A Grammatical Description of Dameli Emil Perder

A Grammatical Description of Dameli Emil Perder
Emil Perder
A Grammatical Description
of Dameli
Emil Perder
A Grammatical Description of Dameli
ISBN 978-91-7447-770-2
Department of Linguistics
Doctoral Thesis in Linguistics at Stockholm University, Sweden 2013
A Grammatical Description of Dameli
Emil Perder
A Grammatical Description of
Dameli
Emil Perder
©Emil Perder, Stockholm 2013
ISBN 978-91-7447-770-2
Printed in Sweden by Universitetsservice AB, Stockholm 2013
Distributor: Department of Linguistics, Stockholm University
‫ماں تا َرف تہ ِاک‬
‫ دامیاں‬،‫توحفہ‬
A gift from me,
Dameli people
Abstract
This dissertation aims to provide a grammatical description of Dameli
(ISO-639-3: dml), an Indo-Aryan language spoken by approximately
5 000 people in the Domel Valley in Chitral in the Khyber Pakhtunkhwa
Province in the North-West of Pakistan. Dameli is a left-branching SOV
language with considerable morphological complexity, particularly in
the verb, and a complicated system of argument marking. The
phonology is relatively rich, with 31 consonant and 16 vowel phonemes.
This is the first extensive study of this language. The analysis
presented here is based on original data collected primarily between
2003-2008 in cooperation with speakers of the language in Peshawar
and Chitral, including the Domel Valley. The core of the data consists of
recorded texts and word lists, but questionnaires and paradigms of word
forms have also been used. The main emphasis is on describing the
features of the language as they appear in texts and other material, rather
than on conforming them to any theory, but the analysis is informed by
functional analysis and linguistic typology, hypotheses on diachronical
developments and comparisons with neighbouring and related
languages.
The description is divided into sections describing phonology,
morphology and syntax, with chapters on a range of individual subjects
such as particular word classes and phrase types, phonological and
syntactical phenomena. This is not intended to be an exhaustive
reference grammar; some topics are only touched upon briefly while
others are treated in more detail and suggestions for further research are
given at various points throughout the work.
Contents
Abbreviations ....................................................................................... vii
On transcription ..................................................................................... ix
Acknowledgements ................................................................................ xi
Background .......................................................................................... 3
1
2
Introduction ................................................................................... 3
1.1
The name of the language ............................................................................ 3
1.2
Genealogical classification ........................................................................... 4
1.3
Dameli and other languages ......................................................................... 7
1.4
Previous research ........................................................................................ 9
1.5
Variation within the language ...................................................................... 10
1.6
Language vitality....................................................................................... 11
1.7
Geographic location ................................................................................. 12
1.8
The Dameli people.................................................................................... 12
Methods and data collection .......................................................... 14
2.1
Field research .......................................................................................... 14
2.2
Text corpus .............................................................................................. 15
2.3
Wordlists ................................................................................................ 17
2.4
Paradigms ............................................................................................... 18
2.5
Questionnaires ......................................................................................... 18
2.6
Other kinds of data ................................................................................... 19
Phonology .......................................................................................... 21
3
Phonemes ................................................................................... 23
3.1
Vowels ................................................................................................... 24
3.1.1
Vowel quality ................................................................................. 25
3.1.2
Vowel length .................................................................................. 25
3.1.3
Nasality ........................................................................................ 27
3.1.4
Diphthongs and vowel sequences ...................................................... 27
3.2
Consonants ............................................................................................. 27
3.2.1
Place of articulation ......................................................................... 28
3.2.2
Manner of articulation ...................................................................... 28
3.2.3
Voicing and aspiration ..................................................................... 31
i
Contents
3.3
4
5
6
Tone ...................................................................................................... 32
Syllable structure .......................................................................... 33
4.1
Onset ..................................................................................................... 33
4.2
Nucleus .................................................................................................. 34
4.3
Coda ..................................................................................................... 34
Prosody ..................................................................................... 36
5.1
Stress ..................................................................................................... 36
5.2
Intonation ................................................................................................ 37
Phonological processes ................................................................. 38
6.1
Vowel sequences ...................................................................................... 38
6.1.1
Elision ........................................................................................... 38
6.1.2
Lengthening ................................................................................... 39
6.1.3
Coalescence .................................................................................. 39
6.1.4
Epenthetic semivowels ...................................................................... 39
6.1.5
Hiatus ........................................................................................... 40
6.1.6
Diphthongs or vocalic sequences ....................................................... 40
6.2
Conditioned vowel lengthening ................................................................... 41
6.3
Word-final processes ................................................................................. 41
6.3.1
Reduction of clusters ........................................................................ 42
6.3.2
Deaspiration .................................................................................. 43
6.3.3
Obstruent devoicing ........................................................................ 43
Morphology ........................................................................................ 47
7
Nouns ....................................................................................... 49
7.1
Noun roots .................................................................................... 49
7.1.2
Compounds ................................................................................... 50
7.2
Animacy ................................................................................................. 50
7.3
Gender................................................................................................... 51
7.3.1
Gender assignment ......................................................................... 52
7.3.2
Gender agreement .......................................................................... 53
7.3.3
Decline of gender ........................................................................... 54
7.4
ii
Noun inflection......................................................................................... 55
7.4.1
Number ........................................................................................ 56
7.4.2
Case ............................................................................................ 58
7.5
8
Noun stems ............................................................................................. 49
7.1.1
Kinship nouns ........................................................................................... 62
7.5.1
Kinship system ................................................................................ 62
7.5.2
Kinship person suffixes ..................................................................... 69
7.5.3
Vocative ........................................................................................ 70
7.5.4
Special plural forms ......................................................................... 70
Pronouns .................................................................................... 72
Contents
8.1
Personal pronouns ..................................................................................... 72
8.1.1
Person........................................................................................... 73
8.1.2
Number ........................................................................................ 73
8.1.3
Distance ........................................................................................ 73
8.1.4
Animacy........................................................................................ 73
8.1.5
Case ............................................................................................ 74
8.1.6
Third person pronouns as determiners .................................................. 77
8.1.7
Relationship to other deictic words...................................................... 78
8.2
Possessive pronouns .................................................................................. 79
The possessive marker sãã ................................................................. 80
8.2.1
The reflexive possessive pronoun taanu ................................................ 80
8.2.2
8.3
9
Indefinite and interrogative pronouns ............................................................ 82
Adjectives .................................................................................. 83
9.1
Variable adjectives.................................................................................... 83
9.2
Invariable adjectives .................................................................................. 84
9.3
Adjective derivation suffixes ........................................................................ 85
−bana/−bani ‘−ish’ ............................................................................ 85
9.3.1
−baṣ ‘able to x’................................................................................ 86
9.3.2
−weela ‘having x’ ............................................................................. 87
9.3.3
−pin ‘full of x’ .................................................................................. 87
9.3.4
10
11
Numerals ................................................................................... 88
10.1
Numeral system ................................................................................... 88
10.2
Cardinal numerals ................................................................................ 90
10.3
Ordinal numerals ................................................................................. 91
10.4
Collective numerals .............................................................................. 92
Verbs......................................................................................... 93
11.1
The verb root....................................................................................... 96
11.1.1
Irregularities in the verb root .............................................................. 97
11.1.2
Vowel-strengthening ......................................................................... 99
11.2
The verb stem .................................................................................... 100
11.2.1
Causativisation ............................................................................. 100
11.3
Finiteness.......................................................................................... 105
11.4
Verb inflection ................................................................................... 105
11.4.1
Subject marking categories ............................................................. 110
11.4.2
TAM categories ............................................................................ 112
11.4.3
Verbs with limited inflection ............................................................. 112
11.5
Finite verb forms ................................................................................ 113
11.5.1
The perfective ............................................................................... 113
11.5.2
The imperfective............................................................................ 113
11.5.3
The future .................................................................................... 114
11.5.4
The indirect past ........................................................................... 114
11.5.5
The potential past .......................................................................... 115
iii
Contents
11.5.6
11.6
11.6.1
Converbs or participles? ................................................................. 116
11.6.2
The infinitive ................................................................................. 118
11.6.3
The present participle ..................................................................... 118
11.6.4
The past participle ......................................................................... 119
11.6.5
The conjunctive participle ............................................................... 120
11.6.6
The inchoative participle ................................................................. 120
11.7
Copula verbs .................................................................................... 121
11.7.1
The animate copula ....................................................................... 121
11.7.2
The inanimate copula..................................................................... 124
11.8
12
Conjunct verbs .................................................................................. 125
Postpositions ............................................................................. 127
12.1
Postpositions that require the locative case .............................................. 127
12.1.1
12.1.2
12.1.3
12.1.4
12.1.5
12.2
The postposition ki ‘to’, ‘for’............................................................. 128
The postposition ṣaži ‘for’, ‘in order to’ ............................................... 128
The postposition ṣawaai ‘by’, ‘through’ ............................................... 129
The postposition mili ‘with’ .............................................................. 129
The postposition ĩĩ ‘appropriate place’ .............................................. 130
Other postpositions............................................................................. 130
12.2.1
12.2.2
12.2.3
12.2.4
12.2.5
13
The imperative .............................................................................. 115
Non-finite verb forms .......................................................................... 115
The postposition ta ‘from’, ‘of’, ‘than’ ................................................ 130
The postposition ṣaa ‘on’ ................................................................. 131
The postposition neẉ ‘under’ ............................................................ 131
The postposition taprei ‘for’.............................................................. 131
The postposition bãĩ ‘towards’ .......................................................... 132
12.2.6
The postpositions tagii ‘from’ and tali ‘until’ ......................................... 132
12.2.7
Nouns that function as postpositions.................................................. 133
Minor or questionable word classes ............................................... 134
13.1
Adverbs ........................................................................................... 134
13.2
Particles ........................................................................................... 135
13.3
Conjunctions ..................................................................................... 135
13.4
Question-words ................................................................................. 136
Syntax ............................................................................................. 139
14
Basic declarative clauses ............................................................. 141
14.1
iv
Valency clause types .......................................................................... 142
14.1.1
Zero transitivity clauses ................................................................... 142
14.1.2
Intransitive clauses ......................................................................... 143
14.1.3
Transitive clauses .......................................................................... 143
14.1.4
Ditransitive clauses ........................................................................ 143
14.1.5
Clauses with copula verbs .............................................................. 144
Contents
15
16
17
18
The noun phrase ........................................................................ 147
15.1
Determiners ....................................................................................... 148
15.2
Attributes .......................................................................................... 148
15.3
Possession ........................................................................................ 149
Tense, aspect and related categories ............................................. 151
16.1
Tense and aspect ............................................................................... 151
16.2
Mood .............................................................................................. 152
16.3
Evidentiality ...................................................................................... 152
16.4
Epistemic modality ............................................................................. 153
16.5
The past tense marker taa .................................................................... 154
16.6
TAM constructions .............................................................................. 154
16.6.1
Simple present .............................................................................. 155
16.6.2
Present result 1 ............................................................................. 155
16.6.3
Present result 2 ............................................................................. 156
16.6.4
Future ......................................................................................... 156
16.6.5
Perfective ..................................................................................... 157
16.6.6
Indirect past ................................................................................. 157
16.6.7
Potential past................................................................................ 158
16.6.8
Continuous past ............................................................................ 158
16.6.9
Habitual past ............................................................................... 159
16.6.10
Inchoative past ........................................................................ 159
16.6.11
Imperative ............................................................................... 160
Adverbials ................................................................................ 161
17.1
Time adverbials ................................................................................. 161
17.2
Spatial adverbials .............................................................................. 162
17.3
Intensifiers ......................................................................................... 163
Coordination ............................................................................ 164
18.1
18.2
18.3
18.4
19
Natural coordination .......................................................................... 164
Conjunctive coordination with ãã ‘and’................................................... 166
Disjunctive coordination with kuu and ya ‘or’ ........................................... 166
Adversative coordination with xu and leekin ‘but’ ...................................... 167
Subordination............................................................................ 169
19.1
Finite subordinate clauses .................................................................... 169
19.2
Non-finite subordinate clauses .............................................................. 171
19.3
Complementation ............................................................................... 172
19.3.1
20
Conditional clauses ....................................................................... 173
19.4
Adverbial clauses ............................................................................... 174
19.5
Relative clauses ................................................................................. 175
19.6
Quotatives ........................................................................................ 176
Other clause types ..................................................................... 178
v
Contents
20.1
21
Questions ......................................................................................... 178
20.1.1
Content questions .......................................................................... 178
20.1.2
Polar questions ............................................................................. 179
20.1.3
Rhetorical questions ....................................................................... 179
20.2
Imperative clauses .............................................................................. 180
20.3
Negation ......................................................................................... 181
Some discourse features .............................................................. 183
21.1
21.2
The topic particles ta and ba ................................................................ 183
Also, −es ........................................................................................... 185
21.3
The question clitic −i ........................................................................... 186
21.4
The vocative particle........................................................................... 186
21.5
Fillers ............................................................................................... 187
21.6
Echo-word formation ........................................................................... 187
21.7
Afterthought constructions ..................................................................... 188
21.8
Tail-head linkage ............................................................................... 189
21.9
Clause chaining ................................................................................ 193
Final matters ...................................................................................... 195
22
Suggestions for further research ..................................................... 197
22.1
The language of women ..................................................................... 197
22.2
Historical linguistics ............................................................................ 198
22.3
Related languages and language contact ............................................... 198
22.4
Variation within the language ............................................................... 198
22.5
Tone................................................................................................ 199
References ........................................................................................ 200
Appendix 1: Verb roots ....................................................................... 206
Appendix 2: Text ............................................................................... 209
Introduction ..................................................................................................... 209
The Patient Woman .......................................................................................... 210
vi
Abbreviations
The language examples in this work are formatted according to the
Leipzig Glossing rules suggested by the Max Planck Institute (2008),
with additions for categories and features in Dameli that are not covered
in the rules. The list below contains all the abbreviations used in the
language examples. Those that are not directly taken from the Leipzig
glossing rules are written in italics.
Table 1: Interlinear glossing abbreviations
Abbreviation
Meaning
1
2
3
First person
Second person
Third person
Animate
Appropriate place
Causative
Second causative
Collective
Copula
Conjunctive participle
Distal
Echo-word
Epenthetic vowel
Ergative
Feminine
Filler
Future
Imperative
Imperfective
Inanimate
Inchoative participle
Indirect past
Infinitive
ANIM
APPR
CAUS
CAUS2
COLL
COP
CP
DIST
ECHO
EP
ERG
F
FILL
FUT
IMP
IMPFV
INANIM
INCHPTCP
INDIRPST
INF
Read about it on
page…
73, 110
73, 110
73, 110
50
130
100
103
92
121
120
73
187
109
60, 74
51, 83, 79, 111
187
114
115
113
50
120
114
118
vii
Abbreviations
INS
KIN
LOC
M
NEG
NOM
OBL
ORD
PFV
PL
POSS
POTPST
PROH
PROX
PRSPTCP
PST
PSTPTCP
Q
REFL
SG
TOPSH
TOPSM
VOC
viii
Instrumental
Kinship
Locative
Masculine
Negation
Nominative
Oblique
Ordinal
Perfective
Plural
Possessive
Potential past
Prohibitive
Proximal
Present participle
Past
Past participle
Question marker
Reflexive
Singular
Shift topic particle
Same topic particle
Vocative
61
69
60, 76
51, 83, 79, 111
181
59, 74
75
91
113
56, 73, 110
79
115
180
73
118
154
119
179
80
56, 73, 110
183
183
70
On transcription
The phonemic transcription used here is a variant of what is known as
the “Standard Orientalist” Transcription (SOT) (Masica, 1991, p. xv),
which in turn is based on the Americanist transcription system. The
rationale for choosing this over the IPA, which would otherwise have
been the natural choice, was twofold: it is the de facto standard in texts
on the languages of the Indian Subcontinent and it is somewhat more
suited to the language than the IPA, especially in having a clear, separate
set of affricate symbols, most of which are well-known from Latin
orthographies used in Eastern Europe. Most other orientalist symbols
correspond to their IPA equivalents, with one important difference: a dot
under a character always indicates a retroflex or retraced phoneme, in
the same position as the base character. (There is one exception: ḷ, which,
represents a velarised alveolar lateral, a sound not used in Dameli but in
several neighbouring languages.) Somewhat non-standard, even within
SOT, is the use of double vowels for long vowels, which allows some
flexibility in the placement of accents. 1
This system is entirely convertible to an IPA-based system, using table
2. In phonetic transcription, when discussing the actual sounds, the IPA
is used throughout. All characters not shown in the table are the same
as in the IPA.
Table 2: Transcription conversion table
Orientalist
ṭ
ḍ
ċ
č
c̣
š
IPA
Description
ʈʂ̑
ʃ
Retroflex voiceless stop
Retroflex voiced stop
Dental voiceless affricate
Post-alveolar voiceless affricate
Voiceless retroflex affricate
Voiceless post-alveolar fricative
ʈ
ɖ
ʦ̑
ʧ̑
1
This feature is not really needed for transcribing Dameli, at least not within the current
analysis, but double vowels are also easier to combine with diacritics than for example
a superscribed line, the other common symbol for long vowels.
ix
On transcription
ž
ṣ
ẓ
ṇ
ʒ
ʂ
ʐ
ɳ
ḷ
ɬ
λ
l ̥, ɫ
ṛ
w
y
ẉ
ɽ
β̞
j
ɻ
Voiced post-alveolar fricative
Voiceless retroflex fricative
Voiced retroflex fricative
Retroflex nasal
Velarised alveolar lateral
approximant
Voiceless alveolar lateral
approximant/voiceless alveolar
lateral fricative
Retroflex tap
Voiced labial approximant
Voiced post-alveolar approximant
Voiced retroflex approximant
In contrast to the phonemic principle usually employed in transcription,
some non-phonemic contrasts have been maintained to increase clarity.
This applies to [ŋ] and [æ], both of which are interpreted as conditioned
allophones (of /n/ occurring before velar stops, and coalescence of /a/ and
/e/, respectively), but which are nonetheless used in transcriptions.
Similarly, the distinction between u and o, and between uu and oo, is not
analysed as a phonemic contrast, but has been maintained in
transcriptions, to facilitate different analyses and comparison with other
languages. When words contain epenthetic semivowels that may or may
not have phonemic status, these are sometimes expressed in the
transcription as though they did.
x
Acknowledgements
This has been a long journey, in more than one sense, and I am so
grateful to all the people who have helped me along the way.
Firstly, to my supervisors, Maria Koptjevskaja Tamm and Eva
Lindström, who have read and reread my texts, challenging me to revise
and clarify and many times helped me rescue my ideas from illegible
convoluted phrases. Östen Dahl also served as my supervisor for the first
period of my doctoral studies, and has given me many valuable insights
and suggestions both then and after, up to the last weeks.
A generous scholarship from Lydia och Emil Kinanders Stiftelse and
P A Siljeströms Stiftelse allowed me several precious months of
undisturbed work when my time was finished but my dissertation was
not. A travel grant from Knut och Alice Wallenbergs stiftelse allowed me
to combine a conference trip to India with my last visit to Pakistan,
effectively funding my travel costs for that journey.
I would not have come to Pakistan in the first place had it not been
for Henrik and Maarit Liljegren, who invited me along to work with
them at the then Frontier Language Institute (now Forum for Language
Initiatives). I am ever indebted to them and their children, Johanna and
Jonathan, for making my stays in Pakistan endurable and enjoyable,
showing me the ropes of living there as a foreigner and hosting me in
their homes for long periods. Henrik has been my colleague and mentor
throughout my studies, and we have spent many pleasant hours
comparing our favourite languages and finding new and interesting
aspects. In the last stages, he also gave me a host of valuable insights,
comments and suggestions on my manuscript.
I would not have started this work, nor been able to carry it out, had
it not been for the enthusiasm, hard work and sharp minds of my main
Dameli co-workers, Asmatullah Khan and Hayat Muhammad Khan. It
was they who first insisted that someone needed to start working with
their languages, and it is through their diligence and insights that I was
able to continue studying it. I am grateful for the time they spent working
with me, for hosting me in their homes and at their workplaces and for
being great friends and co-workers.
xi
Introduction
In a wider sense, I owe a debt of gratitude to the speakers of Dameli
in and outside the Domel Valley. Too many people to fit on these pages
have helped me understand their language, by recording texts,
explaining words, helping me analyse and sort out complicated matters,
as well as hosting and helping me on my visits, serving me delicious food
and caring for my safety and well-being in my visits to the Domel Valley.
My colleagues and friends at the FLI were a great support and
inspiration during my time in Pakistan and afterwards. I am particularly
grateful to Naseem Haider and Fakhrud Din for their hospitality when I
visited their home villages and for their insightful comments on the
languages and cultures of Chitral, and to Muhammad Zaman Sagar for
interesting data on Gawri. To Wayne and Valerie Lunsford, and their
sons Sean and Jordan, I am grateful for opening their home to me on one
of my visits, and for many great times together.
Several people working on languages in the area have also been very
helpful in providing important material and pleasant conversation on
our shared interests, including Ron and Gail Trail, Matthias Weinreich,
Jan Heegård, Joan Baart, and, much missed, Carla Radloff.
Closer to home, I had the privilege of working at the Department of
Linguistics at Stockholm University, surrounded by great friends and
colleagues, who have all contributed in their way to this work. The other
doctorate students, particularly Kerstin Lindmark, Niklas Jonsson,
Niklas Öhrström, Thomas Hörberg, Francesca di Garbo, Yvonne
Agbetsoamedo, Benjamin Brosig and Robert Östling, helped me grow as
a researcher and develop my texts in tutorials and discussions. I am
particularly grateful to Andrea Kiso, who shared most of this long race
with me, and who on a crucial day reminded me to apply for the
scholarship mentioned above, despite the fact that it she herself was
applying as well (although we both got it in the end). Sofia Gustafsson
Čapkova helped me with a lot of things in the last critical days.
Lamont Antieau served as proof-reader and language checker, and did
a wonderful job pointing out errors and inconsistencies. There would be
many more mistakes and illegible or confusing passages had it not been
for his efforts, and those that remain are solely my own responsibility.
I am also grateful to all my dear friends who have helped me through
these years in many different ways, particularly to Mike Löfdahl and
Shalika Marker, for their unceasing friendship and for uncountable
curries and pleasant evenings.
Last but far from least, I am grateful to my family. My parents,
Elisabet and Sören Perder, have always supported me in every possible
xii
The name of the language
way even when it included me being away for many months in a strange
and sometimes dangerous region. I am grateful for their patience and
prayers, help and support, concern and encouragement throughout these
many times difficult years. My sister and brother, Frida Grenholm and
Hannes Perder, have also shared in the joys and difficulties of this
project. I would surely have lost my mind if my sister had not made sure
I kept in touch with reality occasionally by spending time with her and
her husband Tobias Grenholm, and their three children, who were all
born during the time of this work: Elvin, Nelson and Leo, being
reminded that sometimes the first words of a child are more important
than words on a piece of paper, and that life contains not just verbs and
nouns, but also hugs, fairy tales, swings, dinosaurs, lullabies and fencing
with sticks.
Working with Dameli has been a joy and a privilege, and I am
grateful, to God and to my friends, both those mentioned here and those
space did not permit me to name, for this time and all the adventures,
insights and experiences it has given me.
xiii
Background
1 Introduction
In the southwest of the Chitral district in northern Pakistan, just near
the border with Afghanistan, a river-carved valley stretches through the
mountains of the Hindu Kush range. The rest of the world, to the extent
that it knows about it at all, calls this valley Domel or Damel, but its
own people knows it as Daman.
On the slopes of these mountains some 5 000 to 6 000 people have
their homes and their villages, their fields and their pastures. There they
praise and complain, laugh and lament, joke with their friends and insult
their enemies, tell stupid lies and brilliant stories, ask questions and give
answers, lead the prayers through the mosque loudspeakers and whisper
secrets in the dark and do all the other things that words serve to do, in
a language that has been different from any other for hundreds of years.
This book is an attempt to describe that language.
When my friend and colleague, Henrik Liljegren, decided to call his
twice-as-long and thrice-as-detailed thesis on neighbouring Palula
“Towards a Grammatical Description of Palula”, he put me in a
somewhat difficult position. I would have to go to almost ridiculous
lengths to out-humble that, 2 so in the end I resolved not to try, and
decided on the present title. It is true, as far as it goes: it is a description,
and what it tries to describe is mostly grammar. I do not in any way
mean to imply by this that it is an exhaustive description, or a complete
grammar.
1.1
The name of the language
To be honest, the rest of the title is equally problematic. By right, it
should really be “A Grammatical Description of Damiabaasha”, since
the speakers, when they refer to the language while speaking it, call it
“daamiabaaṣa” rather than “Dameli”. The word is a compound of the
2
I considered calling it “A Brief and Rather Disorganised Attempt at Writing Something
that Vaguely Resembles Some Sort of Description of the Grammar of Dameli.”, but my
supervisors didn’t think this had enough pzazz.
3
Introduction
adjective “daamiaa”, which is the name for the people, and the word for
language, “baaṣa”. The Domel Valley is known as “daaman” in Dameli.
The term ‘Dameli’ is probably an exonym, most likely from the
neighbouring language Khowar, but it has been used consistently in
scientific literature. Although this is not the term used in the language
itself, it seems to be relatively accepted as a term in Urdu, English and
other languages, and when the Dameli speak these languages themselves,
they use this term, although some have indicated that they would rather
see some variant of the Dameli term used instead.
Language names seem to be exonyms almost by default, so rather than
introducing a new term and confusing the situation further, I will use the
established term ‘Dameli’, just as I use a foreign word when referring to
my own mother tongue when I write about it in another language.
The ISO-639-3 code for Dameli is dml.
1.2
Genealogical classification
The linguistic heritage of Dameli is a combination of the well known and
the mysterious. On the one hand, Dameli is clearly a part of the IndoEuropean family, the largest (in terms of languages, at least) and most
exhaustively described of the world’s language families, and both the
grammar and the lexicon give abundant evidence of this heritage. On the
other hand, it is very difficult to pinpoint the exact position of Dameli
within this family and to ascertain its closest relatives within it.
Indo-European
European
Iranian
Persian
Indo-Iranian
Nuristani
Pashto
Hindi/Urdu
Indo-Aryan (Indic)
Gujari
"Dardic"
Figure 1: Selected parts of the Indo-European family tree
Within the Indo-European family, there are two nodes that may be
relevant for Dameli: the Nuristani family and the “Dardic” group (see
figure 1).
4
Genealogical classification
In the title of his 1942 article on Dameli, Morgenstierne left the
question of its genealogical affiliation open, by calling it a “KafirDardic” language, i.e., stating that it belonged either to the “Kafir” or
the “Dardic” language group or family.
In a later article (Morgenstierne, 1961), Morgenstierne suggests that
the “Kafir” branch might form a separate branch of Indo-Iranian,
neither Iranian nor Indo-Aryan (p. 139). Strand (1973) develops this
position, but suggests renaming the branch “Nūristāni” 3 to avoid the
derogatory “Kafir” ‘unbeliever’ (p. 297). Figure 1 reflects this view.
The other part of Morgenstierne’s original classification, Dardic, is
almost equally controversial. As a label, “Dardic” is applied to a number
of Indo-Aryan languages spoken in the Hindu Kush mountains in northeastern Afghanistan and north-western Pakistan, usually Dameli (dml 4),
Gawarbati (gwt), Gawri (gwc), Grangali (nli), Indus Kohistani (mvy),
Kalasha (kls), Kashmiri (kas), Khowar (khw), Palula (phl), Pashai (aee,
glh, psi, psh), Sawi (sdg), Shina (scl), Shumashti (sts), Tirahi (tra),
Torwali (trw) and Wotapuri-Katarqalai (wsv) (Morgenstierne, 1961;
1974; Strand, 1973; 2011). Kundal-Shahi (shd), recently described by
Rehman (2011), should probably be added to these.
Opinions vary on whether to use the term Dardic at all, and with what
meaning. Mock (2011) argues eloquently and comprehensively for
abandoning the term altogether, and several authors, including Strand
(2001, p. 251), apply the same solution. In contrast, Zoller (2005, p. 10
ff.) uses the term more or less as a family designation, although he argues
that these languages are best understood using a punctuated equilibrium
model rather than the traditional Stammbaum model. His arguments for
the classification are based on “the preservation of the three OIA
sibilants s, ś, ṣ” (p. 10), i.e., shared retentions, an argument normally not
accepted as evidence of a language family.
Bashir (2003) considers Dardic an “‘umbrella term’ [which] includes
both ‘genetic’ and geographic components. The designation ‘Dardic’
implies neither ethnic unity among the speakers of these languages nor
that they can all be traced to a single stammbaum-model node” (p. 822).
As Bashir (2003) and Liljegren (2008, p. 31) point out, there are
similarities between these languages, but the explanation for this is more
complex than the Stammbaum model allows for: “The similarities of the
Dardic languages today are due to differentially shared retentions,
3
This use follows the name of the Afghan province Nuristan ‘land of light’.
Letters in parentheses are ISO-639-3 codes for the languages. See (Lewis, Simons, &
Fennig, 2013).
4
5
Introduction
innovations affecting various subsets of these languages and contact
(areal) developments” (Bashir, 2003, p. 822).
Morgenstierne later appears to have become convinced that Dameli
belonged in the Dardic group (Morgenstierne, 1961), and Strand (1973;
2001), also support this, based on aspiration, tones, and what he calls
“basic laryngeal posture” . 5 Although this last feature is unconventional,
I agree with Strand that speakers do use an unusual and characteristic
voice quality when speaking Dameli, and I have no reason to doubt his
conclusion that this is a point of similarity that Dameli has with the IndoAryan languages and a point of difference with the Nuristani languages.
What is not yet clear, however, is if this feature is stable enough to be
considered a strong argument for family affiliation. Do languages tend
to maintain their “laryngeal posture” over centuries, or is this a feature
easily acquired through contact?
While the opinions of Morgenstierne and Strand certainly weigh
heavily, a systematic comparison of regular sound changes and
grammatical features would be highly desirable. This is, sadly,
something I have not yet been able to investigate in any depth.
It seems reasonable that Dameli should have a closer relationship with
some languages than with others, even within these languages. Strand
(2001, p. 259; 2011) puts it, tentatively, in the Western Kohistani group,
together with Gawri. While there are many compelling similarities
between Dameli and several different languages, it is often hard to
determine what is due to shared heritage and what is due to later
language contact.
A detailed investigation into the relationship of Dameli to other
languages, perhaps beginning with a systematic comparison of Dameli
and Gawri, the Nuristani languages and probably also Gawarbati 6, is a
key candidate for further research on the subject.
5
Strand describes this as “a characteristic laryngeal quality, the exact physiological
nature of which remains to be determined. The sound quality appears to be ‘anterior
voice’ (Catford 1977: 102 f) with a slight forward displacement of the epiglottis, brought
about by constriction of the upper larynx.” (Strand, 2001, p. 254)
6
Apart from Palula, which, because of its relationship with Shina is unlikely to be closely
related to Dameli, Gawarbati showed the highest percentage of lexical similarity with
Dameli in the comparison made by Decker (1992, p. 120).
6
Dameli and other languages
1.3
Dameli and other languages
The Dameli language bears many signs of intensive contacts with several
other languages throughout its history. Decker (1992) reports lexical
similarities of between 29 and 44% with a few of the surrounding
languages, after comparing a list of 210 words in the respective
languages (p. 120). These similarities are probably due to both shared
heritage and language contact, in varying degrees.
Currently, the strongest influence on Dameli is exercised by Pashto,
the provincial language of the province where Dameli is spoken, the
Khyber Pakhtunkhwa province, formerly known as the North-West
Frontier Province. Pashto is also the main language of Peshawar and
Jalalabad, the nearest major cities, but its principal influence stems from
its position in the immediate locality of Dameli. The closest town, Drosh,
is dominated by Pashtun shopkeepers, and Pashto is the language most
commonly used in the bazaar, although not the only one. More
importantly still, there is a considerable presence of Pashtuns within the
Domel Valley itself. Pashtuns of the Mashmani, Katani and Wardak
clans have been established in the valley for several generations, in
relatively large numbers, and contacts with these groups are close and
frequent, and includes intermarriage.
In all these cases, the dialect with which Dameli speakers are in
contact is the eastern or Ningrahar dialect (Tegey & Robson, 1996, p.
6), although several tribal dialects or subdialects are represented.
Most – probably all – adult Dameli men speak Pashto, probably at
near mother tongue level, and young boys are able to communicate freely
in the language. Among women, there are probably some monolingual
Dameli speakers, but knowledge of Pashto is widespread among the
women as well.
The position of Dameli in relation to Pashto remains strong enough
for many Pashtuns in the valley to learn Dameli at least passively. During
my visits to the valley, it was a popular joke to berate Pashtun
neighbours for having lived there for three generations without learning
the language, when this angreez (foreigner) was speaking it already. The
Pashtuns who were thus attacked defended themselves by claiming that
they understood the language perfectly, although they didn’t speak it, a
claim that seemed well founded.
Influence from Pashto is obvious in a large amount of borrowed
words and expressions. Words for features of modern society and
technology, which typically have Urdu origins, probably often reach
Dameli through Pashto.
7
Introduction
Urdu, being the national language of Pakistan, also plays an
important role for Dameli speakers. Like Dameli, Urdu is an Indo-Aryan
language, and it is sometimes hard to determine if similarities between
the two are due to contact or shared heritage, but it is clear that a large
number of Dameli words have Urdu origins. Urdu is the official medium
of instruction in schools, and the only language most Dameli-speakers
learn to read and write. It is also the language used for wider
communication beyond the Pashto-speaking area.
There are considerable differences between the Urdu spoken as a
mother tongue by a minority, mainly in Karachi, and the Urdu used as a
lingua france throughout Pakistan. The latter is heavily influenced by
Punjabi, and it is mainly this version that Dameli speakers come into
contact with.
How well a person knows Urdu is closely related to their level of
education.
Khowar enjoys high status within the district of Chitral, but its
influence is relatively limited in the south of the district, where Dameli
is spoken. Still, many Dameli speakers have learnt Khowar, and it
remains an important language in the area.
At present, Dameli-speakers intermarry with speakers of Pashto,
Khowar, Palula, Gawarbati, Kataviri and to some extent Gujari . When
a Dameli man marries a woman from these communities, the wife
reportedly usually learns Dameli.
Palula is a language of the Shina group, spoken to the northeast of
Dameli. Until the road to Domel was completed, the closest path to the
outside world went over the Pashtan mountain pass to Ashret, a Palulaspeaking village. Dameli-speakers differentiate between speakers of the
two main dialects of Palula, and refer to them as aċareti ‘Ashreti Palula’
and biuẉyaa ‘Biori Palula’ respectively. Many signs indicate that contact
with Palula has been even closer at some point in history than it is today.
Many words are more similar than what would be expected from the
shared heritage, and in some cases words that are used in both languages
are completely transparent in one, but not in the other.
Gawarbati is spoken in Arandu, southwest of the Domel Valley,
which makes it an important neighbour, but with limited presence within
the Domel Valley. Kataviri is a Nuristani language, spoken in Nuristan
in Afghanistan. A dialect of Kataviri, known as Shekhani, is spoken in
Shekhan Deh in the Bumburet valley; the Dameli interact with speakers
of both dialects. Gujari (or Gojri) is spoken by the Gujar people, spread
over northwestern and central India, northern Pakistan and northeastern
8
Previous research
Afghanistan (Rensch, Hallberg, & O'Leary, 1992, s. 91). Contact and
intermarriage with this group is relatively limited.
The history of the Dameli people, as it is remembered by its elder
members, is full of stories about interactions with “Kalasha”, although
it is unclear whether these are the same as the current Kalasha people,
or some other people, since the name has also been applied to other
groups and languages, e.g., kalaṣa-alā, another name for Waigali, a
Nuristani language (Strand, 1973, p. 299). The Dameli people
reportedly do not intermarry with the Kalasha.
A few well-educated Dameli-speakers speak English, but the language
also has some indirect influence on Dameli, as a source of loanwords
reaching the language through Urdu or Pashto.
1.4
Previous research
For seventy years, the main source of information on Dameli has been
the article “Notes on Dameli, a Kafir-Dardic language of Chitral” by the
Norwegian linguist Georg Morgenstierne (Morgenstierne, 1942).
Although it was published in 1942, the article was based on material
collected more than ten years earlier. Morgenstierne did not visit the
Domel Valley, but worked for a while with two Dameli-speaking
informants. Despite an informant situation that was far from ideal, 7 his
investigation is very comprehensive and informative, thanks especially
to Morgenstierne’s detailed knowledge about the languages in the area
and their history. The article is oriented towards historical comparison,
and contains a wordlist of 1020 entries and a translated narrative.
The only other primary sources of information on the Dameli do not
have the language as such as their main focus. Kendall Decker (1992)
devotes a chapter to Dameli in the fifth volume of the Sociolinguistic
Survey of Northern Pakistan, with a lot of valuable information about
the sociolinguistic situation of the Dameli speakers. This work also
contains a Dameli text and a small standardised wordlist of 210 words.
Augusto and Alberto Cacopardo likewise dedicate a chapter to the
Dameli speakers in their fascinating work on the ethnohistory of
southern Chitral, Gates of Peristan (Cacopardo & Cacopardo, 2001).
Although priceless for the cultural and historical background it provides,
the language itself lies mostly beyond the main scope of the book. It does,
however, contain an appendix by Richard Strand (2001), called The
7
One of the informants was a prisoner in chains.
9
Introduction
Tongues of Peristân, that gives an outline of the linguistic aspects of the
area the book describes, including some very interesting observations
about Dameli, particularly useful because they are at least partially based
on primary information from Dameli speakers.
A few other sources contain some information on Dameli, but are
based on second-hand information, primarily from Morgenstierne. This
includes Turner (1962-85), Grjunberg (1999), and Bashir (2003).
1.5
Variation within the language
Despite the relatively small number of speakers, there are some clear
dialect distinctions within Dameli. Most isoglosses appear to separate on
the one hand the villages of Shinteri and Dondideri, and on the other
Aspar, but this is at best a tentative conclusion, and will require further
study. The geography of the Domel Valley makes such a division likely,
however, since the valley branches into these two side-valleys, with high
mountains separating them further in.
Even Morgenstierne, although he excluded the possibility of dialects
in such a “tiny language community”, had to grudgingly accept that the
language was “not altogether uniform” (Morgenstierne, 1942, pp. 119120). Most of the variation he noted is still present, although it is
unlikely that it always reflects dialect differences.
Notable dialectal distinctions include a preference for pronouncing
[u] rather than [o] in Shinteri/Dondideri, as compared to Aspar, and
some lexical differences. The Aspar dialect favours baloodadi/balooyi over
bap/yei for ‘grandfather/grandmother’, and makes more use of kinship
terms that distinguish relative age, e.g., ẓeṣṭadadi/sureedadi ‘FoB/FyB’
rather than pitri ‘paternal uncle’.
There is also some variation on the basis of social factors such as
education, status and level of exposure to other languages. As has been
observed for South Asia in general (Masica, 1991, pp. 90-93), there is a
lot of difference between speakers in how “foreign” sounds in borrowed
words are treated. For those with more education, maintaining certain
sounds and distinctions not otherwise present in the language becomes
a matter of status. In Dameli, this probably accounts for differing
pronunciation of for example [f]/[pʰ], [ʤ]/[ʒ] or even [k]/[q], at least in
some words borrowed from Arabic, Persian and Urdu.
10
Language vitality
1.6
Language vitality
In UNESCOs Atlas of the World’s Languages in Danger, Dameli is listed
as “Severely endangered” (Elnazarov, 2010). The entry on Dameli was
contributed by Hakim Elnazarov, and was based on information in
Decker (1992).
This classification is based on nine criteria described in a report by
the UNESCO Ad Hoc Expert Group on Endangered Languages (2003).
The first six criteria describe the situation of the language in itself, the
following two assess attitudes to the language and the last concerns the
amount of documentation done on the language.
Since the report is not very explicit on how different criteria have been
weighted against each other, it is hard to evaluate the conclusion, but I
see no reason to argue against it as such.
It is important to note, however, that the situation of Dameli is not
altogether dark. On the first criterion, intergenerational language
transmission, often taken to be the clearest indicator of language vitality,
Dameli would rate as “Stable but threatened”, since just as previous
generations, children born of Dameli parents learn Dameli as their first
language, although many Dameli speakers reach a very high fluency in
one or several other languages, and are forced to use these in some
domains, notably education at all levels and communications with
outsiders. Similarly, Dameli speakers differ from speakers of many of the
surrounding minority languages in showing no reluctancy to speak their
language even in the presence of non-speakers, and appear to lack most
of the negative associations that minority speakers worldwide tend to
attach to their language. Coupled with the enthusiasm and
determination shown in anything that affects the common welfare of the
Dameli people, the future of Dameli is not without hope.
The second criterion is absolute number of speakers. The Atlas cites
the 1987 census, which states 5 500 speakers of Dameli (Elnazarov,
2010), a slightly higher number than the 5 000 estimated by Decker
(1992, p. 118). Using the annual population growth for Pakistan
provided by the World Bank gives us an estimate of almost 9 500, but
the value of such calculations is doubtful, since the population growth
of all of Pakistan is not a good estimate for the population growth of
this very limited area, and since the numbers they are based on have in
any case been cast into serious doubt. The results of the 2011 census,
which includes data on mother tongue (Population Census
Organization, Government of Pakistan, 2012), should provide some
useful information, if and when they become accessible.
11
Introduction
In any case, the Dameli population is relatively small, 8 and therefore
vulnerable to factors affecting populations in this volatile region.
1.7
Geographic location
Dameli is spoken in the Domel Valley, a side-valley of the Kunar Valley,
which leads from Chitral over the border to Afghanistan. The only road
leading into the valley, completed quite recently, takes off from the
Mirkhani-Arandu road. This allows vehicle travel to the nearest town,
Drosh, or across the Afghan border. Most vehicles are operated as
private businesses, taking paying passengers in a semi-organised system
of communal transport.
Footpaths across the Pashtan mountain pass lead to the village of
Ashret, but the journey requires several hours of strenuous walking.
Mobile phone coverage does not extend to the Domel Valley, and
there is no connection to the landline telephone grid. On my last visit, in
2008, there was a single telephone serving the entire valley, powered
with solar panels and utilising a commercially available radio-based
system with low call costs.
1.8
The Dameli people
The speakers of Dameli live mainly on agriculture and herding. Goats,
sheep, cows and hens are kept, and dairy products are an important part
of the diet. The main crops are wheat and maize, but a wide variety of
fruits and nuts are also grown, particularly grapes and walnuts, but also
pomegranates, mulberries, peaches, melons, apricots, pears, apples, figs
and a number of fruits I have never encountered anywhere else, such as
the blue gokhilan, a smaller relative of the persimmon, and berries like
the badar and c̣ʰunc̣ʰo.
Valuable deodar trees grow on the ridges of the mountains, but the
Dameli claim they see little or nothing of the profits from logging, which
is controlled by the government. The tough, stunted hollyoak trees that
grow on the slopes are used for firewood and as fodder for goats, who
eat the twigs and the sharp, spiky leaves.
8
For the purposes of survival, that is. In comparison with the rest of the world’s
languages, it is probably close to the median (Parkvall, 2006, p. 55).
12
The Dameli people
Working outside the valley for longer or shorter periods has become
a common way to earn more money. Dameli speakers have acquired a
reputation as good cooks, and are employed in various functions from
other areas of Chitral to Peshawar, Karachi, Afghanistan and as far as
the Gulf. Some have also settled permanently outside the valley.
The Dameli speakers are Sunni Muslim, and the mosques play an
important role in village life, particularly as the loudspeakers used for
the prayer call also function as a medium of mass communication. Some
boys or young men are sent to study in madrassas, religious schools, in
Peshawar and elsewhere, often as a part of a religious career with the
aim of becoming a mullah, a religious leader on the community level.
To some extent, the principal villages — Aspar, Dondideri, Swato
(Punagram), and Shinteri Kuru — function as organisational units, but
they are also divided into clans that sometimes reside in several villages.
The clan names are based on the names of ancestors to which the clan
members count their lineage, by adding the suffix −dari ‘clan’ to the
name.
Kinship relationship on a level closer than the clan play an important
role in Dameli society, not least because these relationships regulate
inheritance, loyalty and taboos.
Following the customs of surrounding Pashtun tribes, the Dameli
society is governed by a strict division between the sexes, purdah. This
principle is considered an Islamic injunction, and roughly corresponds
to similar divisions in other conservative Muslim societies. In its
character and details, however, this is clearly a Pashtun custom, having
been introduced by contact with the surrounding Pashtun tribes as part
of conversion to Islam. Since purdah is only applied in relation to nonrelatives, kinship relations become especially important.
13
2 Methods and data collection
I first came in contact with Dameli in 2003, as I was working with an
organisation then called Frontier Language Institute (FLI), now Forum
for Language Initiatives (still FLI), in Peshawar, Pakistan. Its purpose,
then and now, was to equip speakers of minority languages in northern
Pakistan to develop and support their own language communities. In this
area, minority language speakers often have a low opinion of their own
language, but the speakers of Dameli had been clamouring for someone
to start working on their language.
2.1
Field research
One of my tasks at the institute was to mentor one Dameli speaker, Mr
Asmat Ullah, through one of the institute’s programs. Together with
Asmat, who was and is one of the most active advocates of his language,
I began collecting material on the language from him and others. When
I returned to Sweden after a year in Pakistan, I used some of this material
for a Master’s thesis in general linguistics entitled “The Sound System of
Dameli”, at the Department of Linguistics of Stockholm University.
When a position as doctoral student opened in the same department that
same autumn, I applied, with the hope of continuing my studies of
Dameli and the people who speak it. To my surprise and great pleasure,
I was accepted and I started as a doctoral student in January 2005.
The data for this thesis has mostly been collected in Peshawar, where
a number of Dameli speakers reside permanently, and where many
others regularly visit. I made three trips to Chitral, and twice visited the
Domel Valley itself. Although these visits were only of a few days each,
they offered invaluable opportunities to collect material and check
conclusions in a more natural context, to understand the physical and
social backdrop to the stories and words I had been collecting, and to
try my hand at speaking the language with people who were not used to
speaking other languages with me.
14
Text corpus
My first year in Pakistan took place between August 2003 and August
2004, and included two trips to Chitral (November 2003 and June-July
2004), and one to the Domel Valley itself (July 2004). I was working as
a teaching consultant at the FLI during this time, but within this work I
had the opportunity to collect a great deal of useful material, and made
my first and most important contacts with Dameli speakers, particularly
with Asmat Ullah Khan and his nephew Hayat Khan, without whom this
thesis would probably never have happened.
My next visit, October-December 2005, came as part of the doctoral
program, with the explicit aim of collecting material for the thesis. Since
this visit took place in winter, when the mountain pass to Chitral is
closed, I did not visit Chitral during this visit, but worked with speakers
in Peshawar.
My next and last visit (yet) to Pakistan took place from March to
August 2008, and included spending July in Chitral, working with Hayat
Khan in Drosh and spending a few more days in the Domel Valley. Apart
from the time in Chitral, this visit involved working with Dameli
speakers in Peshawar, including a Dameli writers workshop held at the
FLI.
Although the bulk of the data collection took place in Pakistan, the
opportunity to communicate with particularly Asmat over telephone,
and later email and chat, even when back in Sweden, has been important
for checking words and word forms, and ascertaining that I was not
entirely incorrect in my tentative conclusions.
2.2
Text corpus
Recorded and transcribed texts form the core corpus used in this thesis.
Most of these were first recorded from a single speaker in one session,
then transcribed, translated and analysed together with the speaker or
another Dameli speaker, primarily Hayat Muhammad Khan and Asmat
Ullah. With some of the later texts, I performed most of the transcription
and analysis myself, and checked unclear passages with a speaker when
I had the opportunity. There is never enough time for this kind of work,
and the texts are in various stages of completion.
Some of the texts were recorded as part of two writers workshops,
where Dameli speakers tried writing in a tentative writing system
constructed on the basis of the Urdu alphabet, which all of them already
knew, with additions for sounds for which there are no letters available.
15
Methods and data collection
In the first workshop, which took place in Drosh from July 12-14 2004,
the speakers wrote their texts and then read them aloud in the recording.
This applies to the texts T0, T1, T4 and T5. In the second one, on the
th
17 of May 2008, the participants were recorded telling the stories that
they planned to write, rather than reading the finished product.
While I have tried to vary the genres and types of texts as much as
possible, the fact that I could not spend longer periods of time in the
Domel Valley itself, and other limitations of time and access to speakers,
has hindered me from exercising enough control over the material to
create a balanced corpus of texts in terms of genre and type of material,
and there’s a disporportionate number of narrative texts, and not nearly
enough dialogues and natural conversations, in the present corpus.
For the earlier texts, I used a regular cassette recorder and a lapel
microphone. By the time of the last field trip, digital recorders had
become more easily available, and I used an Edirol R-09, without an
external microphone, recording 48Khz, 24-bit stereo sound in
uncompressed Wave format.
Table 3 lists the texts used in this thesis. The last column contains an
identifier, which forms the first part of the unique identifier used in all
language examples throughout this work. Thus any example with an
identifier beginning in T4 comes from the text Adil Jesta, etc.
Table 3: Text corpus
Name
Length
Speaker
Date
Id.
Revenge
The Patient Woman
Shinteri
Two Ancestors
Adil Jesta
Pashtan
History
Additions
Genealogy
Love Like Salt
House
Domel
Three
Weeks
in
Domel
Pear Story - Asmat
Pear Story - Hayat
03:12
04:14
03:11
07:02
03:41
02:51
13:02
01:59
04:28
04:29
05:27
03:57
Hayat M. Khan
Asmat Ullah
Gul Ahmad Khan
Sayed Ahmad Jan
Ahmad Nur
Qadir Said
Patijan Hawaldar
Sayed Ahmad Jan
Sayed Ahmad Jan
Asmat Ullah
Muhammad Sayed
Amin Ullah
2004-07-14
2004-07-14
2007-07-20
2007-07-20
2004-07-14
2004-07-14
2007-07-19
2007-07-20?
2007-07-20?
2008-08-13
2004
2004
T0
T1
T2
T3
T4
T5
T6
T7
T8
T9
TA0
TA1
03:45
Asmat Ullah
2008-08-13
TA2
02:16
00:51
Asmat Ullah
Hayat M. Khan
2008
2008
TP0
TP1
16
Wordlists
Workshop
Introduction
Saudagar
Kabul
Moral story
Omar
Gramgal
Difficult
Safarnaama
Gramgal 2
Nat Introduction
First Nat
Second Nat
Ghazal 1
Ghazal 2
Ghazal Bewafa 1
Ghazal Bewafa 2
Joking Ghazal
Watan
Maize
17:39
Asmat Ullah
2008-05-17
TW0
15:46
07:58
04:15
04:31
02:55
04:32
02:46
04:34
00:53
07:02
10:10
01:30
01:40
01:01
01:14
03:03
01:33
00:59
2008-05-17
2008-05-17
2008-05-17
2008-05-17
2008-05-17
2008-05-17
2008-05-17
2008-05-17
2008
2008
2008
2008
2008
2008
2008
2008
2008
2007-07-27
TW1
TW2
TW3
TW4
TW5
TW6
TW7
TW8
TV0
TV1
TV2
TV3
TV4
TV5
TV6
TV7
TV8
TD0
Ripening Fruits
08:35
2008-07-29
TD1
Fairy Tale
02:26
2008-07
TD2
Living room talk
04:44
2008
D0
Phone Call 1
01:53
Xaista Muhammad
Minhaj
Sifat Ullah
Shafiq Ullah
Siraj Ul-Abedin
Asiz Ur-Rahman
Sher Zaman
Waqar
Asmat Ullah
Asmat Ullah
Asmat Ullah
Asmat Ullah
Asmat Ullah
Asmat Ullah
Asmat Ullah
Asmat Ullah
Asmat Ullah
Izat Wali
Muhammad Isa
Khan
Khaista
Muhammad
Hayat M. Khan,
Siraj Ul-Abedin, +1
Asmat Ullah
2008
D1
2.3
Wordlists
A “master wordlist” was collected and added to continuously during the
time I worked with this thesis. The first incarnation of this list was made
by Mr Asmat Ullah, as part of his coursework while I mentored his
participation in a course for mother-tongue language workers. To this
was added nearly every word of Dameli I came across in texts,
conversations or through elicitation. To each entry was added a
recording of the text in a frame, usually “We say … to this” - “ay masa
ki … ganuma.” As with the analysis of texts, there were always more words
than I had time to check with the speakers, and recordings for many
17
Methods and data collection
words are still lacking, particularly those I came across after my last visit
in Pakistan.
Apart from an English translation, most entries were classified
according to semantic domain and word class, and other information
was added when available and relevant, such as references to other
entries, to examples in the texts, to cognates in other languages or to
possible etymologies. At the time of this writing, the wordlist contained
1900 words. For practical reasons, a separate wordlist, and another list
of morphemes, were used for the semi-automatic process of
interlinearisation and glossing.
2.4
Paradigms
Paradigms of grammatical forms were important both as a way of
eliciting new data, a tool in interpreting other forms of data and of
course as part of the presented results. I spent considerable time eliciting
and analysing verb forms and derivation patterns, pronouns, and kinship
terms, among others.
2.5
Questionnaires
Preprepared questionnaires were used to investigate certain subjects,
when such were available and looked promising. A questionnaire
intended to investigate patterns of ergativity and accusativity was
administered by Henrik Liljegren in Drosh in 2005, and it provided
useful data on this and on pronouns and verb forms in general. Examples
from this questionnaire have the identifier Q0.
I also used a picture questionnaire prepared by Bernhard Wälchli to
investigate spatial relationships. In this task, respondents are asked to
describe a series of pictures pairs, in which objects are placed in
particular spatial arrangements, with superimposed arrows indicating
the relevant part of each picture. The pictures, which, for example, show
first a book under a chair, then the same book on top of the chair, were
very valuable in eliciting data for describing e.g., postpositions,
peripheral case and adverbials. Language examples from this
questionnaire have the identifier Q1.
Finally, I made an attempt at collecting data with the tense and aspect
questionnaire presented in Dahl (1985). Unfortunately, I started
elicitation using the English version of the questionnaire, which proved
18
Other kinds of data
to be to difficult for my informants, and I decided to abandon this task
before completing the second part of the questionnaire. While the results
I got have been useful, given the opportunity I would redo the whole
questionnaire using the Urdu translation that is also available. Language
examples from this questionnaire have the identifier Q2.
2.6
Other kinds of data
Apart from the data described above, I kept chat logs and records of
email conversations, a list of elicited example sentences (language
examples from this list appear with the E identifier), a list of idioms
(identifier TI) and a collection of coordinates for place names within the
Domel Valley.
19
Phonology
3 Phonemes
The phoneme inventory of Dameli consists of 47 phonemes (16 vowels
and 31 consonants). There are four basic vowel qualities, but length and
nasality multiply the number of phonemes. Consonants are articulated
in five basic positions and as stops, affricates, fricatives, nasals, trills or
taps, and approximants. There are also voiced and voiceless stops and
fricatives, and a set of aspirated stops. Tonal contrasts are significant,
with (at least) a rising and a falling tone, but this is an area that will
require further study.
Figure 2 shows the phonemes of Dameli, excluding the nasal vowels.
The nasal vowels are articulated in the same places as their oral
counterparts.
Consonants
Bilabial
Stop
p
pʰ
b
Affricate
Fricative
Nasal
Trill or tap
Approximant
m
Dental
t
tʰ
ʦ̑
s
d
z
n
r
l
ʋ
Postalveolar
Retroflex
Velar
ɡ
ʧ̑
ʃ
ʈ͡ʂ
ʂ
k
kʰ
x
ɣ
ʒ
ʈ
ɖ
j
ʐ
ɳ
ɽ
ɻ
Central
Back
Vowels
Front
Close
Mid
Open
i–iː
u
oː
e–eː
a
Figure 2: The phonemes of Dameli - IPA
ɑː
23
Phonemes
To ease comparison, phonemes, but not transcribed words, are written
in the International Phonetic Alphabet (IPA) throughout this chapter.
For historical reasons, a different system, one based on orientalist
transcription, is used for phonemic transcription here and elsewhere in
this work. For phonetic transcription and expressing allophones, the IPA
is always used.
Instructions for converting between the two systems, where they
differ, are found in the section on transcription at the beginning of this
work. For clarity, figure 2 is also repeated here in orientalist
transcription as figure 3.
Consonants
Bilabial
p
pʰ
Stop
b
Affricate
Fricative
Nasal
Trill or tap
Approximant
m
Dental
t
tʰ
ċ
s
d
z
n
r
l
w
Postalveolar
Retroflex
Velar
ɡ
č
š
c̣
ṣ
k
kʰ
x
ɣ
ž
y
ṭ
ḍ
ẓ
ṇ
ṛ
ẉ
Vowels
Front
Close
i–ii
Mid
Open
Central
e–ee
Back
u
uu
a
aa
Figure 3: The phonemes of Dameli – Orientalist transcription
3.1
Vowels
The vowel system of Dameli is based around four basic vowel qualities
and a length contrast. Nasal varieties of all vowels exist, both long and
short, although short nasal vowels are uncommon except as diphthongs
or vocalic sequences, which may hold relevance for their phonemic
24
Vowels
status. Although rhotic vowels appear phonetically, they are interpreted
as sequences of vowels and a retroflex approximant, and are treated
elsewhere (see 3.2.2).
3.1.1
Vowel quality
Four basic vowel qualities make up the Dameli system: /i/, /e/, /a/ and /u/.
Table 4 shows minimal pairs in support of the contrast in both long and
short vowels.
Table 4: Vowel contrasts
Contrast
a|e|i|u
Words
žân 9 ‘snake’
žen ‘eat’ (INF)
žin ‘eat’ (IMPFV.3PL)
žun ‘daughter’ (KIN.2)
ɑː | eː | iː
kaar ‘ear’
keer ‘when’
kiir ‘snow’
ɑː | uː | eː
taa ‘past tense marker’(PST)
tuu ‘you’ (2SG.OBL)
tee ‘that’ (conjunction)
iː | oː
čiir ‘late’
čuur ‘four’
In actual pronunciation, several more qualities come into play, both
because quality is also important for the length distinction and because
of secondary phonological processes.
3.1.2
Vowel length
The length contrast is expressed by the relative duration of the vowel,
but also through vowel quality, and is applied to all the four basic vowel
qualities. Table 5 shows minimal pairs in support of the length contrast.
9
The circumflex (^) represents falling tone. See 3.3.
25
Phonemes
Table 5: Length contrasts within the same quality
Contrast
a | ɑː
i | iː
u | uː
e | eː
Short vowel
daš ‘hand; ten’
ki ‘to, for’
tu ‘you’ (2SG.NOM)
c̣ʰaare ‘throw.IMP’
Long vowel
daaš ‘cubit’
kii ‘who’
tuu ‘2SG.OBL’
c̣ʰaree ‘throw.PFV.3SG’
Differences in vowel quality are important for distinguishing length
contrasts, particularly for the vowels /u/ and /a/. The short /a/ is a front
vowel, [a], whereas the long vowel is a back vowel [ɑː].
The short /u/ is usually a closed sound [uː], whereas the longer variant
is usually more closed-mid [oː]. Here, though, the situation is
complicated by a greater variation. There are many examples of long [uː]
and short [o] in actual pronunciation, but the contrast between the two
sounds does not seem to be phonemic and no minimal pairs supporting
the distinction have been found. Dialectal differences partially explain
the differences in pronunciation, as speakers from Aspar tend to favour
a closed-mid pronunciation [o] in some contexts, whereas speakers from
Shinteri tend to use a more closed pronounciation [u] regardless of length.
Figure 2 reflects the difference in vowel quality, but the symbol u is
preferred in phonemic transcriptions.
This analysis, wherein the two vowel qualities [u] and [o] are treated
as variants of the same phonemes (/u/ and /uu/), is tentative and may
need to be reevaluated at a later stage. For this reason, transcriptions
that distinguish between the two sounds, when they appear to be
pronounced differently, have been kept throughout this work, to
facilitate later reanalysis or comparison with other languages. Note that
this is a deviation from the phonemic principle normally applied in the
transcription. For /i/ and /e/, differences in quality between the long and
the short vowel are less obvious.
Short vowels can be stressed, but stress usually falls on a long syllable
if there is one. It is often difficult to tell different short, unstressed vowels
apart, and it is possible that some contrasts are neutralised in these
contexts and likely that the pronunciation of all unstressed vowels is
more central than otherwise.
Although long and short vowels of similar qualities are often
contrastive, they are in some ways related to each other, as shown by
processes such as conditioned vowel lengthening (6.2), which causes a
short vowel to be exchanged for a corresponding long vowel in some
26
Consonants
contexts. In the phonemic transcription used in this work, long vowels
are written with double vowels (eg., kaar ‘ear’).
3.1.3
Nasality
Nasal vowels are usually either long or found in vocalic sequences,
probably because they originated as nasals that were neutralised between
vowels, as indicated by syncronic variation in pairs such as kʰurdani
‘ankle’ and naasdãĩ ‘nasal bridge’, where the element dani/dãĩ probably
refers to a narrow, connecting part of the body.
Short nasal vowels are uncommon outside of diphthongs or vocalic
sequences, but they do exist (e.g., ãš ‘token’), although possibly not for
all vowel qualities. Table 6 shows all long nasal vowels, and some
examples of nasalised vowel sequences.
Table 6: Examples of nasal vowels
Vowel
ĩː
ẽː
ɑ̃ː
ũː
ãũ
ɑ̃ːĩ
ãĩ
ẽĩ
3.1.4
Example
ĩĩč ‘eye’
ɡrẽẽṭen (t) ‘tie’
mãã ‘my’ (POSS.1SG−M)
žũũ ‘louse’
ṭãũ ‘full, pure (of colours)’
mããĩ ‘my’ (POSS.1SG−F)
muubãĩ ‘towards me’
perẽĩ ‘fairy, djinn’
Diphthongs and vowel sequences
Sequences of vowels occur both within morphemes and as a result of
combinations of morphemes (see 6.1, particularly 6.1.6). Although these
can function as the syllable nucleus, they have not been analysed as a
separate set of diphthong phonemes.
3.2
Consonants
As shown in figure 2, the Dameli consonantal system can be arranged
into five places of articulation and six manners of articulation. This
arrangement is somewhat idealised; each of these labels should be
understood to cover more than a single place or manner of articulation.
Thus both dental and alveolar pronunciations are subsumed under the
27
Phonemes
dental label, the palatal /j/ is listed under postalveolar and the lateral
approximant /l/ is counted together with the other approximants.
The places of articulation used in the table are bilabial, dental,
postalveolar, retroflex and velar, and the manners are stops, affricates,
fricatives, nasals, trills or taps and approximants.
Voicing and aspiration expand the system by two extra sets: there are
voiced and voiceless fricatives, and both voiced, voiceless and aspirated
stops.
3.2.1
Place of articulation
The places of articulation used to describe Dameli consonants are
bilabial, dental, postalveolar, retroflex and velar. The following is a brief
description of the places and any variation that occurs within them. The
individual phonemes are listed in the next section under their respective
manners of articulation.
Bilabial sounds are pronounced with constriction between the upper
and lower lips. There is some allophonic variation among the bilabial
sounds: a labio-dental fricative [f] occurs as an allophone of /pʰ/,
although a bilabial fricative [ɸ] is probably more common.
Dental sounds include both prototypically dental sounds, particularly
the stops, and sounds such as /n/ and /r/, which tend towards a more
alveolar pronunciation.
The postalveolar sounds are relatively few: there are no stops. The
approximant /j/ is more palatal than postalveolar, as mentioned earlier.
The most atypical place of articulation is the retroflex, which is
definitely defined more by the shape of the tongue than by the position
of the primary constriction. In these sounds, the tongue root is retracted
and the tongue tip brought backwards, although not curled as far
backwards as in, e.g., the Dravidian languages. The retroflex
approximant, which is often only realised on the surrounding vowels,
may sometimes involve bilabial coarticulation.
The voiceless velar fricative /x/ has a glottal allophone [h] that is
particularly common word-initially, e.g., hal ‘plough’.
3.2.2
Manner of articulation
The manners of articulation used to describe Dameli consonants are
stops, affricates, fricatives, nasals, trills and taps, and approximants,
including the lateral approximant.
28
Consonants
The stops are the only sounds with both voiceless, voiced and
aspirated sets (see 3.2.3), which makes them the most numerous,
although the fricatives utilise just as many places of articulation. There
are no postalveolar stops, but bilabial (/p/, /b/ and /pʰ/), dental (/t/, /d/
and /tʰ/, retroflex (/ʈ/ and /ɖ/) and velar (/k/, /ɡ/ and /kʰ/).
The set of affricates is limited to three phonemes, the dental /ʦ̑/, the
postalveolar /ʧ/̑ and the retroflex /ʈ͡ʂ/. A voiced postalveolar affricate [ʤ̑]
exists as an allophonic variant of /Ʒ/, particularly in borrowed words.
Aspirated affricates can frequently be heard, although the distinction
between aspirated and unaspirated affricates does not appear to be
phonemic, but form part of the distinction between primarily the
postalveolar affricate, which is usually pronounced without marked
aspiration, and the retroflex affricate, which is usually pronounced with
clear aspiration.
Having affricates in three positions, as Dameli does, is an areal feature
where Dameli is spoken, and such phonemes are present in several of the
neighbouring Indo-Aryan languages, the Nuristani languages,
Burushaski, and several East Iranian languages (Zoller, 2005, p. 14). An
even larger area has the two affricates ʦ̑ and ʧ,̑ but lacks the retroflex ʈ͡ʂ.
(Masica, 1991, p. 132).
There are eight fricative phonemes, the voiceless /s/, /ʃ/, /ʂ/ and /x/, and
their voiced equivalents /z/, /Ʒ/, /ʐ/ and /ɣ/. Of these, /s/, /ʃ/ and /ʂ/
correspond closely to the three affricates in terms of place of articulation.
The fricative [f] and [ɸ] are common as allophones of /pʰ/.
The voiced retroflex fricative, /ʐ/, is used in a very small number of
words, five in my material to date: maẓi ‘female markhor’, reeẓa ‘line,
narrow path’, ẓami ‘brother-in-law (WB)’, ẓanẓer ‘chain’, and ẓekan
‘leather string for tying woolen leggings, bowyang’. Several of these
words have cognates in Palula (Liljegren & Haider, 2011) and Kalasha
(Trail & Cooper, 1999), where the phoneme is equally uncommon.
Nasals are a very limited set, containing the bilabial /m/, the dental /n/
and the retroflex /ɳ/. A velar nasal is common as an allophone of /n/
before velar stops. In some cases, such as the word mraŋ ‘nail’, there is
no trace of a stop, but this is probably not enough to warrant
interpreting this as an independent phoneme. 10
Just as with the velar [ŋ], the retroflex /ɳ/ usually appears together
with other retroflex sounds, but it is common enough on its own to be
10
Investigating this was made considerably more difficult because my standard frame for
single word recordings involved a /ɡ/ immediately after the word. That’s what you get
for not creating a proper frame with vowels on both sides of the word.
29
Phonemes
tentatively included as a separate phoneme. It never occurs wordinitially.
The trill and tap manner consists of the alveolar /r/, usually
pronounced as either a trill [r] or a tap [ɾ]. The retroflex tap /ɽ/ is
infrequent, and used mostly in loanwords. It is quite distinct, and has
been included as a separate phoneme, but may perhaps also be
interpreted as a variant of /ɻ/.
The approximants are a rather heterogenous set, containing the
semivowels /ʋ/ and /j/, the lateral approximant /l/ and the retroflex
approximant /ɻ/.
A couple of particular l-sounding sounds have developed nearby: “the
velarized ɫ found in Kalasha, Khowar and Palula and the voiceless lateral
λ 11 found in Gawarbati, Kalam Kohistani, Wotapuri-Kataraqalai,
Shumashti and Grangali” (Bashir, 2003, p. 906). Neither of these sounds
are found in Dameli, however, despite the fact that they are found in
Khowar and among the immediate neighbouring languages Kalasha and
Gawarbati.
The retroflex approximant is mainly realised on adjacent vowels,
which are pronounced with a rhotic or retroflex quality, primarily
noticeable by a lowering of the third formant. Morgenstierne (1942, p.
122) interpreted this sound as a palatal fricative sound, but since the
tongue tip does not approach the palate, there is very little constriction,
and the sound does not resemble palatal sounds.
It might be possible to analyse each vowel affected by this sound as a
separate phoneme, in effect a series of retroflex vowels, as has been done
for neighbouring Kalasha (Trail & Cooper, 1999), but for Dameli,
where the phenomenon is less widespread, such an interpretation would
be less economical and would not account as well for the distribution of
the sound, which is mostly limited to vocalic sequences. For several of
these occurrences there are cognates in neighbouring languages that have
ɽ between the vowels so that the postulated retroflex approximant
corresponds directly to a retroflex tap in these languages, e.g., aɡeẉi
‘dupatta, lady’s shawl’, cf. Kamviri aɡaṛik; aaŋɡuẉi ‘finger’, cf. Palula
anɡúṛi.
This character represents either a voiceless alveolar lateral approximant [l ̥] or a
voiceless alveolar lateral fricative [ɫ].
11
30
Consonants
3.2.3
Voicing and aspiration
Voicing and aspiration provide an extra dimension of Dameli
consonants, beyond place and manner of articulation. All stops and
fricatives make a distinction between voiced and voiceless phonemes,
and for stops there is also an (over-)aspirated series.
Both voicing and aspiration depend on the property of voice onset
time, the point at which the vocal cords start to vibrate: voiced
consonants are pronounced with vibrating vocal cords throughout their
duration, whereas voiceless consonants are produced without activating
the vocal chords. Aspirated consonants are also voiceless but have an
extra delay before the vocal cords are allowed to resume their function,
which produces an extra puff of air at the end. Voiceless consonants are
not completely unaspirated, but less so than the aspirated set, which have
emphatic aspiration, meaning that the voice onset is delayed longer and
the airflow stronger. I have not been able to distinguish an aspirated
retroflex stop, but all other stops have aspirated versions.
A related phenomenon is voiced aspirated sounds, which are not
present in Dameli. Unlike voiceless aspirated sounds, these depend on
the breathy voice property, in addition to changes in voice onset time, in
order to produce their particular character (Clark, Yallop, & Fletcher,
2007, p. 53). The absence of voiced aspirated consonants in Dameli
contrasts with the norm in the Indo-Aryan languages in general (Masica,
1991, p. 101) but is a feature it shares with some of its neighbours, e.g.,
Gawri (Baart, 1997, p. 20) and Shina (Radloff, 1999, p. 27). Zoller
(2005) sees in the distribution of aspirated stops among the “Dardic”
languages evidence that “language boundaries and boundaries of
isoglosses are frequently not identical”; voiced aspirated stops are
present in Indus Kohistani, Chail Kohistani and Kohistani Shina but
missing in Gilgiti Shina and Gawri (Kalam Kohistani) (p. 12).
In the absence of voiced aspirates, voicing and aspiration form a threeway distinction so that “aspirated” contrasts with “voiceless” and
“voiced”. Table 7 shows minimal and near-minimal pairs in support of
the voicing and aspiration contrast for stops.
Table 7: Voicing and aspiration contrasts
Contrast
p:pʰ:b
Words
paṇ ‘bale’
pʰan ‘path’
ban ‘stick’
31
Phonemes
3.3
t:tʰ:d
taa ‘past tense marker’ (PST)
tʰaa ‘is’ (be.IMPFV.3SG.M)
daa ‘take this!/here you go!’
ṭ:ḍ
ṭaŋɡ ‘drumbeat’
ḍaŋɡ ‘back’ (body part)
k:kʰ:ɡ
kur (t) ‘do’ (verb root)
kurum ‘I am doing’ (do−PFV.1SG)
kuree ‘he/she did’ (do−PFV.3SG)
kʰur−ee ‘on foot’ (foot−INSTR)
ɡurum ‘morning’
Tone
A “significant correlation of tone” was reported for Dameli by both
Morgenstierne (1942, p. 125) and Strand (2001), 12 but this phenomenon
will require further study before I can do much more than confirm their
conclusions. There are minimal pairs distinguished only by tone,
including monosyllabic words with short vowels. Strand limits tone
contrasts to accented vowels (p. 255).
The tones, according to both Morgenstierne and Strand, are analysed
as rising (e.g., žǎn ‘watermill’, sǎn ‘pastures’) and falling (e.g., žân ‘snake’,
sân ‘hole’), and my cursory inspections of spectrograms and pitch curves
appear to bear out this point. 13
A further analysis of the subject would need to go into both how the
tonal system works and what role it plays in the language as a whole,
questions which must remain unanswered for now.
12
It’s not entirely clear from the text if Strand’s conclusions in this matter are based on
his own observations or on Morgenstierne’s article.
13
I am grateful to Niklas Öhrström for helping me with this task, and for some very
useful off-the-cuff observations.
32
4 Syllable structure
The syllable structure of Dameli is rather restrictive. Consonant clusters
are allowed both in the onset and the coda, but only with a very limited
set of phonemes.
The permitted syllable types are shown and exemplified in table 8.
Note that V may represent either a long or a short vowel.
Table 8: Syllable types
Syllable type
V
VC
VCC
CV
CVC
CCVC
CCVCC
Example
i ‘3SG.ANIM.PROX’
ek ‘one’
iṇc̣ ‘bear’
ni ‘not’
nan ‘aunt (MZ)’
brun ‘wall’
ɡraŋɡ ‘hole’
One pronunciation of the word krnaa ‘black’ may suggest a more
complex syllable type (CCCV), but variant pronunciations argue against
making too much of this single example. See 4.2.
4.1
Onset
The syllable onset is the least restrictive position in the Dameli
phonology. Any consonant except /ṇ/ can appear at the beginning of
syllables, and consonant clusters are allowed but follow strict rules.
Table 9 shows some examples of initial consonant clusters.
Clusters are built around a “central” position with a segment of low
sonority, typically a stop, although there are also examples with nasals
and fricatives. Sibilants may precede this position, and /r/ or an
approximant may succeed it.
The vast majority of initial clusters have /r/ in the second position, but
/y/ and /ẉ/ also occur. In my material, the bilabial approximant /w/ and
33
Syllable structure
the lateral approximant /l/ only occur in this position in loanwords, e.g.,
plag, ‘electrical outlet’, swaato ‘Punagram (village name)’. Restricting the
second position in clusters to r-sounds is common in languages across
the world (Ladefoged & Maddieson, 1996, p. 216).
Additionally, initial clusters appear to be restricted to CC, i.e., central
consonants may not have both preceding and succeeding consonants.
There are no such clusters in the material, and historical clusters of this
kind have had prothetic vowels inserted, e.g., ištrii ‘wife’ <Skt. strii, or
been altered in some other way. Even CC clusters with an initial sibilant
appear unstable, and often have variant pronounciations with a
prothetic vowel, as in e.g., skaa/iskaa ‘fat’. 14
Table 9: Some syllable-initial clusters
Word
braa
truida
mraŋɡ
zrax
ɡyu
pẉeyen
ṣpaṣi
4.2
Gloss
Structure
brother
day after tomorrow
deer
active, clever
oil
squeeze
sister’s child
[stop]+[r]
[stop]+[r]
[nasal]+[r]
[fricative]+[r]
[stop]+[approximant]
[stop]+[approximant]
[sibilant]+[stop]
Nucleus
The syllable nucleus consists of a long or short vowel. Consonants do
not usually function as syllable nuclei. One pronunciation of the word
krnaa ‘black’ suggests either syllabic /r/ or /n/, or a more complex initial
cluster, but this is an isolated occurrence and is often pronounced with
an epenthetic vowel as [krɪˈnɑː].
4.3
Coda
The syllable coda is more restrictive than the onset. The only clusters
that are allowed are those consisting of a nasal followed by a stop or an
affricate, and with the exception of the cluster −ng [ŋɡ], even these are
rare.
14
The same tendency to avoid initial clusters can be observed when comparing related
languages, e.g., Dameli: ṣpaṣi – Kalasha: iṣpoṣí ‘sister’s child’.
34
Coda
The word-final position is very restrictive (see 6.3) and only allows
very restricted clusters, and neither aspirated nor voiced consonants. It
is very likely that some of these restrictions are actually restrictions on
the coda, but that secondary processes obscure this. Because the syllable
structure is usually open to reanalysis (resyllabification), allowing the
coda to become onset in a subsequent syllable, the restriction has less of
an effect on clusters within words than word-finally, and is more difficult
to study there.
The word mukdac̣ʰani ‘looking glass’ shows that at least final
deaspiration applies to the coda within words as well as word-finally.
The word is a compound of mukʰ 15 ‘face’ + dac̣−an ‘look−INF’, but the
aspirated kʰ in the first part is not pronounced. The presence of a stop in
the onset of the second syllable blocks kʰ from being moved to this
syllable (*mu.kdac̣ʰani). Since kʰ remains in the coda, final deaspiration is
applied. In the word mukʰaamuk ‘face to face’, in contrast, there is no
consonant in the onset of the second syllable, and the word can be
resyllabified as mu.kʰaa.muk, with intact aspiration.
It is possible that the restriction on voiced obstruents that applies
word-finally (6.3.3) is also a restriction on the coda but that this is
neutralised in the surface form by regressive assimilation from a
subsequent voiced sound. Likewise, the same restrictions on clusters that
apply word-finally probably apply in the coda within words, when they
cannot be avoided by resyllabification.
15
Note that this is the underlying form; the aspiration is lost when used without suffixes,
since the /kʰ/ is then word-final.
35
5 Prosody
Phonological features above the segment level are important but difficult
to analyse, since they often deal less with absolute meaning differences
and more with nuances and discourse functions. Stress and intonation
interact but function on different levels. Stress is mostly a lexical feature
in Dameli and serves to differentiate between word forms, whereas
intonation is used to mark particular sentence types and perform
discourse functions.
5.1
Stress
In Dameli, as in most languages, lexical stress is a composite feature, that
combines duration, loudness and pitch to make syllables more or less
“prominent” (Clark, Yallop, & Fletcher, 2007, p. 339). Stress is mainly
lexically determined and can fall on any syllable. Some suffixes carry
their own stress, altering the stress of the word they form part of. This
is particularly true for verb endings, which often change the stress
pattern considerably.
Stress and long vowels mostly coincide, although it is often impossible
to determine whether the long vowel attracted the stress or the stress
16
caused vowel lengthening. There are, however, enough examples of
words with long, unstressed vowels, or several long vowels, to warrant
positing length and stress as independent features. One clear example of
this comes from the verbal system. For some verbs, stress is the only
feature distinguishing perfective and imperfective forms, despite the
presence of a long vowel: ˈootʰinum ‘I am stopping’ (stop.IMPFV.1SG) vs.
ooˈtʰinum ‘I stopped’ (stop.PFV.1SG), ˈkookinum ‘I am sleeping’
(sleep.IMPFV.1SG) vs. kooˈkinum ‘I slept’ (sleep.PFV.1SG).
16
I’ve also been wondering about the chicken and the egg, lately. Any ideas, people?
36
Intonation
5.2
Intonation
Intonation involves several features, such as phrasal stress, melody and
tempo, which in turn are made up of lower-level phonetic correlates such
as pitch and duration. While a few very general remarks can be made
here, a thorough investigation of intonation patterns must be left for
further research.
The general intonation pattern is raising-falling, where the pitch peak
often emphasizes some part of the utterance. Particular intonation
patterns mark several special clause types, such as polar questions (rising
all the way to the end), conditions, enumeration and lists. Extreme
lengthening of one syllable can be used for narrative effect.
37
6 Phonological processes
The surface form a phoneme takes is determined by the context in which
it appears. Some contexts initiate phonological processes that change the
phonemes involved. Most such processes are subtle changes that are not
consciously perceived by speakers or listeners, but those described in this
chapter have more far-reaching consequences, often altering the surface
form quite drastically. Particularly common are processes that affect
phonemes and clusters that appear at the end of words.
6.1
Vowel sequences
Two or more vowels appearing after each other often result in systematic
changes. Sounds that are phonetically sequences of vowels can occur
where the phoneme structure only contains one vocalic phoneme, as a
result of a diphthongisation processes that affect the pronunciation of
individual vowels (see 6.1.6).
In some contexts, vowels can be reanalysed as semivowels, which can
be used as consonants, to create regular syllables with vowels and
semivowels according to the normal phontactic rules of the language.
This appears to be quite common with /u/ and /i/ in Dameli (turning them
into /w/ and /y/).
Vowel sequences can also appear when a syllable without an onset is
joined to the end of an open syllable, one without a coda, usually when
suffixes are joined to words that end in a vowel. Such sequences can
result in elision, lengthening, assimilation, the insertion of an epenthetic
approximant, hiatus or diphthongs or vocalic sequences.
6.1.1
Elision
Some single-vowel suffixes, such as the locative –a, usually appear only
on words that end in consonants (or vowels of the same quality, see
6.1.2), even when it is otherwise obligatory, such as before some
38
Vowel sequences
postpositions. This may be interpreted as elision of the vowel suffix that
is done in order to avoid a cumbersome vowel sequence.
6.1.2
Lengthening
When two short vowels of the same quality meet, a long vowel of the
same quality is produced, as in madrasaa ‘madrasah 17−LOC’ (madrasa−a).
6.1.3
Coalescence
A similar process occurs whenever /a/ is directly followed by /e/
(regardless of length). Within words, these two vowels do not usually
appear adjacent to each other, but when a suffix or clitic beginning in /e/
is joined to a root or stem ending in /a/, the two phonemes are assimilated
into a single sound, [æ], which lies between /a/ and /e/ in the vowel space
of the language. If either of the vowels combined (or both) is long, the
resulting sound will also be: [æː].
This happens regularly in the third person perfective forms of
causative verbs, where the causative derivation suffix −a precedes the
suffixes −ee ‘PFV.3SG’ or −en ‘PFV.3PL’, e.g., naɡ−a−een ‘(they) went
down’[ˌnaˈɡæːn]. It is also common when the instrumental or ergative
suffix −ee or the clitic −es ‘also’ are joined to words ending in −a, e.g.,
baṣa−ee ‘language−INSTR/in the language’ [ˈbɑːˌʂæː], žaa−es ‘now also’
[ʒæːs].
Coalescence of vowels with the same quality is also common, resulting
in long vowels.
6.1.4
Epenthetic semivowels
While it is by no means physically impossible to produce almost any
series of vowels, some combinations are largely avoided in Dameli,
mostly by the insertion of epenthetic semivowels. Since there is much
variety in how this principle is applied in different words, the semivowels
are usually expressed in the transcriptions, as though they were
phonematic, although they may not always be. There appears to be some
phonetically motivated systematicity, however:
• Epenthetic semivowels are more common with combinations
that involve pairs of vowels that are pronounced relatively far
away from each other in the oral cavity, particularly in terms of
openness (see figure 2).
17
A madrasah (or madrasa) is an Islamic religious school.
39
Phonological processes
• The semivowels that are involved, /y/ and /w/, correspond to /i/
and /u/, and usually appear when these two vowels are involved.
• /y/ appears more often when the sequence is opening up, going
from a more closed vowel to a more open one: iya, but ai.
• /w/ appears more often when the sequence is closing, going from
a more open vowel to a more closed one: awu, but ua.
A slightly different use of this can be illustrated with the derivation of
causative verbs. The second causative suffix −aai is always used after the
causative suffix −a. To preserve the two suffixes intact, an epenthetic −w−
is inserted between the two: nišawaayim ‘sit−CAUS−CAUS2−PFV.1SG/I made
someone sit someone down.’ 18
6.1.5
Hiatus
When no measure is taken to prevent two vowels to meet on a syllable
border, the result is called hiatus. This occurs in Dameli and can be
illustrated by comparing word pairs such as c̣ai ‘body’ (two syllables,
[ʈʂa.ˈi]) and čay ‘tea’ (one syllable, [ʧaj]). The distinction can be quite
difficult to perceive, particularly if, unlike in the example, stress falls on
the first syllable.
6.1.6
Diphthongs or vocalic sequences
Some long vowels are occasionally pronounced, phonetically, as
diphthongs. This is a superficial process that affects pronunciation, but
does not change the phonemic composition of a word. To exemplify,
long /e/ is pronounced with the diphthong [e͡i], as an allomorph of /ee/
used in addition to [eː], in words such as bree ‘girl’, kuree ‘who’. These
are, of course, quite open to reanalysis as combinations of vowels and
semivowels (e.g., as brey, kurey).
Two vowels can also meet within a syllable. These can be analysed as
sequences of vowels and semivowels, or as diphthongs. To exemplify,
the feminine version of the adjective baloo ‘large’ is not ba.loo.i but ba.loy
or ba.loi. Similar cases occur not just as a result of combining morphemes
but within morphemes as well, as in, e.g., auɡus/awgus ‘dragonfly’.
Whether these sounds should be considered true diphthongs or
sequences of vowels and semivowels is largely a matter of interpretation.
Since they cannot, at present at least, be shown to function as a set of
phonemes in their own right, they have been interpreted as sequences of
vowels and semivowels here.
18
This is a real word, cross my heart!
40
Conditioned vowel lengthening
6.2
Conditioned vowel lengthening
In a sense, coalescence of vowels of the same quality may be considered
an instance of vowel lengthening, since it results in a long vowel, but a
more complex process, that affects closed syllables, also occurs: when a
suffix is attached to certain roots, short vowels in the last syllable of the
root are sometimes lengthened. This process is applied to most words
with a short vowel in the last syllable, including loan-words.
Table 10 shows some examples of vowel lengthening, and table 11
some exceptions.
Table 10: Examples of vowel lengthening
Root
bum ‘ground’
mam ‘maternal uncle’
zyat ‘much’
<Pashto: ziat
rimel ‘kerchief’
<Pashto: rumaal
lemp ‘lamp’
<Eng: lamp
mač ‘man’
Suffix
Example
−a ‘LOC’
zyaata ki ‘to [the word] “much”’
−a ‘LOC’
rimeela walisan ‘hidden in the kerchief’
−a ‘LOC’
−a ‘LOC’
−a ‘LOC’
−es ‘also’
buuma ‘on the ground’
mãã maama ṣawaai ‘through my uncle’
leempa alamaaisan ‘hanging from the
lamp’
tanii maačes ‘that man also’
Table 11: Examples without vowel lengthening
Root
put ‘boy’
muṭ ‘tree’
kul ‘house’
sekal ‘bicycle’
Suffix
−a ‘LOC’
−a ‘LOC’
−a ‘LOC’
−a ‘LOC’
Example
puta ki/putra ki ‘to the son’
muṭa ‘in the tree’
kula ‘in the house’
sekala niši ‘sitting on the bicycle’
The difference between the words that do not undergo vowel lengthening
and those that do is still unclear. Tone differences, vowel quality or
otherwise different underlying forms, are possible explanations.
6.3
Word-final processes
The variety of sounds that are allowed in the word-final position is very
restricted. Clusters are severly limited, and both voiced and aspirated
obstruents are avoided. In both cases, this is carried out with processes
41
Phonological processes
that are at least partially still active; underlying forms may still have the
avoided sounds and combinations.
6.3.1
Reduction of clusters
The syllable structure of Dameli is particularly restrictive in the coda (see
4.3): the only clusters allowed are those consisting of a nasal followed
by a stop or an affricate. This restriction is easiest to observe at the end
of words, since resyllabification allows much more maneuverability
within words.
Evidence from a number of words show that this restriction is not just
a distributional fact, but the result of an active process. Some morphemes
have underlying forms that contain final clusters, which are not allowed
in the word-final position, and which are reduced when the morphemes
are used on their own. When they are used in compounds or with
suffixes, the underlying form resurfaces. Table 12 shows some examples
of this.
Table 12: Reduction of final clusters
Underlying form
dašt
putr
prošt
pʰant
aṣt
Independent form
daš ‘hand’
put ‘son’
proš ‘bed’
pʰan ‘path’
aṣ ‘eight’
Example with suffix
daštee ‘with the hand’
putres ‘his/her son’
prošta ‘in bed’
pʰantee ‘on the way’
aṣṭam ‘eighth’
A similar process can be observed in the adaptation of loanwords. See
table 13 for some examples of this. In at least some of these cases, the
cluster may resurface when the word is used with a suffix, e.g., waxt−a
‘time−LOC’.
Table 13: Loan-words with reduced final clusters
Word
bardaš ‘bear, endure’
sax ‘hard’
wax ‘time’
doos ‘friend’
42
Source
Ur: bardašt
Ur: saxt
Ur: waxt
Ur: dost
Word-final processes
6.3.2
Deaspiration
Another process affects aspirated stops at the end of a word. In such
cases, the aspiration is removed, and the phoneme is pronounced as the
corresponding voiceless unaspirated consonant. To exemplify, the word
muk ‘face’ has the underlying form mukʰ, but this appears only when a
suffix is added, such as in mukʰ−a ‘face−LOC/in the face’. Like reduction
of word-final clusters (4.2), this process leads us to assume underlying
forms of morphemes that are more complex than their isolated surface
forms. Only when the morpheme is combined with another morpheme
does the surface form reflect the underlying form fully.
6.3.3
Obstruent devoicing
In the word-final position, voiced sounds are devoiced. This process
affects obstruents, i.e., stops, affricates and fricatives, but not nasals or
other phonemes.
Interestingly, when the voiced velar stop /ɡ/ appears word-finally
preceded by a nasal as [ŋɡ] it is not devoiced. This argues against the
analysis made here, which does not recognise this as a single phoneme
/ŋ/, since nasals are not devoiced under this rule, and stops are.
This process is most obvious as a distributional observation; the only
words that have been recorded in my dictionary that end in voiced stops,
affricates or fricatives are borrowed ones. In fluent speech, these also
tend to be devoiced, even if they were pronounced with the voiced
sounds in careful pronounciation.
Another instantiation of this process can be observed in subject
marking on the verb: suffixes that express plurality usually consist of the
corresponding singular form followed by −a, but this only holds if we
assume that an underlying voiced form of the singular marker is
suppressed in the word-final position but resurfaces when the plural −a
is added: −num ‘IMPFV.1SG’; −numa ‘IMPFV.1PL’, −nap ‘IMPFV.2SG’; −naba
‘IMPFV.2PL’.
It is not immediately clear on which level obstruent devoicing
operates, whether it is purely a surface operation that removes the
voicing from certain sounds, or if the process depends on the phoneme
structure, and exchanges one phoneme for another.
43
Phonological processes
Baarth (1997), in describing Gawri, notes a process similar to the
Dameli one affecting stops, nasals 19 and affricates but insists that an
audible phonetic difference is maintained between devoiced and
voiceless sounds, i.e., that voiced phonemes that have undergone wordfinal devoicing are pronounced differently than inherently voiceless
phonemes. He supports this analysis with acoustic evidence (pp. 13-15).
Radloff (1999) describes a similar process in Gilgiti Shina affecting
voiced plosives, affricates and fricatives but apparently not nasals (p.
32). Based on observations and speakers judgements, she also concludes
that “there is a difference in the pronunciation of devoiced consonants
and their voiceless counterparts” (p. 33).
With a clear accoustic difference between voiceless and devoiced
sounds, as both these languages appears to have, the process is obviously
just one of removing voicing. My suspicion, although I have not been
able to carry out similar accoustic tests, is that the situation in Dameli is
different and that the process there operates on a deeper level.
I have found no signs that speakers make a distinction between
voiceless and devoiced sounds, and transcriptions and texts written by
mother-tongue co-workers do not reflect such a distinction, 20 but this is
something that needs to be studied more carefully.
Unlike Gawri, but like Gilgiti Shina, devoicing in Dameli affects
fricatives, but not nasals. This means that final devoicing affects all
phonemes that have voiceless equivalents, and no other phonemes,
which might be an argument suggesting that the process is dependent on
the phoneme structure involving a change of phonemes when there are
voiceless corresponding alternatives available.
19
Baarth apparently includes nasals in his category of plosives, since the nasal stops have
devoiced variants in his table of phonetic consonant sounds (p. 11), but not in his table
of consonant phonemes (p. 19).
20
The exception is in the domain of loanwords, where a conservative spelling is often
preferred, regardless of pronounciation.
44
Morphology
7 Nouns
Nouns, which typically “refer to things, persons, places, and other more
or less concrete objects” (Tamm, 2006, p. 720), probably form the
largest class of words in Dameli. As an open class, nouns are frequently
borrowed from other languages, but are also derived through productive
word-formation processes.
Nouns consist of a noun stem, to which inflectional suffixes can be
attached. The stem can consist of just the root, but there are also stems
that have a more complex structure, such as compounds or derivations.
Nouns can be inflected according to number and case and refer to
things that are inherently animate or inanimate. All nouns belong to
either the masculine or the feminine gender, but these categories are not
marked on the nouns themselves, although they can affect marking on
other targets, such as adjectives or verbs that show agreement with
nouns. Note, however, that the gender distinction appears to be in
decline in the language and that speakers often have difficulty accessing
the gender of individual words when asked directly (see 7.3.3).
Nouns form noun phrases which fill a variety of syntactic functions
(see chapter 15).
7.1
Noun stems
Noun stems are the basic lexical form of the noun and can consist of a
bare noun root morpheme or a more complex combination with an
internal morphological structure, such as compounds or derivations.
Some forms of the noun consist of just the stem, but some are formed
by adding one or more inflectional suffixes to the stem.
7.1.1
Noun roots
Noun roots are single morphemes used to form noun stems. They usually
range from one to three syllables.
49
Nouns
7.1.2
Compounds
Several lexical morphemes can be combined in a number of ways to form
complex noun stems, ranging from the very transparent, in which the
meaning of the compound is clear from the meaning of the components,
to the very opaque, in which there is little obvious relationship between
the meaning of the components and the meaning of the compound. Case
and other suffixes are applied at the end, on the last root. No overt
markings are used to indicate compounds. Stress tends to fall on the
normally stressed syllable of the second element. Compound nouns can
be formed with roots that are not nouns, particularly in the first position.
Table 14: Some compound nouns
Word
draakmuṭ
banipʰac̣i
naasdãĩ
učʰuṭadadi
aobác̣ʰo
7.2
First element
draak ‘grapes’
bani ‘hollyoak’
naas ‘nose’
učʰuṭa ‘small’
ao ‘water’
Second element
muṭ ‘tree’
pʰac̣i ‘bird’
dani ‘narrow part’
dadi ‘father’
bác̣ʰo ‘calf’
Meaning
grapevine
a kind of bird
nasal bridge
younger uncle
otter
Animacy
Many constructions in Dameli depend on whether or not the participants
are living entities or inanimate objects. Such a division is quite common
in the world’s languages but the categories involved are not necessarily
identical. Cross-linguistically, the animacy distinction is best described
as a division between things considered more or less prototypically
living. Rather than utilising a two-fold distinction, we may talk about a
scale along which individual languages make different cuts. Some
languages, such as Swahili, treat only humans as animate (Lyons, 1998,
p. 210), whereas others, such as Manam, include some higher animals,
especially domesticated animals, in the animate category (Croft, 2003,
p. 130). Languages may even distinguish several levels of animacy. In
Shona, three levels are distinguished: humans, non-human animates, and
inanimates, and these distinctions govern word order. Other languages
still refer to the animacy scale without reference to discrete categories.
In Navajo, for example, elements higher on the animacy scale must
precede lower elements in the word order for a sentence to be
grammatical (Yamamoto, 1999, pp. 52-54).
50
Gender
In Dameli, two categories are distinguished: animate, which consists
of humans and animals, and inanimate, which consists of plants, nonliving objects and abstract concepts. The animate category is not limited
to domesticated animals, as shown by example 1.
The animacy distinction is primarily a property of nouns and
pronouns (see 8.1.4), or rather of the entities they describe. Nouns are
never overtly marked as animate or inanimate, but the distinction is
evident in its effects on other targets.
Describing animacy is somewhat tricky, because it often appears as a
strong tendency rather than as an absolute rule. Inanimates usually do
not trigger plural agreement in verbs, they tend to be avoided as subjects
of transitive verbs, although examples to the contrary can be found, and
possessive pronouns rarely, if ever, refer to inanimates as owners.
Gender agreement is also more often triggered by animate heads than by
inanimate ones.
There are, however, two cases where the difference is fairly
unambiguous:
The possession marker sãã is only used with animate owners.
Possession-like constructions where the “owner” is inanimate are
constructed with the postposition ta ‘of’ (see 15.3). 21
Most importantly, fully inflected copulas with the t root are used with
animate subjects, such as “a bird” in example 1, and the daru/beru
copula, with very limited inflection, is used with inanimate subjects, such
as “name” in example 2.
(1) piŋ ɡan−i ek pʰac̣i tʰui
ping say−CP one bird COP.ANIM.IMPFV.3SG.F
‘One bird is called “Ping.”’ (T3078)
(2) asili
nam daaman daru
original name Daman COP.INANIM.IMPFV.3
‘The real name is “Daman”.’ (T3040)
7.3
Gender
Although the matter is far from simple, Dameli clearly has a system of
“classes of nouns reflected in the behaviour of associated words”
Even here, there are some counterexamples, such as the set expression sãã waǰa, ‘for
this reason’.
21
51
Nouns
(Hockett, 1958, p. 231), i.e., a gender system. In theory, this system
divides all nouns into two groups, masculine and feminine, which in turn
triggers agreement on verbs, adjectives and possessive pronouns and
markers. In actual behaviour, the gender system appears to be in decline,
and gender agreement appears rather erratically. The distribution points
to a relatively complete system, but one that is no longer used all the
time.
7.3.1
Gender assignment
Gender systems always have a semantic component (Corbett, 1991, p.
63). In the case of Dameli, this means that the semantic notions of sex
and animacy play an important role in gender assignment. Nouns that
denote sex-differentiable entities, e.g., terms for men and women and
male or female individuals of any animal, are automatically assigned to
the gender corresponding to their sex. When other factors do not
interfere, non-sex differentiable entities, typically inanimates, tend to be
assigned to the masculine gender. This is consistent with the
development in several other Indo-Aryan languages, where the neuter
gender of Old Indo-Aryan merged with the masculine to form a twogender system (Masica, 1991, p. 220).
There may be additional semantic principles that affect gender
assignment. Body parts, for example, tend to be feminine even when the
form would suggest otherwise, as evidenced by kʰuṭa in table 15.
Another important factor in gender assignment is phonological. The
gender agreement morphemes on other targets are usually −a for
masculine and −i for feminine. These suffixes correspond to the endings
on many nouns belonging to the same gender, a feature known as overt
gender marking (Corbett, 1991, p. 62). Conversely, nouns ending in
either of these sounds may be ascribed to the corresponding gender.
As table 15 below shows, the semantic and phonological principles
described above do not account for all nouns; there are nouns that end
in neither −i or −a nor belong to any recognizable semantic categories.
The inconsistencies in the use of gender described in 7.3.3 prevented me
from completing an exhaustive investigation of this.
Table 15: Gender assignment for some nouns
Word
žu
put
kʰuṭa
52
Translation
daughter
son
knee
Gender
F
M
F
Assignment criterion
semantic
semantic
semantic
Gender
pirããĩ
nikaa
braa
ištrii
abal
nam
7.3.2
shirt; qamis
wedding
brother
wife
joy
name
F
M
M
F
F
M
phonological
phonological
semantic+phonological
semantic+phonological
unknown
unknown
Gender agreement
Evidence for the gender of a noun comes from agreement on other
targets. In Dameli, verbs, adjectives and possessive pronouns and
markers display gender agreement.
Verbs show gender agreement with the subject in the third person, as
shown in example 3.
(3) manee
dac̣i−na
tee
like.this see−IMPFV.3SG.M that
see
tasãã−Ø
žami tʰui
3SG.ANIM.DIST.NOM 3SG.POSS−M wife COP.ANIM.IMPFV.3SG.F
‘Thus he sees that it is his wife.’ (TW1033)
Gender only appears in some tense/aspect forms; it is limited to the
perfective, the imperfective and the indirect past for intransitive verbs,
and to only the imperfective for transitive verbs, both for simple verbs
and for causatives. The future, the potential past and the imperative
never show gender agreement. Some of the non-finite verb forms may
also take adjective-like gender agreement forms when they function as
attributes of nouns.
Subject-marking forms are portmanteau morphs that also contain
other information, but forms expressing gender always build on the
opposition −a / −i. All gender agreement verb forms can be seen in table
27 (page 95).
Only some adjectives, those that belong to the variable group (see
9.1), show gender agreement. Masculine forms are either unmarked or
end in −a. Feminine forms always end in −i.
Possessive pronouns agree in gender with the possessed entity, but not
with the possessor. The forms, which build on the −a / −i opposition, can
be seen in table 20 (page 79). The possessive marker is reminiscent of
the singular possessive pronouns in form, and shows the same kind of
agreement with the possessed object (see 8.2.1).
53
Nouns
7.3.3
Decline of gender
Despite clear evidence of its existence, there are several indications that
the gender distinction is currently declining in Dameli. Most
importantly, speakers of the language consulted in this study often
expressed uncertainty about the gender of inanimate nouns.
Gender is usually immediately accessible to mother-tongue speakers
of a language and can easily be determined by constructing test sentences
with masculine and feminine agreement that are then submitted to
grammaticality judgements. When presented with such material,
however, my Dameli co-workers expressed doubt and contradicted each
other and themselves. This only happened with inanimate nouns, and
mostly with words that did not carry anything that could be interpreted
as overt gender marking, i.e., words that did not end in −a or −i.
Another indication of the decline of gender in Dameli is the lack of
gender marking on pronouns. According to Corbett (1991), gender
distinctions often arise in pronouns, and from there spread to other
categories (p. 310 ff). Likewise, when a system is in the process of losing
gender distinctions, they typically remain longest on the pronouns (p.
259).
Greenberg (1966) even went so far as to postulate this as a universal:
43. If a language has gender categories in the noun, it has gender
categories in the pronoun. (Greenberg, 1966, p. 113)
Despite the fact that Dameli has gender in the noun, pronouns do not
distinguish gender. Possessive pronouns do show gender marking, but
this is a different matter, since it is not the gender of the referent of the
pronoun, but the gender of the possessed object that is expressed. This
is not a unique development, even among the Indo-Aryan languages.
Masica (1991) states:
Not all the languages that show gender agreement of adjectives and
participles have gender-differentiated pronouns: notably Hindi, Punjabi,
“Lahnda”, 22 and Gujarati do not. (Masica, 1991, p. 224)
The inability of speakers to pass grammaticality judgements about
gender could be due to several factors, including the fact that my main
co-workers did not reside in the Domel Valley, although they speak
Dameli at home, and perhaps also to the unnaturalness of the elicitation
situation. Similar indications were obtained, however, from speakers
22
“Lahnda” refers to the language now more commonly known as Siraiki.
54
Noun inflection
residing in the Domel Valley as well, and even in less unnatural language
situations, although no systematic testing has been carried out.
A more likely explanation of both this insecurity and the distribution
of gender in Dameli, is that the system itself is in decline.
The gender system of Proto-Indo-European is believed to have
originated in a system based only on the animacy distinction; sex-based
distinctions were introduced at a later stage, perhaps after the separation
of the Anatolian languages (Ledo-Lemos, 2000, p. 17).
The systems reconstructed for the ancestors of Dameli, for ProtoIndo-European and later for Old Indo-Aryan, both have three gender
systems with masculine, feminine and neuter, where the neuter is
characterised by a large number of inanimate nouns, and the other two
by predominantly animate nouns (Masica, 1991, p. 220; Beekes, 1995).
Being grammatical gender systems, however, these systems showed
considerable exceptions, as do their modern-day descendants. The
languages that now have only two genders have reached this result
through merging the neuter and the masculine (Masica, 1991, p. 221).
What is happening in Dameli is probably a continuation of this
development, in which Dameli gradually loses the grammatical gender
distinctions, but still maintains the semantic distinctions that underly it,
primarily the animacy distinction that has remained central to the system
from its origin.
This development seems already to have taken place in neighbouring
Kalasha and Khowar, both of which have lost gender, but have a
grammaticalised animacy distinction that mainly governs the use of the
copula (Bashir, 1988, pp. 39-40, 125; 2007, pp. 212-213), a distinction
that is already in place in Dameli.
7.4
Noun inflection
Noun inflection is comparatively limited in Dameli. Nouns are inflected
for number and case. Kinship terms constitute a special category and
have several special inflection forms, including a person or anchor
category that is not used with other nouns. With the exception of these,
and some borrowed words that have different plural suffixes, there are
no declensional classes; all nouns are inflected with the same number
and case suffixes, with some phonological restrictions and adaptations.
Number marking is optional, and number and case marking do not
co-occur.
55
Nouns
7.4.1
Number
Nouns can exist in singular or plural form. With the exception of kinship
nouns (see section 7.5.4), the plural form is always formed in the same
way; there are no declensions, although some borrowed words,
described below, appear to retain their plural suffixes.
The standard plural suffix is −nam. It can be applied to any noun, with
the exception of mass nouns that do not have a count noun
interpretation. Example 4 shows a typical use of the plural suffix.
(4) muu−bãĩ
dac̣−i
xu mẽẽ
1SG.OBL−towards see−CP but 3PL.PROX.NOM
baati−nam wail−an−baṣ
tʰ−aa
taa
word−PL hide−INF−able be−PFV.3SG.M
PST
‘You looked at me, but I was able to hide these words.’ (TV5012)
The singular form of nouns is morphologically unmarked, and using the
plural form is optional even when several entities are being referred to.
Because of this, the singular form can be said to refer to objects of
undetermined quantity rather than single objects.
With nouns, the plural can be described as the marked form; it is
overtly marked with a suffix, and it can usually be omitted, if the plural
meaning is irrelevant or can be inferred from the context, as can be seen
by comparing example 5 and 6. 23
(5) waapas aaċ−i
baara aaċ−i
ta
paai mili
back
come−CP away come−CP PART boy with
madad kur−i tee ṭaaŋɡu−nam ral ku−nun
help
do−CP that pear−PL
up do−IMPFV.3PL
‘Returning back, coming from afar, they help the boy pick up the
pears.’ (TP0028)
23
Both examples are retellings of the same visual input (The Pear Story), by two different
speakers.
56
Noun inflection
(6) daara ɡ−aa
fall
go−PFV.3SG.M
ba
duu traa zaatak aaċ−i
TOPSH two three child
come−CP
tas−a
mili ṭaaŋɡu ral ku−nun
3SG.ANIM.DIST−LOC with pear
up do−IMPFV.3PL
‘After [the boy] falls down, two or three children come and pick up the
pears with him.’ (TP1011)
In keeping with this, nouns specified by numerals do not usually take the
plural form, as illustrated by the word zaatak ‘child’ in example 6 above.
A likely origin of the −nam suffix is the OIA genitive plural −ānām
(Masica, 1991, p. 241). Perhaps caused by such a historical origin as a
combined case and number suffix, regular number marking never
co-occurs with case marking. Since number marking is optional, this
does not cause conflicts.
There is a special plural form used only with kinship nouns, −suu, and
a small number of kinship terms have irregular plural forms (see section
7.5.4). The only other exception is that words borrowed from Pashto
sometimes retain their Pashto plural suffixes, as shown in table 16
below.
Table 16: Borrowed Pashto roots with Pashto suffixes
Word
maasum ‘child’
dušman ‘enemy’
musulman ‘muslim’
kapʰar ‘unbeliever’
pulis ‘police(man)’
kom ‘people’
kal ‘year’
Source
Psht: mašum ‘child’
Psht: dušman ‘enemy’
Psht: musulman ‘muslim’
Psht: kaafer ‘unbeliever’
Psht: pulis ‘police’
Psht: qom ‘nation, people’
Psht: kaal ‘year’
Plural
masumaan
dušmanaan
musulmanaan
kaapʰaraan
pulisan
koomuna
kaluna
As several of these examples show, some of these words are not of Pashto
origin (e.g., pulis, musulman). It is likely that many loanwords enter
Dameli via Pashto, as the suffixes indicate these may have done.
There is a curious instance in my material, where a speaker uses the
Pashto plural suffix, −una, with a word that is not from Pashto, as shown
in example 7.
(7) manuu nam−una yat
bin
thus
name−PL remembrance be−FUT.3PL
‘Thus the names are remembered.’
57
Nouns
The Pashto word for ‘name’ is num; nam is the Dameli form. It seems
likely that the Pashto plural is used in order to avoid the cumbersome
*nam−nam ‘name−PL’. Whatever the reason, this example shows that −una
is at least occasionally available as an alternative plural form and that
the borrowed plurals are perceived as morphologically complex, and not
as merely inherently plural entities.
7.4.2
Case
As is common in relatively complex case systems, there are two sets of
noun cases in Dameli, with different functions. Following Blake (1994,
p. 34), we will use the terms core and peripheral case to denote the two
sets. The core cases mark the main arguments of the clause and the
peripheral cases mark various other relationships to elements of the
outside world, functions often filled by adpositions in other languages.
The core cases are a small group of very widely used forms.
Following Dixon (1994), we will use the terms S, A and O 24 to
describe the primitive relations between the core arguments of the verb,
defined in the following way:
S – intransitive subject
A – transitive subject
O – transitive object (Dixon, 1994, p. 6)
The peripheral cases are much more heterogeneous. It is not uncommon
for one morphological form to belong to both sets simultaneously, or for
a core case to develop from a form that was initially used only in a
peripheral function. In Dameli, the core case ergative and the peripheral
case instrumental share the same suffix but are described separately,
since their functions are so widely separated, and since they typically
attach to words from different groups of nouns (animate and inanimate).
In contrast, the locative functions both as an oblique and a traditional
locative, but these two are harder to separate, and are treated together.
Delimiting case in relation to e.g., postpositions is always difficult,
and it would certainly be possible to argue for considering the possessive
marker sãã a case marker. The possessive marker appears to be an
independent word, however, and shares more similarities with the
pronouns than with the case suffixes, in that it agrees in gender.
24
Note that this is in contrast with the Leipzig Glossing Rules (2008), which use P for
O.
58
Noun inflection
7.4.2.1
Split ergativity
The most common pattern of argument marking among the languages
more closely related to Dameli is known as split ergativity. Split ergative
systems are those in which some part of the system adheres to the
ergative principle (i.e., treating subjects of intransitive verbs in the same
way as objects of transitive verbs) and another part to the accusative
principle (i.e., treating subjects of intransitive verbs in the same way as
subjects of transitive verbs). In the case of the Indo-Aryan languages, this
split occurs along the lines of tense/aspect (Masica, 1991, pp. 341-346).
With past or perfective verb forms (the perfective, indirect past and
potential past in Dameli), the system is ergative. With nonpast/imperfective verb forms (the imperfective and future in Dameli), the
system is accusative. This is the basic principle behind the case system of
Dameli. The noun case system is basically summed up by this
description, but the pronoun case system introduces some further
complications (see 8.1.5). The only deviation in the noun case system is
that objects are always unmarked and are therefore not distinguished
morphologically from subjects in the accusative parts of the system.
Note that case assignment is governed by the tense/aspect value of the
verb form, not of the entire verb expression. A complex tense expression
with the imperfective and the past tense marker taa (see 16.5), for
instance, would get a past tense reading but would not trigger ergative
case assignment.
Nouns have two different core case forms in Dameli; although there
are several different core functions, they are carried out by only two
different morphological forms: the unmarked/nominative form and the
ergative form. Likewise, there are two peripheral case forms in the
language, the locative and the instrumental. Kinship terms have a
separate form that may be considered a case, the vocative. This is
described in 7.5.3.
7.4.2.2
Nominative case
With non-past/imperfective verb forms, all core functions, i.e., S, A and
O, are expressed with the unmarked form of the noun, referred to here
as the nominative case. This form is also used to express S and O with
past/perfective verb forms. Since we are talking about a morphological
case form, the term is used to refer to this form regardless of whether it
is used in a “nominative”, “accusative” or an “absolutive” function, or
indeed any other. Example 8 illustrates the nominative case form in a
typical nominative function, as subject in a clause in the imperfective.
59
Nouns
(8) i
žami
c̣ʰiir brikin−ni
3SG.ANIM.PROX woman milk sell−IMPFV.3SG.F
‘This woman is selling milk.’
7.4.2.3
Ergative case
In the past/perfective TAM sets, the subjects of transitive verbs (A) are
expressed by a form marked with the suffix −ee, referred to here as the
ergative case. The subjects of intransitive verbs (S) and the objects of
transitive verbs (O) are still expressed with the unmarked nominative
case form.
(9)
mas−sãã
putr−es−ee
tasãã−Ø
3SG.ANIM.PROX−POSS son−KIN.3−ERG 3SG.POSS−M
ɡram ta ek mač žan−ee
village of one man kill−PFV.3SG
‘His son killed a man from his village.’ (T4016)
The ergative form of nouns is identical to the instrumental (see 7.4.2.5),
but there is little risk of ambiguity, since the ergative is usually applied
only to animates, and the instrumental only to inanimates.
7.4.2.4
Locative case
The locative is formed by adding the suffix −a to the noun stem,
sometimes leading to a lengthening of the vowel in the preceding syllable
(see 6.2).
The locative form functions both as a locative proper and as an
oblique, that is, certain postpositions require their complements, the
noun preceding the postposition, to be in the locative form (see 12.1).
The typical function of the locative, however, is to mark the noun as a
location, as something in or at which something else is, as in example
10. In contrast, goals are not expressed with the locative (unless also
marked with a postposition).
(10) tu
kul−a
tʰop−i
2SG.NOM house−LOC be.IMPFV.2SG−Q
‘Are you at home?’ (D1019)
60
Noun inflection
Just as in the corresponding prepositions in English, the locative can be
used to express a number of figurative “locations,” e.g., at a time, in
grief, in someone’s eye or heart, or in front. Temporal uses of the locative
are very common.
The locative can be used to encode the third argument of a ditransitive
verb such as give, as illustrated in example 11, but this function is often
carried out by postposition phrases as well.
(11) ɡraami sãã-Ø
faisala ta mutaabik šorunḍa paay-a
villager 3SG.ANIM.POSS-M decision of according orphaned boy-LOC
žu-es
prai-tʰee
daughter-KIN.3 give-INDIRPST.3SG
‘Following the decision of the villagers, he gave his daughter to the
orphaned boy.’ (T4021)
7.4.2.5
Instrumental case
The instrumental case is identical in form to the ergative; it is formed by
adding the suffix −ee to the the end of the noun. Instrumental case is
applied to nouns denoting objects being used as tools, i.e., something
that a task is carried out with. Naturally, these are almost exclusively
inanimate. Example 12 shows a typical use of the instrumental.
(12) ay
braadun−ee
wiɡi−num
1SG.NOM catapult−INS
fire−IMPFV.1SG
‘I am shooting with a catapult.’ 25 (E0020)
As noted in section 7.4.2.3, there is virtually no overlapping or
ambiguous cases between the ergative and the instrumental, in spite of
the identical form.
A number of semantically extended uses of the instrumental case exist
with certain groups of nouns. When the instrumental case is applied to
place names, it means ‘by way of’, indicating a place along the way to
somewhere else, a function sometimes referred to as prolative. This is
illustrated in example 13.
25
A slingshot, for American readers. As far as I know, Dameli experiences of Roman
war machines have thankfully been very limited.
61
Nouns
(13) aakuri−ee
Akuri−INS
ṣin
aaċ−i
pʰaarey
mukʰ−a
above come−CP opposite_side front−LOC
ek ɡoor−i duurui
ɡan−numa
one white−F small.valley say−IMPFV.1PL
‘Going by way of Akuri, up on the opposite side, there is a
place we call Gori Durui.’ (T7006)
Instrumental case is also used with words denoting languages to express
how something is said in that language, as in example 14.
(14) ootaa šekʰaan
baaṣa−ee
ɡan−nun
ziat−a
ki
ota
Shekhani language−INS say−IMPFV.3PL many−OBL to
‘In the Shekhani language, 26 they say “ota” for “many.”’ (T8061)
Even beyond such groups, the instrumental case is not limited to concrete
readings, as can be seen in example 15.
(15) ɡan−ee
say−PFV.3SG
tee badšaa yee baati−ee
that king
this word−INS
sax
kʰapa b−aa
hard angry be−PFV.3SG.M
‘Then the king became very angry by these words.’ (T9030)
7.5
Kinship nouns
Kinship nouns, nouns that signify kinship relations, constitute a separate
group among the nouns. They show all the characteristics of nouns in
general, but there are a number of morphological forms that only exist
for kinship nouns. They are very frequent, often serving in contexts
where, in a different cultural setting, personal names or other descriptive
terms would be used.
7.5.1
Kinship system
The kinship term system in Dameli takes as many as four dimensions
into account: relative generation, gender, relative age, and whether the
relationship is by blood or by marriage.
26
The language spoken of here is also known as (Eastern) Kativiri, a Nuristani language.
Shekhani is a term applied to those who convert to Islam. (Decker, 1992, pp. 130-131)
62
Kinship nouns
The generation to which a relative belongs, in relation to the anchor,
is always relevant for kinship terms. Altogether seven generations are
distinguished: the generation of the anchor and three generations
preceding or following him or her. The third generation preceding the
anchor is only used in the term for ‘great-grandfather’, parbap, and the
extent to which extent this term is used is uncertain. Generally, the term
for ‘grandfather’, bap, is used in the sense of ‘ancestor’ as well as in the
literal sense.
All kinship terms distinguish the gender of the relative. For some
terms, there is a masculine and a feminine variant of the same root, such
as nawaasa/nawaasi ‘grandson/granddaughter’; in others, the terms are
completely separate. The terms for kin by marriage also take the gender
of the ego, the person whose relative we are referring to, into account:
the term for a woman’s mother-in-law is not the same as that used for a
man’s mother-in-law. The gender of intermediate relatives can also play
a role: there are different terms for maternal and paternal uncles and
aunts, for example, but no distinction between maternal and paternal
grandparents.
The terms for ‘brother’, ‘sister’, ‘uncle’, ‘aunt’ and ‘brother-in-law’
distinguish relative age, that is, the age of the relative in relation to a
relevant person. In the case of sibling relations, this person is the ego:
there is, for example, a specific word for older sister. For inlaws, it is the
husband/wife, and for aunts and uncles, it is the parent. Dameli culture
ascribes a particular position to the elder brother of a family. In terms of
responsibilities and respect, an elder brother is treated similarly to a
father. An elder brother often takes over the position of the father on his
death and becomes head of the extended family. Two to three levels are
distinguished: there is a term for ‘middle uncle’ and ‘middle aunt’, but
none for ‘middle brother’ or ‘middle sister’. These distinctions are
reportedly used more frequently in Aspar than in the other villages. All
the terms that distinguish relative age also have a general term: braa
‘brother’ can be used for any brother, elder or younger; pitri ‘paternal
uncle’ can be used for any of the father’s brothers, although more specific
terms exist.
The terms for wives and husbands of one’s relatives are often identical
to the term for a corresponding direct relative. Thus, the husband of
one’s nan ‘mother’s sister’ is called mam, which also means ‘mother’s
brother’. When specific terms for affinal kin exist, they are mostly unique
and unrelated to the terms for relatives of blood: the term for ‘brotherin-law’ has nothing to do with the term for ‘brother.’
63
Nouns
A small number of more general terms that are also used to denote
kinship relations behave like kinship terms to some extent and may take
some of the morphological forms that are exclusive to these. These
include the terms for ‘boy’, ‘girl’, ‘child’, ‘tribe’, ‘bride’ and
‘bridegroom’. Figure 4 and figure 5 show the kinship terms used in
Dameli in graphical form; 27 table 17 is a somewhat more extensive list of
the same data.
27
The kinship charts were drawn in the standardised system used in social and cultural
anthropology. Triangles denote male individuals and circles female individuals. An
equals sign (=) denotes that the two individuals are married, vertical lines denote descent
and horizontal lines sibling relationships (i.e., common descent).
64
parbap
bap, baloodadi
pʰaapa pʰapi
pas
nan
yei, balooyii bap, baloodadi
pitri
ẓami bražei
dadi
braa
bawi
nawaasa
kaṛwaasa
yii
EGO
put
nawaasi
kaṛwaasi kaṛwaasa kaṛwaasi
yei, balooyii
nan
mam
ištrii, žami
žuu
mam mamaani
saaraani saaraana
braažei?
zaamaa
nawaasa
nawaasi
kaṛwaasa kaṛwaasi kaṛwaasa
Figure 4: Kinship chart, male ego
pašur
kaṛwaasi
ǰeṣṭaali
ẓami
65
66
parbap
bap, baloodadi
pʰaapa pʰapi
pas
nan
yei, balooyii bap, baloodadi
pitri
saraana bražei
dadi
braa
bawi
nawaasa
kaṛwaasa
yii
EGO
put
nawaasi
kaṛwaasi kaṛwaasa kaṛwaasi
yei, balooyii
nan
mam
bareu
žuu
mam mamaani
žamili
deer?
pašur
preeš
braažei?
deer
zaamaa
nawaasa
nawaasi
kaṛwaasa kaṛwaasi kaṛwaasa
Figure 5: Kinship chart, female ego
kaṛwaasi
Kinship nouns
Table 17: Kinship terms
Term
parbap
bap
baloodadi 29
yei
balooyi
dadi
yii
pitri
ǰeṣṭadadi
mažuma dadi
sureedadi 30
učuṭadadi 31
pʰaapi
žeṣṭeri pʰaapi
mažumi pʰaapi
surui pʰaapi
nan
žeṣṭi
mam
braa
bay
suree
pas
bibi
surei
put
Corresponding
English term
Consanguineal
great-grandfather
grandfather
grandfather
grandmother
grandmother
father
mother
uncle
uncle (elder)
uncle (middle)
uncle (younger)
uncle (younger)
aunt
aunt (elder)
aunt (middle
aunt (younger)
aunt
aunt (elder)
uncle
brother
brother (elder)
brother (younger)
sister
sister (elder)
sister (younger)
son
Description 28
kin
FFF +
FF, MF
FF, MF
MM, FM
MM, FM
F
M
FB
FoB
FmB
FyB
FyB
FZ
FoZ
FmZ
FyZ
MZ (FBW)
MoZ
MB (MZH)
B
oB
yB
Z
oZ
yZ
S, (BS)
28
The abbreviations used in this column are those commonly used in anthropological
descriptions, derived from the first letters of the equivalent term in English, with the
exception of sister, which is represented by Z. A lower-case o means ‘elder’, m ‘middle’
and y ‘younger’. Codes enclosed in parentheses denote secondary, extended uses of
terms, such as when using “daughter” to talk about a niece (brother’s daughter). (m) and
(f) refer to the gender of the ego or anchor, e.g., ZH (f) refers to a woman’s sister’s
husband.
29
Both bap and baloodadi refer to the same person, as do yei and balooyi. The latter forms
are used primarily in Aspar, in contrast with the other dialects.
30
Used mainly in Aspar.
31
Used mainly in Aspar.
67
Nouns
žu
ṣpaṣi
nawaasa
nawaasi
kaṛwaasa
kaṛwaasi
pašur
ǰeṣṭaali
preeš
pʰaapa
mamaani
žami
ištrii
aštrakaa
bareu
mač
ẓami
saaraani
deer
ǰeṣṭa deer
mažuma deer
suruu deer
žamili
bražei
saaraana
bawi
zaamaa
68
daughter
nephew/niece
grandson
granddaughter
great-grandson
great-granddaughter
Affinal kin
father-in-law
mother-in-law
mother-in-law
uncle (by marriage)
aunt (by marriage)
wife, woman
wife
wives, women
husband
husband, man
brother-in-law
sister-in-law
brother-in-law
brother-in-law
(elder)
brother-in-law
(middle)
brother-in-law
(younger)
sister-in-law
sister-in-law
brother-in-law
son-in-law
son-in-law
D, (BD)
ZS, ZD
SS, DS
DD, SD
SSS, DSS, etc.
DDD, SSD, etc.
WF, HF, (WMB)
WM, (WMZ)
HM
FZH
MBW
W
W
W
H
H
WB, ZH (m)
WZ
HB
HoB
HmB
HyB
HZ
BW
ZH (f), WZH
SW, (BSW, ZSW)
DH, (BDH, ZDH)
Kinship nouns
abeeni
dram
paai
brei
kuẉa
zaatak
Special kinship relations
co-wife
HW
friend,
bloodbrother 32
boy
S
girl
D
child
S, D
child
S, D
Kinship terms can be further specified with certain adjectives, notably
abeeni for relationships through a co-wife or a father’s second wife, čoṣṭi
for an only child, and saka for the prototypical or closest kind of relative
when a term is also used in extended senses.
7.5.2
Kinship person suffixes
Each kinship term exists in three different forms: an unmarked form, a
second person form marked by the suffix −un/−in, 33 and a third person
form marked with the suffix −es, both placed directly after the root. The
choice between these forms is determined by whose relative is being
referred to, that is, who the anchor (Dahl & Koptjevskaja Tamm, 2001,
p. 201) or (sometimes) the ego is, that makes the term relevant. The
unmarked form, e.g., žu ‘daughter’, is used generally. The second person
form, e.g., žun, is used when referring to relatives of the second person:
‘your daughter’. The third person form, e.g., žuwes ‘his/her daughter’, is
used when referring to relatives of a third person, i.e., to someone else’s
relatives.
Apart from some slight and unpredictable changes in the vowel, these
forms are mostly regular. The one exception is the word for father, dadi,
which has a suppletive root de in the second and third person forms: deen
‘your father’, dees ‘his/her father’.
Kinship terms with different anchors of this kind are not unique to
Dameli. Several of the neighbouring languages, but by no means all, have
them. In Kalasha, the system is very similar, but the forms are completely
different (Trail & Cooper, 1999, pp. 473-474; Bashir, 1988, pp. 43-45).
32
This refers to a special relationship that is ritually established by sharing the kidneys
of a goat slaughtered for the purpose. It is doubtful if the practice is still in active use.
33
I have found no system in the alternation between these two suffixes.
69
Nouns
In Gawarbati, even the forms appear to be similar 34 (Grierson, 1919, p.
82). A similar system with obligatory possessive prefixes on many
kinship terms and body parts exists in Burushaski (Willson, 1999, p. 7).
The suffixes marking these forms must probably be considered
derivational rather than inflectional, especially since the resulting forms
can be freely combined with inflectional suffixes such as the plural.
7.5.3
Vocative
7.5.4
Special plural forms
A special vocative suffix, −oo, is only applied to kinship terms. Like the
vocative particle (21.2), this form is used when addressing someone
directly.
This form is, by definition, only used about relatives of the speaker,
but it should not be considered the first person form, in analogy with the
kinship person suffixes, as this would imply a parallellism that isn’t
there, or one that is at least incomplete. These forms are only used while
speaking to one’s own relatives. A prototypical first person anchor form
would be one used while speaking to another person and referring to
one’s own relatives, and no special form for this exists in the language.
Interestingly, there is one instance of xudaayoo ‘my God’ in my
material, within the context of a nat, a religious song. It is unclear if this
is representative of wider usage or poetic licence on the part of the
author, but it could be an indication that these forms are not entirely
restricted to kinship terms.
Kinship terms utilise a special plural suffix, −suu. The regular plural
suffix, −nam, can still be used with kinship terms and probably conveys
slightly different semantics, but the special kinship plural is more
common. Like the regular plural, the suffix is invariant and attaches to
any word within the group. Example 16 illustrates the use of the special
kinship plural.
34
At least the third person form is attested in the LSI. Grierson calls these forms “enclitic
pronouns” and does not connect them especially to kinship terms, but all his examples
are kinship terms.
70
Kinship nouns
(16) tasãã−i
traa žu−suu
bai−tʰun
3SG.POSS−F three daughter−KIN.PL be−INDIRPST.3PL
‘He had three daughters.’ (T9002)
For the kinship terms braa ‘brother’, pas ‘sister’, and ištrii ‘wife’, there are
also irregular plural forms: braadi ‘brothers’, pasari ‘sisters’, and aštrakaa
‘wives’.
71
8 Pronouns
Pronouns are a rather heterogenous group in Dameli. The first and
second person personal pronouns are “true” personal pronouns whose
only function is to refer to participants in a speech situation. The third
person forms are used as both personal pronouns and demonstrative
pronouns. The same forms are used, with the same distinctions, for both
functions, and also in a way that closely resembles definite articles.
Another group of pronouns are used both as indefinite and
interrogative pronouns. These forms also show formal and functional
similarities with interrogatives from other word classes (see 13.4).
There are no reflexive personal pronouns; intransitive verbs with
reflexive meanings do not take an overt object. For reflexive possession,
however, there is a specific marker.
8.1
Personal pronouns
The personal pronouns of Dameli form a system of five dimensions:
person, number, distance, animacy and case. Of these, distance and
animacy are only relevant for the third person forms, which are unusual
in other aspects as well (see 8.1.6). Table 18 shows the personal pronoun
paradigm.
Table 18: Personal pronouns
1SG
2SG
3SG.ANIM.PROX
3SG.ANIM.DIST
3SG.INANIM.PROX
3SG.INANIM.DIST
1PL
2PL
3PL.ANIM.PROX
72
Nominative
ay
tu
i
see
ay
bi
mẽẽ
Oblique
mas
tas
yee
see
muu
too
Ergative
amaa
mya
masuu
manii
tanii
Personal pronouns
3PL.ANIM.DIST
3PL.INANIM.PROX
3PL.INANIM.DIST
8.1.1
tẽẽ
mẽẽ
tẽẽ
tasuu
Person
The pronominal system distinguishes between the first, second and third
person, i.e., pronouns referring to, respectively, the speaker, the listener,
and anyone or anything else.
8.1.2
Number
Singular and plural number are distinguished for all persons, but in the
first person nominative, the singular and plural forms are identical. This
conflation of forms is quite superficial, since the number distinction is
maintained in subject marking on the verb.
8.1.3
Distance
In the third person pronouns, there is a distinction between proximal
and distal forms, i.e., forms referring to things nearby and things farther
away. With the exception of the nominative singular, the forms used to
express this distinction are all based on the opposition m−/t−; the forms
are identical except that proximal forms begin in m− and distal forms
begin in t−. This opposition is part of a larger system within the language,
see 8.1.7.
The distinction between proximal and distal forms is not restricted to
the spatial area alone but is used for distance in time as well. The
opposition between previously established referents and new
information, or between background and topic, may also play a role in
the choice of form.
8.1.4
Animacy
In the third person, a distinction is also drawn between living beings and
non-living things, between animate and inanimate forms (see 7.2).
The inanimate pronouns make no core case distinctions, and
possessive pronouns are rarely if ever used to denote inanimate owners.
73
Pronouns
8.1.5
Case
Following Blake (1994, p. 34), we will make a distinction between core
cases, those that encode subjects and direct objects, and peripheral cases,
which encode any other relation. Both the noun and the pronoun case
system contain both types of case forms.
There are three core cases, but the system governing their use is quite
complex. Due to the setup of the argument marking system, labels like
‘nominative’ and ‘absolutive’ are not very useful, but for reasons of
simplicity we will attempt to at least pick the terms for morphological
case forms from among those in common use. The reader must bear in
mind that these are used not as accurate descriptions of how the forms
are used, but as names for the individual forms that may call to mind
some of their functions and also that the rules governing when each form
is used are considerably more complex than the names themselves would
imply. To exemplify, the Dameli case form referred to here as nominative
is used in most of the contexts where nominatives would normally be
used, but also in some contexts where absolutives or accusatives would
be expected.
The three core case forms are the nominative, the ergative and the
oblique. The only non-core case is the locative.
8.1.5.1
Nominative
The term nominative will be used for the case form that encodes the
subjects of both transitive and intransitive verbs in the
non-past/imperfective TAM sets, where they are not distinguished from
each other. This closely corresponds to the prototypical use of the term.
The same form is also used to encode the subjects of intransitive verbs
in the past/perfective TAM sets, where they are distinguished from the
subjects of transitive verbs, a function usually called absolutive.
8.1.5.2
Ergative
The term ergative is used more or less in its prototypical sense, in order
to refer to the case form that encodes the subjects of transitive verbs in
past/perfective TAM sets, where these are distinguished from the
subjects of intransitive verbs. Most ergative pronoun forms have merged
with the oblique so that the two are no longer distinguished except in
the third person animate singular. See 8.1.5.4.
74
Personal pronouns
8.1.5.3
Oblique
The term oblique will be used for the case form that is used to encode
canonical objects anywhere in the system (usually encoded with forms
termed accusative and absolutive).
The oblique form of pronouns is prototypically used when pronouns
form phrases with adpositions. In Dameli, pronouns in this position are
either in the oblique or the locative, depending on the postposition.
The oblique functions more or less as the unmarked form of the
personal pronouns and is the form most commonly used when pronouns
fill any other function than subject. The possessive pronouns of the third
person are formed on the basis of the oblique, as are all locatives. The
first and second person possessive pronouns bear close resemblance to
the corresponding oblique forms, rather than to the nominative forms.
Obliques are also used when pronouns are used as articles specifying
possessors, or as indirect objects.
8.1.5.4
Conflation of the ergative and the oblique
A factor further complicating the system is that the distinction between
the ergative and the oblique is not maintained in all parts of the system.
It seems likely that a separate ergative and oblique form existed at some
earlier point, but that phonological developments caused the two to
merge. Regardless of the reason, synchronically the system maintains the
distinction between ergative and oblique only in the animate singular of
the third person. In the animate plural of the third person, and both the
singular and the plural of the first and second person, the ergative and
the oblique forms are identical. The inanimate third person makes no
case distinctions at all.
The current subsystem, which conflates A and O but distinguishes S
separately, is rather unexpected. The one key function of core casemarking is to clarify which argument is the subject and which is the
object in a transitive clause. Confusion between S and A, or between S
and O, is virtually impossible since they never occur in the same clause,
but being unsure of which argument is A and which is O can lead the
listener to completely misunderstand the clause.
Systems that conflate A and O are sometimes known as doubleoblique systems, since it is often the oblique case form that is extended
to both functions. Naturally, such systems are uncommon, but as the
Dameli case proves, they do exist, although probably never as the only,
or even the dominant, principle of any given language. Payne (1980)
describes a similar system in the Pamir languages, and Haig (2008) notes
75
Pronouns
that “a change which leaves A and O morphologically indistinguishable
is difficult to account for”, but that there is “extremely widespread
attestation of precisely such a change throughout Iranian” (p. 227).
Dixon, remarking on his description, notes that “surely marking A
and O in the same way (differently from S) must be an unstable and
temporary situation, only encountered as a language moves from one
more stable kind of marking to another” (Dixon, 1994, p. 39). There
are no immediate signs of this structure being on the decline in Dameli,
but this is not to say that this may not be the case, simply that I have
found no signs of it being so.
We have precious little diachronic data for Dameli, so it is hard to
make any assumptions about when this system first appeared. The
system seems to have been firmly in place when Georg Morgenstierne
collected his data for his 1942 article in 1929. Morgenstierne does not
discuss it explicitly, but although he separates “agentive” (ergative) from
“oblique” in his paradigm of pronoun case forms, all the “oblique”
forms, except in the third person singular, are simply noted with a ditto
mark (”) to indicate that they are identical to the “agentive” form. The
examples he gives also confirm this interpretation (Morgenstierne, 1942,
pp. 134-135).
At the very least, then, this system has been stable enough to last the
80 or so years that passed between Morgenstierne’s research and my
own.
This is not altogether surprising, although the argument marking of
the pronouns does not distinguish between A and O, and could therefore
be argued to be both meaningless and inadequate. Through the means
of a consistent SOV word order, and through subject marking on the
verb, the subject is consistently marked, whether the verb is intransitive
or transitive, just not by means of case-marking. While the case-marking
system fails to carry out the function of argument marking in this case,
that function is still carried out elsewhere.
8.1.5.5
Locative
The locative is a relatively marginal pronoun case form. It is created by
the addition of a secondary suffix, −a, to the oblique form. The locative
is used to express location, but literal, spatial meanings are probably in
the minority on pronouns, the most common uses being quite figurative.
The locative is also, sometimes, used with the postpositions ki ‘to’, ṣaži
‘in order to’, ṣawaai ‘for’, and mili, ‘with (comitative)’, as illustrated in
example 17.
76
Personal pronouns
(17) ta
tas−a
ki ṭelefun
kur−e
…
TOPSM 3SG.ANIM.DIST−LOC to telephone do−IMP.SG …
‘Call him…’ (D1005)
For phonological reasons, some pronoun forms do not always get a
specific locative; forms ending in vowels do not usually get the suffix,
although there appear to be exceptions, such as example 18.
(18) muu
yee kram ku−tʰim
1SG.ERG this work do−POTPST.1SG
xu muu−a
kaaya
ni aaċe−na
but 1SG.OBL−LOC remembrance not come−IMPFV.3SG.M
‘I may have done this work, but I don’t remember. [Lit.
...but rememberance comes not at me.]’ (E0001)
8.1.6
Third person pronouns as determiners
Apart from filling the prototypical functions of personal and
demonstrative pronouns, i.e., functioning as arguments of the verb in a
clause, forming phrases with postpositions, and performing other
functions that can also be filled by nouns, the third person pronouns are
also used together with nouns, as determiners or articles (see 15.1). Using
pronouns with a noun usually indicates that the modified noun has been
mentioned before or is otherwise previously known, similar to definite
articles, and the choice of form also provides some deictic information
on the nature of the noun, by stating whether it is animate or inanimate,
singular or plural, and proximal or distal. The case form of the
demonstrative gives some information about the syntactic function of
the noun. Examples 19 and 20 show third person pronouns used as
determiners.
(19) tu
mas
paai
2SG.NOM 3SG.ANIM.PROX.OBL boy
‘Why did you kill this boy?’ (T4032)
ku
why
žan−op
kill−PFV.2SG
In example 19, the demonstrative is in the oblique case, since the noun
it modifies, paai ‘boy’, functions as a direct object in the clause. In this
position, case marking on pronouns is more detailed than it is on nouns.
77
Pronouns
(20) tẽẽ
kul
naɡi prat−ee
3PL.INANIM.DIST house fall give−PFV.3SG
‘Those houses fell down.’ (E0101)
In example 20, the form of the demonstrative gives no indication of the
function, since it modifies an inanimate noun, and inanimate third
person pronouns do not distinguish case, although the fact that the
houses are distant in some way is still indicated.
8.1.7
Relationship to other deictic words
The distinction between proximal and distal forms of the pronouns is
part of a larger system that also involves question words, and that
reappears in several other word classes.
Proximal forms begin in m− and distal forms in t−. Question words
(13.4) always begin in k−. The typical cases are those in which all three
forms exist, and where the first phoneme is the only difference between
the words, but the m−, t− and k− also appear where differences between
the forms are more extensive and not all groups have all three words.
Table 19 shows some examples.
Table 19: Deictic distinctions in different word classes.
Proximal
Distal
mas ‘3SG.OBL.PROX’
tas ‘3SG.OBL.DIST’
masãã
‘3SG.PROX.POSS’
manuu ‘thus, in this
way’, ‘this kind’.
manee ‘thus, in this
way’
ayaa ‘here’
matiki ‘so many’
tasãã
‘3SG.DIST.POSS’
tanuu ‘in that
way’ / ‘that kind’
tanee ‘in that
way’
tara ‘there’
tatiki ‘that many’
Question-word
etc.
kii/kurey ‘who?’
keeraa ‘which?’
kya ‘what’
kasãã ‘whose?’
kanuu ‘how?’
kutaal ‘where
to?’
kaa ‘where?’
kati ‘how many’
katiki ‘so many’
Category
pronoun
possessive
pronoun
manner adverb /
adjective
causal adverb /
manner adverb
spatial adverb
quantifying
adjective
Perhaps also relevant is the fact that the first person singular pronouns
begin in m−, while the second person singular pronouns begin in t−.
78
Possessive pronouns
8.2
Possessive pronouns
The possessive pronouns are unique in that they carry double marking,
expressing the person, number and distance of the owner, as well as the
gender of that which is owned. There are no possessive pronouns for the
inanimate third person. Possession-like relationships with inanimate
“owners” can be expressed with the postposition ta ‘of’ (see 12.2.1). The
gender of the possessed is marked by −∅ or −a for masculine, −i for
feminine. Table 20 shows the paradigm for possessive pronouns, and
example 21 illustrates their use.
Table 20: Possessive pronouns
Masculine owned
1SG
2SG
3SG.ANIM.PROX
3SG.ANIM.DIST
1PL
2PL
3PL.ANIM.PROX
3PL.ANIM.DIST
(21) ãã
and
mãã
tãã
masãã
tasãã
amuna
mina
masuna
tasuna
Feminine owned
mããĩ
tããĩ
masããĩ
tasããĩ
amuni
mini
masuni
tasuni
mulaa baait−aai−i
mullah summon−CAUS−CP
tasun−a
nikaa
ɡrẽẽṭ−een
3PL.ANIM.DIST.POSS.PL−M
nika (M) tie−PFV.3PL
‘And having summoned the mullah, they tied their wedding
vows [nika].’ (T1047)
Possessive pronouns are also used with the verb pr (t) ‘to give’ when it is
used in the specialised sense ‘to hit’. The direct object, i.e., the person
being hit, is expressed with a possessive pronoun, as in example 22.
(22) tanii
tasãã−Ø
kaṭeri−yee pra−tʰee
3SG.ANIM.DIST.ERG 3SG.POSS−M knife−INS give−INDIRPST.3SG
‘He hit him with a knife.’ 35 (E0016)
35
Note that the intended meaning is not “He hit with his knife.”
79
Pronouns
8.2.1
The possessive marker sãã
The animate possessive marker sãã is something of a category of its own
but shows paradigmatic agreement that is analogous to the possessive
pronouns. It also appears to function as a separate word rather than a
suffix, which is the motivation for placing it here rather than elsewhere.
These forms are used in attributes to express belonging and similar
relationships when the owner is an animate entity. Due to its function,
it is limited to referring to the third person animate, which limits
agreement to number and gender. The number of the possessor is shown
by the choice between sãã/sããĩ (singular) and suna/suni (plural). The
presence of a final −i expresses feminine gender in that which is owned,
just like with the possessive pronouns. Table 21 shows the paradigm of
the animate possessive marker.
Table 21: The animate possessive marker
Masculine
possessed
SG
PL
8.2.2
sãã
suna
Feminine
possessed
sããĩ
suni
The reflexive possessive pronoun taanu
A special pronoun, taanu ‘own’, is used to indicate that something is
owned by a person already in focus, such as the subject of a clause.
Unlike the personal and possessive pronouns, the reflexive possessive
pronoun does not show person or number agreement but can be used
with any person. In example 23, taanu refers to the first person ‘my own
country’.
(23) tee taanu
mulk−a
ki aaɡ−em
that POSS.REFL country−OBL to come.PFV−PFV.1SG
‘that I am going back to my own country. (TW2007)
The only inflection is gender agreement with the possessed, as shown in
example 24, where the form taanui indicates agreement with the feminine
talwaar ‘sword’ and the form taanu agrees with the masculine kul ‘house’.
80
Possessive pronouns
(24) ɡan−i taanu−i
talwaar
ɡ−i
say−CP REFL.POSS−F sword (F) take.PFV−CP
taanu
REFL.POSS
kul
ta
rawan b−aa
house from start become−PFV.3SG.M
‘Having said this, he took his sword and started out from his
house. (TW4006)
In some set expressions with taanu, the possessive element is more
optional, and the marker is used in the sense ‘self’ to express reflexivity.
This applies to taanu milaai ‘with oneself’ (example 25), tan taanu ‘each’
(example 26), and taanu ki ‘for oneself’ (example 27). The latter two can
be combined into tan taanu ki ‘by themselves’ (example 28).
(25) žan−an ta
bææt tas
učʰuṭa paai taanu
milaai
kill−INF from after 3SG.ANIM.DIST small boy REFL.POSS with
taanu
kul
ɡiɡ−ee
REFL.POSS house take.PFV−PFV.3SG
‘After the killing, she took that small boy with her to her own house.’
(26) traa−i
sãã−Ø
žanii
b−i
three−COLL 3SG.ANIM.POSS−M marriage become−CP
tan taanu
ware baadša *
each REFL.POSS other king
*
tasuu
žanii
kur−een
3PL.ANIM.DIST.ERG marriage do−PFV.3PL
‘The wedding of the three took place and they each
married a king.’ (T9006)
ãã
(27) lee pʰikir−a
very thought−LOC and
lee
very
soč−a
thought−LOC
xux−aai−tʰum
too
taanu
ki
like−CAUS−INDIRPST.1SG 2SG.OBL REFL.POSS for
‘In much thought and consideration, I have liked you for
myself.’ (TV3005)
81
Pronouns
(28) tukuri−nam bum
ṣaa tan taanu
ki daro
basket−PL ground on each REFL.POSS for be.INANIM.IMPFV.3
‘The baskets are by themselves on the ground.’ (Q1044)
8.3
Indefinite and interrogative pronouns
A group of pronouns are used in questions and when referring to an
unknown or undefined entity. The same words are also used as relative
pronouns. These interrogative/indefinite pronouns are kya ‘what’,
kii/kuree ‘who’, keeraa, ‘which one’, and the possessive kasãã ‘whose’. Like
the wh-words of English, all these pronouns begin with the same
phoneme, /k/, and are part of a larger group of question words, which
also includes adverbs and adjectives. See 8.1.7 and 13.4.
82
9 Adjectives
There are two groups of adjectives in Dameli: one that shows gender
agreement with the noun it describes, and one that is completely
invariable. This division is characteristic of adjectives in Indo-Aryan
languages (Masica, 1991, p. 250).
Adjectives show no morphological comparison and no inflection for
categories other than gender. One possible exception was noted by
Cacopardo (2008), who calls the form jest’ero: “a comparative form of
the […] word [j’esta or z’esta ‘elder’]” (p. 404). The word he is referring
to, žeṣṭera in my notation, can indeed function as a comparative
semantically, but there are no indications of a wider pattern of
comparison of adjectives. 36
9.1
Variable adjectives
The inflected, variable group expresses many of the properties that
adjectives usually express, such as basic colour terms, quantities,
large/small, etc. They agree in gender with the word they describe but
show no agreement according to case or number.
Example 29 shows the variable adjective šumaa ‘beautiful’ agreeing in
gender with the place-name daaman ‘Domel’ (or possibly with the head
of the phrase, ɡurum ‘morning’).
(29) muu
1SG.OBL
ta
from
kaaya
remembrance
aaċ−i
daro
come−CP is
mãã−i
šumaa−i
daaman
ta
ɡurum
1SG.POSS−F beautiful−F Domel(F)
of
morning
‘I remember the morning of my beautiful Domel.’ (TV8003)
Table 22 shows some variable adjectives in their masculine and feminine
forms.
36
The corresponding noun, žeṣṭa ‘elder’ is mostly used in the sense of ’village elder’.
83
Adjectives
Table 22: Some variable adjectives
Masculine
krnaa/krinaa
ɡoora
laic̣ʰaa
aarida
niila
čitra
baloo
učʰuṭa
leeṇḍa
kʰušala
čukra
šumaa
suree
awaluka
Feminine
krinii
ɡoori
laac̣ʰi
aaridi
niili
čitri
baloi
učʰuṭi
leeṇḍa
kʰušali
čukri
šumaai
surui
awaluki
Gloss
black
white
red
yellow
green
multicoloured
large
small
bald
intelligent
sour
beautiful
young
old
The most likely source of new variable adjectives is participles,
particularly the past participle. Since these forms may show gender
agreement and express similar meanings as adjectives, they may easily
be reinterpreted as variable adjectives.
9.2
Invariable adjectives
The group of invariable adjectives is probably larger than the variable
group, and is more likely to receive new members through borrowing.
Many borrowed words appear indifferent to gender agreement.
Several South Asian languages derive adjectives by adding the suffix
−i to words of other word classes, e.g., Peshawari, Pakistani. Several such
words have been borrowed into Dameli. Despite the fact that the ending
is similar to the Dameli feminine suffix, these words are treated as
invariable adjectives, as can be seen in example 30.
(30) yee mãã−Ø
axiri faisala
daru
this 3SG.POSS−M final decision (M) COP.INANIM.IMPFV.3
‘This is my final decision.’
It is often difficult to draw a clear line between the invariable group and
other word classes, since there are several cases where other word classes
84
Adjective derivation suffixes
are used in similar syntactic positions as adjectives. Numerals in all
forms, nouns in the locative case, and participles can all be used to
modify nouns; the latter may also display gender agreement.
Likewise, the general decline of the category of gender in Dameli
(7.3.3) makes it difficult to distinguish invariable adjectives from
variable ones. Still, there are some reasonably clear examples of
invariable adjectives, some of which are shown in Table 23.
Table 23: Some invariable adjectives
Word
ɡulabi
axiri
zrax
kʰošan
čoṣṭi
9.3
Gloss
pink
final
clever
happy
without siblings
Adjective derivation suffixes
There are at least four suffixes that are used to derive adjectives from
words of different word classes. These are not general conversion
mechanisms that take any word from one word class and turn it into a
corresponding one of another word class, but rather morphemes with
quite specific meanings that apply a set pattern to the original root,
creating a group of words with limited and predictable meanings.
9.3.1
−bana/−bani ‘−ish’
The suffix −bana/−bani is combined with several kinds of words, to create
adjectives that denote being similar to what that word describes. The
resulting adjectives are variable, with −bana as the masculine form and
−bani as the feminine form. With colour adjectives, the suffix is used to
describe being a peripheral, rather than prototypical example of that
colour, much like the English ‘−ish’ in words such as ‘reddish’, ‘greyish’,
e.g., aaridabana ‘yellowish’ <aarida ‘yellow’, laic̣ʰabana ‘reddish’ <laic̣ʰaa
‘red’, niilabana ‘greenish’ <niila ‘green’. Words of other classes, including
pronouns, can also be combined with −bana/−bani, as shown in table 24.
85
Adjectives
Table 24: Adjective derivation with −bana
Word
muubana ‘like me’
toobana ‘like you’
masbana ‘like him/her/it’ (PROX)
tasbana ‘like him/her/it’ (DIST)
kaasbana ‘like whom’
akabana ‘alike’
trakabana ‘like a footpath’
9.3.2
Source
Source meaning
muu ‘me’ (1SG.OBL)
too ‘you’ (2SG.OBL)
mas ‘him/her/it’ (3SG.ANIM.PROX)
tas ‘him/her/it’ (3SG.ANIM.DIST)
kaa(s?) ‘who’
ek ‘one’
trak ‘footpath’ (probably <Eng.
trek/track)
−baṣ ‘able to x’
The present, past and conjunctive participles of verbs can all be used in
a way that is very similar to adjectives, but there is also a particular
derivation suffix, −baṣ, that is applied to the infinitive form of verbs to
create words that mean ‘able to verb’, where verb represents the verb
root that is being used. The resulting adjective can be understood either
actively, as “able to do x”, or passively “possible to do x to”, compare
ɡewanbaṣ ‘able to walk’ with matrambaṣ ‘legible’ (‘possible to read’). Table
25 shows some examples of such adjectives.
Table 25: Adjective derivation with −baṣ
Word
kuranbaṣa
matrambaṣ
ničinanbaṣ
žananbaṣ
ɡewanbaṣ
Meaning
able to do
legible
unbreakable
able to kill (/be killed?)
able to walk
Source
kur
matr
ni−čin
žan
ɡew
Source meaning
do
read
not−cut
kill
walk
In function, these words closely resemble the abilitative construction in
Kalasha (Bashir, 1988, p. 59) and Khowar (p. 122). In form, however,
they are curiously similar to another construction, the necessitative
construction, which is marked by the suffix −baṣ in Kalasha (Bashir,
1988, p. 59) and by the independent word baṣ in Khowar (p. 122). 37
37
Trail & Cooper (1999) interpret the Kalasha form as independent as well (p. 486).
86
Adjective derivation suffixes
9.3.3
−weela ‘having x’
9.3.4
−pin ‘full of x’
The suffix −weela is applied to nouns and used in the sense ‘with x’ or
‘having x’. They can be used to express ownership when described as a
characteristic of a person (e.g, ɡẉaweela ‘having fields’, ɡaḍiweela ‘having
a car’). A few words created in this way are used to describe weather
(e.g., hawaweela ‘windy’, kiirweela ‘snowy’). There are some examples of
idiomatic expressions with an extended noun phrase that are turned into
adjectives. In these cases, −weela is added to the head, but the meaning is
expressed by the entire phrase (e.g., baloo dašweela ‘generous’, lit. ‘having
big hands’; baloo zaadiweela ‘brave’, lit. ‘having a big heart’.) There is also
a feminine form, −weeli, but the use of this is inconsistent because of the
decline of gender in the language (see 7.3.3). 38
A small number of adjectives are formed by adding the suffix −pin to
nouns and are used to describe the sense of being full of something (e.g.,
skaapin ‘fattened, of animals’) or being soaked with something (e.g.,
loypin ‘bloody’ or xaṭaapin ‘muddy’.) These adjectives are invariable; they
do not show gender agreement.
For ‘a woman who has a car’, both ɡaaḍiweela žami and ɡaaḍiweeli žami were considered
acceptable, for example.
38
87
10
Numerals
Hammarström (2010) defines numerals as “spoken, normed expressions
that are used to denote the exact number of objects for an open class of
objects in an open class of social situations with the whole speech
community in question” (p. 11). Numerals in Dameli are quite typical in
this sense, and satisfy all the conditions of this definition.
Numerals appear in three different forms: cardinal numerals, ordinal
numerals and collective numerals. The cardinal numerals use the bare
root form, and the other two are formed by adding suffixes (-m for
ordinals and –i for collectives).
10.1
Numeral system
The numeral system of Dameli would traditionally be described as
vigesimal, since twenty is the most prominent base.
In the current stage of the language, the numbers between 1 and 20
are monomorphemic, but it is apparent that at least 13-19 originated as
compositions with the base 10. All the latter end in long vowels + −š,
probably from daš ‘ten’ and the first part is similar though not identical
to the corresponding lower number, e.g., pančees 39 ‘fifteen’ < pãc ‘five’ +
daš ‘ten’. 11 and 12 appear entirely unrelated to 1 and 2.
Numbers higher than 20 are formed with 20 as the base, on the
pattern multiplier−base−AND−number, where multiplier and number are
numbers below 20 (e.g., duu−biši−o−aṣtaaš ‘58, 2−20−and−18’). The
numeral 100 can be expressed as either pããč−biši ‘5−20’ or with a special
morpheme (sawa ‘hundred’, a Pashto loan). For higher numbers, the
latter strategy is more common. The borrowed zara (also from Pashto) is
used to express “thousand”. Table 26 provides a list of numerals.
Table 26: Selected numerals
No.
39
Analysis
Numeral
No.
In this case, *panč is probably the older form.
88
Analysis
Numeral
Numeral system
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
1
2
3
4
5
6
7
8
9
10
1+10
2+10
3+10
4+10
5+10
6+10
7+10
8+10
9+10
20
ek
duu
traa
čoor
pããč
ṣoo
sat
aṣ
noo
daš
yaaš
baaš
trooš
čandeeš
pančeeš
ṣoyeeš
santaaš
aṣṭaaš
uneeš
biši
21
22
23
30
31
32
33
40
41
42
43
50
51
52
53
60
80
100
500
1000
20+1
20+2
20+3
20+10
20+1+10
20+2+10
20+3+10
2×20
2×20+1
2×20+2
2×20+3
2×20+10
2×20+1+10
2×20+2+10
2×20+3+10
3×20
4×20
5×20, 100
5×100
1000
bišiyoek
bišiyoduu
bišiyotraa
bišiyodaš
bišiyoyooš
bišiyobaaš
bišiyotrooš
duubiši
duubišiyoek
duubišiyoduu
duubišiyotraa
duubišiyodaš
duubišiyoyooš
duubišiyobaaš
duubišiyotrooš
traabiši
čoorbiši
pããčbiši, sawa
pããč sawa
zara
The notion of base is subject to several conflicting definitions, which may
lead to some confusion.
Hammarström (2010) works with a definition that assigns the status
of a base to several numbers in a numeral system:
The number n is a base iff
1. The next higher base (or the end of the normed expressions) is a
multiple of n; and
2. A proper majority of the expressions for numbers between n and the
next higher base are formed by (a single) addition or subtraction of n
or a multiple of n with expressions for numbers smaller than n.
(Hammarström, 2010, p. 15)
By this definiton 20, 100, and probably also 1000 function as bases in
Dameli, and perhaps also 10.
Comrie (2005) works with the following definition of a base:
By the “base” of a numeral system we mean the value n such that numeral
expressions are constructed according to the pattern ... xn + y, i.e. some
89
Numerals
numeral x multiplied by the base plus some other numeral. (Comrie,
2005, p. 530)
Implicitly, his use of the term assumes a single base as the norm. By his
reckoning, Dameli would be a hybrid vigesimal-decimal system, since
“the system then shifts to being decimal for the expression of the
hundreds” (Comrie, p. 530). One might argue that such a system is not
really decimal except by extension, 100 is not 10, but an exponent of 10,
2
10 , and perhaps a more proper term would be vigesimal-cential.
When applied to Dameli, Comrie’s definition fails to acknowledge the
role of 10 in the creation of the numerals from 11 to 20 but captures a
higher generalisation in giving one base a more prominent position. If
Dameli had been a pure vigesimal system, by Comrie’s definition, it
2
should have assigned no special role to 100, but to 400, as 20 .
Interestingly, Morgenstierne (1942) reports the word a'zâr (>Prs. hazaar
‘thousand’) as ‘four hundred’ (p. 137). None of my Dameli co-workers
recognised this use, although the form is familiar from the Urdu word
for ‘thousand’.
The fact that the current language is structured on 100 and 1000 in
the higher numerals may be a sign that the system is being restructured
under influence from base-10 languages such as Urdu and Pashto,
particularly since the terms for 100 and 1000 seem to have been
borrowed from Pashto.
A number of languages in the area have special rules for numbers
immediately below a multiple of ten or twenty (depending on what bases
are used in the language). These are expressed as base-1 (e.g., Pashto:
yo−kam−derš ‘twenty-nine’ or ‘one−less−thirty’, Urdu: ontiːs ‘twenty-nine’,
based on the ending for 30, tiːs). Urdu and Pashto do this consistently
throughout the system. Other neighbouring languages, such as Palula
(Liljegren, 2008), and Indus Kohistani (Zoller, 2005), do not utilise this
strategy throughout the system, but still have a word for ‘nineteen’ which
is clearly derived from 20 rather than from 10+9. Dameli, like Kalasha
(Trail & Cooper, 1999), shows no traces of such a strategy at all.
10.2
Cardinal numerals
The cardinal numerals are expressed with the bare numeral stem, i.e.,
the morphologically unmarked form of the numerals shown in table 26
above, and are used in counting and when modifying nouns in terms of
countable quantity. Example 31 shows a cardinal numeral in use. Note
90
Ordinal numerals
that nouns modified by numerals above one do not necessarily take
plural forms.
(31) swat yede tara bišiy−o−pããč
kaal kati
Swat go.CP there twenty−and−five year how_much
mudaa
tara ba−i
waapas aaċ−i
some.time there be−CP back
come−CP
‘Having gone to Swat, after staying there for some twenty-five
years, they came back.’ (T3022)
Using a remarkably exhaustive collection of references to numeral
systems in the world’s languages, Hammarström (2010, pp. 26-27)
shows that cardinal numerals have primacy over other types of numerals,
in that other kinds of numerals are morphosyntactically derived from
the cardinal numerals, and that cardinal numerals run higher than other
kinds. Both these points are borne out by Dameli; ordinal and collective
numerals are formed by suffixation on the cardinal numerals, and
although any cardinal numeral can potentially be thus transformed into
an ordinal or a collective numeral, attested, non-elicited cases of these
forms are always of numbers one to five.
Cardinal numerals are also used to express clock time. A form of the
word kʰana ‘part, half’ is used for half hours, and the word baaǰa ‘hour,
o’clock’ is used as a sort of clock time marker, as in example 32.
(32) traa−o
kʰani baaǰa
daro
three−and half o’clock is
‘It’s half past three.’ (E0048)
10.3
Ordinal numerals
The ordinal numerals, comparable to ‘third’ or ‘fourteenth’ in English,
express the position of something in a sequence, and are formed by
adding the suffix −m to the root, and sometimes adding an epenthetic
vowel, as in example 33 below. This formation is productive, and can
be used with any numeral root, although most attested cases in my
material are with numbers below five.
91
Numerals
(33) čoor−am ɡram
ṣiṇṭeri
kuruu daro
four−ORD village Shinteri Kuru is
‘The fourth village is Shinteri Kuru’ (TA1033)
The ordinal numeral for ‘one’, i.e., awal ‘first’, is not formed in this way,
but borrowed from a word derived from Arabic that is found in both
Urdu and Pashto. Suppletive ‘first’-words appear to be the norm in the
world’s languages: In the sample used for the World Atlas of Language
Structure map on ordinal numerals (Stolz & Veselinova, 2005), 237 of
the 321 languages use a word for ‘first’ that differs from the general
strategy of creating ordinal numbers.
10.4
Collective numerals
The collective numerals express groups of a certain number of members,
similar to the English word “both”. Unlike English, however, the
collective numerals are a productive formation in Dameli. They are
formed by adding the suffix −i to any numeral root (except, by definition,
one), again most commonly with the lower numbers (only two and three
are attested in my material, although speakers claim that higher numbers
are possible). The collective numerals can be used to modify nouns, as
in example 34.
(34) tẽẽ
traa−i
žuw−es−suu
lee
3PL.ANIM.DIST three−COLL daughter−KIN.3−KIN.PL very
minaa ta
saati−tʰee
love
TOPSM keep−INDIRPST.3SG
‘These three daughters he loved very much.’ (T9003)
Collective numerals can also be nominalised and used in phrases without
another noun, such as example 35.
(35) tẽẽ
traa−i
xu lee ɣaribanan maaṣ tʰun
3PL.ANIM.DIST.NOM three−COLL but very poor
people be.IMPFV.3PL
‘But those three are very poor people.’ (TW8011)
92
11
Verbs
The verb class in Dameli functions much as verbs are expected to do.
Members of this class describe actions and processes, they are used to
form predicates, and they can be inflected according to a set paradigm.
The core part of a verb is the verb root, a single morpheme that
contains the lexical meaning of the verb.
A complete verb is formed by combining a verb stem with an
inflectional suffix. There are many verbs where the verb stem consists of
just the root, but if the verb contains derivational suffixes, these are
considered part of the stem. The only verbs that contain derivational
suffixes are those that have undergone causativisation, a process that
increases the valency of verbs (see 11.2.1).
The inflectional affixes are grammatical morphemes that define the
verb. Figure 6 shows the structural template for verb forms.
Stem
Root ((Second causative suffix)
Inflection
Causative suffix) Inflectional suffix
Figure 6: Verb structural template
A verb form can contain a maximum of four morphemes; 40 there are four
slots in the template. The root and the inflectional suffix slots must
always be filled, as illustrated are in example 36-38.
In simple verbs such as that in example 36, only the compulsory slots
are filled.
Root 2nd causative Causative
(36) nees
−∅
−∅
exit
−∅
−∅
‘I went out.’ (simple verb)
Inflection
−uma
−PFV.1PL
40
Not including the interrogative suffix, since this appears to be on the level of the clause
rather than the verb itself, although it may attach to the verb when this is the last word
of the clause. Some verb roots, such as ɡirankʰat (i) ‘turn around’, may at least have
originated as compounds, which would introduce yet another morpheme.
93
Verbs
All causative verbs have a morpheme in the causative position, in
addition to the two compulsory slots. See examples 37 and 38.
root
(37)
niš
2nd causative
−∅
−∅
causative
−aai
−CAUS
inflection
−aba
−PFV.1PL
sit
‘You (pl.) are making someone sit down.’ (first causative verb)
The only verb forms to fill all slots are the second causative verbs
(11.2.1.2), which have a morpheme in the second slot, in addition to the
compulsory slots and the causative slot, as illustrated in example 38.
root
2nd causative
causative
inflection
(38) nikʰaa
−aw
−aai
−uma
exit
−CAUS2
−CAUS
−PFV.1PL
‘We made someone take something out.’ (second causative verb)
The inflectional suffixes are polysemous and encode both subject
marking (see 11.4.1) and TAM dimensions. TAM (Tense-Aspect-Mood)
is shorthand for several different semantic dimensions: in Dameli they
include tense, aspect, mood, evidentiality and epistemic modality (see
chapter 16).
Some inflectional suffixes mark verb forms that do not take subject
marking. These are referred to as non-finite (see 11.6), and include the
infinitive (11.6.2), the present participle (11.6.3), the past participle
(11.6.4), the conjunctive participle (11.6.5) and the inchoative participle
(11.6.6).
The finite verb forms comprise six different sets, containing inflection
forms for each person and number 41, and sometimes for gender. These
are used to form verb expressions, both alone and in more complex,
periphrastic constructions. The finite sets are the perfective (11.5.1), the
imperfective (11.5.2) the future (11.5.3), the indirect past (11.5.4), the
potential past (11.5.5) and the imperative (11.5.6).
Table 27 shows Dameli verbs, together demonstrating all the forms a
Dameli verb root can be used in 42.
41
The imperative is by definition second person only and only inflects according to
number.
42
There are 109 unique forms, but 130 positions in the table, as valency distinctions are
not always marked in the suffixes.
94
Collective numerals
Table 27: Verb overview
Form
Gloss
Simple verbs
Causative verbs
Intransitive
Transitive
First causative
play
count
bring down
Second causative
make someone
read/study
Imperfective
Root
muṣ (i) ‘play’
leekʰ (t) ‘count’
naɡ (i) ‘come
down’
matr (t) ‘read’
1SG
2SG
3SG.M
3SG.F
1PL
2PL
3PL
muṣinum
muṣinap
muṣina
muṣini
muṣinuma
muṣinaba
muṣinun
leekʰinum
leekʰinap
leekʰina
leekʰini
leekʰinuma
leekʰinaba
leekʰinun
naɡaayim
naɡaayap
naɡaaya
naɡaayi
naɡaayima
naɡaayiba
naɡaayin
matrawaayim
matrawaayap
matrawaaya
matrawaayi
matrawaayima
matrawaayiba
matrawaayin
Perfective
1SG
2SG
3SG.M
3SG.F
1PL
2PL
3PL
muṣúm
muṣóp
muṣaa
muṣúi
muṣúma
muṣóba
muṣún
leekʰúm
leekʰóp
naɡam
naɡap
matrawam
matrawap
leekʰee
naɡææ
matrawææ
leekʰúma
leekʰóba
leekʰén
naɡaama
naɡaaba
naɡæn
matrawaama
matrawaaba
matrawæn
1SG
2SG
3SG.M
3SG.F
1PL
2PL
3PL
muṣitʰum
muṣitʰop
muṣitʰaa
muṣitʰui
muṣitʰuma
muṣitʰoba
muṣitʰun
leekʰitʰum
leekʰitʰop
naɡaaitʰum
naɡaaitʰop
matrawaaitʰum
matrawaaitʰop
leekʰitʰee
naɡaaitʰee
matrawaaitʰee
leekʰitʰuma
leekʰitʰoba
leekʰitʰen
naɡaaitʰuma
naɡaaitʰoba
naɡaaitʰen
matrawaaitʰuma
matrawaaitʰoba
matrawaaitʰen
1SG
2SG
3SG
1PL
2PL
3PL
muṣitʰim
muṣitʰis
muṣitʰiyo
muṣitʰima
muṣitʰiba
muṣitʰin
leekʰitʰim
leekʰitʰis
leekʰitʰiyo
leekʰitʰima
leekʰitʰiba
leekʰitʰin
Indirect past
Potential past
naɡaaitʰim
naɡaaitʰis
naɡaaitʰiyo
naɡaaitʰima
naɡaaitʰiba
naɡaaitʰin
matrawaaitʰim
matrawaaitʰis
matrawaaitʰiyo
matrawaaitʰima
matrawaaitʰiba
matrawaaitʰin
95
Verbs
Simple verbs
Causative verbs
Person
Intransitive
Transitive
1SG
2SG
3SG
1PL
2PL
3PL
muṣim
muṣes
muṣiyo
muṣima
muṣiba
muṣin
leekʰim
leekʰes
leekʰo
leekʰima
leekʰiba
leekʰin
muṣe
muṣaa
leekʰe
leekʰaa
SG
PL
Future
First causative
Second causative
naɡayim
naɡæs
naɡao
naɡayima
naɡayiba
naɡayin
matrawayim
matrawæs
matrawao
matrawayima
matrawayiba
matrawayin
Imperative
naɡai
naɡaaya
matrawai
matraawaya
Non-finite verb forms
Form
Intransitive
Transitive
Infinitive
muṣan
muṣaal
muṣisan
muṣem
muṣi
leekʰan
leekʰaal
leekʰisan
leekʰem
leekʰi
Present participle
Past participle
Inchoative participle
Conjunctive participle
11.1
First
causative
naɡan
naɡaal
naɡaaisan
naɡæm
naɡaai
Second
causative
matran
matrawaal
matrawaaisan
matrawæn
matrawaai
The verb root
The lexical meaning of the verb is primarily contained in the verb root.
The typical verb root has one or two syllables, but there are a few verb
roots consisting of a single consonant, and one example in my data of a
verb root with three syllables (ɡirankʰat ‘turn around’ 43)
Verb roots normally have a basic valency that determines the number
of arguments they take; they can be intransitive, transitive or ditransitive.
In Dameli, the main morphological distinction is between intransitive
and transitive verbs, which have slightly different inflectional suffixes
(see table 33). Verbs with different valency are sorted into these two
groups with regards to inflection: zero valency verbs are treated as
intransitives, ditransitive and causative verbs like transitives (see also
section 14.1). This is not unexpected; inherent marking of verbs as
43
GM transcribes this as ɡiran khat.
96
The verb root
intransitive or transitive is the norm in NIA languages (Masica, 1991, p.
315).
The verb root in itself is not a complete word, and is never used
without inflectional suffixes. The basic valency of the verb cannot be
determined from the form of the root. For this reason, verb roots will be
marked with their basic valency when discussed in the text following this
pattern: girankʰat (i) ‘turn around’, bin (t) ‘see’, where (i) and (t) stand for
intransitive and transitive, respectively. Some verb roots are only used in
causative verbs (see 11.2.1). Causative verbs are normally verbs that
have undergone a valency-increasing operation, but some verbs follow
this pattern, but no corresponding simple verb exists. Such roots, for
which the basic valency cannot be determined, will be marked with (c):
bigir (c) ‘spread’.
In my material, 152 verb roots have been identified (excluding the
copula). See Appendix 1 for a complete list of these.
There are certainly verb roots I have not yet discovered, but even
taking this into account, the number of verb roots is limited in
comparison with, for instance, English. The usual strategy for expressing
new verb meanings is not to create new roots, but to form conjunct
verbs, combinations of a non-verbal element and a verb, that function as
a single word semantically (see 11.8).
11.1.1
Irregularities in the verb root
A number of common verbs have slightly different alternative roots that
are used when forming perfective forms; these are shown in table 28.
Example 39 is a text example of the same phenomenon. There are some
minor irregularities in other TAM sets as well, mostly changes in the
vowels, that may be phonologically conditioned.
97
Verbs
Table 28: Verbs with different perfective roots
Root
aaċ (i)
ɡi (t)
ɡiaaċ (t)
kook (i)
lai (t)
nat (i)
neeṭ (c)
niš (i)
ootʰ (i)
pr (t)
wali (i)
ẉi (t)
y (i)
ži (t)
Perfective root
aaɡ (i)
ɡiɡ (t)
ɡiaaɡ (t)
kookín (i)
lad (t)
nait (i)
nieṭ (c)
nišiin (i)
ootʰin (i)
prat (t)
wail (i)
ẉin (t)
ɡ (i)
žin (t)
(39) ay
1SG.NOM
ek
one
Meaning
arrive, reach
take, buy
bring (here)
sleep
find
go in
cut hair, shave, shear
sit
stop, remain
give
hide
free, allow, leave
go
eat
waxt−a
daaman
time−LOC Domel
ta
ay
from 1SG.NOM
dir aaɡ−em
Dir come.PFV−PFV.1SG
‘One time I came from Domel to Dir.’
dir
aaċi
…
Dir
come.CP
…
‘Having come to Dir…’ (TW7001-2)
This relatively limited form of suppletion, where the roots are similar
and clearly related to each other, is the most common form. The verb
‘to die’, in contrast, has a completely suppletive root. The root naṣṭ (i) is
used in the perfective, and the root br (i) in all other TAM sets. This
difference is highlighted in example 40 by the use of this verb in a tailhead construction (see 21.8), which is a a device used to keep sentences
together within a narrative, typically by repeating the finite verb of the
last sentence in a non-finite form at the beginning of the subsequent
sentence.
98
The verb root
(40) seepʰurrahman
Sephur Rahman
baadša hawa−ĩĩ
king
air−APP
žaas
ta
aeroplane from
see
naṣṭ−aa
3SG.ANIM.DIST.NOM die−PFV.3SG.M
br−en ta
baat
die.INF from after
paakistan široo
b−aa
Pakistan beginning become.PFV.3SG
‘King Sephur Rahman died in an aeroplane in the air. After
[he] died, Pakistan came into being.’ (T6056-57)
As this example illustrates, bren is used as the infinitive form of naṣṭaa in
the second sentence, within the same tail-head construction.
In other instances still, the root used in simple verbs is slightly
different from the one used in causatives (see 11.2.1), though clearly of
the same origin. In some cases, e.g., wali/wail (i), verbs have the same
variant roots in both causatives and the perfective of simple verbs.
Table 29: Verbs with different causative roots
Simple verb
root
ootʰ (i)
wali/wail (i)
ži/žin (t)
11.1.2
Meaning
stop
hide
eat
Causative
verb root
ooitʰ
wail
žuw
Meaning
stop someone
hide something, steal
feed
Vowel-strengthening
In Sanskrit, there was a process for deriving verbs with different
valencies called vowel-strengthening, a derivational strategy involving
lengthened or otherwise altered vowels (Masica, 1991, p. 319). Though
this process is not productive in Dameli, there are some possible vestiges
of such a process in the two verb pairs shown in table 30:
Table 30: Vowel-strengthening verb pairs
Root
tap (i)
lúṣ (i)
Meaning
become warm
burn, hurt
Root
taap (t)
looṣ (t)
Meaning
heat up
burn something
99
Verbs
11.2
The verb stem
The verb stem can be defined as the part of a verb that remains if all
inflection is removed. Many verb stems in Dameli are so called primary
stems (Masica, 1991, p. 315), in which the stem is identical to the verb
root. Verbs formed with primary stems are referred to as simple verbs.
Roots can also undergo a process called causativisation (see 11.2.1), in
which either one or two suffixes are added as part of the stem to increase
the valency of the verb. These stems are referred to as secondary stems
(Masica, 1991, p. 315).
11.2.1
Causativisation
The valency of a verb root can be increased through a process called
causativisation. By adding a causative suffix to the stem, another
argument is added to the verb’s argument structure. Intransitive verbs
become transitive, transitive verbs get a third argument. Verbs created
in this way are called first causatives. A secondary process adds another
suffix, in addition to the causative suffix, to create second causatives,
with yet another added argument (Masica, 1991, p. 319).
The inflectional affixes that express TAM form and subject agreement
are slightly different for causative verbs than for simple verbs. See table
33 and table 27.
Causativisation is not an inflectional process; causative verbs are
separate lexical entities and can have quite different meanings than
simple verbs with the same root. Some simple verbs have no causative
equivalents, and there is a considerable number of verbs that are
constructed as causatives, but where no simple verbs with the same root
have been found. This is probably not a new development; in Sanskrit,
many verbs that had causative stems did not have causative meanings
(Whitney, 1889).
Causative stems can be used with any inflectional form except the
present participle, and, with some exceptions, the infinitive.
11.2.1.1 First causatives
First causatives are created by adding the causative suffix to the stem.
The causative suffix has two different allomorphs, tied to different TAM
sets and non-finite verb forms. In the imperfective, the imperfective and
the potential past, and in the conjunctive and past participles, the form
is −aai.
100
The verb stem
Sanskrit had a similar causative suffix, −áya, which was also added to
the root in order to create a causative stem (Whitney, 1889, p. 378;
Masica, 1991, pp. 316-317). In Middle Indo-Aryan, the form became
reduced to −e. If the similarity to the Sanskrit form was not incidentally
caused by some later development, the Dameli form represents a very
archaic feature. 44
In the perfective and the future, as well as in the inchoative participle,
the suffix is just −a.
First causatives are usually constructed on the basis of an inherently
intransitive root, and have two arguments, just like an inherently
transitive verb. The meaning is usually related to the corresponding
simple verb, although the addition of another argument may change the
meaning quite radically. To exemplify, the simple verb based on the root
žup (i), seen in example 41, means ‘make oneself’ or ‘become’. The first
causative verb built on the same root, as seen in example 42, means
‘make’ or ‘create’.
(41) tẽẽ
duu maana ek ta
ḍakṭor žup−aa
3PL.ANIM.DIST two of.them one TOPSM doctor make−PFV.3SG
‘Of the two, one made himself [became] a doctor.’ (TW3005)
kya
nat žup−aai−tʰen
(42) … ni ba
… not TOPSH which nat make−CAUS−INDIRPST.3PL
‘…nor has anyone made any nat [devotional song].’ (TV0001)
In many cases, the first causative verb is semantically narrower than its
simple counterpart. While the simple verb may have several related
meanings, causative verbs often single out only one of these.
Inherently transitive roots are uncommon as bases for first causatives;
in these cases there is often only a simple verb and a second causative.
The exceptions usually have first causatives with a meaning that is more
narrow than the simple counterpart and do not necessarily take another
argument. Thus, the simple verb built on the root ɡaṭ (t) has a variety of
polysemic meanings, including ‘win’, ‘want’ (example 43) ‘ask’ (example
44), ‘need’ and ‘arrange a marriage’. Its first causative counterpart
(example 45), however, singles out only the latter meaning and means
‘to arrange someone’s marriage’, which is also a transitive verb with two
arguments.
44
This would not be the only exception to the MIA development, however. In Konkani,
the suffix is −aya (Masica, 1991, p. 317).
101
Verbs
ɡram žup−an
ɡaṭi−numa
ba
…
(43) aɡar taanu
if
POSS.REFL village make−INF want−IMPFV.1PL TOPSH …
‘If we want to make our own village…’ (T4037)
(44) tas
ta
aau ɡaṭi−na
3SG.ANIM.DIST from water ask−IMPFV.3SG.M
‘He asks water from him.’ (T5022)
(45) ay
too
ki ištrii ɡaṭ−aai−m
1SG.NOM 2SG.OBL for wife arrange−CAUS−IMPFV.1SG
‘I am arranging a wife for you.’ (TW1006)
Table 31 shows some first causative verbs with their corresponding
simple verbs to exemplify the meanings that causative verbs usually take.
Table 31: Simple and first causative verb meanings
Simple verb
stem/root
akaṭ (i)
alam (i)
laṭaŋɡ (i)
naɡ (i)
prambal (i)
ṭumbur (i)
wal/wail (i)
žup (i)
ž (t)
ɡaṭ (t)
Meaning
First causative
verb stem 45
Meaning
come together
hang
cross
come down
light up, look
good, start to
burn
roll
akaṭaai
alamaai
laṭaŋɡaai
naɡaai
prambalaai
mix, put together
hang something
take across
take down
light
ṭumburaai
hide
make oneself into
something
eat
want, win, etc.
wailaai
župaai
roll something, fell a
tree
hide something, steal
make
žuwaai
ɡaṭaai
feed
arrange marriage for
someone.
This is the most common stem, although a shorter one, with the affix −a instead of
−aai, is used in the direct past and the future, as described above.
45
102
The verb stem
11.2.1.2 Second causatives
Second causatives always involve three arguments. They are either
created on the basis of an inherently transitive verb or an inherently
intransitive verb that also has a corresponding first causative, so there is
always a corresponding transitive verb.
They are constructed by adding a second causative suffix, −aw, in
addition to the causative suffix −a/−aai, in the position before it (see
example 38). The typical meaning of second causatives, compared to
their transitive equivalents, is indirect action, doing something through
someone else. Compare, for example, the three verbs based on the root
niš (i) ‘sit, wait’.
(46) taɡa
talii ay
aram−a ni niš−im
tee
before? until 1SG.NOM rest−LOC not sit−FUT.1SG that
‘I will not sit down to rest, until…’ (T0028)
(47) ãã muu
keeraa paai madrasa−a
niš−aai−tʰum
ba
and 1SG.ERG which boy madrasa−LOC sit−CAUS−INDIRPST.1SG TOPSH
‘And the boy that I sat in the madrassah…’ (TW3052)
mãã−Ø
maam−a
ṣawaai
(48) ay
1SG.NOM 1SG.POSS−M maternal_uncle−OBL by
mas
niš−aw−aai−m
3SG.ANIM.PROX sit−CAUS2−CAUS−PFV.1SG
‘I sat him down through my uncle.’ [I asked my uncle to calm him
down.] (E0037)
The simple verb in example 46 is intransitive and only involves the
subject. The first causative in example 47 refers to the act of making
someone sit (figuratively) and involves the subject and a second person,
the boy who is made to sit (study) in the madrassah. The second
causative in example 48 involves three people: the subject, who instigates
the action; the object, who is acted upon; and the mediating uncle, who
carries out the action by making him sit down (again figuratively, the
actual meaning is ‘calm down’ or ‘end a conflict).’
The added argument of a second causative does not need to be
explicitly expressed, and the method for expressing it is periphrastic, as
a postpositional phrase, often with the postposition ṣawaai ‘by, through,
for’ (as in example 48).
103
Verbs
Although quite different in principle, second causatives bear some
pragmatic similarities to passives, since both allow a different role to be
marked as the actor, when compared to a clause that hasn’t undergone
the operation. In passives, however, this is done by switching the referent
of the subject and, indirectly, the actor role. In second causatives, it is
done by introducing an actor as a new argument, which changes the
thematic role of the subject into something like causer or instigator.
There is no morphological passive in Dameli, although non-finite verb
forms can be used to similar effects.
Table 32 shows some causative verb stems and their meanings, along
with their corresponding simple verbs.
Table 32: Simple and second causative verb meanings
Second
causative
stem 46
Root
Simple verb
meaning
ɡaṭ (t)
want, win, etc.
ɡaṭawaai
kook (i)
kur (t)
matr (t)
naɡ (i)
nikʰaal (t)
niš (i)
sleep
do
read
come down
take out
sit down
kookiyawaai
kurawaai
matrawaai
naɡawaai
nikʰaalawaai
nišawaai
wal/wail (i)
žan (t)
hide
kill
wailawaai
žanawaai
žup (i)
make oneself
into sth.
župawaai
Meaning
have s.o. arrange or arrange
s.o. a marriage
make s.o. put s.o. to sleep 47
do sth through s.o. else
make s.o. read/study
make s.o. take sth down
make s.o. take sth out
have s.o. calm s.o. down, have
s.o. make s.o. sit down
make s.o. hide sth.
have s.o. killed, make s.o. kill
s.o.
have sth made, make sth
through s.o. else
Second causatives are a common feature among the Indo-Aryan
languages (Masica, 1991, pp. 318-319).
In the direct past and the future, the shorter suffix −a is used instead of −aai.
In the literal sense, as when putting a child to sleep, not in the idiomatic sense of ‘kill
(a pet)’.
46
47
104
Finiteness
11.3
Finiteness
The term ‘finite’ (along with its derivatives, such as ‘infinitive’,
‘non-finite’ or ‘finiteness’) has been the subject of some controversy.
Although it is a central concept in both traditional grammar and several
formal theories of grammar, many authors consider it deeply
problematic. The distinction between finite and non-finite verb forms
developed in the grammatical traditions of European languages,
especially Latin, and it might be argued that it applies reasonably well to
these languages. As Nikolaeva (2007) points out, however, there are no
clear definitions of the term, and most, if not all, attempts to define the
term and apply it uniformly to the world’s languages have turned out to
be incompatible with the variation in these languages.
Dameli is an Indo-European language, however, and although there
is much that differentiates it from “Standard Average European”, the
shared Indo-European heritage assures that there is also much in
common between them, and many terms that have been found wanting
when applied universally, because they were too specifically attuned to
European languages, are perfectly applicable to Dameli.
Such is the case with finiteness. Verb forms in Dameli form two
groups: one contains verbs that show person, number and sometimes
gender agreement with the subject, can appear with an overt subject, and
function independently in sentences, and another contains verbs that do
none of these things. From a strictly language-internal perspective, it
makes sense to group forms that behave similarly together. Using the
terms ‘finite’ for the first group, and ‘non-finite’ for the latter is simple
and convenient, as long as it is understood that this does not imply that
these terms are anything but useful and readily available designations for
distinctions that exist in the language.
The finite verb forms in Dameli are the perfective, imperfective,
future, indirect past, potential past and imperative forms. These are
marked with a set of suffixes that also express subject marking.
11.4
Verb inflection
The final part of a Dameli verb is an inflectional suffix of one or two
syllables. These suffixes are portmanteaus; they contain several
meanings in a single morph.
In finite verbs, suffixes mark the TAM set to which the form belongs.
The Dameli TAM set is a morphologically defined set of verb forms that
105
Verbs
express a number of different semantic dimensions: tense, aspect, mood,
evidentiality and epistemic modality.
The same suffixes also carry subject marking according to several
categories: person, number and in some sets gender.
In non-finite verbs, the suffixes do not carry subject marking, and
although some forms are clearly marked for particular TAM dimensions,
such as being past or completed, their meanings are probably better
defined with reference to the constructions in which they are used. The
non-finite forms are often combined with finite forms of the copula verbs
to form finite complex TAM expressions, described in section 16.6.
The inflectional suffixes are shown in table 33. See also table 27 for
examples.
106
Verb inflection
Table 33: Verb inflection suffixes
Causative
verbs
Simple verbs
Form
Intransitive
1SG
2SG
3SG.M
3SG.F
1PL
2PL
3PL
−úm
−óp
−aa
−úi
−úma
−óba
−ún
Transitive
Perfective
−úm
−óp
−ee
−ee
−úma
−óba
−én
Imperfective
1SG
2SG
3SG.M
3SG.F
1PL
2PL
3PL
−num
−nap
−na
−ni
−numa
−naba
−nun
−num
−nap
−na
−ni
−numa
−naba
−nun
Indirect past
1SG
2SG
3SG.M
3SG.F
1PL
2PL
3PL
−tʰum
−tʰop
−tʰaa
−tʰui
−tʰuma
−tʰoba
−tʰun
1SG
2SG
3SG
1PL
2PL
3PL
−tʰim
−tʰis
−tʰiyo
−tʰima
−tʰiba
−tʰin
−tʰum
−tʰop
−tʰee
−tʰee
−tʰuma
−tʰoba
−tʰen
Potential past
−tʰim
−tʰis
−tʰiyo
−tʰima
−tʰiba
−tʰin
−m
−p
−ee
−ee
−ama
−aba
−en
−m
−ap
−a
−i
−ma
−ba
−n
−tʰum
−tʰop
−tʰee
−tʰee
−tʰuma
−tʰoba
−tʰen
−tʰim
−tʰis
−tʰiyo
−tʰima
−tʰiba
−tʰin
107
Verbs
Causative
verbs
Simple verbs
Person
Intransitive
Transitive
1SG
2SG
3SG
1PL
2PL
3PL
−im
−es
−o
−ima
−iba
−in
−im
−es
−o
−ima
−iba
−in
SG
PL
−ee
−aa
Future
Imperative
−ee
−aa
−im
−es
−o
−ima
−iba
−in
−i
−ya
Non-finite verb forms48
Infinitive
Present participle
Past participle
Inchoative participle
Conjunctive participle
−an
−aal
−isan
−em
−i
As the table shows, there is some systematicity in the form of the suffixes,
and it would be possible to associate meaning to some parts of them, at
a lower level than the entire suffix. For example, we might assume that
−a stands for plural. This could, with some assumptions, be said to hold
true for all first and second person forms throughout the system. It
would, however, ignore the fact that in the present third person singular
masculine, −a is not plural but masculine, and in the third person, plural
is not marked with −a but with −n.
More fundamentally, it would be an uneconomical description, since
we would have to assume that some suffixes were compositional,
whereas others, such as the future third person singular, −o, cannot be.
The correspondences between some forms and meanings are
obviously there, but they are not the whole story. In some places, −a does
indicate plural but if we were to consider it a plural morpheme, we
would have to postulate two separate morphemes with identical form in
48
The non-finite suffixes are identical for intransitive, transitive and causative verbs.
108
Verb inflection
the first/second and third person. In contrast with −im, −ima is indeed
plural because it contains −a, but in contrast with −ni, −na is masculine.
There are some differences between transitive and intransitive verbs
with regard to some of the third person forms. These have been marked
in bold in the table. The differences concern how consistently gender
distinctions are applied in the singular. Gender is always distinguished
in the imperfective, and never in the future or the potential past, but
intransitive verbs distinguish gender in the perfective and the indirect
past, and transitive verbs do not. The third person plural form of these
sets is also different in the two kinds of verbs. The causative inflectional
paradigm shows the same gender distinctions as the transitive paradigm
but differs from the paradigm of the simple verbs in some other ways,
with some different forms in the imperfective, the perfective, and the
imperative.
Between the stem and the inflectional suffix, an epenthetic vowel is
inserted if the combination would otherwise result in a disallowed cluster
(see 4.1.) Usually, the vowel is i, as in example 49, but vowel assimilation
can lead to other alternatives, as in examples 50 and 51.
(49) zaċ−i−na
trickle−EP−IMPFV.3SG.M
‘it trickles, drips’ (elicited)
(50) aaċ−u−num
take−EP−IMPFV.1SG
‘I take’ (elicited)
(51) aaċ−e−na
take−EP−IMPFV.3SG
‘he takes’ (elicited)
Since the causative suffix ends in a vowel, causative verbs never contain
these epenthetic vowels but are often affected by phonological processes
affecting combinations of vowels. When the causative suffix ending in
−a is followed by suffixes beginning with −e, the resulting sound is [−æ],
with length depending on the length of the suffix (see 6.1.3.
Combinations of two short articulations of −a may result in a long −aa
(see 6.1.2).
109
Verbs
11.4.1
Subject marking categories
The inflectional suffixes of finite verb forms contain subject marking
according to the categories of person, number, and gender. In many
clauses, this is the only identification of the identity of the subject, in
others the subject is overtly expressed with a noun or a pronoun. In these
cases, the verb may be said to agree with the overt subject, rather than
express or mark the subject.
11.4.1.1 Person
Finite verbs carry obligatory person marking, indicating whether the
subject of the verb is the first, second or third person. The imperative
shows no person distinctions but should be interpreted as inherently
marked for the second person.
11.4.1.2 Number
The category of number indicates whether the subject is a single person,
marked by the singular as in example 52, or several, marked by the
plural as in example 53.
(52) tas
ta−es
too
ki abut zaan−num
3SG.ANIM.DIST than−also 2SG.OBL for good know−IMPFV.1SG
‘More than that also I love you.’ (T9018)
(53) tee mẽẽ
that 3PL.PROX.NOM
muu
ki kati
1SG.OBL for how_much
abut zaan−nun
ɡan−i
good know−IMPFV.3PL say−CP
‘…how much they love me, he thought.’ (T9009)
Nouns and pronouns can be overtly marked for number, but verb
marking is not dependent on whether this happens or not, as can be seen
in example 54, in which only the verb indicates that the subject, nasal
‘generation’, refers to several persons (or, in this case, generations).
Plural is more consistently marked on verbs than on nouns, and only
some pronouns have plural forms.
(54) ainda nasal−es
amaa
kurei maaf
ni kur−in
future generation−also 1PL.OBL any pardon not do−FUT.3PL
‘And coming generations will not forgive us.’ (T2010)
110
Verb inflection
There is a tendency to avoid expressing number when the subject is
inanimate, using the singular even if several objects are referred to, as in
example 55. The inanimate copula does not have plural forms at all.
(55) mẽẽ
kul(−nam) naɡi prat−ee
3PL.INANIM.PROX house(−PL) fall give−PFV.3SG
‘These houses fell down.’ (E0099-100)
In conjunction with this, it may be worth noting that in several of the
older Indo-European languages, neuter plural subjects required singular
verbs, because they were originally collectives (Fortson, 2006, p. 231).
11.4.1.3 Gender
As explained in section 7.3.3, the gender category in Dameli is a complex
matter. Theoretically, all nouns are divided into two categories,
masculine and feminine, and this distinction is expressed as agreement
on several targets, including verbs, as it is in examples 56 and 57.
(56) mããtẽẽ
around
dac̣i−na
mãã−i
tukuri kii
see−IMPFV.3SG.M 1SG.POSS−F basket who
ɡiɡ−een
ɡan−i
take−PFV.3PL
say−CP
‘He looks around, thinking “Who took my basket?”’ (TP0038)
(57) lee zamaana aaṣanta ek diyoo brei dac̣i−ni
tee
very times
behind one day girl see−IMPFV.3SG.F that
‘Much later, one day the girl sees that… (T1039)
In reality, however, gender is only consistently distinguished with
animate nouns. See 7.3.3)
In many cases, verb inflection is not subject agreement but subject
marking. Many sentences are formed without an overt subject, leaving
the verb form as the only indicator of the identity of the subject. In these
cases, the semantic category sex is probably more relevant than the
morphological category of gender.
Since animate subjects are the norm, the decline of gender has less of
an effect on verbs than on words of other classes. The default gender
seems to be the masculine, which is used with most inanimates. Gender
marking is only present in some TAM sets: perfective, imperfective and
111
Verbs
indirect past for intransitive verbs, and just imperfective for transitive
verbs. For a more exhaustive discussion of gender, see 7.3.
11.4.2
TAM categories
The same suffixes that serve to point out the intended subject also
express other categories that position the event described in relation to
time, information source and speaker knowledge. These categories, tense
and aspect, mood, evidentiality and epistemic modality, are discussed
further in chapter 16.
11.4.3
Verbs with limited inflection
Some verbs are limited as to what kinds of subjects they allow. Verbs
that allow only inanimate subjects have limited inflectional paradigms,
since only the third person is possible. Since gender for inanimates is
rarely distinguished, only the third person masculine form of each set is
used. Most if not all of these verbs fall into two groups.
Members of the first group, shown in table 34, are always
morphologically intransitive and describe movement of liquids. They can
form zero-subject clauses (14.1.1) or be formed with the liquid as the
subject.
Table 34: Intransitive verbs with limited inflection
Root
ċarak (i)
ṭip (i)
zaċ (i)
Meaning
Used about
drip
drip
trickle
rain
any liquid
any liquid
The second group, shown in table 35, consists of first causatives, and
describes bodily sensations caused by some inanimate external agent,
such as temperature, burning nettles, or chili. Although these verbs are
morphologically first causatives, the way they are used is reminiscent of
passives in other languages, since there is often no specified subject.
Table 35: First causative verbs with limited inflection
Root
čičir (c)
ooš (c)
prapaṭ (c)
112
Meaning
Used about
smart, hurt
freeze, stiffen from cold
burn (fig.)
body parts
body parts
mouth
Finite verb forms
11.5
Finite verb forms
The finite verb forms are the perfective, the imperfective, the future, the
indirect past, the potential past and the imperative. They are always
marked with subject-marking suffixes and defined by the TAM
categories described in chapter 16.
The finite verb forms are not entirely independent of each other, and
it is possible to suggest some possible historical developments that cast
some light on the present situation.
The perfective suffixes are the shortest and, with a few exceptions,
represent the least common denominator of the other forms. The
perfective, the imperfective and the future appear to be the oldest, since
the indirect past and the potential past to have been formed by
combining verb stems with different forms of the copula.
11.5.1
The perfective
As suggested above, the perfective is in some ways the most basic verb
form. The suffixes that express the perfective are the least complex ones,
and several other forms appear to have been formed by combining the
perfective with forms of the copula, which have later merged into a single
verb form.
The suffixes that express the perfective are similar for intransitive and
transitive verbs, with the exception of the third person singular and
plural. Intransitive verbs distinguish gender in the third person singular:
−aa ‘PFV.3SG.M’, −ui ‘PFV.3SG.F’, but transitive verbs have a genderneutral form: −ee ‘PFV.3SG’. The third person plural form is also
different, in analogy with the above: −un for intransitive verbs, −én for
transitive verbs. (See table 33.) Causative verbs, being by definition
transitive, also neutralise the gender distinction. As in the imperfective,
but unlike all other TAM sets, the inflectional suffixes for causatives are
a slightly shortened form of those used with simple verbs. The function
of the perfective is described in 16.6.5, but the form also plays a role in
expressing the inchoative past (16.6.10).
Only this form and the imperfective are allowed in conditional
clauses. See section 19.3.
11.5.2
The imperfective
The imperfective is formed with a set of suffixes beginning in n, for both
intransitive and transitive simple verbs, but causatives use a shorter set
113
Verbs
of suffixes, without n. Because the causatives and the tʰ copula have
suffixes that do not begin in n, considering the entire suffix, e.g., −num
‘IMPFV.1SG’, a single portmanteau morph is probably a more economical
solution than assuming that −n is an imperfective suffix to which subject
marking is added.
The imperfective form is used to express the simple present (16.6.1)
and the continuous past (16.6.8), and also plays a role in expressing
completed actions with currently significant results, the present result
expressions (16.6.10 and 16.6.3).
11.5.3
The future
The future is formed with a set of short suffixes, most of which begin in
/i/, that remain the same for all kinds of verbs. There is no gender
distinction. The future suffixes are quite different from those used in the
imperfective, the perfective past and the indirect past, although similar
to those of the potential past, in that the second person singular of both
ends with /s/ rather than /p/, possibly because the potential past was
originally formed by a combination with future forms of the copula.
Morgenstierne (1942) assumed these forms to be older and the forms
beginning in /p/ to be a later development, possibly from a medial suffix
(p. 141). The function of the future form is described in 16.6.4.
11.5.4
The indirect past
The indirect past is formed with suffixes that are identical in form to
what is today the imperfective forms of the copula, although these
appear to have originally been perfective; 49 they begin in /tʰ/ and contain
the same distinctive subject marking elements as the imperfective and the
perfective. It seems likely, therefore, that the indirect past was originally
a periphrastic construction, in which a non-finite form, such as the
conjunctive participle, was combined with the perfective form of the
animate copula, similar to the present result 1 construction (16.6.2). If
this is indeed the case, the forms have now merged and have become
reanalysed as a single verb form.
Section 16.6.6 describes how the indirect past is used.
49
The forms are similar to the perfective forms of all other verbs rather than to the
imperfective. In addition, these forms are used together with the past tense marker taa in
the perfective as well, since actual perfective forms are missing.
114
Non-finite verb forms
11.5.5
The potential past
Like the indirect past, the potential past probably arose as a periphrastic
construction of a non-finite verb form with a form of the copula, in this
case the future, since the suffixes are identical to the future forms of the
animate copula (cf. tʰim ‘COP.ANIM.FUT.1SG/I will be’ and pʰuukitʰim
‘blow.POTPST.1SG/I may have blown’, etc.).
As in the future, gender is not distinguished, so there is no distinction
in the suffixes between transitive and intransitive verbs in this form. The
agreement suffixes used with causatives are also identical to those used
with simple verbs.
Semantically, there are clear parallels with both the indirect past and
the future. Like the indirect past, the potential past describes
unwitnessed events in the past. Like the future, it speaks of things that
may not be true but have the potential of being so.
The potential past is defined primarily by tense and epistemic
modality. See 16.6.7 for a discussion of the function of the potential past
form.
11.5.6
The imperative
The imperative is inherently understood to adress the second person, so
there are no person-marking suffixes; number is the only category that
is overtly marked by suffixes. The suffixes used with simple verbs, −e
‘IMP.SG’ and −aa ‘IMP.PL’, are different from those used with causative
verbs, −i ‘IMP.SG’ and −ya ‘IMP.PL’. The causative forms are used together
with the shorter causative suffix, −a.
The function of the imperative is described in 16.6.10.
11.6
Non-finite verb forms
The non-finite verb forms in Dameli can be defined as word forms
constructed on verb stems, but that do not take subject agreement
suffixes, that do not appear with overt subjects, and that do not function
as the main predicates of sentences. They may be inherently specified as
e.g., ‘past tense’, but there is no inflection paradigm. Some non-finite
verb forms show nominal agreement with the head of a noun phrase they
are part of, such as gender agreement on the infinitive when it is used
adnominally.
115
Verbs
There are five different non-finite verb forms: the infinitive, the
present participle, the past participle, the inchoative participle and the
conjunctive participle, as shown in table 36.
Table 36: Non-finite verb forms
Affix
(Root)
Infinitive
Present participle
Past participle
Inchoative participle
Conjunctive participle
11.6.1
−an
−aal
−isan
−em
−i
kur (t) ‘do’
Example
kuran ‘do’,
kuraal ‘doing’,
kurisan ‘done’,
kurem ‘doing’,
kuri ‘having done’
c̣ʰaar (t) ‘throw’
c̣ʰaaran ‘throw’.
c̣ʰaaraal ‘throwing’
c̣ʰaarisan ‘thrown’
c̣ʰaarem ‘throwing’
c̣ʰaari ‘having thrown’
Converbs or participles?
Haspelmath (1995, p. 4) suggests dividing non-finite verb forms into
three different groups, based on their function:
• Verbal nouns or masdars: verb forms that function as
arguments, like nouns.
• Participles: verb forms that function as adnominal modifiers,
like adjectives.
• Converbs: verb forms that function as adverbial modifiers, like
adverbs.
These terms allow for a clear functional division of labour between the
three different terms. When applied to a language such as Dameli, the
results are predictably messier.
The first problem concerns the term participle. In Haspelmath’s
schema, participles are exclusively adnominal modifiers, but in various
grammatical traditions, the term has not been quite so well delimited
and often denotes forms that modify verbs or clauses, forms that would
be converbs by Haspelmath’s definition. A highly relevant case in point
here is the conjunctive participle in the South Asian tradition, but also
such well-known examples as the English −ing-form, generally referred
to as the present participle, but clearly regarded as a converb by
Haspelmath (1995, p. 17).
But the problem is deeper than this alone: it is not uncommon for the
same form to function both adverbially and adnominally. Compare three
examples of the English −ing form: in “Sparkling Cyanide”, the form is
116
Non-finite verb forms
adnominal; in “Going Postal” it is adverbial; and in “Most Loving Mere
Folly”, it functions as an argument. To resolve the issue, Haspelmath
has to resort to frequency in order to determine whether the −ing form
is a participle or a converb. (1995, p. 20)
The other problematic point with the term is pointed out by Bickel
(1998), who proposes two different types of converbs, the Asian and the
European type. A key function of converbs in Asian languages is
‘narrative chaining’ or ‘clause chaining’ (see 21.9), which expresses
several subsequent events. In Haspelmath’s view, this function “is not a
central, typical use of the converb because it is not really adverbial”
(1995, p. 8). The problem is that in these languages, it is a central
function of these verb forms. Adverbial modification and narrative
chaining consistently reappear in the same forms throughout the area, as
it does in Dameli. And in Dameli, as in many other languages, narrative
chaining cannot be clearly separated from adverbial modification but is
a logical extension of this function. As Coupe shows for Hindi, there are
even cases where both an adverbial and a sequential reading is possible
for the same sentence (2006, p. 147). Likewise, the difference between a
sentence with a single, sequential converb clause and a sentence with two
or more such clauses is pragmatic rather than structural.
Despite the difficulties, the term converb may well be both relevant
and useful. After all, we are far from agreeing on the definition, relevance
and application of such well-established terms as verb, noun, subject,
phoneme or word.
Using new and hopefully more precise terms in the description has
obvious advantages but so does maintaining well-known and
widespread terms, sometimes despite their faults. In an attempt to both
have my cake and eat it I too, will endeavour to use both types of terms.
The terms suggested by Haspelmath and others, particularly ‘converb’,
will be used in the discussion, but despite the apparent contradiction, I
will retain terms that have been established in the local tradition, when
such exist, as designations for individual forms in Dameli.
Applying Haspelmath’s schema would give the following non-finite
forms for Dameli:
Table 37: Non-finite forms according to Haspelmath's schema
Masdar
Participle
Converb
Infinitive
Past participle
Present participle
Inchoative participle
Conjunctive participle
117
Verbs
11.6.2
The infinitive
The infinitive is formed by adding the suffix −an to the verb root and
functions as a verbal noun, a masdar. It is used when referring to an
action as a thing, much like the English or Latin infinitive. As such, it is
the form most commonly used when talking abstractly about a verb, as
a kind of “quotation form”.
(58) lekin see
zamaan−a talim
xasel
kur−an
but 3SG.DIST time−LOC education acquire do−INF
kya
aam
riwaž
ni beru
which common custom not be.PST.INANIM.3SG
‘But at that time it was not common custom to acquire education.’
(T4003)
The infinitive is at least partially nominalised, taking the locative noun
case suffix −a when followed by certain postpositions.
(59) abadi
žup−an−a
ṣaži mutalika kya
muškul daru
building make−INF−OBL for about
which talk
is
‘There are some things to say about making a building.’ (TA0002)
The infinitive is usually built on the root alone, not on the entire stem,
which means that it does not distinguish between simple verbs and
causatives. Thus, both the verb meaning ‘come together’ and the verb
meaning ‘put together’ or ‘mix’ have the infinitive akaṭan, regardless of
the fact that the former has the stem akaṭ and the latter ‘akaṭ−aai/akaṭ−a’.
In some cases, however, the causative suffix may be included in the
infinitive. Thus, the infinitive of the verb meaning ‘do through someone
else’ is ‘kurawan’, with the causative suffix –aw, not just ‘kuran’ as might
have been expected.
The infinitive is commonly used in tail-head linkage constructions.
(See 21.8.) It is also combined with some word derivation affixes, such
as −baṣ, which forms adjectives with the meaning ‘able to…’. (See 9.3.2.)
11.6.3
The present participle
The present participle is formed by adding −aal to the verb root or stem.
Like the infinitive, there is some uncertainty regarding whether the full
stem of second causatives can be used in creating this form, a matter
which will need to be investigated further. There are no instances in my
118
Non-finite verb forms
material of first causative stems being used to form the present participle,
and only a single instance of the second causative stem being used in this
way.
Most uses of the present participle are consistent with Haspelmath’s
definition of a participle (1995, p. 4), a verb form that primarily modifies
nouns. Like the past participle, the present participle is used to describe
persons or objects by attributing them to a process by which they are in
some way affected, as agents or patients or whatever the case may be,
meaning something like “the x who is doing y”, as in example 60.
Unlike the past participle, however, the present participle describes
processes that are still ongoing at the time in focus, whereas the past
participle describes completed processes.
(60) i
laak−aal
mač kʰošan ni tʰ−aa
3SG.ANIM.PROX weep−PRSPTCP man happy not be−IMPFV.3SG.M
‘This weeping man is not happy.’ 50 (E0003)
The present participle can incorporate other parts of a verb phrase and
takes the feminine agreement suffix −i when it is used as an attribute to
feminine nouns, just like a variable adjective. Both these points are
illustrated in example 61.
(61) see
malu−y−aal−i
brei kaa
ɡ−ai
3SG.DIST Malu−go−PRSPTCP−F girl where go −PFV.3SG.F
‘Where did this girl, who [usually] goes to Malu, go?’ (E0012)
The present participle can also be used to form the habitual past. (See
16.6.9.)
11.6.4
The past participle
The past participle is formed by adding the suffix −isan to the verb stem
(either a simple root or a causative stem). Like the present participle, it
can be used attributively and adverbially but is most commonly used
with the copula to form complex verbal expressions such as the present
result 2 (16.6.3).
When used adverbially, the past participle can be treated rather
similarly to nouns and occasionally receives noun inflection, such as the
locative in example 62.
50
Very perceptive, Sherlock.
119
Verbs
(62) kul
ta maaṣ
weeč−aal
nees−i
dac̣i−nun
house of people search−PRSPTCP go.out−CP see−IMPFV.3PL
tee ek muṭ−a
akas−isan−a
lam
cʰi−i
that one tree−LOC climb−PSTPTCP−LOC branch cut−CP
naɡi pra−i
zen baitʰaa
fall give−CP dead be−INDIRPST.3SG.M
‘The people of the house went out searching and saw that,
having climbed up a tree to cut a branch, he had fallen and was
dead.’ (T1012)
11.6.5
The conjunctive participle
The conjunctive participle is a converb, “a non-finite verb form whose
main function is to mark adverbial subordination” (Haspelmath, 1995,
p. 3). I will use the term ‘conjunctive participle’, nonetheless, since it is
the term applied to similar forms throughout the Indo-Aryan languages,
where it is widespread and extremely important. The conjunctive
participle is formed by adding the suffix −i to the verb stem, with the
exception of two highly irregular verbs, y (i) ‘go’ and ɡ (t) ‘bring’, which
have conjunctive participles ending in −de. Since the suffix is added to
the stem and not to the root, the conjunctive participle can be created
from both simple verbs and causatives, with distinguishable forms as a
result. The conjunctive participle takes no agreement suffixes.
As a non-finite verb form, the conjunctive participle forms
subordinate clauses, sometimes including objects or adverbials, but
cannot form complete sentences on its own.
The inherent meaning of conjunctive participles is a completed action
or event, something that has happened or been done before the event
described by the main clause. Typically, this means that it is used as
adverbials that relate an event to previous events, a function that lends
its use to several specialised constructions, such as clause chaining (21.9)
and tail-head linkage (21.8). In combination with the copula, it can be
used to express results of completed events that have current relevance
(16.6.10).
11.6.6
The inchoative participle
The inchoative participle is formed by adding the suffix −em to the verb
root. There does not appear to be a separate causative version. This form
120
Copula verbs
will require further study, to establish more precisely its function and
properties. The inchoative participle is used with the copula to express
the inchoative past (see 16.6.10).
11.7
Copula verbs
The copula verbs are a special case among the verbs. They are very
frequently used and display a number of irregularities. Two different
verbs are used, depending on whether the subject is animate or not. (See
section 7.2 for a description of the category of animacy in Dameli.)
While the animate copula is a fully inflected, highly suppletive verb, the
inanimate copula has only two forms, distinguishing past and present.
Because of this, a grammatically encoded tense/aspect distinction may be
considered the least common denominator defining the category of
verbs.
There is no specific negative copula; copula verbs are negated the
same way as other verbs.
The copula, in both its animate and its inanimate version, is very
widely used, both with non-verbal predicates (14.1.5) and as an auxiliary
used when forming complex verb expressions.
11.7.1
The animate copula
The copula verb used with animate subjects has the root tʰ but is very
irregular in its form and use. It is fully inflected as far as subject marking
is concerned, but the paradigm is defective and does not contain all TAM
sets, and the forms are used in unexpected ways. This is not entirely
surprising, since the regular indirect past and potential past appear to
have been formed by adding forms of the copula to the stem.
Only two finite sets are formed on the tʰ root, resembling,
respectively, the regular perfective and the future, but they are not used
like the forms they resemble. One set is formed by adding the regular
perfective suffixes (−um, −op, etc.), but when used on its own, the
meaning is imperfective. The other is formed by adding the future
suffixes (−im, −is, etc.), but is used in a wider set of contexts. It expresses
wishes, assumptions and hope in a way that resembles the potential past
but is used regardless of tense, instead defined primarily with reference
to evidentiality or epistemic modality. This is illustrated in example 63,
taken from a letter, where the writer does not actually have first-hand
knowledge of the weather in his reader’s country but assumes that it
121
Verbs
must be cold. The adverb žaa ‘now’ clearly sets the sentence in the present
time.
watan lee soor tʰ−iyo
(63) žaa min−a
now 2PL.POSS−M country very cold be−FUT.3SG
‘Your country will be very cold now.’ D3031.
In addition to these, several strategies are used to fulfil functions that the
defective paradigm does not cover. The imperfective forms (which rather
resemble the perfective), when combined with the past tense marker taa,
are used as perfectives. Similarly, verbs formed on the root b (i) ‘become’
overlap semantically with the copula and are often used in its stead.
The conjunctive participle always describes completed action, so the
form bai ‘having become’ can function as the conjunctive participle of
the copula without losing any part of its meaning. Future forms also
assume a change, allowing future forms of the b (i) roots to replace the
future copula, as in example 64.
(64) … yee watan
ta
tasãã−Ø
b−o
… this country PART 3SG.POSS−M be−FUT.3SG
‘…this land will be/become his.’
In contrast, the indirect past of the b (i) root (baitʰum, baitʰop etc.) are
used to replace the indirect past of the copula to mean ‘be’ only, without
usually retaining any sense of ‘change into a state’, as illustrated in
example 65, although in the indirect past, this is not an uncommon
interpretation in any case.
(65) taɡa
zumaana ek
mač
bai−tʰaa…
before times
one man
be−INDIRPST.3SG.M
‘A long time ago there was a man…’ (T0001)
Table 38 shows the different forms of the animate copula, and the forms
and constructions used to fill out the paradigm. No clear imperative or
inchoative forms have been found, perhaps because of semantic
constraints, so these have been omitted in the table.
122
Copula verbs
Table 38: The animate copula
Infinitive
1SG
2SG
3SG.M
3SG.F
1PL
2PL
3PL
tʰ
tʰen
Imperfective
b
ban
tʰum
tʰop
tʰaa
tʰui
tʰuma
tʰoba
tʰun
Perfective
1SG
2SG
3SG.M
3SG.F
1PL
2PL
3PL
tʰum taa
tʰop taa
tʰaa taa
tʰui taa
tʰuma taa
tʰoba taa
tʰun taa
Indirect past
1SG
2SG
3SG.M
3SG.F
1PL
2PL
3PL
baitʰum
baitʰop
baitʰaa
baitʰui
baitʰuma
baitʰoba
baitʰun
Potential past (and other functions)
1SG
2SG
3SG.M
1PL
2PL
3PL
tʰim
tʰis
tʰiyo
tʰima
tʰiba
tʰin
Future
1SG
bim
123
Verbs
boos
boo
bima
booba
bin
2SG
3SG.M
1PL
2PL
3PL
Participles
PRSPTCP
PSTPTCP
CP
11.7.2
baal
baisan
bai
The inanimate copula
With inanimate subjects, a different stem is used, and the inflectional
paradigm is limited to only two forms, the present/imperfective daru (see
example 66) and the past/perfective beru (example 67.) The two forms
bear no clear relation to the animate copula verbs.
(66) see
tʰaana brit
daru
3SG.DIST place border COP.INANIM.IMPFV.3
‘That place is the border.’ (T7005)
(67) yee
tʰaan−a
ɡram beru
3SG.INANIM place−LOC village COP.INANIM.PST.3
tara bai−tʰun
there COP.ANIM−INDIRPST.3PL
‘In this place was a village; there they were.’ (T8036)
The limited paradigm for the inanimate copula can mostly be explained
by the semantic restraints that apply to inanimate subjects. Since only
third persons can be inanimate, there is no inflection according to
person, and since the plural is already uncommon or excluded for
inanimate subjects (see 11.4.1.2), there is no inflection according to
number. The gender system of Dameli is very limited (see 7.3) and seems
to apply mainly in cases where it can be easily assigned on semantic
grounds, which excludes inanimates.
What remains, then, are the TAM distinctions. Since inflection for
these categories is expressed by the choice of a set of agreement forms,
these distinctions cannot be expressed in the ordinary way either and
have been reduced to the simple past/non-past distinction.
124
Conjunct verbs
Diachronically, these forms may have evolved from the third person
singular forms of a suppletive verb, much like the animate copula. The
agreement suffixes of the future and the potential past, which probably
reflect an older system, end in −o in the third person singular, which may
explain the endings of these forms.
11.8
Conjunct verbs
The number of verb roots in Dameli is quite limited, and a large portion
of verb meanings are not expressed by specific verb roots. As in most of
the South Asian languages, a construct known as conjunct verb allows
the creation of a large number of verbs for which there are no specific
roots. Conjunct verbs are combinations of two phonological words, a
verb complement and a light verb, that function as a single verb
semantically. Verb complements can be nouns, adjectives or elements
that are not used except in these conjuncts. Light verbs are a small group
of verbs with general meanings. They function as ordinary verbs in their
own right, but in conjunct verbs they retain very little of their lexical
meaning and are reduced to the function of causing the combination to
be interpreted as a verb. The most common light verbs in Dameli are y
(i) ‘go’, kur (t) ‘do’, pr (t) ‘give’, laɡ (c) ‘do?’ 51 and b ‘become’. There are
no formal differences between a general combination of object and verb
and a conjunct verb, which makes it impossible to generalise across all
occurrences. Table 39 shows some conjunct verbs.
Table 39: Some conjunct verbs
Verbal complement
aaŋɡuẉi ‘finger’
krããka ‘snores’
miskol ‘missed call’
ral ‘up’
soč ‘thought’
ṣũũki ‘whistle’
taŋ ‘trouble’
ṭelefun ‘telephone’
andaza ‘guess’
kabza ‘?’
Light verb
kur (t) ‘do’
kur (t) ‘do’
kur (t) ‘do’
kur (t) ‘do’
kur (t) ‘do’
kur (t) ‘do’
kur (t) ‘do’
kur (t) ‘do’
laɡ (c) ‘do?’
laɡ (c) ‘do?’
Meaning
give the finger
snore
make a missed call
take up
think
whistle
trouble
telephone
guess
take by force
51
This verb, which is probably borrowed from Pashto, is a special case, since it is
probably only used as a light verb.
125
Verbs
pata ‘knowledge’
aara ‘song’
baṣ ‘rain’
bawee ‘swim’
ḍak ‘lie’
ɡuuḍe ‘?’
naɡi ‘fall’
kuta ‘search
daara ‘fall’
čalii
abut ‘good’
pui
rawan
široo
milau ‘meeting’
laɡ (c) ‘do?’
pr (t) ‘give’
pr (t) ‘give’
pr (t) ‘give’
pr (t) ‘give’
pr (t) ‘give’
pr (t) ‘give’
y (i) ‘go’
y (i) ‘go’
y (i) ‘go’
zaan (t) ‘know’
b (i) ‘become’
b (i) become’
b (i) become’
b (i) become’
know
sing
rain
swim
lie (down)
wash, bathe
fall (animate subject)
search
fall (inanimate subject)
go into exile
love
understand
start < Psht: rawandal
begin
meet
Conjunct verbs are the most common strategy for creating new verbs in
Dameli, and many borrowed words have this structure (e.g., miskol kur
(t) <Eng. ‘make a missed call’, ṭelefun kur (t) <Eng. ‘telephone’ and soč kur
(t) <Urd. soč, ‘thought, consideration’).
Unlike many of its neighbours, Dameli does not have compound
verbs, combinations of two verbs that function as a single unit, usually
a non-finite form and a finite form. There may, however, be single-word
units that have arisen historically from old compound verbs, such as
ɡiaaċ (t) ‘bring’, which may have originally consisted of ɡi ‘having
brought’ and aaċ (i) ‘come’.
There are no obvious conjunct verbs with the inanimate copula or the
animate copula tʰ, although this is by necessity a matter of interpretation.
The two complements wee and teer show evidence of having
undergone grammaticalisation, as they show very little lexical meaning
and a closer phonological attachment to the verb they are used with
(particularly wee). Both are used to add a sense of completion to a verb;
teer is probably borrowed from Pashto. For instance, weeb (from wee + the
copula b) means ‘reach’ and teer b means ‘pass’, but both wee and teer can
also be used with kur ‘do’ to mean ‘complete’. 52
52
The verb weeč (t) ‘seek’ is probably unrelated to this.
126
12
Postpositions
The function of a postposition is to form a unit with the noun or phrase
that comes before it, relating it to something else. Spatial, social and
temporal relationships are prominent.
Some postpositions require that the nouns they join with take the
locative case; others do not.
Postposition phrases can be used within both noun phrases and verb
phrases, and even as clause adverbials.
Postpositions vary in their degree of phonologic independence. Some
postpositions, especially ĩĩ ‘appropriate place’ and baĩ ‘towards’, are
usually cliticised to the preceding word; there is no pause between the
two parts, and only one syllable is stressed.
The distinction between postpositions and case suffixes is difficult to
make. In this analysis, which is by no means the only one possible,
independent words, as well as clitics or suffixes that can be applied on
top of a case suffix, 53 are considered postpositions, leaving the locative
and the instrumental as the only peripheral cases. Another difficult
distinction to make is that between postpositions and certain nouns.
Nouns in certain case forms sometimes function exactly as postpositions,
as a consequence of their usual function, but it is also possible that some
of these become grammaticalised as proper postpositions, with changes
in meaning and use as a result.
12.1 Postpositions that require the locative
case
Five postpositions require the nouns that precede them to be in the
locative case: ki ‘to’, ṣaži ‘in order to’, ṣawaai ‘for’, mili ‘with (comitative)’,
and ĩĩ ‘appropriate place’.
53
Another possible interpretation would be to include two or more layers of case
suffixes, as done in Masica (1991, p. 231).
127
Postpositions
12.1.1
The postposition ki ‘to’, ‘for’
The postposition ki is one of the most common ones and is used to signify
a target, both in the literal, spatial sense (as in example 68) and in a more
figurative sense (as in examples 69 and 70).
(68) mẽẽ
swaat minɡauri ki ɡ−een
3PL.PROX.NOM Swat Mingora to go−PFV.3PL
‘They went to Mingora [in] Swat.’ (T6032)
A common use of ki is to indicate someone who receives or benefits in
some way from an action, as in example 69. This is also used to indicate
the recipient in clauses with the verb ‘to give’ and the addressee in clauses
with the verb ‘to say’.
(69) tu
muu
ki kati
abut zaan−ap
2SG.NOM 1SG.OBL for how.many good know−IMPFV.2SG
‘How much do you love me.’ (T9014)
(70) see
tʰaana paš tʰ−a−aw−aal
3SG.DIST place trap be−CAUS−CAUS2−PRSPTCP
bai−tʰun
ɡoora ki čoẉweela ki
be−INDIRPST.3PL gora for hoepoe
for
‘In that place, they were putting traps for goras and for hoepoes
[birds]. (T3080)
12.1.2
The postposition ṣaži ‘for’, ‘in order to’
Partially overlapping semantically with ki is the postposition ṣaži ‘for’,
which is used to indicate the purpose or beneficiary of an action, in the
sense ‘for someone’s sake’ (example 72) or ‘in order to’. In the latter
sense, it often appears with infinitives, as in example 71.
(71) abadi
žup−an−a
ṣaži mutalika kya
muškul daru
building make−INF−LOC for about
which talk
is
‘In order to build a house, there are some things to talk about.’ (TA002)
Unlike the other postpositions of this group, nouns preceding ṣaži are not
consistently rendered in the locative case; exceptions such as example 72
also occur.
128
Postpositions that require the locative case
(72) tu
keeraa ɡram ṣaži ek čoṣṭi
put žan−op
2SG.NOM which village for one only.child son kill−PFV.2SG
‘For which village did you kill an only son?’ (T4042)
12.1.3
The postposition ṣawaai ‘by’, ‘through’
The postposition ṣawaai ‘through’ is used primarily with second causative
verbs (11.2.1.2) to describe the third argument of the verb indirectly. A
second causative verb often has the person who requested, paid for, or
otherwise initiated an action as the subject, but the person who actually
performs the action is then described with ṣawaai. In all the examples of
this in my material, the perfomrer is an animate person, as in example
73:
(73) ay
mãã−Ø
maam−a
ṣawaai
1SG.NOM 1SG.POSS−M maternal_uncle−OBL by
mas
niš−aw−aai−m
3SG.ANIM.PROX sit−CAUS2−CAUS−PFV.1SG
‘I sat him down through my uncle.’ [I asked my uncle to calm him
down.] (E0037)
Judging by the form, ṣawaai could be a conjunctive participle of a second
causative verb, possibly based on the root ṣa ‘send’, although this is very
speculative, and no corresponding second causative verb has been found.
In Palula, the cognate ṣaawaá is used in a similar function, described
as ‘manipulee’. There, however, the form is quite transparent as a
converb of the verb ṣaawóo ‘turn on, light, take up, dress, etc.’, which is
in very general use. (Liljegren & Haider, 2011, p. 138)
12.1.4
The postposition mili ‘with’
The postposition mili is used to express comitative meanings, as in doing
something, or being, ‘with someone’. It is primarily used with animates,
such as aštrakaa ‘women’, in example 74; when used with inanimate
complements the meaning is usually spatial, as in ‘stand with/by/next to’.
129
Postpositions
(74) ni man−ee
not accept−PFV.3SG
ta
sapun sootii
aštrakaa
TOPSM all
together women
mili nanawaat ɡe−n
tas−a
ki
with nanawati take−INF 3SG.ANIM.DIST−OBL to
‘When he didn’t accept, they all went, with their women, to
seek nanawati 54 (mercy) from him.’ (TW5008)
12.1.5
The postposition ĩĩ ‘appropriate place’
The postposition ĩĩ is used to indicate that something is in its
“appropriate” place, in the place where it belongs or is expected to be.
In this, it is somewhat more abstract than most of the other pospositions.
In example 75, “on a jug” is in some way the appropriate or expected
place for a jug lid to be (see also example 40).
(75) ǰaɡ ta ṣaa ǰaɡ−a−ĩĩ
daro
jug of head jug−LOC−APP is
‘The lid of the jug is on the jug.’ (Q1006)
Even clearer than most other postpositions, ĩĩ is always cliticised to the
preceding word, perhaps since it consists of only a vowel.
12.2
Other postpositions
Several other words function as postpositions but do not require their
complements to be in the locative case. Many of these may actually be
nouns or phrases in various stages of grammaticalisation, and it can be
difficult to determine their exact status. While some have clearly been
reanalysed as postpositions, others may still act as nouns or other
constituents in certain syntactic positions.
12.2.1
The postposition ta ‘from’, ‘of’, ‘than’
Although it does not require the locative, ta ‘from, of’ is clearly a
postposition. There is no obvious grammaticalisation source or more
concrete meaning. It can be used spatially to mean ‘from’ but is also used
in a construction that is similar to possession, but which only allows
54
Nanawati is a Pashtun concept of conflict resolution in which you seek mercy from
your opponent by demonstrating your complete surrender.
130
Other postpositions
inanimate “owners” (see 15.3). Like the English of-possessive, it is used
very widely and with very general meaning, encompassing uses such as
“the morning of Domel”, “intention to go”, “the scenes of the story”,
“at the time of the insurrection”, i.e., part-whole relationships,
properties and results.
There is also a topic particle ta ‘previously mentioned topic’ (see 21.1),
which may easily be confused with the postposition, although the
function is quite different. The same applies to some extent to the past
tense marker (16.5), which is also very similar in form but is usually
pronounced with a long vowel.
12.2.2
The postposition ṣaa ‘on’
The postposition ṣaa ‘on, on top of’ is identical in form to the word
meaning ‘head’ and has clearly been grammaticalised from this word,
most likely perceived as a locative form, ‘on the head’ (the locative suffix
−a cannot be distinguished in words ending in long −aa.) The fact that ṣaa
is used with items that obviously have nothing that can be described as
a head, such as bum ‘ground’ in example 76, provides evidence of the
degree to which ṣaa has become reanalysed as a postposition.
(76) šey−nam bum
ṣaa daro
thing−PL ground on COP.INANIM.IMPFV.3
‘The things are on the ground.’ (Q1033)
12.2.3
The postposition neẉ ‘under’
The postposition neẉ means ‘under’ or ‘below’ and is used as in example
77.
(77) kitap kursi neẉ
daro
book chair under COP.INANIM.IMPFV.3
‘The book is under the chair.’ (Q1001)
12.2.4
The postposition taprei ‘for’
The postposition taprei is used in the sense ‘for’, ‘on someone’s behalf’,
‘because of something’. It may have originated as a combination of the
conjunctive participle of the verb pr (t) ‘give’ and the postposition ta ‘of’.
Regardless of its origins, taprei is clearly used as a postposition, as
illustrated in example 78.
131
Postpositions
ẉ−en
zuruuri
tʰaa
(78) mãã−Ø
1SG.POSS−M leave−INF necessary be.IMPFV.3SG.M
‘It is (was?) necessary for me to leave.’ [Lit. ‘My leaving is necessary.’]
see
taprei muu
…
3SG.INANIM.DIST for
1SG.ERG …
ay−es
nees−um
see
kʰoptan−a
1SG.NOM−also go_out−PFV.1SG
3SG.INANIM.DIST afternoon−LOC
‘For this [reason/purpose], I also went out, that afternoon.’ (TW21022103)
12.2.5
The postposition bãĩ ‘towards’
The postposition bãĩ ‘towards’ is used to indicate direction towards
someone or something. Like ĩĩ, it is always cliticised to the preceding
word.
(79) serf ek nazar muu−bãĩ
dac̣−i …
only one sight 1SG.OBL−towards
see−CP …
‘Only one glance you threw my way…’ (TV4006)
(80) dir ta
ay
širiŋɡal−bãĩ
rawan b−um
Dir from 1SG.NOM Shiringal−towards start
be−PFV.1SG
‘From Dir I started out towards Shiringal.’ (TW7002)
12.2.6
The postpositions tagii ‘from’ and tali ‘until’
The postpositions tagii ‘from’ and tali ‘until’ are often used in the same
clause in order to describe the extent of something, either in time or
space, by stating its starting and finishing point, as in example 81.
(81) yee daaman takriban daman nisaar taɡii lačiɡraam ãã piṣotʰan tali
this Domel about Domel Nisar from Lachigram and Pishothan until
‘This Domel [stretches] approximately from Domel Nisar to Lachigram
and Pishothan.’ (TA1047)
When used alone, these postpositions tend to retain the same sense of
describing a starting or stopping point, even when the other end is less
clear, as in example 82, where an area is described as ranging from a
given point and “in this direction”.
132
Other postpositions
(82) see
pʰar−ee
paaštʰan taɡii yee−bãĩ
3SG.INANIM.DIST over.there−INS Pashtan from 3SG.INANIM.PROX−towards
‘From that farther Pashthan and in this direction.’ (T3083)
Of the two, tali is more common than taɡii, perhaps because the more
general ta ‘from’ (see 12.2.1) covers much of the same same meaning as
tagii, and is often used in its place.
12.2.7
Nouns that function as postpositions
To exemplify how difficult it is to determine whether a word is to be
considered a postposition or belonging to some other word class, let us
look at the word manora ‘in the middle’. In example 83, which
exemplifies how the word is usually used, it functions exactly as a
postposition, synctactically and semantically.
(83) see
manora
3SG.INANIM.DIST middle
tanii
dos−ee
3SG.ANIM.DIST.ERG friend−ERG
ɡan−ee
tee
…
say−PFV.3SG
that
…
‘In the middle of this, that friend said that…’ (TW1024)
If we had only such examples, we might assume manora to be a
postposition and thus to have an invariable form. A couple of other
examples, however, such as 84, in which other forms of the word are
used, 55 show that manora is really the locative form of manor ‘middle’,
and therefore a full noun with the potential to be used in any case form.
(84) see
ek manor−ee paaštʰan ek ba
pʰar−ee paaštʰan
3SG.DIST one middle−INS pass
one TOPSH far−INS pass
‘There is one middle pass, and one farther pass.’ (T3077)
Naturally, this flexibility of use can also provide a source for new
postpositions, as nouns are gradually grammaticalised. Also compare
with 12.2.2. In both cases, the locative form is used for postposition-like
functions.
55
The use of the instrumental here indicates ‘a way through somewhere’, which is quite
typical for a mountain pass. See 7.4.2.5.
133
13 Minor or questionable
word classes
A number of words cannot readily be included in any of the major word
classes, yet do not necessarily form classes of their own, either because
they are too hetereogenous to form a coherent group or because they
consist of too few words.
13.1
Adverbs
Adverbs are words used to modify verbs, clauses, adjectives and other
adverbs. It is unclear whether Dameli possesses a separate word class to
fill these functions, although there are some indications to that effect.
The Dameli case system regularly allows nouns to perform adverbial
functions, and even beyond this, the language is relatively flexible in
using nouns, non-finite verb forms and words of indeterminate status as
adverbial modifiers. For this reason, most candidates for an adverb class
turn out upon closer inspection to be nouns, particularly in the
instrumental or locative case. This may be true even for a number of very
common words that are only used in adverbial functions, such as žaa
‘now’, ayaa ‘here’ and tara ‘there’. All these end in −a, which may be the
locative suffix −a.
Still, there are words that may with some justification be considered
adverbs: words that are only used in adverbial functions and that show
no indications of ever having been nouns. Of course, if an adverb class
exists, some words that may have started out as nouns, such as the
examples above, may well have developed into adverbs later.
Some of the most important functions for adverbs are to describe the
manner of an action, time, space, and modality, and to intensify other
meanings. Within these fields, expressions are often defined in contrast
or opposition to others. Words and phrases of various types may often
form part of such a system of oppositions, as in the temporal expressions
described in 17.1.
134
Particles
Table 40 shows some possible members of an adverb class.
Table 40: Some possible adverbs
Adverb
manuu
apatʰaar
pʰirki
žaa
mudya
itrii
čiir
baat
ayaa
tara
puču
bun
oor
pʰar
ṣin
baar
lee
ṭãu
pak
šãũ
kideši
13.2
Gloss
Semantic field
in this way
quickly
again
now
today
day before yesterday
late
after
here
there
up
down
near
far
up
down
very
pure
pure
pure
maybe
manner
manner
manner
time
time
time
time
time
space
space
space
space
space
space
space
space
intensifier
intensifier
intensifier
intensifier
modality
Notes
on a slope
on a slope
general use
modifies colour
modifies colour
modifies colour
Particles
Dameli has a number of short words with mostly grammatical functions
that do not readily fit into other word classes. These have mostly been
described elsewhere in this work, along with their functions. Some
examples are the topic particles ta and ba (21.1), and the vocative
particles a and e (21.4).
13.3
Conjunctions
While there are several different strategies to carry out coordination and
subordination, not all of which utilise individual function words for this
purpose, there appears to be a limited group of words specifically used
135
Minor or questionable word classes
to join clauses to each other. This includes the coordinating conjunction
ãã ‘and’ (18.2), the disjunctive conjunction ku ‘or’ (18.3), the adversative
coordinations xu ‘but’ and lekin ‘but’ (see 18.4), the subordinating
conjunctions, or subordinators, tee ‘that’ kuitee ‘because’ and nitee ‘that
not/lest’ (see 19.1) and the conditional conjunctions agar ‘if’ and agarka
‘if’ (see 19.3).
13.4
Question-words
Question-words, or interrogatives, in Dameli are similar both in form
and function: they all begin in k− and are used to form content questions
(see 20.1.1). The problem with considering them a word class of their
own, however, is that many members fit the definition of other word
classes as well. This is, of course, caused by the particular function of
these words and is not particular to Dameli; since each question-word is
used to ask about a particular group of constituents, it tends to behave
like these constituents syntactically and often morphologically as well.
Of course, Dameli question-words share these features with the whwords of English and their equivalents in other Indo-European
languages, to which they are most certainly related.
Apart from forming questions, some of these words can also be used
as indefinite pronouns (8.3) when forming relative clauses (19.3). Some
of the question-words form part of a system that also includes deictic
distinctions (e.g., manuu ‘in this way’ tanuu ‘in that way’ kanuu ‘how’).
See 8.1.7.
Table 47 shows the question-words of Dameli and the word class with
which each word most closely aligns.
136
Question-words
Table 41: Question-words
Question-word
kii
kuree
keeraa
kasãã
keer
kanuu
ku
kaa
kutaal
kati
kya
Meaning
who
who
which
whose
when
how
why
where (general)
where (direction)
how many
what
Word class
pronoun
pronoun
pronoun
pronoun (possessive)
adverb
adverb
adverb
adverb
adverb
adjective
noun
137
Syntax
14
Basic declarative clauses
The clause type used for declarative main clauses is unmarked, both in
the sense that it is the most common type and in the sense that all other
types can be described as minor deviations from this type.
The typical pattern, for all clause types, is SOV, with the final verb as
the most important, and the only obligatory part. Other clause types are
morphologically marked, so that an added word or morpheme marks
the type rather than changes in word order: the question marker −i is
added to create polar questions, question-words to create content
questions, verb morphology marks most subordinate clauses and
imperative clauses, and the negator ni ‘not’ marks negative clauses. This
does not mean that changes in word order do not occur, only that they
are not the main means for creating other clause types. Operations such
as fronting or afterthought constructions (21.7) do cause changes in
word order, and some positions, especially the subject, can be moved
around quite freely.
A basic, declarative main clause in Dameli can contain three main
sections: a clause adverbial, a subject, and a predicate, which consists of
the verb and any objects. Only the verb itself is obligatory and must be
finite in order to form a main clause. All other parts can be omitted; a
single finite verb can form a full sentence. Example 85 shows the order
of a typical main clause.
Clause adverbial
Subject
Object
(85) bun
nag−i
see
ek tukuri
down come_down−CP 3SG.DIST one basket
‘Having come down, he fills a basket.’ (TP011)
Predicate
Verb
por−na
fill−IMPFV.3SG
The first section of the clause, the clause adverbial, is used to situate the
clause in relation to the world and the narrative context. As in the
example, this position is often filled by tail-head linkage (21.8), which
relates the clause to previously described events. More than one clause
141
Basic declarative clauses
adverbial can be used in this position; the number appears to be limited
only by pragmatic concerns, as evidenced by clause chains (21.9).
Clause adverbials are followed by a noun phrase functioning as
subject, commonly a pronoun. The subject is also indicated by subject
marking on the verb, and overt subjects are often omitted.
The subject is followed by the predicate, which always include a verb
at the end. Any objects precede the verb. Indirect objects, or postposition
phrases that function like indirect objects, are usually placed before
direct objects.
The head of the predicate is the verb, which is placed in the final
section. For clauses with non-verbal predicates, the copula verb fills this
position, and the non-verbal predicate element is placed in the same
position as objects. Modifications of the verb, such as adverbials, are
placed in this section, before the verb itself. It is not uncommon,
however, for elements of the clause to be placed after the verb, in an
“afterthought” construction (see 14.1.5).
14.1
Valency clause types
Basic clauses come in a number of variants that are caused by differences
in the type of the main verb. One major factor is the transitivity of the
verb, which affects the number of arguments of the verb. Dameli verbs
carry morphological marking of transitivity; verbs can be
morphologically intransitive, inherently transitive, first causative or
second causative. These four types of verbs are used to form clauses with
different transitivity. There is a regular relationship between
morphological transitivity and clause type, but it is not entirely
straightforward. Finally, copula verbs are used in clauses that do not
express verbal action, and require different types of arguments.
14.1.1
Zero transitivity clauses
Explicit subjects are not a requirement in complete sentences, but subject
marking on the verb is compulsory. In some cases, especially when
describing impersonal processes like the weather or the movement of
fluids, there may not be anything involved that a subject could refer to.
In these cases, subject marking reverts to a sort of “default option”: the
third person singular masculine.
The zero transitivity, or impersonal, clauses formed with these verbs
thus have no arguments of the verb and typically consist of a single word,
142
Valency clause types
the verb itself, which usually comes from the group of verbs with limited
inflection (11.4.3). Example 86 shows a zero-transitive clause.
(86) ċarakina
raining
‘It is raining.’
14.1.2
Intransitive clauses
Intransitive clauses are formed with morphologically intransitive verbs
(see 11.1) and take a subject as their only argument, as in example 87.
(87) tẽẽ
lee kʰasar−un
3PL.ANIM.DIST.NOM very tire−PFV.3PL
‘They tired greatly.’ (‘They became very tired.’) (TW8006)
14.1.3
Transitive clauses
Transitive clauses can be formed with inherently transitive verbs (11.1)
or with first causatives (11.2.1.1). Both take a subject and a direct object.
Example 88 shows a transitive clause with an inherently transitive verb,
ku (t) ‘do’, and example 89 a transitive clause with a first causative
formed on the root ṣa (c) ‘send’.
ku−ni
(88) pay kaaṇ
goat sound do−IMPFV.3SG.F
‘The goat makes a sound.’ (TP1003)
ki ǰarɡa ṣa−a−een
(89) ɡan−i tas−a
say−CP 3SG.ANIM.DIST−LOC to Jirga send−CAUS−PFV.3PL
‘Having said this, they sent a jirga (delegation) to him.’ (TW5006)
14.1.4
Ditransitive clauses
Ditransitive clauses can be formed with ditransitive verbs such as ‘give’,
that take a subject, a direct object and an indirect object. Ditransitive
verbs are only distinguished semantically; there is no morphological
marking to set them apart from other transitive verbs. The indirect object
can be formed with a postposition phrase, usually with the postposition
ki ‘to/for’ or with a core argument in the locative or oblique case. The
subject in these clauses is usually an agent, the object a theme and the
143
Basic declarative clauses
indirect object a recipient/benefactive. Example 90 shows a ditransitive
clause with the verb pr (t) ‘give’ and the indirect object as a postposition
phrase.
ki pããč lak rupay muu pr−es
(90) ãã too
and 2SG.OBL to five lac rupees 1SG give−FUT.2SG
‘And I will give five lakh rupees to you.’ (TW1072)
Ditransitive clauses can also be formed with second causatives, which
take a subject and a direct object as their core arguments, and a
postposition phrase as a third argument, usually with the postposition
ṣawaai ‘through by’. In these clauses, the subject is typically the person
who causes or instigates the action, the object usually something
inanimate that is being acted upon, and the third argument the one who
actually carries out the action. The third argument appears before the
direct object, in the same position as an indirect object (see example 48
for an illustration).
14.1.5
Clauses with copula verbs
Non-verbal clauses are typically avoided in Dameli; 56 words that are not
verbs are usually accompanied by a copula verb in order to function as
predicates and form clauses, and clauses that are usually non-verbal in
many languages, such as examples 91 and 92, are mostly rendered with
a copula verb.
(91) ay
ta
ḍakṭor tʰum
1SG.NOM PART doctor be.IMPFV.1SG
‘I’m a doctor.’ (TW3021)
(92) ãã bi
ba
mãã−Ø
put−suu
tʰoba
and 2PL.NOM TOPSH 1SG.POSS−M son−KIN.PL be.IMPFV.3SG
‘And you are my sons.’ (TW3030)
Adjectives that function as predicates are placed before the copula, as in
example 93. Variable adjectives show gender agreement with the subject.
56
As always, this is a matter of interpretation. Counterexamples can be found,
particularly in enumerations, but I would argue that they are not complete clauses but
cases of ellipsis.
144
Valency clause types
(93) see
mač aluuna
tʰ−aa
3SG.DIST man tasteless−M be−IMPFV.3SG.M
‘That man is tasteless (not serious)’. (TI0013)
Nouns that function as predicates are placed before the copula, just as
adjectives are, as shown in example 94, and take the unmarked
nominative case.
(94) yee
ek tʰaana daro
3SG.INANIM.PROX one place be.INANIM.IMPFV.3
‘This is one place.’ (TW5034)
A common type of non-verbal predicates consist of locative predicates,
which state where something or someone is by use of a noun in the
locative case, as in example 95, or a postposition phrase, as in example
96.
(95) baaṭ buum−a
daro
stone ground−LOC be.INANIM.IMPFV.3
‘The stone is on the ground.’ (Q0017)
(96) kitap kursi ṣaa daro
book chair on be.INANIM.IMPFV.3
‘The book is on the chair.’ (Q1002)
Existential predicates, predicates that state the existence of something,
are often very similar in form to locative predicates. Although it is not
required, many clauses with existential predicates also contain a locative.
The difference is in the topic structure: a locative predicate states where
something is; an existential predicate states that something exists there.
This is reflected in the word order of the clause.
In Dameli sentences, the theme or topic of the sentence is usually
placed first, with the rheme or comment, the information ascribed to it,
coming after it. This is true for regular basic clauses, where the subject
comes first, followed by the predicate, and for other, less typical clauses,
such as those with non-verbal predicates.
In a sentence such as example 95, the subject baaṭ ‘stone’ appears first,
since it is also the theme, and is followed by the locative predicate. In
clauses with existential predicates such as example 97, in contrast, the
145
Basic declarative clauses
locative predicate is the theme rather than the subject and is therefore
placed first in the clause.
(97) ɡilaas−a san daro
glass−LOC hole be.INANIM.IMPFV.3
‘There is a hole in the glass.’ (E0025)
Possessive predicates are also non-verbal in Dameli and are expressed
with possessive attributes (see 15.3) and copulas, as seen in example 98.
traa žu−suu
bai−tʰun
(98) tasãã−i
3SG.POSS−F three daughter−KIN.PL be−INDIRPST.3PL
‘He had three daughters.’ [Lit. ‘His three daughters were.’] (T9002)
Possessive predicates also differ from existential predicates in terms of
information structure, although the word order is identical.
As the plural verb marking shows, the entire noun phrase tasããĩ traa
žusuu ‘his two daughters’ functions as the subject of the clause. The
information structure of the clause, however, does not match the clause
structure. Just as in English have-clauses, the possessor is the theme of
the clause, as can be seen in sentences containing one of the topic
particles, ta ‘same-topic’ or ba ‘shift-topic’, as in example 99.
Theme
Rheme
Subject
Predicate
(99) tasãã−Ø
ba
ek čoṣṭi
put bai−tʰaa
3SG.DIST.POSS−M TOPSH one only.child son be−INDIRPST.3SG.M
‘He had an only son.’ (T4002)
In example 99, the shift-topic particle ba is used directly after tasãã ‘his’
and not after the entire NP, reflecting the fact that although tasãã isn’t
the whole subject, it is the topic/theme of the clause. The rest of the
sentence makes up the comment/rheme.
146
15
The noun phrase
Noun phrases are units that serve in the same functions as nouns or
pronouns, but which may be expanded with modifications of different
kinds. Nouns, infinitives (which are nominal derivations of verbs, see
11.6.2), and pronouns can head noun phrases, and serve in this function,
although pronouns are usually used without modifications. Common
modifiers are demonstrative pronouns, adjectives, possessives,
postposition phrases and counting words.
Noun phrases typically function as arguments of the verb but can also
serve as e.g., attributes or adverbials, together with postpositions or
independently. Noun phrases show a left-branching structure, like most
aspects of the language, where the head is the last element of the phrase,
and less central elements are placed to the left of the head. Example 100
shows a noun phrase divided into its main components.
Attribute
Head
(Adverbial)
(100) ek
lee
ɣarib mač
poor man
one
very
‘A very poor man’ (T4001)
Determiner
Two different kinds of modifiers can be added to expand the noun
phrase, determiners and attributes. Determiners include demonstrative
pronouns, which function somewhat like definite articles, and the
numeral ek ‘one’, which is used like an indefinite article. Indefinite
pronouns are also used in this position. Attributes are adjectives and
other words or phrases used to further modify the head. In noun phrases
that contain both determiners and attributes, the attributes are usually
placed between the determiner and the head. Possessive pronouns and
phrases may be placed there or in the beginning of the phrase, preceding
any determiners. Some attributes and determiners display agreement
with the head: demonstratives agree with the head in regard to animacy,
number and case, and some adjectives and possessives agree with the
gender of the head.
147
The noun phrase
15.1
Determiners
Demonstrative pronouns are often used together with nouns,
functioning quite similarly to definite articles. There are examples of
nouns with definite functions used without determiners, but they are
quite rare. Demonstratives agree with the animacy, number and case of
the noun they are describing but are not marked for gender.
When a noun is being introduced into the discourse, the numeral ek
‘one’ is used as an indefinite article. Plural nouns can be introduced by a
higher numeral or by indefinite expressions such as ek kati ‘a few’ or
keeraa ‘some’, as in example 101. Example 102 shows a demonstrative
pronoun used as a determiner.
Determiner Attribute
Head
(101) keeraa
foren
maaṣ
some
foreign people
‘Some foreign people’ (T2013)
Determiner
(102) mas
this (3SG.ANIM.PROX.OBL)
‘this boy’ (T2013)
15.2
Head
paai
boy
Attributes
The head of a noun phrase can be further modified by attributes, which
provide additional descriptions. Attributes are formed with either
adjectives (9), participles 11.6), possessives (15.3), postpositions (12) or
other nouns, particularly in the locative case (7.4.2.4). Possessive
pronouns and phrases (see 15.3) may be placed in the beginning of the
phrase, preceding any determiners, but other attributes are usually
placed just before the head, as illustrated in example 103.
(103) ek lee xofnak
baas teer kur-ee
one very dangerous night pass do-DIRPST.3SG
‘He spent a very dangerous night.’ (T5017)
148
Possession
15.3
Possession
Possession, the notion that something in some way “belongs” to
something or someone else, is usually expressed as an attribute, although
possessive attributes show greater variety in their placement than other
attributes do.
Prototypical possession, expressing ownership in the strictest sense of
the word, naturally requires an animate owner, but a similar
construction is used with inanimate “owners” to express relationships
that bear some resemblance to ownership, such as part-whole
relationships, properties, etc. Although similar, two different
constructions are used for possessive constructions with animate and
inanimate “owners”. With inanimate possessors, a postposition phrase
formed with ta ‘from, of, than’ (12.2.1) is used.
The possessive marker used with animate possessors, sãã (see 8.2.1),
is harder to classify. While postpositions are usually invariant, sãã agrees
in number with the possessor; the plural form is sun−. It also agrees in
gender with the possessed, in the same way as an adjective agreeing with
the head of the noun phrase. Example 104 illustrates both these points.
In both these regards, the marker is similar to the possessive pronouns.
(104) paai sãã−i
yii
c̣aŋ pre−i
boy 3SG.ANIM.POSS−F mother (F) cry give−CP
ɡan−ni
tee …
say−IMPFV.3SG.F that …
‘The boy’s mother cried out, saying that…” (T1020)
Possessive personal pronouns (section 8.1) are also used to express
possession, but are usually limited to animate possessors. Like the
marker of animate possession, they inflect for the gender of the
possessed, as in example 105.
(105) ãã
and
ɡram ta
maaṣ sãã−Ø
zaadi−a
village POSS people POSS−M heart−LOC (M)
tasãã−Ø
kadar
lee
ziat b−aa
3SG.POSS−M respect (M) very much become−PFV.3SG.M
‘And in the hearts of the people of the village came great respect
for him.’ (T4050)
149
The noun phrase
Possessive predicates are formed by using possessive phrases with the
construction used for existential predicates, involving the copula (see
14.1.5).
These constructions can be used to express all kinds of possession,
including relationships that are often described as inalienable, such as
kinship and the “ownership” of body parts, but for kinship terms, there
is also a system of person suffixes that takes on some of the functions of
possession (see 7.5.2).
150
16 Tense, aspect and related
categories
Describing events in Dameli involves situating them in several
dimensions, such as time, the origin of the information, and how much
of it is known by the speaker. The categories involved are tense and
aspect, mood, evidentiality and epistemic modality. When the
abbreviation TAM is used in this chapter, it refers to all these categories,
not just to tense, aspect and mood.
On one level, these categories are expressed with the same sets of
markers that also express subject marking on finite verbs and with the
non-finite verb suffixes. Only some of these verb forms can be used
alone, however, and some change their meaning when combined with
the copula or the past tense marker taa. The resulting constructions, both
simple and complex, are used to describe more fine-grained TMA
distinctions.
16.1
Tense and aspect
In Dameli, tense, i.e., the time of the event, and aspect, i.e., the temporal
structure of the event, are difficult to separate, but both are relevant for
explaining both finite and non-finite verb forms. An important line
divides, on the one hand, the imperfective and non-past forms and, on
the other hand, the perfective and past forms. In the finite verbs, this is
reflected in the case system, which applies ergative alignment to clauses
with past/perfective forms, and accusative alignment to clauses with
imperfective/non-past forms (see 7.4.2.1).
The next division relies primarily on tense, dividing the finite verb
forms into a three-way distinction between the past, the present and the
future, as in table 42.
151
Tense, aspect and related categories
Table 42: Three-way tense distinction
Past
Present
Future
Perfective
Indirect past
Potential past
Imperfective
Future
While this seems to be a clear example of tense, aspect meanings such as
completedness probably affect the choice of form as well, in any given
situation. The contrast between the perfective and the indirect past also
involves evidentiality, but the forms are often used to signify a difference
in tense as well, with the indirect past being used for things that
happened longer ago. In these cases, the distinction is four-way rather
than three-way. The constructions described in 16.6 provide even finer
divisions involving tense and aspect and other categories.
16.2
Mood
Mood is a category concerned with the kind of speech act an utterance
is intended to perform. In Dameli, it is primarily relevant in defining the
imperative 11.5.6). The imperative is a very limited verb form, used to
express actions that the speaker wants the listener to perform, that is,
instructions or orders. In this, it speaks not of what is currently true, but
of what the speaker wants to become true. All other verb forms may be
considered to be indicative, that is, they imply that what is said is
intended as a statement, unless other elements of the clause indicate
otherwise. See chapter 20 for a discussion of different clause types.
16.3
Evidentiality
Evidentiality is the grammatical category that expresses source of
information. Common examples include distinctions between statements
for which there is evidence and statements that lack such evidence;
between information that was directly observed, inferred by the speaker,
or rests on information from someone else (Aikhenvald, 2003, p. 1); and
between information based on visual, auditory or other sensory
information.
In Dameli, evidentiality is one of the contrasts that form the
distinction between the perfective and the indirect past. The perfective is
152
Epistemic modality
used for events (in the past) that were witnessed by the speaker, the
indirect past for events (likewise in the past) that are hearsay, surmise or
second-hand reports. The two forms are also used about recent and
remote past events, but in cases like example 106 and 107, the
evidentiality contrast is quite clear.
(106) tanii
yee kram doos
kur−ee
3SG.ANIM.DIST.ERG this work yesterday do−PFV.3SG
‘He did this work yesterday.’ [The speaker saw this.] (E0054)
(107) tanii
yee kram doos
kur−tʰee
3SG.ANIM.DIST.ERG this work yesterday do−INDIRPST.3SG
‘He did this work yesterday.’ [The speaker did not see this.] (E0055)
The potential past is primarily defined in terms of epistemic modality
but can also be said to contrast with the perfective in terms of
evidentiality, since using the potential past precludes any direct
knowledge of the event.
16.4
Epistemic modality
Epistemic modality, as a category, refers to the grammatical coding of
the speaker’s attitude to the event being described. As such, it is closely
related to mood and evidentiality, but works on a different level.
Whereas mood typically determines the kind of speech act the utterance
is interpreted as, and evidentiality the source of the information,
epistemic modality typically occurs in statements, and expresses whether
the speaker is asserting these statements as truths.
In Dameli, epistemic modality is the contrast most important in
defining the potential past. The potential past is used when describing
situations that might have taken place in the past, when the speaker does
not want to commit to the truth value of the proposition, e.g., when
describing an event that might have taken place without being sure that
it in fact did. Compare example 108 below with examples 106 and 107
above for an illustration of this.
153
Tense, aspect and related categories
(108) tanii
yee kram doos
ku−tʰiyo
3SG.ANIM.DIST.ERG this work yesterday do−POTPST.3SG
‘He may have done this work yesterday.’ [The speaker does not
know whether this is actually the case.] (E0056)
16.5
The past tense marker taa
The marker taa is used as a past tense marker when forming complex
TMA constructions. It can be used with verb forms that are otherwise
interpreted as non-past, such as the imperfective, to create past tense
complex verb expressions such as the continuous past (16.6.8), where
the aspectual meaning is maintained, but the tense is changed. Use of taa
is not limited to situations where a past tense reading is an exception;
the marker is very common with verbs in the indirect past tense, where
a non-past reading would not be expected in any case (see example 109
for an illustration).
(109) doos
see
baazar ye−tʰaa
taa
yesterday 3SG.DIST bazaar go−INDIRPST.3SG.M PST
‘Yesterday he went to the bazaar.’ (Q0022)
16.6
TAM constructions
All the verb forms of Dameli have inherent TAM interpretations, but the
more fine-grained TAM distinctions are expressed not by the verb forms
in themselves, but by the constructions they form. These constructions
can consist of the verb forms alone, in some cases, or of combinations of
verb forms with forms of the copula verbs and/or with the past particle
taa. Thus, the imperfective form (11.4.3) can be used on its own to form
the simple present (16.6.1), but it can also be used with the past particle
taa to form the continuous past (16.6.2). The inchoative participle
(11.6.6), in contrast, appears only to be used in combinations with the
copula.
The constructions described here are the simple present, the present
result 1, the present result 2, the future, the perfective, the indirect past,
the potential past, the continuous past, the habitual past, the inchoative
past and the imperative, but there are probably a number of other
constructions yet to be discovered.
154
TAM constructions
16.6.1
Simple present
The simple present is expressed by the present form of a verb and is
prototypically used to describe situations that take place at the moment
of speaking to describe actions or states that are currently happening.
(110) mãã−i
luṭi
bawi
laaki−num
1SG.POSS−F young;girl daughter_in_law cry−IMPFV.1SG
‘I’m crying for my young daughter−in−law.’ (T1022)
The present can also be extended to describe immediate intentions, or
even things happening in the past, when the time-frame is anchored in
some other way, as in the context of a story, as example 111 shows.
(111) waali
tas
paay−a
aau
pre−na
shepherd 3SG.ANIM.DIST boy−LOC water give−IMPFV.3SG.M
‘The shepherd gives the boy water.’ [The author retells an experience
from his childhood.] (T5023)
16.6.2
Present result 1
The present result construction is formed with the conjunctive participle
and a present form of the copula. This construction is used to describe a
completed event, the results of which remain significant.
(112) tu
kaa
niši−i
tʰop
2SG.NOM where sit−CP
be.IMPFV.2SG
‘Where are you sitting.’ (Q0002)
The basic meaning of the verb niš (i) ‘to sit’ is punctual rather than
continuous, as in ‘to sit down’ rather than ‘to be sitting’ but the present
result lets the speaker describe this as a currently ongoing process. In
example 112, thus, the person has already sat down, and is still sitting
at the time of speaking. While it is conceivable that a similar construction
could be formed with any finite form of the copula, only examples in the
present have been found.
155
Tense, aspect and related categories
16.6.3
Present result 2
A very similar construction, both in form and meaning, is formed by
combining the past participle with an imperfective form of the copula,
as in example 113.
(113) ay
ayaa niš−isan
tʰum
1SG.NOM here sit−PSTPTCP be.IMPFV.1SG
‘I’m sitting here.’ (Q0001)
There may be a slight difference in meaning and extension between the
two forms, in that the past participle can be used in situations where the
action described by the participle is not immediately observable any
longer, as in example 114.
(114) ay
baloo ḍããš tali
ye−isan
tʰum
1SG.NOM big
rock until go−PSTPTCP be.IMPFV.1SG
‘I have gone all the way to Baloo Dash [the big rock]’ (TI0002)
The past participle construction is more common than the conjunctive
participle one.
The difference is consistent with the way that conjunctive participles
are usually used to describe events immediately preceding other events,
whereas past participles describe completed actions and their results.
This is a very tentative observation, however, and one that needs further
study before it can be confirmed and expanded upon.
The present result construction bears close resemblance to a passive.
By describing an event as a result rather than an action, it allows the
subject to be removed from the subject position, as in example 115,
which requires a passive translation.
(115) yede
... laka rooza waar−isan
tʰum
taa
go.CP ... like fast
catch−PSTPTCP be.IMPFV.1SG PST
‘Having gone, well, I was caught by/in the fast.’ (TW2019)
16.6.4
Future
The future is formed with the bare future form of the verb, and is used
to describe situations that are predicted or expected to take place in the
future, as in example 116.
156
TAM constructions
(116) ainda
coming_times
nasal−es
generation−also
amaa
1PL.OBL
kurei
which
maaf
ni kur−in
forgiveness not do−FUT.3PL
‘And coming generations will not forgive us.’ (T2010)
16.6.5
Perfective
The perfective verb form can be used on its own, mostly to describe
situations that had already happened at the time of speaking and that
the speaker participated in or witnessed. Prototypically, these situations
are expected to have taken place more recently than situations described
by the indirect past.
(117) hila
mãã−∅
zaadi soor
kur−op
now 3SG.POSS−M heart cold
do−PFV.2SG
‘Now you have made my heart cold.’ [‘stilled my desire’] (T0041)
16.6.6
Indirect past
The indirect past can be used on its own, or together with the past tense
marker taa (16.5) The addition of taa does not appear to have much
effect on the meaning. See 16.5 for a discussion of this.
In either case, the expression is used to convey both evidential and
temporal meanings. According to the speakers’ intuitions about its use,
it sometimes signals that the events described were not witnessed by the
speaker but are hearsay or second-hand reports, as in example 118. In
other cases, it is used with respect to witnessed events that took place
longer ago, as in example 119.
(118) tanii
yee kram doos
kur−tʰee
3SG.ANIM.DIST.ERG this work yesterday do−INDIRPST.3SG
‘He did this work yesterday.’ [The speaker did not witness this.] (E0054)
(119) ay
aaċi−tʰum
1SG.NOM come−INDIRPST.1SG
‘I came.’ [Presumably a long time ago.] (E0053)
This is an area that will require further research. There are cases such as
the one illustrated in example 120 that seem to contradict both
157
Tense, aspect and related categories
alternatives, since the indirect past is used, in the first person, when
describing events that took place yesterday.
(120) doos
muu
c̣ʰiir brikin−tʰum
taa
yesterday 1SG.ERG milk sell−INDIRPST.1SG PST
‘Yesterday I sold the milk.’ (Q0053)
This may mean that the distance-in-time is relative so that yesterday in
some cases may be sufficiently long ago to warrant the use of the form.
More probable is that there are more factors complicating the use of the
form, such as considerations about narrative foreground and
background.
16.6.7
Potential past
The potential past can be used on its own, and then describes an event
in the past that the speaker reports on, without wanting to commit to
the truth value of the proposition, as in examples 121 and 122.
(121) taanu
mulk−a
ki ? ṭuṭu
POSS.REFL country−OBL to ? request
ya
aaċi−tʰaa
see
pra−i
ɡiaaċi−tʰin
give−CP bring−POTPST.3PL
bre−es
aaċ−i
or
come−INDIRPST.3SG.M 3SG.DIST brother−KIN.3 come−CP
‘To their own land, perhaps they brought him by asking [him], or he
came [on his own], this brother had come.’ (T3023)
(122) see
kookiy−aal
ye−tʰiyo
3SG.DIST.NOM sleep−PRSPTCP go−POTPST.3SG
‘He may have gone to sleep.’ (E0015)
16.6.8
Continuous past
The continuous past is formed by combining the imperfective form of
the verb with the past tense particle taa and is used to describe events of
a continuous nature that took place in the past.
158
TAM constructions
(123) preešbãĩ
before
sapun c̣atral ta
maaṣ maɡool ware aċareti
all
Chitral TOPSM people Kho
other Ashreti
lee mas−a
peeɡoor
very 3SG.ANIM.PROX.OBL−LOC taunt
pre−nun
taa aruniaa
give−IMPFV.3PL PST Gawar people
‘Before, all the people from Chitral, the Kho and others, the Ashreti
(Palulo) taunted them a lot, the Gawar people [as well].’ (TW5031)
Many occurrences of the continuous past are instances of the animate
copula tʰ ‘be’ (see 11.7.1), that are used as equivalents of perfective forms
that are missing in the paradigm of this root.
16.6.9
Habitual past
The habitual past is formed by combining the present participle with the
indirect past form of the verb b ‘become’, which function as suppletive
forms of the animate copula, and describes habitual acts or repeated
events in the past, as in example 124.
(124) see
tʰaan−a
paš tʰ−aw−aal
bai−tʰun
3SG.DIST place−LOC trap be−CAUS2−PRSPTCP be−INDIRPST.3PL
ɡoora ki čoẉweela ki
Gora for Chuwela for
‘In that place they were putting traps, for Goras and Chuwelas
[birds].’ (T3080)
16.6.10 Inchoative past
The inchoative participle can be used with the copula to convey the sense
of an action being initiated or an event starting to take place.
(125) taanu
ṣaa lum aċap−aai−i
laak−em
b−ui
POSS.REFL head hair pull.out−CAUS−CP weep−INCHPTCP be−PFV.3SG.F
‘Pulling her hair, she started weeping.’ (T1016)
The examples of this form are scanty in my material, and more research
will be needed to provide a more detailed description.
159
Tense, aspect and related categories
16.6.11 Imperative
The imperative form is used, on its own, to form imperative clauses
(20.2). The subject is always implicitly the second person, so there are
usually no overt subjects, but objects are typically expressed if the verb
is transitive. Example 126 shows two subsequent imperative clauses: the
first with an inherently transitive verb; the second with a causative.
(126) tayaar kur−e
šumaa kur−i pač−a−i
ready do−IMP.SG beautiful do−CP cook−CAUS−IMP.SG
‘Make it ready, make it pretty and cook it.’ (T9049)
In negative imperatives, such as when forbidding or warning someone,
a particular negator or prohibitive marker, ma ‘don’t’, is used before the
imperative verb form, as in example 127.
(127) ware sãã−i
dašt−ee žan ma žan−e
other 3SG.ANIM.POSS−F hand−INS snake don’t kill−IMP.SG
‘Don’t kill a snake with someone else’s hand.’ (TI0012)
160
17
Adverbials
Adverbials are modifiers, usually of verbs or entire clauses, and can be
expressed with words from several different classes. A separate adverb
class would be relatively small, and the case for distinguishing it rather
weak, but some likely candidates are discussed in the chapter on minor
or questionable word classes (13.1). Most adverbials are formed with
either nouns, particularly in the locative (7.4.2.4) or instrumental case
(7.4.2.5), with postposition phrases or with non-finite subordinate
clauses 19.4). Adjectives can often modify verbs or clauses as well as
nouns, and some non-finite verb forms can fulfil similar functions.
Example 128 shows a locative noun functioning as an adverbial, and
example 129 two postposition phrases performing the same function.
(128) ay
baazar−a
c̣ʰiir brikin−numa
1PL.NOM bazaar−LOC milk sell−IMPFV.1PL
‘We are selling milk in the bazaar.’ (Q0044)
(129) dir ta
ay
širiŋɡal−bãĩ
rawan b−um
Dir from 1SG.NOM Shiringal−towards start become−PFV.1SG
‘From Dir I started towards Shiringal.’ (TW7002)
Adverbials describe a variety of different meanings. In Dameli,
adverbials describing time, location, and movement in space are
particularly relevant, but these are by no means the only kinds of
adverbial. Of particular interest, because they form relatively coherent
subsystems with restricted but extensive dimensions, are the time
adverbials, spatial adverbials and intensifiers.
17.1
Time adverbials
Time adverbials are interesting from a semantic point of view, since they
deal with a relatively delimited field, with clearer dimensions than most
words. In particular, there is a set of adverbial expressions used to talk
161
Adverbials
about particular days that ranges over up to four days before and after
the present day. The expressions used involve both nouns and noun
phrases, postposition phrases and adverbs. Simple, single-word
expressions are used for ‘today’ and ‘yesterday’, with increasingly
complex constructions used as the division between the moment of
speaking and the time referred to by the expression becomes greater.
Table 43: Tomorrow, yesterday and beyond
seek čooṭ diyoo
čooṭ diyoo
itrii
doos
mudya
beraa ki (also ɡurma ki ‘in the morning’)
truida ki
čooṭa ki
seek čooṭa ki
four days ago
three days ago
day before yesterday
yesterday
today
tomorrow
on the day after tomorrow
in three days
four days from now
Other important temporal adverbials are žaa ‘now’, kya waxta ‘at the time
when’, and yee ta baat/aaṣanta ‘after this’.
17.2
Spatial adverbials
The interaction between language and geography make spatial
adverbials a particularly fascinating field in Dameli and among the
languages surrounding it. 57
Unfortunately, I have not been able to study this as closely as the
complexity of the subject requires. Pending further research, several
points must remain tentative, and I am convinced that a more thorough
investigation would render a much more complex picture.
In Dameli, adverbials that express location and direction often form
antonym pairs that express a variety of contrasts. The most general pair
is the adverbs ayaa ‘here’ and tara ‘there’. The adverb oor ‘near’ is more
specific, typically denoting a location closer to or on the same side as the
speaker. The opposite is pʰar ‘far’, which is used for things nearer to
someone else or for things on the opposite side, typically used with
respect to a river or valley.
57
See e.g., Heegård Petersen (2006) on Kalasha, particularly section 12.2.8.1, and Bashir
(2000) for Khowar.
162
Intensifiers
In the vertical direction, the general pair is puču ‘up’ and bun ‘down’,
whereas the more specific ṣin ‘up’ and baar ‘down’ are used in relation to
locations higher or lower on a slope. Not all contrasts can be neatly
arranged in antonym pairs, however. For example, although bun ‘down’
is said to be the opposite of puču ‘up’, it often contrasts with ṣin ‘up on a
slope’ or ralee/rala ‘up’ as well.
17.3
Intensifiers
The general intensifier in Dameli is lee ‘very/many’, a word that can be
used both as an intensifier ‘very’ and a quantifier ‘many’, and can modify
verbs, adjectives, and nouns (in which case it functions an attribute
rather than an adverbial).
A particular group of verbs, the colour adjectives, have specific
“lexicalised intensifiers” (Liljegren, 2008, p. 160) that are used to signify
a “pure” or “true” colour. These are shown in table 44.
Table 44: Colour intensifiers
Intensifier
ṭãu ‘pure’
pak ‘pure’
šãũ ‘true’
Used with
krinaa ‘black’
laic̣ʰaa ‘red’
ɡoora ‘white’
aarida ‘yellow’
niila ‘blue’
Gloss
pure black
pure red
pure white
pure yellow
real blue
163
18
Coordination
Coordination, according to Haspelmath (2007), “refers to syntactic
constructions in which two or more units of the same type are combined
into a larger unit and still have the same semantic relations with other
surrounding elements” (p. 1).
In Dameli, such constructions are usually formed with conjunctions,
or coordinators in Haspelmath’s terminology − words that are used to
mark coordinating constructions. There are several types of coordinators
and also constructions in which no coordinator is used. The units that
are coordinated are called coordinands (Haspelmath, 2007, p. 2).
Some constructions connect units of more or less similar type without
being coordination in the strictest sense. See for example the section on
clause chaining (21.9), tail-head linkage (21.8), subordinate clauses (19),
and the clitic −es ‘also’ (21.2).
18.1
Natural coordination
Natural coordination is a special kind of conjunctive coordination,
which only applies to certain items, namely “coordination of items
which are expected to co-occur, which are closely related in meaning,
and which form conceptual units” (Wälchli, 2005, p. 5),
Natural coordination in Dameli can be expressed either as simple
juxtaposition, without any coordinator, or with the clitic −o ‘and’. Both
constructions are applied to words only, not to larger phrases. When
juxtaposition, or asyndetic coordination (Haspelmath, 2007, p. 7), is
used, the coordinands are simply enumerated after each other. This can
be applied to two or more words. Table 45 shows some examples of
natural coordination with juxtaposition. The examples include reversed
perspective antonyms and taxonomic sisters, particularly among kinship
terms.
164
Natural coordination
Table 45: Natural coordination with juxtaposition
Example
yei bap
yii dadi braa pas
yen aaċan
ɡẉa ṭooki
Meaning
grandmother and grandfather
mother, father, brother and sister
going and coming
fields and plots
The coordinator −o ‘and’ is postpositive; it is attached at the end of the
first coordinand, in this case the first word of the coordinated word pair.
Table 46 shows some examples of word pairs coordinated with −o.
Table 46: Coordination with −o
Example
aan−o baraan
baaṭ−o kʰaṭ
žuwaar−o ɡoom
aaru−o draak
abut−o kʰača
punaɡraamo aspar−a
mas−o tas−a
šuŋɡaar−o zin−a ki
Meaning
inside and outside
stones and wooden beams
maize and wheat
peaches and grapes
good and bad
in Punagram and Aspar
this and that
in summer and winter
Most pairs coordinated with −o clearly express natural coordination, but
there may be exceptions, signs that the construction is used more widely,
as some pairs do not appear to form a conceptual unit. In the context
where it appears, the pair aaru−o draak ‘peaches and grapes’ only refers to
two fruits that happen to ripen at the same time. There connection
between the two is arguably not very close, and the expression itself does
not signify anything more general. Even less conceptual are pairs of
personal names or titles, which also occur with −o.
Numerals are also coordinated with −o, when creating complex
numbers (bišiy−o−traa ‘twentythree’), when saying ‘and a half’
(duu−o−kʰana ‘two and a half’), and when expressing approximate
numbers (pããc−o čoor ‘four−five’, lit. ‘five and four’).
165
Coordination
18.2 Conjunctive coordination with ãã
‘and’
Conjunctive coordination is constructed with the coordinator ãã ‘and’.
Almost any constituents, from single words or phrases to full clauses,
can be coordinated in this way (example 130 illustrates both).
(130) tas mač sãã−Ø
put baloo
that man 3SG.ANIM.POSS−M son big
ek
kaabil
one capable
b−i
become−CP
ãã baaṣaɡui
mač žup−aa
and well_spoken man make−PFV.3SG.M
ãã
see
ɡram ta maaṣ−e
tasãã−Ø
and 3SG.DIST village of people−ERG 3SG.DIST.POSS−M
žeṣṭaɡari
prai−tʰen
chieftainship give−INDIRPST.3PL
‘When the son of that man had grown up, he turned into a
capable and well-spoken man and the people of the village made
him chief.’ (T4007)
18.3 Disjunctive coordination with kuu
and ya ‘or’
Examples of disjunctive coordination, i.e., coordination with ‘or’, are
few and far between in my material. There are two disjunctive
coordinators, kuu ‘or’ and the (probably) borrowed ya ‘or’ (< Ur. ya).
Examples 131 and 132 illustrate their use. In general, ya is more common
than kuu in my material.
(131) tu
kur−isan
suwal−una ta
2SG.NOM do−PSTPTCP question−PL of
say
right
ǰawaab
answer
daro
be.INANIM.IMPFV.3
kuu
ni
ček
kur−i muu ki
kʰuṇḍ−e
or
not
check
do−CP 1SG to
tell−IMP.SG
‘Check if the answers to the questions you have made are
good or not and tell me.’ D3122
166
Adversative coordination with xu and leekin ‘but’
(132) yee
kya
afsana ya kahani ni daro
3SG.INANIM.PROX which legend or story not be.INANIM.IMPFV.3
‘This is not a legend or a story.’ (TW5001)
18.4 Adversative coordination with xu
and leekin ‘but’
Dameli uses two adversative coordinators, or “but-coordinators”: leekin
and xu. Interestingly, both these forms are clearly borrowed. They are
used in different positions, but the meaning is roughly the same: they are
ways to signal that a constituent somehow contradicts expectations that
the listener might have. Both of them typically coordinate clauses or verb
phrases.
The coordinator leekin ‘but’ (<Ur. lekin) is used prepositively, at the
beginning of the second coordinand (see example 133).
(133) mãã−Ø
iŋɡliš
ta baaraa fikir
ni kur−an ta
lee
1SG.POSS−M English of about thought not do−INF TOPSM very
‘Don’t worry about my English, thank you very much.’
šukriya leekin ay
lee koošiš
kun−um
thank but
1SG.NOM very attempt do−IMPFV.1SG
tee mãã−Ø
iŋɡliš
war−es
abut bo
that 1SG.POSS−M English other−also good become−FUT.3SG
‘But I am trying a lot so that my English will become good.’ (D2067)
The other coordinator, xu ‘but’ (<Psht. χu), can be used both
prepositively, like leekin (see example 134), or postpositively, by placing
it after the first word of the second coordinand. (The second occurrence
of xu in example 135 illustrates this.) This follows the use of χu in Pashto.
(134) too
ki pata
daro
2SG.OBL to knowledge be.INANIM.IMPFV.3
tee ay
lee sax
that 1SG.NOM very hard
masruf tʰi−num
xu koošiš
kur−i ṭeem nikʰaal−num
busy
be−IMPFV.1SG but attempt do−CP time take.out−IMPFV.1SG
‘You know that I am very busy, but I am trying to make the time.’
(D4009)
167
Coordination
(135) awal xu mãã−Ø
imayl aḍres
ni tʰaa
taa
first but 1SG.POSS−M email address not be.IMPFV.3SG.M PAST
žaa xu mãã−Ø
imayl aḍres−es
daro
now but 1SG.POSS−M email address−also be.INANIM.IMPFV.3
‘At first, though, I did not have an email address, but now I have an
email address too.’ (D3011)
Example 135 also illustrates another point. Both leekin and xu can be
used in situations that do not involve coordination in the proper sense,
where there aren’t necessarily two items that can be joined together. In
these cases, they tend to express the sense of contradicting the listener’s
expectations, without assuming that those expectations are the result of
the previous clause or phrase, similar to English “however” or “though”.
The first occurrence of xu in example 135 illustrates this.
168
19
Subordination
The concept of subordination is a controversial subject, that I will not
attempt to address in any depth here. For the purposes of this
investigation, any clause that fills a syntactic function within another
clause is considered a subordinate clause. 58
In terms of form, there are few elements distinguishing subordinate
clauses from main clauses, unless a subordinator or a non-finite verb
form is used. Notably, there is no difference in word order between main
clauses and subordinate clauses. Questions and commands can be
subordinated just as well as declarative clauses, and like these, they
maintain the same word order and other characteristics.
From a language-internal perspective, the distinction between finite
and non-finite clauses is arguably more important than that between
main clauses and subordinate clauses. Finite clauses, clauses that are
formed with a finite verb form, can function as both main clauses and
subordinate clauses, without difference in form, although they can also
contain a subordinator (subordinating conjunction). Non-finite clauses,
clauses that are formed with a non-finite verb form, are always
subordinated, but never contain subordinators.
This chapter begins with a discussion of finite and non-finite
subordinate clauses, and continues with four sections describing
different functional categories.
19.1
Finite subordinate clauses
Finite subordinate clauses are formed in the same way as main clauses,
although they are sometimes initiated by a subordinator. There are no
other differences in form.
The main subordinator, or subordinate conjunction, is tee ‘that’.
Although tee belongs to the subordinate clause syntactically, it is often
58
Of course, what it means to “fill a syntactic function within another clause” is almost
as controversial.
169
Subordination
grouped with the preceding superordinate clause prosodically. Example
136 shows a typical use of tee, at the beginning of a subordinate clause.
bin−um
tee too
naɡi prat−ep
(136) muu
1SG.ERG see−PFV.1SG
that 2SG.ERG fall give−PFV.2SG
‘I saw that you fell down.’ E0093
There are a number of special subordinators, that indicate the semantic
relationship of the subordinate clause to the superordinate clause it
belongs to. Causes or reasons can be indicated with the subordinator
kuitee ‘because’. This probably arose as a combination of ku ‘why’ and
tee ‘that’. Example 137 shows the use of kuitee.
(137) yee
abut soč
kur−een
3SG.INANIM.PROX good thought do−PFV.3PL
ɡ−aa
go−PFV.3SG.M
ba
ek muškil kram daro
TOPSH one difficult work be.INANIM.IMPFV.3
kuitee yee
ek liṭreyčer ta kram daro
because 3SG.INANIM.PROX one literature of work be.IMPFV.INANIM.3
‘This, if you think well about it, is a difficult work, because this is a
work of literature.’ T2004-T2005
Risks and things to be avoided can be expressed with the subordinator
nitee ‘lest’, which probably arose as a combination of ni ‘not’ and tee
‘that’. The idea conveyed by this is that the subordinate clause describes
an undesirable outcome that might ensue if the instructions of the main
clause are not heeded, as in example 138.
(138) ware ki ma kʰuṣ−e
nitee tu
naɡi pr−es
other for PROH dig−IMP.SG lest 2SG.NOM fall give−FUT.2SG
‘Don’t dig for another, so you yourself won’t fall.’ (T0042)
Another special subordinator is aɡar ‘if’, which is used to form
conditional clauses (see 19.3).
Although subordinate clauses are identical to main clauses in terms of
word order, the subordinate clauses themselves are treated differently in
sentences than other elements that fulfil similar functions are. Objects,
for example, are usually placed before the finite verb, but when a
subordinate clause is used as an object, it is placed after the finite verb,
as in example 136 above.
170
Non-finite subordinate clauses
19.2
Non-finite subordinate clauses
Non-finite verbs never form main clauses; non-finite clauses are always
subordinated and they are usually more reduced than finite clauses, with
the most common type consisting of just the verb itself, although there
are also non-finite clauses that include objects and adverbials. Overt
subjects are not included in non-finite clauses.
Non-finite verb clauses straddle the border between syntax and
morphology, since other elements of the clause, such as objects and
adverbials, are sometimes incorporated into the verb itself. This is most
common with present participles in attributive functions, as in example
139, in which the proper noun malu ‘a place in the Domel Valley’ 59 forms
a part of the verb form. In the corresponding non-finite clause, e.g., “She
goes to Malu”, malu would fill the function of an adverbial.
(139) see
malu−y−aal−i
brei kaa
ɡ−ai
3SG.DIST Malu−go−PRSPTCP−F girl where go−PFV.3SG.F
‘Where did that Malu-going girl go?’ (E0012)
These verb forms, even when they incorporate large parts of a clause into
the word, can also be accompanied by independent words from the
underlying clause they represent, as in example 140 below, where the
adverbial muṭa ‘in trees’ is an independent word. This argues in favour
of interpreting these forms as complete clauses, even when they consist
of a single word.
(140) see
muṭ−a
ral−b−aal−i
brei
3SG.DIST tree−LOC up−become−PRSPTCP−F girl
kaa
ɡ−ai
where go−PFV.3SG.F
‘Where did that tree-climbing girl go?’ (E0013)
Clauses headed by a conjunctive participle (11.6.5) often serve to
connect a sentence in a narrative to the preceding one (see 21.8), placing
the main clause in time, in relation to other described events (as in “after
doing this, they did that”). Clauses headed by a past participle (11.6.4)
can be used in similar ways, as subordinate clauses embedded in the main
clause, also with clearly temporal meanings, although the focus is less on
59
Incidentally, Malu is where the photo on the cover of this work was taken.
171
Subordination
the sequence of events, and more on the result of the event described,
when compared to conjunctive participle clauses.
The infinitive (11.6.2), being a verbal noun, is most commonly used
either in complementation or with postpositions.
No instances of the inchoative participle (11.6.6) functioning as
anything else than part of a complex verb expression have been found in
my material.
19.3
Complementation
Complementation, wherein an entire clause functions as the argument of
the verb in the main clause, is carried out primarily by non-finite clauses
formed with infinitives and by finite clauses formed with the
subordinator tee ‘that’. Since arguments of a clause are normally roles
filled by nominal word classes, it is not surprising that the infinitive,
being a verbal noun, is used in this way. That this is not simply a matter
of derivation, whereby the verb is turned into a noun, in a regular noun
function, is shown by the fact that it is not just the verb that is used.
Objects, adverbials and other parts of the clause can be included, as
shown in example 141, where the object imtihan ‘test’ is included in the
subordinate clause headed by the infinitive ɡen ‘take’.
(141) …
…
dees−ee
tasuu
ta
father.KIN3−ERG 3PL.ANIM.DIST.OBL from
imtihan ɡ−en
ɡaṭ−ee
test
take−INF want−DIRPST.3SG
“… the father wanted to take a test from them.” (TW3007)
Verbs such as “see” and “say” often require a clause rather than a noun
phrase as an object. These are typically formed as finite verb clauses
formed with the subordinator tee ‘that’, as in example 142.
(142) manee dac̣i−na
tee see
tasãã−i
žami tʰui
like.this look−PRS.3SG.M that 3SG.DIST 3SG.DIST.POSS−F wife be.PRS.3SG.F
‘Thus he sees that it is his wife.’ (TW1033)
172
Complementation
19.3.1
Conditional clauses
Conditional clauses are a particular kind of complementation and are
formed by finite clauses that function as conditions for a second clause,
much like the English “if…then” construction. These clauses are marked
by a contrastive intonation pattern, rising on the first clause and falling
on the second, and by the contrastive particle ba at the end of the
conditional clause. Example 143 shows a typical conditional clause.
(143) tu
muu
ki ṭelefun
kur−an ɡaṭi−nap
ba
2SG.NOM 1SG.OBL to telephone do−INF want−IMPFV.2SG TOPSH
‘[If] you want to call me,’
beera
reeṭ−a
yaaš
baǰa
pakistani ṭeem−a
kur−e
tomorrow night−LOC eleven o’clock Pakistani time−LOC do−IMP.SG
‘do it tomorrow night at eleven o’clock, Pakistani time.’ (E0104)
The main verb in conditional clauses is restricted to only two verb forms,
which are determined by the aspectual meaning of the clause. Conditions
that involve telic acts, such as doing or understanding something, are
composed with the perfective, whereas conditions that involve atelic
acts, such as wanting or being something, are composed with the
imperfective.
In some examples, a conjunction, aɡar or aɡarka ‘if’, is used, as in
example 144. It seems likely that this construction represents influence
from other languages, particularly Urdu (cf. Ur: aɡar ‘if’ 60).
The form aɡarka ‘if’ is quite strange, but may be a combination of Ur. aɡar ‘if’ and
either Psht ka ‘if’ or another Urdu word.
60
173
Subordination
(144) aɡar taanu
ɡram žup−an
ɡaṭi−numa
ba
if
POSS.REFL village make−INF want−IMPFV.1PL TOPSH
‘If we want to make [save] our village,’
mas
mač zaruur ɡiaaċ−ima
3SG.ANIM.PROX man must bring−FUT.1PL
‘we must bring this man.’ (T4037)
19.4
Adverbial clauses
Adverbial clauses are subordinate clauses that function as adverbials, in
that they modify the verb or the entire clause. These are often formed
with the conjunctive participle (11.6.5), as in example 145.
(145) mulaa baaitaa−i
amun−a nikaa kur−aa
mullah summon−CP our−M nika do−IMP.PL
‘After calling the mullah, perform our nika [wedding ceremony]!’
(T1045)
Adverbial clauses with conjunctive participles are used to form clause
chains (21.9), where they often function more like main clauses than
typical adverbials pragmatically. Tail-head linkage (21.8) is also a kind
of adverbial clause, and can be formed both with the conjunctive
participle and with finite subordinate clauses and the topic shift particle
ba.
Non-finite clauses with infinitives often form part of postposition
phrases, which in turn function as adverbials. While these are not
adverbial clauses in the strict sense of the term, and the way they are
used does not differ from other postposition phrases, the end result is
very similar to non-finite adverbial clauses both in terms of form and
function. Example 146 illustrates this.
174
Relative clauses
(146) tas
šorunḍa
paai sãã−Ø
mukʰ−a
3SG.ANIM.DIST orphaned boy 3SG.ANIM.POSS-M face-LOC
nees−an
ṣaži taanu
žu−n
go.out−INF for REFL.POSS daughter−KIN.2
mas−a
pr−es
3SG.ANIM.PROX.OBL−LOC
give−FUT.2SG
‘To come before this orphaned boy, you must give him your daughter.’
(T4020)
19.5
Relative clauses
A relative clause, according to Andrews (2007), is “a subordinate clause
which delimits the reference of an NP by specifying the role of the referent
of that NP in the situation described by the [relative clause]” (p. 206).
Finite relative clauses in Dameli are constructed with questionwords/indefinite pronouns (20.1.1). The clause is placed after the
element it describes, and the pronoun takes the place of this element
within the relative clause. It is not uncommon for the head to be repeated
after the relative clause. This is probably a matter of pragmatics: if the
distance between the head of the subject and the predicate becomes too
great, the head is repeated to make understanding easier. Example 147
shows a relative clause describing the subject see ‘he’, which is then
repeated after the relative clause.
(147) see
keeraa ek čoṣṭi
bai−tʰaa
3SG.ANIM.DIST.NOM which one only.child be−INDIRPST.3SG.M
see
lee kʰušala
ãã
zrax
bai−tʰaa
3SG.ANIM.DIST.NOM very intelligent and clever be−INDIRPST.3SG.M
‘He who was an only child, he was very intelligent and clever.’ (T0005)
Non-finite verb forms that function attributively typically form relative
clauses, but are much more similar syntactially to other attributes such
as adjectives. Example 140 above is an example of this.
Just as with adverbials, non-finite subordinate clauses formed with
infinitives often form postposition phrases, which in turn are used as
attributes, in a way that does not differ from other postposition phrases,
but closely resembles relative clauses. Example 148 illustrates this.
175
Subordination
(148) žaa ba
makẉui ɡaaṣṭee muk
now TOPSH monkey like
face
wail−an ta
hide−INF of
zururat kya
daro
necessity which be.INANIM.PRS.3
‘But now, what need of hiding the face like a monkey is there?’ (TV7012)
19.6
Quotatives
A special case of subordinate clauses, rather difficult to classify, is that
of quotatives, clauses that express reported speech or thoughts. These
clauses usually function as complements, but can also fulfil adverbial
functions. Quotatives are often marked with ɡani, a grammaticalised
conjunctive participle form of the verb ɡan (t) ‘say’.
While ɡani can still be used as an ordinary conjunctive participle, it is
also often used sentence−finally, a position normally reserved for the
finite verb. This is made possible by reanalysing sentences like 149,
where the quote is embedded as an adverbial within a main clause, as
sentences like 150, where the quote is treated as an independent clause
with ɡani added to mark it as a quote.
ta
kurei ni tʰ−un
(149) dac̣−i
see−CP TOPSM who not COP−PFV.3PL
ãã tee yee
and that this
tukuri por−isan
daro ɡan−i ootʰ−ina
basket fill−PSTPTCP is
say−CP stop−IMPFV.3SG.M
Having seen, having thought ”There is no-one here, and this
basket is full.”, he stops. (TP0016)
mãã−i
tukuri
(150) mããtẽẽ dac̣i−na
around see−IMPFV.3SG.M 1SG.POSS−F basket
kii
ɡiɡ−een
ɡani
who
take−PFV.3PL
QUOT
’He looks around, thinking (or ‘and thinks’) “Who took my
basket?”’ (TP0038)
In both these examples, ɡani is extended to mean ‘think’ rather than ‘say’.
This semantic extension may also be a sign of a grammaticalisation
process, since it occurs only with the conjunctive participle. The lack of
temporal meaning in example 150 may also be an indication of this.
Since ɡani comes at the end, it is not possible to interpret it as taking
176
Quotatives
place before the next clause in the sentence, as would be expected with
the conjunctive participle.
Another related function is its use in marking something as a name
for something or someone, much like the English ‘called’, as in ‘a man
called Tom’, illustrated in example 151.
(151) aaċ−i
baloo daaš
ɡan−i
ek baaṭ daro
come−CP big
stone
say−CP
one stone COP.INANIM.IMPFV.3
‘Having come there, there is a big stone called the great rock.’ (T3003)
Quotative particles that have developed from words for ‘say’ are
common in Dravidian languages and languages adjacent to the
Dravidian-speaking area, such as Bengali, Marathi and Dakkhini Urdu,
but they are also common among the closer neighbours of Dameli,
including Kalasha, Shina, Palula, Balti, Khowar and Burushaski (Bashir,
1996, pp. 193, 197).
According to Bashir (1996), opinions among researchers differ about
whether this should be considered a result of influence from Dravidian
languages, a development within Sanskrit, or simply a convergence of
independent developments in the various languages (pp. 195-196).
Quotatives developing from ‘say’ are attested in various languages the
world over (Heine & Kuteva, 2002, pp. 267-268).
177
20
Other clause types
Basic declarative clauses can be altered to serve other purposes than just
stating facts, to express other kinds of illocutionary force, using a variety
of strategies. Questions are expressed with the use of intonation,
morphological markers and question-words, imperative clauses by an
imperative verb form, and negation with a special negator.
20.1
Questions
Questions in Dameli come in two main types: content questions (also
known as wh-questions or question-word questions) and polar
questions. In both cases, the word order remains unchanged in relation
to declarative clauses; there is no wh-movement.
20.1.1
Content questions
Question-word questions or content questions are questions about
specific information, formulated with a question-word (13.4). To form
content questions, one of the constituents of a clause is replaced by a
question-word, without changing the word order. The intonation
pattern used in content questions places focus on the question-word,
similar to other clauses in which focus is put on a particular element.
Question-words are a limited group of words with several
characteristics in common. Most notably, they all begin with /k/. Since
these words are used to ask questions about various parts of a clause,
they belong to different word classes, although they share many features.
In this way, they are similar to the wh-words of English and their
equivalents in other Indo-European languages to which they are most
certainly related. Table 47 is a list of the question-words in Dameli.
178
Questions
Table 47: Question-words
Question-word
kii
kuree
keeraa
kasãã
kya
keer
kanuu
ku
kaa
kutaal
kati
Meaning
who
who
which
whose
kya
when
how
why
where (general)
where (direction)
how many
Word class
pronoun
pronoun
pronoun
pronoun (possessive)
kya
adverb
adverb
adverb
adverb
adverb
adjective
Question-words are not limited to questions but are also used as
indefinite pronouns when forming relative clauses (19.3).
20.1.2
Polar questions
Polar questions, questions that are expected to be answered with yes or
no, are formed like ordinary statements but are marked with the
question clitic −i (21.3), which is placed on the last regular element of the
sentence, whether this is the verb itself or the past tense marker. Example
152 illustrates this.
(152) tu
kul−a
tʰop−i
2SG.NOM house−LOC be.IMPFV.2SG−Q
‘Are you at home?’ D1019
20.1.3
Rhetorical questions
Questions are often used as narrative devices in Dameli, as ways to
present new information or to stress a certain point. Although these are
syntactically complete questions, they have a very different pragmatic
function. The speaker often immediately fills in the answer, separating it
with the shift-topic particle ba. There is an uninterrupted prosodic
pattern covering both question and answer, as the tone rises on the
question, culminating on the answer, and falls toward the end. Example
153 shows a question–answer structure used to introduce a new subject.
179
Other clause types
(153) daaman ta
Domel of
asli
paidawaar keeraa daru
original production which COP.INANIM.IMPFV.3
ba
zaŋɡalat daru
forestry is
‘The real produce of Domel, what is it? It is forestry.’ (T1053)
TOPSH
20.2
Imperative clauses
Imperative clauses express instructions and commands, and are formed
with imperative verb forms (11.5.6), which are restricted to referring to
the second person, and only distinguish between singular and plural.
Example 154 shows an imperative clause.
(154) mulaa baaitaa−i
amun−a nikaa kur−aa
mullah summon−CP our−M nika do−IMP.PL
‘Summon the mullah and hold our wedding.’
[Lit. ‘Having summoned the mullah, hold our wedding.’] (T1045)
Imperative clauses usually contain no overt subjects, although there are
exceptions, as illustrated in example 155.
(155) tu
dac̣−e
ɡale see
kook−isan
tʰaa−i
2SG.NOM see−IMP.SG ?
3SG.DIST sleep−PSTPTCP be.IMPFV.3SG.M−Q
‘You look, is he sleeping?’ (E0014)
Negative imperatives, or prohibitives, expressions telling someone not to
do something, are formed in the same way as other imperatives but with
the prohibitive marker ma ‘don’t’, as illustrated in example 156. The
prohibitive marker is placed immediately before the finite verb, in the
same way as the negator ni ‘not’.
180
Negation
(156) ware sãã−i
dašt−ee
žan
ma
žan−e
other 3SG.ANIM.POSS−F hand−INS snake PROH kill−IMP.SG
‘Don’t kill a snake with someone else’s hand!’ (TI0012)
20.3
Negation
The negator ni ‘not’ can be used to negate any clause except an
imperative. It is normally placed just before the verb. The clause
undergoes no other changes. Example 157 shows a negated clause.
(157) mãã−Ø
putr−oo too
ni laaki−num
1SG.POSS−M son−VOC 2SG.OBL not weep−IMPFV.1SG
‘My son, I am not crying for you.’ (T1021)
The position of the negator is relatively fixed and therefore not available
as a device for expressing the scope of negation.
The negator can also be used with the topic shift marker ba and the
topic marker ta to express two or more negated clauses, similar to the
English “neither … nor …” construction. In these cases, the negator is
placed before the discourse particle, at the beginning of each clause, as
in example 158.
(158) daamia baaṣa
mudiya diyoo talii
Dameli language today day until
ni
not
ta
TOPSM
kii
who
nat
nat
prai−tʰen
give−INDIRPST.3PL
ni
ba
kya
nat žup−aai−tʰen
not
TOPSH
which nat make−CAUS−INDIRPST.3PL
‘Until today, no one has sung nats (religious songs) in the
Dameli language, nor have they made any nats.’ (TV0001)
Additional clauses can be added at the end of this, each initiated by ni
ba. Example 159 is a continuation of the utterance in example 158.
181
Other clause types
(159) ni ba
kii
yee daamia baaṣa
nat
not TOPSH who this Dameli language nat
lik−an
ta koošiš
ku−tʰen
write−INF of attempt do−INDIRPST.3PL
‘Nor has anyone tried to write down a nat in Dameli.’ (TV0002)
Prohibitives, commands or instructions not to do something, are formed
with a special marker, ma ‘don’t’. See 20.2.
182
21
Some discourse features
It would not be feasible to give a full description of everything that
occurs in Dameli discourse in this dissertation, but some elements of the
grammar that include pragmatic or discourse-level features have been
included in this chapter. These are features of language that primarily
concern the relation between the sentence they are found in and its
context, either narrative or social. These features include single words or
suffixes, and more complex constructions.
In the first category are the ubiquitous topic-marking particles ta and
ba, the intriguing echo-word formation, fillers, the clitic −es ‘also’, and
the vocative particle. The question marker clitic −i also falls into this
category.
The second category comprises the “afterthought” construction, tailhead linkage and clause chaining.
Naturally, many other features that have been described elsewhere
also perform discourse-related functions. This is particularly true for the
different clause types (chapter 20), some elements of coordination
(chapter 18), and some forms of the kinship terms (7.5).
21.1
The topic particles ta and ba
The particles ta 61 ‘previously mentioned topic’ and ba ‘new topic’ or
‘topic shift’, are among the most common words of the language, each
at least twice as common as, e.g., ãã ‘and’, in my material. They both fill
similar but contrastive, functions crucial to the information structure of
Dameli clauses.
Common to both is the function of delimiting different sections of the
clause, essentially splitting a clause into two parts with different
functions. The most typical use is to split the sentence into the subject
and the predicate.
Note that the particle ta is homonymous with several other words, such as the
postposition ta ‘from, of, than’ (12.2.1).
61
183
Some discourse features
The first position in the clause, preceding the particle, is thus marked
as a topic position. Other arguments of the verb or adverbials can also
be placed in the topic position, in order to mark them as the topic of the
clause. Consequently, whatever comes after the particle is marked as
comment, as additional information being ascribed to the topic.
Where the two particles differ is in the relationship to previous
clauses. Whereas ba is usually used when there is a switch in topic, when
the subject is different from the subject of the preceding clause (as in
example 160) or when another topic is introduced, ta is usually used
when the subject or topic is the same. Both ba and ta are commonly used
in tail-head linkage (21.8).
(160) ɡram ta zaatak−nam muu
ki peeɡoor pre−nun
village of child−PL
1SG.OBL to taunt
give−IMPFV.3PL
ãã
tu
ba
wail−aai−ap
ni kʰuṇḍi−nap
and 2SG.NOM TOPSH hide−CAUS−IMPFV.2SG not tell−IMPFV.2SG
‘The children of the village are teasing me, and you are hiding
[something] and don’t tell.’ (T0024)
Interestingly, two particles with identical form and similar functions are
also found in neighbouring Palula, despite the fact that the two
languages are not closely related (Liljegren, 2008, pp. 305-306, 377383).
There are some instances of ta appearing first in a sentence. This is
relatively common (30 times in my material), and probably fills the
function of a conjunction, meaning more or less ‘then’. ba never occurs
in this position, but compare also taa ba.
Both ta and ba can be used in a particular construction, which
capitalises on their contrastive meaning to create a sense of opposed
meanings, as illustrated in example 161.
184
Also, −es
(161) tẽẽ
duu maana
ek ta
ḍokṭor
3PL.ANIM.DIST two of_them one part doctor
žup−aa
make−PFV.3SG.M
ãã
and
tasuu
maana
3PL.ANIM.DIST.ACC of_them
ek ba
pulis
one part police
‘Of the two one made himself a doctor, and one of them a police.’
(TW3005).
21.2
Also, −es
The clitic −es means ‘too’ or ‘also’ and can be attached to nouns,
including proper nouns, postpositions, adverbs, adjectives, and probably
other word classes as well. Phonologically, it functions as a part of the
word it is attached to, as in example 162, where the marked word is
pronounced [taˈræs] (see 6.1.3).
(162) tara−es
čoor kom−una tʰun
there−also four tribe−PL be.IMPFV.3PL
‘There are four tribes there as well.’ (TA1034)
In some cases, this functions as a kind of coordination, with −es attached
to the second element that is coordinated. In others, it functions as an
afterthought, adding something to a clause after it has been completed.
Syntactically, it is often a larger unit, such as a phrase, that is modified
by −es, although the clitic is attached to the head of the phrase (or to the
end of the phrase, which is usually the same thing), as in example 163,
where −es is attached to the phrase.
185
Some discourse features
(163) žaa ba
yee ɣam lee baloo daro
now TOPSH this grief very big
is
tee
that
muu−a
tãã
nasar−es ni daro
1SG.OBL−LOC 2SG.POSS sight−also not COP.INANIM.IMPFV.3
‘Now this grief is very great, that your sight as well is not with me.’
(TV3004)
21.3
The question clitic −i
The question clitic −i turns a statement into a polar question (20.1.2), a
question for which the expected answer is either “yes” or “no”. It is
more limited in its distribution than −es and is always placed as a clitic
on the last word of the clause, usually either the verb itself or the past
tense marker taa.
21.4
The vocative particle
A vocative particle, a or e, is used with names or other appellations when
directly addressing someone. Unlike the vocative form of kinship terms
(7.5.3), this particle can be used together with any word designating a
person and is placed directly before that word, as in example 164.
(164) a
bree kaa
yi−nap
VOC girl where go−IMPFV.2SG
‘Girl! Where are you going?’ (E0105)
Morgenstierne (1942) reports form ē (p. 134) but not a. It is unclear what
governs the use of the two forms, but e could be more common in less
relaxed styles, such as songs and poetry, as in example 165 below, which
is from a ghazal, a poem.
186
Fillers
(165) e
toti
tãã−Ø
waš
uštruun−i
VOC parrot 2SG.POSS−M sound hear−CP
watan
y−en
ta iraada
kur−um
country go−INF of intention do−PFV.1SG
‘Oh parrot, hearing your sound I want to go to my country.’ (TV8001)
21.5
Fillers
When a speaker is unsure about how to complete a sentence or needs
some extra time to think, a filler is often inserted. The most common
word in such situations is iŋɡa. No etymology or semantic meaning of
the word has been discovered.
Another common filler is laka ‘like’, which can be used to indicate
inexactness but is often used primarily as a filler. Unlike iŋɡa, laka is
borrowed and widespread in the area, particularly in Pashto.
21.6
Echo-word formation
In Dameli, one often hears rhyming nonsense words in the middle of
ordinary conversation. These form a particular construction that
expresses semantically expanded concepts, in which a real word is
repeated, but the second time the first element is replaced with /m/. The
meaning of the complete expression is usually a broader concept that
includes the meaning of the real word. To exemplify, čay−may ‘tea−ECHO’
might mean “tea and something to eat” or “teabreak”.
Stolz (2008) argues that this phenomenon, which he calls echo-word
formation (p. 109), has spread widely as a result of contact, and that the
source is Turkish, or, possibly Persian. In both these languages, the initial
sound is m−, as it is in both Pashto and Dameli, but there is considerable
variation in this regard on the Indian Subcontinent (e.g., v− in Hindi, š−
in Punjabi and Urdu, at least in Pakistan).
In Dameli, the words that are most commonly echoed are nouns, but
verbs and other word classes occasionally appear in this construction as
well. Like other similar devices, such as diminutives in many languages,
the use is spreading, and echo words might be used with little change in
meaning from the original word, for pragmatic reasons or as a matter of
course.
Example 166 shows a relatively typical use of echo-word formation.
187
Some discourse features
(166) manuu
thus
daru−i
is−Q
tee
that
boṇḍri~moṇḍri
border~ECHO
amun−a
ek kisma ṣaa
POSS.1PL−M one kind
over
kulau daru
yin−numa
aaċu−numa
open COP.INANIM.IMPFV.3 go−IMPFV.1PL come_back−IMPFV.1PL
‘Thus it was that the border and all that was kind of open for us and we
went and came back.’ (T6069)
Words that begin in vowels have the m− added without removing any
part of the original word, e.g., aaṭ−maaṭ ‘flour and such things’.
Consonant clusters have only the first phoneme replaced, as long as the
resulting cluster is a valid syllable-initial cluster, e.g., braa−mraa ‘brothers
and such’.
21.7
Afterthought constructions
The great exception to any rule of word order in Dameli is what we
might call an “afterthought construction”. It is quite common for
elements that should, according to normal word-order rules, appear
somewhere in the middle of a sentence to be added at the end instead, as
though the speaker had not thought of them before speaking or decided
that additional clarification was needed. In many cases, that is probably
precisely what happens. Since the construction is so common, it is likely
that it is also used to emphasise certain elements of a clause, much like
fronting or other kinds of topicalising are used in other languages.
Almost any element can be subject to this construction. Example 167
shows an attribute in the form of a possessive placed at the end, after the
finite verb, instead of just before the noun to which it functions as an
attribute.
(167) i
ba
kʰur duu−i
rux daro masãã−i
3SG.ANIM.PROX TOPSH foot two−COLL well is
3SG.POSS−F
‘But the feet are both well, his.” (T3013)
Example 168 shows how this construction is used to satisfy grammatical
constraints. The example consists of two sentences. Since the referent of
the subject doesn’t change, no explicit subject would normally be used
188
Tail-head linkage
in the second sentence. But because the subject of the first sentence, see
‘he’, is in the wrong case form to serve as a subject of the second
sentence, the speaker probably feels obliged to add the subject tanii ‘he’,
to avoid confusion. Although they refer to the same person, see is
nominative, which is appropriate for the intransitive verb baa ‘became’,
whereas tanii is ergative, which is required by the transitive past tense
form pratʰee ‘gave’.
(168) prambal−aai−i
aan
muu
bin−i
see
burn−CAUS−CP inside 1SG.OBL see−CP 3SG.ANIM.DIST.NOM
ɡabrau
b−aa
frightened become−PFV.3SG.M
aan
luspiker−a
baaŋɡ−es
ni pre−i
inside loudspeaker−LOC prayer_call−also not give−CP
baraan nees−i
baraan baaŋɡ
outside go.out−CP outside prayer_call
pra−tʰee
tanii
give−INDIRPST.3SG 3SG.ANIM.DIST.ERG
‘Having lit [it] he, having seen me inside, became frightened. He did not
give the prayer call inside, through the loudspeakers, [but] went outside
and gave the prayer call, he.’ (TW7020)
21.8
Tail-head linkage
Tail-head linkage is a device used to link together sentences in a narrative
(Coupe, 2006, p. 151) by echoing the verb phrase of a preceding sentence
at the beginning of the next sentence. The whole verb phrase does not
need to be repeated, and the main verb usually changes verb form. Tailhead linkage is very common in Dameli, and throughout the region.
The constructions used to express tail-head linkage are different if the
topic or subject is the same in the two linked sentences or if it is different.
Same-topic linkage is always expressed by non-finite verb forms. The
most common construction utilises the conjunctive participle, as in
example 169. This construction is used to join two sentences with the
same topic in focus, which usually translates to having the same subject.
A conjunctive participle is formed on the same root as the finite verb that
concludes the first sentence and is placed at the beginning of the second
sentence. Since the conjunctive participle, being non-finite, does not
permit a subject, the subject is never repeated in the joining clause, but
189
Some discourse features
other parts of the verb phrase, such as objects and adverbials, may be
carried over along with the verb, as shown in example 169.
(169) ačooṭ−aal
b−aa
ba
neepʰar ta
do_one_by_one−PRSPTCP be−IMPFV.3SG.M TOPSH below from
ek učʰuṭa paai sekal−a
aaċe−na
one young boy bicycle−LOC come−IMPFV.3SG.M
‘While he is picking, a boy comes from below on a bicycle.’
aaċ−i
see
paai tara see
dac̣i−na
come−CP 3SG.DIST boy there 3SG.DIST see−IMPFV.3SG.M
‘Having come, that boy there sees that.’
dac̣−i
ta
…
see−CP
TOPSM …
‘Having seen, …’ (TP0014-16)
Like clause chaining (21.9), this construction is consistent with the core
use of the conjunctive participle, i.e., expressing concluded events, but
has a pragmatically specialised function. Repeating the main verb of the
preceding sentence does not primarily modify the clause by adding new
information but rather serves as a pragmatic device to keep the narrative
together.
Almost as common in this function as the conjunctive participle is a
construction combining the infinitive and one of two words meaning
‘after’, bat (example 170) and aaṣanta (example 171) in a postposition
phrase with ta ‘from’.
(170) …
alim
žup−aa
mullah make−PFV.3SG.M
‘…[he] became a mullah.’
alim
žup−an
ta bat
…
mullah
make−INF
of after …
‘After becoming a mullah…’ (TW3006-7)
190
Tail-head linkage
(171) zaŋɡal−a
traa diyoo traa reet safar
kur−i
forest−LOC three day three night journey do−CP
tẽẽ
lee kʰasar−un
3PL.ANIM.DIST.NOM very tire−PFV.3PL
‘After travelling three days in the forest, they got very tired.’
kʰasar−an ba
kʰasar−an ta aaṣanta …
tire−INF TOPSH tire−INF
of after
…
62
‘Tired, after getting tired…’ (TW8006-7)
Formally, this is a regular postposition phrase, functioning as a clause
adverbial. A similar construction with manora, ‘middle’ or ‘while’, is
used when the events described by the two sentences take place
simultaneously.
(172) suǰai mulk−ee
yee
tʰaana
Sujai Mulk−ERG 3SG.INANIM.PROX place
‘Sujai Mulk captured this place.’
kabza kuree
capture do−3SG.PFV
kuran manora amun−a
bap−suu
…
do−INF while
POSS.1PL−M
grandfather−KIN.PL …
‘While [he was] doing [that], our ancestors…’ (T6043-44)
When sentences with different topics are joined, typically when the
subject shifts, a slightly different construction is used. This construction
uses a finite verb form in the second sentence, and the topic-shift particle
ba (21.1) always follows the verb. The repeated verb can have the same
form as in the preceding sentence, as in example 173; a different finite
form, as in example 174; or can even be expressed with a complex verb
expression, as in example 175. The factors governing the choice of verb
form will require further study.
The repetition “kʰasaran ba” is probably due to hesitation and is not part of the typical
construction.
62
191
Some discourse features
(173) see
ba
baara yede ootʰi−na
3SG.DIST TOPSH away go.CP stop−IMPFV.3SG.M
‘Having gone away, he stops.’
ba
ootʰi−na
ek baara ɡaa
stop−IMPFV.3SG.M TOPSH one away go−PFV.3SG.M
ba
ek baat−a
san
ba−i
see
sekal
TOPSH one stone−LOC touch become−CP 3SG.DIST bicycle
daara yi−na
fall
go−IMPFV.3SG.M
‘Having stopped, and gone away, that bicycle touches a stone and falls.’
(TP0024-25)
(174) daara yede tasãã−Ø
see
ṭaaŋɡu−es
fall
go.CP 3SG.POSS−M 3SG.DIST pear−also
sapun preeṣ biy−ãã
all
spill become−IMPFV.3SG.M
‘Having fallen, those pears of his also all become spilled.’
preeṣ baa
ba
tẽẽ
zaatak−nam
spill become−PFV.3SG.M TOPSH 3PL.ANIM.DIST child−PL
waapas bin−nun
back look−IMPFV.3PL
‘[The pears] having become spilled, those children look back.’
(TP0026−27)
(175) ačooṭ−aal
do_one_by_one−PRSPTCP
b−aa
be−PFV.3SG.M
ba
neepʰar ta
TOPSH below
from
ek
učʰuṭa
paai
sekal−a
aaċe−na
one young
boy
bicycle−LOC
come−IMPFV.3SG.M
‘While he is picking, a boy comes from below on a bicycle.’ (TP0014)
Some speakers occasionally use taaba ‘then’ in place of tail-head linkage
constructions. Although this word probably arose as a
grammaticalisation of the past tense marker taa (16.5) and the subject
shift particle ba (21.1), it is freely used in series of same-subject clauses.
Example 176 shows a typical use of taaba.
192
Clause chaining
(176) taaba buṭo aaɡ−aa
then bhutto come−3SG.M
‘Then Bhutto came.’ T6067
21.9
Clause chaining
A construction known as clause chaining or narrative chaining (Coupe,
2006, p. 146) is used to relate sequences of events happening one after
another. This construction is an extension of the core function of the
conjunctive participle (11.6.5), describing preceding events. Rather than
using a sentence to describe each event, a series of conjunct participle
clauses are stacked in the same sentence, each describing one event in the
sequence. The sentence is concluded by a final clause headed by a finite
verb.
This construction is common in the languages of South, Central and
East Asia, and also in the Semitic languages of Ethiopia (Coupe, 2006,
p. 147). It is different from the ordinary adverbial function of the
conjunctive participle in that it allows clauses headed by a conjunctive
participle to be foregrounded, rather than functioning as mere
modifications of the main verb of the clause. The number of clauses that
can be used after each other in this way is limited only by pragmatic
concerns; as long as the listener can understand what events are being
referred to, the speaker can continue piling event after event on top of
each other. Other parts of the verb phrase referred to, such as objects or
adverbials, may be included in this construction as well, but the subject,
which is not necessarily overtly expressed, is understood to be constant
throughout the entire chain. Example 177 shows a clause chain with two
events expressed with conjunctive participles (drink, find/catch) and a
final event expressed with a finite verb (reach).
(177) aau
pi−i
žeṣṭaŋɡal ta sarax ta pʰan waar−i
water drink−CP Jestangal of road of way catch−CP
aaspar kul
weebi−ãã
Aspar house reach−IMPFV.3SG
‘He drinks water, finds the path of the Jestangal road, and
reaches the house in Aspar.’ (T5033)
193
Final matters
22 Suggestions for further
research
An investigation of this kind is a rather humbling experience, where
every new discovery reveals a disheartening number of new areas crying
out for a deeper investigation, and there is never enough time or material
to fully exhaust any field. The areas mentioned here are some of those I
hope to have the opportunity to investigate to a greater extent in the
future, or that someone else would take up such an investigation. This is
particularly true for the larger fields that I have had to leave out almost
entirely.
22.1
The language of women
The Dameli society is divided along gender lines, but these lines are never
clearer than when a male stranger is involved. As a rule, women do not
interact with men who are not relatives, but in a society where people
are usually surrounded by relatives, who are not subject to the rules of
purdah, this does not need to be a very noticeable element of culture.
For me, however, who was related to no one, the rules were in full force,
and thus, I had no direct access to the language of women. That is not
to say that women and the language of women have had no influence on
this work, only that this influence was by necessity always mediated
through men. Countless times my most difficult questions on grammar,
words and usage was met with the reply “I’m not sure, I must ask the
people in the house.”
It is likely that even a female researcher would meet with cultural
restrictions that make studying the language of women difficult. To
record a woman’s voice, for example, even if it was done by a relative,
was considered too sensitive. Still, a female researcher or a researcher
from within the group (ideally, both), would undoubtedly be able to add
many important observations, insights and perspectives to this work, on
197
Suggestions for further research
language use among women in particular and on the Dameli language in
general.
22.2
Historical linguistics
Dameli belongs to a group of Indo-Aryan languages that have been
isolated and neglected for a long time. Most of the material that is
presented here has not been presented in any form before. What does the
case of Dameli mean for the description and reconstruction of ProtoIndo-Aryan, and, by extension, Proto-Indo-European?
22.3 Related languages and language
contact
A related question concerns the relationship of Dameli to its
neighbouring languages and its later history. What languages are the
closest relatives of Dameli? Are they to be found among the geographical
neighbours of the Dameli people, such as Kalasha or Gawarbati, or
among languages spoken farther away, such as Gawri, as suggested by
Strand (2011)?
Equally interesting are the traces of contact with some of the other
languages in the area, that are probably not evidence of direct
genealogical relationships. Unsurprisingly, there is ample linguistic
evidence of extensive contact with Palula, and similar observations can
be made about Kalasha and the Nuristani languages as well. Studying
the particulars of these relationships would likely generate a great deal
of interesting information.
The relationship between Dameli and the Nuristani languages is of
special interest. The extent of similarity between these languages is so
great that it led Morgenstierne to consider placing Dameli in the
Nuristani family (Morgenstierne, 1942, p. 147).
22.4
Variation within the language
As discussed in 1.5, Dameli is not entirely uniform across the valleys
where it is spoken, to the extent that different dialects can be identified
and described. The present research has not involved enough speakers
198
Tone
from different parts of the Domel Valley or studied them in a manner
systematic enough to determine the extent and characteristics of these
dialects in more than a cursory fashion.
In the same way, social variation is probably widespread, especially
as it pertains to exposure to other languages, but investigating this would
require more data on the language of people who reside within the
Domel Valley and have little contact with the outside, the very people
most difficult to contact and work with.
Some of the questions that remain are what the main dialects of
Dameli are, and what characterises them. What are the main isoglosses,
and how do they follow geographical and societal features?
22.5
Tone
Another issue that has only been touched upon briefly is tone. Dameli
has a system of contrastive tones, but this system will need to be
thoroughly investigated and described. How is the tonal system
constructed, and what functions does it fill?
199
References
Aikhenvald, A. Y. (2003). Evidentiality in typological perspective. In A.
Y. Aikhenvald (Ed.), Studies in Evidentiality (pp. 1-31).
Amsterdam: John Benjamins.
Andrews, A. D. (2007). Relative clauses. In T. Shopen (Ed.), Language
Typology and Syntactic Description (Second edition ed., Vol. 2,
pp. 206-236). Cambridge: Cambridge University Press.
Baart, J. L. (1997). The Sounds and Tones of Kalam Kohistani.
Islamabad: Summer Institute of Linguistics and National
Institute of Pakistan Studies, Quaid-i-Azam University.
Bashir, E. (1988). Topics in Kalasha Syntax: An Areal and Typological
Perspective. Ann Arbor, Michigan: University of Michigan.
Bashir, E. (1996). Mosaic of Tongues: Quotatives and
Complementizers in Northwest Indo-Aryan, Burushaski, and
Balti. In W. L. Hanaway, & W. Heston (Eds.), Studies in
Pakistani Popular Culture (pp. 187-286). Lahore, Pakistan:
Sang-e-Meel Publications.
Bashir, E. (2000). Spatial Representation in Khowar. Chicago
Linguistic Society (CLS 36) (pp. 15-29). Chicago: Chicago
Linguistic Society.
Bashir, E. (2003). Dardic. In G. Cardona, & D. Jain (Eds.), The IndoAryan Languages (pp. 905-990). London: Routledge.
Bashir, E. (2007). Contact-induced change in Khowar. In S. Saeed
(Ed.), New Perspectives on Pakistan: Contexts, Realities and
Visions of the Future (pp. 205-238). Karachi: Oxford
University Press.
Beekes, R. S. (1995). Comparative Indo-European Linguistics.
Amsterdam: John Benjamins.
Bickel, B. (1998). Review: Converbs in cross-linguistic perspective.
Linguistic Typology, 381-397.
Blake, B. J. (1994). Case. Cambridge: Cambridge University Press.
200
References
Cacopardo, A. (2008). The Dameli of Southern Chitral: An
Ethnographic Inquiry. In Israr-ud-Din (Ed.), Proceedings of the
Third International Hindu Kush Cultural Conference (pp. 397420). Karachi: Oxford University Press.
Cacopardo, A. M., & Cacopardo, A. (2001). Gates of Peristan:
History, Religion, and Society in the Hindu Kush. Rome:
Instituto Italiano per l'Africa e l'Oriente.
Clark, J., Yallop, C., & Fletcher, J. (2007). An Introduction to
Phonetics and Phonology (3rd ed.). Oxford: Blackwell
Publishing.
Comrie, B. (2005). Numeral Bases. In M. Haspelmath, M. S. Dryer, D.
Gil, & D. Comrie (Eds.), The World Atlas of Language
Structures (pp. 530-531). Oxford: Oxford University Press.
Corbett, G. (1991). Gender. Cambridge: Cambridge University Press.
Coupe, A. R. (2006). Converbs. In K. Brown (Ed.), Encyclopedia of
Language and Linguistics (pp. 145-152). Oxford: Elsevier.
Croft, W. (2003). Typology and Universals. Cambridge: Cambridge
University Press.
Dahl, Ö. (1985). Tense and Aspect Systems. Bath: Basil Blackwell.
Dahl, Ö., & Koptjevskaja Tamm, M. (2001). Kinship in Grammar. In
I. Baron, M. Herslund, & F. Sørensen (Eds.), Dimensions of
Possession (pp. 201-225). Philadelphia: John Benjamins.
Decker, K. D. (1992). Sociolinguistic Survey of Northern Pakistan,
Volume 5: Languages of Chitral. Islamabad, Pakistan: National
Institute of Pakistan Studies and Summer Institute of
Linguistics.
Department of Linguistics, Max Planck Institute for Evolutionary
Anthropology and Department of Linguistics, University of
Leipzig. (2008, February). The Leipzig Glossing Rules.
Retrieved July 13, 2010, from
http://www.eva.mpg.de/lingua/resources/glossing-rules.php
Dixon, R. M. (1994). Ergativity. Cambridge: Cambridge University
Press.
Elnazarov, H. (2010). Dameli. In C. Moseley (Ed.), Atlas of the
World's Languages in Danger (3rd ed.). Paris: Unesco
Publishing. Retrieved 10 24, 2012, from
http://www.unesco.org/culture/en/endangeredlanguages/atlas
201
References
Fortson, B. (2006). Proto-Indo-European Syntax. In K. Brown (Ed.),
Encyclopedia of Language & Linguistics (pp. 228-235).
Oxford: Elsevier.
Greenberg, J. H. (1966). Some Universals of Grammar with Particular
Reference to the Order of Meaningful Elements. In J. H.
Greenberg (Ed.), Universals of Language (Second Edition ed.,
pp. 73-113). Cambridge, Massachusetts: M.I.T. Press.
Grierson, G. (1919). Indo-Aryan Family, North-Western Group,
Dardic or Pisacha Languages (including Kashmiri) (Vol. VIII).
Delhi: Motilal Banarsidass.
Grjunberg, A. (1999). Dameli Yazik. In Dardskie i Nuristanskie Yaziki.
Moscow: Институт языкознания РАН,Российская академия
наук.
Haig, G. (2008). Alignment Change in Iranian Languages: A
Construction Grammar Approach. Berlin: Mouton de Gruyter.
Hammarström, H. (2010). Rarities in numeral systems. In J.
Wohlgemut, & M. Cysouw (Eds.), Rethinking Universals: How
Rarities affect Linguistic Theory (Empirical Approaches to
Language Typology 45) (pp. 11-60). Berlin: Mouton de
Gruyter.
Haspelmath, M. (1995). The converb as a cross-linguistically valid
category. In M. Haspelmath, & E. König (Eds.), Converbs in
Cross-Linguistic Perspective. Berlin: Mouton de Gruyter.
Haspelmath, M. (2007). Coordination. In T. Shopen (Ed.), Language
Typology and Syntactic Description (Second ed. ed., Vol. II, pp.
1-51). Cambridge: Cambridge University Press.
Heegård Petersen, J. (2006). Local case-marking in Kalasha.
Copenhagen: University of Copenhagen.
Heine, B., & Kuteva, T. (2002). World Lexicon of
Grammaticalization. Cambridge: Cambridge University Press.
Hockett, C. F. (1958). A Course in Modern Linguistics. New York:
MacMillan.
Ladefoged, P., & Maddieson, I. (1996). The Sounds of the World's
Languages. Oxford: Blackwell Publishers.
Ledo-Lemos, F. J. (2000). Femininum Genus. München: Lincom.
Lewis, P. M., Simons, G. F., & Fennig, C. D. (Eds.). (2013).
Ethnologue: Languages of the world (Electronic ed.). Dallas:
SIL International. Retrieved from www.ethnologue.com
202
References
Liljegren, H. (2008). Towards a grammatical description of Palula - An
Indo-Aryan language of the Hindu Kush. Stockholm:
Department of Linguistics, Stockholm University.
Liljegren, H., & Haider, N. (2011). Palula Vocabulary. Islamabad:
Forum for Language Initiatives.
Lyons, C. (1998). Definiteness. Cambridge: Cambridge University
Press.
Masica, C. P. (1991). The Indo-Aryan Languages. Cambridge:
Cambridge University Press.
Mock, J. (2011). Dards, Dardistan, and Dardic: an Ethnographic,
Geographic, and Linguistic Conundrum. Retrieved 12 2, 2011,
from John Mock and Kimberley O'Neil:
http://www.mockandoneil.com/dard.htm
Morgenstierne, G. (1942). Notes on Dameli. A Kafir-Dardic Language
of Chitral. Norsk Tidsskrift for Sprogvidenskap, 12, 115-187.
Morgenstierne, G. (1961). Dardic and Kafir Languages. In The
Encyclopedia of Islam (New ed., pp. 138-139). Leiden: E.J.
Brill.
Morgenstierne, G. (1974). Languages of Nuristan and surrounding
regions. In K. Jettmar (Ed.), Cultures of the Hindukush Selected Papers from the Hindu-Kush Cultural Conference held
at Moesgård 1970 (pp. 1-10). Wiesbaden: Franz Steiner Verlag.
Nikolaeva, I. (Ed.). (2007). Finiteness: Theoretical and Empirical
Foundations. Oxford: Oxford University Press.
Parkvall, M. (2006). Limits of Language. London: Battlebridge.
Payne, J. R. (1980). The Decay of Ergativity in Pamir Languages.
Lingua, 51, 147-186.
Population Census Organization, Government of Pakistan. (2012).
Census 2011. Retrieved April 17, 2013, from Population
Census Organization, Government of Pakistan:
http://www.census.gov.pk/census2011.php
Radloff, C. F. (1999). Aspects of the Sound System of Gilgiti Shina.
Islamabad: Summer Institute of Linguistics and National
Institute of Pakistan Studies, Quaid-i-Azam University.
Rehman, K. A. (2011). Language Shift in the Neelam Valley: A Case
Study of the Kundal Shahi Language (Unpublished doctoral
dissertation). Islamabad: Department of Anthropology, Quaidi-Azam University.
203
References
Rensch, C. R., Hallberg, C. E., & O'Leary, C. F. (1992).
Sociolinguistic Survey of Northern Pakistan, Volume 3: Hindko
and Gujari. Islamabad: National Institute of Pakistan Studies
and Summer Institute of Linguistics.
Stolz, T. (2008). Total reduplication vs. echo-word formation in
language contact situations. In P. Siemund, N. Kintana, P.
Siemund, & N. Kintana (Eds.), Language Contact and Contact
Languages (pp. 107-132). Amsterdam: John Benjamins.
Stolz, T., & Veselinova, L. (2005). Ordinal numerals. In M.
Haspelmath, M. S. Dryer, D. Gil, & D. Comrie (Eds.), World
Atlas of Language Structures. Oxford: Oxford University Press.
Strand, R. (1973). Notes on the Nuristani and Dardic Languages.
Journal of the American Oriental Society, 93(3), 297-305.
Strand, R. (2001). The Tongues of Peristân. In A. M. Cacopardo, & A.
S. Cacopardo, Gates of Peristan; history, religion and society in
the Hindu Kush (pp. 251-257). Rome: Instituto Italiano per
l'Africa e l'Oriente.
Strand, R. (2011, 12 08). Languages of the Hindu Kush. Retrieved 08
23, 2013, from Richard Strand's Nuristân Site:
http://nuristan.info/lngIndex.html
Tamm, M. K. (2006). Nouns. In K. Brown (Ed.), Encyclopedia of
Languages and Linguistics (pp. 720-724). Oxford: Elsevier.
Tegey, H., & Robson, B. (1996). A Reference Grammar of Pashto.
Washington, D.C.: Center for Applied Linguistics.
Trail, R., & Cooper, G. (1999). Kalasha Dictionary - with English and
Urdu. Islamabad: Summer Institute of Linguistics and National
Institute of Pakistan Studies.
Turner, S. R. (1962-85). A Comparative Dictionary of the Indo-Aryan
Languages. London, United Kingdom: School of Oriental and
African Studies.
UNESCO Ad Hoc Expert Group on Endangered Languages. (2003).
Language Vitality and Endangerment. Paris: UNESCO.
Whitney, W. D. (1889). Sanskrit Grammar (2 ed.). Cambridge, MA:
Harvard University Press.
Willson, S. R. (1999). Basic Burushaski Vocabulary. Islamabad:
Summer Institute of Linguistics and National Institute of
Pakistan Studies, Quaid-i-Azam University.
204
References
Wälchli, B. (2005). Co-Compounds and Natural Coordination.
Oxford: Oxford University Press.
Yamamoto, M. (1999). Animacy and reference (Electronic ed.).
Amsterdam: John Benjamins. Retrieved from
http://site.ebrary.com
Zoller, C. P. (2005). A Grammar and Dictionary of Indus Kohistani.
Volume 1: Dictionary. Berlin: Mouton de Gruyter.
205
Appendix 1: Verb roots
Simple verbs form an almost closed class in Dameli, with a limited
number of verb roots. In my material, 152 verb roots have been
identified (excluding the copula), which are shown here. Although these
are certainly not all there are to find, the total number is likely to be
relatively low in comparison with, for instance, English. The usual
strategy for expressing new verb meanings is not to create new roots but
to form conjunct verbs, combinations of a non-verbal element and a
verb, that function as a single word semantically (see 11.8).
Since the basic valency of a verb cannot be determined from the root
alone, each verb root is marked with either (i) for intransitive, (t) for
transitive or (c) for roots that are only used in causative verbs.
206
Appendix 1: Verb roots
Table 48: Verb roots
Root
Meaning
Root
Meaning
aċap (c)
ačooṭ (t)
akás (i)
akaṭ (i)
alam (i)
apʰ (c)
aẉp (c)
ãs (i)
aaċ (i)
aaštoor (t)
aaẉ (i)
bait (c)
ban (t)
bas (i)
baw (t)
bay (c)
baait (c)
biɡir (c)
bil (i)
bin (t)
birand (i)
biṣam (i)
biž (t)
br (i)
brikin (t)
buṣ (t)
buṭ (t)
ċap (c)
ċarak (i)
ċĩĩt (c)
čap (t)
čičir (c)
čoop (t)
čul (i)
čũũṣ (t)
čʰin (t)
c̣am (c)
c̣ip (t)
pull out
do one by one
climb
gather
hang
cover
thin (maize)
smile, laugh
arrive, reach
spread out (cloth)
get stuck
fold
dress
stay overnight
plant
happen, be able to
call, summon
spread
melt
see
moan
rest
pour
die
sell
stab
braid
chew
drip (of rain)
sharpen
eat (hard things)
smart, hurt
dip
rock to and fro
suck
cut
consider, opine
wash (clothes)
c̣ʰaar (t)
c̣ʰun (t)
dác̣ (t)
dral (t)
draaṣ (t)
dumuk
duw (c)
ḍrõõk (i)
ḍuŋɡ (c)
ɡan (t)
ɡaṭ (t)
ɡeew (i)
ɡi/ɡiɡ (t)
ɡiaaċ (t)
ɡirankʰaṭ (i)
ɡoor (t)
ɡrẽẽṭ (t)
iṣir (i)
kaċ (c)
kook (i)
kreeš (t)
throw
thresh (wheat)
look
scratch, rake
comb
push
wash
low
fall asleep
say
want, win, ask for
walk
take, buy
bring (here)
turn around
peel
tie
break
show
sleep
soften by mixing
with a liquid
scratch
do
tire
grow, cultivate
tell
dig
do/put?
find
cross
weep
drop
count
write
burn
burn
rub
kuč (c)
kur (t)
kʰasar (i)
kʰiṣ (t)
kʰuṇḍ (t)
kʰuṣ (c)
laɡ (c)
lai/lad (t)
laṭaŋɡ (i)
laak (i)
laas (t)
leekʰ (t)
lík (t)
looṣ (t)
luṣ (i)
makʰ (t)
207
Appendix 1: Verb roots
Root
Meaning
Root
Meaning
man (t)
matr (t)
maar (t)
muuṣ (i)
naɡ (i)
nat (i)
naṣṭ (i)
naaš (t)
nees (i)
neeṭ/nieṭ (c)
nikʰaal (t)
niš (i)
ootʰ (i)
ooš (c)
ooẉɡ (c)
pač (i)
palaam (t)
paš (c)
pičil (i)
pi (t)
poor (t)
prambal (i)
pramuṣṭ (t)
prapaṭ (c)
pražar (i)
preec̣ (c)
pr/prat (t)
puč (c)
pũũc̣ (c)
pẉei (t)
pʰac̣al (i)
pʰak (t)
pʰam (i)
pʰaal (t)
pʰirk (i)
pʰus (c)
pʰuuk (t)
raṭ (t)
saat (t)
accept
read
kill
play
go down
go in
die
destroy
go out
cut (hair)
take out
sit
stop
freeze
boil
cook, ripen
pass
clean
slip, slide
drink
pour
light
forget
burn
become ill
throw out
give
ask
scratch (someone)
squeeze
waste away
eat (small things)
swell
split
go around
lose (something)
blow
bury
keep
suuw (t)
šaam (t)
šẽẽk (i)
šook (t)
ṣa (c)
tap (i)
taap (t)
trap (i)
truk (t)
ṭamb (c)
ṭip (i)
ṭʰook (t)
ṭraṭ (c)
ṭumbur (i)
ṭup (c)
undas (i)
undre (i)
urušt (i)
uštruun (t)
wali (i)
waar (t)
weeb (i)
weeč (t)
wiɡ (t)
wiṭik (c)
ẉak (i)
ẉ (t)
yaṇḍ (c)
y/ɡ (i)
zaċ (i)
zaan (t)
žalak (c)
sew
castrate
breathe
dry
send, scoop (food)
become warm
heat up
run
snap, kill (lice)
thresh (maize)
drip
hammer, beat
drive (cattle)
roll
cover, hide
cough
fly
stand up
hear
hide
catch
reach
seek
fire, hit
cut off
bleat
free, allow, leave
beat
go
trickle
know
throw down
(mulberries)
be frightened
kill
be seen
eat
sneeze
make
208
žalaŋɡ (i)
žan (t)
žaap (i)
ži /žin (c)
žišt (i)
žup (i)
Appendix 2: Text
Introduction
The text presented here, The Patient Woman, was choosen for several
reasons, not the least of which is that the story itself is quite good.
This text has been an important part of my material since 2004, when
it was written by Asmat Ullah, who as been one of my main Dameli
co-workers throughout this entire project, as part of a writers workshop
arranged by the Frontier Language Institute (now Forum for Language
Initiatives) to test the experimental writing systems for Palula and
Dameli with which we had been working.
Analysing interlinear texts is always a work in progress, and this text
is nowhere near as ready as I would like it to be, but since it has been
with me so long, it is among the most well-analysed texts I have and also
one of those that have had the most impact on the work presented here,
as can be seen from the disproportionate number of examples retrieved
from it.
It is also an excellent illustration of the dynamic of the culture in
which Dameli is spoken. In the story, we are treated to a moving and
traditional illustration of patience, loyalty and steadfastness, but we are
also given a keen description of how a young woman can use the strength
of her character and her willingness to endure difficulties to shape her
own situation. At the end of the story, the protagonist has not only
escaped the stigma and hardships of widowhood, she also has a husband
who is probably closer to her own age than is usual in this culture, and,
perhaps most importantly, a good and caring mother-in-law. As anyone
familiar with the virilocal societies that characterise much of South Asia
knows, this may be the single most powerful person in the life of a young
wife, and the one most likely to make her days a living misery if she
decides to.
209
Appendix 2: Text
The Patient Woman
ek žami
sãã−Ø
duu put−suu
bai−tʰun
one woman 3SG.ANIM.POSS−M two son−KIN.PL be−INDIRPST.3PL
‘One woman had two sons.’
ek put liliwaak
bai−tʰaa
ek ba
učʰuṭa bai−tʰaa
one son young.man be−INDIRPST.3SG.M one TOPSH small be−INDIRPST.3SG.M
‘One son was a young man, and one was small.’
žami
woman
sãã−Ø
3SG.ANIM.POSS−M
lee harman
very desire
tʰaa
tee
be.IMPFV.3SG.M that
mãã−Ø
žeṣṭera put−a
ki žanii
kur−im
1SG.POSS−M elder
son−LOC for marriage do−FUT.1SG
‘The woman had a great desire to arrange a marriage for her elder son.’
tas žeṣṭera
that elder
put−a
ki žami
ɡaṭ−aw−a−ee
son−OBL for woman ask−CAUS2−CAUS−PFV.3SG
ãã žanii
šuruu
kur−ee
and marriage beginning do−PFV.3SG
‘She found a wife for the elder son and started the wedding.’
žaniibrei kul
ɡiaaɡ−een
bride
house bring−PFV.3PL
‘They brought the bride to the house.’
duura
ta
pʰračaa
ɡ−een
far
from guest
go−PFV.3PL
‘The guests who had come from far away left.’
taanu
kul
ta
maaṣ ootʰin−un
REFL.POSS house from people remain−PFV.3PL
‘The people of the house remained.’
210
Appendix 2: Text
see
žanii
diɡar−a
3SG.DIST marriage afternoon−LOC
žaniibra
bridegroom
banii
ki
ɡ−aa
oakwood for
go−PFV.3SG.M
‘On the afternoon of the wedding, the groom went to bring oakwood.’
lee
very
čiir
late
ni
not
aaċ−an
come−INF
sãã−Ø
3SG.ANIM.POSS−M
waǰa
reason
kul
ta maaṣ
sax
parešan
bun
house of people hard worried
become.PFV.3PL
‘Since it became very late, but he still didn't come, the people of the house
became very worried [and wondered]’
tee mač ku ni aaɡ−aa
that man why not come−PFV.3SG.M
‘why the man didn’t come.’
kul
ta maaṣ
weeč−aal
house of people search−PRSPTCP
tee
that
ek muṭ−a
one tree−LOC
nees−i
go.out−CP
dac̣−i−nun
look−EP−IMPFV.3PL
akas−i−san−a
lam
čʰi−i
climb−EP−PSTPTCP−LOC branch cut−CP
naɡi pre−i
zen bai−tʰaa
fall give−CP dead be−INDIRPST.3SG.M
‘The people of the house went out searching and saw, that climbing
up to cut a branch, he had fallen down and died.’
naaċap
suddenly
zanaaza ɡiaaċ−i
corpse bring−CP
kul−a
house−LOC
tʰ−a−een
be−CAUS−PFV.3PL
tee
sapun kʰušaali ɣam−a
badal
b−aa
that all
joy
grief−LOC exchange become−PFV.3SG.M
‘As they suddenly brought the corpse and put it in the house, all the joy was
turned into grief.’
211
Appendix 2: Text
paai sãã−Ø
yii
boy 3SG.ANIM.POSS−M mother
sãã−Ø
žanii
3SG.ANIM.POSS−M marriage
diyoo
day
taanu
REFL.POSS
žaniibra
bridegroom
zanaaza
corpse
bin−i
see−CP
put
son
apʰaamee
ɡ−ai
unconscious go−PFV.3SG.F
‘The mother of the boy, having seen the corpse of her bridegroom son on
his wedding day, fainted.’
lee čiir−ee pʰikir−a
aaċ−i
c̣aŋ pr−ee
very late−? thought−LOC come−CP cry give−PFV.3SG
‘When she came to herself, very late, she cried out.’
taanu
ṣaa−lum aċap−aai−i
laak−em
b−ui
REFL.POSS head−hair pull.out−CAUS−CP weep−INCHPTCP become−PFV.3SG.F
‘Pulling her hair, she started weeping.’
mudiya ta
awal diyoo daru
žanii
ta
today TOPSM first day be.INANIM.IMPFV.3 marriage of
‘“Today is the first day of the marriage.’
žaniibrei kaasa
ki ẉi−i
ɡ−eep
bride
for.whom for having_left−CP go−PFV.2SG
‘For whom have you left your bride?”’
paai sãã−i
yii
c̣aŋ pre−i
ɡan−ni
tee
boy 3SG.ANIM.POSS−F mother cry give−CP say−IMPFV.3SG.F that
‘Crying out, the mother of the boy said:’
mãã−Ø
putr−oo too
ni laak−i−num
1SG.POSS−M son−VOC 2SG.OBL not weep−EP−IMPFV.1SG
‘“My son, I am not crying for you.’
mãã−i
luṭi
bawi
laak−i−num
1SG.POSS−F young.girl daughter−in−law weep−EP−IMPFV.1SG
‘I am crying for my young daughter−in−law.”’
212
Appendix 2: Text
yee
baati baarbaar
kʰuṇḍ−i−ni
3SG.INANIM.PROX word again.and.again tell−EP−IMPFV.3SG.F
tee žaniibrei hairan
b−ui
that bride
wondering become−PFV.3SG.F
‘As she said these words over and over again, the bride started to wonder.’
yii
ta
taanu
put sãã−Ø
ɣam kur−an−baṣ−a
mother TOPSM REFL.POSS son 3SG.ANIM.POSS−M grief do−INF−able−LOC
mãã−Ø
ɣam ku−ni
1SG.POSS−M grief do−IMPFV.3SG.F
‘“The mother could be grieving for her own son, but she grieves for me!”’
yee
3SG.INANIM.PROX
baati−a
word−LOC
bawi
sãã−Ø
daughter−in−law 3SG.ANIM.POSS−M
zaadi−a
ibrat
naɡ−aa
heart−LOC good_lesson come.down−PFV.3SG.M
‘Hearing these words the daughter−in−law learned the good lesson in her
heart.’
tanii
taanu
zaadi−a
yee
faisala
3SG.ANIM.DIST.ERG REFL.POSS heart−LOC 3SG.INANIM.PROX decision
kur−ee
do−PFV.3SG
‘She made this decision in her heart.’
ay
mãã−Ø
mas
učʰuṭa deer
1SG.NOM 1SG.POSS−M 3SG.ANIM.PROX.ACC small brother−in−law
sãã−Ø
naam−a
niš−im
3SG.ANIM.POSS−M name−LOC sit−FUT.1SG
‘I will wait for the name of this my young brother−in−law.’
kya
which
waxt−a
time−LOC
zuaani−a
prat−ee
young.man−M give−PFV.3SG
ba
TOPSH
mas−a
mili nikaa ɡrẽẽṭ−aw−a−im
3SG.ANIM.PROX.ACC−LOC with nika tie−CAUS2−CAUS−FUT.1SG
‘When he becomes a young man, I will marry him.’
213
Appendix 2: Text
lee waxt aaṣanta preeš−es−ee
ɡan−ee
tee
very time after mother−in−law−KIN.3−ERG say−PFV.3SG that
‘A long time later her mother−in−law said:’
tu
taanu
dadi sãã−Ø
kul
y−et
2SG.NOM REFL.POSS father 3SG.ANIM.POSS−M house go−IMP.SG
‘“Go to your father’s house.’
amun−a tarafta tu
azaad tʰop
our−M side 2SG.NOM free be.IMPFV.2SG
‘From our side, you are free.”’
bawi−es−ee
ɡan−ee
tee
daughter−in−law−KIN.3−ERG say−PFV.3SG that
‘Her daughter−in−law said:’
tu
tãã−Ø
put ni laak−i
muu laak−i−nap
ba
2SG.NOM 2SG.POSS−M son not having.cried−CP 1SG weep−EP−IMPFV.2SG PART
‘“You did not cry for your son, you cried for me.’
ay
ba
matiki beɡarati
ni tʰum
1SG.NOM TOPSH so
dishonourable not be.IMPFV.1SG
‘I am not so dishonourable,’
tee
that
tãã−Ø
2SG.POSS−M
kul
house
ẉi−i
leave−CP
ware
other
sãã−Ø
3SG.ANIM.POSS−M
kul−a
ki y−im
house−LOC to go−FUT.1SG
‘that I would leave your house to go to someone else’s house.’
žaa ba
ay
tãã−Ø
učʰuṭa put−a
ki niš−i−num
now TOPSH 1SG.NOM 2SG.POSS−M small son−LOC for sit−EP−IMPFV.1SG
‘Now I am waiting for your younger son.’
yee
mãã−Ø
axiri faisala daru
3SG.INANIM.PROX 1SG.POSS−M final decision be.IMPFV.INANIM.3
‘This is my final decision.”’
214
Appendix 2: Text
lee zamaana aaṣanta ek diyoo brei dac̣−i−ni
very times
after
one day
girl look−EP−IMPFV.3SG.F
tee paai
mrãč
muṭ−a
akas−isan
tʰaa
that boy
mulberry tree−LOC climb−PSTPTCP be.IMPFV.3SG.M
‘A long time later, one day, the girl sees that the boy has climbed a mulberry
tree.’
brei paai muṭ−a
bin−i lee kʰošan b−ui
girl boy tree−LOC see−CP very happy become−PFV.3SG.F
‘The girl, having seen the boy in the tree, became very happy.’
paay−a ki kaaṇ
kur−ee
boy−LOC to sound do−PFV.3SG
muu
ki
1SG.OBL to
mrãč
žalak−ai
mulberry throw.down.mulberries−CAUS.IMP.SG
‘She called to the boy: “Throw down some mullberries for me.”’
paay−a brei ki mrãč
žalak−a−ee
tee
boy
girl to mulberry throw.down.mulberries−CAUS−PFV.3SG that
‘The boy threw down mullberries, then’
pʰak−i
apʰataar ba−i
preeš
sãã−Ø
eat.small.things−CP quickly become−CP mother−in−law 3SG.ANIM.POSS−M
aaɡ−ai
come−PFV.3SG.F
‘After eating, she quickly came to her mother−in−law.’
umbaarak
beebu tãã−Ø
put zuaan
bai−tʰaa
congratulations to.you 2SG.POSS−M son young.man be−INDIRPST.3SG.M
‘“Congratulations to you, your son has become a young man.’
mulaa baaitaa−i
amun−a nikaa kur−aa
mullah summon−CP our−M nika do−IMP.PL
‘Summon the mullah and perform our wedding.”’
215
Appendix 2: Text
preeš
yee
baati uštruun−i
mother−in−law 3SG.INANIM.PROX word hear−CP
lee kʰošan b−ui
very happy become−PFV.3SG.F
‘Upon hearing these words, the mother−in−law became very happy.’
ãã mulaa baaitaa−i
tasun−a
nikaa ɡrẽẽṭ−een
and mullah summon−CP 3SG.DIST.POSS−M nika tie−PFV.3PL
‘And, having called the mullah, they performed their marriage.’
brei
girl
sãã−Ø
3SG.ANIM.POSS−M
sabur
azmãã waxt
patience trial time
aaxir xatam
b−aa
finally finished
become−PFV.3SG.M
‘The days of trying the girl’s patience were finally at an end.’
natiža aɡar iraada
mazbut tʰaa
ba
muškilat
result if
intention strong be.IMPFV.3SG.M TOPSH difficulties
ta
bawuǰut insan
taanu
manzala
ki
wee−b−in
of
having human REFL.POSS
destination to
reach−be−FUT.3PL
‘Sens Morale: If your will is strong, you will reach your destination, in spite of
all difficulties.’
216
‫بَس‬
Emil Perder
A Grammatical Description
of Dameli
Emil Perder
A Grammatical Description of Dameli
ISBN 978-91-7447-770-2
Department of Linguistics
Doctoral Thesis in Linguistics at Stockholm University, Sweden 2013
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

Related manuals

Download PDF

advertisement