e.g., 1PL), the latter’s features are
copied to T, producing the following result:
At this point, the movement of P to Spec,TP becomes redundant, since all -features
have already been copied to T, and it does not make sense to project a specifier that
would duplicate the same set of features. We thus predict that pro should not appear in
The ban of pro from Spec,TP has nothing surprising if we dispense with EPP. For
example, according to Kučerová (2014), EPP could be replaced by a more general TExtension Requirement:
If Merge(T,α) applies, Merge(T,β) must be the next step of the derivation,
where T is a projection of T and β belongs to the same phase domain as T
(Kučerová 2014: 137).
To rephrase, (22) roughly states that, if T has been merged with some phrase (e.g., vP),
another instance of Merge should also take place. The second instance of Merge can,
but—crucially—does not need to end up with an XP in Spec,TP. What is relevant for our
discussion is that T in (21) can now be extended: either by an XP, which would result in a
specifier, or by C, in which case no specifier would be projected. Whatever is the next
instance of Merge, once the -features of P are copied to T, they become accessible to
C. Even if an XP moves to Spec,TP, C is still the category that, at some point, is merged
with TP and a specifier of T is not an obstacle between C and TP (and all features of T
are supposedly available in its projection, TP). In other words, we run into an empirical
problem: we do not expect intervention effects between C and P, since T is the head
(local to C) that makes the copied features readily accessible to C. We falsely predict that
structures (13) and (14b) should be compatible with both referential and arbitrary
readings (see the above discussion of (10) and (11)).
To remedy the problem created, we should ensure that pro can still move to
Spec,TP without inducing a redundancy between such movement and its features copied
to T. The moving phrase could be, for example, a superset of features that are copied in a
probe-goal relation. There are two possible ways to achieve this result. One of them is to
specify T in such a manner that it does not require all features to be copied to T. Another
way is to add an additional structural layer to P. These two options are dealt with in
sections 3.2.1 and 3.2.2, respectively.
PAMAPLA 38 / ACALPA 38 (2014)
3.2.1 Present vs. past tense
As mentioned in section 1, the past tense in Russian shows agreement in number and
gender, whereas the present (and future) tense shows agreement in person and number.
Thus, instead of (18b), we should have different -geometries for T[PRESENT] and T[PAST], as
in (23).
Not all -features are copied to T. T[PRESENT] looks for person and number values, except
[SG].6 T[PAST] would trigger copying either [PL] or [SG].
Now, if we replace T in (21) by T[PAST] from (23b), we would have a configuration
in which only [PL] is copied to T. In (24), P is a superset of the -features in T (cf. (21)).
Movement of P in (24) would not create a specifier with an identical set of -features as
T. Moreover, moving P to Spec,TP brings [] to a higher position, making it accessible
to C. [] can now be linked to C in accordance with the C/edge linking generalization in
In the present tense, on the other hand, the situation is the same as in (21). [] is
already present in T, and this feature will either stay as such (3rd person) or it would
dominate [1] or [2] if these values are copied from the goal (P). To put it differently,
T[PRESENT] would basically have the same set of -features as P. That is, we predict an
asymmetry between T[PRESENT] and T[PAST]: the latter, but not the former, should induce
intervention effects discussed in section 2. As we can see in (25), this prediction is not
accurate: the topic (Mašu) does not allow a null subject.
Kak vy
otnosites’ k
how you.NOM feel
towards Maša.DAT
‘How do you feel about Maša?’
očen’ ljubim.
Maša.ACC we.NOM very like.1PL
‘Maša, we like her very much.’
To reiterate, if pro were just an instance of the -features in T[PRESENT], it would be
accessible to C with or without an XP in Spec,TP. Even though T[PRESENT] and T[PAST] do
have different -feature specifications, this difference does not seem to affect the
syntactic distribution of the null subjects in Russian.
Recall that we aim to ensure that a copy of -features in T (as a result of a probegoal/T-P relation) does not make P accessible to C. Otherwise, Russian would have a
Romance-type pro-drop. P should move to Spec,TP, and this movement should not be
redundant with regard to -features in T (i.e., features copied to T should be a subset of
features of the category moved to Spec,TP). Adding a structural layer to the moving P is
a plausible solution.
3.2.2 P vs. DP
As pointed out in section 2, an arbitrary pro is likely to remain in Spec,vP, whereas a
referential one is likely to move to Spec,TP. This syntactic difference can be attributed to
a structural weight of the pronoun itself. For example, the arbitrary-referential dichotomy
can be formalized as a P-DP structural distinction.
Arbitrary 3PL
Referential 3PL
The D(eterminer) layer adds other features such as ‘specific’, ‘definite’ and ‘deictic’.
Unlike Camacho (2013), I do not represent these features as part of the -feature
geometry. D ensures the referentiality of the -set, which otherwise does not have its own
reference. According to Holmberg (2005), D is a feature of T in pro-drop languages and a
null pronoun is a P. Thus, when P is probed by T, P is interpreted as a referential
pronoun. Since Russian is not a pro-drop language of the Romance type, T does not have
a D-feature. A referential pronoun is a DP itself.
As it stands, the structural difference in (26) can also be related to spell-out.
Under the assumption that D is the head that is targeted by overt pronominal exponents at
the time of vocabulary insertion (going back to Postal’s (1966) suggestion that pronouns
are determiners), the structure in (26a) is always phonetically empty, whereas the one in
(26b) can be either null or not. That is, a referential pro is nothing but a lack of phonetic
PAMAPLA 38 / ACALPA 38 (2014)
material in D; otherwise it is not syntactically different from an overt pronoun, like oni
‘they’ (in the case of the 3PL feature specification). The overt vs. covert distinction in the
referential pronominal system is just a matter of spell-out regulated at the sensorimotor
Now we can also assume that a DP (but not a P) has to move to Spec,TP by
virtue of the additional D layer. It is D that has to be linked to C in accordance with the
C/edge linking generalization in (4). Consequently, P always stays in situ without being
affected by an intervening XP. If we return to the structures in (14), we should have a DP
(= referential pro) in (14a) and a P (= arbitrary pro) in (14b), and no overlap between
these two structures is expected either with T[PAST] or T[PRESENT].
To conclude, Russian patterns with a Germanic topic-drop (e.g., Icelandic) that is
recognized by intervention effects, which are absent in canonical pro-drop languages
(e.g., Italian). To have a contextually linked pro in a topic-drop language, nothing should
intervene between C and pro. Russian also has a non-referential (arbitrary 3PL or
impersonal 3SG) pro, which is not subject to C/edge linking generalization (Sigurðson
In section 3, two approaches (A and B) were considered. Under approach A
(shown in (27)), pro is a bundle of -features (P) that are valued in syntax. The locus of
-features for P is C (C’s -features are contextually linked). If there is an XP
intervening between C and P, the latter cannot be probed by C and receives an
arbitrary/impersonal reading (Tsedryk 2015).
(27) Approach A
Referential pro
Non-referential pro
Under approach B, P has a predetermined set of features. In this case, we are led to
postulate the structural distinction shown in (28) (see section 3.2 for motivation).
(28) Approach B
Referential pro
Non-referential pro7
In (28), referential pro is syntactically identical to an overt pronoun; the only difference is
the lack of a phonetically visible exponent in D, which is a matter of post-syntactic
vocabulary insertion.
Approach B has two conceptual advantages. First, it does not have to rely on
feature interpretability as a driving force of syntactic agreement (in light of Preminger
2014). Second, movement to Spec,TP can now be related to the fact that the moving
category is a DP, as opposed to a bare P (definiteness effect). Under approach A, the
only motivation for moving P to Spec,TP is EPP. However, this movement loses ground
if EPP is to be reconsidered in terms of a more general T-Extension Requirement
(Kučerová 2014; see (22)).
From an empirical point of view, approach A establishes a direct relationship
between XP fronting and non-referential pro: if an XP moves to Spec,TP, pro remains in
situ and receives an arbitrary/impersonal reading. We thus expect impersonal and
arbitrary constructions to have a fronted XP. This prediction is only indirectly implied by
approach B. XP fronting might correlate with the occurrence of a bare P (since the latter
is not supposed to move due to its referential defectiveness), but a bare P could in
principle occur independently from a fronted XP. Thus, as far as I can see, approach A
has a stronger predictive power. Moreover, approach B raises a question related to
intervention effects, presented in section 2. If referential pro is syntactically
indistinguishable from overt pronouns, it is not clear why the latter do not induce
intervention effects. We still need to add a stipulation that would differentiate referential
pro from overt pronouns at some level of representation (if it is not syntax). Other things
being equal, I leave both approaches as plausible options for the analysis of null pronouns
in Russian.
1. Object pronouns in Russian usually scramble to the left of the verb. Thus, leaving eë
after the verb in (8) is not a viable option, even though A2 could be somewhat improved
if the pronoun were used post-verbally.
PAMAPLA 38 / ACALPA 38 (2014)
2. According to Gribanova (2013), there is an aspectual projection sandwiched between
TP and vP. Here I use a slightly simplified structure.
3. In (14), EPP could potentially be satisfied by pro. However, if pro moved from vP, it
would not prevent the topic from moving to a higher position (the topic is supposed to be
structurally prominent). In a structure with multiple specifiers, the topic can still move to
an outer Spec,TP, and we would again prevent pro from being linked to C. See section 3
for further discussion.
4. In Tsedryk (2015), the root node in -feature geometry is labeled as [D].
5. In Tsedryk (2015), I propose that feature values that are not linked to discourse/context
(i.e., non-D-linked [3PL] and [3SG]) are transferred from C to T. In fact, I assume that
only an edge head can have a D-linked set of features.
6. Only [PL] is copied under the [#] node of T[PRESENT]. Copying [SG] without copying [M]
or [F] would be impossible, since it is a snippet of the -geometry that is copied, not a
random value (see Preminger 2014). If [SG] is copied as well, we expect that T[PRESENT]
also has agreement in gender, which is not the case. In other words, singular in present
tense inflection is a privative feature, implying the absence of [PL], rather than the
presence of [SG].
7. It is possible that a 3SG non-referential (impersonal) pro in Russian does not have []
at all, which would explain the fact that this pro is non-agentive and inanimate.
Alternatively, animacy could be part of the -feature geometry, following Harley and
Ritter (2002) and McGinnis (2005).
Avrutin, Sergey, and Bernhard Rohrbacher. 1997. Null subjects in Russian inverted
constructions. Formal Approaches to Slavic Linguistics 4: 32-53.
Bailyn, John. 2012. The Syntax of Russian. Cambridge: Cambridge University Press.
Camacho, José A. 2013. Null Subjects. Cambridge: Cambridge University Press.
Chomsky, Noam. 2000. Minimalist inquiries: the framework. In Roger Martin, David
Michaels, and Juan Uriagereka (eds.), Step by Step: Essays on Minimalist Syntax in
Honor of Howard Lasnik, 89-155. Cambridge, MA: MIT Press.
Choo, Sukhoon. 2003. The decline of null pronominal subjects in Old Russian. Ph.D.
dissertation, Indiana University.
Dyakonova, Marina. 2009. A Phase-based Approach to Russian Free Word Order. Ph.D.
dissertation, University of Amsterdam. LOT Dissertation Series 230.
Franks, Steven. 1995. Parameters of Slavic Morphosyntax. New York: Oxford University
Gribanova, Vera. 2013. Verb-stranding verb phrase ellipsis and the structure of the
Russian verbal complex. Natural Language and Linguistic Theory 31: 91-136.
Harley, Heidi and Elizabeth Ritter. 2002. Person and number in pronouns: A featuregeometric analysis. Language 78: 482-526.
Holmberg, Anders. 2005. Is there a little pro? Evidence from Finnish. Linguistic Inquiry
36: 533-564.
Kučerová, Ivona. 2014. The syntax of null subjects. Syntax 17(2): 132-167.
Lindseth, Martina. 1998. Null-Subject Properties of Slavic Languages: With Special
Reference to Russian, Czech and Sorbian. München: Verlag Otto Sagner.
McGinnis, Martha. 2005. On markedness asymmetries in person and number. Language
81: 699-718.
McShane, Marjorie. 2009. Subject ellipsis in Russian and Polish. Studia Linguistica
63(1): 98-132.
Pesetsky, David, and Esther Torrego. 2007. The syntax of valuation and the
interpretability of features. In Simin Karimi, Vida Samiian, and Wendy K. Wilkins
(eds.). Phrasal and Clausal Architecture, 262-294. Amsterdam: John Benjamins.
Postal, Paul. 1966. On so-called ‘pronouns’ in English. In Francis Dinneen (ed.). 19th
Monograph on Languages and Linguistics, 177-206. Washington, DC: Georgetown
University Press.
Preminger, Omer. 2014. Agreement and its Failures. Cambridge, MA: MIT Press.
Rizzi, Luiggi. 1986. Null objects in Italian and the theory of pro. Linguistic Inquiry 17:
Rizzi, Luiggi. 1990. Relativized Minimality. Cambridge, MA: MIT Press.
Sigurðsson, Ármann Halldór. 2011. Condition on argument drop. Linguistic Inquiry 42:
Tsedryk, Egor. 2015. Deriving null pronouns: a unified analysis of subject drop in
Russian. Formal Approaches to Slavic Linguistics 23: 342-361.
PAMAPLA 38 / ACALPA 38 (2014)
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