University of Southern California Dissertations and Theses

A descriptive grammar of San Bartolome Zoogocho Zapotec

(USC Thesis Other)

A descriptive grammar of San Bartolome Zoogocho Zapotec

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content NOTE TO USERS
This reproduction is the best copy available.
®
UMI
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
A DESCRIPTIVE GRAMMAR OF SAN BARTOLOME ZOOGOCHO ZAPOTEC
by
Aaron Huey Sonnenschein
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment o f the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(LINGUISTICS)
December 2004
Copyright 2004 Aaron Huey Sonnenschein
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UMI Number: 3155481
Copyright 2004 by
Sonnenschein, Aaron Huey
All rights reserved.
INFORMATION TO USERS
The quality of this reproduction is dependent upon the quality of the copy
submitted. Broken or indistinct print, colored or poor quality illustrations and
photographs, print bleed-through, substandard margins, and improper
alignment can adversely affect reproduction.
In the unlikely event that the author did not send a complete manuscript
and there are missing pages, these will be noted. Also, if unauthorized
copyright material had to be removed, a note will indicate the deletion.
®
UMI
UMI Microform 3155481
Copyright 2005 by ProQuest Information and Learning Company.
All rights reserved. This microform edition is protected against
unauthorized copying under Title 17, United States Code.
ProQuest Information and Learning Company
300 North Zeeb Road
P.O. Box 1346
Ann Arbor, Ml 48106-1346
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Dedication
This dissertation is dedicated with love to
my daughter,
Cordelia Maria Beatrice Sonnenschein
and my wife,
Ruth Anne Crossley.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Acknowledgements
I would like to take this opportunity acknowledge the many people without whose help and
assistance I would not have been able to complete this dissertation.
First and foremost, 1 would like to thank the people o f San Bartolome Zoogocho who have
graciously allowed me to learn their language. Angela Mendoza Marcial, my first teacher and
longtime friend, deserves a great deal o f credit for first introducing me to the language. In Los
Angeles, Samuel Morales Ruiz, and Enrique and Verna Calete gave my work a sense o f purpose and
helped me down the first part o f the continuing process o f teaching me what I had not learned and o f
correcting what I misunderstood about the grammar o f San Bartolome Zoogocho Zapotec. In the
actual village o f Zoogocho, Alberta Martinez Marcial has been a great teacher, helper, and friend. Her
family, especially her husband, Eduardo Vasquez Garcia, and children Eduardo and Narcedalia, and
also her uncle Fidencio and mother Angela deserve a great deal o f credit for humoring me and helping
me to feel that I belonged in Zoogocho when undertaking fieldwork. 1 would also like to thank more
generally all o f the village for accepting, with reservations, this crazy gringo who came out o f the
blue. I would like to especially thank the rural school teachers who helped me give something back to
the community in the form o f English classes, by opening their primary school to me.
I am very grateful to the linguistics department at the University o f California at Berkeley,
for having inspired me to study linguistics, and to concentrate on descriptive linguistics. I also would
like to thank the University o f California Education Abroad Program for giving me the opportunity to
study for a semester at the National Autonomous University o f Mexico, and to conduct fieldwork in
the highlands o f Chiapas. 1 knew from that point forward generally what I wanted to do. I am also
very thankful to Terrence Kaufman for having introduced me to Zapotecan linguistics.
I would also like to thank the members o f my committee, Bernard Comrie, John A. Hawkins,
and Pamela Munro, for having stuck with me over all o f these years. I have taken great inspiration
from each o f them in different ways and have learned much o f being a linguist and academic. 1 also
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
appreciate their willingness to consider the manuscript at various unpolished points in its evolution,
and the thorough comments which have guided my work.
I would like to thank all o f the teachers and colleagues that 1 have had in Los Angeles. 1
would like to particularly thank the attendees and presenters at the American Indian Seminar at
UCLA, and my fellow students at USC. I also would like to thank my colleagues at the Landelijke
Onderzoekschool Taalwetenschap and at the Max Planck Institute for Psycholinguistics in Nijmegen
for their support and help making it through the cold dark Dutch winters.
I am very grateful to the Writing Program at the University o f Southern California for
providing the majority o f the financial support for my graduate studies. The Max-Pianck Institute for
Psycholinguistics in Nijmegen deserves a great deal o f credit for funding the majority o f the fieldwork
which led to the current dissertation and for providing an intellectually stimulating environment while
I was in Holland. I am also very grateful to the Graduate School for the Summer Finishing Fellowship
which enabled me to do just that.
Finally, I would like to take this opportunity to thank my family for their help and support
with the dissertation. My mother, father, stepmother, and sisters have been a great source o f
inspiration and support. My aunt, Susan Sonnenschein, helped me greatly with the formatting and in
converting/making legible the two maps. Most importantly, my daughter Cordelia Maria Beatrice
Sonnenschein, and her mother, Ruth Anne Crossley, deserve many thanks for being my inspiration to
write, to take time o ff to lie in the grass and collect my energies, and to breathe.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
V
Table of Contents
Dedication ii
Acknowledgments iii
List o f Tables x
List o f Figures xiv
Abstract xv
Abbreviations and glossing conventions xvi
1 Introduction 1
1.1 Basic information about San Bartolome Zoogocho Zapotec 1
1.2 The sociolinguistic situation
2
1.3 Genetic and dialectal information 4
1.4 Previous research 8
1.5 Non-Zapotec languages spoken in the region 9
1.6 Other geographic and ethnographic information 10
1.7 Typological characteristics 12
1.8 Theoretical framework 12
2 Phonetics and phonology 14
2.1 Orthography and inventory o f sounds 14
2.2 Consonants 17
2.2.1 Examples 18
2.2.1.1 Stops 18
2.2.1.2 Nasals 19
2.2.1.3 Laterals 20
2.2.1.4 Fricatives 21
2.2.1.5 Affricates 21
2.2.1.6 Glides 22
2.2.1.7 The glottal stop 22
2.2.2 Loans 22
2.2.3 Labialization 22
2.2.4 Consonant clusters 23
2.3 Vowels 30
2.3.1 Phonation 32
2.3.2 Tone 32
2.3.3 Intonation 33
2.4 Syllable structure 33
2.5 Stress 34
2.6 Phonological word 34
3 Pronominal forms 37
3.1 Introduction 37
3.2 Independent pronouns 37
3.3 Dependent forms 41
3.4 Experiencer and instrumental forms 44
3.5 Dependent pronominal forms for the object 48
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
vi
3.6 Emphatic pronouns 50
3.7 Co-occurrence restrictions between subject and object clitics 51
3.8 The classification o f the forms described above 54
4 Morphology 63
4.1 Introduction 63
4.2 Nominal morphology 63
4.2.1 Possession 63
4.2.2 Lexical compounds 66
4.2.3 Nominal template 66
4.3 Verbal morphology 67
4.3.1 General overview 67
4.3.2 Primary aspect 67
4.3.3 Other, more rarely encountered preverbal aspectual markers 99
4.3.4 Plural marking for third person subjects 101
4.3.5 Secondary aspects 103
4.3.6 Andative/venitive 107
4.3.7 Imperatives 109
4.3.8 Infinitives 110
4.3.9 Valence changing operations 111
4.3.10 Incorporated nouns 115
4.3.11 Adverbial suffixes 116
4.3.12 Pronominal clitics 117
4.3.13 Verbal template 118
4.4.1 Adjectives 118
4.4.2 Quantifiers 119
4.5 The morphological typology o f SBZZ 119
5 Simple constructions 125
5.1 Introduction 125
5.2 Basic word order 125
5.3 The noun phrase 128
5.3.1 Adjective-noun order 128
5.3.2 Cardinal numbers and quantifiers 130
5.3.3 Plural marker 132
5.3.4 Demonstratives 132
5.3.5 Possession 133
5.3.5.1 Introduction 13 3
5.3.5.2 A brief overview o f attributive possession in Zoogocho Zapotec 13 3
5.3.5.3 Discussion o f attributive possession in the literature 136
5.3.5.4 A processing approach to the formal marking 142
5.3.5.5 The use o f attributive possession in three Zoogocho Zapotec texts 146
5.3.6 Determiners 151
5.3.7 The directional clitic 152
5.3.8 Overall NP ordering 153
5.4 Declarative clauses 155
5.4.1 Intransitive clauses 155
5.4.2 Transitive clauses 156
5.4.3 Ditransitive clauses 156
5.4.4 Negation 158
5.4.5 Adverbs 160
5.4.5.1 Time 160
5.4.5.2 Place 161
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
VII
5.4.5.3 Manner
5.4.5.4 Time/Manner/Place
5.5 Pragmatic fronting o f elements
5.5.1 Topicalization
5.5.2 Focus
5.5.3 Fronting o f non-core arguments
5.6 Interrogative sentences
5.6.1 Overall
5.6.2 Yes/no
5.6.3 Content questions
5.6.3.1 no
5.6.3.2 bate’
5.6.3.3 ga/gan
5.6.3.4 bal/bale’
5.6.3.5 kaka
5.6.3.6 nake’
5.6.3.7 bicheen
5.6.3.8 bi
5.6.4 Interrogative clauses-a synopsis
5.7 Zoogocho Zapotec and Greenberg’s universals
6 Complex Constructions
6.1 Causatives
6.1.1 Morphological causative
6.1.2 Syntactic causative
6.1.3 Differences between the two constructions
6.2 Instrumentals and comitatives
6.3 IO ‘lowering’ and benefactive constructions
6.3.1 IO ‘lowering’
6.3.2 The basic expression o f beneficiaries
6.3.3 Beneficiaries or recipients?
6.3.4 What is whose?
6.4 The reflexive o f possessive construction
6.4.1 Introduction
6.4.2 Binding conditions
6.5 Constructions with multiple verbs
6.5.1 Introduction
6.5.2 Infinitival constructions
6.5.3 Verbs borrowed from Spanish
6.5.4 Auxiliary constructions
6.5.5 Complementation
6.5.6 Argument sharing
6.5.7 Relative clauses
6.5.8 Coordination
6.5.9 Adverbial clauses and indirect questions
6.6 Comparatives
6.7 Conditionals
7 Lexical classes
7.1 Introduction
7.2 Verbs
7.2.1 Auxiliary verbs
7.2.2 Positional verbs
161
162
164
164
164
165
166
166
166
167
167
170
170
172
175
175
176
176
179
180
193
193
193
194
195
196
198
198
199
201
203
206
206
208
210
210
212
213
215
217
221
222
227
231
232
234
236
236
242
244
244
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
v i i i
7.3 Adverbs 251
7.4 Nouns 252
7.4.1 Nouns 252
7.4.2 Reflexives and reciprocals 254
7.4.3 Pronouns 256
7.4.4 Generic nouns 257
7.5 Adjectives 258
7.6 Quantifiers 2 6 1
7.7 Noun and verb adjuncts 262
7.8 Demonstratives 264
7.9 Question words 268
7.10 Conjunctions 269
7.11 Other closed classes 2 7 1
7.12 The relational noun/locative preposition continuum 277
7.12.1 Introduction 277
7.12.2 Heine etal. 279
7.12.3 Mixtecan 283
7.12.4 MacLaury and Ayoquesco Zapotec 285
7.12.5 An initial description and textual exploration o f relational nouns in 288
Zoogocho Zapotec
7.12.6 Non-body part derived prepositions 294
7.12.7 The cognitive development o f body part terms in Zoogocho 296
Zapotec
7.12.8 Lexical classification o f body part terms 302
7.12.9 Comparison with Valley Zapotec 312
7.12.10 Conclusion 316
8 A performance approach to word order in SBZZ 3 19
8.1 Introduction: Basic word order properties 319
8.2 Structural approaches to the typology o f verb initial languages 323
8.2.1 Payne’s restatement o f Keenan 323
8.2.2 Dryer 3 4 1
8.2.3 Polinsky 343
8.3 A pragmatic approach to word order variation in verb initial 350
languages: Du Bois and ‘The Discourse Basis o f Ergativity’
8.4 A processing approach to VSO order 352
8.5 Textual data from SBZZ 357
8.5.1 Methodology 357
8.5.1.1 Corpus 357
8.5.1.2 Coding 358
8.5.2 Data 360
8.5.2.1 Types o f data 360
8.5.2.2 Quantitative data and analysis 369
8.5.2.3 Comparison with similar studies 396
8.6 Topic and focus 405
8.7 Conclusions 4 12
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
9 Conclusion 419
9 Conclusion 419
Bibliography 422
Appendix: Text 431
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
X
List of Tables
2.1 SBZZ Vowels 15
2.2 SBZZ Consonants 16
2.3 Stops 18
2.4 Nasals 19
2.5 Laterals 20
2.6 Fricatives 21
2.7 Affricates 21
2.8 Glides 22
2.9 The glottal stop 22
2.10 Vowels 30
3.1 Independent pronouns 37
3.2 Forms for third person 41
3.3 Dependent pronouns 42
3.4 Experiencer forms 45
3.5 Dependent object forms 49
3.6 Yalalag Zapotec co-occurrence restrictions 52
3.7 Yatzachi Zapotec co-occurrence restrictions 53
3.8 Zoogocho Zapotec co-occurrence restrictions 54
3.9 Classification o f pronominal elements in Zoogocho
Zapotec
59
4.1 Conjugation in gw- 76
4.2 Conjugation in y-, completive subclass in b- 77
4.3 Conjugation in y-, completive subclass in gd- 79
4.4 Conjugation in y-, completive subclass in gw- 80
4.5 Conjugation in g-, completive subclass in go- 81
4.6 Conjugation in g-, completive subclass in g- 83
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
x i
4.7 Conjugation in g-, completive subclass in b- 84
4.8 Conjugation in g-, completive subclass in be- 85
4.9 Conjugation in g-, completive subclass in bi- 86
4.10 Conjugation in 0-, completive subclass in b- 88
4.11 Conjugation in 0-, completive subclass in gu- 89
4.12 Conjugation in 0-, completive subclass in gud-, regular 91
continuative
4.13 Conjugation in 0-, completive subclass in gud-, 92
continuative in dxy-
4.14 Conjugation in 0-, completive subclass in gulh- 94
4.15 Conjugation in kw-, completive subclass in gulh- 95
4.16 Conjugation in 0-, completive subclass in gu 97
4.17 Conjugation in sh-, completive subclass in gy- 98
4.18 -o-plural forms 102
4.19 -e- plural forms 102
4.20 Plural forms with completive in gw- 103
4.21 Plural forms for vowel initial roots 103
4.22 Frequentative aspect 106
5.1 Methods o f possession in Zoogocho Zapotec 146
5.2 Possession in Tigr 148
5.3 Possession i n ‘G ’ 149
5.4 Possession in Miner 149
7.1 Positional verbs 246
7.2 Distributions and categories o f demonstratives 265
7.3 Textual usage o f body part terms 293
7.4 Noun-preposition cline 3 1 1
8.1 EIC and VO languages 356
8.2 Overall count o f clause types 369
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
x i i
8.3 Overall transitivity 372
8.4 The VS type 373
8.5 The SV type 374
8.6 The SVS type 376
8.7 Intransitive clauses 377
8.8 The VAO type 378
8.9 The OVA type 380
8.10 The AVO type 383
8.11 Miscellaneous transitives 384
8.12 Transitive clauses 386
8.13 The VAO 102 type 387
8.14 The O l V A 02 type 388
8.15 Ditransitive clauses 390
8.16 Postverbal subjects 391
8.17 Preverbal subjects 391
8.18 Pre- and postverbal subjects 392
8.19 Zoogocho Zapotec s{ vp[V] S vp[0]} 395
8.20 Syntactic role in clauses with one lexical argument in
SBZZ
397
8.21 Syntactic role in clauses with one lexical argument in
Yagua
397
8.22 Pre- and postverbal subjects and objects in SBZZ and
Yagua
399
8.23 Number o f lexical arguments (transitive and intransitive
clauses conflated)
400
8.24 Number o f lexical arguments per clause and transitivity
in SBZZ
401
8.25 Number o f lexical arguments per clause and transitivity
in Sacapultec
402
8.26 Grammatical role and syntactic type 404
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
x i i i
8.27 Given/new reference by syntactic role 405
8.28 Du Bois' Table 9 'Dimensions and constraints o f 413
Preferred Argument Structure'
8.29 Accessibility Marking Scale 414
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
List of Figures
1.1 Map o f Approximate Locations o f SBZZ-speaking
Communities
1.2 Map o f Language Groupings
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
XV
Abstract
In this dissertation, I provide a grammatical description o f San Bartolome Zoogocho
Zapotec, an endangered Otomanguean language spoken in the southern Mexican state o f Oaxaca. The
initial six chapters are concerned with providing a description o f the major grammatical features o f
the language, while the final two examine two major current theoretical issues: parts-of-speech and
word order.
The first six chapters provide descriptions o f the ethnographic and sociolinguistic situations
o f the Zoogocho Zapotec community, the sounds o f the language, the pronominal system, the
morphology, and the syntax o f the language. While no particular theoretical framework is used, the
inspiration for much o f the description comes from the typological universal grammar research
program. SBZZ (as I will refer to San Bartolome Zoogocho Zapotec) is a tonal language which can be
complex phonologically. It is an agglutinative, slightly fusional language. It is a prototypical VSO
language having prepositions, NAdj, NDem, NGen, and NRel orders. Various means o f combining
clauses exist, including complementation, coordination, and relativization.
Chapter Seven is an examination o f the lexical classes present in SBZZ. While I try to define
necessary and sufficient conditions for each lexical class, it can be difficult to find conditions which
are both necessary and sufficient. I try to rely, therefore, on multiple definitions which, while
informed by a variety o f cross-linguistic data, are based on and presented by the SBZZ grammar. I
devote much o f the discussion to the grammaticalization o f relational nouns, a topic which has
received a great deal o f discussion in the literature, both specifically for Zapotecan and Otomanguean
languages and more generally. I conclude that relational nouns are a separate category from
prepositions and regular nouns in SBZZ, while sharing characteristics with both.
In Chapter Eight, I examine verb initial word order, place SBZZ in two typologies o f verb
initial languages, and compare a study o f word order in two SBZZ texts with other textual studies o f
word order in verb initial languages. Chapter Eight also confirms a hypothesis about the processing o f
VSO languages which comes from the processing theory o f John A. Hawkins.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Abbreviations and glossing conventions
List of abbreviations
lplexcl first person plural exclusive
Iplincl first person plural inclusive
lsg first person singular
1sgexp first person singular experiencer
1 sgfsf first person singular fast speech form
2pl second person plural
2sg second person singular
2sgexp second person singular experiencer
2sgfsf second person singular fast speech form
2sgm second person singular male to male form
3 an third person animal
3 f third person formal
3fexp third person formal experiencer
3fo third person formal object form
3inan third person inanimate
3 inf third person informal
A agent o f transitive clause
adv adverb
an animate
and andative
caus causative
clan classifier for animates
clinan classifier for inanimates
clsm classifier for babies and small things
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
comp completive
cond conditional
cont continuative
cop copula
demadv demonstrative adverb
demdist distal demonstrative
demmed medial demonstrative
demprox proximate demonstrative
dim diminutive
dir directional
DO direct object
dp definite past
dub dubitative
emph emphatic
freq frequentative
genan animate generic noun
geninan inanimate generic noun
gensm small generic noun
inan inanimate
IO indirect object
indef indefinite
inf infinitival form
instr instrumental
int intensifier
itr intransitive
neg negative
negan negative for animate
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
x v i i i
neginan negative for inanimate
num numeral
O object o f transitive clause
Pi
plural
poss possessive
pot potential
rel relative pronoun
rep repetitive
S subject o f intransitive clause
s.o. someone
Sp. Spanish
stat stative
s.t. something
tr transitive
V verb
ven venitive
Glossing conventions
I followed the Leipzig Conventions for glossing. Those conventions will be followed strictly
in the final draft for typing. The hyphen (-) is used to indicate a word internal morpheme boundary
and an equal sign (=) is used to indicate clitic boundary. A period (.) is used in those cases where the
morphemes are separate but not segmentable (e.g. a suppletive form or some type o f fusion). The
ordering o f fused morphemes should be in the order which corresponding non-fused morphemes
normally would take. A _ is used when a single SBZZ lexeme corresponds to more than one English
lexeme. In addition, plurality is not marked for the third person pronouns, as either a separate
morpheme marks plurality (often not adjacent to the pronominal form) or it is recoverable from
context.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Chapter One: Introduction
1.1 Basic information about San Bartolome Zoogocho Zapotec
San Bartolome Zoogocho Zapotec1 (or dizha zxon ‘Zxon language’ as it is
referred to in SBZZ) is an Otomanguean language primarily spoken in San
Bartolome Zoogocho in the Northern Sierra of Oaxaca at 170’ 15” north by 96’ 1 ” E.
As of the year 2000, there were fewer than 45 monolinguals out of a total population
of 638 according to the Mexican National Census
(http://www.ini.gob.mx/indica2000/mpo/oax83.htm).
The number of speakers in the village has dropped considerably over the
years. In 1980, according to the Mexican National Census, there was a population of
750 speakers out of a total population of over 2200 total residents in Zoogocho. This
number had dropped to 560 total bilingual speakers and 72 monolingual speakers out
of a total population of 716 residents according to the 1990 census as reported in
Long and Cruz (1990:11).
There has been significant migration out of the village over the past fifty
years to Oaxaca City, Santa Cecilia and Santa Teresa Nigromante in the state of
2 • ' *
Veracruz , Mexico City and the surrounding areas, especially the state of Mexico.
and, even as far away as the United States, primarily Los Angeles, California (p.c.
Alberta Martinez Marcial). This migration has often been as the result of government
programs such as the Bracero program in the United States and Mexican programs
such as one which gave young Zapotecs training in Mexico City to be barbers.
According to Zoogochense in Los Angeles, there are more than 500 people from
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Zoogocho in Los Angeles. O f these 500, not everyone speaks Zapotec and some are
probably cross-listed in Zoogocho, maintaining legal status both in Los Angeles and
in Zoogocho. In Los Angeles, the Zoogochense do not live in one single area but arc
rather dispersed. The following map shows the approximate location of the primary
SBZZ-speaking communities mentioned so far.
Figure 1.1: Approximate Locations of SBZZ-speaking Communities
Map Abbreviations
LA: Los Angeles
DF: Distrito Federal
EdM: Estado de Mexico
O: Oaxaca
Z: Zoogocho
N: Santa Teresa
Nigromante
DF \
EdM g N
1.2 The sociolinguistic situation
As already noted, the great majority of SBZZ speakers are no longer
monolingual, and many note that the language has changed as a result of contact with
Spanish during their lifetimes. Most of the children of whom I know in the village
grow up learning both Zapotec and Spanish in the household, and the influence of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Spanish is definitely increasing, along with the increase in the number of households
which have television, often with satellite programming.
There are still, however, many households where Zapotec is the primary
language used; and Zapotec is still used in the market, some business transactions, in
the city government, and the local intemado (government boarding school). The
internado serves many other communities, both Zapotec and non-Zapotec, and is not
attended by every Zoogochense student. The other primary and secondary schools in
Zoogocho all use Spanish as the medium of instruction. There is a radio station in
Guelatao which broadcasts in Zapotec occasionally, but rarely if ever is this in the
Zoogocho or even the Villa Alta variety. However it is an important and encouraging
resource to speakers, who are happy to hear any variety of Zapotec being broadcast.
From what I know of the situation among speakers who have migrated out of
Zoogocho, as based on personal communication and observation during the saint’s
day festival for San Bartolome, an occasion which sees many Zoogochense return
from afar, SBZZ is not actively learned by the majority of the first generation born
outside the village. In general, I feel that, especially with a concerted effort on the
part of native speakers (which is very possible) and with an incredibly improved
local economy (which is conceivable), there is some potential for the language to
still be spoken in 50 years. I would suppose that the situation was similar in most of
the towns within the Zxon dialect grouping.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4
1.3 Genetic and dialectal information
SBZZ is a member o f the Northern grouping of the Zapotecan sub-family of
the Otomanguen stock. Note that Zapotec comprises anywhere between 5 languages
(Kaufman 2004) to perhaps as many as 58 languages, as per the classification on
Ethnologue (Grimes et al 2000).
The Northern grouping encompasses at least four distinct dialect groupings as
a conservative estimate, and perhaps as many as five. The way the language
grouping is divided by the authors of the Diccionario Zapoteco-Esparto I: Reglas
para el entendimiento de las variants dialectales de la sierra: Hechos par zapolecos
de la variante del sector xhon3 (Castellanos et al 1995), a bilingual dictionary
compiled by native speakers of the-speakers of the Zxon dialect grouping which
includes Zoogocho Zazpotec, is as follows.
The lhe’ja dialect grouping includes varieties spoken in the district of Ixtlan
de Juarez (such as Chicomezuchitl, Atepec, Analco, Aloapam, Amatlan, San Miguel
del Rio, Macuiltianguis, Yarenia, Ixtepeji, Lachatao, Jaltianguis, Yavesia,
Soquiapan, and Teococuilco) and is said to be the most distinct of all the varieties of
the Sierra (ibid. 12).
Xidsa is spoken in both the district of Ixtlan in the towns of Yaneria, Josaa,
Tiltepec, Ysagila, Teolaxco, Yagavila, Tepanzacoalco, Cacolotepec, and Yotao: and
the district of Villa Alta in the towns of Juquila Vijanos, Yaee, Yatzona,
Temacalapa, Camotlan, Lalopa, Roayaga, Tanetze, and Villa Alta (ibid. 13).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Xhon, the variety to which SBZZ belongs, is spoken primarily in the Villa
Alta district in the towns of Solaga, San Andres Yaa, Yatzachi el Bajo, Zoogocho,
Lachirioag, San Francisco Cajonos, Taba, San Mateo Cajonos, Betaza, Yaganiza.
San Pedro Cajonos, Yalina, Zoochila, Xagacia, and Yalalag. Castellanos et al also
note the existence of the community in Nigromante, Veracruz. I will obviously come
back to this language grouping later in this section (ibid. 14).
Xan, considered to be ‘una extension de la variante xidsa’ ‘an extension of
the xidsa variety’ (ibid. 14), is spoken in the district of Choapan in the towns of
Comaltepec, Lachixoba, Jaltepec, Latani, Hahuive, Yaveloxi, and Choapan (ibid.
15).
Finally, welab, is spoken in ‘algunos pueblos del Districto de Tlacolula, que
se encuentran en la colindancia con los pueblos Cajonos y son conocidos como las
Albarradas’, ‘some towns in the Tlacolula District which are found neighbouring
Cajonos towns known as the Albarradas’ (ibid. 15), Santa Maria Albarradas. Santo
Domingo Albarradas, San Lorenzo Albarradas, and San Bartolo Albarradas. It is
very doubtful whether this dialect grouping belongs with the other Northern Zapotec
languages or rather should be classified as belonging to the Valley Zapotec
subgroup, as has been mentioned to me by Pamela Munro (Munro p.c.). The authors
of the dictionary do not discuss the intelligibility of welab with any of the other
groupings listed.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
In general, the authors claim, outside of welab, that the lhe’ja dialect
grouping is the most unintelligible to speakers of other Sierra Zapotec language
(ibid. 18). They state that speakers of xhon, xidsa, and xan can communicate if need
be, especially speakers of xhon and xidsa, but only on a basic level.
The following map is significantly adapted from maps obtained from
www.maps-of-mexico.com and from ‘Mapa 2’ in Castellanos et al (1995). It shows
the geographic distribution of the language groupings discussed above.
Figure 1.2; Map of Language Groupings
Cametli
Sources: www-maps-of-mexico.com (5/1/2004), Castellanos et al. (1995)
However, as the authors of the dictionary note, these are not discrete
groupings. There is a great diversity within these dialect groupings where an
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
individual language might pattern like a member of another dialect grouping. For
example, within the zxon grouping, there is ‘ the tendency for Tabaa and Yojovi to
make (>retain 4a.s.) polysyllabic words which are characteristic of the xidsa and
Ihe’ja sectors, the tendency of San Pedro Cajonos, Betaza and Laxopa to use tz as
used in the xidsa and xan sectors, (and) tendency of Yalalag to use k as they do in
the xidsa and xan sectors’ (ibid. 20). They also note lexical differences, such as
pronominal forms. There are also significant syntactic differences between the
different zxon languages. For example, there are differences between Yalalag,
Yatzachi, and SBZZ, in the co-ocurrence of subject and object pronominal clitics, the
retention of pronouns in relative clauses (first mentioned in Marlett (1990,1993)),
plurality (Marlett and Picket 1999) and focus and topic constructions.
The nature of and causes for the dialect differentiation are interesting topics
for future research. It needs to be determined what exactly the situation is. There
could be a dialect chain, where one dialect fuses into the next without any clear-cut
boundaries. There could be discrete groupings in which individual members of the
grouping might pattern on some levels with members of another grouping. Perhaps
most likely, the situation is a combination of the two above, where, based on contact
with the nearest neighbor, speakers of one dialect grouping can understand the
dialect of their closest neighbors, and, to varying degrees, those of the rest of the
dialect grouping and perhaps close neighbors belonging to other dialect groupings,
but where, by and large, the groupings do constitute different languages.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
8
Depending on the reasons for contact, and the geographic proximity, any
two varieties will have differing degrees of mutual intelligibility. For example, as my
collaborators have repeatedly pointed out to me and has been reported in Ethnologue
(Grimes et al 2000), because of the weekly market in Zoogocho, speakers from
Yatzachi, who are forced to come to Zoogocho to buy products, can communicate
with Zoogocho Zapotec speakers. Speakers from Zoogocho, however, generally
report difficulty in understanding those from Yatzachi and report using Spanish. If
one examines the rate of mutual intelligibility among languages of the zxon dialect
grouping as reported in Ethnologue, one can see that there seems to be no one
particular factor (such as geographic proximity) which is the determining factor for
intelligibility. In addition, the access to transportation which no longer requires
people to travel to and through near-by villages in order to conduct business with the
outside world is in all likelihood aiding in the destruction of the social and linguistic
networks which used to exist between related dialects. A systematic study of the
dialectal differences based on geography, and also crucially on social networks and
other social factors is a necessary step in the further description of the language
family.
1.4 Previous linguistic research
In terms of work done on the zxon dialect, the earliest work done on the
dialect known to me was done by Jaime de Angulo (1926). In the 1940s, there was
work done by Eunice Pike (1949) and Mary and Otis Leal (1954). Later. Inez Butler
has done considerable work on Yatzachi to which we all, as Long points out, are
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
indebted. Marlett and Pickett, working both together and separately, have also
included data from various zxon dialects in various comparative papers. When I
began working on Zoogocho Zapotec, there were only two articles written on the
Zoogocho Zapotec variety (Long (1985), Butler (1985)). However, since that time, a
formidable dictionary Cruz and Long (1999) with a grammatical sketch has been
published. The current dissertation represents the most thorough grammatical
description of Zoogocho Zapotec to date.
1.5 Non-Zapotec languages spoken in the region
There are non-Zapotec languages spoken in relatively close proximity to
SBZZ, most notably Mixe and Chinantec. Because of the large amount of money that
is brought in from outside the village and outside of the region, San Bartolome
Zoogocho is relatively affluent in comparison with many of these other communities.
As a result of this, many Mixes and a few Chinantecs come to work as migrant
laborers in Zoogocho because of the much higher wages. In general, my primary
consultant’s husband, Eduardo Vasquez, a Mixe, has reported to me that, circa 2000,
Mixes made only 10-15 pesos a day without food being included if they work in the
Mixe region, whereas they will earn around 35 pesos a day plus meals, if they work
as day laborers in Zoogocho. Most of the Mixes and Chinantecs who 1 know do not
learn Zapotec, communicating in Spanish instead. However, some, who stay for an
extended period, such as Eduardo, do pick up some Zapotec (p.c. Eduardo Vasquez
2000).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1.6 Other geographic and ethnographic information
San Bartolome Zoogocho is situated on the side of a mountain in a pine scrub
forest. Its lands extend to the top of the mountain and down to the hot, dry valley
below. As a result of this wide range of microclimates, the cultivation of a great
variety of produce is possible. Peaches and pears can be local, as can mangos and
bananas. The mountainous terrain does make it difficult to have horses, or large
herds of cattle.
The primary economy in Zoogocho is agricultural. At this point, it is mostly
subsistence level farming. Coffee is frequently grown as a cash crop, and, in the past,
there were government programs encouraging the growth of silkworms as a potential
crop, but few people currently raise silkworms. Livestock are raised by some in the
village to sell. Most of the population, however, keeps livestock for personal use.
There are also some artisans involved in trades such as weaving, and pottery. There
are not as many artisans today as there once were, given the availability of cheap
imported products. Also, as the regional market town, with a weekly market on
Thursdays which draws people in from all over the region, there is also a thriving
mercantile sector. There are also a number of individuals who have thrived outside
of the village as engineers, teachers, sociologists, anthropologists, business people,
and so on.
In terms of the belief system of the average Zoogochense, it would be fair to
say that it was syncretic Catholicism; Catholicism which shows traits of indigenous
beliefs. Even though, as with many other traditions, such practices are on the wane.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
there are still various practices which predate the conquest. Many popular myths
involve the ‘lord of the woods’, or other mythical figures. There is also a fair amount
of talk of witchcraft.
There are very few that remember traditional forms of healing in all of its
forms. However there are still native healers who are widely used by both people
from the village and outsiders. Some beliefs persevere, such as that of ‘susto’ or
‘fright’, in which by being startled or frightened people, especially small children,
can become spiritually and physically ill. However, some of the elaborate rituals
which used to be performed in such cases are no longer performed, some abbreviated
version being performed if anything at all.
Similarly, many of the other traditions which still existed in the early part of
the twentieth century have also fallen by the wayside. See Julio de la Puente’s
description o f Yalalag (1977), to see a description of many of these.
Unfortunately, there are very few current comprehensive ethnographies of
San Bartolome Zoogocho or the near-by communities of which I am aware. There is
a great discussion of the market economy from the 1970’s written by Ralph Berg
(1974), a description of traditional medicine by Zoogocho native Filemon Beltran
Morales (1982) and a monograph about the Mountain Zapotecs as a whole and a
collection of essays about the region by Zoogocho native Manuel Rios (1994).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
12
1.7 Typological characteristics
Zoogocho Zapotec is a tonal, agglutinating VAO/VS language with noun
adjective, noun genitive, noun determiner, noun relative clause, and numeral noun
orderings. I will describe its typological characteristics in much greater depth over
the course of the dissertation.
1.8 Theoretical framework
For the most part, 1 will attempt to be as theory neutral as possible in the
actual description of the language. O f course, this is easier said than done. I will
attempt to define those terms which are unclear or vary considerably
crosslinguistically. I will strive to use definitions which correspond to the grammar
of Zoogocho Zapotec as much as possible while retaining some degree of cross-
linguistic applicability.
In a few chapters of the dissertation, I will present some arguments based on
the theory of grammaticalization (See Bybee et al. (1994), Heine et al. (1991),
Lehmann (1982), etc.). I will refer heavily to theories of word order typology such as
Greenberg (1963), Comrie (1989), and others; and will also appeal to the processing
approach to the explanation of linguistic universals as espoused in Hawkins (1994) at
various points in the dissertation. Overall, I intend the theoretical framework to be a
loose one, and am more concerned with the description of the language than the
promotion of theories.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 H enceforth referred to as SBZZ, and Z oogocho Zapotec. 1 will continue to refer to the language as Z oogocho
Zapotec, even though this is a name given from outside the speech com m unity because o f the fact that is specifies
the specific dialect, and dizha zxon has a w ider usage, also referring to the dialect grouping to w hich Z oogocho
Zapotec belongs.
2 In the I9 4 0 ’s, after either a plague o f locusts according to som e stories or as a result o f econom ic necessity
according to others, there was a governm ent sponsored m igration o f speakers from a num ber o f towns, including
San Bartolom e Zoogocho, to Nigrom ante, Veracruz. W hen there, the speakers reportedly form ed a sort o f koine
based on the various V illa A lta dialects. It is this researcher’s intention to som eday travel to N igrom ante to
docum ent and analyze this variety.
3 In the orthography used by the authors o f the dictionary, xhon corresponds to zxon in the orthography' used for
the purposes o f the present work.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 4
Chapter Two: Phonology/Phonetics
2.1 Orthography and inventory of sounds
The orthography which will be used in this dissertation, while sharing many
points of similarity with other orthographies in use for Zoogocho Zapotec and related
languages, is solely intended for the current dissertation. It is a practical orthography
originally devised by Terrence Kaufman and consequently revised by speakers of the
language and myself.
There are various other orthographies that have been devised for the language
grouping, including various orthographies devised by native speakers, and by the
Summer Institute of Linguistics. The only reason why I use this particular
orthography is that I am familiar with it and that it does adequately represent the
phonemes of the language.
In the following exposition of the orthography, I will write segments in the
orthography which I am using surrounded by slashes, and will put the International
Phonetic Alphabet symbol which corresponds to the phoneme in the orthography in
square brackets in those cases where there is a difference between the standard IP A
symbols and the current orthography.
There are five vowels in SBZZ. Note that these correspond to a four-vowel
system, with the addition of the vowel /u/ in loanwords. Doubled vowels represent
creaky voice /VV/ [Y], vowels with an /h/ between them represent breathy voice
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
15
/VhV/ [Y], and vowels followed by an apostrophe represent checked vowels /V ’/
[V2]. These phonation types will be discussed later in the current chapter.
Table 2.1 SBZZ Vowels
i (u)
e o
a
Phonation types: VV,VhV,V’
While it is noted that tone is indeed a very important part of the language and
is phonemic, I will only mark tone as it is relevant to the discussion of the
morphosyntax at hand. Tone is not marked in most native orthographies, instead
being recovered from context. Unfortunately, I have not completed a systematic
study of tone in all of its myriad aspects at this point. For the purposes of the
description of the phonetics and phonology of this language, and for later parts of the
dissertation I will represent the five tones which are tentatively identified (high, mid,
low, rising, falling) as in the IP A, with [v] corresponding to high tone, a plain [v]
corresponding to mid tone, [v] corresponding to low tone, [v] corresponding to
falling tone, and [v] corresponding to rising tone.
The following table represents the native phonemic consonantal inventory of
SBZZ.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
16
Table 2.2 SBZZ Consonants
Labial Alveolar Palatal RetroflexVelar
Fortis stops
P
t k
Lenis stops b d
g
Fortis affricate ch [tj]
Lenis affricate dx [d3]
Flap r [r]
Fortis fricatives s sh [f] x [g]
Lenis fricatives z zh [3] zx [zj
Fortis nasals m nh
Lenis nasal n
Fortis lateral lh
Lenis lateral 1
Approximates w
y[j]
Labiovelar Uvular Glottal
kw ' [V |
gh \v \
The following sounds are found in loan words from Spanish: /f/, /r/[r], /rr/[r],
/n/[r|], and /j/ [x]. /r/[r] also shows up in variation with the native phoneme /l/.
The fortis/lenis distinction in non-resonants is represented by the standard
voiced/voiceless symbols for stops. For resonants, the fortis version is marked with
an /h/ following the /l/ or /n/, as in /lh/ and /nh/.
/w/ can also represent labialization of the preceding consonant, as will be
discussed in greater detail below.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
17
2.2 Consonants
While I have represented the fortis/lenis distinction for the stops
orthographically as a voicing distinction (fortis being voiceless and lenis being
voiced), it is not the case that voicing is the primary distinction which is being made
in fortis/lenis pairs. As discussed in Ladefoged and Maddieson (1996: 98), Jaeger
(1983) ‘concluded that in two quite different languages she was investigating,
Zapotec and Djauan, the phonetic factors underlying the contrast were primarily
duration, glottal width and possibly closure width. She suggests that in both these
languages, the proto-typical fortis obstruent is long, voiceless, has no variation in
stop closure and has higher amplitude noise; the prototypical lenis obstruent is short,
usually voiced but often voiceless, has much variation in closure type, and lower
amplitude noise.’ The length of the preceding vowel is often greater for lenis
consonants. Lenis consonants can also become voiced fricatives intervocalically.
These observations might all follow from the fact that lenis consonants are said
above to have a higher variation in the type of closure and in voicing. I will discuss
the fortis/lenis distinction in sonorants in greater detail below. The exact
characterization of the distinction between fortis and lenis consonants in Zoogocho
Zapotec requires greater study.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
18
2.2.1 Examples
2.2.1.1 Stops
Table 2.3 Stops
Initial position Medial position Final position
¥
N o native ex. (see /kw/) la p a' ‘hat’ tap ‘lo u r'
/t / te ’grey’ gata ’ '1 will die.’ hat "when"
/k/ ke ‘ ‘really (tag question)’ dxaka’ ‘I can do ... ’ nak ‘how '
/b/ be ‘anim al classifier’ laba ’ ‘drop (n )’ Icont ‘base’
/d/ dehe ‘ash ’ xada ’ ‘flattened, sm ashed' y id ‘skin, h id e’
¥
g e ‘interrogative particle-w here o n ’ lh a g a' ‘le a f zahag ‘co ld ’
/kw/ kwan ‘edible herb’ none1 none‘
Lenis consonants are often devoiced word-initially, especially when
immediately followed by fortis consonants. At this point, I have only found
nonderived examples of /kw/ word initially. In all other positions, putative examples
of /kw/ all appear to be derived.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 9
2.2.1.2 Nasals
Table 2.4 Nasals
Initial position Medial position Final position
/ m / me ’ ‘baby sheep’ zh o m e ''basket' dam 'o w l’
Iml is fortis and has no lenis counterpart.
I n i nez ‘path, road’ bena ’ ‘1 m ade ... ’ Ion ‘to he hungry’
/nh/ nhe ‘n o ’ b en h a’ ‘Ig a v e ... ’ lonh 'long, tali'
One way of distinguishing between fortis and lenis nasals is by assimilation
in place of articulation of the lenis nasal to the following morpheme.
(2)
benhbe’ to yet
b-enh=be’ one tortilla
comp-give=3inf one tortilla
‘He gave a tortilla.’
bembe’ to yet
b-en=be’ one tortilla
comp-make=3 inf one tortilla
‘He made a tortilla.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2.2.1.3
/l/
/lh/
(3)
(4)
2 0
Laterals
Table 2.5 Laterals
Initial position Medial position Final position
l e ’ ‘yo u ’ nile ‘this w ay’ y e l ‘sandal'
Ih e' ‘him /her’ nil he ' cooked co m ’1 belli ‘fish’
The lenis lateral often turns up in free variation with the flap as seen below
da lis
clinan small
‘small’
da ris
clinan small
‘small’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2.2.1.4 Fricatives
Table 2.6 Fricatives
Initial position Medial position Final position
/si setw ‘squah plant/vine’ dxasa ’ i gel up.’ dxas ‘fall v itr.'
/z/ zee ‘w all’ d x a za' ‘I p la n t... ' dxaz ‘begin (aux)'
/sh/ shi ‘or, i f g w a sh a' ‘1 left.’ yash 'lo o se'
/zh/ zhe ‘night’ dxonlazha‘ 1 am lying.’ yazh ‘plum '
/x/ xeene ’ ‘saliva’ yixe ’ ‘woods, w ild’ hex ‘tomato"
/zx/ zxen ‘big ’ dxazxa ’ ‘1 am calling ... ' yizx ‘grass'
/gh/ ghed ‘chicken, h en ’ w e g h e' ‘each one’ begh 'sp rin g '
The retroflex/non-retroflex distinction in the fricatives is sometimes
accompanied by a slight rhotacization of the vowel preceding the retroflex fricative.
2.2.1.5 Affricates
Table 2.7 Affricates
Initial position Medial position Final position
/eh/ che ‘of, poss’ bichia ’ ‘my child lhach ‘even though'
/dx/ dxee ’ ‘ She says’ x la d x a' ‘my clothes’ nadx ‘then'
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2.2.1.6 Glides
Table 2.8 Glides
Initial position Medial position Final position
/w/ w e g o ’ ‘youth’ zxaw ed ‘stick insect' NA
/y/ yeg o ‘river’ lh a y e ' ‘green onion’ lim y ‘cichicastle’
/w/ can represent a [{I] or even an [f] in some speaker’s varieties, /y /is preaspirated
by some speakers, even becoming a [9] before /i/.
2.2.1.7 The glottal stop
Table 2.9 The glottal stop
Initial position Medial position Final position
/’/ no initial ble ’ ida ’ ’I see’ l.ha ’ ‘O axaca-
While I am currently representing it as a separate symbol, 1 should note that
the glottal stop is a feature of the vowel as it does not occur word-initially and it does
not occur directly following consonants.
2.2.2 Loans
Sounds which are only found in loan words in SBZZ are: /f/, /n/, /rr/, and /j/
as seen below in (5). /r/ is primarily found in loans, although, as mentioned above, it
does alternate with the native lenis lateral in some words.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(5) febrer ‘February [Sp.]’
kafe ‘coffee [Sp.]’
rranshw ‘ranch [Sp.]’
karrw ‘car [Sp.]’
jarrw ‘pitcher [Sp.]’
2.2.3 Labialization
Labialization occurs in environments where historically, a labial vowel had
occurred. Words that had the shape CYCV, with stress falling on the penultimate
syllable often reduced the first vowel. If the first vowel was labial, this resulted in
labialization of the first consonant, as in the following.
(6) gwbizh ‘sun (from inf.dry)’
Also, there are examples, especially in the case of Spanish loan words, where post
tonic labial vowels are the source of labialization, as in (7)
(7) platw ‘plate [Sp.]’
Labial consonants are not further labialized.
There are also aspectual prefixes that become labialized in certain
environments. This will be discussed in greater detail in Chapter Four.
2.2.4 Consonant clusters
Initial consonant clusters are generally of the shape, CCV. Consonant cluster
generally come from earlier CYCV morphemes in which the pretonic vowel has
been lost. There are also cases where an earlier prefix has lost a vowel resulting in a
CCV structure. If a consonant cluster begins with a stop, then the following
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
consonant could either be a fricative as in (8), a resonant, as in (12), a glide as in (9),
or potentially another stop. I will discuss the case of when it is a stop-stop cluster
following the examples. By far, stop-resonant clusters are the most frequently found
stop initial clusters in the lexicon.
(8) bsia ’ ‘eagle’
(9) byinhe ’ ‘bird
(10) bzhina ’ ‘ foam ’
(11) Dmingw ‘Sunday [Sp.]’
(12) glan ‘spotted’
(13) platw ‘plate [Sp.]’
All examples of initial clusters beginning with /t/ will be described in the following
section.
(14) kleka ’ ‘than (used in comparative constructions)’
In those cases of stop-stop clusters, there are a few explanations. As with all
other clusters, they all appear to come from lexemes which historically had the shape
CVCV, where the initial CV has simplified to a C, or, in the case when the V was
/u/, a CW. The original CV seems to have frequently been another morpheme
diachronically, especially be ‘animal classifier’ for roots beginning with /b/, or to
‘one’ for all of the roots beginning with /t/. There are also cases where loan words
from Spanish have been reduced as well. There is often either a bit of aspiration
following the initial stop (a reflex of its syllabic origin), and, as already mentioned,
labialization, when the original vowel was /u/ or /o/.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 5
(15) bdao 4 ‘baby’
(16) bgope ’ ‘armadillo’ (from be=gope)
(17) gden ‘chain [Sp.]’ (from Spanish cadena)
(18) gwbizh ‘sun’ (from inf.dry)
(19) tdia’ ‘one generation’ (from to+dia’)
There are some examples of affricate initial clusters, but these are all verbs, where
the initial affricate is a continuative aspect marker as in (20). (When the following
consonant is a fortis consonant dx- becomes sh-.) Affricates are not otherwise part of
consonant clusters.
(20) dxbab libren ’
dx-bab libr=en’
cont-be.counted book=definite marker
‘The books were counted.’
Resonant-stop order is also found.
(21) lhbaha‘v ine’
(22) Itonh’ ‘appetite’
(23) mba’ ‘happy’
(24) nga ‘demonstrative (medial)’
Fricative-stop order is also frequently encountered. Note that gh does not occur
initially in consonant clusters.
(25) steb ‘stale coffee’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(26) shbal ‘dawn’
(27) zxdan ‘pretty, beautiful’
(28) zban ‘ugly/bad’ (most o f the clusters which begin with /z/ are either
adjectives or adverbs)
(29) zh ta ’ ‘W
There are also resonant-resonant clusters, as in the following.
(30) Inha ’ ‘bright (color)’
(31) nlha ’ ‘to be seen’
Resonants more frequently begin clusters with either following stops, as seen
above, or fricatives, as in the following.
(32) Ighezh ‘reciprocal pronoun’
(33) Ihshil ‘hallway’
(34) Ihzxozh ‘dangerous’
(35) nzha' ‘different’
/l/ can also occur before an affricate, as in the following.
(36) Ichegho ’ ‘your skirt’
(37) Ichelh ‘rotation’
The stative marker n- can also occur before many consonants, assimilating in place
to the following consonant, as in the following.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 7
(38) mbane ’
n-ban=e’
stat-live=3f
‘He lives.’
Fricatives can occur before stops, as seen above, before resonants, as seen
immediately below, and before other fricatives and before glides, as 1 will show
later in this section.
(39) shlaa ‘side’
(40) shlak ‘while’
(41) shnegh ‘once’
(42) zxnha’ " red’
(43) zxlapa ’ ‘my hat’ (There are many examples with either /x/ or /zx/ preceding
a consonant in their use as possessive prefixes.)
Fricatives can also occur before glides, as in the following.
(44) vxwaye ’ ‘malamujer (plant species)’
(45) xya ‘damage’
(46) zw ia’ ‘cocoa bean’
(47) ixwaga ’ ‘my firewood’
(48) xwe ’ ‘lunch’
Fricatives may also precede other fricatives. They do not precede fricatives o
the same place of articulation.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 8
(49) sshaa ‘hot’
(50) sshag ‘noisy’
(51) shgheza ’ ‘cacomixtle’ (cat or raccoon like mammal)
(52) zxgheda ’ ‘my hen’
(53) xsil ‘breakfast, morning’
(54) z_xi ‘salty’
Example (54) shows a potential problem with the orthography currently
being used and with other similar orthographies, in that it represents a cluster of two
fricatives and not a single voiced retroflex fricative. Where necessary 1 will represent
the fact that there are two phonemes by placing an underline between. While glides
do not normally occur as the initial consonant in a consonant cluster, there is one
notable exception.
(55) ygho ‘the area behind a house’
I will now turn to syllable final consonant clusters. Most syllable final
consonant clusters are only two consonants in length. There is nowhere near as wide
a range of potential consonant clusters as initially. The most common consonant
cluster final segment in syllable final consonant clusters is /gh/. Stops may show up
after a resonant, fricative, or glide as in the following.
(56) bant ‘sash [Sp.]’
(57) gasgh ‘black’
(58) bolgh ‘egg yolk’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Stops may also occur before fricatives, though this seems to primarily be restricted to
the uvular fricative.
(59) guditgh Lalo ‘Lalo played.’
(60) yapgh ‘chayote(edible plant)’
Affricates can also occur before /gh/.
(61) yichgh ‘head’
Fricatives can generally only occur before /gh/, as in the following.
(62) gasgh ‘black’
Resonants can occur before /gh/.
(63) bechgh ‘pottery figurine’
While it may appear that there are consonant cluster final glides in syllable
final consonant clusters, these are always the result of labialization.
(64) setw ‘squash vine’
(65) bankw ‘bench [Sp.]’
(66) belghw ‘cloud’
Word medial consonant clusters are viewed in the current work as either
being syllable initial clusters, syllable final clusters, or combinations of both.
(67) dxzozxghbizxghwa’ ‘I crumble s.t.’
Example (67) can be viewed as being.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 0
(68) dx-zozxgh-bizxghw=a\
cont-tear-small= 1 sg
‘I crumble something.’
2.3 Vowels
Zoogocho Zapotec has a five vowel system synchronically, with /u/ only
occurring in loans. Unlike Yatzachi Zapotec, SBZZ does not have phonemic schwa.
With the exception of the word for year, iz (which could be interpreted as being
underlying y/z), there are no native vowel initial words. Some examples of minimal
and near minimal pairs for vowels follow.
Table 2.10 Vowels
/a/ ba ‘to m b ’ y a a ‘raw, green, unripe’
Id be ‘classifier-anim al’ yee ‘excrem ent’
/i / bi ‘negative’ yiinh ‘coffin’
lot bolhgh ‘y o lk ’ y o o ‘house’
There are also the following sequences of vowels:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 1
(69) /ao/ bao’ ‘charcoal’, dao ‘maguey’
/ai/ gwbaai ‘broom’
/ae/ none attested
/ea/ ghea ‘rooster’
/ei/ beida’ ‘I sense ... ‘
/eo/ shgheondapart ‘You will go to sue me.’
/ia/ lia/ria’ ‘Ms.’
/ie/ chie ’ ‘his’
/io/ chio ’ ‘your’
/oa/ dxoabe ’ ‘his mouth’
/oe/ dxoe ’ ‘he gave’
/oi/ moises ‘Moses [Sp.]’
/o/ is sometimes pronounced as a /u/, apparently in free variation. As
mentioned above, lenis consonants sometimes tend to lengthen the previous vowel.
This appears to be the only case whereby vowels are long. The vowels themselves do
not appear word-initially, although they can occur word finally.
There are also cases where vowels are nasalized. This only occurs in vowel-
final stems when they are followed by the definite article, -n 7-na Vowels are also
slightly nasalized preceding regular nasal segments.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 2
2.3.1 Phonation
Vowel quality may be creaky, breathy, checked, or regular. The following
are the ways the various types of phonation types are being represented. V'
represents a checked vowel (such as ya \ 'reed'), VV represents creaky voice (such as
yaa 'TemazcaT (steam bath)), and VhV represents a vowel with breathy voice (such
as yaha, 'iron, rifle').
2.3.2 Tone
As mentioned above, there are three level tones in SBZZ and one rising tone
and one falling tone. There also appears to be tone sandhi, especially with the
personal prefixes. The first person clitic -a' appears to raise the tone of the preceding
root, for example. This is something which needs to be more systematically
examined in the future.
There is not enough information about the interaction of phonation type with
tone to say that one is predictable based on the other. It has been claimed for other
Zapotec languages that tone and phonation type are linked and it has even been
tentatively claimed for Villa Alta Zapotec (Pike 1946). It is well known that
consonants can influence tone. Jean-Marie Hombert in on “Consonant Types, Vowel
Quality, and Tone" Hombert in Fromkin (1978) states that, "In the Lolo-Burmese
family, Burmese high tone corresponds to the Jingpho glottal stop ... and Lahu high
rising tone developed through glottal dissimilation ... (and that) the development of a
falling tone from a post vocalic [h] has been observed in two cases (Vietnamese and
Middle Chinese)” (Hombert 93). While breathy voice might have a lower tonal
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
33
quality, and checked vowels, a higher tonal quality, it appears that they are distinct
phenomena and that tone is not predictable from phonation type. This is not entirely
decided and it will be interesting to determine how they do influence tones, even if it
turns out that they are distinct. The following set shows examples of words with the
same phonation type and different tones and vice versa.
(70) yaa 'plaza'
(71) yaa 'hill'
(72)
Z f
ya
‘reed’
(73) yaha 'weapon'
(74) yahd 'temascaf
(75) zhaha 'day'
A systematic analysis of tone and phonation type is of the utmost urgency.
2.3.3 Intonation
In addition to tone, there is a rising intonation used for yes/no questions.
2.4 Syllable structure
Syllable structure in Zoogocho Zapotec is as follows. The basic syllable is
generally of the shape:
(76) (C)CV(V)(C)(C)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 4
There are no restrictions as to the type of phonation on the vowels in the
schema above.
2.5 Stress
Stress in Zoogocho Zapotec primarily falls on the penultimate syllable of the
stem, except for those cases where a word is composed of more than one root
(incorporated noun or adverb) in which case the accent falls on the second root.
Here, the stressed syllable is marked in bold.
(77) nilhe ‘nixtamal’
(78) yeten’ ‘the tortilla’
(79) zoalawa’ ‘I begin.’
2.6 Phonological word
Dixon and Aikhenvald (2002) offer the following definition of phonological
word, noting that there was no single criterion among the following which would
apply to all languages:
A phonological word is a unit larger than the syllable (in some languages it
may minimally be just one syllable) which has at least one (and generally
more than one) phonological defining property chosen from the following
areas:
(a) Segmental features - internal syllabic and segmental structure;
phonetic realizations in terms of this; word boundary phenomena;
pause phenomena.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
35
(b) Prosodic features - stress (or accent) and/or tone assignment;
prosodic features such as nasalisation, retroflexion, vowel harmony.
(c) Phonological rules - some rules may only apply within a
phonological word; others (external sandhi rules) apply specifically
across a phonological word boundary, (ibid 13)
All phonological words in Zoogocho Zapotec consist of at least a CV(C(V))
base. The stress is on the penultimate syllable. The place assimilation to the
following consonant which lenis nasals undergo does not occur across word
boundaries.
(80) dxombe’ to
dx-on=be’ to
cont-make=3.inf one
‘He made a house.’
(81) dxon bedo’ to yoo
dx-on Bedo’
cont-make Pedro
‘Pedro built a house.
In addition, phonological words can have a great deal of prefixed materials,
as will be seen in 4.5. These prefixed materials vary depending on the conjugation
and do not alter the stress placement. There are some adverbial suffixes in the
language. These also do not affect the stress placement.
yoo
yoo
house
to yoo
one house
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
36
There are a number of enclitics in the language, such as the pronominal clitic
seen above. While phonologically these are not words, grammatically they behave as
independent words, as will be discussed in 4.5.
1 There are som e exam ples which show m edial /kw/, but these, such as zxikwa ’ are alw ays the result o f
labialization before a vowel.
2 O nce again, word-final /kw / is the result o f the labialization o f a /k/ where a word final labial vowel had been
lost, as in yegh sikw ‘onion’.
3 For corn dough for tortillas [Sp. nixtamal].
4 This com es from the SBZZ w ord bidao ‘ch ild ’. The reduction indicates that it is a very small child, or a baby.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 7
Chapter Three: Pronominal forms
3.1 Introduction
In this chapter, I will describe the pronominal elements in SBZZ. I will look
at a small selection of what has been claimed for other Zapotecan languages with
respect to pronominal clitcs/suffixes, and come to an initial morphological
classification of these elements. This is an issue which has received much attention,
and I hope to shed new light on the subject, or at least clarify the issue for Zoogocho
Zapotec. While parts of this chapter could potentially be part of the chapter on
lexical classes and other parts in the morphology chapter, I feel that, given their
central role in the grammar of SBZZ, pronominal forms deserve their own chapter.
3.2 Independent pronouns
The following are the independent pronouns.
Table 3.1 Independent pronouns
Singular Plural
1 (excl.) neda’ neto
1 (inch) dxioo
2 lee (loo) le
3 (formal) Ihe’ lhegake’
3 (informal) lhebe’ lhegakbe’
3 (animals) lheba’ lhegakba’
3 (inanimate) lhen lhegaken
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The first person inclusive includes the second person, which is excluded in
the exclusive form. Independent pronouns are used in isolation, as subjects (quite
often preverbally as in (1)), objects of predicates in cases where a dependent
pronominal form would be infelicitous (10), such as when the subject is a full noun
phrase or is of the identical person or a person lower on the person hierarchy1 , and
objects of prepositions which cannot take clitic forms2 (15).
(1) neda’ b-shaaga-na=a’
1 sg comp-j oin-hand= 1 sg
‘I got married.’
(2) gakate gd-izxghw=e’ lee
when comp-pay=3f 2sg
‘When did they pay you?’
(3) na’ gon sh-da-lenh=be’ lh=e’
and no_more cont-walk-with=3 inf base=3f
‘She only walks with him.’
(4) to-z lhe=be’ n-de-kse
one-emph base=3inf stat-lie-emph
‘Only one is lying down.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
39
(5) kate’ g-zoa fayo che lhe=be\
when comp-stand error of base=3inf,
lhe=be’
base=3inf
‘When he messes up, I will admonish him.'
(6) na’ zhia to benhe
and stat.sit one person
‘And a person is sitting on it.’
(7) na’ lhe=n gu-zh=be’
and base=3inan comp-say=3inf
‘And tell him this now ... ’
(8) no b-en lhe=n
who comp-make base=3inan
‘Who made that there?’
(9) na g-ak-lenh=fo=be’
and pot-be-with=lplexcl=3inf
‘And we will help him.’
(10) dx-os-tee-shke=ba dxioo
cont-caus-tire-emph-3an lplincl
‘It makes us very tired.’
dx-shash-lenh=a’
cont-complain-with=l sg
lhe=ba’
base=3an
na’
now
na’
demdist
neto
Iplexcl
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4 0
(11) gu-yaa-lenh=a’ le
pot-dance-with= 1 sg 2pl
‘I’m going to dance with you all.’
(12) na ka y-eyozh g-oso-yaa
and demadv pot-finish pot-pl-dance
‘And like that they finish dancing.’
(13) gage lhe-gak=be’ g-oso-a=be=nda’
neg base-pl=3inf comp-pl-carry=3inf=lobj
‘They aren’t the ones who took me, you know?’
(14) ga dx-ogh lhe-gak=e’
where cont-exit base-pl=3inf
‘Where do they leave?’
(15) dxi to be’ko’ trasde lhe-gak=en’
stat.sit one dog behind base-plural=3inan
‘A dog is sitting behind them. (A group of trees.)’
The form for the second person singular in the chart above in parentheses,
loo, while reportedly used more generally in other towns such as Betaza, is said to be
something which would be used primarily by men by my primary consultant. I can
confirm that I have also only heard men use this form in Zoogocho. Men and boys
also refer to each other using the form normally reserved for animals, Iheba
Ihe-gak=be’
base-pl=3 inf
ke
really
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4 !
(16) lez kate weghe g-saka che=le=n\ ke, loo
equal when each one comp-cost of-2pl=det tag, 2sgm
‘All of your all’s cost the same, didn’t they, you?!’
(17) lheba’ n-ak bixag
base=3an stat-be deputy
‘He’s the deputy.’
While speech act participant pronouns are pretty clearly monomorphemic, the
forms for the third person require explanation. As described in Pickett and Marlett
(1999) and Marlett (1993), the third-person independent forms in most Zapotec
languages are constructed out of a pronominal base with the dependent form attached
as an enclitic. In SBZZ, the form of the base is Ihe-. -gak- can also appear by itself
as a plural marker. In the following table, I have segmented these independent forms.
Table 3.2 Forms for the third person
3 (formal) lhe=e’ lhe-gak=e’
3 (informal) lhe=be’ lhe-gak=be’
3 (animals) lhe=ba’ lhe-gak=ba’
3 (inanimate) lhe=n lhe-gak=en
3.3 Dependent forms
The following chart shows the dependent pronominal forms used for
subjects of verbs, possessors of nouns (in possessive constructions), objects of
relational nouns, objects of prepositions, adjectives, and quantifiers. The following
chart shows these forms. I will return to the question of whether to consider these
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4 2
forms to be clitics or affixes later in the current chapter in section 3.8 and also, from
a perspective based on a processing analysis of quantitative textual data, in 8.7.
Table 3.3 Dependent pronouns
Singular Plural
1 (excl.) =a’ =to
1 (inch) =dxo
2 =o’ =le
3 (formal) =e’ =(gak=)e’
3 (informal) =be’ =(gak=)be’
3 (animals) =ba’ =(gak=)ba’
3 (inanimate) =(e)n =(gak=)(e)n
The vowel in parentheses for the third person inanimate forms does not occur
following vowel final roots. The morphophonology of the pronominal forms and
certain stem changes that occur to verbal roots with the first and second persons will
be discussed briefly in 4.3.12. The forms used for the third person are the same in the
singular and the plural in most instances. Plurality, as will be seen in later chapters,
is a sporadically marked category in Zoogocho Zapotec, and the forms with T gak
are often not seen, as discussed above. What follows are examples of the dependent
subject forms. I have put in bold and in parentheses the part of speech to which the
dependent forms are attached.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(18) zoalao=dxo (Verb)
stat.begin=lplincl
‘We begin.’
(19) sh-daa=tont=o’ (Verb)
cont-walk=fool=2sg
‘You go around like a fool’
(20) gasgh=ba’ (Adjective)
black=3an
‘It is black.’
(21) to=be’ (Quantifier)
one=3inf
‘one of them’
(22) che=be’ (Preposition)
of=3inf
‘his/her’
(23) lao=to’ (relational noun/body part)
eye=lplexcl
‘our eyes/in front of us’
(24) chizxghw=a’ (noun)
poss.tortilla=lsg
‘my tortilla’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4 4
The following is an example of the plural marker - gak= in use. This is the
only occurrence of -gak= not used in an independent pronoun found in a corpus of
over 2000 clauses.
(25) shi nhe=e’ y-ese-zi-ks=e’ che=gak=be’
if pot.say=3f pot-pl-buy-emph=3f of=pl~3inf
‘If theyj say what theyi bought from them,...’
3.4 Experiencer and instrumental forms
What follows is the set of dependent pronominal forms used with a particular
subset of verbs which either have experiencer subjects or are the causativized version
of one of these verbs. These verbs include -en ‘hear’, -een ‘want/desire’, -le 7 ‘see’. -
Ihei ’to show’ -lan ’tener asco/feel sick’3 , -eghnii ‘ understand’-zeghnii ‘teach how
to do something’, -ak ‘ be or be able to do something, when used in the sense of, to
believe’, -nez ‘ know’, -zhel ‘find\-g e e ‘ to hate s.o. or s.t.'-soozh ‘be drunk’, -
zhee ‘feel hot’, -zue ‘be capable of enduring, stand up to’, -sed ‘to teach’, -Ihee ‘to
smell’ and probably many others4.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4 5
Table 3.4 Experiencer forms
Singular Plural
1 (excl.) =da’ =to
1 (inch) =dxo
2 =do’ =le
3 (formal) =de’ =(gak=) e’
3 (informal) =be’ =(gak=)be’
3 (animals) =ba’ =(gak=)ba’
3 (inanimate) =(e)n =(gak=)(e)n
These forms can also be used to introduce a new argument, most often an
experiencer, as seen in the following examples.
(26) sh-lhee=be’ kushin
cont-smell=3inf piggy
‘He smells bad.’
(27) sh-lhee=da’ to be’ko’
cont-smell= 1 sgexp one dog
‘I smell a dog.’
There are also phonologically identical forms which are used to introduce an
instrument5. The experiencer subject form is used in (28) and the instrumental
marker in (29). Unfortunately this instrumental marker only shows up with the first,
second, and third formal singular, exactly as with the experiencer form, making it
difficult to distinguish between the two.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4 6
(28) b-le’i=da’ bidao lizh=a’
comp-see=lsgexp child poss.house=lsg
‘I saw a child in my house.’
(29) b-en-d=a’ lizh=a’ martiyw
comp-make-instr=lsg poss.house=lsg hammer
‘I made my house with a hammer.’
One potential way of distinguishing between these two constructions is
afforded by the reflexive-of-possessor6 construction. Example (30) shows a normal
clause which uses the experiencer form of the subject. In (31) through (33), the
reflexive-of-possessor construction allows the subject to be expressed as the
possessor of the object, but does not allow any evidence of the experiencer form to
show up in any of the places where it might be expected to, as seen in (34)-(36).
(30) b-le’i=da’ lish=a’
comp-see= 1 sgexp poss.paper= 1 sg
‘I saw my paper.’
(31) b-le’i lish=a’
comp-see poss.paper=lsg
‘I saw my paper.’
(32) dx-eene kuin=a’
hab-like refl=lsg
‘I like myself.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4 7
(33) b-le’i kuin=a’
comp-see self=lsg
‘I see m yself
(34) *ble’id lisha’
(35) *ble’i kuinda
(36) * ble ’ id kuinda’
So, if there is no pronominal marking on the verb, there is no evidence of the
experiencer form. In the following one can see that, unlike the cases with the
instrumental, the -d= in the instrumentals can be separated from the pronominal
clitic, which enables one to distinguish between the experiencer and the instrumental.
(37) b-en-d lizh=a’ martiyw
comp-make-instr poss.house=lsg hammer
‘I made my house with a hammer.’
In fact, if the d is not left behind with the verbal root in these cases as in (39)
it turns out to be ungrammatical as in, it turns out to be ungrammatical as in (40) and
(41).
(38) b-en-d=a’ lizh=a’
yag
comp-make-instr= 1 sg poss.house = 1 sg wood
‘I made my house with wood.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4 8
(39) b-en-d lizh=a yag
comp-make-instr poss.house=lsg wood
‘I made my house with wood.’
(40) *ben lizha yag
(41) *ben lizha’ martiyw
Neither the experiencer form nor the instrumental form can be used as the
possessor of an object as seen above in (36) and below in (42).
(42) *ben lizhda’ martiy
3.5 Dependent pronominal forms for the object.
The following are the dependent pronominal forms used for the object. For
the most part, they are identical to the regular dependent forms with the following
exceptions. The first and second person forms are fast speech forms of the full forms
of the pronoun and perhaps should not be included with the other forms on this chart,
as they are reflexes of the person hierarchy which will be discussed later in the
current chapter7. The third person formal form might potentially be analysed as
showing the only real case marking in the language. However, as all of the other
subject and object forms begin with either a glottal stop or with a vowel, the nasal in
this form could be seen as being epenthetic. These forms are all used after dependent
pronominal subjects. The examples which follow the chart also show how these
markers do not show up if the corresponding object is present. They also do not
occur if the subject is not a dependent pronominal form. These forms can be used for
either indirect or direct objects.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 3.5 Dependent object forms
Singular Plural
1 (excl.) =nda’ =nto
1 (incl.) =dxo
2 =le’ =le
3 (formal) =ne’ =(gak=)e’
3 (informal) =be’ =(gak=) be’
3 (animals) =ba’ =(gak=) ba’
3 (inanimate) =(e)n =(gak=)(e)n
(43) b-i=a’ bidao’ to libr
comp-give=lsg child one book
‘I gave the child the book.’
(44) b-i=a=n bidao’
comp-gi ve= 1 sg=3 inan child
‘I gave it to the child.’
(45) b-i=a=be’ to libr
comp-give= 1 sg=3 inf one book
‘I gave her a book.’
(46) b-i=a’ to libr thebe’
comp-give=lsg one book 3 inf
‘I gave a book to her.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(47) *b-i=a’ to libr=be’
comp-give= 1 sg one book=3 inf
‘I gave a book to her.’
(48) b-i=a=be=n
comp-give= 1 sg=3 inf=3 inan
‘I gave it to him.’
(49) b-i=a=ba=be’
comp-gi ve=l sg=3an=3inf
‘I gave it to her.’
As seen in examples (48) and (49), the order for the pronominal elements is V-
SUBJ-DO-IO.
3.6 Emphatic pronouns
There are also forms which are used emphatically following an identical
clitic pronoun or in-lieu o f a full form for an object pronoun when the clitic is not
allowed because of the person hierarchy. These are only used for the first and second
person singular forms. The use of these forms will be discussed in Chapter Fight.
(50) bibi b-en=o’=le’ bi g-ot=o’
neg comp-make=2sg=2sgfsf gensm pot-sell=2sg
‘You didn’t make what you were going to sell.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
51
(51) na b-id= a-nda’
and comp-come=lsg=l sgfsf
‘and I came’
(52) ba b-elha=a'=nda' na
already comp-arrive= 1 sg= 1 sgfsf there
‘ I already arrived there.’
(53) pero ka dx-ak=da’=nda’ nool chi=e’=n’
but so hab-be= 1 sgexp= 1 sgfsf woman o f; 3Ldel
ba gw-et
already inf=die
‘But I think that his wife was already dead.’
(54) dx-bezh=ks=a’=nda’ lao dio
hab-cry=emph= 1 sg= 1 sgfsf eye/face god
‘I cried in front of God.’
(55) g-on=a’=nda’ danh
pot-make= 1 sg= 1 sgfsf geninan
‘I will make it.’
3.7 Co-occurrence restrictions between subject and object clitics
I include here three tables, showing the co-occurrence restrictions between
subject and object clitics in Yalalag Zapotec, Yatzachi Zapotec, and Zoogocho
Zapotec. The first chart shows the co-occurrence of subject and object clitics in
Yalalag Zapotec, as reported in Lopez and Newberg (1990: 9). It shows the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
following person hierarchy where 1st and 2n d person singular objects cannot appear
as clitics with any subject, and must instead show up as independent forms. Third
formal objects cannot occur as clitics with third informal, animals, or inanimate
subjects. Third informal objects cannot occur as clitics with third animal or
inanimate subjects. Third animal object clitics can show up with everything except
third animal and inanimate subject clitics, and, finally third inanimate object clitics
can show up in every case except when the subject is a third inanimate. In all cases
where the object clitics cannot occur, a full form of the pronoun must be used. In alt
of these charts, an X is used to indicate that a particular combination of subject and
object clitics is not allowed. (I have converted their U to dx, 1 to Ih, and n to nh.)
Table 3.6 Yalalag Zapotec co-occurrence restrictions
1,2 3f 3 inf 3 an 3inan
lsg X -a’-e’ -a’-be’ -a’-ba1 -a’-nh
lplincl X -dxo-e’ -dxo-be’ -dxo-ba’ -dxo-nh
lplexcl X -to-e’ -to’-be’ -to’-ba’ -to'-nh
2sg X -o-e’
I
1
cr
-o ’-ba’ -o ’-nh
2pl X -lhe-e’ -lhe-be’ -lhe-ba’ -Ihe-nh
3f X X
l
cr
-e’-ba’ -e’-nh
3 inf X X X -be’-ba’ -be’-nh
3 an X X X X -ba’-nh
3inan X X X X X
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
53
In the second chart, which shows the co-occurrence restrictions as shown in
Yatzachi Zapotec by Butler (1980: 175-179), a slightly different story emerges, with
the differences being that, with the exception of the first and second persons which
cannot occur as object clitics, third person animal and third person inanimate subject
pronouns, which cannot co-occur with any of the object clitic pronouns, and cases
where the subject and object are of the same form and must be dissimilated, one can
pretty much combine subject and object clitics willy-nilly. The third person formal
object form in Yatzachi, like the corresponding form in SBZZ is =«e ’ and as such is
different from the subject form and does not block the combination, unlike the form
for the 3rd informal object which is identical to the subject form and therefore blocks
the combination.
Table 3.7 Yatzachi Zapotec co-occurrence restrictions
1,2 3 f 3 inf 3 an 3inan
lsg X Ok Ok Ok Ok
lplincl X Ok Ok Ok Ok
lplexcl X Ok Ok Ok Ok
2sg X Ok Ok Ok Ok
2pl
X Ok Ok Ok Ok
3 f X Ok Ok Ok Ok
3inf X Ok X Ok Ok
3 an X X X X X
3inan X X X X X
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
In Zoogocho Zapotec pretty much the same pattern as in Yatzachi Zapotec
emerges with the sole exception being that 3rd singular animal subjects do not
disallow the object clitics. Other than the person hierarchy which disallows first and
second person dependent pronominal forms to be used as objects with third person
dependent pronominal subjects, the other combinations on the chart which are
disallowed are instances of dissimilation where two identical morphemes would
otherwise co-occur.
Table 3.8 Zoogocho Zapotec co-occurrence restrictions
1,2 3 f . 3 inf 3 an 3inan
lsg X =a=ne’ =a=be’ =a=ba’ =a=n
lplincl X =dxo=ne’ =dxo=be’ =dxo=ba’ =dxo^n
lplexcl X =to=ne’ =to=be’ =to=ba’ t o n
2sg X =o=ne’ =o=be’
* C 3
_o
I I
o
l i
=o=n
2pl X =le=ne’ =le=be’ =le=ba’ =r|e=n
3 f X =e=ne’ =e=be’8 =e=ba’ Ic 1 1
3 inf X =be=ne’ X =be=ba’ be n
3 an X =ba=ne’ =ba=be’ X ba~n
3inan Xs * X X X X
3.8 The classification of the forms discussed above
Stephen Marlett (1993) classifies Zapotec pronouns into three major types;
prosodically independent pronouns, syntactically independent pronouns, and
syntactically dependent pronouns. He begins the paper by mentioning the wide range
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
55
of terminology used in discussions of Zapotec to describe the personal pronouns,
noting that linguists have traditionally described only two types.
Prosodically independent pronouns ‘occur(s) in one (or more) of three
positions: (a) in isolation, as a simple utterance, such as in answer to a question; (b)
preverbally without a phonological host; [and A. S. ] (c) as object of a Spanish
preposition’(Marlett 83). These correspond to the independent pronouns in
Zoogocho Zapotec described above.
He then goes on to discuss ‘syntactically independent pronouns’. His main
criterion is that the ‘pronoun may occur in object of verb position FOLLOWING A
NONPRONOMINAL SUBJECT’ (ibid 88) (emphasis mine). In Zoogocho Zapotec.
all of the prosodically dependent pronouns are also syntactically dependent, so this is
not relevant to the current discussion.
The next type which he looks at, which are incredibly important to the
current discussion are, the syntactically dependent pronouns. Marlett claims that all
third person pronouns in Zapotec are prosodically dependent. He asserts that these
pronouns are the unmarked type of pronoun. He maintains that syntactically
dependent pronouns are like French clitic pronouns, a paradigm case of a special
clitic in the terms of Zwicky and Pullum (1983).
He goes on to state that ‘(F)or the most part, pronouns and noun phrases
which are coreferential are mutually exclusive in Zapotec’ (ibid. 91). In SBZZ, as the
following examples show, this is not the case. While pragmatically marked,
pronouns and coreferential noun phrases can co-occur, especially in copular
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
5 6
sentences, as in the following example. This issue will be discussed in greater detail
in Chapter Eight.
(56) na’ n-ak=ba’ to kabayw shish
and stat-be=3an one horse white
‘And there was a white horse.’
Marlett claims that syntactically dependent pronouns must be adjacent to a
projection of the head. This means that they do not have to follow the head directly,
but can instead follow some element which is final in the phrase. Thus, he explains
that they can follow adverbs. In Zoogocho Zapotec, dependent pronominal forms
cannot follow non-suffixal adverbs. He finishes this section by saying that there are
two different versions of his hypothesis concerning adjacency: a strict version and a
lax version.
The strict version states that syntactically dependent pronouns ‘must directly
follow a projection of the head’ (ibid 95). The lax version states that syntactically
dependent pronouns must directly follow a projection of the head if they are not third
person, but can be separated from the head by other syntactically dependent
pronouns if they are third person (ibid 95). As has been demonstrated in the previous
section, the lax version is applicable to SBZZ.
Marlett’s classification partially covers the SBZZ forms which have been
discussed in this chapter. The independent forms described in section 3.2 correspond
to both Marlett’s syntactically independent pronouns and his prosodically
independent pronouns. The syntactically dependent pronouns with the lax version of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
the adjacency hypothesis correspond to the dependent subject forms which were
discussed in section 3.3. The dependent experiencer subject forms described in
section 3.4 also correspond to Marlett’s syntactically dependent forms with an even
stricter version of his adjacency hierarchy: the experiencer forms must immediately
follow a verb. There are then two types of pronominal forms found in SBZZ which
do not correspond to any described by Marlett: the dependent forms for the object,
and the emphatic pronouns.
Before going any further, I would like to introduce a number of terms that 1
find to be useful at this juncture. As described in Zwicky and Pullum (1983), the
distinction between simple and special clitics (which was also made in Zwicky
(1977) as well) is as follows: “(T)he basic property of SIMPLE clitics is that their
distribution in sentences is exactly the same as that of associated full forms...” (ibid
510). “All other clitics are SPECIAL clitics... either no corresponding full forms
exist... or else the clitics do not have the same distribution as the corresponding full
forms... as in the pronominal clitics of many Romance and Slavic languages...’ (ibid
510). Stephen Anderson (1992: 223) makes the point that special clitics can
potentially be viewed as being phrasal affixes, affixes which apply at the phrase-
level rather than the word-level. While there will be a more in-depth discussion of
grammatical word in the following chapter (4.5), I will tentatively define
independent word as a stress bearing lexical item. I will also define a fast speech
form as being a phonologically reduced form that is phonologically more complex
than the other clitics and which is a result of phonological reduction in fast speech.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
58
To sum up, Marlett describes three classes of pronominal elements in
Zapotecan languages: syntactically and prosodically independent pronouns which
occur either in isolation (e.g. as an answer to a question), preverbally as an argument
of a verb, or as an object of a Spanish preposition and cannot co-occur with a
coreferent noun phrase; syntactically independent and prosodically dependent
pronouns which occur as the object of a verb following a non-pronominal subject
and which also cannot co-occur with a coreferential noun phrase; and, finally,
syntactically and prosodically dependent pronouns which must be adjacent to a
projection of a head (strict and lax versions) and which cannot co-occur w ith a
coreferential noun phrase. Using the terms which were introduced in the previous
paragraph, one could view the syntactically and prosodically independent pronouns
as being independent words, the syntactically independent and prosodically
dependent pronouns as being simple clitics (occurring where the corresponding full
noun phrases would and in complementary distribution with full noun phrases), and
the syntactically and prosodically dependent pronouns as being special clitics (not
occurring in exactly the same distribution as the corresponding full noun phrases).
The following chart is a classification of the pronominal elements found i n
Zoogocho Zapotec. I will take up the issue of whether or not to consider the
dependent forms to be agreement markers or not in Chapter 8, and will treat them as
clitics through the course of the dissertation.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
5 9
Table 3.9 Classification of pronominal elements in Zoogocho Zapotec
Primary syntactic
environments
Co-occurrence
restrictions
Classification
per Marlett
Morphological
classification
Independent
pronouns
a) Preverbaily
b) Object due to
person
hierarchy
c) Object
following full
noun phrase
subject
d) Following
some
prepositions,
especially
Spanish loans
(see 7.12.6)
N Syntactically
and
prosodically
independent
pronouns
Independent
word
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6 0
Table 3.9 Classification of pronominal elements in Zoogocho Zapotecfcont.)
Primary syntactic
environments
Co-occurrence
restrictions
Classification
per Marlett
Morphological
classification
Dependent
subject/
possessive
\forms
a) Following the
verbal complex
b) Following a
possessed
object (used as
a subject in the
subject-
possessor o f
object
construction)
c) As a possessor
d) Following a
relational noun
e) Following a
preposition
f) Following a
quantifier or
adjective
N Syntactically
dependent and
prosodically
dependent
Special Clitics-
cannot co-occur
everywhere the
corresponding
full form does.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
61
Table 3.9 Classification of pronominal elements in Zoogocho Zapotecfcont.)
Primary syntactic
environments
Co
occurrence
restrictions
Classification per
Marlett
Morphological
classification
Experiencer
subject forms
Following a small
subset o f verbs
N No
corresponding
form in Marlett’s
classification.
Affixes-They
are moved as
whole-there are
obvious gaps in
the coverage,
and there are
selectional
restrictions
which look
more affix-like
than clitic-like.
Object forms Following
pronominal
subjects. See 3.4
and 3.6
Y Syntactically
dependent and
prosodically
dependent
pronouns
Special clitics,
cannot co-occur
everywhere the
corresponding
full form does.
Emphatic
forms
1s t and 2n c l persons
only, following
pronominal
subjects
N No
corresponding
form in Marlett’s
classification
Fast speech
clitics.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6 2
1 I will discuss the person hierarchy and its application to pronouns later in the current section.
2 The types o f prepositions will be discussed in som e depth in the lexical classes chapter.
3 N ote that the causativized form - lhan ‘dar asco/m ak s.o. feel sick’ does not take this form, unlike other
causativized form s here.
4 Unlike in Yatzachi (B utler 1980 p. 65), these form s do not seem solely used only in certain phonological
contexts, but rather are used with verbs w hich have an experiencer subject.
5 The instrum ental usage will be discussed in greater detail in 6.4.
6 See the com plex constructions chapter for an in-depth discussion o f this construction.
71 will also discuss the fast speech form s later in the current chapter.
8 U nfortunately, I have no data on experiencer subjects at this point.
9 Interestingly, in the case o f the third singular inanim ate subject clitics, if a first or second person pronom inal
form is the object, an independent form o f the first or second person pronoun m ust be used, even the last speech
form s are disallowed.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6 3
Chapter Four: Morphology
4.1 Introduction
In this chapter, I will give a brief sketch of the morphology of the language. 1
will focus on nominal and verbal morphology, noting where it may be extended to
other word classes.' The structure of this chapter will be as follows: I will begin by
discussing the nominal morphology of the language. I will then go on to discuss in
greater detail the verbal morphology. 1 will treat the morphology associated with
pronouns separately after discussing the verbal morphology, and will finish this
chapter with a brief discussion of the morphological classes present in the language
and a brief discussion of where the morphology of SBZZ fits into overall
morphological typologies.
4.2 Nominal morphology
4.2.1 Possession
A much fuller description of possession is found in 5.3.5. Inherent possession
(possession of items which, while alienable, are still closely associated with the
possessor) is marked prenominally with the prefix x- and the possessed nominal is
followed by either a pronominal clitic or a noun phrase, as in the following
examples.
(1) x-migw=a’
poss-friend=lsg
‘my friend’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6 4
(2) x-kuzh lalo
poss-pig lalo
‘Lalo’s pig’
(3) xtaobe’
x-dao=be’
poss-corn.tassel=3inf
‘his com tassel’
While for the most part, the possessive prefix only causes a following lenis
consonant to become fortis, as in (3) above, there are also examples such as the
following which show a great deal of assimilation.
(4) zxwikw=a’
poss.dog=lsg
‘my dog’
Presumably, this form comes from a combination of the possessive prefix x- and the
word for ‘dog’ be ’ ko ’ , as in the following:
(5) x-be’ko’=a’->zxwikwa’
The change caused by the possessive prefix, however, is by no means a
productive process in the language, as there are also examples such as the following:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6 5
(6) xpexa’
x-bex=a
poss-tomato=T sg
‘my tomato’
For a small, closed class of inalienably possessed nouns (which are always or
almost always possessed), possession is marked solely by the juxtaposition of a noun
phrase (7) or a pronoun (8).
(7)
yichgh lalo
head lalo
‘Lalo’s head’
(8)
yichgh=a’
head=lsg
‘my head’
The possessive marker which is used syntactically to mark possession of
alienable nouns, che= ‘o f has the allomorph chi= before vowel initial pronominal
clitics. Its use will be exemplified along with more of the idiosyncrasies of the
possessive construction will be discussed in 5.3.5. See also Sonnenschein (pending)
for another description. I hope to provide more in-depth discussion of all of the
idiosyncrasies relating to individual lexical items, including information about
suppletive forms, in the monolingual dictionary/trilingual lexicon 1 am preparing for
Zoogocho Zapotec.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6 6
4.2.2 Lexical compounds
There are also a number of lexical compound nouns, such as the following:
(9) lizhya
lizh=ya
poss.house=iron
‘Jail’
(10) yetextil
yete=xtil
tortilla=castillian (or foreign [from Sp.caslellano])
‘bread’
(11) bidao’
bi’=dao’
child=dim
‘child’
4.2.3 Nominal template
The following is the very basic template for nominal morphology in
Zoogocho Zapotec:
(12) (poss-) noun (=pronominal clitic)
I will discuss the pronominal clitics in greater depth later in the current
chapter in section 4.5.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6 7
4.3 Verbal morphology
4.3.1 General overview
In this section, I will give an initial sketch of what is by far the richest part of
the morphology of SBZZ: the verbal morphology. I will attempt to give both formal
descriptions and demarcations, and also to discuss the uses of the verbal
morphology. This section of this chapter will be primarily concerned with aspect,
though there will also be discussions of the fossilized passive and causative marker,
andative/venitive markers (directional markers), frequentative, repetitive, the
marking of plural subjects and objects, the marking of infinitives, the marking of
imperatives, deverbalizations, incorporation, adverbial clitics, and pronominal clitics,
among other things. Where necessary, I will cross-reference more in-depth
discussions in other chapters.
4.3.2 Primary Aspect
I will begin this section with a brief discussion of the meanings and uses of
the four primary aspects. Before going any further it is imperative to emphasize that
there is no tense in this language. Like many other Mesoamerican languages, aspect
is much more salient than tense in the grammar of the language2. In order to discuss
temporal placement or sequencing, Zoogocho Zapotec relies on temporal particles
such as za ‘ just, barely, hardly (Spanish apenasy, ba ‘already, at this/that point
(Spanishy a )\ or na ’ ‘now’; adverbs such as gxe ‘tomorrow’, or neghe ‘yesterday’;
or discourse knowledge.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
To begin the discussion, consider the continuative aspect. As its name
indicates, the continuative aspect is used to discuss events which are either still
occurring or which occurred or will occur over a period of time. While it is often
used in environments where other languages might use a present tense, it can be used
in non-present environments. The continuative aspect is marked by the prefix dx-, as
seen in the following examples. I am following Butler (1980) for Yatzatchi, and
Long and Cruz (1999) for Zoogocho who have called it the ‘aspecto continuativo',
‘continuative aspect’. It has also been called the ‘presente’, ‘present’ by Lopez and
Newberg (1990) for Yalalag. The following are a few initial examples. Note that this
aspect is mostly translated into Spanish with either the simple present or with the
present progressive.
(13) dx-e-ban-e’ yogo zhaha
cont-freq-wake=3f every day
‘She wakes up every day.’
(14) bi dx-aog zxoana
whatcont-eat Juan
‘What is Juan eating?’
Note that the continuative aspect is quite often also used to describe events
which took place in the past, but which regularly occurred, as in the following
sentence where a speaker described a feature of a job that she had for a number of
years.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6 9
(15) waana dx-eyozh=a’ dx-on=e’ danh dx-aw=a’
imagine cont-finish=lsg cont make=3f geninan cont-eat=lsg
xsil=en’
breakfast=det
‘Imagine that, when I finished there, they would make me breakfast.’
In another life history, a speaker described how children would play with avocado
seeds.
(16) b-en=to=n strom
comp-make=lpl(excl)=3inan top
‘We made them into tops.’
dx-on=t=on ka’
cont-make=lplexcl=3inan demadv
‘That’s what we did with them.’
Note that the continuative aspect can also be used with a future temporal
reading as in the following.
(17) dx-on=a’ shinhgxe
cont-make= 1 sg work tomorrow
‘I work tomorrow.’
The next aspect marker which will be discussed is the completive aspect,
usually marked with a b-, gw-, gud-, g-, or another marker. The choice of form will
be discussed below. The completive aspect (called ‘completivo’ or ‘completive’ by
Butler (1980) and by Long and Cruz (1999), but called the ‘preterito’ ‘preterite’ by
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
7 0
Lopez and Newberg (1990)) is used to discuss events that have or will come to
completion or ended or will end. This completion could be occurring as the sentence
is being said or could occur in the future or in the past, which is why a tense based
description of this aspect would be erroneous. The following are examples of the
completive aspect.
(18) b-e-ban=e’
comp-freq-wake=3 f
‘She used to wake up.’
(19) ga gud-ao=be’ yet
where comp-eat=3inf tortilla
‘Where did he eat the tortilla?’
The following example from a recipe shows a non-past usage of the completive
aspect.
(20) ba b-eyozh go-k gelatina kate
alreadycomp-finish comp-become gelatin when
0-zozxgho=dxo wi
pot-tear_up=lpl(incl) orange
‘When the gelatin is finished (boiling), we tear an orange up into pieces.’
Similarly, we see a present perfect use of the completive in the following example.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(21) per le galgh mil=en’ dx-onh=e’ kanate
but foe 20 thousand=det cont-give=3f before
‘but they used to give 20 thousand before’
dx-onh=e’ kanate galgh mil=en’
cont-give=3f before 20 thousand=det
‘Before they used to give 20,000.’
galhgh peso ba b-ey-on=en na’
20 peso already comp-freq-be=3inan now
‘20 pesos, it’s now become.’
The completive is also used for imperatives, as will be discussed in detail
later in the current chapter.
The potential mood, which sometimes corresponds to a future time reference
and sometimes corresponds to a subjunctive mood, is marked by gu-, gw- (gw is
realized as a w when it is word initial preceding a vowel (#gwV— >#wV-), g-, y-, and
w-). It also occurs with a zero allomorph or as a change to the root-initial consonant.
The choice of form will be discussed below. In general, it refers to an event that
either has not occurred yet or which is not specific. I call it the potential aspect
following the term ‘aspecto potencial’ as used by Butler and by Long and Cruz and
more generally in the literature on other Zapotecan languages and Mixtec (Macaulay
1996). Lopez and Newberg refer to this aspect in Yalalag as the ‘future tense’. The
following are examples of this aspect.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
7 2
( 2 2 ) gw-e-ban=e’
pot-freq-wake=3 f
‘She will wake up.’
(23) dx-bez=a’ w-aow bidao yeth
cont-hope= 1 sg pot-eat child tortilla
‘I hope the kid eats the tortilla.”
(24) bate’ y-egh=o’ nis
when pot-drink=2s water
‘When are you going to drink water?’
(25) shows an irrealis use of the potential marker. (26) shows a past use of
the potential marker, and also, as seen elsewhere here, the importance of temporal
adverbs in locating expressions in time.
(25) kage bi juguet ba dee 0-chitghe=dxo’ na lenh
neg no toy ya exist pot-play=lplincl demdist 3inan
gu-zh-be’ na’
pot-say=3inf now
‘There were no toys to play with, tell him this now.
(26) nadxe y-egh=dxo kafe lizh=e’
afterwards pot-drink=lplincl coffee poss.house=3inf
‘Afterwards we drank coffee at her house.’
(27) shows the use of the potential marker in a non-future context, as in (26) as well.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
7 3
(27) shi ba gu-zhed=o’ le gu-zheb=o’
if ya comp-get.late=2sg cause pot-get_frightened ;2sg
dx-zxit=o'
cont-jump=2sg
‘If you’ve gotten yourself so late that you get startled by your hurry.’
The fourth aspect is the stative aspect, which is referred to as such by all the
authors mentioned above. The stative aspect seems to have a number of uses,
including the expression of states and conditions (28) and (29), and habitual meaning
(30). The stative is marked with the prefix n- or by nothing at all. Note that the lenis
nasal will assimilate in place of articulation to the following consonant, as in (28).
The following are examples of this aspect.
(28) m-ban=a’
stat-live=lsg
‘I am alive.’
(29) kuzh la n-dxe=be’ ke?
pig focus stat-carry=3sginf no
‘He’s carrying a pig, right?’
(30) nake gu-ditgh=le kate n-ak=le bidao=na’
how comp-play=2pl when stat-be=2pl child=det
‘How did you all use to play when you were children?’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
7 4
(31) Maria n-ak-dx=e’ benhe zxen ka xoan
Maria hab-be-more=3f person large than Juan
‘Maria is larger than Juan.’
I will now go on to discuss the classification of the verbs in the Zoogocho
Zapotec lexicon according to the aspectual forms. I will not talk about the
classification of the verbs with respect to the difference between verbs which take
the =a ’ set of pronominal clitics or verbs in = da ’ which take the =d a ' set of
pronominal clitics as this has already been discussed in the chapter on the
pronominal forms. Similarly, we will not discuss other ways in which verbs might be
classified here, such as by the type of pluralization they take, but will instead leave
that discussion until later in the current chapter.
There are many ways in which one could potentially group SBZZ verbs
according to their aspectual forms. One way which is quite common (as used by
Butler (1980), Lopez and Newberg (1990), Cruz and Long (1999), and others) is to
form groups of verbs which have the same potential form (which for this dialect
grouping would be four main conjugations) and then to make smaller subgroups
based on the form of the completive. Lopez and Newberg arrive at 44 separate
conjugations. As mentioned to me by Pamela Munro, one could just as easily classify
the verbs according to the completive forms and then form subgroups based on the
potential. (In fact, on a broad level, while this would be more compact with three
main groupings, the subgroupings would be even more confusing than the morass 1
present below.) In general, the person trying to learn this language will have to learn
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
7 5
the forms for each verb on a verb by verb basis. Unfortunately there are neither
formal or functional means of predicting what aspectual form an individual verb will
have. In what follows, I will follow previous researchers in making the primary
groupings based on the potential form, and then the subgroupings based on both the
potential and completive forms. Except where noted, the continuative is always dx-
and the stative is always n-.
The first group which I will discuss are the verbs which take, as their
potential marker, gw-. gw- is pronounced as gu- before consonants and w- before
vowels. This set is the most regular; all of the verbs take b- as the completive marker
and dx- as the continuative marker. All of these verbs are transitive.
(32) gu-yaa=be’
pot-dance=3inf
‘He will dance.’
b-yaa=be’
comp-dance=3inf
‘She danced.’
dx-yaa=be’
cont-dance=3inf
‘He is dancing
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
7 6
(32cont.) n-yaa=be’
stat-dance=3inf
‘She dances.’
Table 4.1 Conjugation in gw-
Conjugation in gw-
Potential gw-
Completive b-
Continuative dx-
Stative n-
The next aspectual verb class which I will discuss is composed of verbs
which take y- as the potential marker. This is made up of three subclasses; verbs that
take b- as the completive marker, verbs which take gw- as the completive marker,
and verbs which take gd- as the completive marker.
The first subclass to be discussed are those which take b- as the completive
marker. Note that there are no real semantic generalizations which can be made here
unique to this subclass. Formally, one can observe that all of these roots are vowel
initial.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
7 7
(33) y-eyalh=a’
pot-get_cold=lsg
‘I will get cold.’
b-eyalh=a’
comp-get_cold= 1 sg
‘I got cold.’
dx-eyalh=a’
cont-get_cold= 1 sg
‘I am getting cold.’
n-eyalh=a’
stat-get_cold=lsg
‘I get cold.’
Table 4.2 Conjugation in v-, completive subclass in b-
Conjugation in y-, completive
subclass in b-
Potential
y-
Completive b-
Continuative dx-
Stative n-
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The next subclass is composed of those verbs which take y- as the potential
marker and have as the completive marker gd-. Note that this subclass is made up of
one verb. (It is a result of my following Butler’s classification initially; more verbs
may turn up at a later date.)
(34) y-eb=a=n
pot-swallow= 1 sg=3 inan
‘1 will swallow it.’
gd-eb=a=n
comp-swallow=l sg=3inan
‘I swallowed it.’
dx-eb=a=n
cont-swallow= 1 sg=3inan
‘I am swallowing it.’
n-eb=a=n
stat-swallow= 1 sg=3 inan
‘I swallow it.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
7 9
Table 4.3 Conjugation in v-, completive subclass in gd-
Conjugation in y-, completive
subclass in gud-
Potential gw-
Completive b-
Continuative dx-
Stative n-
The final subclass of verbs which take y- as the potential marker are those which
take gw- as the completive marker. This too is a small subclass.
(35) y-egh=a=n
pot-drink=l sg=3inan
‘I will drink it.’
gw-egh=a=n
comp-drink= 1 sg=3 inan
‘I drank it.’
dx-egh=a=n
cont-dr ink= 1 sg=3 inan
‘I am drinking it.’
n-egh=a=n
stat-drink= 1 sg=3 inan
‘I drink it.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
8 0
Table 4.4 Conjugation in v-, completive subclass in gw-
Conjugation in y-, completive
subclass in gw-
Potential
y-
Completive gw-
Continuative dx-
Stative n-
The next class of verbs which will be discussed are those verbs which take g~
as the potential marker. There are 5 subclasses: verbs which take g- in the completive
concomitant with a change in the vowel of the verb from -a- to -o- (this subclass and
those which similarly involve a completive beginning with g- and having a change
from -a - to -o - should be considered to actually have gw- as the completive marker
with the rule (gw-a->go)), verbs which take g- as a completive marker, verbs which
take b- as the completive marker with no change to the vowel in the verb, verbs
which take b- as the completive marker along with a change to -e- of the vowel in
the verb, and finally verbs which take b- as the completive marker along with a
change to the vowel to -i- in the verb.
The first subclass is made up of those verbs which take g- as the potential
marker and have, as their completive marker, g- along with a change from -a- to -o-
in the verbal root. There is no semantic explanation I can come up with for this
grouping. All of these roots are a- initial.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
81
(36) g-azgh=a’
pot-bathe=lsg
‘I’m going to bathe myself.’
go-zgh=a’
comp-bathe=lsg
‘I bathed myself.’
dx-azgh=a’
cont-bathe=lsg
‘I am bathing myself.’
n-azgh=a’
stat-bathe=l sg
‘I bathe myself.’
Table 4.5 Conjugation in g-, completive subclass in go-
Conjugation in g-, completive subclass
in go-
Potential
g-
Completive go- (Root initial vowel is deleted.)
Continuative dx-
Stative n-
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The following is the sub-class which takes g- as the potential marker and
takes g- as the completive marker. The only way these forms can be differentiated is
by the tone. The potential has a higher tone than the completive. Note that there is no
semantic generalization to be made here, nor are there any phonological
generalizations which can be made about this subclass.
(37) g-ozxghe yish=en’
pot-rip paper=det
‘The paper will rip.’
g-ozxghe yish=en’
comp-rip paper=det
‘The paper ripped.’
dx-ozxghe yish=en’
cont-rip paper=det
‘The paper rips.’
n-ozxghe yish=en’
stat-rip paper=det
‘The paper is ripped.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
8 3
Table 4.6 Conjugation in g-, completive subclass in g-
Conjugation in g-, completive
subclass in g-
Potential g- (Higher tone than the
completive.)
Completive
g-
Continuative dx-
Stative n-
The following subclass is made up of verbs which take g- as the potential
marker and have b- as the completive marker. With the exception of gosia' ‘I will
scream’, all members of this class are transitive, and there are quite a number of
morphologically causativized verbs.
(38) g-os-bizh=a=n
pot-caus-dry=l sg=3inan
‘I’m going to dry it’
b-os-bizh=a=n
comp-caus-dry=l sg=3inan
‘I dried it.’
dx-os-bizh=a: =n
cont-caus-dry=l sg=3inan
‘I am drying it.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
8 4
(38 cont.)n-os-bizh=a=n
stat-caus-dry= lsg=3 inan
‘I dry it’
Table 4.7 Conjugation in g-, completive subclass in b-
Conjugation in g-,
completive subclass in b-
Potential
g-
Completive b-
Continuative dx-
Stative n-
The following subclass consists of /o/ initial roots which take g- as the
potential marker and which take b- along with a change from /o/ to Id as the
completive marker. There are no semantic or formal generalizations which can be
made here.
(39) g-ot=a’ to kuzh
pot-sell=lsg one pig
‘I will sell a pig.’
be-t=a’ to kuzh
comp-sell=lsgone pig
‘I sold a pig.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
8 5
(39 cont.)dx-ot=a’ to kuzh
cont-sell=lsg one pig
‘I am selling a pig.’
n-ot=a’ kuzh
pot-sell=lsg pig
‘I sell pigs.’
Table 4.8 Conjugation in g-, completive subclass in be-
Conjugation in g-, completive
subclass in be-
Potential
g-
Completive be-
Continuative dx-
Stative n-
The following subclass is composed of /o/ initial verbs which take g- as the
potential marker and b-, accompanied by a change in the vowel from o to i, as the
completive marker.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
8 6
(40) g-olh=a’
pot-sing=lsg
‘I will sing.’
bi-lh=a’
comp-sing=lsg
‘I sang.’
dx-olh=a’
cont-sing=lsg
‘I am singing.’
n-olh=a’
stat-sing=l sg
‘I sing.’
Table 4.9 Conjugation in g-, completive subclass in bi-
Conjugation in g-, completive
subclass in bi-
Potential
g-
Completive bi-
Continuative dx-
Stative n-
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
I have saved the most difficult aspectual verb class for last, those that have
potential forms which do not have a prefix . All of these verb roots are consonant
initial. The first division is rather easy: there are two sub-classes, based on whether
they have completive forms beginning with b- or beginning with gw-. T he second of
these subclasses is rather complex and is divided into further subclasses. I will
discuss this subclass in greater detail after first having a brief discussion of the
subclass which has as its completive marker b-. This subclass is quite regular,
although there are no semantic or phonological generalizations which can be made.
(41) 0-ganh=a’
pot-stay=lsg
‘I will stay.’
b-ganh=a’
com p-stay-lsg
‘I stayed.’
dx-ganh=a’
cont-stay=lsg
‘I am staying.’
n-ganh=a’
stat-stay=lsg
‘I am staying.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
8 8
Table 4.10 Conjugation in 0-, completive subclass in b-
Conjugation in 0-, completive
subclass in b-
Potential 0-
Completive b-
Continuative dx-
Stative n-
Now I will look at the subclass consisting of verbs that have no prefix or a
change from lenis to fortis for the potential form of the verb and which take some
form of gw- for the completive aspect. The subclasses of this subclass will be based
on either whether the potential causes a change to the initial consonant or what the
form of the completive or, for one subclass, the continuative is.
The first subclass involves those verbs for which the potential is zero marked
and for which the completive begins in gw- and which are otherwise regular. While 1
can offer no real semantic generalizations (other than the fact that there are many
verbs which express states here, but not enough to say that it is really a trend), all of
these roots are consonant initial.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
89
(41) 0-bizh=a’
pot-dry=lsg
‘I’m going to be very dry.’
gu-bizh=a’
comp-dry=lsg
‘I was very.’
dx-bizh=a’
cont-dry=lsg
‘I am very dry.’(Right now)
m-bizh=a’
stat-dry=lsg
‘I’m very dry.’(In general)
Table 4.11 Conjugation in 0-, completive subclass in gu-
Conjugation in 0-, completive
subclass in gu-
Potential 0-
Completive gu-
Continuative dx-
Stative n-
The next subclass in which the potential conditions a change from lenis to
fortis for the initial consonant is basically defined by being those which have as their
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
90
completive form gud- (with the omission of the root initial consonant), and has two
further subclasses; one which has the regular form of the continuative dx-, and one in
which the continuative is contrasted with the potential by being marked with dxy-.
(42) 0-kap=a=ne’
pot-slap=lsg=3fo
‘I will slap him.’
gud-ap=a=ne’
comp-slap=l sg=3fo
‘I will slap him.’
dx-gap=a=ne’
cont-slap= 1 sg=3 fo
‘I am slapping him.’
n-gap=a=ne’
stat-slap=lsg=3fo
‘I slap him.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
91
Table 4.12 Conjugation in 0-, completive subclass in gud-, regular continuative
Conjugation in 0-, completive
subclass in gud-, regular
continuative
Potential 0- (accompanied by a change in the
initial consonant of the root from
lenis to fortis)
Completive gud- (accompanied by the
omission of the root initial
consonant)
Continuative dx-
Stative n-
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
9 2
Table 4.13 Conjugation in 0-, completive subclass in gud-, continuative in dxy-
Conjugation in 0-, completive
subclass in gud-, continuative in
dxy-
Potential 0- (accompanied by a change in
the initial consonant of the root
from lenis to fortis)
Completive gud- (accompanied by the
omission of the root initial
consonant)
Continuative dxy- (accompanied by the
omission of the root initial
consonant)
Stative n-
(44) 0-chitgh=a’
pot-play=lsg
‘I will play.’
gud-itgh=a’
comp-play=lsg
‘I played.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(44cont.) dxy-itgh=a’
cont-play=lsg
‘I am playing.’
n-dxitgh=a’
stat-play=lsg
‘I play.’
The next subclass are those which have as their completive form, gulh-
(which involve the omission of the root initial consonant) and also has two further
subclasses, those which alternate between fortis and lenis for the initial consonant for
the potential and those which alternate between kw- for the potential and -b- for the
continuative.’ ' The sub-subclass mentioned first only has stems beginning in the
potential with k and in the continuative with g. There are no semantic generalizations
to be made here either.
(45) 0-ko=a’
pot-climb=lsg
‘I will climb.’
gulh-o=a’
comp-climb=lsg
‘I climbed.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(45cont.) dx-go=a’
cont-climb=lsg
‘I am climbing.’
n-go=a’
stat-climb=lsg
‘I climb.’
Table 4.14 Conjugation in 0-, completive subclass in gulh-
Conjugation in 0-,
completive subclass in
gulh-
Potential 0- (accompanied by a
change in the initial
consonant of the root from
lenis to fortis)
Completive gulh- (accompanied by the
omission of the root initial
consonant)
Continuative dx-
Stative n-
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 4.15 Conjugation in k\v-, completive subclass in gulh-
Conjugation in 0-,
completive subclass in
gulh-
Potential 0-(The initial consonant of
the root is kw-)
Completive gulh- (accompanied by the
omission of the root initial
consonant)
Continuative dx-
Stative n-
(46) 0-kwez=a’
pot-wait=lsg
‘I will wait.’
gulh-ez=a’
comp-wait=lsg
‘I waited.’
dx-bez=a’
cont-wait=lsg
‘I am waiting.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
96
(46cont.) m-bez=a’
stat-wait=lsg
‘I wait.’
The next subclass involves all those verbs which have the potential marked
by a fortis consonant and which have the completive subclass marked by gw-. There
are no semantic generalizations which can be made.
(47) 0-si=a’ to be’ko’
pot-buy=lsg one dog
‘I will buy a dog.’
gu-zi=a’ to be’ko’
comp-buy=lsg one dog
‘I bought a dog.’
dx-zi=a’ to be’ko’
cont-buy=lsg one dog
‘I am buying a dog.’
n-zi=a’ be’ko’
stat-buy=lsg dog
‘I buy dogs.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
9 7
Table 4.16 Conjugation in 0-, completive subclass in gu-
Conjugation in 0-,
completive subclass in
gu-
Potential 0-(accompanied by a
change from lenis to
fortis in the initial
consonant of the root)
Completive gu-
Continuative dx-
Stative n-
The final class which I will mention here are those which have potentials
which begin in sh- and completives which begin in gy-.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 4.17 Conjugation in sh-, completive subclass in gy-
Conjugation in sh-,
completive subclass in gy-
Potential sh-
Completive
gy-
Continuative dx-
Stative n-
(48) sh-oo=a’ to yoo
pot-enter=lsg one house
‘I will enter a house.’
gy-oo=a’ to yoo
comp-enter=lsg one house
‘I entered a house.’
dx-oo=a’ to yoo
cont-enter=lsg one house
‘I am entering a house.’
y-oo=a’ yoo
pot-enter=lsg house
‘1 enter houses.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
9 9
In general, as already mentioned, the determination of the aspectual forms for
an individual verb is a lexical task. I hope here to have given an initial idea of what
patterns there are.
4.3.3 Other, more rarely encountered preverbal aspectual markers
In addition to the primary aspects discussed so far and secondary aspects
which will be discussed in 4.3.5, there are a few other preverbal aspectual markers
which have been found by previous researchers which I have not investigated as of
yet and deserve to be mentioned. The first of these is the dubitative for which we
have found the following examples.
(49) w-ak=a’5
dub-be=lsg
‘I might be able to.’
(50) w=ak-z=e’
dub-can-emph=3 f
‘He can !?’
As mentioned in Long and Cruz (1999: 429-430), this aspect is only used
with a handful of verbs.6 Its specific use7 and forms are issues which require further
research.
There is also a form which is discussed in Butler (1980: 113) and Long and
Cruz (1999: 451-452) which they both describe as being an interrogative form. When
I first began studying Zoogocho Zapotec, I found no evidence of this form, and 1
believed that it was no longer in use. I have found however the following evidence of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
100
use by older monolingual speakers. The form is an invariant z-. As described by
Butler, and Long and Cruz, affirmative responses use the same form, as in the
following examples.
(51) z-ag=o’ yet yelha’
dp-eat=2sg tortilla banana
‘Did you eat banana tortillas?’
z-agu=a’ yet yelha’
dp-eat=lsg tortilla banana
‘Yes, I ate banana tortillas.’
However, I have also found a few examples like the following, which seem to
indicate to me that it could perhaps be considered to be a definite past.
Unfortunately, these forms could also be viewed as being long distance answers to
questions. There were around five lines of intervening materials.
(52) bi gud-aow=a=nda’ yet dao=n’
neg comp-eat=lsg=lsgfsf tortilla com JasseU def
‘I didn’t eat com spike tortillas.’
z-agw=a=nda’ yet yelha
dp-eat=l sg=T sgfsf tortilla banana
‘I did eat banana tortillas.’
There are two additional forms which Long and Cruz mention which 1 have
not encountered to date, a form which indicates termination, indicated by -ed- in the
secondary aspect slot8 and a form which means ‘can’ (which has too many
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
101
allomorphs to mention presently, though note that two of the forms are just formed
by taking the experiencer form of the personal pronoun). I have not found either of
these forms in either elicitation or in the texts, although I was not aware of them until
recently and have not had a chance to consider them as a possibility. They both
deserve further research.
4.3.4 Plural marking for third person subjects
Plurality is an sporadically marked category in Zoogocho Zapotec. If it can
be recovered from context, it is very typically not marked. When it is marked
verbally, it is marked with a prefix which is fused with aspectual markers. As
mentioned in Butler for Yatzachi (1980: pp. 77-78), there are two major classes of
plural marking of 3rd person subjects: those which have as their completive b-, and
those which have as their completive marking g-. Whether or not a verb belongs to
one of these categories is lexically determined.
The first conjugation that we will discuss is the forms which have completive
beginning in b-. Note that, as mentioned in (Long and Cruz 435), there are two
subclasses of this type of plural conjugation. The first of these subclasses involves
those verbs which have b- as their completive marker in the singular and gw- for the
potential. The 3rd person plural paradigm for this subclass is as follows, the forms to
the right are the singular forms.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 4.18 -o- plural forms
Plural forms Singular forms
zghan-sede=be’ ‘They are studying.’ (Stative) n-sede=be’ ‘He studies.’
dxosc-sede=be’ ‘They are studying.’(Continuative) dx-sede=be’ ‘He is studying.’
boso-sede=be’ ‘They have studied.’(Completive) b-sede: :be' ‘He studied.’
yoso-sede=be’ ‘They will study.’(Potential) gw-sede=be’ ‘He will study.’
The second major class also involves verbs which have b- in their singular
completive forms, but which lack the potential prefix. Note that the main difference
between this fonn and the last is that in this conjugation the vowel is an e whereas it
is an o in the previous conjugation.9
Table 4.19 -e- plural forms
Plural forms Singular forms
zghan-zxit=be’ ‘They are jumping’ (Stative) n-zxit=be’ ‘He jum ps.’
dxese-zxit=be’ ‘They are jumping.’(Continuative) dx-zxit=be’ ‘He is jumping.’
bese-zxit=be ‘They jumped.’(Completive) b-zxit=be’ ‘He jumped.’
yese-zxit=be’ ‘They will jum p.’(Potential) 0-zxit^be’ ‘He will jum p.’
The next type of plural marking which we will discuss are those verbs which
take g- in the plural. This class consists of those verbs which in the singular, are
consonant initial and have completives which begin in gw- (Butler 76-77).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 0 3
Table 4.20 Plural forms with completive in gw-
Plural forms Singular forms
zghan-nheb=be’ ‘They are asking’ (Stative) no form
dxese-nheb=be’ ‘They ask.’ (Continuative) dx-nheb=be’ ‘He is asking.’
gose-nheb=be’‘They have asked.’ (Completive) gw-nheb=be’ ‘He asked.'
yese-nheb=be’ ‘They will ask.’ (Potential) nheb=be’ ‘He will ask
Finally there are those roots which are vowel initial.
Table 4.21 Plural forms for vowel initial roots
Plural forms Singular forms
(No stative found) n-olh=be’ ‘He sings.’
dxes-olh=be’ ‘They sing.’ dx-olh=be’ ‘He is singing.’
gos-olh=be’ ‘They sung.’ b-ilh=be’ ‘He sang.'
yes-olh=be’ ‘They will sing.’1 0 g-olh=be’ ‘He will sing.’
4.3.5 Secondary aspect
What has been called the frequentative in the literature will be the first
secondary aspect to be discussed here. The semantics of this morpheme are, in most
cases, relatively transparent. This morpheme is used to express actions which are
repeated. It might be considered to be an iterative, but I will follow the tradition in
the literature and call it a frequentative. A good example which Long and Cruz use is
the following ‘el verbo chyib (dxyib- in our orthography) lavar no se usa en el
aspecto frecuentativo cuando se refiere a lavar ropa, pero es siempre frecuentativo
cuando se refiere a lavar platos’ ‘ the verb chyib- to wash is not used in the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
104
frequentative aspect when it refers to washing clothes, but it is always frequentative
when it refers to washing plates’ (ibid. 442). There are some further uses which
deserve to be discussed here.
Similar to the example cited, we see examples like (53) below.
(53) b-o-s-ol=a=n
comp-freq=caus=go_out= 1 sg=3inan
‘I put it out over and over again’ (As though it is a light bulb and 1 am turning
it off and it keeps coming back on.)
We also see examples like the following, which indicate repetition of an
already completed task.
(54) g-o-sha=a’ chizx=o’
pot-freq-heat=l sg poss.tortilla=2sg
‘I’m going to reheat your tortilla.’
With verbs of arrival, the use of the frequentative can indicate that a person
has arrived at the same place where they left from. Note the following examples.
(55) b-le=be’
comp-arri ve=3 inf
‘He arrived (at a different place than he left from).’
(56) b-e-le=be’
comp-freq-arri ve=3 inf
‘He arrived (at the same place that he left from).’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
When used with the verb dxzhel ‘to find’, the frequentative indicates that
someone has been searching for the item. Without the frequentative, it can have an
almost accidental quality, potentially being glossed ‘bumped in to’.
(57) b-zhel=da’ Lia Lank
comp-fmd=lsgexp Dona Angela
‘I bumped into Dona Angela.’
(58) b-e-zhel=da’ Lia Lank
comp-freq-find=lsgexp Dona Angela
‘I found Dona Angela.’ (Like I couldn’t find her and I was looking for her)
Finally, one can see the following example, which, like many verbs when
used in the frequentative, is not semantically transparent
(59) y-ey-ak=a’
pot-freq-feel=lsg
‘I’m going to get myself better.’
The frequentative has two basic forms, one with the prefix -e- and one with
the prefix -o-. They must both be followed by a -y- if the root is vowel initial. The
forms of the aspectual marker that these verbs are used with are basically the verbs
with a potential in y- subclass in b- as mentioned above with the exception that the
stative begins in z-. The following table, adapted from Long and Cruz (1999: 442),
shows the basic paradigm of the frequentative aspect.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
106
Table 4.22 Frequentative aspect
Frequentative Cont.-Freq. Stat.-Freq. Compl.-Freq. Pot.-Freq.
e
3r d plural
dxe(y)-
dxesye(y)-
ze(y)-
zesye(y)-
be(y)-
besye(y)-
ye(y)-
yesye(y)-
0
3r d plural
dxo(y)-
dxosyo(y)-
z°(y>-
zosyo(y)-
bo(y)-
bosyo(y)-
yo(y)-
yosyo(y)-
The other secondary aspect which will be discussed here is what is called the
repetitive in the literature. Verbs do not take both the repetitive and the frequentative
at the same time. The repetitive form is formed with the prefix -ez- or -oz- preceded
by dx- for continuative aspect, z- for stative, b- or g- for completive, and y- for the
potential. Verbs which take -ez- take -esez- for the 3rd person plural and verbs which
take -oz- take -osoz-. The repetitive is used when an action is repeated after a long
while. Whereas the frequentative can indicate rapid repetition, the repetitive
indicates that there is a period of time before the action is repeated, ft is quite rare in
the texts. This example, from the first text I transcribed, remains one of my best
examples of the repetitive.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
107
(60) Na despues b-ez- beyah=a’ yezx, b-eyah=a’
Then afterwards [Sp.] comp-rep go=lsg village comp-go -1 sg
Mexico lenh familia chi=a’
Mexico with family of=sg
‘Afterwards, I went to the village again, and went to Mexico City with my
family.’
4.3.6 Andative/venitive
There are also andative and venitive markers in this language. The andative
marker, -gh-, is used to describe actions where the subject is ‘going (literal) to do
something’. The venitive marker, -ede-, is used when the subject is ‘coming to do
something.’ We will discuss each in turn. The andative marker could potentially be
derived from the verb zegh= ‘to go’, and the venitive from the verb -id- to come.
The andative marker -gh-becomes -gha- before a consonant, while -gha-
becomes -ghase- for the third person plural in front of a consonant and -ghas- before
a vowel. The aspect markers used with the andative are: dx- (continuative), z-
(definite past)/stative), 0- (completive), and zh- (potential). For an in-depth
discussion of the forms of the andative marker and the combination of the andative
marker with the frequentative and the repetitive markers see Butler (1980), or 1 x>ng
and Cruz (1999). The use of the andative marker indicates that someone is going to
do something, but may not actually arrive and do it. (They may of course, but they
also may not.) For example, compare the following question and answer.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(61) ga zegh=be?
where comp.go=3inf
‘Where did he go?’
If one were to give the following answer, it would state that he went to sing
(where ever he would go to sing), but that he might not have made it there.
(62) z-gh-elh=be’
stat-and-sing=3 inf
‘He goes to sing.’
If one wanted to specify that the person actually did go somewhere and sang, one
would use a construction like (63).
(63) zegh=be’ skwel na’ g=olh-be’
comp.go=3 inf school there com p-sing3inf
‘He went to the school and sang there.’
Other constructions which need to be differentiated here include the infinitival
construction (64) and the participial construction (65).
(64) zegh=be’ golhe’
comp.go=3inf inf.sing
‘He went (in order) to sing.’
(65) zegh=be’ dx-olh=be’
comp.go=3inf cont-sing=3inf
‘He goes singing.’ (He goes while singing.)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 0 9
The basic form of the venitive is -ede- before consonant initial roots and -
edey- before verb initial roots. The plural forms are -edese- before consonant initial
roots and -edes- before verb initial roots. The aspectual markers used with the
venitive are: dx- (continuative), z- (stative), b- (completive), and y- (potential). I
refer the readers once again to Butler (1980) and Long and Cruz (1999) for a more
in-depth discussion of the forms of the venitive and of its interaction with other
secondary aspect markers. I will finish the discussion of the venitive with a pair of
textual examples.
(66) b-id=a’ b-ede-zo=a’ xono bio
comp-come-lsg comp-ven-be=lsg 8 month
‘I came and stayed for 8 months.’
(67) g-os-id=e’ b-ede-s-elag=e’ dxioo
comp-pl-come=3inf comp-ven=pl~run.off=3f lplincl
‘They came to run us off.’
4.3.7 Imperatives
Imperatives are formed using the completive form of the verb (68), or, in the
case of plural imperatives, by taking the potential form of the verb (69). In neither of
these cases is a subject present. Negative imperatives are formed using the potential
and bi ‘negative word’ (70). In this case, a subject is generally present.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
10
(68) b-enh to dulc chi=a’
comp-give one candy of=lsg
‘Give me a candy!’
(69) g-onh to dulc chi=a’
pot-give one candy of=lsg
‘Give me a candy you all!’
(70) bi g-onh=o=be’ to dulc
neg pot-give=2sg=3inf one candy
‘Don’t give him a candy!’
4.3.8 Infinitives
Infinitives are constructed using the completive form of the verb root
preceded by g- before an o or w- in all other cases. Infinitives are used in complex
constructions which will be described in greater detail in 6.5.2, mostly following the
verb ‘to go’; as deverbal adjectives; and as deverbal nouns.
(71) g- g o s‘to plant/sow’
sha=a’ g-os zaha
stat.go=lsg inf-plant beans
‘I go to plant beans.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
111
(72) w- waow ‘to eat’, wen’ to do’
benhe w-en shinh
person inf-do work
‘workers’
(73) w-e-yaa
inf-freq-dance
‘the dance’
4.3.9 Valence changing operations
There are a number of morphological valency changing operations in SBZZ.
Here I will concentrate on the causative.
First, however, I will dispense with the notion of the passive. In both Butler
(1980) and in Long and Cruz (1999), there is mention of a passive. It is my opinion
that, given the small number of examples that are cited by both authors and the fact
that this is not a productive process in the language, this should be relegated to the
lexicon. As further evidence of this, unlike the case of the causative which will be
examined shortly, native speakers do not recognize the connection between the
putative passive and active forms. Causatives are easily recognized and produced
when elicited. While experiments with novel stems might be in order to confirm this
it appears to be the case that Zoogocho Zapotec does not have a productive passive.
An example which follows the patterns described in Butler and in Long and Cruz is
the following.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(74) gw-chexo=a=n
comp-toast= 1 sg=3 inan
‘I will toast it’
(75) y-exo=n
pot-pass.toast=3inan
‘It will be toasted.’
Speakers have informed me, however, that the first was a causativized
version of the second, and, if anything, the more transitive form is morphologically
more marked than the less transitive form, confirming those intuitions.
1 will now move on to discuss the causative. The syntax of causatives will be
discussed in greater depth in 6.1. Although speakers for the most part will produce
analytic causatives in elicitation, both native speakers and myself recognize synthetic
causatives. I will therefore tentatively consider them to be a productive part of the
grammar. O f course, psycholinguistic experiments using novel stems would be of
use in verifying the previous statement.
There are a number of ways in which the causative may be identified. If the
root is vowel initial, then the causative can be formed by adding a -w- initially as in
the following example.
(76) dx-aog bi chi=a’ yet
cont-eat child of=lsg tortilla
‘My child is eating tortillas.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 1 3
(77) dx-w-aog=a’ bi chi=a’ yet
cont-caus-eat=lsg child of=lsg tortilla
‘I’m feeding my child tortillas.’
In other cases where the verbal root is vowel initial, the addition of a -s- or a -
z- causativizes a root.
(78) g=ozxgh yish=en’
comp-rip paper=det
‘The paper ripped.’
(79) b-z-ozxgh=a’ yish=en’
comp-caus=rip= 1 sg paper=det
‘I tore the paper.’
Note that it can cause changes in the frequentative as in the following.
(80) b-e-ban=a’
comp-freq-wake= 1 sg
‘I woke up.’
(81) b-os-ban=a’=le
comp-caus-wake= 1 sg=2sg(o)
‘I woke you up.’
There are also cases of verb initial roots where a change in the root signals the
causative. This change is often a change to a labial vowel as in the following.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 1 4
(82) y-eyee=n
pot-cook_itr=3 inan
‘It will cook.’
(83) g-oya=a=n
pot-cook_tr=l sg=3inan
‘I will cook it.’
Verb roots which are consonant initial can be causativized in a number of
ways. One of the most common ways for consonant initial verb roots to be
causativized is via a change in the initial consonant from lenis to fortis, as in the
following example.
(84) 0-lulh=a’ dxeelhe
pot-roll= 1 sg downwards
‘I’m gonna roll downwards (down the hill).’
(85) 0-lhulh=a=n dxeelhe
pot-caus-roll=1 sg=3 inan downwards
‘I’m gonna roll it downwards (down the hill).’
Another common way of forming a causativized form of a consonant initial verbal
root is through the addition of -os- or -oz-, as in the following example.
(86) 0-biz=a’
pot-get_wet=lsg
‘I will get wet.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 1 5
(87) g-os-biz=a=le
comp-caus-get_wet=l sg=2sg
‘I will get you wet.’
The final way in which a consonant initial verbal root is causativized is by some
other change than fortition to the root initial consonant, as in the following example.
(88) dx-bab=dxo
cont-itr.count=lplincl
‘We are counted.’
(89) dx-lab=dxo zaha
cont-tr.count=lplincl beans
‘We count beans.’
4.3.10 Incorporated nouns
I will briefly mention a few facets of noun incorporation here, as noun
incorporation is not synchronically a productive process in the language. Note that
incorporated nouns must directly follow the verbal root, as in (89). Adverbial
suffixes cannot intervene, as seen in the contrast between (90) and (91).
(90) zoa-lao-tek=a’
stat.stand-eye-alot=l sg
‘1 start out a lot.’
(91) *zoateklaoa’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
116
Most cases of noun incorporation are fossilized incorporations of body-parts
such as (92).
(92) bi dx-on-lazh=o’, aron
no cont-make-liver/heart=2sg, Aaron.
‘Don’t lie Aaron!’
More lexical and historical work is needed on noun incorporation both in
SBZZ and in Zapotecan in general.
4.3.11 Adverbial suffixes
Below I will discuss a few of the more prominent adverbial suffixes. 1
consider them to be suffixes on the grounds that they do not correspond to any full
form, and that they only attach to one lexical class (verbs), obligatorily before the
pronominal clitics. The first adverbial suffix I will discuss will be the suffix -Igha-
which is used to express hearsay or doubt. The following is an example o f -Igha-.
(93) b-en-lgha=o’ to legh
comp-make-doubt=2sg one fence
‘I heard you made a fence.’
By far, the most common type of adverbial suffix are the adverbial suffixes
which express quantity or emphasis, these include, but are not limited to -tek-, -z,(e)-,
-dxgwa, -dx-, etc. An example follows.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
117
(94) konte’ b-en-dx=a’ ganh
so that comp-do-more=l sg earn [Sp.]
‘ . so that I earned more... ’
Multiple adverbial suffixes can potentially co-occur Zapotecanists generally
distinguish between primary adverbials which occur directly following the root and
secondary adverbials which follow primary adverbials if the primary adverbials are
present.
(95) b-en-tek-dxgw=a’ lizh=a’
comp-make-more-emph=l sg poss.house=l sg
‘I really did build my houses.’
4.3.12 Pronominal clitics
Pronominal clitics have already been discussed in some detail in the previous
chapter. They are placed after all other material has already been placed on the verb.
The vowel initial pronominal clitics do cause some changes to the verbal root. See
Long and Cruz (1999: 418) for a summary of the main changes. The first and second
person also can cause changes to the verbal root, typically in speech or transfer
verbs. The following examples show different forms for the verb ‘to say’.
(96) dx-e=be=ne’
cont-say=3 inf=3 fo
‘He says to him ...’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
118
(97) dx-ap=a=ne’
cont-say=l sg=3fo
‘I say to him
4.3.13 Verbal template
In this section, I will describe briefly what a fully inflected verb looks like.
The following verbal diagram represents a fully loaded verb.
(98) Aspect-andative/venitive-secondary aspect-plural-Causative-VERB-
Incorporated Noun-primary adverbial suffix-secondary adverbial suffix^Clitic
pronoun (Su)=Clitic pronoun (Obj).
An example of an almost fully inflected verb is as follows.
(99) b-edey-ey-os-ban-tont-tek=e=nda’
comp-ven-freq-caus-live-fool-really=3 f= 1 sg
‘He keeps on coming and causing me to foolishly wake up.’
4.4.1 Adjectives
We have but two things to say about predicative adjective. They can take
adverbial suffixes and pronominal clitics, as in (100) and (101).
(100) tonhe-dxgw=o’
tall-emph=2sg
‘You are really tall.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(101) tonhe-dxgwa n-ak=e’ jef=en’ kleka benhe yeto
tall-emph stat-be= 3f boss=det than person other
‘The boss is taller than the other person.’
4.4.2 Quantifiers
Quantifiers can also take adverbial suffixes and pronominal clitics, as in
(102) and (103) and are sentence initial.
(102) to=ba’ dxi na’
one=3an stat.sit there
‘One of them is sitting there.’
(103) to-z=ba’ dxi na’
one-only=3an stat.sit there
‘Only one of them is sitting there.’
4.5 The morphological typology of SBZZ
Zoogocho Zapotec is an agglutinative, rarely fusional VAO/VA language.
While it is low on the index of fusion to use the terms introduced by Comrie (1981:
43-52), ZZ is moderately high on the index of synthesis, often containing many
morphemes in one word. Zoogocho Zapotec is a primarily prefixing language,
although there are adverbial suffixes and noun incorporation is post-verbal. T here at
a number of enclitics in the language, such as the pronominal clitics discussed in
Chapter Three. While phonologically these are not words, grammatically they
behave as independent words.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
In contrast with their definition of phonological word (cf 2.6.), Dixon and
Aikhenvald (2002) believe that the following is a universal definition of
grammatical word.
A grammatical word consists of a number of grammatical elements which:
(a) always occur together, rather than scattered through the clause (the
criterion of cohesiveness);
(b) occur in a fixed order;
(c) have a conventionalised coherence and meaning (ibid. 19).
(d) Morphological processes involved in the formation of words tend to
be non-recursive. That is, one element will not appear twice in a word
(ibid. 21).
(e) There will be just one inflectional affix per word (ibid. 22).
(f) A speaker may pause between words but not within a word (ibid. 23).
(g) A word may constitute a complete utterance, all by itself (ibid. 24).
In terms of these criteria, I arrive at the following definitions: a grammatical
word in Zoogocho Zapotec consists of a root or a compound and may potentially
have prefixes or suffixes attached to it. These prefixes and suffixes only appear
attached to some element. (Thus, although it would be phonologically acceptable,
one never sees the plural marker -ese- in isolation.) The roots and affixes occur in a
fixed order. (Thus one always sees x-kuzh=a ’ ‘my pig’ and never a ’ -x-kuzh or
kuzh=a ’ -x.) Together they have a conventionalized coherence and meaning. (Thus
zoalao means ‘to begin’.) These first three are the crucial criteria for defining a
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
grammatical word in SBZZ. Additionally, there are no recursive word formative
processes. (Thus we do not see b-ese-ese-zxit=be ’ ‘They jumped.’ but only b-ese-
zxit=be ’ ‘They jumped.’.) Criterion (e) does not apply. Speakers do pause between
words but not within a word. (The bold underlines below represent pauses for the
current purposes.)
(104) dxombe’ to yoo
dx-on=be’ to yoo
cont-make=3inf one house
‘He made a house.’
(105) dxon bedo’ to yoo
dx-on Bedo’ to yoo
cont-make Pedro one house
‘Pedro built a house.’
(106) *dxom___ be’ to yoo
dx-on=be’ to yoo
cont-make=3inf one house
‘He made a house.’
Finally, grammatical words may constitute a complete utterance. This criterion only
applies to nouns and verbs.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
122
(107) A:no b-zxit?
who comp-jump
‘Who jumped?’
B: bedo’
Pedro
‘Pedro’
(108) b-zxit=be’
comp-j ump=3 inf
‘He jumped.’
If one considers the types of elements which have been found in SBZZ so far.
one can see that all of the preverbal inflectional and derivational elements which
have been seen earlier in the current chapter are neither phonological words (not
being stressed) nor grammatical words (not being able to pause between them, not
being able to occur in isolation). Similarly, the possessive prefix and adverbial suflix
are neither grammatical nor phonological words. Note that all of the elements
discussed above in the current paragraph either only attach to one type of word class
or to single words.
The only elements which act as proclitics in the language are the classifiers,
which, although they can occur on their own as an independent phonological word,
are phonologically weak and often attach to the following wordThe directional clitic
and the determiner clitic are not phonologically words, but they attach to the noun
phrase as a whole, either attaching to the noun if the noun is the only word present in
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
the NP or attaching to following material such as an adjective, thus I will consider
them to be special (phrasal) clitics. I will treat similarly the sentence-level emphatic
discourse particle enclitic -x, which attaches to the last word of an utterance.
Finally, I will discuss the pronominal clitics'1 from this perspective. As they
have no independent stress, they are not phonological words. As already seen, they
form a part o f a phonological word with a verb. As seen below, they also do with a
possessed noun.
(109) xbembe’
x-ben=be’
poss-finger=3inf
‘her finger’
By the criteria given above, the pronominal clitics are not grammatical words
either. However, the grammar o f SBZZ almost requires one to look at them as
grammatical words. In pragmatically neutral contexts, the pronominal clitics occur in
complementary distribution with full N P’s.
(110) ?b-zxit=be’ bidao
comp-j ump=3sg child
‘The child jumped.’ (marginally ok with emphatic reading)
(111) * x-kuzh=be ’ bidao
poss-pig=3inf child
‘the child’s pig’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 2 4
I will return, in 8.7, to a fuller explanation of the status of the pronominal subject
clitics.
1 The m orphology w hich occurs with other w ord classes such as adjectives and quantifiers can be related lo
nom inal and verbal m orphology.
2 See B ohnem eyer (1998) for a com prehensive discussion o f this issue for Yucatecan Mayan.
3 W hile there is no prefix used for the potential form s, I will m ark potentials with a zero prefix. The potential can
cause alternations in the root initial consonant for som e o f these verb classes.
4 N ote that this is quite interesting given certain reconstructions o f Proto-Zapotec which reconstruct an
alternation betw een fortis kw and lenis b.
5 The potential form for this verb is gok.
6 The verbs w hich are listed are: -ak ‘can ’, -agw ‘eat’, -at ‘die’, -dxogh ‘take ou t’, and ezhaa ‘be h o t’ (Long
and C ruz 430).
7 In B utler (1980:114) and Long and C ruz (1999: 450), an apparently identical form is claim ed to be used for
interrogatives in the future tense, although neither investigator notes that the form is identical.
8 An interesting possibility w ould be to investigate potential connections between this form and the gwd- form
o f the com pletive.
9 I can see no reason for the variation betw een /o/ and I d in this and other conjugations at this point.
1 0 N ote that this is an interesting point o f com parison betw een Yatzachi Zapotec and Zoogocho Zapotec in that
the potential o f the Y atzachi form has no ye-.
1 1 A s seen in C hapter Three, the experiencer pronom inal form s, should be considered to be affixes, as they have
no corresponding full form s and obligatorily attach to a small subclass o f verbs.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Chapter 5: Simple Constructions
5.1 Introduction
In this chapter, I will give an in-depth sketch of the primary features of
simple constructions in Zoogocho Zapotec. I am using the term ‘simple
constructions’ to refer to all those syntactic constructions which are at or below the
level of the clause. Multi-clause constructions and constructions with additional noun
phrases which are introduced either by morphological derivation or by prepositions
will be discussed in the following chapter.
I will begin this chapter by giving a brief sketch of the basic features of
Zoogocho Zapotec word order, go on to a more in-depth discussion of simple
constructions, and will then go on to use the (near) universals put forth in Greenberg
(1966) as the basis for further discussion. Note that a more comprehensive discussion
of the contribution Zoogocho Zapotec makes to the knowledge of word order
typologies, especially to those that deal primarily with verb initial languages, will he
given in Chapter Eight.
5.2 Basic Word Order
Zoogocho Zapotec is a relatively rigid VSO language. SVO, and OVS do
occur as alternates, in focused or topicalized constructions as seen in (1) through (7)
below, which will receive more discussion in section 5.5 below. OSV never occurs,
and SOV never occurs in elicitation1 .
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
126
(1) Y S O
dx-aogo be’ko’ yet
cont-eat dog tortilla
‘The dog is eating tortillas.’
(2) S V O
be’ko’=n’ dx-aogo yet
dog=det cont-eat tortilla
‘It’s the dog that’s eating tortillas.’
(3) S V=s O
be’ko’ dx-aogo=ba’ yet
dog cont-eat-3an tortilla
‘The dog, it’s eating tortillas’
(4) O V S
yet dx-aogo be’ko’
tortilla cont-eat dog
‘Tortillas, the dog is eating.’
(5) S O V=s
*be’ko’ yet dxaogoba’
(6) O S V
*yet be’ko’ dxaogoba’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 2 7
(7) V O S
*dxaogo yet be’ko
Descriptive adjectives, genitives, and demonstratives all follow their head
nouns, numerals and quantifiers precede the nouns which they modify, and relational
nouns and prepositions precede the noun phrase they modify, as seen in the
following examples.
(8) Noun Adjective
be’ko’ gasgh
dog black
‘Black dog’
(9) Noun Genitive
yichgh bedw
head Pedro
‘Pedro’s head’
(10) Noun Demonstrative
be’ko’ nga’
dog dem.med
‘This dog’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 2 8
(11) Relnoun Noun
lao bedw
eye Pedro
‘Pedro’s eye’/ ‘in front of Pedro’
(12) Num Noun
chupe be’ko’
two dog
‘Two dogs’
(13) Preposition Noun
lenh acha
with ax
‘With an ax’
5.3 The Noun Phrase
5.3.1 Adjective-noun order:
The ordering of adjectives and nouns is noun-adjective as in the following
examples.
(14) yoo shish
house white
‘White house’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(15) be’ko’ gasgh
dog black
‘Black dog’
(16) yag gasgh zxen
tree black big
‘The big, black tree’
(17) bekozxo shnaa zxen shtaha
shawl red big pretty
‘pretty, big, red shawl’
Verbs can be deverbalized by using the infinitival form to give deverbalized
adjectives such as w-ate ’ ‘dead’ or bchog ‘cutoff.
(18) dx-oso-kwash=e’ ghea w-at=en
cont-pl-inter=3f hen inf-dead=def
‘The buried the dead hen.’
(19) n-di=e’ to pantalon b-chog
stat-wear_on_legs=3f one pants inf-cut
‘He’s wearing a pair of cut-offs.’
Note that when asked to cite adjectives in isolation native speakers will
invariably put da ‘inanimate classifier’ (which will be discussed in greater detail in
7.7), in front of the adjective as in (20).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 3 0
(20) da shish
clinan white
‘White’
Furthermore, da ‘inanimate classifier’, bi ‘classifier for small things', and be
‘animate classifier’ can occur in sentences agreeing with the head noun. The reason
behind and conditioning for this remain to be determined. The following
comparative sentence exemplifies this use.
(21) n-ak-dx bdxee be lis kleka’ be’ko’
stat-be-more ant clan small comp dog
‘Ants are smaller than dogs.’
5.3.2 Cardinal Numbers and Quantifiers
Cardinal numbers and quantifiers precede the noun
(22) to bidao’
one child
‘A child’
(23) ye-to ghed
some-one chicken
‘Another chicken’
(24) ye2 ghed
some chicken
‘some chicken’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
131
(25) do kafe=n’
all coffee=det
‘All the coffee’
(26) toto bidao
every child
‘Every child
(27) yogo ghed
all chicken
‘All chickens’
(28) balhe benhe bio
some person masc
‘Some men’
(29) zgha-nita=be’ shlaa wegh=be’
plur.stat-be=3inf side eachone=3inf
‘They are each on their own side.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
5.3.3 Plural marker
While the plural marker is not always present in plural noun phrases, and
plurality is either recovered from context or from the verbal marking, there are a
couple of examples in the text where the marker ka is used to mark a plural noun
phrase. Note that this is very rare, occurring 2 times in over 2000 clauses which 1
investigated. The following are the two examples.
(30) na yego ka
and river plural
‘and the rivers’
(31) shgh-een=a’ benhe bila ka
cont.and-visit=lsg person sister of woman plural
‘ I went to visit my sisters.’
5.3.4 Demonstratives
Demonstratives follow their nouns, and either appear by themselves as in
(32) or with a classifier as in (33). See the discussion of demonstratives in 7.8 for
further explanation of the use of demonstratives.
(32) bia na’
animal demdis
‘This animal’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 3 3
(33) be’ko’ be=nga
dog clan=demmed
‘That dog’
5.3.5 Possession
5.3.5.1 Introduction
Zoogocho Zapotec is a language which could potentially be characterized as
making a distinction between alienable and inalienable possession; however, one of
the main points of this subchapter will be that there is a great need to refine these
terms and be more careful in the description of this phenomenon. 1 will refine these
terms over the course of this subchapter. The structure is as follows: I will begin by
giving a brief description of attributive possession in Zoogocho Zapotec, give a brief
overview of what has been said about possession in the literature, delve into a brief
discussion of a potential processing explanation, and finish by looking at the use of
attributive possession in SBZZ texts.
5.3.5.2 A brief overview of attributive possession in Zoogocho Zapotec
There are various ways in which one could potentially describe possession in
Zoogocho Zapotec. In terms of functional characteristics, one might take into
account the semantics of the noun, the frequency the individual noun is possessed,
and many other factors, most of which are, as the two which were mentioned,
interrelated. For example, many Zoogocho Zapotec kinship terms and body-parts arc
usually possessed and use no formal marking to indicate that they are indeed
possessed, such as examples (34) and (35). In these instances in Zoogocho Zapotec,
possession is marked solely by the juxtaposition of possessee and possessor. In cases
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 3 4
like (35) where the possessor is marked by a pronominal clitic, it is important to note
that the pronominal clitics which are used are the same forms which are used for
marking subjects.
(34) yichgha’5
yichgh=a’
head=lsg
‘My head’
(35) tao lalo
grandmother Lalo
‘Lalo’s grandmother.’
On the other end of the spectrum, there are nouns which belong to a large,
open class of nouns which are not possessed very often (material objects, celestial
objects, etc) and which take the preposition che followed by the possessor in order to
mark possession. I will call this syntactic possession.
(36) libr chebe’
libr che=be’
book of=3inf
‘His/her book’
In between lies a potential source of confusion. There is an open class of
nouns, which, while frequently possessed, can take either the possessive prefix x-
(quite often resulting in a great deal of change to the root or an altogether different
suppletive form), or can use syntactic possession. The semantic difference, which, on
first approach seems to be the difference between inherent possession (possession
which, while potentially alienable, is strongly associated) and non-inherent
possession (possession which is not already presupposed), can initially be seen in the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
135
difference between (41) and (42), below. I will return to these differences later in the
current subchapter.
(37) x-kuzh=a’
poss-pig=lsg
‘My pig’
(38) kuzh chia’
kuzh che=a’
pig of-1 sg
‘My pig’
(39) yet chia’
yet che=a’
tortilla of=lsg
‘My tortilla’
(40) chizxa’
chizx=a’
poss.tortilla=lsg
‘My tortilla’
(41) yoo chia’
yoo che=a’
house of=lsg
‘My house’
(42) lizha’
lizh=a’
poss.house=lsg
‘My house’,’My home’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
136
As I have shown in this brief sketch of possession in Zoogocho Zapotec,
there are three formal classes motivated by functional considerations; one of which
shows no formal change to the root and which involves a small, closed class which is
always possessed4, one with purely syntactic marking of possession involving a
large, open class of nouns, and one, involving a large, open class of nouns which
shows both types of marking, depending on whether the possession is inherent or
not. I will now move on to discuss how attributive possession has been discussed
more generally in the literature.
5.3.5.3 Discussion of attributive possession in the literature
As analytical frameworks, I will rely on Hansjakob Seiler’s Possession as an
Operational Dimension o f Language (1983) and Bernd Heine’s Possession (1997).
Heine (1997) distinguishes between attributive and predicative possession by
pointing out of attributive possession that:
(a)((I)t) presents typically presupposed rather than asserted information
b) it involves object-like, time stable contents rather than event-like
contents; and
c) it involves phrasal rather than clausal syntax (ibid. 143)
Consider the following two examples from English, in which one can see the
distinction which Heine is making.
(43) My money (Attributive possession)
(44) I have money. (Predicative possession)
Example (43) shows a presupposed object-like entity with phrasal syntax and
in (44) an event-like assertion with clausal syntax is seen. For the purpose of this
chapter, I will only be considering examples in Zoogocho Zapotec which correspond
to (43), however as a brief excursion, consider the following Zoogocho Zapotec
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
examples which correspond to the English examples seen above and show the use of
positional verbs for predicative possession.5
(45) xmedxoa’
x-medxo=a’
poss-money=lsg
‘My money.’
(46) zehe xmedxoa’
zehe x-medxo=a’
hang poss-money= 1 sg
‘I have money.’
Heine also gives the following table of potential diachronic sources for the
grammaticalization of attributive possession (Table 3.1, p 144) as ‘A formulaic
description of source schemas used for the expression of attributive possession"
(ibid. 144).
Formula Label o f event schema
Y at X Location
Y from X Source
Y for/to X Goal
X with Y Companion
(As for) X, X ’s Y Topic (ibid. 144)
Zoogocho Zapotec possessive constructions may potentially be related to the
‘Goal Schema’. Heine cites the ‘ pal periphrastic possessive’ that Campbell describes
for Pipil, as being evidence for the presence of the Goal Schema because of its use in
benefactive constructions (ibid. 147). Zoogocho Zapotec also uses possessive
constructions for benefactive constructions, as in the following.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(47) Bzoghale’ to kart.
b-zogh=a’=le’ to kart,
comp-write^ 1 sg^2sgo one letter
‘I wrote you a letter.’
(48) Bzogha’ to kart chio’.
b-zogh=a’ to kart che=o’
comp-write=lsg one letter of=2sg
‘I wrote a letter for you.’ (As though the second person cannot write, and the
first is doing her a favor. Can be followed with another clause like ‘which you sent to
the municipal authorities’.)
(49) Bzoghale’ to kart chio’
b-zogh=a=le’ to kart che=o’
comp-write=lsg=2sgo one letter of=2sg
‘ I wrote you a letter.’ (No potential benefactive reading)
In (47) and (49), there is no potential benefactive reading in the sense of (48)
because the second singular person is specified as a recipient. Possession in general
is used in many syntactic constructions, as will be shown in depth in Chapter Six.
Whether the Goal Schema is really present in Zoogocho Zapotec remains to be seen.
I will now move on to discuss Hansjakob Seiler’s definition of possession. .
He begins by defining possession as:
the representation of a relationship between a substance and another
substance. Substance A, called the POSSESSOR, is prototypically
[=animate], more specifically [=human], and still more specifically [=EGO'|
or close to the speaker....Substance B, called the POSSESSUM is either
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 3 9
[=animate] or [-animate]. It prototypically includes reference to the
relationship as a whole and to the POSSESSOR in particular (Seiler 4).
He also defines possession as being crucially ‘bio-cultural’ and differentiated from
other linguistic relations such as valence and location in the following ways.
VALENCE is the relationship between an action or process or state and its
participants. The number of participants can range from zero to three or four,
whereas POSSESSION is a strictly binary relationship (ibid. 4).
He continues to state that although location is also a binary relationship, it,
like valence, crucially relies on a relator (ibid. 4). That is, possession can consist
solely of possessor and possessee, but location relies on a figure, ground, and
relation between the two. This is different from what has already been shown for the
case of possession as possession can often go unmarked as in the first class of nouns
discussed in 5.3.5.2.
Seiler then goes on to distinguish between inherent possession and
established possession. Inherent possession for him implies that ‘(S)emantically this
kind of representation implies more intimate POSSESSION: Prototypically, of 'se lf
to his kinsmen, his body parts, etc’ (ibid. 5). In other words, this is what is normally
considered to be inalienable possession.
Established possession is ‘established by explicit means, which are, in
principle, means of predication’ (ibid. 5). Furthermore Seiler claims that:
‘The more explicit, more predicate-like expressions are marked vis-a-vis the
less explicit, more inherent expressions. On the other hand, the latter are
more grammaticalized, more morphologically expressed, while the former
are more syntactically expressed and less grammaticalized.’ (ibid. 6)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
140
The explicit type of possession is often called alienable possession. Of
interest to Zoogocho Zapotec and to Otomanguean languages in general is the
distinction which Seiler is making here between the more ‘morphologically
expressed’ inalienable possession and the more ‘syntactically expressed’ alienable
possession. As will be shown later there are potential processing explanations for
why these functional groupings receive these types of formal expression. However
first, I will examine what Heine has to say about these distinctions.
Heine defines the alienable/inalienable distinction as follows:
Superficially, the distinction is a straightforward one: Items that cannot
normally be separated from their owners are inalienable, while all others are
alienable. Thus, items belonging to any of the following conceptual domains
are likely to be treated as inalienable:
(a) Kinship roles
(b) Body-parts
(c) Relational spatial concepts, like ‘top’, ‘bottom’, ‘interior’, etc
(d) Parts of other items like ‘branch’, ‘handle’, etc.
(e) Physical and mental states like ‘strength’, ‘fear’, etc. (cf. Lichtenberk
1985:105)
(f) Nominalizations, where the ‘possessee’ is a verbal noun, for example
‘his singing’, ‘the planting of bananas’ (ibid. 11).
He goes on to note that in individual languages there are other terms which
might end up being treated as inalienable. However, as he notes and 1 will show later
when trying to explain the distribution of the formal marking of this phenomenon,
the specific items which are alienable or inalienable in a given language might have
less to do with their specific semantics and more to do with their occurrence in
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
141
discourse. Before going on to this point, however, I feel that it is useful to discuss
how categories (a) through (f) above relate to Zoogocho Zapotec.
Zoogocho Zapotec fits neatly into Heine’s categorization of inalienably
possessed items. Kinship terms (with the exception of loanwords which will be
discussed at the end of this section) are inalienables, as in the following:
(50) xa=be’
father=3inf
‘Her father’
It is interesting to note that, other than kinship terms (category a), all of the
other applicable categories up to and including (d) are expressed with body-part
terms. As is widespread in the language family and in the area, most spatial concepts
are expressed with the use of body-part terms as relational nouns. In Zoogocho
Zapotec, these are used for most of the relational spatial concepts and for the
constituent parts of most items as well. Note that even those terms such as ladghw
‘in between’ which are not currently associated with a body part in Zoogocho
Zapotec and might be claimed to be prepositions also behave like possessed nouns.
(51) lao=be’
eye=3inf
‘her eye’
(52) lao xa-xna=be’
eye father-mother=3inf
‘in front of her parents’
(53) lao plum
eye pen
‘tip of the pen’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 4 2
Mental and physical states are often expressed using an incorporated form of
a word like lazhe= ‘liver’, so they also potentially correspond to this categorization,
if tangentially. I have nothing to say at this point about nominalizations where the
possessee is a verbal noun, except that this is an interesting issue for further
investigation.
One final thing which should perhaps be discussed here is the issue of loan
words. Loan words are a good indicator of the limitations of a purely blind semantic
approach to the issue of inalienability. Loan words in Zoogocho Zapotec, such as tio
‘uncle [Sp.]’ are possessed with the preposition che= as in example (54) below.
(54) tio chia’
tio che=a’
uncle[Sp.] of=lsg
‘My uncle’
However, one could justifiably say that this word has not had the time to be
fully grammaticalized, as is quite common for both loan words and kinship terms not
related to the nuclear family cross-linguistically. This brings up an important issue:
what is responsible for the differentiation of the formal marking of possession in
Zoogocho Zapotec (and other languages)? One could potentially claim that this is an
issue of iconicity (as discussed in, say, Haiman (1983)) in which the lack of marking
on inalienables corresponded directly to their closer semantic bond. However, there
is another possible explanation that will be investigated in the next section.
5.3.5.4 A processing explanation of the formal marking
Having seen the distribution of the differing means of marking possession in
Zoogocho Zapotec, and some of the functional theoretical explanations and
definitions of the different ways of marking attributive possession in the languages
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
143
of the world and the alienable/inalienable distinction, it remains to work towards an
explanation of the different means of marking possession.
Johanna Nichols rightly points out that:
inalienable possession is not primarily a semantic distinction but the
automatic consequence of the closer formal bonding that results in head-
marked possession: inalienables typically include kin terms, part/wholes
and/or body-parts, nouns which are most likely to occur possessed in
discourse, and the formal marking of inalienability simply grammatical izes
that possession. (Nichols 1992:121-122)
Suarez (1985) also hints at the importance of frequency of possession in the
following quote, which discusses various means of marking possession in
Mesoamerican languages.
Nouns are very frequently obligatorily possessed, optionally possessed and
unpossessable. Nouns referring to parts of the body, personal belongings (e.g.
clothes), and kin terms are obligatorily possessed, but in Tlapanec, for
instance, only kin terms are obligatorily possessed, and in Classical Nahuatl
probably any noun could occur, at least in quotation form, as unpossessed
and, in this case, marked with the absolutive suffix. The class of nouns that
never occurs possessed seems to be determined by largely non-linguistic
factors; nouns that are usually unpossessable are those referring to natural
phenomena such as the sun, wind, etc. (Suarez 84)
More generally, Hawkins posits the following principle in “Efficiency and
Complexity in Grammars: Three General Principles” (Hawkins 2003)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Minimize forms
The human processor prefers to minimize the formal complexity of
each linguistic form F (its phoneme, morpheme, word or phrasal units) and
the number of forms with unique conventionalized property assignments,
thereby expanding the compatibility of F with a larger set of properties {P}.
These minimizations apply in proportion to the ease which a given P 1 can be
assigned in processing to a formally reduced F with expanded property
compatibilities, (ibid. 135)
Hawkins mentions various examples (such as pronominalization to name but
one) to support this claim which basically says that the human processor prefers to
process more reduced forms as long as the meaning is still easily recoverable. He
then goes on to make the following predictions.
Form Minimization Predictions
a) The formal complexity of each F is reduced in proportion to the
frequency of that F and/or the processing ease of assigning a given PI to a
reduced F (e.g. to zero).
b) The number of unique F:P1 pairings in a language is reduced by
grammaticalizing or lexicalising a given F:P1 in proportion to the frequency
and preferred expressiveness of that PI in performance, (ibid. 137)
For my purposes, (a) provides an easy explanation of the lack of marking of
possession on nouns which are obligatorily possessed such as body-part and kinship
terms. In the case of body part terms and other lexical items which are always or
almost always possessed there is no need to mark anything other than the possessor,
because the listener already knows that they are possessed. The only thing that varies
is the possessor which is marked. This is directly parallel to markedness hierarchies
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
145
or feature hierarchies (ibid. 11). Consider for a moment, the well-known markedness
hierarchy related to number marking in (55).
(55) Singular>Plural>Dual>Trial/Paucal (From (ibid. 11), (23))
As is well known (Greenberg 1966, Croft 1990), a category high in a
hierarchy such as this one will be much more frequent than a category low on this
hierarchy. Conversely, a category low on this hierarchy will be much more likely to
be marked than a category high on this hierarchy. In other words, singular nouns are
more frequent than plural nouns; thus, one would expect to see explicit marking of
plurality in a given language more often than explicit marking of singularity. This
leads to Hawkins’s ‘Quantitative Formal Marking Prediction’.
Quantitative Formal Marking Prediction
For each hierarchy H the amount of formal marking (i.e. phonological and
morphological complexity) will be greater or equal down each hierarchy
position, (ibid. 140)
Now I will return to the tentative classification of possessive marking from
5.3.2, restated here in the form of a table. If one takes the labels on the left to be the
terms in a hierarchy, one can see similar results to that mentioned above for number.
The more frequently possessed nouns are less marked, and vice versa. However a
number of questions remain, some of which will be answered in the next
section.How frequently are nouns from these three classes possessed in texts? What
is the distribution of the middle class? (The nouns which can show either type of
marking.)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 4 6
Table 5.1 Methods of possession in Zoogocho Zapotec
Possession unm arked
(inherent possession)
Possession m arked
m orphologically
(Inherent possession)
Possession m arked
syntactically
(non-inherent
possession)
A lw ays possessed,
sm all closed class
(Inalienables)
yichgha’
yichgh=a’
h ead = lsg
‘m y head’
n/a n/a
Frequently possessed, n/a xkuzhe' kuzh ch ie’
large closed class x-kuzh=e’ kuzh che= e’
(M ore frequently poss-pig=3f pig of= 3f
possessed alienables) ‘her p ig ’ (one she has at
the house)
‘her p ig ’ (which
she m ight be
selling at the
m arket)
N ot very frequently
possessed, large open
class (Less frequently
possessed alienables)
n/a n/a tigr ch eb e’
tigr che= be’
tiger o f= 3inf
‘her tig er’
5.3.5.5 The use of attributive possession in three Zoogocho Zapotec texts
In this section, I will discuss the marking of possession in three texts, one, an
instructional text, which I will call ‘Tigr’, another, a conversation, which I will call
‘G’, and another, a narrative history, which I will call ‘Miner’. These texts were
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
transcribed and translated with the help of my primary collaborator, Alberta Marcial
Martinez. Any errors are obviously my own. Rather than look at every noun in the
texts, I have examined nouns from each of the three types of possession exemplified
in Table One. I will begin by briefly discussing the first text in which a community
leader instructs a group of children in a traditional dance (‘La danza de los tigres' )
which they then perform in the village’s saints day festival.
This text consists of 760 Zoogocho Zapotec utterances6. 1 will examine five
words: xna= ‘mother’, ni= ‘foot’, yoo/lizh= ‘house’, son ‘song [Sp.]’, and tigr ‘tiger
[Sp.]’ The first two, xna=7 and ni=, are both not separately marked for possession
with a possessive marker and belong to the first class which was discussed in section
II. Correspondingly they are always possessed in this text. xna= occurs 14 times in
the text, and is possessed every time. Similarly ni= occurs 28 times in the text and it
is possessed every one of those times. Interestingly, only the possessed form of the
word for ‘house’, lizh=, occurs three times. The other form never occurs in the text.
son occurs 18 times in the text. Four times it is possessed (son chedxo ’ ‘our song’.
son che tigr ‘ the tiger’s song’). 14 times it occurs on it’s own. Tigr is present 15
times in the text. Not surprisingly, it is never possessed.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 4 8
Table 5.2 Possession in Tigr
Possessed Not possessed Total
xna= 14 0 14
ni= 28 0 28
lizh= 3 0 3
yoo (che=) 0 0 0
son (che=) 4 18 22
tigr (che=) 0 15 15
In the second text, a conversation between Alberta and a monolingual 83-
year-old speaker of Zoogocho Zapotec, to whom I will refer as G, the participants
discussed how things had changed in G’s lifetime. The text consists of 1450 Zapotec
utterances. The word xna= ‘mother’ shows up 26 times in those 1450 lines,
possessed each and every time. Similarly, the word for head, yichgh=, occurs three
times, each time possessed and the word for neck, lbaha=, occurs four times,
possessed in each instance. The word for maize, zxoa, and the word for bean, zaha,
occur unpossessed in the text 13 and 12 times respectively. Now, to move on to the
interesting and not so interesting cases, the word for coffee, kafe, occurs 13 times in
total, and is unpossessed 11 of those times, and the word for house, yoo when
unpossessed, and lizh= when possessed, shows up a grand total of 47 times in the
text. It shows up unpossessed 28 times (as yoo) and possessed 19 times (as lizh ).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
149
Table 5.3 Possession in ‘G’
Possessed Not possessed Total
xna= 26 0 26
yichgh= 3 0 3
Ibaha= 4 0 4
lizh=/yoo 19 28 47
zxoa(che=) 0 13 13
zaha (che=) 0 12 12
kafe (che=) 2 11 13
Finally, I will briefly examine a third text, from a 70-year-old bilingual
speaker of Zoogocho Zapotec. This text, which recounts a time when American
miners came to Zoogocho in search of ores, consists of 350 SBZZ utterances. 1 will
examine three words: xna= ‘mother’, lizh=/yoo ‘house’, andyegh ‘rock’. xna=
occurs five times, always possessed, yoo occurs seven times unpossessed, and once
possessed (yoo cheto ‘our house’). I will return to the one time when it occurs
possessed in a moment. The word lizh= occurs four times in the text, possessed
every time. Yegh shows up six times, never possessed.
Table 5.4 Possession in Miner
Possessed Not possessed Total
xna= 5 0 5
lizh= 4 0 4
yoo (che=) 1 7 8
vegji 0 6 6
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Before finishing this section, it will be useful to examine the use of the words
for house. As has already been noted, there are two ways of possessing the word yoo
‘house’, one is by using the suppletive form lizh=, the other is to use the word yoo
and the possessive preposition che=. When the word is unpossessed, it is generally
when the speaker was referring to something that happened to the house, such as (56)
below, from the G text. When yoo is used with che - it is also when discussing
something which has happened to the house as in (57) below, from the Miner text.
When it shows up possessed, as lizh=, it is generally when the referring to something
which occurred in the location or to the owners of the house, as in (58) below, from
the G text.
(56) na’ kate bento yoo nga
na’ kate b-en=to yoo nga
and when comp-make=lplexcl house demmed
‘and when we made this house.’
(57) na’ gonteto yoo cheto
na’ g-on=te=to yoo che=to
and pot-make=int=lpl(excl) house of=lplexcl
‘We are going to make our house.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(58) nadxe yeghdxo kafe lizhe’
nadxe y-egh=dxo kafe lizh=e’
afterwards pot-drink=lplincl coffee poss.house o f
‘Afterwards we drank coffee in his house’
The hypothesis stated in the previous section appears to have been tentatively
confirmed. It seems that those items which can be possessed without having an overt
possessive marker do occur always possessed in the texts examined so far. Those
that are possessed syntactically appear to occur mostly unpossessed. Those that show
both syntactic and morphological marking of possession seem to be more variably
possessed. Additionally, those belonging to the middle class show the morphological
marking for possession when they are more inherently possessed and the more
syntactic marking for possession when they are non-inherently possessed, such as
when they are in the process of being constructed and when they are not currently
being resided in and therefore are not intimately associated with the possessor.
5.3.6 Determiners
The determiner is a clitic which has three main variants; one which occurs
following a non-nasal consonant as in (59) is =en ’ , one which occurs with words
ending in n or nh is =na ’ , as in (60), and, finally, the one which occurs after a vowel
* S
is =n or =na ’ in free variation as in (61). Determiners occur at the end of a noun
phrase, as in (62).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
152
(59) yet=en’
tortilla=det
‘the tortilla’
(60) dizha zxon=na’
tongue Zxon=det
‘Zxon language’
(61) a)zxoa=n
com=det
‘the com’
b) zxoa=na’
com=det
‘the com ’
(62) mbis shish=en’
cat white=det
‘the white cat’
5.3.7 The directional clitic
The clitic =le is used to indicate direction towards a location. It is always
phrase final. The following are examples of the directional clitic.
(63) soalaga=le
soalaga=dir
‘Towards Soalaga’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 5 3
(64) yaa zxen=le
mountain big=dir
‘Towards the big mountain’
5.3.8 Overall NP order
NP word order is relatively straightforward. As already seen, numerals and
other quantifiers come before the noun and adjectives come after, as in the following.
(65) to sita’ zito’ zxen
one country distant large
‘A large, distant country’
(66) yogo manzana ga’ zxen
all apple green big
‘All big, green apples’
(67) xon libr nga
three books demmed
‘These three books.’
(68) xon libr exo’
three books old
‘Three old books’
(69) yogo libr nga
all books demmed
‘All these books’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 5 4
(70) yogo libr exo nga
all books old demmed
‘All these old books’
The only combination that I have seen disallowed is when a demonstrative precedes
a numeral, as in the following.
(71) * yogo xon libr
all three books
‘all three books’
One reason I can point to for this at this point is that numerals, when taken by
themselves, already quantify the NP they modify and are often interpreted as being
the totality of the set as in the following.
(72) taph bi chi=e’ nita’ ni’
four child poss=3f stat.be demprox
‘All four of his children are here.’
The ordering for a maximal NP is therefore: Quantifier/Numeral Noun
Adjective Possessive Demonstrative/Determiner as seen in (73).
(73) shone be’ko’ xo chi=a’ ni’
three dog old of=lsg demprox
‘these three old dogs of mine’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 5 5
5.4 Declarative clauses
Before going on to discuss the various types of declarative clauses in SBZZ,
it will be useful to discuss what the main grammatical relations in the language are.
Subjects are those N P’s which, in non-exceptional cases (cf. 6.4), immediately
follow the verb and are semantically the agent, experiencer, or undergoer. Direct
objects are the semantic patient/stimulus/addressee/causee in a transitive clause.
Direct objects immediately follow the subject in non-exceptional clauses (cf. 6.4).
Indirect objects are recipients, instruments, or former direct objects which have been
demoted by a causative (cf. 6.1). Note that the strict VSO order found in most non-
pragmatically marked transitive clauses is evidence for SBZZ being a
nominative/accusative language. Grammatical relations will be discussed in greater
detail in Chapter Eight.
5.4.1 Intransitive clauses
The most basic, non-pragmatically marked clause in Zoogocho Zapotec
consists of an inflected verb with either a pronominal clitic or a full noun phrase as
the subject following the verb, as in (74) and (75) below.
(74) dxi=a’
sit=lsg
‘I sit.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
156
(75) b-zxit lalo
comp-jump lalo
‘Lalo jumped.’
5.4.2 Transitive clauses
In clauses which are not pragmatically marked, direct objects directly follow
the subject as seen in (76) through (79) below.9
(76) kate b-edey-a mansia yish=en’
when comp-ven-take Amansia grinding_stone=det
‘When Amansia came to take the grinding stone...’
(77) bi gud-aw=a’ yet=en’ dao
neg comp-eat=lsg tortilla=det com tassel
‘I didn’t eat tortillas made from com tassels.’
(78) n-chee=ba’ lheba’
stat-take=3an 3 an
i t takes them.’
(79) g-zxi=to=n
pot-buy= 1 plexcl=3 inan
‘We buy it.’
5.4.3 Ditransitive clauses
Simple indirect objects, such as the one in (80), which are not introduced by
another morpheme, are found in non-derived ditransitive clauses. Note that the
canonical order for ditransitive clauses is: V SU IO DO. The order of indirect object
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
! 51
and direct object is quite free when they are non-pronominalized as seen in (81) and
(82). Also, if either the 10 or DO is pronominalized, it will attach to the subject
clitic, as in (83) and (84). I will discuss other indirect objects which are introduced
by prepositions, verbal suffixes, or causatives in greater depth in the chapter on
complex constructions.
(80) b-enh bidao neda’ to libr
comp-give child lsg one book
‘The child gave me a book.’
(81) b-enezxghw=a' to libr to bidao
comp-give=lsg one book one child
‘ I gave a book to a child.’
(82) b-enezxghw=a’ to bidao to libr
comp-give=lsg one child one book
‘I gave a child a book.’
(83) b-enh=a=be’ to libr
comp-give=lsg=3inf one book
‘I gave him a book
(84) b-enh=a=n lalo
comp-give= 1 sg=3 inan lalo
‘I gave it to Lalo.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 5 8
5.4.4 Negation
Negation of a clause is achieved by including a preverbal particle as seen in
(85).
(85) bi dx=aog mbis yet=en’
neg cont=eat cat tortilla=det
‘The cat is not eating the tortilla.’
(86) bi g-ak y-id-e’
neg pot-be pot-come-3f
‘He cannot come.’
(87) bi b-it=be’ neghe’
no comp-come=3inf yesterday
‘He didn’t come yesterday.’
To negate a noun phrase, another negative word is oftentimes used in
conjunction with the regular negative marker. The following two examples show the
construction bi...neto. In Cheryl Black’s dissertation, she claims that a similar form
to neto in Quiegolani Zapotec is actually a verb (Black 1994). These may not be
more of an emphatic negative than anything else.
(88) bi b-daow-a’ neto yet
neg comp-eat=Tsg neg tortilla
‘I ate no tortillas.’ (I didn’t eat even one tortilla.)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(89) bi b-le’id-a’ neto bis
neg comp-see=lsg neg cat
‘I saw no cat.’
Note that neto can be used on its own as in the following example.
(90) neto zha kana zha=a’ tiend=en’
neg day that cont.go=lsg store=det
‘There isn’t one day that I don’t go to the store.’
The word, nono ‘nobody’, is used as in the following.
(91) nono zoa’
nobody stand
‘There is no one here.’
In addition, the word gag can be used preceding a quantifier, thereby negating the
quantifier.
(92) gag yogo benhe dx-ak=de’ be’ko’
not all people cont-like=3fexp dog
‘Not everyone loves dogs.’
‘Never’, bite occurs before the verb.
(93) bite’ dx-ib=e’ ni’
never cont-come=3f demprox
‘He never comes here.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
160
5.4.5 Adverbs
The adverbial suffixes have already been discussed in Chapter Four. The
emphasis, now, will be on prosodically independent adverbs and their placement in
the sentence.
5.4.5.1 Time
While temporal adverbs may potentially be placed between the verb and its
non-pronominal subject as in (94) or between a verb and its object as in (95), they
are often found in sentence final position as in (96), and in sentence initial position
as in (97) and (98).
(94) dx-on dezd octubr xoan to yoo
cont-make since[Sp.] October Juan one house
‘Juan’s been building a house since October.’
(95) b-en=a’ neghe to yoo
comp-make=lsg yesterday one house
‘I built a house yesterday.’
(96) b-id zxoan gxe’?
comp-come Juan tomorrow
‘Will Juan come tomorrow?’
(97) gxe’ ba zan zxoan
tomorrow already stat.com e, Juan
‘Juan will come tomorrow.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
161
(98) mil novecientos cuarenta dx-on=a=n
1940 cont-make=l sg-"3inan
‘In 1940,1 made it.’
5A.5.2 Place
Locative adverbs are often found sentence finally as in (99).
(99) na chup=be’ sh-naa-te=be’ yaa /xen le
and two=3inf cont-look-emph=3inf mountain bigv dir
‘And the two of them are looking towards the big mountain.’
Locative adverbs can also occur sentence initially as in (100).
(100) yaa zxen=le ze-naa=be
mountain big=dir pl.cont-look=3inf
‘Towards the mountain they look.’
5.4.5.3 Manner
Many manner adverbs are adverbial clitics, but those that are not tend to be
relatively free in their placement as seen in (101) through (104) below.
(101) be-na’ dx-lonhgh=e’ sholazhe
clan=demdist cont-run=3f slowly
‘That person runs slowly.’
(102) sholazhe dx-longh be=na’
slowly cont-run clan=demdist
‘That person runs slowly.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
162
(103) be=na’ sholazhe dx-longh
clan=demdist slowly cont-run
‘That person runs slowly.’
(104) dx-lonhgh be=na sholazhe
cont-run clan=demdist slowly
‘That person runs slowly.’
The following example shows that manner adverbials are preferentially not placed
between verb and subject.
(105) * dxlongh sholazhe bena’
Similarly, both manner adverbial and subject may not both be fronted.
(106) * sholazhe bena’ dxlongh
5,4.5.4 Time/Manner/Place
When multiple adverbs co-occur, the preferred order seems to be as seen in
(107).
(107) neghe dx-longh benhebio bedaones dxeele
yesterday cont-run person masc quickly below
‘Yesterday the man ran quickly down below.’
While (107) is acceptable to native speakers, it is preferable to break this up into two
clauses, as in other expressions of destination, as in (108).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(108) neghe dx-longh benhe bio bedaones, zegh=e’
dxeele
yesterday cont-run person masc quickly stat.go=3f
below
‘Yesterday the man ran quickly down below.’
In (108), bedaones could potentially belong to either clause. Further research is
necessary to determine which clause it actually belongs to. Also, while either the
locative or the manner adverb can be placed verb initially by themselves, only one
adverb may be fronted. Also, the temporal adverb neghe ‘yesterday’, can be placed
after the verb, but not with both the manner and place adverb as well.
(109) bedaones dx-longh benhe bio
quickly cont-run person masc
‘The man runs quickly.’
(110) *bedaones neghe dx-longh benhe bio
(111) * dxeele neghe dxlongh benhe bio
(112) *bedoanes dxeele dxlongh benhe bio
(113) dx-longh benhe bio neghe
cont-run person masc yesterday
‘The man ran yesterday.’
(114) * dx-longh benhe bio bedaones dxeele neghe
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
164
5.5 Pragmatic Fronting of Elements
5.5.1 Topicalization
Topicalization, which will be discussed and defined in 8.6, involves the
dislocation of a noun phrase to the left of the verb, with a corresponding resumptive
clitic or independent pronoun following the verb as seen in (115) below. Both
subjects and objects, as seen in (116), may be topicalized.
(115) benhe g-onh=e’ yete=n’ na
person pot-give=3f tortilla=det demdist
‘The people give food there.’ (In a reference to a place where, in addition to
the daily wage, people also feed their workers.)
(116) yet=en’ dx-on=a=n
tortilla=det cont-make= 1 sg=3 inan
‘Tortillas, I make.’
Note that when a subject is a full noun phrase, an independent pronoun is used.
(117) bedo b-et zxoan lhebe
Pedro comp-hit Juan 3 inf
‘Pedro, Juan hit him.’
5.5.2. Focus
Focus will also be discussed in greater detail in 8.6. The noun phrase that is
focussed is also found in sentence initial position without a corresponding
resumptive pronoun, as seen in (118) below. Once again, both subjects, and objects,
such as in (119), can be focused.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(118) primer yag=en’ zoa
first tree=det stat.stand
‘The tree is what goes first.’ (This sentence occurred in a text where one
speaker was asked to recreate a scene that another speaker was describing.
This was in response to ‘where does the tree go?’ The speaker was trying to
specify that the tree, as opposed to other potential referents, went in such and
such a place.)
(119) pelot-en’ dxy-itghe-d=e’
ball-det cont-play-instr=3f
‘She played with the ball.’
5.5.3 Fronting of non-core arguments
Other elements which are not core arguments can also be fronted, as in the
following example.
(120) lenh yaa wag dx-ogo=a’ wage=n’
with iron firewood cont-cut=lsg firewood=det
‘With an axe, I cut firewood.’
In these cases, the fronting serves to emphasize the element being fronted. 1
will discuss fronting in greater detail in Chapter Eight.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
166
5.6 Interrogative sentences
5.6.1 Overall
Questions normally1 0 involve a rising intonation, or the sentence final particle
ke for yes/no questions or a sentence initial interrogative for content questions. In the
case of content questions, much as is the case with relative clauses, as 1 will show in
the following chapter, there is no resumptive pronoun in the position questioned
unless it serves to disambiguate a potentially ambiguous VX structure. Also, many
question words like no ‘who’, and bi ‘which’ result in the pied-piping of the
argument that is being questioned resulting in gaps in the sentential structure. This
too will be taken up later in this section.
5.6.2 Yes/no questions
In example (121) below, shows the use of rising intonation in order to ask a
yes/no question.
( 121) ------------------------ /
sh-tas bidao’ chi=a’
cont-sleep child of=lsg
‘Is my child sleeping?’
Example (122) seen below shows the use of the tag question formative, ke.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
167
(122) na zghe-zhia=dxgwa tushe nahago=ba, ke?
and plur.stat-stand=emph pointy ear=3an, right?
‘And their ears are very pointy, right?’
The semantic difference between these two expressions is that the basic yes/no
question with ke seems to be more emphatically asking for an answer.
5.6.3 Content questions
5.6.3.1 no
In order to ask who did something one uses the word no which means ’who'.
I have a few examples of bi the word which means ‘what’ being used to ask ‘who’ as
well, but those other examples will be shown later in the present chapter. Arguments,
which would otherwise have to be encoded post verbally, cannot be if they are the
element which is being questioned unless they are redundantly present as a means of
disambiguating the sentence. This will be discussed in greater detail at the end of this
section.
(123) no b-lee
who comp-arrive
‘Who came?’
(124) no zoa yoo
who stand house
‘Who is at the house?’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(125) no b-e-gan yoo
who comp-freq-remain house
‘Who remained at the house?’
(126) no zegh tiend
who comp.go store
‘Who went to the store?’
(127) no g-olghe
who com p-bebom
‘Who was bom?’
The following show examples of questions with the direct object as the object of the
question word. Note that there is no direct object following the verb.
(128) no b-le’i=do’
who comp-see=2sgexp
‘Who did you see?’
(129) no b-dizxgh-o’
who comp-pay=2sg
‘Who did you pay?’
The following example shows the use of no as something that would roughly
translate to ‘which’ in English. Note that in this case the noun phrase no benhe bio
‘which man’ comes before the verb.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
169
(130) no benhe bio b-le’i=do’
what person masc. comp-see=2sgexp
‘Which man did you see?
The following examples show the use of no as ‘whose’. Note the order no
xna ’ , the opposite from what one would expect from a possessive construction in a
non-interrogative environment, and the lack of an obligatory possessor on the
normally obligatorily possessed NP.
(131) no x-na zoa Los Angeles
who poss-mother stand Los Angeles
‘Whose mother is in Los Angeles?’
(132) no kuzhe b-le’i=do’=be’
who back comp-see=2sgexp=3inf
‘Behind whom did you see him?’
This example could equally well be translated, if awkward in English, as ‘At
whose back did you see him?’
Furthermore, in order to question an indirect object, no is used with the
possessive marker che= as in the following example.
(133) no chi=e’ b-eselh=o=n
who of-3f comp-send=2sg=3inan
‘To whom did you send it?’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 7 0
This is an example of the indirect object ‘lowering’, which I will discuss in the
following chapter in section 6.3. Note the following corresponding declarative
sentences.
(134) b-eselh=a=n
comp-send= 1 sg=3 inan
‘I sent it to Lalo.’
(135) b-eselh=a’ libr
comp-send= 1 sg book
‘I sent a book to Lalo.’
5.6.3.2 bate’
In order to ask ‘when’, one uses bate ’ . The following is an example of this.
(136) bate’ y-egh-o’ nis
when pot-drink=2sg water
‘When are you going to drink water?’
5.6.3.3 ga/gan
The following examples use ga/gan1 1 to ask ‘where’.
(137) ga b-le ’ i=de=ne ’
where comp-see=3fexp=3fobj
‘Where did he see him?’
lalo
lalo
che lalo
of lalo
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
171
(138) ga b-le’i=de=nda’
where comp-see=3 fexp= 1 sgobj
‘Where did he see me?’
The basic locative question has two forms. The first fonn, which seems to be more
common, is as follows.
(139) WHERE Positional Figure
Gan dxi be’ko’?
Where sit dog
‘Where is the dog?’
The answer to this type of locative question generally presupposes the figure and
position and therefore does not include the positional verb or the subject and consists
solely of the relational noun and the ground.
(140) lhoo yixe
in grass
‘In the grass’
An alternative form doesn’t include the positional and is as follows.
(141) WHERE Figure
gan be’ko’
where dog
‘Where is the dog?’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The answer to this type of question is generally a fully formed construction
which includes a positional verb.
(142) dxi=ba lhoo yixe
stat.sit=3an in grass
‘It’s sitting in the grass.’
For more on locative constructions, see 7.2.1.
5.6.3.4 bal/bale’1 2
In order to ask ‘how many’, one uses the question word, bale Note that
questions like ‘How many trees are there?’ require the subject to directly follow
bale However, examples (144) and (147) below both show that if the entity is
nonspecified, bale ’ is all that is necessary. Note however that these examples do
appear to show the optional retention of a pronoun. I will come back to this point in
chapter eight.
(143) bale’ plum n-ak(=en)
how m any pens stat-be(=3inan)
‘How many pens are there?’
(144) bale’ n-ak(=en)
how m any stat-be(=3inan)
‘How many are there?’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 7 3
(145) bale’ bidao n-ak(=be’)
how m any children stat-be(=3inf)
‘How many children are there?’
(146) bale’ yag=en n-ak(=en)
how m any trees=det stat-be(=3inan)
‘How many trees are there?’
(147) bale’ zak
H ow m any stat.come
‘How many are coming?’
(148) bal dxioo ba gu-dag=dxo
how m any lplincl already comp-eat=lplincl
‘How many of us already ate?’
Bale ’ is related to the word for some, seen below.
(149) bale’ benhe bio zghe-nita zxwikwe’
some person masc. Plural-exist poss-dog=3f
‘Some men have dogs.’/ ‘How many men have dogs?’
(149) is an ambiguous sentence, meaning either of the two glosses which 1 have
given above.
Note that there are many examples such as (150) which do not include a
resumptive pronoun.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(150) bale’ vaca g-os-ede
how m any cows comp-pl-passed
‘How many cows have gone by?’
However, similar to that which will be seen later in this section and in 6.5.7, the
pronoun may be retained if it serves to disambiguate the clause. Thus, in (151), - e ’
is retained, as without it, although a bit awkward, (152), in which the second person
is interpreted as the subject would be the result.
(151) bale’ benhe bio g-os-ot=e’ le
how many person masc comp-pl-hit=3f 2sg
‘How many men hit you?’
(152) bale’ benhe bio g-os-ot le
how many person masc comp-pl-hit 2sg
‘How many men did you hit?’
There are also examples such as (153) which show similar retention of a pronoun in
order to disambiguate the clause.
(153) bale’ bidao g-os-ot=be’ le
how m any children comp-hit=3inf 2sg
‘How many children hit you?’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
175
5.6.3.5 kaka
The following examples show the use of kaka to ask ‘how much’. The
difference between bale ’ and kaka is that one questions count nouns (bale ’) and one
questions mass nouns (kaka).
(154) kaka nis yozh lhoo vaso
how m uch water is.inserted in glass
‘How much water is there in the glass?’
(155) kaka b-dizxgh=o ’
how m uch comp-pay=2sg
‘How much did you pay?’
(156) kaka tsaka=n plum
how m uch cost=3inan pen
‘How much does the pen cost?’
5.6.3.6 nake*
The following are examples of nake ’ which is used to ask ‘how’.
(157) nake’ b-dizxgh=o’
how comp-pay=2sg
‘How did you pay?’
(158) nake’ dx-igib-dx=o’ ladxe
how cont-wash-int=2sg clothes
‘How do you wash your clothes?’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(159) nake’ b=lazho taso=na’
how comp-fall cup=def
‘How did the cup fall?’
5.6.3.7 bicheen
The following two examples are ways in which one can say ‘why’, using bicheen.
(160) bicheen b-dizxgh=o’
why comp-pay=2sg
‘Why did you pay?’
(161) bicheen b-dxogh=o ’
why comp-leave=2sg
‘Why did you leave?’
5.6.8 bi
I have left potentially the most complicated set of data for last. The following
section will describe how bi is used in a way that corresponds to ‘what’, bi ‘what’ is
segmentally the same as bi ‘ neg’.
(162) bi b-zi=o’
what comp-buy=2sg
‘What did you buy?’
(163) bi dx-aog=e’ zxoan
what cont-eat=3f Juan
‘What is Juan eating?’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
177
Example (164) shows that when an object is questioned using bi it is pied-
pied to sentence initial position.
(164) bi or=en’ zaa=o’
what hour=det stat.leave=2sg
‘At what time do you leave?’
The following four examples show the potential confusion between bi meaning
‘what’ and the negative bi.
(165) bi b-le’i=do’
what comp-see=2sgexp
‘What did you see?’/ “You didn’t see.’
(166) bi bi b-le’i=do’
what neg comp-see=2sgexp
‘What didn’t you see?’
(167) bi b-zi=o’
what comp-buy=2sg
‘What did you buy?’/ ‘You didn’t buy.’
(168) bi bi b-zi=o’
(what neg) comp-buy=2sg
‘What didn’t you buy?’
Examples like (169), (170), and (171) show that the correct analysis should be that
the first bi is the one meaning ‘what’ and that the second bi is the negative. One can
see from examples (169)-(171) that when an object is questioned, it follows the bi
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
meaning what. (171) is what crucially confirms this fact, as it does not have any
potential ambiguity with respect to a potential negative interpretation.
(169) bi camion dx-o’=o’
what bus cont-take=2sg
‘Which bus do you take?’
(170) bi camion bi dx-o’=o’
what bus neg cont-take=2sg
‘What bus do you not take?’
(171) bi pelicula bi b-lei=do’
what movie neg comp-see=2sgexp
‘What movie didn’t you see?’
(172) through (174) provide additional confirmation of this.
(172) pelicula=na’ bi b-le’i=do’
picture-det no comp-see=2sgexp
‘The picture, you didn’t see it.’
(173) bi pelicula b-le’i=do’
what movie comp-see-2sgexp
‘What movie did you see?’
(174) bi libr bi b-zi=o’
what book neg comp-buy=2sg
‘What book didn’t you buy?’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
179
5.6.4 Interrogative clauses- a synopsis
As seen in (175) below, content questions can potentially be ambiguous.
There are two potential interpretations, one in which Pedro is the subject and one in
which Pedro is the object. Normally, the first NP (bedo) found after the verb is
interpreted preferentially as the subject, the primary interpretation of the position of
the gap is that of the object, and a pronoun can be optionally retained (as in (176)) in
order to disambiguate the clause. However, as seen in (177), a resumptive pronoun
can be included in the subject position to assert that Pedro is the object. The general
nature of SBZZ as a verb initial language lends itself to potential ambiguities of this
sort.
(175) No gu-dap_bedo _?
Who pot-slap _ pedro _?
‘Who will Pedro slap?’/ ‘Who will slap Pedro?’
(176) No gu-dap=e’ bedo?
Who pot-slap=3f pedro
‘Who will slap Pedro?’
(177) No gu-dap bedo (lee)?
who pot-slap pedro 3f
‘Who will Pedro slap?
Multiple wh-questions appear to be ruled out by the grammar, at least among the
speakers whom I have asked and on the basis of the texts that I have analyzed.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
180
5.7. Zoogocho Zapotec and Greenberg’s universals
• 13
Mitla Zapotec was one of the original languages in the sample which
Greenberg used for his seminal paper, ‘Some universals of grammar with particular
reference to the order of meaningful elements’ (Greenberg 1966). SBZZ, on the
surface level of word order, corresponds exactly to his Type 1 of Appendix II. being
VSO/Pr/NG/NA (ibid. 87). It also could very well stand in as a double for the exact
variety exemplified in Appendix I: VSO with prepositions, Noun Adjective, Noun
Demonstrative, and Numeral Noun. Having looked at basic ways in which SBZZ
corresponds to the variety of Zapotec used in Greenberg’s sample, I will now move
on to look at some of the specific claims made in his collection of ‘Universals".
Universal 1 which states that ‘(I)n a declarative sentence with nominal
subject and object, the dominant order is always one in which the subject precedes
the object’ (ibid. 61) is easily verified. Although OVS orders exist, they are, as will
be seen later, pragmatically marked structures.
Universal 2 states that ‘(I)n languages with prepositions, the genitives almost
always follow the governing noun, while in languages with postpositions it almost
always precedes’ (ibid. 62). This is also validated by the data seen above from
Zoogocho Zapotec.
Universal 3 which states that ‘languages with VSO order are always
prepositional’ (ibid. 62) is also verified by the data. Universals 4 (‘With
overwhelmingly greater than chance frequency, languages with normal SOV order
are postpositional’ (ibid. 62)) and 5 (‘ If a language has dominant SOV order and the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
181
genitive follows the governing noun, then the adjective likewise follows the noun’
(ibid. 62)) are both inapplicable given that SOV is neither the dominant nor normal
order. However, SBZZ is also consistent with universal 6 which states that ’All
languages with dominant VSO order have SVO as an alternative or as the only
alternative basic order’ (ibid. 63). Universal 7 (‘If in a language with dominant SOV
order, there is no alternative basic order, or only OSV as the alternative, then all
adverbial modifiers of the verb likewise precede the verb’ (ibid. 63)) does not apply
to VSO languages such as SBZZ. Universal 8 is borne out. As seen in example
(178), in a yes-no question which is not a tag question, the intonational patterns do
seem to be ‘reckoned from the end o f the sentence rather than from the beginning’
(ibid. 63).
(178) ^
zoa yag yedx=en’?
stand tree pine=det
‘Is there a pine tree?’
Universal 9 which states that ‘(W)ith well more than chance frequency ,
when question particles or affixes are specified in position by reference to the
sentence as a whole, if initial, such elements are found in prepositional languages,
and, if final, in postpositional’ (ibid. 64) is not verified by the data from Zoogocho
Zapotec. This bears on Greenberg’s statement that ‘Zapotec (I/Pr) has either an
initial particle alone or this same particle in conjunction with a final particle’ (ibid.
64). Zoogocho Zapotec does not have an initial particle and the status of the final
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
particle, which indicates a tag question more than anything else, has been discussed
in greater depth above.
(179) kuzhle n-dxe=be’, ke?
pig focus stat-carry=3inf, really
‘He’s carrying a pig, isn’t he?’
Note the focus marker following pig. This focus marker is related to the base
for the third person pronominal forms.
The fact that SBZZ, as a VSO language, only has sentence initial particles
which precede the word they question and does not have question particles or affixes
which follow the word they question confirms Universal 10 (‘Question particles or
affixes, when specified in position by reference to a particular word in the sentence,
almost always follow that word. Such particles do not occur in languages in
dominant order VSO’ (ibid. 64)).
Given that the basic word order of SBZZ is verb-initial, Universal 11 which
states that ‘(I)nversion of statement order so that the verb precedes subject only
occurs in languages where the question word or phrase is normally initial. This same
inversion occurs in yes-no questions only if it occurs in interrogative word
questions.’ (ibid. 65) is inapplicable.
Universal 12 is also definitely confirmed by the data from SBZZ. Universal
12 states that VSO languages have interrogative words in sentence initial position, as
has been seen earlier in the current chapter (ibid. 65).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
183
Universal 13, ‘(I)f a nominal object always precedes the verb, then verb
forms subordinate to the main verb also precede it’ (ibid. 66), is not applicable to the
normally verb-initial Zoogocho Zapotec.
Universal 14, ‘In conditional statements, the conditional clause precedes the
conclusion as the normal order in all languages’ (ibid. 66), finds no opposition in
SBZZ as seen in example (180).
(180) CONDITIONAL CONCLUSION
shi dx-een=de, dx-o-dxo avante
if cont-want-3fexp, cont-get=lplincl ahead
‘If we want to, we can get ahead.’
Universal 15 which states that ‘(I)n expressions of volition and purpose, a
subordinate verbal form always follows the main verb as the normal order except in
those languages in which the nominal object always precedes the verb’ (ibid. 66) is
also verified as seen in the following example.
(181) dx-eene=be’ y-id=be’ ni’
cont-want=3inf pot-come=3inf demprox
‘He wants to come here.’
Universal 16 which states that ‘(I)n languages with dominant order VSO, an
inflected auxiliary always precedes the main verb’ (ibid. 67) can be seen in the
following example in which the auxiliary -ak= which is inflected for aspect precedes
the main verb.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
184
(182) bi g-ak y-id=e’
neg pot-be pot-come=3f
‘He cannot come.’
See the discussion of auxiliary constructions in 6.5.4 for more details.
Universal 17 which can be stated ‘(I)f VSO then in all likelihood NA' is also
valid for SBZZ as seen above. Universal 18 (‘When the descriptive adjective
precedes the noun, the demonstrative, and the numeral, with overwhelmingly more
than chance frequency, does likewise’ (ibid. 68)) does not apply given the
postnominal nature of descriptive adjectives in SBZZ.
Universal 19, which states that ‘when the general rule is that descriptive
adjective follows, there may be a minority of adjectives which usually precede' (ibid.
68), is valid if one considers the following examples applicable.
(183) be zaan be’ko’
clan many dog
‘many dogs’
(184) be’ko’ be zaan
dog clan many
‘many dogs’
Universal 20 ‘(W)hen any or all of the items-demonstrative, numeral, and
descriptive adjective- precede the noun then they are always found in that order. II"
they follow, the order is either the same or its exact opposite’ (ibid. 89) is verified for
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
185
the descriptive adjective and demonstrative, as the following example shows. With
only two items, this is really a moot point.
(185) Quantifier N DescAdj Dem
shone libr exo nga
three books old demmed
‘these three old books’
Adverbs do not follow the adjective they modify but precede them, as seen in (186)
below, however ‘the qualifying adjective follows the noun and verb precedes its
nominal object’ (ibid. 69), which makes Universal 21 irrelevant to the current
discussion.
(186) leka fwert n-ak kafe=n
much strong stat-be coflee=det
‘The coffee is very strong.’
SBZZ is prepositional and the order of comparison of superiority is adjective-
marker-standard, as one would expect from Universal 22 (‘If in comparisons of
superiority, the only order or one of the alternative orders, is standard-marker-
adjective, then the language is postpositional. With overwhelmingly more than
chance frequency if the only order is adjective-marker-standard, the language is
prepositional’ (ibid. 71)), and as seen in (187) below.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
186
(187) adjective marker standard
n-ak=dx bdxee be lis kleka’ be’ko’
stat-be=more ant clan small comp dog
‘Ants are smaller than dogs.’
Universal 23 states that ‘If in apposition the proper noun usually precedes the
common noun, then the language is one in which the governing noun precedes its
dependent genitive. With much greater than chance frequency, if the common noun
usually precedes the proper noun, the dependent genitive precedes its governing
noun’ (ibid. 71). As already seen, the dependent genitive precedes its governing
noun, but as seen in the following examples, the common noun precedes the proper
noun, thereby verifying this claim as well.
(188) yezx zxghozho
town zoogocho
‘The town of Zoogocho’
(189) yego estudiante
river student
‘Guelatao river’
Universal 24 which talks of relative clauses preceding their head noun does
not apply to SBZZ. Universal 25, which states that if the pronominal object fo llo w s
the verb, so does the nominal object, also holds for SBZZ.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
187
(190) y-egh=a’ ms
pot-drink=lsg water
‘I will drink water’
(191) gu-nopa lia lank neda’
pot-kiss lady angela water
‘Dona Angela will kiss me.’
To sum up the previous section on syntax, SBZZ is, as will be shown in
much more depth in Chapter Eight, a prototypical VSO language, and could have
stood in for Mitla in Greenberg’s original sample. While this is all very unsurprising,
it is a good thing to be able to confirm. There have been over 35 years of work done
on typology and on linguistic universals since Greenberg wrote the paper 1 have been
discussing. There have been numerous refinements to this theory such as Hawkins
(1983, 1994, 2001), Dryer (1992, 1996), etc. However, this is the initial starting
point of modem typology and I am using it, along with other works, as one of the
initial starting points of this typological descriptive grammar.
I will now briefly examine the claims which Greenberg made about
morphology, continuing with the list of universals. Universal 26 (‘If a language has
discontinuous affixes, it always has either prefixing or suffixing or both’ (ibid. 73));
this, discontinuous affixes (infixes and circumfixes), does not apply to Zoogocho
Zapotec given the lack of discontinuous affixes.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
188
Universal 27, ‘(I)f a language is exclusively suffixing, it is postpositional; if
it is exclusively prefixing, it is prepositional’ (ibid. 73), does not apply, as S13ZZ is
neither exclusively prefixing or suffixing. Nor do ‘both the derivation and inflection
follow the root or.. .both precede the root’ (ibid. 73) making Universal 28 (‘ If both
the derivation and inflectional follow the root, or they both precede the root, the
derivation is always between the root and the inflection’ (ibid. 73)) inapplicable.
Universal 29 (‘If a language has inflection, then it always has derivation’ (ibid. 73))
is confirmed, as SBZZ has both inflection and derivation as seen in chapter four.
Universal 30, which claims that ‘ If the verb has categories of person-numbcr
or it has categories of gender, it always has tense-mode categories’ (ibid. 73), is
worthy of a very brief bit of discussion. SBZZ verbs do indeed have tense-mode
categories, and a limited number category, but no real productive person or gender
categories, as I have demonstrated in Chapter Four. Thus, given the lack of person or
gender categories, I must conclude that SBZZ has little to contribute to discussions
of this universal.
Universal 31 states that ‘ (I)f either the subject or object noun agrees with the
verb in gender, then the adjective always agrees with the noun in gender’ (ibid. 74).
Given the fact that I have shown that there is no clear person or gender agreement
between the noun and the verb, this universal is not applicable to Zoogocho Zapotec.
Universal 32, ‘(W)henever the verb agrees with a nominal subject or nominal object
in gender, it also agrees in number’ (ibid. 74), similarly does not apply.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
189
Universal 33 which states that ‘when number agreement between the noun
and verb is suspended and the rule is based on order, the case is always one in which
the verb precedes and the verb is in the singular’ (ibid. 74) is inapplicable, as the rule
is never based on order. Examples like (192)-(195) below show that order is not
relevant and thus the universal is not applicable to SBZZ.
(192) chupe blozh=dao nkwaa=ba
two frogs=dao stat.be_stacked_up=3an
‘Two little frogs are stacked up there.’
Similarly, one can see the following.
(193) chupe blozh=dao zghe-nkwaa=ba
two frogs=dim stat.pl.bestackedup =3an
‘Two little frogs are stacked up there.’
(194) na zghe-nkwaa pur
demdist stat.pl.bestackedup only
‘There’s only little frogs in there’
(195) nkwaa chupe blozh=dao
stat.b estack ed u p two frog=dim
‘There’s two little frogs here.’
SBZZ has neither a morphological dual nor a trial, so Universal 34. u (N)o
language has a trial number unless it has a dual. No language has a dual unless it has
a plural’ (ibid. 74), does not apply. Universal 35, ‘(T)here is no language in which
the plural does not have some non-zero allomorphs, whereas there arc languages in
ni
demprox
ni
demprox
blozh=dao
frog=dim
n i
demprox
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
190
which the singular is expressed only by zero’ (ibid. 74) is valid-there are some non
zero allomorphs of the plural(at least with respect to verbal morphology), and the
singular is expressed only by zero.
Universal 36, ‘(I)f a language has the category of gender it always has the
category of number’ (ibid. 74), is also not relevant given that SBZZ does not have
the category o f gender. Universal 37, ‘(A) language never has more gender
categories in non-singular numbers than in the singular’ (ibid.75), is valid as there
are not more gender categories in non-singular numbers than in the singular.
Universal 38, which states that ‘(W)here there is a case system, the only case
which ever has only zero allomorphs is the one which includes among its meanings
that of the subject of the intransitive verb’ (ibid. 75) is marginally relevant,
especially given the dubious nature of the case system of Zoogocho Zapotec: as has
been seen in the discussion of the pronominal system, the only case which is marked
is that for pronominal objects, and the only one which is clearly marked (as opposed
to being a fast speech phenomenon) is that of the third person formal pronominal
object clitic. Universal 39, ‘(W)here morphemes of both number and case are present
and both follow or both precede the noun base the expression of number almost
always comes between the noun base and the expression of case’ (ibid. 75), is not
applicable, as the noun is not inflected. Similarly, Universal 40, ‘(W)hen the
adjective follows the noun, the adjective expresses all the inflectional categories of
the noun. In such cases the noun may lack overt expression of one or all of these
categories’ (ibid. 75), is not applicable.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Being a verb initial language, ZZ has nothing to contribute to Universal 41
which states that, 4 (I)f in a language the verb follows both the nominal subject and
the nominal object as the dominant order, the language almost always has a case
system’ (ibid. 75). Universal 42, ‘(A)ll languages have pronominal categories
involving at least three persons and two numbers’ (ibid. 75), is definitely true, given
that there are more than three persons and exactly two numbers included in the
pronominal categories of SBZZ. There are no gender categories in then Zoogocho
Zapotec noun, making Universal 43(‘(If) a language has gender categories in the
noun, it has gender categories in the pronoun’ (ibid. 75)) inapplicable. Universal 44.
‘(I)f a language has gender distinctions in the first person, it always has gender
distinctions in the second or third person, or in both’ (ibid. 76), is similarly
inapplicable, as there are no gender distinctions made in the first or second person.
Finally, there are gender differences in the plural of the pronoun and in the singular,
thereby verifying Universal 45 which states that ‘(I)f there are any gender
distinctions in the plural of the pronoun, there are also some gender distinctions in
the singular also’ (ibid. 76).
1 See C hapter Eight for a fuller discussion o f these facts.
2 ye ‘som e’ is a m orphem e w hich either occurs on its own, as in (24) or occurs with a num eral, as in y e to in (21),
It has been m entioned to me (R osem ary Beam de A zcona p.c.) that this m ight be related to the potential m arker
ye-
3 N ote that both the suppletive form and the preposition can cooccur when the noun is further modi tied by an
adjective, as in lizh golh ehi= a’
poss.house old o f= lsg
‘M y old house’.
One could also express this as in the following:
lizh= a’ da golh
poss.house = ls g clinan old
‘M y old house.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
192
4 Speakers do not offer these form s w ithout a possessor. In questions, the question word stands as a preposed
possessor. There are a few w ords such as xsil ‘breakfast’ w hich could potential com e from a form like
x-zil
poss-m om ing
‘breakfast (o f the m orning)’. N ote that this form is alm ost alw ays used with the verb To e a f .
5 This use will be described in slightly greater depth in chapter seven.
6 W hen transcribing and translating the text, A lberta and 1 broke up each text into individual lines based on (he
intonation patterns o f the speaker w hose text we were transcribing. These could be m ade up o f one sentence, one
noun phrase, interjection, etc.
7 W hile it m ay appear that the w ord xna= includes the possessive prefix x- it is not the case. Xna never occurs
on its own, and the vocative form for ‘m other’ is ma. N ote that the word nowe is used as the vocative form for
grandm other, and is potentially related.
8 N ote that for w ords w hich end with a labialized consonant, such as conejw ‘rabbit |S p .|\ the addition o f a
determ iner, causes the form to change to conejo=n ‘the rabbit’, evidencing partially the nature o f labialization.
9 There are constructions where the object can precede the subject, but these are pragm atically m arked and will
be discussed below , for the case o f topic and focus constructions and in the follow ing chapter, for the case o f the
reflexive o f possessor construction.
1 0 Recall from the previous chapter (4.3.3) that an obsolete m orphological is used by som e older speakers,
although it is not currently used by the m ajority o f speakers
1 1 At this point, I can com e up with no reasons for the different forms. This issue will require further research.
121 can see no reason for the difference betw een these tw o form s at the present time. M ore research is needed.
1 3 This w as not listed in G reenberg’s paper, but alw ays referred to as ‘Z apotec’ (H aw kins p.c. 1998).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Chapter Six: Complex Constructions
6.1 Causatives
There are two causative constructions in Zoogocho Zapotec: a morphological
construction which involves affixes or a change to the verb root (described in 4.3.9)
and a syntactic construction which involves the root Vn ‘to make’.
As has already been discussed in 4.3.9, a morphological causative is marked
by a change from lenis to fortis in the verb root, the addition of -os- or .v - or -z- or
another consonantal or vocalic change. The net effect of this is that a formerly
intransitive or transitive verb becomes, respectively, transitive or ditransitive, as seen
in the following examples. This can apply to verbs with experiencer subjects as well
as seen in (3) and (4).
( 1) gu-zizh campan
comp-ring bell
6.1.1 Morphological causative
‘The bell rang.’
(2) b-sizh=a’ campan
comp-caus.ring= 1 sg bell
‘I rang the bell.’
(3) b-le’i=da’ to yag yelha
comp-see= 1 sgexp one tree banana
‘I saw a banana tree.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
19 4
(4) b-lhe’i=da=ne’ to yag yelha
comp-caus.see=lsgexp=3fo one tree banana
‘I showed him a banana tree.’
Inanimate subjects can also be causers, as seen in (5) and (6).
(5) dx-zxiz-edx-gw=a’
cont-shake-much-int=l sg
‘I shiver a lot.’
(6) leka ch-xiz da zague neda
much cont-cause.shake clinan cold lsg
‘The cold makes me shiver a lot.’
6.1.2 Syntactic causative
The syntactic causative is based on a structure like (7), with the verb ben ‘to
make or do’ directly followed by the causer, sometimes followed by the word ga ‘so
that’, followed by the second verb and then followed by the causee and optionally
another object. In the following schema, ga is optional, as is the object. Subj 1 refers
to the causer and subj2 refers to the causee.
(7) cause subjl (ga) verb subj2 (obj)
(8) b-en=a’ ga b-sizh bidao campan
comp-make=lsg so th a t comp-caus.ring child bell
‘ I made the child ring the bell.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
195
(9) b-en=e’ b-lab bidao to libr
comp-make=3f comp-read child one book
‘He made the child read a book.’
There seem to be no restrictions on what verbs can occur in this construction,
not even potential blocking by the morphological causative.
These two constructions differ basically in how direct the causation is. As is
common crosslinguistically (cf. Haiman 1985), the morphological causative
indicates that the causer had a direct role in the main event. Conversely, the syntactic
causative generally indicates that the causer caused someone else to perform the
action or made the situation such that the action could come to pass. Note that the
causee can be implied.
( 10) b-en=a" ga gu-zizh campan
comp-make=lsg s o th a t comp-ring bell
‘ I made it so that the bell rang.’ (I got someone else to do it or set it up such
that the bell rang. I did not directly ring the bell.) (Compare with (2) above.)
(11) b-en=a’ ga b-e-yegh Jose
comp-make=Tsg so that comp-rep-go jose
‘I made Jose leave."
6.1.3 Differences between the two constructions
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
196
6.2 Instrumentals and comitatives
Once again, there are both syntactic and morphological strategies which
increase the valency of a verb, introducing an instrumental noun phrase. Both of
these strategies are productive. The morphological form includes the use of a suffix
d= and has a pure instrumental reading. Example (12) below exemplifies the use of
d=. See 3.4 for a more in-depth discussion of its morphological characteristics.
( 12) benda’ lizha’ martiyw
b-en-d=a’ lizh=a’ martiyw
comp-make-instr= 1 sg poss.house= 1 sg hammer
‘I made my house with a hammer.’
The comitative/instrumental involves the use of lenh ‘with’ as in (13). lenh is also
used to conjoin N P’s as in (14) below.
(13) dx-aw=a’ lenh sede
cont-eat= 1 sg with salt
‘I eat with salt.’
(14) dx-een=da’ yinha’ lenh sede
cont-want=lsgexp chile with salt
‘I want chile with salt’
(15) b-le’i=da’ lia lank lenh lia bert
comp-see=lsgexp donaangela with dona alberta
‘I saw Dona Angela and Dona Alberta’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 9 7
In general, the semantic difference between these two is as seen in the following.
(16) dx-awa-d=a’ naa’
cont-eat-instr= 1 sg hand
‘I am eating with my hand.
(17) dx-awa=a’ lenh naa’
cont-eat=lsg with hand
‘I eat with my hands.’
(18)a) dx-awa=a’ lenh plat
cont-eat = 1 sg with plate
‘I am using a plate to eat.’
b) *dxawada’ plat
c) *dxawada’ to plat (putting the plate in the mouth, not normally done with a
plate)
The difference between the two constructions is that the morphological
instrumental implies a much closer use of an object, either in terms of physical
proximity or in terms of amount of usage.
Furthermore, in other contexts, the instrumental marker -d- implies direct use
as opposed to the comitative meaning of lenh.
(19) dx-aw=a’ lenh bi chi=a’
cont-eat=lsg with child of=lsg
‘ I am eating with my child.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
198
(20) dx-awgo-d=a’ bi chi=a’
cont-eat-instr= 1 sg child of-1 sg
‘I am using my child in order to eat.’ (For example, if one’s child was
working and one was eating the fruits of their labor, one would use this
form.)
6.3 IO ‘lowering’ and benefactive constructions
6.3.1 IO ‘lowering’
As described in Croft (1985), ‘(i)ndirect object ‘lowering’ is the realization of
a recipient or a benefactive argument as the possessor of the direct object NP’ (ibid.
41). It is debatable if indirect object ‘lowering’ constructions in Zoogocho Zapotec
should be considered the same as beneficiary constructions as will be discussed
below. The following are examples of indirect object lowering.
(22) b-enh=a=le’ to libr
comp-give=Tsg=2sgfsf one book
‘I gave you a book.’ (But not necessarily permanently.)
(22) b-enh=a’ to libr chi=o’
comp-give=lsg one book of=2sg
‘I gave a book to you.’ (Permanent transfer.)
Note that in example (22), the transfer is not necessarily permanent, but in
(23), with indirect object lowering, the transfer is permanent. Beneficiaries are also
marked as the possessor of the direct object in SBZZ, but do not have a
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
I'«
corresponding form including the beneficiary as a core argument as seen in the
following section.
6.3.2 The basic expression of beneficiaries
There are two basic ways in which to express a beneficiary in Zoogocho
Zapotec: the use of par ‘for (from Sp. para ‘for’)’ (as in (23)) which gives an
unambiguous beneficiary reading and the use of a possessive construction, as in (24).
It is the use of the possessive construction to express beneficiaries (as in (24)) with
which we will be primarily concerned.
(23) Bchexoa’ to yet par le’.
b-chexo=a’ to yet par le’
comp-toast=lsg one tortilla for 2sg
‘ I toasted a tortilla for you.’
(24) Bchexoa’ chizxghwo’.
b-chexo=a’ chizxghw=o’
comp-toast=lsg poss.tortilla=2sg
‘ I toasted a tortilla/tortillas for you.’ / ‘I toasted your tortilla.’
Trying to differentiate between these two constructions is not easy, a potential clue is
seen in the contrast between (25) and (26).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
200
(25) Bia’ to zhome chio’
b-i=a’ to zhome che=o’
comp-carry=lsg one basket of=2sg
‘ I carried a basket for you.’ (As though I went over to your house and carried
a basket for you, perhaps glossed best as ‘I carried your basket for you.’)
(26) Bia’ to zhome par le’
b-i=a to zhome par le’
comp-carry=lsg one basket for 2sg
‘ I carried a basket for you.’ (As though I carried and brought it for you from
Oaxaca City for you, perhaps ‘I carried a basket for you.’)
Note that example (27) below shows the effects of oversaturation of the
argument structure of the verb. -chexo=, ‘to toast’, is a verb which normally takes
two core arguments, the subject (a semantic agent, the toaster) and the object) a
semantic patient, the object which is being toasted). Through the use of either of the
prepositions par or che= or through an inherently or inalienably possessed NP, an
additional NP, the beneficiary, can be added to the clause. However, it cannot be
included as a pronominal clitic on the verb, as the default interpretation will be that
the object clitic will be interpreted as the patient and not as the beneficiary. Thus,
sentences like (27) are judged to be unacceptable (on both semantic and syntactic
grounds).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(27) *Bchexoale’ chizxghwo’.
b-chexo=a’=le’ chizxghw=o’
comp-toast= 1 sg=2sgo poss.tortilla=2sg
(Too many arguments, would mean ‘ I toasted you.’ without the object NP.)
However, as we will see in the next section, there is a potential for ambiguity
with verbs which can take three core arguments.
6.3.3 Beneficiaries or recipients?
For some verbs such as -zogh ‘to write’, there is a potential ambiguity as to
whether the non-agentive animate argument is a benefactive or a recipient. Consider
the following three examples. In example (28), the second person object pronoun is
clearly a recipient. There is no way of interpreting it as a beneficiary.
(28) Bzoghale’ to kart.
b-zogh=a’=le’ to kart.
comp-write=lsg=2sgo one letter
‘I wrote you a letter.’
Conversely, in (29), the second person is interpreted as a beneficiary and cannot be a
recipient.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
202
(29) Bzogha’ to kart chio’.
b-zogh=a’ to kart che=o’
comp-write=lsg one letter of=2sg
‘I wrote a letter for you.’ (Literally, ‘I wrote your letter.’ As though you can't
write, and I’m doing you a favor. Can be followed with another clause like .
‘which you sent to the municipal authorities’.)
However, in (30) the second person, which is marked both as an argument on
the verb and as the possessor of the letter must be interpreted as a recipient. One
could not follow it with another clause like ‘which you sent to the municipal
authorities.’
(30) Bzoghale’ to kart chio’
b-zogh=a=le’ to kart che=o’
comp-write=lsg=2sgo one letter of=2sg
‘I wrote you a letter.’
In the following section, a further explanation of the potential difficulties in
differentiation between recipients and benefactives will be developed.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6.3.4 What is whose?
The following sentence is ambiguous.
(31) Gde Zxoan to tas kafe chia’
g-de Zxoan to tas kafe che-a’
comp-pour John one cup coffee of-lsg
‘ John poured a cup of coffee for me.’
The coffee could potentially be mine and John just poured it for me, or, in
what is considered a more natural reading, it is John’s and he is giving me a cup of
coffee. In order to specify that it is indeed my coffee first, one must use a definite
construction as in the following.
(32) Gde Zxoan to tas kafe chian’
g-de Zxoan to tas kafe che=a’=n’
comp-pour John one cup coffee of-1 sg~det
‘Juan poured me a cup of my coffee.’/’John poured a cup of my coffee’ (He
could have then potentially given it to someone else.)
This may somehow be analogous to the difference between the following
English examples.
(33) John poured me a cup of coffee.
(34) John poured a cup of my coffee.
The use of the definite marker can make for some interesting constructions
like (35) which clearly show the benefactive use of the che-- construction.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(35) Gde Zxoan to tas kafe chian’ chio’
g-de zxoan to tas kafe che=a’=n’ che=o'
comp-pour John one cup coffee of=lsg=det of=2sg
‘John poured a cup of my coffee for you.’
One could also have an example like (36), in which the first person is made
definite as above, this is however viewed as being a bit odd by native speakers .
(36) Gde Zxoan to tas kafe chian’ chia’
g-de zxoan to tas kafe che=a’=n’ che^a’
comp-pour John one cup coffee of=lsg=det of=lsg
‘John poured a cup of my coffee for me.’
Of course, (37) would be a more natural way of expressing this.
(37) Gde Zxoan to tas kafe chian’ par neda
g-de zxoan to tas kafe
comp-pour John one cup coffee
‘John poured a cup of my coffee for me.’
Note that (35) can also be expressed as in (38).
(38) Gde Zxoan to tas kafe chian’ par le’
g-de zxoan to tas kafe
comp-pour John one cup coffee
‘John poured a cup of my coffee for you.’
che=a’=n’ par neda
of=lsg=det for lsg
che=a,:=n’ par le’
of=l sg=det for I sg
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 0 5
Also, note that the use of the determiner to differentiate between potential possessors
is seen outside of beneficiary constructions as in the following examples.
(39) Gzxia’ to libr chio’
g-zxi=a’ to libr che^o’
comp-buy-lsg one book of-2sg
‘I bought a book for you.’/’I bought one of your books.’
Note that both of the potential readings of (39) could be interpreted as being
beneficial to the person whose book was bought. (In one case, the person will be the
recipient and in the other case the person will profit from it.) In order to specify that
it was not bought with the intention of giving it to the person one uses a construction
like (40). The use of the definite marker here, as above, specifies that the book was
the second person’s possession before the first person purchased it.
(40) Gzxia’ to libr chion’
g-zxi=a to libr che=o’=n’
comp-buy one book of=2sg=det
‘ I bought one of your books.’
One can also use the Spanish loan word par ‘for’, in order to specify that the
book was bought and intended to be given to someone as in (41).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
206
(41) Gzxia’ to libr par le’
g-zxi=a to libr par le’
comp-buy one book for 2sg
‘ I bought a book for you.’
6.4 The reflexive-of-possessor construction
6.4.1 Introduction
In Zoogocho Zapotec a construction, first described by Inez Butler for
Yatzachi Zapotec (Butler 1976) and since described by Cheryl Black (Black 1995,
1996), exists which appears to violate binding theory. I will follow Butler in
referring to this construction as the ‘reflexive of possessive’. In this construction, as
seen below in (42), a subject may be omitted if it is coreferential with the possessor
of the object.
(42) b-en lizh=a’
comp-make poss.house=lsg
‘I built my house.’
Reflexives and reciprocals are formed using the terms kwin+ ‘self and
lghezh+ ‘each other’ as possessed nouns in this construction .
(43) sh-naa kwin=a’
cont-wash self=lsg
‘I washed myself.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
207
(44) sh-naa lghezh=dxo
cont-wash fellowof=lplincl
‘We washed each other.’
The reason why this construction is puzzling to Zapotecanists is that
Zoogocho Zapotec (and other languages which take part in this construction1 ), as
with most if not all Zapotecan languages, is a staunchly VSO verb initial language,
as seen in the previous chapter, and does not allow apparent VOS constructions in
any other environment. As a result of this, it may be necessary to posit a null element
in the subject position as in (45) or (46) below, this null element would, however, be
restricted to this construction. Outside of this construction, Zoogocho Zapotec is not
a language which readily allows for null structural arguments.
(45) b-en=pro lizh=be’
comp-make=nullpro house.poss=3f
‘She made her house.’
(46) b-en=pro lizh Lalo
comp-make=nullpro poss.house Lalo
‘Lalo made his house.’
Furthermore, in terms of both linear precedence and the structure of the
clause, the position of the anaphor in a superior position to its antecedent would
appear to violate binding condition A. I will examine the binding conditions in more
depth in the following section.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 0 8
6.4.2 Binding conditions
In this section, I will largely follow the exposition of Black (1996). The
familiar binding conditions are restated below (Haegeman 1991: 216).
A. An anaphor(e.g. reflexive or reciprocal element) must be bound in its
governing category.
B. A pronoun must be free in its governing category
C. r-expressions(names) are free.
These conditions were largely devised in order to account for the following
core types of data (based on Black (1996), example (11) pg. 77).
(47) John, slapped himself.
(48) * Himself, slapped John,.
(49) John/ slapped him 7 /*,.
(50) John, slapped John,/*,
(51) He,/*, slapped John/.
Principle A would account for the ungrammaticality of (48) given the unbound
anaphors, B for the ungrammaticality of the coreferential version of (49) given the
bound pronoun, and C for the lack of potential coreferential readings in (50) and (5 1)
given the impossibility of the binding of r-expressions.
Before looking at the actual Zoogocho Zapotec examples, it might be useful
to investigate what the core examples used to corroborate the Binding T h e o ry in
English would be like if reordered to mimic the strict VSO syntax which the majorit v
of Zoogocho Zapotec clauses conform to.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 0 9
(52) Slapped John, himself,.
(53) * Slapped himself, John,
(54) Slapped John, him/*/.
(55) Slapped John, John/*,
(56) Slapped he/*, John,.
With the exception of (59), Zoogocho Zapotec behaves exactly as we would expect it
to with respect to these judgements.
(57) gud-ap Ron, kwin=be,’
comp-slap Aaron self=3sgf
‘Aaron, slapped himself,.’(Emphatic reading)
(58) * gud-ap kwin+be, R on/
comp-slap self+3sgf Aaron
(59) gud-ap kwin Ron
comp-slap self Aaron.
‘Aaron slapped himself.’
(60) gud-ap Ron, lhebe/*,
comp-slap Aaron him
‘Aaron/ slapped him/*, ’
(61) gud-ap Ron, Ron/*,
comp-slap Aaron/ Aaron/*/
‘Aaron, slapped Aaron/*,.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
210
(62) gud-ap lhebe, Ron,,*,
comp-slap him/ Ron//*/
‘He slapped Aaron.’
As we can see, except for (59) above, ZZ neatly conforms to the predictions
which one might make for a strict VSO language. We might want to consider (59) to
have a structure as in (63) below.
(63) gud-ap 0/ kuin Ron/
comp-slap 0, self Aaron,
‘Aaron slapped himself.’
However, I do not want to analyze this construction as being one which
involves pro-drop, as it would be very anomalous with the rest of the grammar of
Zoogocho Zapotec. I will not go into GB internal theoretical depth here. However, it
seems that one could appeal to reconstruction effects, movement, or even backwards
binding as proposed by Black (1996) to explain this data within the GB theoretical
framework. Even though I have a great deal of data on this construction, the
discussion of these data is outside of the scope of the current chapter. 1 will return
briefly to this construction in Chapter Eight.
6.5 Constructions with multiple verbs
6.5.1 Introduction
In the following sections, I will discuss various constructions which all
involve multiple verbs which show varying degrees of integration. The following
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
cline (adapted from Thomas Payne (1997)), shows the range of possibilities
human languages.
High degree of grammatical integration
One clause
Serial verbs
Complement clauses
Adverbial clauses
Clause chains
Relative clauses
Coordination
Two separate clauses
No grammatical integration (ibid. 307)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2.12
In terms of what will be shown in the following sections, the following cline
is more appropriate for Zoogocho Zapotec.
One clause
Infinitival constructions
Auxiliary constructions
Complement clauses
Argument sharing
Relative clauses
Coordination
Separate clauses
6.5.2 Infinitival constructions
The following construction is one in which a verb with an aspect marker and
a pronominal enclitic is followed by an infinitive. An infinitive is a verb with special
aspectual marking, as discussed in 4.3.8. An infinitival verb form is not required to
have an overt subject in subject position. The subject of both verbs is the same, and
only one of the verbs can be negated. While I am not aware of any particular
restrictions on which verbs can be used, the verb zegh ‘to go’ seems to show up quite
often in this construction. Examples of this construction follow.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 11
(64) zegh=be’ g-os zaha
comp.go=3inf inf-plant bean
‘S/he went to plant beans.’ (But might not have made it there, this would be
an answer to a question such as ‘Where did s/he go?’)
Note that only the main verb can be negated, as in the following.
(65) bi zegh=be’ g-os zaha
neg comp.go=3inf inf-plant bean
‘She didn’t go to plant beans.’
Examples like the following are viewed as being ungrammatical.
(66) *zegh=be’ bi g-os zaha
comp.go=3inf neg inf-plant bean
‘She didn’t go to plant beans.’
(67) zhaa-dxo no w-e-shib kafe
stat.say=lplincl indef inf-freq-cut coffee
‘We’re saying we cut/harvested coffee.’
6.5.3 Verbs borrowed from Spanish
There are also the following examples which show how Spanish loan-verbs
are incorporated into SBZZ.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(68) dx-on=e’ pensar
cont-make=3f pensar
‘He thinks.’
(69) dx-on=e’ to pensar
cont-make=3f one pensar
‘She thinks. She has a thought.’
(70) *dx-on=e’ chupe pensar
cont-do=3f two pensar
‘He thinks. He has two thoughts.’
So far, based on the possible quantification of the Spanish verb, it looks lik
the Spanish verb has possibly been borrowed as a noun as opposed to being
borrowed as a verb. See the following similar examples.
(71) dx-on=e’ to shinh
cont-do=3f one work
‘He has a job.’
(72) dx-on=e’ chupe shinh
cont-do=3f two work
‘She has two jobs.’
(73) dx-on=e’ shinh
cont-do=3f work
‘He works.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6.5.4 Auxiliary constructions
I will call constructions that consist of two verbs with aspectual prefixes, the
first of which has no subject and the second of which has a subject, auxiliary
constructions. The verbs which make up the first element of this construction are a
restricted set of verbs some of which might be considered to be modals in languages
such as English or Dutch, having meanings such as ‘can’, and ‘should' and others
which are not such as ‘begin’, or ‘finish’. These verbs can take a full range of
primary aspect markers, but typically do not have the requirement that verbs have a
subject, even when used by themselves, as seen in (74) below. They can potentially
co-occur with subjects as seen in (75) below.
(74) na kate gu-zoalao
and when comp-begin
‘and when it began’
(75) gu-zoalao=a’
comp-begin=lsg
‘I started
Normally, auxiliary verbs are used with other verbs with more specific
meanings. In this case they do not usually cooccur with a subject noun phrase, when
they do it is always of the same person as the main verb. However, they do not
necessarily agree with the main verb in terms of aspect.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(76) na’ y-eyozh g-onh=dxo da~na’
and pot-finish pot-make=lplincl clinan=demdist
‘Now that we have finished making this...’
(77) na’ kate gu-zoalao dx-a=a’
and when comp-begin cont-go=lsg
‘And when I began to go.’
(78) per yogo benhe benh dx-ake dx-ak dx-on=o’ rmed
but all people genan cont-be sick cont-be cont-make^2sg remedy
‘But you can cure everyone who is sick?’
(79) dx-eyala si-i=e’ shon gayoa
cont-should pot.grab=3f three hundred
‘She should grab three hundred.’
Normally, only one of the verbs, the auxiliary, is negated. However, as seen below in
(81) the main verb can also be the one which is negated. This does have slightly
different semantics, as seen in the gloss.
(80) bi dx-ak gu-ta=a=n
neg cont-be pot-play=Tsg=3inan
‘I cannot play it (a musical instrument).’
(81) dx-ak bi gu-ta=a=n
cont-be neg pot-play=lsg=3inan
‘I alone cannot play it (a musical instrument).’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Both auxiliary and main verb can be individually negated at the same time, as in
(82). This seems to give an emphatic negative.
(82) bi dx-ak bi gu-ta=a=n
neg cont-be neg pot-play=lsg=3inan
‘There is really no way can I play it (a musical instrument).’
In the textual study I completed for Chapter Eight, there were some examples of
auxiliaries appearing with apparently no subject whatsoever, as in the following,
already seen above in (74).
(83) ka bi g-ak
demadv neg pot-be
‘It’s not going to be possible.’
This is a most unusual construction, and one which deserves further research.
6.5.5 Complementation
In this section, I will discuss complementation. Complement clauses in
Zoogocho Zapotec are fully inflected for tense and have the normal argument
structure requirements. The majority of verbs which have complement clauses as
arguments can also have full N P’s satisfying the verb’s argument structure
requirements. Compare the following.
(84) dx-een=da’ to dulc
cont-want=lsgexp one candy
‘I want a candy.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
218
(85) dx-eene=be’ y-egh=be’ kuan wizh=en
na’
cont-want=3inf pot-drink=3inf herb ‘thepill’^det
demdist
‘She wanted to take the contraceptive herb there.’
(86) dx-eene=be’ zegh=be’
cont-want=3inf cont.go-3inf
‘He wants to go.’
The majority of complement clauses have the same subject as the main clause
as seen above. There are cases, however, in which the subject of the complement
clause differs from that of the main clause. Many such clauses show little difference
from normal, verb-initial clauses as in the following.
(87) dx-bez=a’ g-aow bidao yet
cont-hope= 1 sg pot-eat child tortilla
‘I hope the kid eats the tortilla.’
(88) dx-bez=a’ bi g-ak yegho
cont-hope=lsg neg pot-be rain
‘I hope it doesn’t rain.’
Notice the topicalized example seen below.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(89) Maria n-dxeen=da’ go-yeb-e’ zaha
Maria stat-want=lsgexp pot-cook=3f beans
‘I want Maria to cook beans.’
The subject of the complement clause can also be raised, although a
resumptive pronoun remains in the complement clause. This raising is optional for
most verbs, as seen below.
(90) dx-eene pan; g-ab lia lankj to yel
cont-want Panfila pot-weave Dona Anglea one shawl
chi=e’j
of=3f
‘Panfila wants Dona Angela to weave her a shawl’
(91) dx-eene pan; lia lankj g-ab=e’j
cont-want Panfila Dona Anglea pot-weave=3f
to yel chi=e’j
one shawl of=3f
‘Panfila wants Dona Angela to weave her a shawl.’
The meaning difference between these two options is the following, with the
raised subject and resumptive pronoun giving a more emphatic reading.
(92) n-ez=da’ shegh=o’ gxe
stat-know=lsgexp stat.go=2sg tomorrow
‘I know that you are going to leave tomorrow ‘
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
220
(93) n-ez=da’ lee shegh=o’ gxe
stat-know+lsgexp 2sg stat.go=2sg tomorrow
‘I know that you are going to leave tomorrow ‘
Raising can even be ruled out. Examples like the following are ungrammatical.
(94) dx-ene=be’ zegh benhe bio
cont-want=3inf comp.go person masc.
‘He wants the man to go.’
(95) *dx-ene=be’ benhe bio zegh(=be’)
cont-want=3inf person masc comp.go(=3 inf)
‘He wants the man to go.’
With ditransitive clauses, this raising is obligatory, as seen below. Example (96)
shows a simple ditransitive clause, and examples (97)-(100) show that raising is
obligatory for complement clauses in these constructions.
(96) zeghnii Pan Lia Lank to yegh
stat.show Panfila Dona Angela one flower
‘Panfila showed Dona Angela a flower.’
(97) zeghnii Pan Lia Lank g-ab=e’ to yel chi^e'
stat.show Panfila Dona Angela pot-weave=3f one shawl of=3f
‘Panfilaj showed Angelaj how shej was going to weave her, shawl.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(98) *zeghnii Pan g-ab Lia Lank to yel
stat.show Panfila pot-weave Dona Angela one shawl
chi=e’
of=3f
‘Panfilaj showed Angelaj how shej was going to weave herj shawl.’
(99) b-lhe’i lia lank chach gu-lhab-'-be’
comp-caus.see Dona Anglea Chacho comp-read 3inI'
‘Dona Angela taught Chacho to read’
(100) * b-lhe’i lia lank gulhab chacho
6.5.6 Argument Sharing
In this section, I will discuss what I am tentatively labelling argument
sharing. In these constructions, which differ from relative clauses in not having a
relative pronoun, an argument (either a subject or an object) is shared by two clauses.
These are sometimes translated as relative clauses and sometimes as conjoined
clauses.
(101) bizx ka dx-ghe-za-kse benhe zghe-noa
which demadv cont-pl-walk-emph person stat.pl-take
bidao’
child
‘How is it that people go around and take children?
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Note that (101) is odd, largely because the second verb zghenoa does not have a
subject. This looks on some level like it is an internally headed relative clause, but
the preverbal positioning of benhe would be extremely odd. Consider, as more
evidence, the following example.
( 102) le bate’ gu-za=be’ nox=e’
foe when pot-walk=3inf pot.grab=3f
‘It’s when she walks that they grab her.’
Once again, it appears that the subject of the first verb is fulfilling an
argument role in the lower clause. In the following example, the translation which
was given was unequivocally a relative clause the first time I elicited it. The second
time I elicited it, I got the second reading.
(103) zxoa’ dx-on=o’ yet=en b-dxogh wen.
com cont-do=2sg tortilla=det comp-tumout good
‘It’s corn you make the tortillas that turn out good.’/ ‘Its corn you make
tortillas with and they turn out well.’
6.5.7 Relative clauses
In this section, we will examine relative clause formation and see how SBZZ
should potentially be classified according to the criteria set forth in Keenan and
Comrie (1977). The first two examples are relative clauses where the head noun is
the subject of the relative clause. Relative clauses in Zoogocho Zapotec are
postnominal. The generic noun berth2 introduces the relative clause, and appears to
serve the function of satisfying the argument structure requirements of the verb
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
dxban. benh and the other generic nouns play an analogous role to relative pronouns
in spoken Spanish. The form of the generic noun is invariant with respect to the case
of the role it is filling.
(104) b-le’i=da’ benhe bio benh dx-ban
comp-see=Tsgexp person masc genan cont-steal
‘I saw the man who was stealing.’
(105) b-le’i=da’ benhe bio benh dx-bezh
comp-see=T sgexp person masc. genan cont-cry
‘I saw the man who was crying.’
In the following examples, the direct object is the position relativized on.
(106) gage n-ak=en yet go danh dx-es-on-e'
neg stat-be=3inan tortilla root geninan cont-pl-make<3f
naa
now
‘They are not like the tamales that they make now.'
(107) na’ b-edey-aa-te=e’ dao clri- a" danh
and comp-ven-take-emph=3f com tassels of-1 sg geninan
b-edey-e=e’
comp-ven-take=3 f
‘And they came to take the corn tassels of mine that they had come to take.’
In the following, an indirect object is relativized on.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 2 4
(108) benhe bio benh b-i=o’ libr sh-tas-e’
person masc. genan comp-give=2sg book cont-sleep-3f
‘The man to whom you gave the book is sleeping.’
For instrumentals, we see examples such as the following:
(109) machet danh dx-ogo=d=a’ yelhe ni
machete geninan cont-cut=instr=lsg m aizefield demprox
de=ks=en
stat.lie=rep=3inan
‘The machete with which you cut the maize field is here.’
In the following, we see an example of a relative clause in which a possessor
is relativized. The possessor no longer appears in the relative clause.
( 110) b-le’i=da’ bidao benh dx-zxite x-kabayo lho
comp-see=lsgexp child genan cont-jump poss-horse belly
tronkw
trunk
‘I saw the child whose horse jumped over the tree trunk.’
Finally, we will see how Zapotec relativizes on ‘objects of comparison’3. The
following is an example of a regular comparative sentence.
(111) Maria n-ak-dx=e’ benhe zxen kazxoan.
Maria stat-be-more=3f person large than Juan
‘Maria is larger than Juan.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 2 5
The following example is a relative clause formed on an object of
comparison.
( 112) b-le’i=da’ benhe bio benh n-ak zxen ka
Maria
comp-see=lsgexp person masc genan stat-be big demadv
Maria
‘I saw the man who is bigger than Maria.’
I will now briefly see how these constructions correspond to the statements
made about relative clauses in Edward Keenan and Bernard Comrie’s 1977 article
‘Noun Phrase Accessibility and Universal Grammar’. There are two issues which arc
important to the classification set forth in that paper: 1) whether a particular
grammatical relation is relativizable and 2) whether the strategy used gives some
indication as to the grammatical role in the relative clause of the NP being relativized
on. As has been seen above, all of the potential positions which could possibly be
relativized on, are relativized.
Zoogocho Zapotec does not normally retain pronouns in relative clauses,
using a gap strategy instead in order to indicate the grammatical role of the
relativized NP. However, SBZZ is furthermore a VSO language. As a result of these
two factors a potential ambiguity arises in sentences such as (113).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
226
(113) b-le’i=da’ benhe bio benh gud-ap Maria
comp-see=lsgexp person masc. genan comp-slap Maria
‘I saw the man who slapped Maria.’ ‘I saw the man who Maria slapped.'
This ambiguity comes from the fact that the sentence could be interpreted as
having either of the following structures (where a _ indicates the position of the gap).
(114)bleida’ benhe bio benh gudap_ Maria ‘I saw the man who slapped Maria.’
(115) bleida’ benhe bio ben gudap Maria _ ‘I saw the man who Maria slapped.’
In order to avoid this type of ambiguity, one can retain a resumptive pronoun4
in ZZ.
(116) b-le’i=da’ benhe bio benh gud-ap=e’ Maria.
comp-see=lsgexp person masc. genan comp-slap=3f Maria.
‘I saw the man who slapped Maria.’
(117) b-le’i=da’ benhe bio benh gud-ap Maria lhee
comp-see=lsgexp person masc. genan comp-slap Maria 3f
‘I saw the man who Maria slapped.’
In (116), the bound pronoun must be the subject because of its immediate
post verbal position and therefore ‘Maria’ must be the object, and similarly in (! 17),
‘Maria’ must be the subject given the immediate post-verbal positioning and the
pronoun must be the object.
There are also headless relatives such as the following.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
221
(118) dx-on=a’ danh g-ao bi chi=a’
cont-make=lsg geninan pot-eat child of=lsg
‘I make what my child eats.’
6.5.8 Coordination
In this section, I will briefly investigate the interpretation of pronouns in
constructions which are formally made up of two clauses, conjoined both with the
conjunction na ’ as in (119), and with two juxtaposed clauses as in (120), which I will
refer to as the ‘conjunction construction’5 and the ‘ juxtaposition construction’
respectively.
(119) b-et zxoan bedo na’ gu-chu-e’
comp-hit Juan Pedro and comp-cough -31 '
‘Juan, hit Pedroj and hej coughed.’
( 120) b-et zxoan bedo gu-chu=e’
comp-hit Juan pedro comp-cough=3f
‘Juanj hit Pedroj and hej coughed.’
Although these two constructions are formally almost identical, the pronouns
in the second clause of both constructions can have very different interpretations. In
many but not all of the instances of constructions similar to example (119), the
intended antecedent of the second clause is the patient of the preceding clause.
The interpretation of pronouns in the second clause of constructions such as
those exemplified in example ( 120) varies between the patient and the agent of the
first clause. This construction is sometimes glossed with a participial construction in
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 2 8
Spanish, thus making it falsely reminiscent of a control construction. In the
remainder o f this section, I will concentrate on the properties of each of these
constructions separately.
As a first step to clarifying the constructions with na it may be helpful to
first describe a few of the differences and similarities between this construction and
corresponding constructions in English. Conjunction reduction is not allowed in
Zoogocho Zapotec. Thus, one of the primary means of clarifying structures
corresponding to the ZZ example (119) above is not allowed. For example, although
the following sentence might have as its preferred reading (in my dialect) that John is
the one who is coughing, it is still potentially ambiguous, being more likely
determined by real world knowledge. (For example if we knew that Paul had a
cough, we would be more likely to interpret Paul as being the one who was
coughing.)
(121) John, hit Paulj and hej/? j coughed.
However, the following example with the use of conjunction reduction is
unambiguous.
(122) Johnj hit Paulj and coughedj.
John must be the one who coughed. The interpretation of the subject in conjoined
sentences with conjunction reduction varies depending upon the language. Thus, in a
syntactically ergative language like Dyirbal, the interpretation of the omitted subject
would be Paul.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
229
One of the potential explanations for the interpretation of the pronoun in
(119) which jumped to my mind at first was that the language was syntactically
ergative, even though it shows no morphological traces of ergativity. ZZ is
surrounded by morphologically but generally not syntactically ergative languages
(such as Mixe6) and one of the typological characteristics of the Mesoamerican area
(as posited by Campbell et al (1985)) is, of course, ergativity. Traces o f syntactic
ergativity have also been found in other Otomanguean languages such as Chinantec
(Foris 1994). Unfortunately, examples such as (123) came up as quickly as I started
to investigate the phenomenon more carefully.
(123) gw-eene Maria Zxoan na’
comp-want Maria Juan and
‘Maria wanted John and she screamed.’
It began to become clear that perhaps the interpretation of (119) above had
more to do with real world knowledge than with grammatical patterning, although
every speaker with whom I checked the example gave the same interpretation.
People are more likely to cough if they have been hit than if they have not, perhaps,
and this is what probably led to this particular interpretation.
The juxtaposition construction is perhaps a bit more understandable, but still
provides some difficulties in its analysis. I would have liked to be able to
differentiate between different types of these constructions as in the prototypical
control construction in a language like English, having a verb of cognition or
b-osi=e’
comp-scream^Jf
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 3 0
perception and those constructions which have an agentive verb (such as ‘to hit’), but
the data do not seem to lead in any general direction with respect to this issue.
Although there are a number of cases where the reference changes, it is by no
means regularly determined by the type of verb, as seen in the following examples.
In the juxtaposition constructions, there tends to be a great deal of variation on the
interpretation of the pronoun in the second clause, which is co-dependent upon,
among other things, the aspectual markings on both verbs. See the following
examples.
(124) gw-eene maria zxoan dx-osi=e\
comp-want maria juan cont-scream=3f
‘Maria wants Juan screaming.’(Juan is screaming. (Maria might think it to
be cute.))
(125) dx-eene maria zxoan b-osi=e’
cont-want maria juan comp-scream=3f
‘Juan screamed before Maria wanted him.’
(126) y-eene maria zxoan b-osi=e’
pot-want Maria Juan comp-scream=3f
‘Maria is going to scream after wanting juan.’
(127) y-eene maria zxoan dx-osi=e’
pot-want Maria Juan cont-scream=3f
‘Juan is going to scream because M wants him.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The aspectual markers of both the first and the second verb and the
interpretation of the enclitic pronoun of the lower verb are definitely co-dependent,
as seen in the above examples. It is rather difficult at this point to determine exactly
how they are dependent however, as discourse knowledge and other factors clearly
play some role as well. This is an area which deserves a great deal of investigation at
some point in the future.
6.5.9 Adverbial clauses and indirect questions
Adverbial clauses and indirect questions are formed in much the same way.
Adverbial clauses are introduced by a complementizer such as kate ’ ‘when’ as in the
following:
(128) g-on tibo ka kate’ bi na
pot-make Primitivo demadv when neg demdist
y-esey-ela lhegakbe’
pot-pl-arrive 3plinf
‘Primitivo will do it like that when they get over there.’
Indirect questions are introduced by an interrogative form such as no ‘who’ as in
(129):
(129) per bi n-ez=da’ no b-en
but[Sp.] neg stat-know=lsgexp who comp-make
ka’
demadv
‘But I don’t know who did it.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6.6 Comparatives
In the following two examples we see two examples of the comparative
where it is a quality which is being compared. The basic construction appears to be
with the first noun either before or after the copula:
(130) (N) COPULA (N) ADJ (kle)ka’ NP
QUAL MARKER STANDARD
The following examples exemplify this.
(131) Maria n-ak-dx=e’ benhe zxen ka Juana
Maria stat-make-more=3f person large than Juan
‘Maria is larger than Juan.’
(132) n-ak-dx bdxe’ be lis
stat-be-more ant clan small
‘The ant is smaller than the dog.’
The adjective can be fronted as in (133).
(133) tonhe-dxgwa n-ak=e’ jef=en’
tall-emph stat-be= 3f boss=det
‘The boss is taller than the other person.’
Although the adverbial clitic -dx= is often found in this construction it is not
absolutely necessary as seen in (134).
kleka’ be’ko’
than dog
kleka benhe veto
than person other
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(134) n-ak nise ye yanha’ kleka’ kafe
stat-be water com_porridge thick than coffee
‘Com porridge is thicker than coffee.’
One can make statements of equality such as the following, with the word for same
being optional:
(135) NP COPULA (same) kanak NP
(136) yixe’ che=be n-ak=en (lebze) kanak chi=a'
field of-3 inf stat-be=3inan (same) as of-Tsg
‘His field is the same (size) as mine.’
One can also compare actions, in which case the structure is as follows with the
subject optionally appearing before or after the verb:
(137) (NP) Y (NP) ADV kleka’ NP
QUAL MARKER STANDARD
(138) avion=na’ dx-ghe-dx=en bedaones kleka’
airplane=det cont-go-more=3inan quick than
camion=na
bus=def
‘The plane goes faster than the bus.’
(139) be’ko’ zaa-dx=ba bedaones kleka’ bdxe’
dog stat.walk-more=3an faster than ant
‘Dogs walk faster than ants.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6.7 Conditionals
Conditionals are normally in the order if-then, as seen in the following
example.
(140) shi dx-eene’, dx-o-dxo avante
if cont-want, cont-get=lplincl ahead
‘If we want to, we can get ahead.’
They can however also follow.
(141) bi dx-i=o’ shi bi g-ak b-ei^dxo
neg cont-say=2sg if neg pot-make comp-sense^lplincl
shoo
pot.enter
benhe lho yoo dx-ap=a=ne
person in house cont-say=lsg=3f
“‘Don’t say anything if we notice people going into the house”, I said to
him.’
1 See A velino et al. (2004) for a description o f this and sim ilar constructions in Zapotecan languages.
2 See the lexical classes chapter for an explanation o f why benh is considered to be a generic noun.
3 See the follow ing section on com paratives for a larger discussion o f comparatives.
4 C ontent question could also potentially have this am biguity and have the am biguity resolved in a sim ilar fash
as seen in 5.6.4.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 3 5
5 N ote that there are a num ber o f other m eans o f conjoining two clauses, lor instance, using words such as ie
‘because’, or p e r ‘bu t’(these two are coordinating conjunctions, in spite o f the English gloss). 1 chose the Ivvo
discussed in this section because they are both glossed in the m ajority o f cases as being sim ple conjoined
sentences in Spanish. W hile one can, o f course, force m ore exact definitions, these seem to be the basic
interpretation o f these two. N ote that in personal com m unication with Rebecca Long. I m entioned these
constructions in order to get her opinion and she told me that there was definitely som ething going on with
respect to change o f reference, but that she did not feel as though she had the training to classify them. (This is all
basically to confirm my suspicion that these tw o should be grouped together and to indicate that 1 am not the first
to be confounded by these constructions.)
6 Though note that Olutec, a M ixe-Zoque language described by Roberto Zavala (2000), show s ergalivity in its
w ord order with VS and AVO ordering (Zavala p.c.).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 3 6
Chapter Seven: Lexical Classes
7.1 Introduction
Before proceeding to the actual classification of the parts of speech or lexical
classes in Zoogocho Zapotec, I will first discuss the general means by which I will
attempt to rigorously and consistently classify the categories and category
boundaries of SBZZ.
Munro (2002) lists the following criteria:
The number of separate parts of speech one recognizes (whether or not we count
particles!) depends on a number of factors. My analysis in (1), for any target
word, is based on consideration of the criteria listed in (3):
(3) Some criteria for determining part of speech
syntactic factors: what other types of words a target word can combine with,
in what types of phrases;
morphological factors: what affixes the target word is used with;
semantic factors: what the target word and the phrases it is used in means.
(Munro 2002: 5)
In another discussion of how a linguist might classify a lexical item,
specifically the English word, walking, Thomas Payne mentions the three following
possibilities:
1. Taking some nominal properties as criteria. For instance, we could simply
define noun for English as a form that can refer to its only argument with a
genitive pronoun. In this case walking is a noun. However, if we decided that
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ability to pluralize or take a wide range of descriptive modifiers were the
criteria for nounhood, then walking would not be a noun. We would just have
to make a somewhat arbitrary decision and stick to it consistently.
2. Making up a different grammatical category for each complex of nominal
features instantiated by some form or forms in the language. In this case only
those lexical items that have all nominal properties would be considered
nouns. Forms such as walking would be considered something else, such as
present participles...
.. .3. Acknowledging that the difference between nouns and verbs is a
continuum, and that verbs with the -ing suffix fall somewhere in between the
two extremes. (Payne 1997: 35-36)
Payne then goes on to reject any single one of these three potential means of
determining whether walking should be classified as a noun or a verb, taking the
position that a combination of 2 and 3 is preferable from the perspective of the
descriptive linguist. Solution 1 is unacceptable because ‘(C)riterial definitions are
inherently questionable for supposedly universal categories because there is always
the possibility that some language might not display a particular criterion’ (ibid. 36).
He goes on to state that often these criterial definitions are biased towards the better
known languages. The other problem is that, while ideally one would be able to find
necessary and sufficient conditions for the placement of a particular lexical item or
class in a particular category, this is a difficult prospect.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
23 X
‘Most descriptive linguists use solution 2. However, unless great care is taken
in the definition, this can lead to confusion among readers’ (ibid. 37). It can
potentially lead to the overproliferation of lexical classes and to a lack of
generalizations which otherwise should be made. As will be seen later, the terms
‘preposition’ and ‘relational noun’ are often bandied about in descriptions of
Zapotecan languages without any real definition1 . It can also lead to over
specification. While one might want to differentiate verb classes in Zapotec based on
the form of aspectual prefixes which each verb takes, this differentiation is, in all
likelihood, the result of historical residues, not having any synchronic semantic or
formal basis, and therefore, from the point of view of lexical classes, irrelevant .
Solution 3 is, as Payne points out, what ‘reflects most accurately the nature of
linguistic categorization’ (ibid. 37). However, it is incredibly difficult, even for
native speakers of a given language, to come up with rigorous clines which ideally
would include, with justification, all of the many thousand words of an individual
language and doing so would be of little use to the general reader.2
Arbitrary, language-specific categories, based on the type of categorization
which Munro mentions above, are therefore necessary. It is also necessary to
acknowledge that forms which straddle these arbitrary categories exist and provide a
description of them. Such a multi-factorial approach, as explicitly taken by Comrie
(1989) in his discussion of definitions and categories and adjectival and substantival
properties of Russian numerals, which can be used to ‘establish criteria that correlate
with the focal values’ (ibid. 107) of clearly definable categories, will also enable the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 3 9
description of the ‘continuum separating those prototypes from one another, much as
with colour terms, even though here we are clearly dealing with grammatical
categories’ (ibid. 109). The grammar will therefore be both static and self-contained,
yet also self-referencing and ‘emergent’ in the sense of Hopper (1987). I will attempt
to both discuss and acknowledge what parts of speech or lexical categories are
already grammaticalized and what the basis for that grammaticalization is and also
where select individual lexical items which are difficult to categorize fall on the cline
and what the reasons for their positions are.
I will also discuss Paul Schachter’s (1985) seminal paper, ‘Parts-of-speech
systems’, which will be used to provide definitional assistance over the course of this
chapter.
While it is assumed here that the assignment of words to parts-of-speech
classes is based on properties that are grammatical rather than semantic, and
often language-particular rather than universal, it is also assumed that the
name that is chosen for a particular parts-of-speech class in a language may
appropriately reflect universal semantic considerations. Thus, although the
familiar notional definition of noun mentioned above does not always
provide an adequate basis for deciding whether or not a given word is a noun,
once the words of a language have been assigned to parts-of-speech classes
on grammatical grounds and it is found that one of these classes includes the
preponderance of words that are the names of persons, places, and things,
then it is perfectly reasonable to call this class the class of nouns, and to
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
compare the class so named with the similarly named classes of other
languages. (Schachter 1985: 4)
An important concept to be used in the current discussion is that of the
necessity and use of necessary and sufficient conditions, as discussed in Matthews
(2003) and Dixon and Aikhenvald (2003). Necessary conditions are those which all
entities within a particular class possess, but which might also be possessed by
another class. Sufficient conditions are those conditions which are only possessed by
a particular class, but which may not be possessed by every member of that class. If
one can find individual conditions within a language for the definition of a particular
element which are both necessary and sufficient, then the task is relatively easy.
Unfortunately, it is very rare that one can find conditions or properties which are
both necessary and sufficient to define a particular class. For example, one might
describe candy as being sweet. This is indeed a property of candy, even if individual
candies also have the property of being very sour or spicy. Flowever, it is also a
property of other things, such as ice cream, cake, and prawns in coconut milk with
walnuts, and so will not serve to define candy by itself. One might try again to come
up with a definition of candy as being made from boiled sugar and then flavored with
peppermint oil. All things which match this definition are, in my book, candy.
Unfortunately (or quite fortunately as the case may be), although this definition
might be necessary and sufficient to define all peppermint candies, there are also
candies made from cocoa butter, sugar, and chocolate liqueur which would not lit
this definition. One must be careful then in our definitions to be neither too broad nor
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
too narrow. One shall, therefore, strive to find a combination of necessary and
sufficient conditions which define an individual class, taking into account Whorfis
(1945) admonition not to look to the use as the primary criterion, but rather to look to
other independent facts. In all likelihood, I will be unable to come up with such
criteria and will have to come up with a list of conditions, both necessary and
sufficient, which, combined, can be said to be prototypical of a particular category.
For the most part, I will not distinguish between lexical classes on the basis
of whether or not they are able to enter into constructions with pronominal clitics, as
there would be one category which does and subsumes everything except for
demonstratives, adverbs, interjections, and particles. As suggested in the quotation
by Munro above, I will crucially use three factors in determining category
membership: syntactic (what combinations the particular lexical item can be a part
of), morphological (what morphological marking there is), and semantic (what the
meaning is and how it interacts with other lexical items). I will sporadically follow
Schachter (1985) in my presentation and, where possible, use his data as a means of
comparison. In general, I will try to state whether or not a particular part of a
characterization of a lexical class is necessary or sufficient by placing it in
parentheses. Unfortunately, it is very rare that I have found definitions which are
simultaneously both necessary and sufficient.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
7.2 Verbs
The large open class of verbs can be defined in Zoogocho Zapotec as
consisting of those lexical items which take aspectual prefixes (see 4.3 for more
information) (necessary and sufficient). Most verbs also require the presence of a
nominal or pronominal subject. The following examples exemplify the preceding
statements.
(1) y-e-zhinh=dxo
pot-freq-arrive= 1 plincl
‘We arrive...’
(2) b-e-ban=e’
comp-freq-live=3 f
‘S/he woke up.’
(3) (a) b-zhinh=a’
comp-arrive=l sg
‘I arrived.’
(b) *b-zhinh
Schachter characterizes verb as ‘the name given to the parts-ol-speech class
in which occur most of the words that express actions, processes, and the like. The
characteristic function of verbs is as predicates’ (Schachter 9). However, he does cite
examples like the following Tagalog data to show that verbs can be used in some
languages as arguments. In the first sentence, the verbpinanood ‘watched’ requires a
subject and an object. In this case, the object is the verb sumasayaw.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(4) Pinanood ko ang mga sumasayaw.
watched I TOP PL were dancing
‘I watched the ones who were dancing.’
Cf. Sumasayaw ang mga tao
Were dancing TOP PL person
‘The people were dancing.’(Schachter’s 13) (ibid. 9)
He does note that it is more common to use a verbal noun (meaning a noun
morphologically related to a verb) as an argument, as in the following example he
gives from Akan (5).
(5) MehwEE asaw no
Iwatched dancing the
‘I watched the dancing.’(Schachter’s 14) (ibid. 9)
If we consider the following examples from Zoogocho Zapotec, we see in (6) the
verb root -yaa- used as a verbal predicate.
(6) gu-yaa-lenh=a=le zha=be’
pot-dance-with= 1 sg=2pl stat.say=3inf
‘I will dance with you all, he said.’
In (7) and (8), we see it used as a verbal noun.
(7) 0-chaz=a’ weyaa
pot-participate= 1 sg inf. dance
‘I will take part in the dance.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
244
(8) 0-wi=a’ weyaa
pot-see_performances= 1 sg inf.dance
‘I will see the dance.’
Schachter notes that ‘(T)he categories for which verbs may be specified
include tense, aspect, mood, voice, and polarity’ (ibid. 10) and that many languages
have a subclass of ‘ copulative verbs’ (ibid. 11). As seen in the Chapter Four,
Zoogocho Zapotec verbs are specified at least for aspect. There is one copulative
verb, as seen in the following example.
(9) benhe n-ak noolh wego
person stat-be woman virgin
‘The person is a young woman.’
7.2.1 Auxiliary verbs
Auxiliary verbs, a small, closed class, are the only verbs which do not have to
have an obligatory subject (necessary and sufficient). See 6.5.4 for further
discussion.
7.2.2 Positional verbs
In this section, I will discuss a particular closed subclass of verbs, positional
verbs. Positionals differ from other verbs, forming their own subclass, in not having
an aspect marking on the stative form of the verb (necessary) when taking part in the
Basic Locative Construction3 (sufficient). The Basic Locative Construction is
defined to be the default response to a question which asks where something is.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
245
Many verbs can take stative marking in the form of a lenis nasal prefix for which the
place of articulation assimilates to the following consonant as in (10).
(10) m-ban=e’
stat-live=3f
‘S/he is alive.’
The positionals and the existential (among a few other verbs) do not use the nasal
prefixal stative form. Also, when used to describe location the positionals are always
in the stative aspect. As Zoogocho Zapotec has no tense marking, the stative can be
used for past, present, and future temporal domains. When used in the Basic
Locative Construction, regardless of temporality, the stative form of the positional
verb is used. The following table shows the basic positional verbs.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
246
Table 7.1 Positional verbs
Dxi Zhia Zee Nala Zehe Yoo
‘sit’ ‘sit’ (not ‘stand’ ‘hang’ ‘hang’ ‘be
(persons used with animates (especially inserted
or persons at a higher in, wrap
animals, and altitude, around'
but also animals like an
things) except if a
person is
mounted
on a
horse)
electrical
wire)
Daa Zxoa Nkwaa Ndobe Ndosa Shtulhe
‘be stuck ‘lie in an ‘to be ‘to be ‘to be ‘to be
on’ extended heaped’ folded or upside lying on
fashion’ wrapped
around’
down’(as in
a pot lying
on its
opening)
its side'
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
247
In addition there is a word dee which means ‘to lie (but not necessarily he
spread out)’, a word nase which means ‘to be spread out (multiple objects such as
beans)’ and a doubtful existential which will be discussed later.
The Basic Locative Construction is as follows:
(11) (dem) Positional Figure (Relational Noun Ground)
(na’) dxi be’ko’ (lho yixe)
(there) stat.sit dog (in weeds)
‘A dog is in the weeds.’
As indicated by the parentheses, either a demonstrative or the relational noun
and ground must be present, or potentially both. The basic locative question has two
forms. The first form, which seems to be more common, is as follows.
(12) WHERE Positional Figure
Gan dxi be’ko’?
Where stat.sit dog
‘Where is the dog?’
The answer to this type of locative question generally does not include the
positional verb and consists solely of the relational noun and the ground as in the
following.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
248
(13) Relnoun Ground
lho yixe’
in weeds
‘In the weeds’
An alternative form doesn’t include the positional and is as follows.
(14) WHERE Figure
gan beko’
where dog
‘Where is the dog?’
The answer to this type of question is generally a fully formed construction such as
in (11) above.
I have some difficulties in determining what to consider the verb zoa to be in
the general scheme of the positionals. It is quite often glossed as ‘to be standing’
when used with inanimate objects such as trees or mushrooms as in the following
example.
(15) na’ zoa to bi’a’
demdist exist one mushroom
‘There was a mushroom standing there.’
Note that the corresponding word zee ‘to stand (of an animate) cannot be used with
inanimate objects.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
240
(16) na’ zoa to bas lho m es-en’
demdist stat.stand one glass stomach table=det
‘There is a glass standing on the table.’
(17) *na’ zee to bas lho mesen’
Zoa can also be glossed quite often as ‘to be’ as in the following4.
(18) ni zoa to zhome’
here stat.exist one basket
‘Here is a basket.’
It also quite often means ‘to live’ as in the following
(19) gan zoa=be’
Where stat.exist=3inf
‘Where does he live?’
It is noteworthy that the majority of the pictures in elicitation tasks such as
the Bowerman-Pederson book, which I used while working at the Max-Planck
Institute for Psycholinguistics in Nijmegen and which were described with this verb
showed inanimate figures in a vertical position. Note that the verb dee can also be
used in such a way as well, but is much more common in negative constructions.
The general locative zoa is used to describe location in a space such as a
town.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
250
(20) zo=a’ yezh
stat.exist= 1 sg town
‘I am in town’
The specific positionals are used to indicate location, as we have already seen.
However, the positionals dxi, nala, and zehe can be used to assert the
existence of abstract nominals, as in the following examples.
(21) zehe-dxgwa yizhwe
stat.hang-much sickness
‘There are many sicknesses.’ (From time to time)
(22) dxi-dxgwa yizhwe
stat.sit-much sickness
‘There are many sicknesses.’ (Most of the time.)
(23) dxi-dxgwa dizha’
stat.sit-much words
‘There are many words.’
(24) nala magia
stat.hang magic
‘There is magic’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(25) zehe x-medxu=a’
stat.hang poss-money= 1 sg
‘I have money.’ (It used to be that money would be hung from the rafters in
small ceramic containers.)
(26) dxi yelha justis yezh=en’
stat.sit dream justice town=det
‘ There is justice in the town.’
7.3 Adverbs
There are two classes of adverbs in Zoogocho Zapotec: a closed class of
adverbial suffixes, which have been discussed in 4.3.11, and an open class of full
adverbs. Full adverbs are distinguished from other lexical categories in the following
ways: they cannot take pronominal clitics (necessary) (cf. (27)), and they have a
relative freedom of ordering in the sentence (sufficient) (cf. (28)-(30)). Adverbs are
an open class, especially locative adverbs which can be easily created from most any
place name or location with the directional marker discussed in 5.3.7.
(27) *bedaones=a’
quickly=Tsg
(28) b-zxit lalo bedaones
comp-jump Lalo quickly
‘Lalo jumped quickly.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(29) bedaones b-zxit lalo
quickly comp-jump Lalo
‘Lalo jumped quickly.’
(30) b-zxit bedaones lalo
comp-jump quickly Lalo
‘Lalo jumped quickly.’
Schachter notes many difficulties in defining adverbs and comes to the
conclusion that, in order to avoid limitations which a definition of adverbs as
‘modifiers of verbs, adjectives, and other adverbs’ might have with respect to
adverbs which operate on the level of the verb phrase or sentence, that it is best to
‘say that adverbs function as modifiers of consitituents other than nouns’ (ibid. 20).
Temporal particles such as za ‘barely, hardly (Spanish ‘apenas’)’ or ba ‘now, at this
point (Spanish ‘ya’) also should be considered to be adverbs by the current
definition.
7.4 Nouns
7.4.1 Nouns
The large, open class of nouns crucially can cooccur with demonstratives
(sufficient) (cf. (32) below), can be modified by adjectives (sufficient) (cf. (32)
below), and do not take aspect markers (necessary) (cf. (33) below).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(31) x-kabayw=a’
poss-horse=lsg
‘my horse’
(32) mbis ni’
cat demprox
‘This cat’
(33) nis zahag
water cold
‘cold water’
(34) *gw-mbis
comp-cat
Schachter states the most common function of nouns is to act as argument s,
and this is what nouns (and noun phrases) in Zoogocho Zapotec do most commonly
as well. He notes that they can also function as predicates, as in the following
examples, either with copulas (35) and (36) or without.
(35) They are teachers.
(36) Su malamai ne. (Hausa)
they teachers COP
‘They are teachers.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
254
(37) Mga guro sila (Tagalog)
PL teachers they
‘They are teachers.’
(38) Oni u(ch)itelja (Russian)
they teachers
‘They are teachers.’ (Schachter’s (3)-(6)) (ibid. 7)
Zoogocho Zapotec uses either a copula or juxtaposed noun phrases, as in the
following example.
(39) a) n-ak=be maestr
stat-be=3inf teacher
‘They are teachers.’
b) maestro lhegakbe’
teacher 3plinf
‘They are teachers.’
7.4.2 Reflexives and Reciprocals
Reflexive pronouns are pronouns ‘which are interpreted as coreferential with
another nominal, usually the subject, of the sentence or clause in which they occur'
(ibid. 27). Schachter notes that there are ‘many languages (,) where reflexive forms
are analyzable as a head nominal modified by a pronominal possessive agreeing with
the subject.... There are also languages such as Malagasy, (69) (our (40) A.S.), that
use a common noun without a modifying possessive’ (ibid. 28). Zoogocho Zapotec.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
as seen in (41) could be viewed as being one such language, with the inalienably
possessed noun kwin= ‘self acting as a reflexive marker and the inalienably
possessed noun lghezh= ‘fellow’ acting as the reciprocal marker. This is a form of
the reflexive-of-possessor construction, which has been discussed in 6.4. The reason
for separating these elements from the rest of the nouns is that these elements
obligatorily take part in the reflexive of possessor construction. (Necessary and
sufficient.)
(40) Namono tena Rabe
Killed body Rabe
‘Rabe killed himself.’
(41) gud-ap kwin ron
comp-slap self Aaron
‘Aaron slapped himself.’
‘ Reciprocalpronouns, like reflexive pronouns, are interpreted as coreferential with a
co-occurring nominal, but are used to express mutual actions, conditions, etc’(ibid.
29). See the following example from Zoogocho Zapotec.
(42) shone nia’ gw-a Ighezfodxo
three feet/times pot-carry recip=lplincl
‘We are going to carry each other, three times.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
256
7.4.3 Pronouns
The closed class of personal pronouns can be defined as being the class of
words which, in clitic form, can cliticize to nouns, verbs, adjectives, and quantifiers
(necessary and sufficient). In non-cliticized forms, pronouns are not modified by
adjectives (necessary and sufficient).
(43) dxoalao=a’
moutheye(face)= 1 sg
‘my face’
(44) gw-dezh=a’
comp-cry=lsg
‘I cried.’
(45) gol=to
old=lplexcl
‘We are old.’
(46) yogo=ba
all=3an
‘all of them’
(47) *neda zahag
The personal pronouns have been discussed in greater depth in Chapter
Three.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
257
‘Indefinite pronouns are pronouns like English someone, something, anyone,
anything’ ’ (ibid. 30).
A prominent indefinite pronoun in Zoogocho Zapotec is the following: no
‘someone, something’, However, this is used much more commonly as a quantifier,
meaning ‘some’, or as a question word. In the following example, the word no is
used meaning someone. It appears that the part of the sentence following no might
possibly reflect an earlier way of forming a disjunction.
(48) zegh no zegh bi zegh
comp.go someone comp.go no comp.go
to-z=be’
one-alone=3inf
‘Someone went who (normally) might or might not go by themself.’
7.4.4 Generic nouns
Generic nouns are a small closed class of nouns derived from the classifiers
which will be discussed below in 7 .7 .1 am following Marlett (1985) in calling them
generic nouns, although they are generally used in reference to a previously
mentioned entity as in (49). They are distinguished from nouns and regular pronouns
in that they cannot, by themselves, serve as subjects, as seen in (50). (Necessary and
sufficient.) They differ from the classifiers in that they can occur on their own in a
sentence, not requiring either a demonstrative or adjective unlike the classifiers and
occur after the noun which they modify.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(49) bi gud-aw=a’ danh
neg comp-eat=lsg geninan
‘I didn’t eat that.’
(50) *b-zhinh danh/benh/bi
comp-amve geninan/ genan/gensm
Generic nouns crucially serve the function of introducing a relative clause and as
such are also used as relative pronouns. As Schachter writes:
‘Relative pronouns are pronouns like English who and which in (51).
(51) The man who wrote that was a genius.
The book which he wrote was brilliant.’ (ibid. 31)
ZZ has the following corresponding example. I choose to call the elements
benh, danh, and binh relative pronouns when used as such because they agree in
gender with the head noun, even if they do not otherwise show any grammatical
relation within the relative clause. (See Keenan (1985), Payne (1997)).
(52) b-zhinh noolhe=n’ benh dx-on rmed
comp-arrive woman=det who cont-make remedy
‘The woman arrived who cures.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
259
7.5 Adjectives
Adjectives can be distinguished from the other lexical categories in the
following ways: adjectives co-occur with nouns (necessary) (53), follow the noun
(54) unless used as a predicate (55) (necessary (this differentiates them from
demonstratives, determiners, and quantifiers)), do not take aspectual markers (56)
(necessary (differentiates them from verbs)), and never show the morphosyntax of
possession (57) (necessary ( this distinguishes them from nouns)). It is very difficult,
in Zoogocho Zapotec to find a single criterion which is sufficient to classify an
adjective. However, the criteria which have already been stated are enough to
distinguish them from all other lexical classes in SBZZ.
The adjectival class is a relatively large, open class, containing recent loan
words such as font ‘foolish’. Along with the nouns, and verbs, these arc the three
main open class lexical categories in SBZZ. In San Lucas Quiavini Zapotec, as
described by Pamela Munro (2002), there is a lenis nasal alveolar prefix which
accompanies some adjectives, and for which we find the occasional reflex in
Zoogocho Zapotec (such as the SBZZ word, mba ‘happy’). Unfortunately, it is not
very common and is fossilized in SBZZ. See Munro (2002) for an enlightening
description of the difficulties in determining parts of speech in a number of
languages.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
26(1
(53) bekozxo shish
rebozo white
‘White rebozo’
(54) be’ko’ gasgh=en’
dog black=det
‘the black dog’
(55) lhaa-dxgwa kushiyo=n’
sharp-very knife=det
‘The knife is very sharp.’
(56) *gu-gasgh=a’
comp-black=lsg
*’I was black.’
(57) *x-gasgh=a’, *gasgh chi=a’
Schachter notes that other researchers (Jespersen (1924) and Lyons (1971))
have mentioned the shortcomings of a definition of adjectives as being ‘the class of
words denoting qualities or attributes’ (ibid. 13), and that ‘adjectives have usually
been defined at least in part in functional terms as words which modify nouns’ (ibid.
13). He also states that along with their function modifying nouns (as we have seen
above), adjectives can also be predicates and can either occur in some languages
with a copula, in other languages without a copula, or in some languages with and
without a copula (ibid. 13), as seen in the following examples from Zoogocho
Zapotec.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
26 i
(58) yag=en n-ak=en da ga
tree-det stat-be=3inan clinan green
‘The tree is green.’
(59) tonhe-dxgw=a’
long/tall-int=lsg
‘I am really tall.’
7.6 Quantifiers
Quantifiers form a distinct lexical class because they are invariably
prenominal nominal modifiers (sufficient) which do not take aspectual markers
(necessary). This is a small closed class, although in the case of numerals, members
of this class, they can be combined to create a potentially infinite number of
members.
(60) yogo mbis
all cat
‘all cats’
(61) yogo=ba
all=3an
‘all of them’
(62) *gw/y/dx/n/...-yogo=ba
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
262
(63) shone be’ko’
three dog
‘three dogs’
7.7 Noun and verb adjuncts
In this section, I will discuss what Schachter calls noun adjuncts, ‘several
classes of words that typically form phrasal constituents with nouns’ (ibid. 35) and
verb adjuncts ‘two classes of words that form phrasal constituents with verbs:
auxiliaries and verbal particles ’ (ibid. 35). He differentiates four classes of noun
adjuncts: role markers, quantifiers, classifiers, and articles (ibid. 35). 'Role markers
include case markers, discourse markers, and (other) prepositions’ (ibid. 35). As will
be seen in Chapter Eight, discourse function in SBZZ is marked via word order, as is
syntactic and semantic role. One is then left with the prepositions and relational
nouns, as described below in 7.12. Quantifiers ‘consist of modifiers of nouns that
indicate quantity or scope’ (ibid. 38) and have been discussed above. Classifiers ‘are
words which are required by the syntax of certain languages, when a noun is also
modified by a numeral’ (ibid. 39).
At this point, I will discuss the classifiers. There are three classifiers in the
language. These differ from the classifiers which Schachter discusses in that they are
not of the numeral or mensural type described by Schachter, but are instead more
like a small system of noun classes as in Bantu languages. Classifiers correspond to
the generic nouns which, with the exception of the alternate form hi, are constructed
from classifiers by the addition of a fortis nasal /nh/. Classifiers are sometimes used
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
263
with adjectives, and form demonstrative pronouns with the demonstrative adjectives
(see 7.8 below) (necessary and sufficient). The following are the classifiers.
(64) be ‘animal classifier’
(65) da ‘inanimate classifier’
(66) bi ‘classifier for small things (both animate and inanimate)
(67) n-ak-dx bdxee be lis kleka’ be’ko'
stat-be=more ant clan small comp dog
‘Ants are smaller than dogs.’
Finally, Schachter discusses articles in which category he includes, ‘in
addition to the words usually identified as definite and indefinite articles (e.g.
English the, an), words that are sometimes identified as demonstrative adjectives or
modifiers (e.g. this in this man, that in that woman)’. His reasoning is that
demonstrative adjectives and modifiers often have the same distribution as articles,
which is true of SBZZ if one considers the examples below, where all follow both
noun and adjective. Determiners are distinguished from other postnominal modifiers
in that they cliticize to the final word in the noun phrase (necessary and sufficient).
(68) be’ko’ gasgh nga
dog black demmed
‘this black dog’
(69) be’ko’ gasgh=en’
dog black=det
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
26-1
‘the black dog’
However, note that to which is often translated as an indefinite article, shows up
prenominally.
(70) to be’ko’ gasgh
one dog black
‘one black dog’
Schachter goes on to talk about verb adjuncts, primarily auxiliaries and
verbal particles. ‘Auxiliaries are words that express the tense, aspect, mood, voice,
or polarity of the verb’ (ibid. 41). As has been discussed in greater detail in 6.5.4, -
ak among other verbs act as auxiliaries in SBZZ. Note that they are differentiated
from regular verbs in that they can occur without the normally obligatory subject.
(71) dx-ak dx-on=o’ shinh
cont-can cont-make=2sg work
‘You can work.’
The other class of verbal adjunct which Schachter mentions are verbal particles, 'a
closed class of uninflected words that co-occur with certain verbs’ (ibid. 45). These
are elements such as the up in English hurry up. There are no verbal particles in
Zoogocho Zapotec.
7.8 Demonstratives
As discussed in Diessel (1999), there are two important facets which one
must take into account when examining demonstratives from a categorial
perspective: the distributional characteristics of the item, and separately the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
265
categorial status of the item. He notes four distributions and corresponding
categories, his Table 1, modified and repeated here:
Table 7.2 Distributions and categories of demonstratives
Distribution____________________ Category
pronominal demonstrative demonstrative pronoun
adnominal demonstrative demonstrative determiner
adverbial demonstrative demonstrative adverb
identificational demonstrative demonstrative identifier (ibid. 4)
While they do not on first consideration form a unified lexical class, unlike
the classes which I have discussed so far, I choose to discuss them together in the
current section because they are derived from the same roots. The base
demonstrative would be considered to be a demonstrative adverb5, which can be
used as a nominal modifier6 as well, as seen below. Note that they are homonymous
and I see no reason to separate them into different classes.
To begin with, the base demonstratives ni ‘proximate demonstrative’, nga
‘medial demonstrative’, and n a ’ ‘distal demonstrative’ can either be used as nominal
modifiers, as in (72) or as verbal modifiers (locational deictic) as in (73). I use the
terms proximate, medial, and distal, because they seem to be the best terms to use.
When I elicited them, ni ’ has been used to describe objects or actions that happen in
the immediate vicinity of speaker (say within an arm’s length), nga to describe
objects or actions that occur somewhat close , and na ’ to describe objects which
occur further away. When lining up three pencils, ni ’ is used to describe the one
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
266
closest to the speaker, nga the one in the middle, and na ’ the one furthest. All three
pencils could potentially be described as being ni ’ or nga. This having been said,
however, it is my intention to more rigorously test these meanings.
Note that, when used as a nominal modifier, its position in the noun phrase is
fixed (as seen by the ungrammaticality of (74)), but when used as a verbal modifier,
its position is as free as that of any other locational adverb (normally7 not being able
to intervene between verb and subject) (as seen in (75)).
(72) benhe xo nga
person old demmed
‘This old person.’
(73) nga O-nkwaa-shka dao
demmed stat-liestacked-emph corn_ tassels
‘The com tassels lay out there stacked up.’
(74) * nga benhe xo
(75)a) O-nkwaa-shka dao nga
stat-lie stacked-emph tassel demmed
‘The com tassels lay out there stacked up.’
b) * O-nkwaa-shka nga dao
stat-liestacked-emph demmed tassel
‘The com tassels lay out there stacked up.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
One issue here which is of interest with respect to Diessel’s typology is that,
in Zoogocho Zapotec, adnominal and adverbial demonstratives seem to class
together in opposition to the derived pronominal demonstratives, which is a pattern
which Diessel does not mention. The addition of a classifier (bi, be, and da (cf. 7.7))
allows the demonstrative determiner to be used in pronominal demonstrative
contexts8 , as seen in the following examples:
(76) bi=ni this one (proximate) (used for small things)
bi-nga this one (medial) (used for small things)
bi=na ’ this one (distal) (used for small things)
be=ni this one (proximate) (used for animates)
be=nga this one (medial) (used for animates)
be=na’ this one (distal) (used for animates)
da=ni this one (proximate) (used for inanimates)
da=nga this one (medial) (used for inanimates)
da=na’ this one (distal) (used for inanimates)
(77) bi dx-■een=da’ be=nga dx-een
neg cont-want=l sgexp clan=demmed cont-want=lsgexp
be=na’
clan=demdist
‘I don’t want this one, I want that one.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 6 8
Demonstratives are not used in the identificational sense which Diessel
mentions, such as the French cela.
Now, as promised, I will discuss the categorial status of demonstratives in
Zoogocho Zapotec. It would appear that, with respect to the base demonstrative
which serves adnominal and adverbial functions, there are two potential options with
respect to its lexical classification. One could differentiate between the two functions
and classify the locative deictic with other adverbs and the demonstrative adjective
by itself (it cannot be used predicatively); or, in my view more correctly, one could
claim that there was one lexical class of demonstrative, and that, necessarily and
sufficiently, its ability to be used adverbially and adnominally serves as a means of
distinguishing it from other lexical classes.
7.9 Question Words
I include a class of question words, even though, as Schachter mentions,
question words or interrogative proforms as he refers to them, cut across many
lexical classes. There interrogative pronouns like English who or SBZZ no ‘who’,
interrogative adverbs like English how or SBZZ nak, and interrogative adjectives
like English which or SBZZ bi. These and the rest of the set are discussed in greater
detail in 5.5.3. Note that there is overlap between these. For example no can be used
as an interrogative adjective and bi can be used as an interrogative pronoun. The only
conditions one needs initially to separate this whole lot as a lexical class are that they
must be placed preverbally, which is necessary to define individuals of this lex ical
class as belonging to a separate lexical class (unfortunately it is not sufficient, as the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
269
focus marker le also must occur sentence initially), and to state that they are used in
content questions (which is sufficient, but unsatisfactory as it is notional). Note that
while these are syntactic criteria and could be said to define the class as a syntactic
position, it is consistent with the general framework set out in 7.1 to consider them to
be a separate class based on their syntactic uniqueness.
(78) bi sh-laa-dxgwa?
what cont-smell-emph?
‘What really smells?’
(79) *sh-laa=dxgwa bi ?
*cont-smell=emph what?
‘What really smells?’
7.10 Conjunctions
Schachter defines conjunctions as ‘words that are used to connect words,
phrases, or clauses’ (ibid. 46) distinguishing between coordinating (which assign
equal weight to the coordinated elements (such as English and or SBZZ n a ')) and
subordinating conjunctions (in which one clause or element is marked as being
subordinate to the other one (like English that or SBZZ ga, as seen below)) (ibid.
48). In SBZZ, there are coordinating conjunctions which conjoin two noun phrases,
and coordinating conjunctions which conjoin both nouns and verbs. No element
cliticizes to a conjunction (necessary) and conjunctions require two arguments
(necessary and sufficient). I will briefly address the reasons for distinguishing
between conjunctions and prepositions below in 7.12.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 7 0
(80) yina’ o sede
chile or (Sp.,) salt
‘Chile or salt’
(81) g-za=a’ na’ gu-ya=a’
comp-walk= 1 sg and comp-go= 1 sg
‘I walked and I went.’
(82) yina’ na’ sede
chile and salt
‘Chile and salt’
Schachter goes on to differentiate three main types of subordinating conjunctions:
complementizers, relativizers, and adverbializers. ''Complementizers mark a clause
as the complement of a verb (cf. (ibid. 138) [repeated as (83) below AS]), noun, or
adjective ‘(ibid. 50). Example (84) shows the use of the complementizer ga ‘where’.
(83) Itinanong ko kung nasan sila Tagalog
asked I where they were
‘I asked where they were.’ (ibid: 50)
(84) ba n-zhaga-lao=dxo ga zegh=dxo
already stat-encounter-eye=lplincl where stat.go~-1 plexci
‘We are already rushed to where we go.’
‘Relativizers are markers of relative clauses ‘which crucially do not include
relative pronouns which crucially serve a nominal function within the relative clause'
(ibid. 51). There are no relativizers in Zoogocho Zapotec. However, the issue of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
resumptive pronouns (discussed in 6.5.7) muddies the water a little bit, as in those
cases, the relative pronoun could potentially be said to not play a central nominal
role in the relative clause.
‘Adverbializers mark clauses as having some adverbial function, such as the
expression of time, purpose, result, etc’ (ibid. 51). Note the following use of an
adverbializer in SBZZ.
(85) nake gu-ditgh=le kate n-ak=le bidao^na'
how comp-play=2pl when cont-be=lpl child=det
‘How did you all play when you were children?’
7.11 Other closed classes
Schachter mentions a number of other closed classes, including ‘clitics,
copulas and predicators, existential markers, interjections, mood markers, negators,
and politeness markers’ ’ (ibid. 53). Clitics have been discussed in Chapters Three and
Four.
‘Copulas are words used to indicate the relation between a subject and a
predicate nominal or adjective’ (ibid. 55). In Zoogocho Zapotec, these words arc a
subset of the verbs. Note that while it is difficult to find any criteria which de fine
copulas as opposed to other verb classes, they do tend to have repeated subjects and
preverbal subjects more than other verbs and without the pragmatic implications of
focus or topicalization which such preposing or repetition would have for other
verbs. See Chapter Eight for more on the issues of the repetition and preposing of
subjects.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 7 2
(86) bidao n-ak=be’
child cont-be=3inf
‘He is a child.’
Schachter also makes a distinction between predicators which are used when
there is no overt subject. Zoogocho Zapotec does not have a separate set of
predicators.
Emphasis markers/clitics which serve to emphasize the predicate such as -
dxgw=, are found in Zoogocho Zapotec. I have discussed the morphological status
and use of such markers in greater depth in Chapter Four. A verbal predicate,
adjectival predicate, or quantifier can be emphasized.
‘ Existential markers are words which are equivalent to English there is!are
etc’ (ibid. 57). He gives the following example form Hausa.
(87) Akwai littafi a kan tebur
EXIST book at top table
‘There is a book on the table.’ (Schachter’s (175) (ibid. 57)
He then goes on to mention that ‘(L)anguages that do not have existential
markers often use verbs meaning ‘be (located)’ to express equivalent meanings' , as
in the following example from Akan.
(88) sika bi wo me foto mu
money some islocated my bag in
‘There is some money in my bag.’ (Schachter’s 178) (ibid. 57)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The reader is referred to the earlier discussion of positional verbs in this
chapter, to see how locative constructions are used to express existential meanings.
Negative existentials are formed in the following way. The forms bibi (used
with inanimates) and nono (used with animates) do not occur outside of the
following construction which involves bibi or nono followed by the word dee or
another positional. Note that they only appear in this constructions.
(89) bibi dee zaha
neginan lie(exist) beans
‘There are no beans.’
(90) bibi dxi justis
neginan stat.sit justice
‘There is no justice.’
(91) nono dee doktor
negan exist doctor
‘There is no doctor.’
As an interesting side note, SBZZ speakers consider the following lexical items to be
animate, by this classification, and by the use of the classifiers.
(92) nono dee juguet
negan exist toy
‘There are no toys.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
274
(93) nono dee camion
negan exist bus
‘There is no bus.’
Schachter also states that it is relatively common for there to be a relationship
between existential and possessive constructions. He cites the following example
from Tagalog which I have adapted.
(94)(a.)Mayroon-g libro sa mesa
EXIST/POSS-LINK book on table
‘There is a book on the table’
Wala-ng libro sa mesa
EXIST/POSS(NEG)-LINK book on table
‘There isn’t a book on the table’
(b.)Mayroon-g libro ang bata
EXIST/POSS-LINK book TOP child
‘The child has a book.’
Wala-ng libro ang bata
EXIST/POSS(NEG)-LINK book TOP child
‘The child doesn’t have a book.’ (Schachter’s 180) (ibid. 57-58)
Note the following Zoogocho Zapotec examples, using positional verbs.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(95) dxi libr=en’
stat.sit book=det
‘There is a book./The book is there.’
(96) dxi libr chi=a’
stat.sit bookof=lsg
‘My book is there.’/ ‘I have a book.’
‘Interjections are words, often of an exclamatory character, that can
constitute utterances in themselves and that usually have no syntactic connection to
any other words that might occur with them’ (ibid. 58). The following is an example
of an interjection in Zoogocho Zapotec.
(97) bi gu-yegh=o?
not comp-drink=2sg
‘You didn’t drink?’
od
‘no’
‘ Mood markers are words that indicate the speakers attitude or that solicit the
hearer’s attitude, toward the event or condition expressed by a sentence’ (ibid. 58).
One such marker in Zoogocho Zapotec is sheka which marks that the speaker is not
entirely sure about the veracity of the statement, as seen in the following exam ple.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
276
(98) sheka bi g-ak yogh
belief neg pot-be rain
‘I don’t think its going to rain.’
‘Negators are words like English not, which negate a sentence, clause, or
other constituent’ (ibid. 59). I will now just say that they are differentiated in the
following way: they occur immediately prior to the element which they negate.
Consider the following example of a negator in Zoogocho Zapotec.
(99) bi gu-daw=a’
neg comp-eat=lsg
‘I didn’t eat.’
Schachter mentions politeness markers, which ‘express a deferential attitude
towards the person addressed’ (ibid. 60). There are no such markers in Zoogocho
Zapotec. There are formal and informal forms of the third person pronominal forms,
but these are among the only grammaticalized ways one can express such attitudes.
Among the other lexical classes which Schachter mentions as potentially
occurring in a language are proforms such as; pro-sentences, ‘words like English ye.v
and no which are used in answering questions and are understood as equivalent to
affirmative and negative sentences respectively’ (ibid. 32) (as seen in examples
( 100a) and ( 100b) below (note that these are both interjections)), pro-clauses, as in
the question tag ke in example (101) below, pro-verbs (which do not apply to ZZ).
pro-adjectives, pro-adverbs (see example ( 102) below to see examples of the use of
ka in SBZZ as a pro-adverb), and finally interrogative pro-forms which as he notes
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
277
cut across lexical classes, and which were discussed earlier in the current chapter.
While I am describing them here, they do not form part of a discrete lexical class and
are included here because they do not fit discretely anywhere else.
(100) zegh=be’
comp. go=3 inf
‘Did she go?’
(a) 06 ‘no’
(b)uhuh’ ‘yes’
(101) gu-da=lenh=e=be’, ke?
comp-walk-with=3f=3inf, really?
‘She went with him, didn’t she?’
( 102) dx-osia bsia=n ‘ ka’
cont-scream eagle=det demadv
‘The eagle screams like that.’
7.12 The Relational Noun/Locative Preposition Continuum.
7.12.1 Introduction
The grammaticalization of prepositions from relational nouns (body part
terms for human and animal bodies used in spatial description) is an issue which has
received a great deal of attention from descriptive and comparative linguists. In
Zoogocho Zapotec, human body part terms form a nascent grammatical category
which has not undergone the amount of semantic bleaching (abstraction) which has
occurred in many languages (members of this class are mostly still not used in a
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
grammatical sense, which, as we will see, is a major defining criterion for the
category o f preposition), and which is not uniform in the status of its members. Some
body part nouns do have generalized locative uses; yet even in these cases, there is
still both the potential for interaction with the base metaphorical system and the
canonical, fixed use which makes them differ from locative prepositions in a
language like English, or the varieties of Valley Zapotec discussed in Lillehaugen
(2003) and Munro (2002). As has been mentioned in other contexts, including the
discussion in Hollenbach to which I will come later, the presence of positional verbs
which indicate, in part, the relation of figure to ground and thus share some of the
relational burden of the body part term, both help to be the source of and also to
explain the variability in the status of the individual body part terms.
In this section, I will discuss the continuum which exists between relational nouns
and prepositions in Zoogocho Zapotec. I will begin by an examination of one
example of what has been said crosslinguisticallly on the issue, namely Heine et al.'s
(1991) position on the grammaticalization of body part terms, largely based on data
from African languages. I will then go on to a discussion of the issues in one Zapotec
language (MacLaury (1989)) and Mixtecan (based mostly on Hollenbach (1995)). I
will then propose a synthesis of the various approaches for the relational
noun/preposition cline, apply this synthesis to the relational body part terms of
Zoogocho Zapotec, and end with a comparison of a group of Zapotec languages in
which the body part terms have been much more grammaticalized than in SBZZ
(Lillehaugen (2003) and Munro (2002)).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
7.12.2 Heine et al.
Heine et al. (1991) list the following stages as a typical conceptual path
through which body parts develop into spatial concepts in African languages. This is
meant to model the cognitive development from concrete object to spatial
description, and is not meant to represent categorial changes.
(103)
STAGE CONCEPTUAL DOMAIN
0 Body part of X OBJECT
I Subpart o f X, spatially defined OBJECT/SPACE
1 1 Space as part of and adjacent to X SPACE/OBJECT
III Space Adjacent to X SPACE (Heine et al. 130)
Examples from Swahili are given for stages I-III. Note that no example of
stage 0 is given, as the word mbele (which originally came from ‘the lexeme *-hele
‘breast” (ibid. 131)) no longer functions as a body part term. In the following
example, Stage III has adverbial syntax. The importance that I am placing on it is the
(extremely) general semantics associated with it.
(104) Stage I:
mbele ya gari lake ni nyeusi
front of car his is black
‘The front part of his car is black.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 8 0
(104 cont.) Stage II:
taa ziko mbele ya gari
lamps are front of car
‘The lamps are on the front part of the car.’
or
mbele ya gari lake ni peusi
front of car his is LOC. black
‘The space in front of his car is black (e.g. in a garage).'
Stage III:
gari liko mbele
car is front
‘The car is in front/ahead.’ (ibid. 131)
Heine et al. then go on to list a number of positions which Africanists have
used in the description of prepositions derived from nouns. They are repeated in
(105) below.
(105) (a) Adpositions, or ‘prepositions’, are words that can be translated by
prepositions in a given matrix language, like English, German, or
French.
(b) They are homophonous with or similar to nouns.
(c) They are nouns or form a distinct subclass of nouns.
(d) They are cognate with nouns.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(105 cont.)(e) They are historically derived from nouns, (ibid. 132)
They go on to state a typology for the ‘continuum of decreasing nominality
along which any given adposition may be located’ (ibid. 132).
(106) (a) ability (+) versus inability(-) to express a morphological number
distinction, that is, typically, to take a plural marker (PL);
(b) ability (+) versus inability(-) to take a demonstrative (DEM);
(c) ability(+) versus inability(-) to take adjectival qualifiers (ADJ);
(d) ability (+) versus inability (-) to permit relativization when not being
qualified by a genitive noun phrase (REL);
(e) ability (+) versus inability (-) to form the sentence subject when not being
qualified by a genitive noun phrase(SUBJ; cf.. h below)
(f) ability (+) versus inability (-) to take first- or second-person
possessive pronouns as modifiers (PRON)
(g) presence (+) versus absence (-) of a genitival/subordinating
morphology (GEN);
(h) ability (+) versus inability (-) to permit relativization when
qualified by a genitive noun phrase (REL GEN);
(i) ability (+) versus inability (-)to form the sentence subject as the head of a
genitive noun phrase (SUBJ GEN;cf.. e above);
(j) ability (+)versus inability (-) to take third-person possessive pronouns as
modifiers (PRON 3RD; cf.. f above) (Heine et al. 133)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 8 2
I will come back to these criteria in greater depth later in the current section.
One issue which might be considered at this junction is the issue of metaphor in the
grammaticalization of body part terms. When used metaphorically, the original
characteristics of a set of source concepts will not necessarily transfer to the target
dominain. Thus, one would potentially be remiss to say, in a discussion of the
metaphorical transfer from the human body to locative constructions, that because a
certain thing which might have been able to be said about the human body cannot be
said in the locative construction that this means that one apriori claims that two
different lexical classes exist. For example, while one may talk of the foundations of
a theory and one might construct theories like buildings, one generally does not
construct tall theories (Lakoff and Johnson 1987). Or, perhaps in a better example, if
one was using the metaphor anger is heat, exemplified by statements like That really
steams me or His blood boiled, and found that statements like That really quick (or
quickly) steams me or something in that sense to be nonsensical, one would not want
to necessarily say that a categorical shift has occurred. Metaphorical mappings are
not one to one, onto mappings (otherwise known as isomorphisms), to use the
mathematical terms. There are likely to be some semantic cooccurrence restrictions
on the interaction of the source and target domains of any metaphorical mapping,
and one should not be surprised if a mapped term does not have all the properties of
the original. I will come back to this point briefly when reformulating these
characteristics.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 8 3
7.12.3 Mixtecan
In this section, I will focus on the 1995 IJAL paper, ‘Semantic and Syntactic
Extensions of Body part terms in Mixtecan: the Case o f ‘Face’ and ‘Foot” , by
Barbara Hollenbach. In this paper, Hollenbach shows how syntactic reanalysis and
semantic shift has led to a wide range of meanings and functional use of those body
part terms in Mixtecan languages. She lists the following range of uses of words
diachronically derived from Proto-Mixtecan words for ‘face’: face, front of, top of,
in front of, on top of, in presence of, to, in place of, than, place (where), time (when),
when, and if (Hollenbach 170). The uses of words derived from the word for ‘foot’
are: foot, bottom of, beginning of, basis for, at foot of, at beginning of, for benefit of,
on behalf of, about, in exchange for, and because.
She claims the following with respect to the path which these words have
taken. First, ‘(W) ithin the spatial domain, the core meaning of a body part of a
person or animal is extended to some analogous part o f an inanimate ob ject. T he
mechanism that accounts for this change is metaphor, i.e., the mapping of an image
from one domain into another’ (ibid. 171). Next, ‘(A) second extension moves from
a part of an object to the space that projects out from that part’ (ibid. 171). T his is
described by the term ‘projecting space’ (ibid. 171), or ‘adjacent location’ by
MacLaury (1995). This change comes about via metonymy, however, rather than a
part/whole approach to metonymy, Hollenbach takes the approach that metonymy is
‘the use of a word for something associated with its original meaning’(Hollenbach
171). Following this, ‘(A) second path of semantic extension moves from the spatial
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 8 4
domain to the temporal’ (ibid. 172). Then, ‘(A) nother kind of change moves from
the spatial domain to the domain of logical entailment’ (ibid. 172). This could
include a detour through the temporal domain. I will not discuss the semantic
extensions further here, but will discuss the semantic extensions more fully in our
discussion o f MacLaury and Ayoquesco Zapotec and Zoogocho Zapotec.
Syntactically, ‘they [body part terms AS1 .] move from the major lexical class
noun, the members of which prototypically refer to concrete objects, to other parts of
speech, the members of which prototypically mark grammatical relations’ (ibid.
172). While Hollenbach posits two different types of syntactic change to account for
all of the various meanings that face and foot have taken, we will concentrate on the
change from noun to preposition, as the change from noun to subordinating
conjunction is not relevant to the discussion of Zoogocho Zapotec. The mechanisms
which she posits for the syntactic change are reanalysis and then generalization
based on the reanalysis (ibid. 173).
As Hollenbach notes, ‘syntactic function is shown mainly by word order and
not by case...(T)here are therefore no grammatical signals that distinguish the
schema preposition +complement NP... from the schema body part noun t - possessor
NP’ (ibid. 177). This could also be said of Zoogocho Zapotec. As will be seen,
possessive noun phrases and body part terms which are reanalyzed are
indistinguishable. However, the ‘trigger’ for the reanalysis is, as Hollenbach rightly
points out, ‘to be found in the nature of the verbs with which the phrases occur'
(ibid. 178). One such set is the positional verbs, which ‘link a subject and the place
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
where it is located’ (ibid. 178), and ‘express most of the relational notions that are
expressed by spatial prepositions in English and other European languages’ (ibid.
178). ‘The locational element in the sentence frequently contains a body part noun
and its possessor’ (ibid. 178). We will come back to this later in our discussion of the
Zoogocho Zapotec data. Before moving on to the description of the reanalysis and
the generalizations, we shall note that Hollenbach also states that ‘(V)erbs of
movement, transport, placement, and change of possession’ (ibid. 178) are also
similar in that they ‘express spatial concepts and all of them commonly occur with
possessible noun phrases containing body part nouns’ and ‘ the relation is expressed
by the verb’ (ibid. (178). With the high frequency of body part nouns being used in
such constructions, ‘(I)t is, however, also possible to view such sentence more like
English, with the relation found in the body part noun, as well as in the verb’(ibid.
179). Once such reanalysis has taken place, the door is opened to the generalization
of this construction to other verbs. The change in the syntactic category of the lexical
item has occurred (ibid. 180).
7.12.4 MacLaury and Ayoquesco Zapotec
In the important 1989 IJAL paper, ‘Zapotec Body part Locatives: Prototypes
and Metaphoric Extensions’, Robert E. MacLaury discusses the grammaticalization
of body part terms in Ayoquesco Zapotec and compares this with the
grammaticalization o f similar elements in Trique and Mixtec. He notes that
Ayoquesco Zapotec differs from the Mixtecan cases because, in addition to the fact
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
286
that Ayoquesco Zapotec uses only the human body as a metaphorical model lor the
description of space:
[Ayoquesco] Zapotec does not use body part terms as prepositions, nor even
as markers of goal, path and source; its dative extensions are restricted to
verbs of speaking, which locate one person in front of another. Whereas
Mixtecan generalizes some of these directional functions, Zapotec has
innovated use of body part words to differentiate whether a figure is close to
its ground or in contact. (MacLaury 120)
Even though he acknowledges that other Zapotecan languages also have a
class of prepositions, he reasons that the body part locatives are not prepositions on
the following grounds:
Unlike English prepositions, they are identical in form to the nouns applied to
body organs, their use in syntax is optional, they only add specificity to other
locative expressions, they do not complicate syntax, they do not denote
direction, and they do not mark grammatical relations as do case markers,
(ibid. fn. 3, 120)
MacLaury notes that these nouns behave exactly like possessed noun phrases
in the language. He also mentions that ‘face’ is the only body part term that
‘functions as a dative marker and as an expression of abstract location’ (ibid. 121).
While it is the most widely used term in Zoogocho Zapotec, we would be amiss to
say that it is used as a dative marker in SBZZ. He discusses, as I will note later in the
discussion of Zoogocho Zapotec, how the human body is used as a metaphorical
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
model for the description of the location of most objects, even for objects such as flat
mats or pieces of paper which are difficult to map the human body to or featureless
items such as balls, even using occasional novel body parts in such a function.
When body part terms specify the shape of a location and its relation to a
ground, they appear to mark goal, path, or source of motion. Nevertheless, all
information regarding direction and manner of motion is encoded by verbs,
such as ‘go’, ‘pass’, and ‘leave’, (ibid. 137)
MacLaury also makes the following distinction, which also will be relevant
to the later discussion, when discussing how the use of naming a part of an object
and naming a space adjacent to an object can lead to potential confusion.
Removed location versus location-in-contact
(107) b-zaby-ma gik yag
C-fly-3+animal head tree
(a) ‘The bird flew over the tree.’
(b) ‘The bird flew to the treetop.’ (37 in (ibid. 143))
(108) bzaby-ma gik lo yag
C-fly-3+animal head face tree
‘The bird flew to the treetop.’ (38 in (ibid. 143))
Similar ambiguity to that in (107) can be found in Zoogocho Zapotec and will
be discussed later.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 8 8
One final claim which MacLaury makes for Ayoquesco Zapotec which does
not apply to Zoogocho Zapotec is that Ayoquesco Zapotec does not have locational
verbs without ‘further specification of containment’. As I will show in the discussion
of Zoogocho Zapotec and has already been seen in Hollenbach’s discussion of the
issue in Mixtecan, the combination of positional verb and body part term or other
locative expression is absolutely necessary in the description of locative relations.
7.12.5 An initial description and textual exploration of relational nouns
in Zoogocho Zapotec.
I have examined and quantified the use of the words : ni ‘foot, below’; yichyji
‘head, above’; dxoalao ‘face, around’; kwit ‘side, beside’; lee ‘belly, middle’; Iho
‘intestines, inside’; kwitlee ‘middle of side’; dxoa ‘mouth, in front o f ; lao ‘eye, in
front of, to’; kuzhe ‘back, behind’; and zxan ‘buttocks, below’ in a variety of contexts
which I will immediately enumerate in a corpus of over 2000 clauses. 1 looked at
their primary use as body part terms, as in (109)-(110) below.
(109) gw-e-le’i-kse=do’ dxoalao=be’ do gxe do
pot-freq-see-emph=2sgexp face=3inf indef tomorrow indef
wizhghe
dayaftertom orrow
‘You will see his face tomorrow or the day after tomorrow.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(110) na’ zha yeshe ye nia=dxo
and stat.prick thistles foot=lplincl
‘And the thistles pricked us in the feet.’
I examined their use when used to describe a part of an item (cf. (111)-(112) belo
(111) na’ b-zu=e’ yetgha dxoa trapish^en’
and comp-put=3f cane mouth mill=det
‘And they put the sugar cane in the mouth of the mill.’
(112) za b-zu=e’ azulejo yichgh=en’ na’
just comp-put=3f tile head=3inan demdist
‘He had just put tiles on the roof (its head).’
I also examined their use when used to describe a location in relation to a
of an item (cf. (113) and (114) below).
(113) na g-loo=be=ba’ Iho danh
now comp-insert=3inf=3an inside geninan
‘She put it in it.’
(114) nak g-on=to y-e-dxogh=to yichgh=e’
how pot-do=lplincl pot-freq-exit^ 1 plincl head-^3 f
‘How are we going to do it so that we leave there by his head?’(This was
taken from a text in which children are instructed in a traditional dance, ‘1
danza de los tigres’, and in which they are going to exit the stage by
someone’s head. Thus, it is in an extended relation to the body part.)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
290
I also looked into those cases where, instead of just marking a location in
relation to a part, it marked a more general location.
(115) b-e-zhinh=be’ lao xa-xna=be
comp-freq-arrive=3inf face fat her-mother^ 3
‘She arrived in front of her parents.’
(116) dx-bezh-ks=a’=nda’ lao dio
cont-cry-emph= 1 sg= 1 sgfsf face god
‘I cried in front of god.’
I also tried to look for instances in which the body part term could potentially
be deemed to be a grammatical preposition, and found none. In all cases, the use of
lao in constructions such as (115) and (116) is constrained to situations where the
object of the relational noun is either directly in face to face contact or
metaphorically in face-to-face contact. The term lao and all other body part terms do
not have abstract dative like uses.
Body part terms can also quite frequently become parts of noun-noun
collocations which have fixed meanings, as in the following.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
291
(117) b-e-z=e’ zgh-ed=e’ kapiya chi=e’ da
comp-freq-go=3 f comp.and-arrive=3 f chapel of=3f clinan
zoa dxoa bla’o
stat. stand mouth zapote
‘She left and went to her chapel which is at Zapotesmouth (a location in the
village).’
Finally, I examined the incorporation of body part terms into verbs, quite
common both within Otomanguean and crosslinguistically.
Some of these uses are relatively transparent, as in (118).
(118) to bi kunadchi=a’ 0-zhiaha-lao=be’
one clsm inlaw of=lsg pot-go-eye=3inf
‘One of my in laws went in front.’
Note the repetition of lao in (119) used below as a means of specifying location.
(119) 0-zhia-lao meka=n’ lao=a’
pot-go-face Mika=det eye=l sg
‘Mika went in front of me.’
Other transparent uses are as in the following.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
20?
(120) kabi 0-sue=dxo 0-za-nia=dxo
neg pot-handle=lplincl pot-walk-foot=lplincl
0-shinh=dxo tlaclul=le
pot-arrive=lplincl Tlacolula=dir
‘We aren’t going to handle walking to get to Tlacolula.’
Note that the meaning of these incorporated roots can become quite opaque
as in ( 121), which also exemplifies the use of lao in phase verbs.
( 121) na’ ze-lao g-lez=en
demdist stand-eye comp-standup=3inan
‘There it stopped (standing).’
I noted each use of the body part terms I have been discussing in a corpus o f
over 2000 clauses and classified their use according to the criteria I have just
discussed in an attempt to see the degree to which each individual item had been
lexified and to see as a group what use they had. I have included Iho here and above
as a potential point o f comparison, even if synchronically it is not a body part term.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
295
Table 7.3 Textual usage of body part terms
bp
P
loc rel prep n=bp v=bp Total
nia 9 2 11
yichgh 3 3 8 14
dxoalao 7 7
kwit 7 7
lee 2 2 3 4 11
kwitlee 1 9 10
dxoa 2 7 3 1 13
zxan 2 2 16 10 30
kuzhe 5 10 8 1 24
lao 2 6 20 24 35 87
lho 2 1 5 26 2 36
Total 35 31 79 65 0 3 37 250
In the chart above, bp=body part, p=part of object, loc=locative in relation to
an object, rel=more generalized relational noun, prep=prepositional with
grammatical uses, n+bp =nounbody part compound, and v+bp - verb body part
compound.
As seen in the chart above, and, as I will show later, confirming other studies
of body part terms in Zapotecan languages, lao is used by far the most of any of the
body part terms currently in use and should be considered to be the most
grammaticalized body part term. Furthermore, it is used less referring to the body
part and more referring to a location. It is by far the most grammaticalized of any of
the body part terms. Next was Iho which is no longer a body part term, followed by
zxan and kuzhe. Note that these are the next most grammaticalized members of this
set, and their predominance in this study is probably a result of this. It would seem
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 9 4
that it is much rarer to talk of body parts than it is to talk of locative relations.
Furthermore, it would appear, however, that nia and dxoalao were among the least
grammaticalized elements of this set, followed by yichgh, kwitlee, dxoa, and finally
lee. With respect to their uses, it is hard to claim any real generalizations. It was very
difficult differentiating between locations and relations. This is something which
needs to be cleared up in the future. The SBZZ body part terms do show the range of
variation which one would expect from a lexical class which was in the process of
being grammaticalized.
7.12.6 Non-body part derived prepositions
As already mentioned there is also a class of non-body part derived
prepositions, both locational such as those very partially listed in ( 122) below (which
also include various Spanish loan words) and purely relational such as the comitalive
marker lenh and the possessive marker che.
( 122) ladgho ‘between’, gadxol ‘in the center o f, galha ’ ‘near, entr ‘between
[Sp.]\fuerle ‘outside of [Sp.]’, trasde ‘behind [Sp.]’, etc
The first task is to distinguish a class of prepositions. This is not a trivial task.
They could be said to be words which introduce a new phrase into a clause (123).
Unfortunately, this is a property which is also shared by body part locatives, some
possessed nouns, and place names as seen in (124). Thus, this is a necessary
condition, but is not sufficient in and of itself to classify the prepositions of the
language. I choose to classify all of these items as prepositions (as opposed to all
other lexical classes with the exception of relational nouns, which will be discussed
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
later) because they occur in a prenominal position (necessary), they crucially express
a relation between one or more noun phrases or a noun phrase and the predicate
(necessary), and their application is not based intrinisically, meaning that they vary
with respect to the positioning of the figure or ground, crucially distinguishing them
from many of the relational nouns we will describe below and also
conjunctions(sufficient). I am excluding the body part terms at this point, even
though they could easily meet many (but not all) of these criteria.
(123)(a) sh-cho=a’
cont-cough=l sg
‘I coughed.’
(b) *sh-cho=: a’ bedo
cont-cough= 1 sg Pedro
*’I coughed Peter.’
(c) sh-cho=a’ galha bedo
cont-cough=lsg near bedo
‘I coughed near Peter.’
(124)(a) sh-cho=a’ lizha=a’/tiend chi=a’
cont-cough= 1 sg house= 1 sg/store o f = 1 sg
‘I coughed in my house/store.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
296
(124cont.) (b) sh-cho=a’ yezh=en
cont-cough= 1 sg to wn=det
‘I coughed in town.’
It is also not necessarily an easy task to distinguish between various classes
of prepositions within the language. The primary grammatical distinction that we can
make is between prepositions which take pronominal clitics (which include many of
the locative prepositions and the two relational prepositions mentioned above) (125)
and prepositions (many of which are Spanish borrowings) which cannot take
pronominal clitics (126).
(125) ladgho=dxo
between=lplincl
‘between us’
(126)(a) trasde neto
behind lplexcl
‘behind us’
(b) *trasde=to
behind= lplexcl
7.12.7 The cognitive development of body part terms in Zoogocho
Zapotec
Before I attempt to discuss the classification of the body part terms discussed
above, I would like to point out how they correspond to the conceptual continuum set
forth above for Swahili. If we were to consider similar examples in Zoogocho
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 9 7
Zapotec, we would see the following. Once again note that we are discussing the
cognitive stages here, not the categorial stages.
(127) STAGE CONCEPTUAL DOMAIN
0 Body part of X OBJECT
1 Subpart of X, spatially defined OBJECT/SPACE
II Space as part of and adjacent to X SPACE/OBJECT
III Space Adjacent to X SPACE (Heine et al. 130)
(128) Stage 0:
(a) nadxen 0-sala yichgh=to ka’
afterwards pot-throwback head-lplexcl dcmadv
‘Afterwards we will throw our heads back like this.’
Stage I:
(b) za b-zu=e’ azulejo yichgh=en na’
just comp-pit=3f tile head=3inan demdist
‘He had just put tiles on the roof (its head).’
Stage II:
(c) nak g-on=to y-e-dxogb=to yichgh-e’
how pot-do= lplexcl pot-freq-exit= lplexcl head“ 3f
‘How are we going to do it so that we leave there by his head?'
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
298
(128 cont.) Stage IV:
(d) to gonh pintw zoa yichgh=en
one bull spotted stand head=3inan
‘One spotted bull is standing there in front of it.’ 9
Before I go on to classify the body part terms, it will be useful to first discuss
the metaphorical mappings which occur. To begin with, the basic model is indeed
that of the human body, as mentioned by MacLaury above. Novel body part terms
based on terms for an animal’s body are for the most part resoundingly rejected by
native speakers1 0 . Objects described in relation to animals can use the canonical
orientation of the animal as a model for the description. For example, the word kuzhe
can refer to either a human back or to an animal back. As humans canonically are
upright and their back is behind their point of view, the generalized use of kuzhe in
describing spatial relations refers to things which are behind other things (with in
front of and behind defined either in reference to the speaker or to a reference point
like the door of a house). In contrast, an animal is canonically on all fours and as
such its back is upwards. Thus things that are on an animal’s back will be described
as being kuzhe=ba ’ ‘on its back’ as in example (129). Note first that this is an
instance of stage I above, and furthermore that one would not be able to say this in
any other way using the body part terms. (All of the other terms would potentially be
ambiguous.)
na’
demdist
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
299
(129) to=ba’ dxi kuzhe=ba=n’
one=3an stat.sit back=3an=det
‘One of them is on the other’s back.’
The human model is predominant and is used in most instances which describe
animals, as in the following.
(130) na’ pshina’ yixe=n’ zeghe=ba’ dx-zxlonhgh-ba’
and deer wild=det comp.go=3an cont-run=3an
kuzhe=ba’
back=3an
‘ and the deer went running behind it(a dog).’
It is also important to note that the model of the human body, when
metaphorically mapped to objects in this construction is often mapped in an
incomplete fashion. When I was first attempting to learn about the use of relational
nouns, I asked one of my teachers if I could describe some crows on top of a tree as
being yichgh yag ‘head of tree’, and was told that that would be impossible given
that ‘trees do not have heads’. As it turns out, a large part of the problem was with
the choice of tree. It was a pine tree and as such one would describe those crows as
being zxiine yag ‘nose of tree’ or punt che yag ‘point of the tree’. One could
potentially say yichgh yag if they were above the tree, but that was still deemed to be
odd and it would be better to say lao yag ‘eye tree’ or one of the two discussed
above.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 0 0
In a similar vein, one can see body part terms which are used not normally
used in locative constructions being used in locative constructions for individual
items. As an example of one such extension, take the word for ‘nose’, zxiin= just
discussed in the previous paragraph. In San Bartolome Zoogocho, many of the
traditional sandals or yelh come to a point in the front, like a pair of cowboy boots.
When one is describing that part one can use the phrase zxiin yelh ‘nose of sandal’.
One can then use that term to describe something which is on top of that area or
directly in front of it. Now, interestingly, when one puts a sandal up on its nose
(holding it there, of course), and something is below the nose, one can still say that it
is zxiin yelh. Similarly, if the sandal is placed on its back, and something is either
suspended above it or put on the tip of the shoe one can still say zxiin yelh." Note
that this term goes through all of the stages which Heine defines.
(131) STAGE CONCEPTUAL DOMAIN
0: Body part of X OBJECT
(a) dx-ak=da’ zxiin=a’
cont-feel=lsg nose=lsg
‘My nose hurts.’
I: Subpart of X, spatially defined OBJECT/SPACE
(b) puntiagud n-ak zxiin yelh
pointy cont-be nose sandal
‘The point of the sandal is pointy.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 0 1
(131 cont.) II: Space as part of and adjacent to X SPACE/OBJECT
(c) dxi to caj yes zxiin yelh
stat.sit one box cigarettes nose sandal
‘ A pack of sandals is on/in front of the sandals."
Ill: Space Adjacent to X SPACE (Heine etal. 130)
(d) zehe to yishe zxiin yelh
stat.hang one paper nose sandal
‘A piece of paper is hanging above the sandal."
This point brings up the final point which I will discuss here with respect to
the conceptual origin and limitations of relational nouns; the issue of canonical
relations or intrinsic reference. Certain items, such as the shoes which we have just
described, or, for example a leaf which has a pronounced curve and a spine, like a
banana leaf, have parts that are typically labeled in a certain way.
For example the banana leaf, lahaga yelha'2 , can be described as having a
front and a back, labeled Ihee lahaga ‘stomach leaf/front of leaf (without spine)’ and
kuzhe lahaga ‘back leaf/back of leaf (with the stem running down the leaf)’. Once it
is so labeled, things described with respect to the leaf will always refer to these parts,
regardless of the orientation of the leaf. If the leaf is placed on top of a package of
cigarettes with the stem-side down, and one asks where the cigarettes are, they will
be described as being kuzhe lahaga. Similarly, if the smooth side is placed down on
top of the cigarettes, the cigarettes will be described as being Ihee lahaga. Once
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 0 2
again, the use of other body part terms is possible, but these are what I was told were
preferable.1 3
7.12.8 The lexical classification of body part terms
I will attempt to discuss ways in which one might go about classifying the
body part terms we saw above. To begin with, note that nouns and prepositions form
a continuum, with nouns as described above on one end of the continuum, and
prepositions as described above on the other end of the continuum. One might
consider this continuum to be something like the following
(132) nouns locational adverbs relational nouns prepositions
Returning now to Comrie’s discussion of defining categories in terms of
prototypes, I will attempt to come up with a chart similar to the one which Heine el
al. proposed above, by examining the definitions of nouns and prepositions which
have been arrived at and then positing them as being on either end of the continuum.
The nominal criteria which will be considered are the following: possession,
cooccurrence with demonstratives, ability to be the sentential subject either with or
without being possessed, ability to be modified by an adjective, and quantification.
The last three are a bit tricky. While they make sense in Heine et al.’s discussion of
Hausa, there are some inherent difficulties in the case of SBZZ. To begin with, when
used as parts of the body, these nouns are always possessed, so whether or not they
are able to be the subject without taking genitival modification is a moot point as
they will always take genitival modification.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
In a different vein, I should note that, when used to describe location, it is
very difficult to come up with contexts where relational nouns can be described with
adjectives. It is questionable whether relational nouns could be modified in this way
at any of the points in their historical, cognitive, and linguistic development. If one
considers the discussion above on metaphor theory, it would be a case where one
does not map all of the elements or potential combinations of a source domain to a
target domain. The test is, however, very useful in determining whether a relational
noun/preposition still can be used nominally, as all of the relational nouns can, by
themselves, be used with adjectives when being used as body parts or parts of an
object.
The prepositional criteria I will consider are: prenominal position, expression
of a relation between one or more noun phrases and each other or the predicate, and
freedom of application. Finally, I will distinguish between prepositions which can
take pronominal clitics and prepositions which cannot.
Before going any further, consider the following chart, in which, using the
criteria we have discussed above, the criterial status of prepositions in Zoogocho
Zapotec is considered. In this chart, we will include as items we are testing:
(133) I. Body part terms used to label parts of the human body
yichgh=a’
head=lsg
‘my head.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(133 cont.) II. Body part terms used to describe objects
zxan yishe
buttocks quern (grinding stone)
‘bottom of the quern’
III. Body part term used to describe part of and space adjacent to an object.
nak g-on=to y-e-dxogh=to yichgh-e’
how pot-do= lplexcl pot-freq-exit= lplexcl h e a d ? f
‘How are we going to do it so that we leave there by his head?'
IV. Locations
lizh=a’
poss.house=lsg
‘my home’
V. Body part terms used to describe locative relations
zxan mes=en’
buttocks table=det
‘below the table’
VI. Locative prepositions not related to body part terms
ladgho nia=be’
between foot=3inf
‘between his feet’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
VII. Relational prepositions
lenh=a’
with=lsg
‘with me’
I will be testing the following criteria.
Property A, a nominal property, is whether the item can be the sentential
subject.
A. Sentential subject1 4
(134) (a) nala nia=be’
stat.hang foot=3inf
‘His feet hung.’
(b) *nala lenh=a’
stat.hang with= 1 sg
*’With him hung.’1 5
Property B, also a nominal property, has to do with whether an item is able to
be modified by an adjective.
(135) B. Adjective
(a) yichgh zxen chi=a’
head big of=lsg
‘my big head’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(135 cont.) (b) *gadxol zxen nia=be’
between big foot=3inf
Property C, yet another nominal property, deals with whether a
particular lexical item can be quantified.
(136) C. Quantification
(a) chupe ni=a’
two foot=lsg
‘My two feet’
(b) * chupe gadxol bedo1 6
two between Pedro
The next two criteria which will be tested are actually irrelevant, given the
post nominal syntactic structure of demonstratives and determiners. Unfortunately
one cannot know what is being modified, i.e. whether it is the whole PP or RelNP or
whether it is the noun.
(137) D. Demonstrative
lizh=o’ na’
poss.house=2sg demdist
‘Your house there’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(138) E. Determiner marker
yezh=en’
town=det
‘the town’
The next criterion which will be tested will be whether a pronominal clitic
can attach to the lexical item. Most of the lexical items which are being discussed
will be positive for this test.
(139) F. Pronominal clitics
(a) ladgho=dxo
between=lplincl
‘between us’
(b) trasde neto
behind lplexcl
‘behind us’
(c) *trasde=to
behind= lplexcl
The next criterion I will test, a property of prepositions, will be whether or
not the particular lexical item will be able to be inserted into a sentence whose core
argument structure is already filled.1 7
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(140) G. Freedom in syntax
(a) sh-cho=a’
cont-cough=lsg
‘I coughed.’
(b) *sh-cho=a’ bedo
cont-cough= 1 sg Pedro
*’I coughed Peter.’
(c) sh-cho=a’ galha
cont-cough=lsg near
‘I coughed near Peter.’
(d) sh-cho=a’ lizha=a’
cont-cough= 1 sg house= 1 sg
‘I coughed in my house.’
(e) sh-cho=a’ yezh=en’
cont-cough= 1 sg to wn=det
‘I coughed in town.’
(f) sh-cho=a’ lao=o’
cont-cough= 1 sg eye=2 sg
‘I coughed in front of you.’
bedo
bedo
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 0 4
(140 cont.) (g) b-id=a’
comp-come=lsg
‘I came.’
(h) b-id=a’ zxghozxo
comp-come= 1 sg Zoogocho
‘I came to Zoogocho’
(i) *b-id=a’ bedo
(j) b-id=a’ lao bedo
comp-come=lsg eye pedro
‘I came to Peter.’
The final test which I will apply will be whether the lexical item, in
expressing a relation between two items, expresses an intrinsic relationship as
described above.
(141) I. Intrinsic
(a) kuzhe bedo
back Peter
‘at Peter’s back’ (no matter what orientation Peter has to a
speech act participant)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(141 cont.) (b) trasde bedo
behind Peter
‘behind Peter’ (could vary depending on Peter’s positions with
respect to the speaker or hearer, much like English
prepositions)
I will repeat the labels once again here:
(142) Lexical items tested
I. Body part terms used to label parts of the human body
II. Body part terms used to label parts of objects
III. Body part term used to describe part of and space adjacent to
an object.
IV. Locations
V. Body part terms used to describe locative relations
VI. Locative prepositions not related to body part terms
VII. Relational prepositions
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(143) Nominal and prepositional criteria
A. Sentential subject
B. Adjective
C. Quantification
D. Demonstrative
E. Determiner marker
F. Pronominal clitics
G. Freedom in syntax
H. Intrinsic
Table 7.4 Noun-preposition cline
I II III IV V VI VII
A + + + +/- - - -
B + + + + - - -
C
+ + + +
?
- -
D + + + +
* * *
E + + + +
* * *
F + + + + +
+/-
+
G - - + + + + +
H
* *
+
*
+ - -
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
312
In the previous table, + means that a particular criterion was successfully
tested, a - means that the test was not successful, +/- means that a particular group of
lexical items varies, a * means that a particular test was inapplicable and a ? means
that the information is incomplete.
Note that items which are used to describe a part of/space adjacent to seem to
pattern both like nouns and like prepositions. This is what one would expect for a
transitional lexical class. While it would appear that body part terms used to
describe locative relations pattern more closely with prepositions, it must also be
noted that they also pattern with locations. One possibility, as mentioned above is
that the metaphorical extension, possible before a change of category has taken
place, applies only to a limited portion of the source domain. In general, it seems like
this is a class that is definitely in transition. It might be useful to compare briefly, the
situation in Zoogocho Zapotec with the situation in Valley Zapotec, based on the
very thorough master’s thesis of Brook Lillehaugen.
7.12.9 Comparison with Valley Zapotec
In this section, I will briefly compare our findings in SBZZ with what has
been said about ‘The Categorial Status of Body Part Prepositions in Valley Zapotec’.
a 2003 UCLA Master’s Thesis by Brook Lillehaugen. In this insightful work, Ms.
Lillehaugen comes up with a variety of compelling reasons to consider the body part
terms in Valley Zapotec to be prepositions, and not relational nouns. Her data
compares nicely with what has been discussed so far. Her main reasons for
considering these terms to be prepositions in Valley Zapotec are the following.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 13
(144) a)Their ability to be used with intransitive verbs, where normal NP’s lack that
ability (Lillehaugen 2003:14).
b) The fact that certain verbs (especially positional verbs) require body part
prepositions (and not just any body part term but only those are
grammaticalized) as complements to express the ground (ibid. 17).
c) The fact that body part prepositional phrases enters into coordinate
structures with other prepositional phrases (ibid. 18).
d) The fact that certain uses of body part terms are infelicitous as descriptions
of parts of objects but can be used as prepositions with those same objects
(ibid. 18-19).
e) The fact that no novel body part terms can be extended to become
prepositions (ibid. 20).
f) The fact that when parts of objects are named in Valley Zapotec languages,
these names do not correspond to locative descriptions (ibid. 20-22).
g) The fact that the canonical orientation of an object does not seem lo affecl
locative constructions (ibid. 22-23).
h) The fact that certain structures are structurally ambiguous (ibid. 23-24).
i) The fact that directional verbs require prepositional complements (ibid.
25).
Note that, with respect to her first argument, intransitive verbs in SBZZ can
also take locations, both relational nouns and non-relational nouns. Her second
argument also does not apply to SBZZ locative verbs, as they can either take body
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 14
part nouns (and novel body part terms as well), demonstratives, or can appear bare. 1
have no data on her third argument at this point. Her fourth argument is partially
valid for SBZZ body part terms; however, in those cases where the body part terms
are infelicitous, most speakers do prefer to use other ways of describing the location
without using the infelicitous terms. Her fifth argument is not valid at all for SBZZ.
Novel body part terms can be extended in Zoogocho Zapotec. Her sixth argument is
also not valid for SBZZ body part terms as when they are used to name parts of
objects, this naming can then be extended to locative descriptions. With respect to
her seventh argument, it is definitely the case that in SBZZ the canonical orientation
of an object affects the way in which location is described.
Her ninth argument deserves a bit more explanation. First, consider the data,
which she bases this on.
(145) Naa' ca-cwaa=a' laa'iny yudooo'. (San Juan Guelavia Zapotec)
I PROG-paint=ls in church
'I am painting in the church.'
‘I am painting the inside of the church.’ (Her 16) (ibid. 24)
Examine also the following data from Heine et al. (Heine et al. 135).
(146) me-kpo e-me
lsg.see 3sf.POSS-lN
(i) ‘I saw its interior’
(ii)’I saw inside it.’ (Their 8c)(ibid. 135)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
There is similar evidence from Zoogocho Zapotec as well, as in (147)
(147) sh-na=ba kuzhe=ba
cont-look=3an behind/back=3an
‘Itj is looking at itSj back’
‘Itj is looking behind itj.’
Heine et al. analyze this as being ‘an inherent characteristic of transitional stages in
grammaticalization: when a new structure (i.e. an adverbial morphosyntax in this
example) is introduced, the old structure (a nominal morphosyntax) is generally still
in use, the result being overlapping’ (ibid. 135-136). One might therefore consider
this a similar case. Finally, her tenth argument is not valid for SBZZ body part terms.
As has been seen above, verbs like go and come in SBZZ which encode directional
information do not necessarily require a prepositional complement.
One other argument which Pamela Munro has used for the San Lucas
Quiavini Zapotec word for face, loh, (Munro 2002:23), is that there non-locative
uses, as in the following.
(148) Loh Jwaany b-zi=a-ih. 'I bought it from Juan'
face/from Juan perf-buy=ls=3s.prox (her 52)
(149) B-zhiiu'azh=a'gueht loh bee'cw. 'I tore up the tortilla for the dog'
perf-tear=ls tortilla face/for dog (her 53)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 16
(150) Zyuua'll=ru' Rrodriiegw loh Lia Oliieb
tall=more Rodrigo face/than Ms. Olivia
'Rodrigo is taller than Olivia’ (her 54) (ibid. 23)
Note that in Zoogocho Zapotec, there are no such dative uses. Based on these
comparisons, I feel confident in saying that the corresponding terms in Zoogocho
Zapotec are not prepositions, but are rather something else, something we will label
‘relational nouns’.
7.12.10 Conclusion
In conclusion, I have found that, while they are definitely a separate lexical
class from garden variety nouns, the lexical class I have been calling body part
locatives and will now call relational nouns are also distinct from prepositions in the
language. That they share adverbial morphosyntax is unimportant. I consider the
semantic and cross-linguistic generalizations to have shown a nascent lexical class,
and one which fits in its own well defined point on the noun-preposition continuum,
and shares many qualities with non-body part, non prepositional locatives.
Comparison both within the Zapotec language family and outside the family leads
me to call these terms ‘relational nouns’, being careful to keep in mind that they
form a chain, as described in Heine et al. and are definitely being grammaticalized
on their way towards being prepositions, but have not yet gotten there.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
317
1 Though see Lillehaugen (2003) for a very good example o f a description which does provide a
definition.
2 This would however be potentially o f great use to historical linguists and people who specialize in
grammaticalization for, as we will see later, the status o f an individual lexical item on an individual
cline could potentially help to further our knowledge o f the nature o f linguistic change.
3 1 am using this term as it is used at the Max-Planck Institute for Psycholinguistics in Nijmegen,
Holland.
4 As is quite common crosslinguistically, see the roots o f Spanish estar.
5 1 have described what Diessel refers to as manner demonstratives using Schachter’s term pro
adverbs in the section on proforms below.
6 It is interesting to note that the determiners, discussed above in 7.7 appear to be historically related
to the demonstratives.
7 Though see 6.4 to see some examples o f locative adverbs intervening between verb and subject.
8 1 will classify these forms as being like independent pronouns and therefore in the noun category.
9 Once again note that this is from a task in which speakers were asked to describe scenes to each
other (the man and tree task) and in which there is a bull in front o f a cart. The front o f a cart is its
head, and thus the present description.
1 0 For example some words, such as xkogoba' ‘its neck’ cannot be used in describing objects at all,
whereas other terms such as xbanhba’ ‘its tail’ can only be used in very restrictive environments (in
one case, to describe a single gully coming o ff o f a ravine).
1 1 O f course, one could potentially say zxanyelh ‘under the sandal’ or yichgh yelh ‘above the sandal'
to describe either o f these situations as well.
1 2 1 will refer to it as lahaga ‘le a f and not specify that it is a banana leaf, as it is redundant.
1 3 Note as a point o f comparison, in English, if one is describing an object with respect to a person
who was standing with their side facing the speaker, to say that that object was behind someone could
mean two things: it could mean that it was behind the person with respect to the speech act
participants, or it could mean that it was to the rear o f the person (at the person’s back). In Zoogocho
Zapotec, one would preferably say kuzhe ‘behind the person’ preferably for those instances where the
item was at the person’s back, as it does not have the generality which English behind has. There has
been a great deal o f work done on absolute versus relative systems for spatial descriptions at the Max
Planck Institute for Psycholinguistics in Nijmegen and elsewhere. (See Levinson (1996) among
others.)
1 4 Note that, among the word types are being tested, all o f them with the exception o f some o f the
locations and some o f the non-body part prepositions must be possessed in order to use them as a well
formed complement.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
318
1 5 Interestingly, although the possessive preposition by itself cannot act as the sentential subject, if it
is definite it can. *nala chia ’ (ok) nala chi=a ’=n ’ ‘mine hung’
1 6 Note that I have collected data like the following:
(A) lho chupe shaa dao
inside two casseroles dim
‘inside two little casseroles’
(B) chupe lho shaa dao
two inside casseroles dim
‘Inside two little casseroles’
This is for the most grammatical ized o f the putative relational nouns. Also, note that this does not
mean ‘the two insides o f the little casseroles’. Otherwise, constructions like this are unheard of.
However, I will be forced to put a ? in the box corresponding to relational nouns for this criterion until
I have more data.
1 7 For example, an intransitive verb for which there is already a subject.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 1 9
Chapter Eight: A performance approach to VSO order in
Zoogocho Zapotec
8.1 Introduction: Basic word order properties
Before delving into the more detailed typologies I will present in the next
section, I will begin by reviewing the basic word order universal tendencies which
were proposed by Joseph Greenberg (1963) and which were presented above in 5.7.
To begin with, SBZZ is a type I language (VSO) as per Greenberg’s typology. It is a
VSO language as seen in (1):
Verb Subject Object
( 1) gud-ap delia xa=be’
comp-slap Cordelia poss.father=3inf
‘Cordelia slapped her father.’
It has prepositions/prenominal relational nouns, as seen in (2).
Preposition Noun
(2) yichgh be’ko’=n’
head dog=det
‘above the dog’
The ordering of noun and adjective is noun followed by adjective as in (3).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Noun Adjective
(3) yoo zxen
house big
‘big house’
The order of noun and genitive is noun genitive as in (4).
Noun Genitive
(4) lizh Franco
poss.house Franco
‘Franco’s house’
Relative clauses follow the noun which they modify.
Noun Relative clause
(5) na b-esey-eyu=e’ metal=en’ danh g-os-of=e'
and comp-pl-carry=3f metal=det geninan com p-pl-grind3f
‘and they carried the metal they had ground...’
The order in comparative sentences is adjective marker standard as in (6).
Adj-Mkr-Std
(6) n-ak-dx Maria benhe zxen ka’ zxoan
cont-make-more Maria person large demadv Juan
‘Maria is bigger than Juan.’
Auxiliaries precede the main verb as below.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(7) bi dx-eyalha shegh=dxo=x
neg cont-should stat.go=lplincl=well
‘Well, we shouldn’t really go.’
Contrary to what one might expect and to what Greenberg predicted, there is
no sentence initial question particle for polar questions, the only particle involved in
yes/no questions is the tag formative ke as in the following1 .
(8) na zghe-zhia-dxgwa btushe nahago=ba, ke?
and stat.pl-stand-emph pointy ear=3an, right?
‘And their ears are very pointy, right?’
Interrogative words do appear sentence initially as in (9);
(9) balhe benhe gringo na b-ese-laak-s=e’
how-many people gringo demdist comp-pl-arrive-emph=3f
‘How many gringos came?’
As seen in the morphology chapter, most of the inflectional and derivational
morphology is in the form of prefixes, although there are some suffixes as well.
Finally, as will be seen in much greater depth over the course of this chapter,
Greenberg’s universal # 6 is entirely applicable to SBZZ.
Universal #6: All VSO languages have SVO as an alternate order.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 2 2
(10) s V o
delia gu-dap xa=be’
Cordelia comp-slap poss.father=3inf
‘Cordelia slapped her father.’
(11) S V=pro O
delia gu-dap=be’ xa=be’
Cordelia comp-slap=3inf poss.father =3inf
‘Cordelia slapped her father.’ (I will translate both of these sentences with the
same English gloss for the time being, and come back to them with individual
translations which better indicate their meanings later in the current chapter.)
In the remainder of the chapter, I will investigate various claims which have
been made in much greater detail about verb-initial languages and then present a
corpus study I have conducted on SBZZ. I will investigate claims which have been
made from a variety of functional and structural perspectives, although I will limit
myself to claims which have been made within typological universal grammar. This
exploration will begin with an investigation of structural statements which have been
made about verb-initial languages, will continue with an exploration of
discourse/pragmatic claims which have been made, and will pause briefly at a
potential processing explanation for VSO and alternate structures before reaching the
main destination, an examination of 1942 clauses in two SBZZ texts with
comparison of other textual investigations of verb-initial languages.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 2 3
8.2 Structural approaches to the typology of verb initial languages
In this section, I will discuss three different approaches to the typology of
verb initial languages which have been expounded in the literature. I will begin with
a discussion of Doris Payne’s 1990 restatement of Keenan’s (1977) and (1979)
characterization of verb initial languages2 , and will discuss the extent to which SBZZ
conforms to these characterizations. I will then briefly discuss the reasoning which
Matthew Dryer uses in collapsing VSO and VOS into one type (his VS and VO type)
(Dryer 1996). Finally, I will end this section with an in-depth discussion of claims
made by Maria Polinsky (1997) about verb initial languages.
8.2.1 Payne’s restatement of Keenan
I will begin this section, by exploring how Zoogocho Zapotec conforms to
the Verb Initial Norm which Doris Payne describes3 in The Pragmatics of Word
Order-Typological Dimensions of Verb Initial Languages (1990). Payne bases this
on ‘a number of observations extracted from Keenan’s (1977) ‘Summary of word
order typologies’ and from his 1979a manuscript on ‘Word order typologies: the
verb initial typology’ (Payne 1990: 10). I will discuss briefly how Zoogocho Zapotec
corresponds to the claims made in this typology as I go along.
1. General. Verb initial languages are largely, though not entirely, the
mirror image o f verb final languages.
2. Morphology Verb initial languages evidence significant prefixing, though
normally there is some suffixing as well. There is a possibility o f arnbi-
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 2 4
fixing (discontinuous affixes), and a somewhat greater than chance
tendency fo r discontinuous demonstratives.
Zoogocho Zapotec does have both ‘significant’ prefixing and ‘some’
suffixing, especially if one takes the post-verbal pronominal clitics as evidence of
suffixing. I see no evidence of ambifixing, and no discontinuous demonstratives.
2.1. Verb initial languages may be agglutinative and polysynthetic.
While SBZZ is agglutinative, it is definitely not polysynthetic. See 4.5 for
discussion of this issue.
3. Basic word order
3.1. Verb initial languages are comprised o f the following types:
[1] verb initial plus free order o f full N P ’ s (Tagalog)
[2] V-DO-S-Obl (Fijian, Toba Batak)
[3] V-Do-Obl-S (Malagasy, Tzeltal)
[4] V-S-DO-Obl (Celtic, Eastern Nilotic, Polynesian, Jacaltec)
Type [4] is by fa r the most common.
SBZZ is firmly a type 4 language if we only consider simple transitive
clauses. The most common ordering of verb, subject, and object is as seen below in
(12). I will demonstrate later that although this is by far the most common ordering,
there is a great deal of variation, and, amazingly, every possible ordering with the
exceptions of VOS and OSV is encountered.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(12) dxy-izxgho benhe gringo=na’ dxioo
cont-pay person gringo=det lplincl
‘The gringos paid us.’
The following examples of variation with respect to direct object (patient) and
indirect object (recipient) order are found.
(13) b-i=a’ bidao to libr
comp-give-lsg child one book
‘I gave the child a book.’
(14) b-i=a’ to libr bidao
comp-give-lsg one book child
‘I gave a book to the child.’
Note that the ordering for this particular verb4 is verb subject indirect object
direct object when the nouns and objects are pronominalized.
(15) VSIODO
b-i-a-be-n
comp-gi ve-1 sg-3 inf-3 inan
‘I gave him it.’
On the basis of simple transitive sentences, SBZZ should be considered to be
Type 4. On the basis of ditransitives, it is not as clear.
3.2. Freedom. Fronting o f subject NPs to the left o f the verb is always a
possibility, though often it is morphologically marked in some way (not
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
necessarily on the NP). The order after the verb is frequently rigid, though
sometimes quite free as in Tagalog and, to a lesser extent, in Chinook.
Subject NPs may be fronted in SBZZ, however there is often greater
morphological marking, with a coreferential pronominal clitic on the verb, as in (16).
There are also other fronting constructions, which lack this coreferential pronominal
clitic (as in (17) and (18)). I will discuss the pragmatics of these constructions later
in the current chapter. The order after the verb is quite fixed, with the possible
exception of the reflexive of possessor constructions, as has been discussed in
chapter 4.
(16) na da dolor=en dx-e=e=ne’
and deceased dolores=det cont-say=3f=3fo
‘And the late Dolores said to him....’
(17) to=be’ zegh
one=3f stat.go
‘One of them went.’
(18) pelot-en’ dxy-itghe-d=e’
ball-det cont-play-instr=3f
‘She played with the ball.’
4. Sentence level syntax
4.1. Topicalization. Topicalization may be done by fronting, though there is
a tendency in Nilotic to move old information to the end o f the clause
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 2 7
Topicalization is done by fronting as in (16). I will discuss the definitions for
focus and topic in SBZZ towards the end of the current chapter, and for now will use
the term topic to refer to those constructions where there is a repeated pronominal
clitic after the verb and focus to those constructions where there is no repeated
pronominal clitic after the verb.
4.2. Focussing. Focussing o f information as in a cleft or information
question is done by fronting. Often this may be accompanied by
particles separating the subject from the rest o f the clause. The residt is
always pragmatically marked, i.e. emphatic, contrastive, focussed, etc.
Focussing is done by fronting. What is meant by ‘particles separating the
subject from the rest of the clause’ is unclear, however I can say that focuss.ed nouns
are often marked by quantifiers or definite markers as in (17).
4.3. Comparisons. The comparative form precedes the standard. The
comparative marker is commonly a verbal form, or else an adposition.
Thus, ’ John is taller than Bill ’ may be expressed as ‘Tall John from
Bill or as ’ Tall John exceed Bill ’ .
In SBZZ, comparatives are formed with an adverb (the ka ’ in (18) or the
kleka ’ in (19)) both of which are accompanied by the adverbial suffix -dx- on the
verb as seen in the following examples.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(18) n-ak-dx Maria benhe zxen ka’ zxoan
stat-make-more Maria person large demadv Juan
‘Maria is bigger than Juan.’
(19) Bdxe n-ak-dx=ba’ be lis kleka’ be'ko'
Ant stat-be-more=3an clan small than dog
‘Ants are smaller than dogs.’
4.4. Questions
4.4.1. In yes-no questions, the question particle, if any, occurs
sentence initially.
As seen above (in (8)), the tag question formative is sentence final. There is
no question particle.
4.4.2. In NP questions, a questioned NP is always frontable and this
is the normal pattern. It is possible, but less normal, to leave
the questioned NP in the position questioned. A few cases o f
rightward movement o f question words are attested, but there
is no attested tendency fo r the question word to attract to the
normal DO position (as is the case fo r verb final languages).
Questioned NPs are fronted, as seen in (20) and (21).
(20) no b-le’i=do’
who comp-see=2sgexp
‘Who did you see?’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(21) balhe benhe gringo=na’ b-ese-laak-s=e’
how many people gringo=det comp-pl-arrive-emph=3f
‘How many gringos came?’
4.5. Subordinate clauses and sentence complements
4.5.1 It is very common fo r many types o f subordinate clauses to be finite.
As I will show below, subordinate clauses are generally finite.
4.5.2. Subordinating markers such as complementizers, nominalizers, and
subordinate conjunctions precede their clauses.
As seen in (22), complementizers precede their clauses.
(22) zxenlazh=a’ leka=n’ dx-on=dxo kate’
stat.get_used_to=lsg much=det cont-do=lplincl when
n-ak=dxo bidao
stat-be=lplincl child
‘I got used to it as we do when we are children’
4.5.3. Sentences which are subordinate to verbs, adjectives, or nouns
invariably follow the element to which they are subordinate.
Subordinate clauses follow their head in SBZZ as seen in (22).
4.5.4. Adverbial subordinate clauses usually follow their main clauses. For
example ‘Will leave John because is tired M ary’ occurs fo r ‘ John will leave
because Mary is tired’ . However, frontability o f conditionals is likely
universal (cf. Greenberg 1963)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Adverbial subordinate clauses do indeed usually follow their main clauses, as
seen above. Conditionals are indeed optionally fronted, as in (24).
(24) Shi ba b-zhinh=dxo na b-le’id=a’
if already comp-arrive=lplincl dem comp-see=lsg
‘If we’ve already gotten there, then I’ll see.’
4.6. Coordinate sentences are commonly expressed as [S and SJ. [S, S and]
is not attested. Perhaps the existence o f overt coordinate conjunctions at
the S level, especially or, is less well attested than in verb medial
languages.
While coordinate sentences are very often expressed as [S and S] as seen in
(25), they can also be expressed without an overt coordinate conjunction. As I have
shown in 6.5.8, this potentially affects coreference possibilities. The structure [S, S
and] does not occur. The Spanish loan word o ‘or’ is most commonly used for or.
(25) dx-eene zxoan noole=n’ na zegh=e’
cont-want Juan woman=det and stat.go=3f
lawe yaa
face plaza
‘Juan wants the woman and she went to the market’
4.7. Speech act indicators (e.g. question particles, etc.) are normally
sentence initial, though other positions are possible.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
As already seen, this is the case for interrogatives (cf. (20)), but not for the
yes/no question particle (cf. (8)).
5. The noun phrase
5.1. Case marking
5.1.1. All major NPs may be case marked (Tongan, Nandi), but it is very
common fo r most major NPs to carry little or no nominal case marking.
Where affixal case marking occurs, it is more likely to be prefixal than in
verb final languages, but suffixing is still fairly common.
There is no case marking o f full NPs in Zoogocho Zapotec. For the third
person singular respectful forms of the enclitic pronoun, there is a distinction
between -e ’ for 3f subjects as seen in (25) and -ne ’ for 3f objects as seen in (26).
(26) dx-ap=a=ne’
cont-say=lsg=3f
‘I told her.’
This difference has been discussed in 3.6.
5.1.2. Where case marking exists it is normally done by prepositions (though
some Amerindian languages are the exceptions here, such as Machiguenga
and Quileute, which have postpositions).
Instrumentals can be marked by a preposition, and benefactives and indirect
objects can be marked by possession, as discussed in 6.2 and 6.3. However, oblique5
case is neither marked prepositionally nor with case marking.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(27) sh-chog=a’ yag-en lenh yaa wag
cont-cut-lsg tree-det with iron wood
‘I cut the tree with an ax.’
(28) b-en=a’ lizh=o’
comp-make=lsg poss.house=2sg
‘I made a house for you’
5.1.3. Verbal case marking is attested to a very significant degree. That is,
verbs carry affixes indicating that an instrumental, goal, locative, benefactee,
etc. is present and the corresponding fu ll NP ’ s carry no adpositions or
distinctive case marking.
In (29), there is evidence of an instrumental affix in SBZZ. However, note
that this is the only such potential verbal case marking present in the language.
(29) b-en=d lizh=a’ martiyw
comp-make=instr poss.house=lsg hammer
‘I built my house with a hammer.’
One might also consider the experiencer set of pronominal enclitics to serve a
similar function as in (30).
(30) b-lei=da’ to be’ko’
comp-see= 1 sgexp one dog
‘ I saw a dog.’
See 3.4 for more detail.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
5.1.4. As with verb final languages, but in distinction to verb medial
languages, case marking (and verb agreement) may follow an ergative
pattern.
This statement is inapplicable to Zoogocho Zapotec. As briefly discussed in
5.3 and 6.5.8 and as will be discussed below, Zoogocho Zapotec is most definitely
not an ergative language.
5.2. Adjectives
5.2.1. The demonstrative, numeral, and qualifying adjective follow the
common noun in that order or its mirror image (Adj=Num=Dem).
Zoogocho Zapotec is an exception to this pattern in that all three of these
modifers do not follow the verb. In SBZZ, the ordering is Num N Adj Dem as seen
in (31).
(31) Num N Adj Dem
shon libr xo nga
three books old demmed
‘These three old books’
5.2.2. There is probably less agreement with common nouns than in verb
final languages, especially case agreement.
There is no nominal agreement in SBZZ.
5.2.3. Adverbs follow adjectives (but this needs further checking).
Adverbs appear to precede the adjective they modify, as seen in (32) below.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(32) leka fwert n-ak kafe=n’
much strong stat-be coffee=det
‘The coffee is very strong.’
5.3. Articles
5.3.1. The presence o f definite articles distinct from demonstratives is much
more common than in verb final languages.
As already seen, demonstratives are distinct from definite articles (compare
(31) and (32)).
5.3.2. The existence o f several articles (definite, indefinite, specific, plural,
proper noun) is much more common than in verb final languages (e.g. Maori.
Fijian).
While there is a definite article, and an indefinite article, this appears to be
the extent of it.
5.4. Possessors: with great regularity Possessor NPs follow the head NP, as
in father o f John rather than John’ s father.
This is the case for both alienable and inalienable possession as seen in (33) and (34)
respectively.
(33) be’ko’ che delia
dog of Cordelia
‘Cordelia’s dog’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(34) nahag lalo
ear lalo
‘Lalo’s ear’
5.5. Relative clauses
5.5.1 The dominant order is always postnominal.
Relative clauses are postnominal as in (35).
(35) na b-esey-eyu=e’ metal danh g-os-ot^e’
and comp-pl-carry=3f metal geninan comp-pl-grindr 3 f
‘They took the metal which they ground.’
5.5.2. Occurrence o f personal pronouns in positions relativized is fairly
common, though relativization by deletion is still the most common strategy’
While deletion is the standard practice, personal pronouns can appear in then
position relativized, if they disambiguate the clause. See 6.5.7 for a discussion of this
issue. Note that this is a general strategy among VI languages, especially those that
do allow free WO after the verb. (See Thomas Payne (1997) for more discussion.)
5.5.3. In distinction to verb final languages, co-relatives are not attested.
This is true of SBZZ.
5.5.4. Like verb final languages, but in distinction to verb medial languages,
relative pronouns which code the case o f the position relativized are rare. It
is less rare than in verb final languages, however (e.g. Tamazight, Berber).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Relative pronouns do not code the case of the position relativized; see 6.5.7
for more information.
5.5.5. Relative pronouns which agree with the head noun in noun class and
sometimes even case are attested (e.g. Classical Arabic, Nandi).
As seen in Chapters Five and Seven, relative pronouns agree with the head noun in
noun class. See section 7.4.4.
5.5.6. In distinction to verb final languages, internally headed relatives are
not attested, though the phenomenon is not well studied.
Internally headed relatives are not attested.
6. The verb phrase
6.1. Tense/aspect, passive, inchoatives, causatives, negation, modals,
disideratives, and volitionals may appear marked on the verb. There is
significantly more prefixing in verb initial languages than in verb final ones,
and very possibly more ambifixing and infixing. There is, to K eenan’ s
knowledge, always some suffixing, however.
Aspect and causatives appear marked on the verb. There is also more prefixing than
the some suffixing which is present. See the 4.5 for confirmation of this statement.
6.2. I f expressed by morphemically independent forms, modals, auxiliaries (if
such exist), negative particles or words, disideratives and volitionals always
precede the main verb and may themselves have independent verbal
morphology. (This may also be true fo r tense/aspect, passive, inchoatives.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
and causatives.) The strength o f the order correlation here is better than its
converse fo r verb final languages.
This is most definitely true. Consider the following in which an auxiliary,
inflected for aspect but with no marker for the subject, precedes a main verb.
(36) dx-ak dx-le’i=da’
cont-be cont-see= 1 sgexp
‘I can see.’
See 6.5.4 for a fuller discussion of auxiliaries.
6.3. Manner adverbs follow the verb if they are a distinct category (which
often they are not).
As seen in the following, manner adverbs do indeed follow the verb.
(37) b-en=a’ sholazhe chup lizh=a'
comp-make=lsg slowly two poss.house=Tsg
‘I slowly built my two houses.’
6.4. Sentential objects always follow the subject and are very commonly finite
as opposed to the more usual non-fmite/nominalized treatment they receive in
verb final languages.
As seen in 6.5.5 this is indeed the case.
6.5. Sentential objects are never embedded. They normally follow the main
sentence but may precede, especially in direct quote contexts.
Sentential objects follow the main sentence, as in (38).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(38) dx-een=da’ dx-aog=a’ to yet
cont-want=lsgexp cont-eat=lsg one tortilla
‘I want to eat a tortilla.’
Direct quotes generally precede the verb of speaking, as seen in (39). It is possible,
however, for direct quotes to follow as in (40).
(39) “na y-e-yo=a=n” dx-ap=a=ne’
demdist pot=freq=carry=lsg=3inan cont-tell=lsg=3fo
‘ “I’ll carry it there”, I told him.’
(40) per dx-ee-te xna=a’ le “sha ka
but cont-say-emphposs.mother=lsg 3sgf “cond demadv
shegh=o’”
stat.go=2sg”
‘But my mom said to her, “You shouldn’t go like that.’”
6.6. Verbal forms subordinate to the ‘ main ’ verb (e.g. complements o f verbs
like want, try, etc.) always follow the main verb, and are commonly finite.
This is true, as seen below in (41).
(41) dx-eene=ba’ xob=ba’ bishede=n’
cont-want=3an pot.want=3an beehive=det
‘It wants to grab the beehive.’
6.7 . Causativized verbs follow the causativizing verb.
Causativized verbs do follow the causativizing verb, as seen below.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(42) dx-on=a’ sh-le’i bidao
cont-make=lsg cont-see child
‘I made the child see.’
6.8. ‘ Backward’ equi-deletion may occur. That is, ’ want John go ’ or ‘ want go
John ’ may occur fo r ‘ John wants to go This is never a possibility in verb
final languages.
‘Backward’ equi-deletion does indeed occur, as seen below. See 6.5.4 for a fuller
discussion of this and related constructions.
(43) dx-ak dx=aog x-kuzh=o’
cont-can cont=eat poss-pig=2sg
‘Your pig can eat.’
6.9. There is possibly less rich means fo r nominalizing and definitizing verb
phrases than in verb final languages. On the other hand, in many but not all
verb initial languages, the verbal complex seems historically to be a nominal
construction, at least in part (Middle Egyptian, Welsh, Malagasy, Philippine
languages, Mayan)
I would not say that there are incredibly ‘rich means for nominalizing’ verb
phrases, yet would say that there are some, though the following is a deverbal noun
and not an action nominalization. See 4.3.8 for more information.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(44) gubanh
inf.rob
‘thief
(45) dx-banh=e’ zxikw=a’
comp-rob=3f poss.dog=lsg
‘He’s stealing my dog!’
At this juncture, I do not have much to contribute to Payne’s second
statement, except to state that, while it would be interesting to study the historical
development of the verbal complex in Zapotecan and Otomanguean languages, I do
not know enough about its sources to comment at this point.
6.10. Verbal initial languages always have a passive voice and it is almost
always marked in the verbal morphology (rather than by a serial verb
construction as in Chinese, fo r example). It may be marked by a verb plus
nominalization as in ‘ John receive hitting from B ill’ (Tzeltal, Mayan).
As discussed in 4.3.9, there is no productive passive voice.
6.11. With possibly greater than chance frequency, the verb in verb initial
languages either agrees with no NPs, or with two NPs (both subject and
direct object, or sometimes subject and indirect object).
While I will return to this issue in greater depth later in this current chapter
and have already discussed it in 3.8 and 4 .5 ,1 would say that, if the pronominal
clitics are seen to be a form of agreement, then there is indeed agreement with the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
341
subject. However, as already discussed, it is questionable if it is desirable to consider
it to be agreement on par with affixal agreement.
6.12. Verb initial languages normally have no overt copula.
There is indeed an overt equational copula However, there are also other
ways of expressing copular relations, as discussed in 7.2.
8.2.2 Dryer
I will now briefly summarize and discuss the arguments which Dryer has
made for rejecting the traditional six-way word order typology. Within this paper,
Dryer has also proposed collapsing VSO and VOS languages into a VS&VO type.
The first argument which he presents for the desirability of collapsing these two
types is that it provides for an easier way of classifying the many languages o f the
world which, while verb initial, possess relatively free word order after the verb
(Dryer 1996: 74-75). His second argument is that VSO and VOS languages pattern
quite similarly when one takes into account the general word order properties which
have been claimed for VSO languages since Greenberg’s original paper (ibid. 75-
76). His third argument is that if one looks within individual genetic groupings
which show propensities for being verb initial (such as Otomanguean), there are both
languages which are VSO (such as Zapotec) and languages which are VOS (such as
certain varieties of Otomi such as Estado de Mexico Otomi (Grimes 2000)), which
he takes to indicate that the possibility of changing from one language type to the
other is relatively easy, further showing that the difference between these two types
is rather insignificant (ibid. 76-77).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 4 2
This last point appears to be valid on some very general level, but it seems
that one could make a similar argument for VSO and SVO languages and language
families which have both VSO and SVO daughter languages. Also, there are very
few Otomanguean languages which are VOS and many Otomanguean languages
which are VSO, like SBZZ, are relatively strictly VSO. As shown by Polinsky
(1997), there are certain ways in which VOS languages differ from VSO languages
as well. His fourth argument, one which I will examine in greater depth later in this
chapter, is that clauses which contain both a noun subject and a noun object are very
rare. Thus, it is much more common to find either VO structures or VS structures in
most languages which are characterized as being either VSO or VOS, but it is very
rare to find clauses which contain both subject and object as full nouns.
Dryer goes on to discuss other reasons for his proposed reformulation of the
word order typology, most important for the purpose of this chapter being the
differences between transitive and intransitive subjects in terms of their word order
properties. He claims that there is quite frequently variation about the positioning of
verb and subject dependent on the transitivity of the verb (ibid. 87-89). 1 will
investigate the validity of this claim for Zoogocho Zapotec when discussing the
textual data. He also notes, as noted by Du Bois, that transitive clauses are more
likely to have pronominal subjects than intransitive clauses (ibid. 90). This too will
be discussed. The importance of his typology is that it will allow for more exact
classifications of individual languages. One language which he cites as an example is
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 4 3
Salinan which is SYO for clauses with two lexical arguments but predominantly VO
and VS for languages with only one lexical argument (ibid. 93).
8.2.3 Polinsky
As the final approach to verb initial typologies which will be examined in the
current section, Polinsky (1997) differentiates four basic word order types:
VSO/*VOS
VOS/*VSO
VSO/VOS
VOS/VSO (ibid. 254)
The last two types are differentiated by what the default interpretation is of
what would be potentially ambiguous clauses in languages which allow both VSO
and VOS orderings like:
(46) kicked George Harry
VSO/VOS languages preferentially interpret George as being the subject and
VOS/VSO languages preferentially interpret Harry as being the subject.
She then goes on to note an important restriction for verb-initial languages:
they do not allow ‘non-referential quantified NPs such as ‘nothing’ or ‘anybody’ in
the postverbal subject position’ (ibid. 254). This is definitely the case in Zoogocho
Zapotec. Consider the following example.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 4 4
(47) a.nozxono dx-az-tek yag yinha
nobody cont-plant-emph plant chile
‘Nobody plants chiles (anymore).’
b. * dx-az-tek nozxono yag yinha
cont-plant-emph nobody plant chile
*’Nobody plants chiles.’
She proceeds to discuss a theory of information structure which is useful to
consider. She differentiates between topic and focus and follows Reinhart in defining
sentence topics as ‘referential entities under which propositions are classified in the
context set’ (ibid. 255). Thus, topics are presupposed to exist and be semantically
specific (ibid. 255). Focus is defined as being what the proposition is asserting, and
is more likely to be an adjunct. She also states that this makes focus the licenser of a
Wh-phrase, which considering the following examples, seems to make sense. In (48),
an example of a topicalized subject is shown, in (49) a focussed subject, and in (50) a
Wh-phrase. Note that both (49) and (50) do not have a subject clitic following the
verb. I will discuss my reasoning for considering (48) to be topicalization and (49) to
be focus later in the current chapter.
(48) na da dolor=en dx-e=e=ne’
a n d d e c e a s e d d o lo r e s = d e t c o n t-s a y = 3 f= 3 fo
‘And the late Dolores said to him....’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(49) to=be zegh
one=3f stat.go
‘One of them went.’
(50) no zegh
who stat.go
‘Who went?’
She then goes on to argue that, because post verbal subjects must be referential in
verb initial languages, there is an obligatory mapping between post verbal subject
and topic (ibid. 256). This leads her to the first statement of the SO/OS hypothesis.
SO/OS HYPOTHESIS: FIRST APPROXIMATION
(i) VSO and VOS languages share a predominantly head-initial/right-
branching structure but differ in the order of those elements that are not in the
head-complement relation, namely, the relative ordering of different nominal
constituents, adverbials, and sentential complements (arguments and
adjuncts).
(ii) The linear order of these constituents reflects more general principles of
the ordering of topic and focus constituents. Specifically, VSO languages are
predominantly ordered topic-before-focus, while VOS languages are focus-
before-topic. (ibid. 257)
However, as Polinsky notes, VSO languages can reflect either an earlier SVO
stage or an earlier VOS stage and as such are not as rigid as VOS languages in the
application of this rule. While Polinsky gives evidence for the rightness constraint
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
346
based on YOS languages, I will contrast the evidence given for the rightness
constraint with data from SBZZ which confirms the leftness constraint.
Polinsky’s first point of comparison is that in transfer DO constructions, ‘the
recipient in a VOS language follows the patient’ (ibid. 261).
(51) 7a li Xun-e ba y-ak’-be chi tom li
TOP ART Xun-e go A3-give-APPL pig ART
7antz-e
woman
‘Xun went to give the pig to the woman.’ Tzotzil (Polinsky’s example (17)
(ibid. 261) taken from (Aissen 1987:105))
(52) tya7 rpaq aa’ xwan
3:2:give money youth Juan
‘Give Juan his money.’ Tzutujil (Polinsky’s (18) (ibid. 261) taken from
(Dayley 1985: 323))
In SBZZ, the relative order of recipient and patient in ditransitive transfer
constructions can vary, but the basic order is VSRP, with the exception of those
constructions exhibiting indirect object ‘lowering’, thus, as might be expected, this is
the opposite of a VOS language in this particular way.
Similarly ,’(I)f a VOS language has an applicative or a causative, the applied object
in this construction follows the basic object’ (ibid. 261).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(53) no-aso-api te bae te iai-no
3-sell-APPL ART rice ART sibling-3POSS
‘He sold his brother some rice.’ Tukang Besi (her 19 (ibid. 261) taken from
(Donahue 1995:236))
For the causative construction, in Zoogocho Zapotec, the applied object precedes the
basic object, as in the following.
(54) b-sede maestr to bidao to kwent
comp-caus.leam teacher one child one story
‘The teacher taught the child a story.’
‘Similarly, reflexives are controlled “right to left” in VOS languages, for instance, in
Tzotzil (see also Kekchi (28) below):’ (262)
(55) 7ep 7i-s-tak-be s-ba vun li xun-e
lotsASP-AGR-send-APPL AGR-self paper ART Xun-CL
‘Xun sent himself lots of letters.’ Tzotzil (Her 24 (ibid. 262) taken from
(Aissen 1987:135))
lit: ‘Lots sent self papers Xun.’
(56) x-Or-il r-ib(i) li al(i) sa’ lem
TNS-AGR-seeAGR-self ART boy in mirror
‘The boy saw himself in the mirror.’ (her 28 (ibid. 262) taken from
(Berinstein 1985:184))
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
348
Similar constructions in Zoogocho Zapotec have been discussed in 6.4. On
some levels, it appears that Zoogocho Zapotec is even more subject to the rightness
constraint than most VOS languages in this respect, as both the reflexive marker and
the verb are not marked for agreement.
(57) sh-na kwin lalo lho spejw
cont-see self lalo in mirror
‘Lalo saw himself in the mirror.’
She also notes that, ‘(P)ossessive reference is tracked right-to-left and the
following example from Tzutujil shows that such tracking may be rather long’(ibid.
262).
(58) xinwijljun r-wach r-hajab’ r-k’aajool nb’-esi’ino
found a its-strap his-shoe his -son my-neighbor
‘I found a strap of my neighbor’s son’s shoe.’ Tzutujil (Her 25 (ibid. 262) from
(Dayley 1985:286))
Possessive reference is also tracked right to left to the same extent in
Zoogocho Zapotec, as in the following.
(59) na baca che xna bidao=na
hand cow of mother child=det
‘the child’s mother’s cow’s hoof
Polinsky predicts that according to the Rightness Constraint, adjuncts should
precede their arguments. However, as she notes, this is not always the case as seen in
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
349
the difference between (60) and (61) below (ibid. 263). One reason for this is the
semantic dependency between a verb and its arguments.
(60) x-Ox-q’ue r-e lix mar li utz’u’uj
TNS-AGR-give AGR-DAT ART Mary ART flower
‘He gave the flowers to Mary.’ Kekchi (Her (26) (ibid. 263) taken from
(Berinstein 1985:191))
(61) x-Or-il r-ib(i) li al(i) sa’ lem
TNS-AGR-seeAGR-self ART boy in mirror
‘The boy saw himself in the mirror.’ Kekchi (Her (27) (ibid. 263) taken from
(Berinstein 1985:184))
Zoogocho Zapotec, as a VSO language consistently places arguments before
adjuncts as seen above in the simple constructions chapter.
A difficulty with Polinsky’s account, which she notes, is that, in OS
languages, ‘ the linear precedence of focus to topic apparently contradicts the left-to-
right parsing strategies.’ (264)
She notes that OS languages might opt for strategies which result in a ‘shortening
(of) the segments over which the Rightness Constraint operates’(264).
Note that all the generalizations and constraints proposed above have been
arrived at on the basis of segments with overt nominal constituents. This
suggests that if the conflict between the processing needs and the language-
internal information structure requirements cannot be resolved entirely, it can
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 5 0
nevertheless be alleviated by keeping the number of arguments per utterance
to a minimum, (ibid. 265)
Polinsky then goes on to show that, among the null-subject (pro-drop)
languages which she surveyed, those languages which were VOS tended to have
fewer clauses with more than one overt argument than others and conversely more
clauses with one overt argument (ibid. 265). This claim will be discussed briefly
below in reference to the SBZZ data.
8.3 A pragmatic approach to word order variation in Verb initial
languages: Du Bois and ‘The Discourse Basis of Ergativity’
In his groundbreaking 1987 paper, ‘The Discourse Basis of Ergativity’, John
Du Bois showed that, in Sacapultec Maya, the subjects of intransitive clauses (S’s)
and the objects of transitive clauses (O’s) not only shared the same verbal cross-
referencing techniques in contrast with the subjects of transitive clauses (A’s), but
also shared the same patterning of information status, S’s and O ’s were much more
likely to be new information than A ’s. This then gave a potential discourse/pragmatic
explanation for the ergative/absolutive, nominative/accusative morphological split
which had long been known6 but for which previously there had been no real
explanations put forth.
In this section, I will not go into great depth about his findings, but will
merely summarize. Du Bois analyzed a relatively large corpus based on the
following criteria, a) morphological type (whether the mention of a referent was
lexical, pronominal, or afflxal) (Du Bois 1987:814), b) inherent semantic class
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 5 1
(whether the referent was human, inanimate, or grammatical) (ibid. 814)7 , c)
Grammatical role (each mention was classified as A,S,0, oblique, or possessor, but
he focussed, as I will on the first three)(ibid. 814), and d) Information Status, (he
classified each mention as GIVEN if the mention was a speech act participant or it
was mentioned less than 20 previous noun phrases, NEW if it referred to a mention
which was not previously referred to, was not a speech act participant, or was not
part of ‘a previously evoked, entity-based frame’, or ACCESSIBLE if it was part of
that already mentioned entity-based frame or it was mentioned more than 20 noun
phrases previously (ibid. 815-816).
Du Bois findings, some of which I will discuss in greater depth in reference
to my own, were remarkable. He found that, to begin with, the majority of transitive
clauses either included one lexical argument or none but very rarely included two
(ibid. 818-819). Furthermore, he found that the majority of lexical arguments were
either S’s or O’s (ibid. 821-22).These findings were summarized in two principles:
‘Avoid more than one lexical argument per clause’ (ibid. 819), and ‘Avoid lexical
A ’s’ (ibid. 823).
In terms of the pragmatics of the clause, he made the following findings: not
a single one of the clauses in his corpus contained two new arguments,leading to the
statement ‘Avoid more than one new argument per clause’(ibid. 826); and, the
majority of new arguments are introduced as either the object of a transitive clause,
the subject of an intransitive clause, or an oblique, leading to the formulation ,’Avoid
new A ’s’ (ibid. 827) (ibid.825-827). I will examine the figures in greater depth in
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 5 2
discussing the SBZZ findings. Du Bois then goes on to show that a wide range of
languages, both nominative/accusative and ergative/absolutive, show these same
patterns of discourse.
8.4 A processing approach to VSO languages
In this section, I will propose a processing theory for VSO ordering based on
the theories of John A. Hawkins (1994). He states the basic intuition behind his
Performance theory of order and constituency as follows:
(I believe that) words and constituents occur in the orders they do so that
syntactic groupings and their immediate constituents (ICs) can be recognized
(and produced) as rapidly and efficiently as possible in language
performance. (Hawkins 1994: 57)
He gives the following two sentences as illustrative examples of what he means by
this statement.
(64) a. I V p[gave np[ the valuable book that was extremely difficult to find] pp[ to
Mary.]]
1 2 3 4 5 6 7 8 9 10 11
(Hawkins’ example (3.1a))
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
353
(65) b. I vpfgave pp[to Mary] n p[the valuable book that was extremely difficult to
find.]]
1 2 3 4
(Hawkins’s example 3.1b)
Example (3.1b). provides a more rapid presentation of the three ICs of the VP
(V, NP, and PP) than (3.1a). The verb gave is the first IC of the VP in both
examples and signals to the parser that a VP should be constructed. The PP is
a two word IC here. Its positioning to the left of the lengthy NP in (3.1b)
makes it possible for all three daughter ICs to be recognized within a short
viewing window, (ibid. 57)
Hawkins goes on to define the Constituent Recognition domain as follows:
Constituent Recognition Domain (CRD) (ibid. 58-59):
The CRD for a phrasal mother node M consists of the set of terminal and
non-terminal nodes that must be parsed in order to recognize M and all ICs
of M, proceeding from the terminal node in the parse string that constructs
the first IC on the right, and including all the intervening terminal nodes and
the non-terminal nodes that they construct.
He goes on to state the following assumptions about constituent structure in the
following definition o f immediate constituent.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The ICs of each phrasal node M will be all and only the phonetically realized,
lexical, functional, and phrasal categories most immediately dominated by M
in surface structure, excluding any phrasal categories that are non-maximal
projections of the head. (ibid. 73)
In order to calculate the immediate constituent to non-immediate constituent ratio,
one uses the following procedure.
The IC-to-non-IC ratio for a CRD is calculated by dividing the number of ICs
in the domain by the total number of non-ICs (or words alone) in that
domain, expressing the result as a percentage. The ratio for a whole sentence
is the aggregate of the scores for all CRDs within the sentence, (ibid. 77)
Finally, Hawkins defines Early Immediate Constituents as follows:
The human parser prefers linear orders that maximize the IC-to-non-IC ratios
of constituent recognition domains. Orders with the most optimal ratios will
be preferred over their non-optimal counterparts in the unmarked case; orders
with non-optimal ratios will be more or equally preferred in direct proportion
to the magnitude of their ratios. For finer discriminations, IC-to-non-IC ratios
can be measured left to right, (ibid. 78-79)
This leads to the following with respect to verb initial languages.
I shall therefore assign one more word to O than S in VO languages, e.g. two
to S and three to O. With these assignments, EIC’s scores for the three VO
orders (of (5.70) are given in table 5.12, assuming a VP (and VP
discontinuity in VSO languages)...Notice that VSO has a better ratio than
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
VOS, even though there is discontinuity between the ICs of VP, V, and O.
This discontinuity extends the VP domain and results in a low ratio for VP.
But the intervening subject produces a perfect ratio for the S (i.e. clause: the
V constructs VP, the first IC of S; and the subject can then immediately
construct the second IC, i.e. 2/2=100%. By contrast, VOS has a very low
ratio for the S domain because of the long initial VP (40%), and although the
VP is optimal, the aggregated score is lower than that for VSO. (ibid. 330)
Note that if subject and object were taken to be equal in length, the
aggregated scores would be equal for VSO and VOS languages. In any event,
on the basis of the previous paragraph and on Tomlin’s (1986) sample, the
following table is obtained.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 5 6
Table 8.1 EIC and VO languages
(Adapted from Hawkins (1994: 331 Table 5.12))
Assume: mS8 =2 words, mO=3, V=1
VP dominates V and mO (even when discontinuous)
V or O constructs VP
Total n in sample=402 languages (217 VO languages), from Tomlin 1986
Structure Agg. IC to word ratio N in sample % of VO languages
1. s[mS vp[V mO]]
S CRD: 2/3=67% 84% 168 77%
VP CRD: 2/2=100%
2. s[vp[V]mS vp[mO]]
S CRD:2/2=100% 75% 37 17%
VP CRD:2/4=50%
3. s[vp[V mO] mS]
S CRD:2/5=40%
VP CRD: 2/2=100% 70% 12 6%
On the basis of this theory, I have obtained the following hypothesis.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 5 7
Hypothesis: The length of subjects in VSO structures will be shorter than or
equal to the length of objects, in order to facilitate the recognition of the VP.
This hypothesis is as it is because if it were the case that subjects were longer
than objects, say if subjects had a mean aggregate length of three and objects had a
mean aggregate length of two; then, even though the constituent recognition domain
for the subject would remain perfect, the constituent recognition domain for the verb
phrase would be 2/5 and the overall aggregate IC to word ratio would be 70%, much
poorer than if the subject were equal (75%) or shorter than the object (84% if mS 1
and mO=2). I will now investigate this hypothesis on the basis of a quantitative study
of textual data from SBZZ.
8.5 Textual data from SBZZ
8.5.1 Methodology
In this section, I will briefly discuss the methodology I have used in this
study. While I feel that this study is adequate for the present purposes, it could, as
always, be improved upon.
8.5.1.1 Corpus
The data in the present study were compiled from two texts and three
speakers. The first text I coded was from a 68 year old bilingual Zapotec dominant
speaker and was a historical narrative monolingual. This text was the shorter of the
two texts at 351 clauses. My main consultant, a 38 year old bilingual speaker of
Zapotec and Spanish aided in the collection and elicitation of this text and the
following one, however in the first text, her contributions were at a minimum.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 5 8
The second text was basically a conversation between a 77 year old
monolingual speaker of Zapotec and my primary consultant and its topics range from
historical narrative to gossip. As a result, its style is different from that of the first
text. It was the longer of the two texts, comprising 1591 clauses.
Both texts were transcribed and translated by my primary consultant and
myself, and any errors are solely my responsibility. I questioned the syntactic and
morphological structures and semantics as we were transcribing these texts in order
to get an idea of the potential variation and to aid my understanding of the texts.
8.5.1.2 Coding
I coded for the following types of data. I coded Subject, Primary Object, and
Secondary Object. Obliques, including prepositional phrases, were not coded unless
they were subcategorized for by the verb (in which case they were treated as objects)
or they intervened between a verb and its core arguments (which occurred very
rarely). Adverbs did not intervene between the verb and core argument in either of
these texts, even though that is a grammatical possibility in the language. I assigned
a value of one to each element in a particular noun phrase, including pronominal
clitics, even though they are phonologically reduced, on the reasoning that they are
syntactically very salient, being able to satisfy argument structure requirements by
themselves, and excluding determiner and directional clitics based on the reasoning
that they do not contribute to the argument structure. I did code for determiners and
directional clitics, so this data is recoverable if need be.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 5 9
I also coded the information status of each noun phrase, based on a
given/new distinction. I coded an element as new if it was not mentioned in the
previous 20 noun phrases mentioned9 and as given otherwise. Note that I have not
otherwise coded for referential distance at this point. I coded the ordering of these
elements and the verb. I did not count auxiliary verbs unless they had a specified
subject. 101 also, for the purpose of the present study, excluded questions as they have
a syntacticized, pragmatically marked structure.
Similarly, although I did count the elements in relative clauses to come up
with the length of individual noun phrases modified by the relative clause, I did not
count relative clauses as clauses in their own rights for similar reasons to the reasons
I excluded questions. I also excluded imperatives and did not count objects of
infinitives as they did not vary in terms of their ordering. I also excluded the
argument sharing constructions discussed in 6.5.6. All of these elements are coded
and easily recoverable as well. I excluded incomplete clauses, such as clauses which
were started and then abruptly ended because of an interruption of some sort or
another, and those clauses which, in some cases because of the brief intermi ttent
periods of heavy rain on the corrugated aluminum roof during the recording were
incomprehensible to my primary consultant and myself.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 6 0
8.5.2 Data
8.5.2.1 Types of data
I will now give brief examples of the types of data which were encountered
and counted in the texts. I will start by describing those sentences which were verb
initial. As will be seen, by far the two most common types of clauses present in the
text were intransitive clauses with a pronominal subject and transitive clauses with
pronominal subject and object, as seen in the following two examples:"
Vs
(66) na b-ese-dxogh=e’
demdist comp-pl-leave=3f
‘They left there.’
Vao
(67) bi=dx b-le’i=da=n
neg=emp comp-see= 1 sgexp=3 inan
‘I really didn’t see it.’
The next most common type of clauses in these texts were intransitive
clauses with full noun subjects and transitive clauses with pronominal subject and
nominal object as seen below.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
361
VaO
(68) dx-ot=e’ kuzh=en’
cont-sell=3f pig=det
‘They sold pigs.’
VS
(69) na gu-s-ak shone cocineras ka’
and comp-pl-be three cooks demadv
‘And three cooks were there...’
Transitive clauses with two non-pronominal arguments were very rare. Ditransitivc
clauses with three lexical arguments did not occur.
VAO
(70) kate b-edey-a mansia yish=en’
when comp-ven-take Amansia grinding_stone=det
‘When Amansia came to take the grinding stone...’
Also rare were ditransitive clauses with pronominal recipient and nominal patient, as
in the following. Note that none of these encoded repeated objects. Note that (71) has
a fast speech form of the pronoun, and (72) has a full pronoun. These two types
should probably be combined.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
362
V aol02
(71) na dxy-izxw=e=le’ tapa rrel
and cont-pay=3f=2sg 4 real
‘And they pay you four reals(50 centavos)’
VaO 102
(72) b-enh=e’ neda shi rrel
comp-give=3f lsg 10 reals
‘They gave me 10 reals.’
Slightly more common are those instances of the reflexive-of-possessor
constructions, as seen in the following example.
Reflexive-of-possessor
(73) na b-zhelh gadx x-medxu=e’
there comp-fmd much poss-money=3f
‘He found a great deal of (his) money there...’
Excluding for the moment those verb initial clauses which included repeated
subjects (which will be discussed later in the current section), these were the main
types of verb initial structures which were found in the texts. I will now discuss those
instances of fronted subjects. Note that I will not discuss their pragmatic status at this
juncture, but will wait to do so until later in this chapter. The most common type of
intransitive construction with a fronted subject is one without a coreferential pronoun
following the verb.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(74) da dolor chiquito b-en rezgar
deceased doloreschiquito comp-do dared (Sp.)
‘It was the deceased Dolores Chiquito who dared...’
A close second were those clauses which have fronted subjects and a coreferential
pronoun following the verb as in the following.
SVs
(75) per to-z=e’ dx-egh=e’
but one-emph=3f cont-go=3f
‘But she goes alone...’
Transitive sentences with fronted nominal subjects and verbs followed by
pronominal subjects and objects were also found as seen below.
AVao
(76) na’ da dolor=en’ dx-e=e=ne’
and deceased dolores=det cont-say=3f=3fo
‘And the late Dolores said to him....’
There were also a number of transitive sentences with non-pronominal fronted
subjects and non-pronominal postverbal objects.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
364
AVO
(77) lhe na b-et metal-en’
focus demdist comp-grind metal-det
‘This is what ground the metal.’
Similarly, a number of sentences with non-pronominal preverbal subjects and non-
pronominal postverbal objects and postverbal pronominal subjects also occurred in
the corpus.
AVaO
(78) benhe g-onh=e’ yet=en’
person comp-give=3f tortilla=deT
‘People give tortillas.’
A number of copular sentences with pre- and post-verbal nouns as in the following
were also found.
Copular sentences (S1VS1)
(79) benhe n-ak noolhwego
person stat-cop woman young
‘The person is a young woman.’
One very unexpected type of sentence which was found was a sentence with
fronted subject and object. These are very odd and require further research. Note the
preposed headless relative clause in (81). (82) shows a negated subject preceding the
clause. The preverbal ‘subjects’ in both of these clauses could be considered separate
constituents.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 6 5
AOVa
(81) danh de ka ne nilhe ka
geninan cop demadv not cookedcorn demadv
b-et=en
comp-grind=3 inan
‘The thing like this cannot even grind the nixtamal.’
(82) ne xna=o’ ka b-le’i=de’
not poss.mother=2sg demadv comp-see=3fexp
‘Not even your mother saw it like that.’
Another unexpected find was the following, a clause which looks like a reflexive of
possessor construction, but which has what is translated into Spanish as a preposed
non-pronominal subject and object. In all, fairness, this example should be
considered to be a construction with the first NP (zxoan) a separate non-argument (a
left dislocated topic), the second NP (to gosh) the grammatical subject, and the
postverbal NP (yeenbe’) a location.
SOVLocs
(83) na zxoan na’ konto gosh zxoa yeen=be’
and juan there like one sack stat.lie neck=3inf
‘And Juan there has a sack propped up on his neck.’ (He is carrying the bag
on the back of his neck.)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
By far the most common object initial clause type are those clauses which
have a non-pronominal fronted object and a post-verbal pronominal clitic subject, as
seen in the following.
OVa
(84) pur dizha zxon=na’ gu-ne=e’
only tongue/language Zapotec=det comp-speak=3f
‘They only spoke Zapotec.’
There are also constructions which include a pronominal copy of the preverbal
object, as in (85) and (86).
QlVaol
(85) galgh pes g-onh=e=n
twenty pesos pot-give=3f=3inan
‘Twenty pesos they give.’
(86) nisfero=na’ dx-ap=a=ne’
nisfero=det cont-say=lsg=3fo
‘Nisfero, I said to him...’
There are also a few cases of a preverbal non-pronominal object and a postverbal
non-pronominal subject.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 6 7
OVA
(87) lelez gashgho=n dx-eyalha sii maxa
because half=3inan cont-should pot.grab maximino
‘Because Maximino should take half of it...’
Finally, before turning to those sentences which exemplify repeated subjects,
there are a number of ditransitive clauses which have one fronted object. These
clauses either have a non-pronominal subject like (88) or a pronominal subject like
(89).
Q1VAQ2
(88) chupe gayoa yoo gasgh gu-zxi da berte blahagen
lenh
two hundred fifty comp-pay deceased alberto fuentes
3inan
‘Two hundred fifty the late Albert Fuentes paid for this.’
0 1 VaQ2
(89) yez le zxis dx-om=ba’ yel=en’
mazorca focus stick cont-make=3an m ilpa-def
‘It made the milpa into little scraps of mazorca.’
There was also one instance where a demonstrative pronoun coreferent with a
preverbal object was found following the verb subject complex.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 6 8
O lV aO l
(90) por menos yet dao le bi gud-aw=a’ danh
at least tortilla tassels focus not comp-eat=lsg geninan
‘At the very least I didn’t eat tortillas made from corn tassels.’
Finally, I encountered a fair number of sentences which contained repeated
subjects. For the most part these were emphatic pronominals, but there were a few
cases of non-pronominal repeated subjects. I will not go into any great detail about
these at this point, except to state that the repetition of the subject gives an emphatic
reading for the subject, as one might expect.
Repeated subjects
VsS
(91) na sh-da=a=nda
demdist cont-walk=Tsg=Tsg
‘There I was walking.’
VaoA
(92) b-le’i-dx=da’=n neda’
comp-see-emph=lsgexp=3inan lsg
‘I still saw it.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
369
VaAO
(93) bi g-daw=a=nda’ yet dao
neg comp-eat= 1 sg= 1 sg tortilla com tassels
‘I didn’t eat tortillas made from com tassels.’
OYaA
(94) pur dizha zxon na ne=dxo
only language zapotec now stat.speak=lplincl
dxop=dxo
two=lplincl
‘Only Zapotec we’re going to speak now, the two of us’
Note that the majority o f the examples which had non-pronominal repeated subjects
were like (94) and had a quantifier followed by a clitic pronoun. There was one
instance of a proper name as in (95).
(95) nhe ka gu-ne-z=be’ da liaventura
neg demadv comp-speak-emph=3inf deceased Buenaventura
‘The late Buenaventura didn’t say that.’
8.5.2.2 Quantitative data and analysis
Table 8.2 Overall count of clause types
Clause type Number of clauses of this
type
Percentage of total
Vs 805 41.5%
VaO 315 16.2%
VS 308 15.9%
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 8.2 Overall count of clause typeslcont.)
3 7 0
Vao 204 10.5%
OVa 75 3.9%
VsS 40 2.1%
SV 39 2.0%
SVs 25 1.3%
SVS 19 1%
Reflexive-of-possessor 18 .9%
VAO 16 .8%
AVaO 11 .5%
AVao 11 .5%
O lV aol 10 .5%
AVO 9 .5%
V a o l0 2 9 .5%
VaAO 8 .4%
OVa A 4 .2%
OVA 3 .2%
0 1 V a 0 2 3 .2%
AO Vs 3 .2%
Vao A 2 .1%
V a 0 1 0 2 2 .1%
AOVLOCa 1 <.1%
0 1 V A 0 2 1 < .1%
O lV aO l 1 <.1%
TOTAL 1942 slightly less than
100%
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Before I go on to examine the individual patterns which have been found,
allow me to begin with a brief discussion of the data as a whole. The reader can see
that five of the first six most frequent clause types are verb initial, and comprise 86.2
percent o f the total overall number of clauses. This is not to say that the other orders
are not significant, just that verb initial ordering is indeed the norm in Zoogocho
Zapotec. Nor does this say anything about the grammaticality of the other clause
types. In addition, as I will discuss in greater depth below, within the top 8 most
frequent clause types (combined, a whopping 92.1% of the grand total) are no
clauses with more than one lexical argument, if the definition of lexical argument
excludes clitic pronouns. I will only reiterate what was said above in stating that this
says nothing about the grammaticality or significance of the other types. It does say
something about the overall discourse patterning of language, which will be returned
to later in this chapter.
Having looked at the overall totals, the data will be broken down more
specifically, before being compared with data from the studies by Du Bois and
Payne. To begin with, examine the following chart.1 2
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 7 2
Table 8.3 Overall transitivity
Transitivity Number Percentage
Intransitive 1236 63.6%
Transitive 691 (Including reflexive of
possessor and AOVLOCa)
35.5%
Ditransitive 15 < 1%
As is easily seen, intransitive sentences made up by far the bulk of the corpus
which was studied. This correlates with other studies of this sort. For example, in
Doris Payne’s investigation of word order in Yagua, 69% o f the clauses were
intransitive, the remainder being transitive (Payne 1990: 222). The only potential
surprise is that there were any ditransitives, given their rarity or nonexistence in Du
Bois’s and Payne’s studies.
I then proceeded to analyze the clauses by various factors: first off by
transitivity, compiling the average length of the arguments and the number o f given
and new mentions. I then proceeded to break down each of the three major
transitivity classes (Intransitive, Transitive, and Ditransitive) further into subtypes,
based on the basic ordering of verb, subject, and object(s) and then compiled
individual tables for each of these sub-types. For example, for intransitive clauses,
there were three basic possibilities; there were sentences of the type: VS, sentences
of the type SV, and sentences of the type SVS. For sentences of the VS type, I could
further break them down into sentences with a pronominal argument (Vs), sentences
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 7 3
with lexical arguments (VS) and sentences with a repeated subject (VsS). The fourth
possible permutation (VSs) was ruled out by the grammatical conventions of the
language.
Now I will move on to examine the individual data sets, beginning with
intransitives. It should be mentioned that each of the NPs in the copular sentences
was counted.
Table 8.4 The VS subtype
VS type # Average Length of
Subject
Given/New
Subject
Vs 805 1 805/0
VS 308 1.65 196/112
VsS 40 1.23 33/7
Total: 1153 1.18 1034/119
The VS subtype demonstrates a number of things. The first thing is that clauses like
(95) (67 repeated), were by far the most common and pragmatically least marked
clause type in the entire corpus.
Vs
(95) na b-ese-dxogh=e’
dem comp-pl-leave=3f
‘They left there.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 7 4
Next, (68), repeated here as (96), exemplifies the third most common clause
type in the corpus; and one of the types with the highest proportion of new mentions
and consequently one of the largest average subject lengths.
VS
(96) na gu-s-ak shone cocineras ka’
and comp-pl-be three cooks pi
‘And three cooks were there...’
Finally, (91), repeated here as (97), exemplifies the use of the repeated subject in
such constructions.
VsS
(97) na sh-da=a=nda’
demdist cont-walk= 1 sg= 1 sgfsf
‘There I was walking.’
Table 8.5 The SV type
SV type # Average Length of
Subject
Given/New
Subject
SV 39 1.23 34/5
SVs 25 1.28 23/2
Totals 64 1.25 57/7
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 7 5
A much less common type than the VS type, the SV type includes two
subtypes, the first type, SV, will be provisionally considered to be a focus
construction. Note that this example is made up of a question and an answer.
SV
(98) Panfila no dx-i=o’ ka’
who cont-tell=2sg demadv
‘Who were you telling that to?’
Alberta moises dx-ap=a’ ka’ ke, lee
moises cont-tell=lsg demadv no, 2sg
‘It was Moises I told that to, you know?’
For the time being, the second type being discussed here, with a repeated subject
clitic after the verb, will be tentatively called topicalization. Example (75), repeated
here as (99) illustrates this. I will come back to the discussion of these terms later in
the current chapter.
SVs
(99) per to-z=e’ dx-egh=e’
but one-emph=3f cont-go=3f
‘But she alone goes...’
Now I will turn to copular sentences of the type SVS, as seen in the chart
below. I count each S separately, the S in parentheses being the S which was
counted.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 7 6
Table 8.6 The SVS type
SVS type # Average Length of
Subject
Given/New
Subject
(S)VS 19 1.48 11/8
SV(S) 19 1.52 13/6
Totals 38 1.5 24/14
There is not much to say about the SVS type, with the exception that it is
quite frequently used to introduce new discourse entities and as such, is quite often
longer. The first NP mentioned was often the new NP. It is interesting to note that
there were no copular sentences of the sort VSS. Copular sentences of the sort VsS
were counted as VsS’s.
The table below represents all of the intransitive clauses in the corpus.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 8.7 Intransitive clauses
3 7 7
Intransitives # Average Length of
Subject
Given/New
subject
Vs 805 1 805/0
VS 308 1.65 196/112
VsS 40 1.23 33/7
SV 39 1.23 34/5
SVs 25 1.28 23/2
(S)VS 19 1.48 11/8
SV(S) 19 1.52 13/6
Totals 1255 1.19 1115/140
Now moving on to the transitive clause, the VAO type is the first table which
I will investigate, as seen below.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 7 8
Table 8.8 The VAO type.
VAO
type
# Average
Length of
Subject
Average
Length of
Object
Given/New
Subject
Given/
New
Object
VaO 315 1 1.78 315/0 179/136
Vao 204 1 1 204/0 204/0
VAO 16 1.31 1.19 11/5 13/3
VaAO 8 1 1.85 8/0 2/6
Vao A 2 1 1 2/0 2/0
Totals 545 1.01 1.47 540/5 400/145
By far, as seen in the data, the most common VAO type is the VaO subtype
as seen in (100) ((69) repeated from earlier). In (100), one can see that objects in
such constructions are quite often definite. Furthermore, there is a high proportion of
new mentions for objects of this subtype, with corresponding increased length.
VaO
(98) dx-ot=e’ kuzh=en’
cont-sell=3f pig=det
‘They sold pigs.’
The next type, seen in (68), repeated here as (101), is a non-pragmatically marked
pronominalized clause, and is among the most frequent clause types in the entire
corpus.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 7 9
Vao
(101) bi=dx b-lei=da=n
neg=emp comp-see=1 sgexp=3 inan
‘I really didn’t see it.’
The VAO subtype (or VSO as it is often referred to) was really not that common.
Note that subjects were quite frequently new information and were longer than the
corresponding objects by a small amount. This would appear to contradict the
hypothesis, if the fact that the overall length for all of the VAO types subjects was
far less than that o f the objects was ignored.
VAO
(102) kate b-edey-a mansia yish=en’
when comp-ven-take Amansia grinding_stone=def
‘When Amansia came to take the grinding stone...’
The last two types of the VAO type involve repeated emphatic subjects, as in the
following examples.
Vao A
‘I still saw it.’
The next example came at a point in the text when someone was discussing
how land was going to get divided, and stated that she thought that someone else
(103) b-le’i-dx=da’=n neda’
comp-see-emph=l sgexp=3inan lsg
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
should get half of the land and that she should get the other half. This is typical of
repeated subjects.
VaAO
(104) na sii=a’=nda’ ye gashgho
and pot.grab=lsg=lsg other half
‘And I am going to take the other half.’
Table 8.9 The OVA type
OVA
type
# Average
Length of
subject
Average
Length of
Object
Given/New
Subject
Given/
New
Object
OVa 75 1 1.64 75/0 53/22
O lV aol 10 1 1.9 10/0 4/6
OVa A 4 1.25 2.25 4/0 2/2
OVA 3 1 2 3/0 3/0
O lV aO l 1 1 2,1 1/0 1/1
Totals 93 1.01 1.7 93/0 63/30
The OVA type’s most numerous subtype was OVa, as seen in (83), repeated
as (105). Note that many tokens of objects in this subtype were new mentions and
that there was a correspondingly high length. This one appears to correspond to the
focussed subject construction above.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
381
OVa
(105) pur dizha zxon=na’ gu-ne=e’
only tongue/language Zapotec=det comp-speak=3(f)
‘They only spoke Zapotec.’
The next most common subtype, O lV aol, is seen below. The object is definitely
emphatic in these constructions. Note that this type appears to correspond to the
subject topicalization construction discussed above in 5.5.
O lV aol
(106) galgh pes g-onh=e=n
twenty pesos pot-give=3f=3inan
‘Twenty pesos they give
(107) nisfero=na’ dx-ap=a=ne’
nisfero=det cont-say= 1 sg=3 fo
‘Nisfero, I said to him...’
I encountered only three examples of the following type. As will be shown
later, clauses with two or more non-clitic arguments are very rare in SBZZ. In all of
the examples of this type which I have found in the corpus, the subject, even though
a full NP, was given information, as was the object. This appears to be focussing the
object.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 8 2
OVA
(108) lelez gashgho=n dx-eyalha sii maxa
because half=3inan cont-should pot.grab maximino
‘Because Maximino should take half of it...’
The following could actually have potentially been classified with the
O lV aol sub-type given the use of the demonstrative as the second occurrence of the
object. The speaker was emphatically stating that they did not eat that particular type
of tortilla, generally eaten in times of great scarcity. Notice that the focus was on the
type of tortilla which was eaten.
O lV aO l
(109) porm enos yet dao le bi gd-aw=a’ danh
at least [Sp.] tortilla tassels focus not comp-eat=lsg geninan
‘At the very least I didn’t eat tortillas made from com tassels.’
Finally, the repeated subject form of this sub-type is shown below.
OVaA
(110) pur dizha zxon na ne=dxo
only[Sp.] language Zapotec now stat.speak=lplincl
dxop=dxo
two=lplincl
‘Only Zapotec we’re going to speak now, the two of us’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 8 3
Table 8.10 the AVO type
AVO
type
# Average
Length of
Subject
Average
Length of
Object
Given/New
Subject
Given/
New
Object
AVaO 11 1.36 1.18 10/1 4/7
AVao 11 1.73 1 7/4 11/0
AVO 9 1.6 1.6 6/3 5/4
Totals 31 1.56 1.24 23/8 20/11
The AVO type is of interest for various reasons. There is no particular reason
to worry about the order of talking about the various subtypes, as they are all more or
less of the same frequency. Note that the fronted subjects are longer than the subjects
of the other subtypes discussed so far. This might be because they are new
information and are therefore more likely to be longer.
The first sub-type which I will discuss at this point is the AVaO subtype.
Note that the following example was part of a list of different ways people pay for
goods that a speaker was selling. Again, the repetition of the pronoun immediately
following the verb appears to be an instance of topicalization or at least topic-stating.
AVaO
( 111) benhe g-onh=e’ yet=en’
person comp-give=3f tortilla=def
‘People give tortillas.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 8 4
Identical claims can be made about the next subtype, which exemplifies the
AVao subtype.
AVao
( 112) na da dolor=en’ dx-e=e=ne’
and deceased dolores=det cont-say=3f=3f(obj)
‘And the late Dolores said to him....’
Finally, consider the following, focussed AVO subtype.
AVO
(113) lhe na b-et metal-en’
focus dem comp-grind metal-det
‘This is what ground the metal.’
Table 8.11 Miscellaneous transitives
Miscellany # Average
Length of
Subject
Average
Length of
Object
Given/New
Subject
Given/New
Object
Reflexive-
of-possessor
18 1 1.33 18/0 12/6
AOVLOCa 1
1,1 3,1 1/1 0/2
AOVa 3 2.33 1
2/1
2/1
It is telling that all of the subjects in the reflexive of possessor construction
were given information. I will exclude all three of these constructions from all future
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 8 5
quantification within this chapter, but will come back to discuss them later in the
current chapter. I must however briefly discuss the AOVa constructions, as seen
below in (114) and (115). These were very unexpected in terms of everything else
which was known about the language, never having been accepted when attempted
in elicitation. At this point, I feel that, especially given the pronominal subject clitic,
it is best to interpret the sentence initial noun phrase as being dislocated from the rest
of the sentence. The only examples I have for this sentence type are negative.
AOVa
(114) dan de ka nhe nilhe ka
geninan cop demadv not nixtamal neg
b-et=en
comp-grind=3 inan
‘The thing like this cannot even grind the nixtamal.’
(115) nhe xna=o’ ka’ b-le’i=de’
not poss.mother=2sg demadv comp-see=3fexp
‘Not even your mother saw that.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
386
Table 8.12 Transitive clauses
Transitive
Clauses
(Excluding
reflexive-of-
possessor
and
AOVLOCa)
# Average
Length of
Subject
Average
Length of
Object(s)
Given/
New
Subject
Given/New
Object
VaO 315 1 1.78 315/0 179/136
Vao 204 1 1 204/0 204/0
OVa 75 1 1.64 75/0 53/22
VAO 16 1.31 1.19 11/5 13/3
AVaO 11 1.36 1.18 10/1 4/7
AVao 11 1.73 1 7/4 11/0
O lV aol 10 1 1.9 10/0 4/6
AVO 9 1.6 1.6 6/3 5/4
VaAO 8 1 1.85 8/0 2/6
OVa A 4 1.25 2.25 4/0 2/2
OVA 3 1 2 3/0 3/0
AOVa 3 2.33 1 2/1 2/1
Vao A 2 1 1 2/0 2/0
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 8 7
Table 8.12 Transitive clauses (cont.)
O lV aO l
(counted 01
of
0 1 vs0 2 ....)
1 1 0 1 =2,
02=1
1/0 0/ 1(0 1 )
1/0(0 2 )
Totals 672 1.04 1.48 658/14 483/189
I will now move on to discuss ditransitive clauses. The first type which will
be discussed will be the V A 0102 type.
Table 8.13 The VAQ1Q2 type
V A 0 1 0 2
type
# Subj
Ave
O l
Ave
0 2
Ave
Given/
New
Subj
Given/
New
O l
Given/
New
0 2
V a o l0 2 9 1 1 2.1 9/0 9/0 3/6
V a 0 1 0 2 2 1 1 1 2/0 1/1 0/2
Totals 11 1 1 1.9 11/0 10/1 3/8
I will not discuss this type in any great depth. These are garden-variety
ditransitive clauses. One thing of interest is that all of the Ol ’s and subjects are given
information and are short and the longer, new information comes later.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 8 8
Vao 102
(115) na dxy-izxw=e=lhe’ tapa rrela
and cont-pay=3f=2sg 4 real
‘And they pay you four reals (50 cents).’
Table 8.14 The Q1VAQ2 type
0 1 V A 0 2
type
# Subj
Ave
O l
Ave
0 2
Ave
Given/
New
Subj
Given/New
O l
Given/
New
0 2
0 1 V a 0 2 3 1 1.67 1.67 3/0 0/3 2/1
01 V A 0 2 1 3 4 1 0/1 0/1 1/0
Totals 4 3 2.25 1.5 3/1 0/4 3/1
The other major types of ditransitive clauses are as follows. The first two
examples show pronominalized agents. In (116), it might be possible to interpret the
two objects as being coreferential (referring to different stages of the same entity),
but I have chosen not to.
0 1 Va0 2
(116) yez le zxis dx-om=ba’ ye Hen
mazorca focus stick cont-make=3an milpadef
‘Unfortunately, it made the milpa into little scraps of com plants.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 8 9
(117) dan dx-gape=d=a’ yet
geninan cont-pat_out=instrument tortilla
‘That is what I make tortillas with.’
O f interest in the second example is that the fronted object is far longer than
the secondary object.
Q1VA02
(118) chupe gayoa yoo gasgh gu-zxi da berte blahagen
lhen
two hundred fifty comp-pay deceased alberto fuentes
3inan
‘Two hundred fifty Albert Fuentes paid for this.’
Upon examination, the following chart shows that the object occurring first is
considerably shorter, on average, than the second. It also is much more likely to be
new information.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 9 0
Table 8.15 Ditransitive clauses
Ditransitives # Average
Length
of
Subject
Average
Length of
Objects
Given/New
Subjects
Given/New
Objects
V a o l0 2 9 1 0 1 = 1,
0 2 =2.1
9/0 9/0(01)
3/6(02)
0 1 V a 0 2 3 1 01
&02=1.67
3/0 0/3(01)
2/ 1(0 2 )
V a 0 1 0 2 2 1 0 1 &02=1 2/0 1/ 1(0 1 )
0/2(0 2 )
0 1 V A 0 2 1 3 01=4,
02=2
0/1 0/ 1(0 1 )
1/0(0 2 )
Totals 15 1.13 01=1.33
02=1.86
14/1 01=10/5
02=6/9
Before proceeding to compare the data obtained from SBZZ with the classic
studies of Payne and Du Bois, I will first discuss the following tables. First, table
8.16 shows the average length of post verbal subjects for intransitive and transitive
clauses and the number of given versus new mentions.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 8.16 Postverbal subjects
3 9 1
Postverbal subjects Average Length of
Subject
Given/New
Intransitive 1.18 1034/119
Transitive 1.01 632/7
Combined 1.12 1666/126
Similarly, table 8.17 shows the length of preverbal transitive and intransitive
subjects.
Table 8.17 Preverbal subjects
Preverbal subjects Average length of
subject
Given/new
Intransitive 1.25 57/7
Transitive 1.6 25/9
Combined 1.38 82/16
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 9 2
In the following chart, similar data for pre- and postverbal objects is shown.
Table 8.18 Pre- and postverbal objects
Pre- and postverbal
objects
Average Length of
Object
Given/New
Postverbal objects 1.47 418/155
Preverbal objects 1.68 64/31
The most important data to note in this instance is that transitive postverbal
subjects have an average length o f 1.01 words and that postverbal objects have an
average length of 1.47, which actually is the same as the average length of subjects
and objects in VAO structures. This average length of subjects and objects in VAO
structures easily confirms the first hypothesis stated above, that the average length of
subjects would be less than the average length of objects in VAO structures.
However, there is no significant difference in the position of transitive and
intransitive subjects with respect to the verb, which does not confirm the hypothesis
forwarded by Dryer which was discussed above that stated that there was likely to be
a difference in the placement of transitive and intransitive subjects. It is also
important to note that preverbal subjects and objects were much more likely to be
given information and to be longer than their postverbal counterparts.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
393
I will end with a brief condensation o f the data presented in this section. In
his (1994) A performance theory of order and constituency, John A. Hawkins
presents various predictions for the weight of sentential elements in Polish1 3 . There
are two general types of predictions: wzY/zm-structure predictions (predictions which
apply to only one structural type) and across-structure predictions (predictions which
are based on comparison of various structures) (Hawkins 1994: 175). The within
structure predictions are the following. The aggregate length of the VP (V+O) will
be greater than or equal to the length of the subject in SVO structures (based on the
principle of short before long). Similarly the mean length of O will be greater than or
equal to the V in SVO structures. Subjects should be longer than verb phrases in
VOS structures (short before long). Objects should also be longer than or equal to the
verb in VOS structures. The mean length of the VP should be greater than or equal to
the length of the S in SOV structures. The mean length of the V should be greater
than or equal to the length of the object in SOV structures. The mean length of the O
in a VSO structure should be longer than or equal to the mean length of the S which
in turn should be greater than or equal to the mean length of the V. In OSV
structures, it is expected that the mean length of the V be greater than or equal to the
mean length of the S. Finally, the principle of short before long predicts that the
mean length of the S be greater than or equal to the mean length of the V.
The acrav,v-structure principles are the following. The VP in a SVO structure
will be greater than or equal to the VP in a VOS structure. This is because the VP in
the SVO structure should be equal to or longer than the S and the VP in a VOS
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
394
structure will be shorter than or equal to the S; and therefore, assuming that the
length of pre and postverbal S’s are commiserate, one would expect by transitivity of
formulas that the VP would be longer. The S in a SVO structure will similarly be
predicted to be longer than a S in a VSO structure. The 0 in a SVO structure will be
predicted to be greater than or equal to the O in a VOS or a SOV structure. The mean
length of a S in a VOS structure will be predicted to be longer than the S in a SVO or
a SOV structure. Finally, the mean length of an O in a VSO structure will be greater
than or equal to the mean length of an O in a VOS or a SOV structure. Thus, 1
arrived at the final table in this section, which collapses many of the charts above.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 8.19 Zoogocho Zapotec s( vpfVl S vpIOH (adapted from Hawkins
(1994:177)) (frequency=proportion of transitive clauses)
Order: S VO V O S S O V
(S O V s)
Mean lengths:
Frequency:
Structure(s):
1.56 1 1.24
5%
(a) [S [V O ]]
n/a
(b) [[V O ] S]
2.33 1 1
<1%
(c) [S [O V]]
EIC predicts: (i) VP > S
[2.24>1.56]
(ii) 0 > V
[1.24>1]
(iii) VP > VP in (b)
n/a
(iv) S > S in (d&e)
[1.56>1.01]
(v) O > O in (b)
n/a
(vi) 0 > 0 in (c)
[1.24>1]
(i) S > VP
(ii) O > V
(iii) S > S in (a)
(iv) S > S in (c)
(v) S > S in (d&e)
(vi) S > S in (f&g&h)
(i) VP > S
[2<2.33]*
(ii) V > O
[1 = 1]
Order: V S O O S V O V s
Mean lengths:
Frequency:
Structure(s):
1 1.01 1.47
80%
(d) [VSO] or
(e) [yp[V]S v p [ 0 ] ]
n/a
(f) [OSV]
(g) [ v p [ 0 ] S V p[V]] or
(h) [O [S V p[ V]]]
1.7 1 1.01
14%
(i) [ V p [ O V] SI
(j) [ 0 V S]
(k) [O [V S]]
EIC predicts: (i) O > S
[1.47>1]
(0 V > S (i)S>V
[1.01>1]
(ii) S > V
[ 1.01> 1]
(iii) O > 0 in (b)
n/a
(iv) O > 0 in (c)
[1.47>1]
Only 10 of the predictions made above apply to SBZZ (the structures VOS and OSV
did not occur). O f these predictions, the Zoogocho Zapotec data confirms nine out of
the possible 10; six out of seven within-structurQ predictions, and three out of three
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3%
acms.s-structure predictions. The one wY/im-structure prediction which failed (that
the VP should be greater or equal to the S in a SOV structure) is based on a very
small amount of data (three examples amounting to less than one percent of the tot al
data), and would require a much larger corpus to find resolution.
8.5.2.3 Comparison with similar studies
In this section, I will compare the SBZZ data with data which Doris Payne
has presented for Yagua and which John Du Bois has presented for Sacapultec. The
biggest difficulty in comparing this data and similar data with the data which Du
Bois presents for Sacapultec and Payne presents for Yagua is that in both o f those
languages agreement is affixal, and so it is easier to classify what counts as a lexical
argument. I will treat the cliticized pronominal forms as agreement in most of the
following tables, indicating specifically where I do not. This really has little effect on
the tables discussed so far. For example, if one treats the clitic pronouns as affixes in
table 8.8 above, one would arrive at a mean length for subjects in VAO structures of
1.19 and a mean length for objects of 1.75, which confirms the hypothesis even more
strongly. I will, for the purpose of comparison with the other studies, treat the clitic
pronouns as affixal, noting where this leads to difficulty. The affix vs. clitic question
will be discussed in the conclusion to this chapter.
I will start by comparing the data which Payne (1990) presents with the
Zoogocho Zapotec data. The first comparison which will be made will be the
syntactic role played by the lexical argument in clauses with one lexical argument.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 9 7
Table 8.20 Syntactic role in clauses with one lexical argument in SBZZ
Syntactic role in clauses
with one lexical
argument
# %
S 412 49.9%
A 13 1.6%
O 401 48.5%
Total: 826 100%
Table 8.21 Syntactic role in clauses with one lexical argument in Yagua
Syntactic role in clauses
with one lexical
argument in Yagua
(adapted from
Payne: 1990:224)
# %
S 292 55%
A 39 7%
O 203 38%
Total: 554 100%
While the percentages for S and O are relatively similar, the reader can see
that there are far fewer instances of single lexical A ’s. This is an artifact of the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
grammar of SBZZ. In order for the A to be the only lexical item in the clause, the
structure of the clause would have to be either AVao, Vao A, or VA. However, the
last, unattested case, would either be ungrammatical or interpreted as intransitive.
However, the overall tendency towards O’s and S’s being lexical is by-and-large
confirmed.
Similarly, as seen in Table 21, the data for preverbal and postverbal lexical
subjects and objects is proportional to the data which Payne presents (ibid. 225). In
the following, sentences of the SVS type were taken to include preverbal subjects as
per (Payne 1990:224-225). However, in ditransitive clauses with both pre- and
postverbal objects disjoint in reference from each other, I consider each object
separately; when coreferential, the preverbal object was counted.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 9 9
Table 8.22 Pre- and postverbal subjects and objects in SBZZ and Yagua
Pre- and postverbal subjects and objects in SBZZ
Preverbal subjects 117 23.5%
Postverbal subjects 382 76.5%
Preverbal objects 100 21%
Postverbal objects 376 79%
Pre-and postverbal subjects and objects in Yagua
Preverbal subjects in
Yagua
122 32%
Postverbal
subjects(Payne
1990:225)
257 68%
Preverbal objects in
Yagua
65 26%
Postverbal
objects(Payne
1990:225)
186 74%
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4 0 0
I will now proceed to compare the SBZZ data with the data which Du Bois
presents for Sacapultec and which led to his conclusions regarding discourse patterns
which will be discussed below. The first data of interest to compare is the number of
lexical arguments. Not counting the clitic pronouns as lexical arguments, 1 obtained
the following chart.
Table 8.23 Number of lexical arguments (transitive and intransitive clauses
conflated)
0 1009 52.4%
1 835 44.4%
2 (incl SVS) 78 3.2%
3 1 <.1
These figures compare nicely with Du Bois’s figures for Sacapultec (47.6%
for 0 arguments, 51.2% for 1, and 1.1% for 2) (Du Bois 1987:818). This also follows
Polinsky’s data stating that VOS languages were more likely to have one or fewer
arguments than VSO languages (Polinsky 265). While there are considerably more
clauses with no arguments, this is an artifact of the methodology of excluding clitic
pronouns, though note that if the clitic pronouns had been included as ‘lexical’, this
would have been a meaningless exercise, as there would have been no clauses with
no lexical arguments. In any event, there are at least two significant differences, the
presence in the SBZZ texts of almost three times as many clauses with two lexical
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
arguments, and the presence of one clause with three lexical arguments. This is
further emphasized when tables 23 and 24 are compared.
Table 8.24 Number of lexical arguments per clause and transitivity in SBZZ
0 1 2 3 TOTAL
Intransitive 805
65.1%
412
33.3%
19(including
copulas)
1.6%
0 1236
Transitive 204
30.3%
414
61.6%
54
8.1%
0 672
Ditransitive 0 9
60%
5
33.3%
1
6.7
%
15
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4 0 2
Table 8.25 Number of lexical arguments per clause and transitivity in
Sacapultec
Sacapultec
(Adapted
from Du
Bois
(1987:819)
0 1 2 3 TOTAL
Intransitive 127
48.1%
137
51.9%
0 0 264
Transitive 84
46.9%
90
50.3%
5
2.8%
179
Ditransitive 0 0 0 0 0
Table 25 shows that SBZZ is a much more lexical language than Sacapultec:
there is a much larger proportion of clauses with two lexical arguments in S BZZ
(even excluding the copular clauses). A grammatical reason for this is that if a clause
in SBZZ has a lexical A then the O must be lexical as well, given that object clitics
only occur if there is a subject clitic already present. There were also a number of
ditransitive clauses including one with three lexical arguments. However, on another
level, the proportion of intransitive clauses with no lexical arguments to those with
lexical arguments was much greater than in Sacapultec. This is a result of there being
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4 0 3
no affixal agreement in Zoogocho Zapotec. One would imagine, although the
Sacapultec data is not presented, that there are a number of intransitive clauses with
overt pronominal subjects. These would be counted as lexical arguments in
Sacapultec, but would not in SBZZ, as they would, in speech, be cliticized. However,
the overall difference in lexicality is possibly due to the nature of the texts which
were studied. I will discuss this issue in more depth below.
In the following, I examined the type of data which led Du Bois to his
dictum, ‘Avoid Lexical A ’s’ (Du Bois 823). I included both objects in ditransitive
clauses and the one double object clause as a case of doubling. The data compares
interestingly with that of Du Bois. His data on Sacapultec has 48.1 % for lexical S’s,
compared with 29.6% for Zoogocho Zapotec. Although the difference is slight, it
would appear that the nominal/accusative bias in the SBZZ grammar lead to this
difference. However, as a whole, lexical A ’s were avoided.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4 0 4
Table 8.26 Grammatical role and syntactic type.
Morphological Lexical and free
pronoun
Clitic pronoun Doubled
S(ubject of 366 805 65
intransitive) 29.6% 65.1% 5.3%
A(gent of 29 619 39
transitive) 4.2% 90.1% 5.7%
0(bject of 445 216 11
transitive) 66.2% 32.1% 1.6%
0(bjects of 20 10
ditransitives) 66.7% 33.3%
Total 860 650 76
54.2% 41% 4.8%
Finally, as one last piece of comparison, examine Table 26. In Table 26, the
reader can see very similar data to that which is seen in Du Bois (ibid. 826). Keep in
mind that he had a third category, ‘accessible’, though this was really not significant
from this perspective for the core argument roles.1 4 Du Bois’ data appears in bold
following the Zapotec data. One issue which needs to be explained at some future
point is the difference between the SBZZ figures and the Sacapultec figures for
intransitive subjects. It is possible that the type of text plays a factor in this
difference. The longer of the two SBZZ texts was a conversation, and the first was an
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4 0 5
informal narrative history with some conversation. The texts which Du Bois
analyzed were all Pear Stories (Chafe et al 1980). The conversational style exhibited
by the SBZZ texts is much more contextual and as such there were many more given
intransitive subjects.
Table 8.27 Given/New Reference (By Syntactic Role)
Given New
S(ubject of intransitive) 1115 (89%) (72.5%) 140(11%) (22.5%)
A(gent of transitive and
ditransitive)
659 (95.9) (96.3%) 28 (4.1%) (3.2%)
0(bject of transitive) 483 (71.9%) (65.3%) 189 (28.1%)
(24.7%)
0(bjeets(primary and
secondary) of
ditransitive)
0 1 - 10, 02=6
01 =66.7%,02=40%
01=5, 02=9
01=33.3% ,02=60%
8.6 Topic and focus
Before going any further, it is imperative to discuss those structures which 1
have been tentatively describing as topic and focus. Before discussing and
classifying the actual examples in SBZZ, I will first present more generally a few
notions of focus and topic as discussed in the literature. First, I will present the
definitions which Thomas Payne has given (1997: 267-272).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
406
First, in defining focus, he describes the following ways in which the term ‘focus* is
used.
1) ‘Focus is a term applied to some morphosyntactic operation or category
whose function has not been adequately analyzed.
2) ‘Focus’ is a term applied to one element of every clause. In this approach,
focus can pretty much be equated with ‘new information’ or ‘asserted
information.’
3) ‘Focus’ describes a condition of some pragmatically marked clauses, other
clauses can be ‘focus-neutral’ or ‘unfocused.’ (ibid. 267)
While at present we are not interested in focus which applies to an entire clause, but
are interested in constituent focus, of which there are various types, including
assertive focus, conter-presuppositional focus (contrastive focus), and exhaustive
listing focus.
Assertive focus. S believes H has no knowledge of the information
‘They brought me this bowl of this thick, green, mushy stuff.
Counter-presuppositional focus ‘contrastive focus’
‘Sally and Robert came over last night, but SF1E got drunk.’
Exhaustive listing focus. That information which S asserts is unique in that
the rest of the clause is true only with respect to it and false with respect to all
other possible information
‘I drank only Pepsi at the party.’ (ibid. 269)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4 0 7
Furthermore, Payne states that:
[A] prototypical contrastive focus presupposes:
(a) a particular event E (taken loosely to mean any state of affairs)
occurred;
(b) there is a group of entities that might have had a role, R in E
(c) the addressee ‘incorrectly’ (in the eyes of the speaker) believes
that one of the entities did in fact have the role R. (ibid. 269)
The contrastive focus clause then asserts:
(a) the ‘correct’ identity of the entity involved, according to the
perception of the speaker;
(b) the proposition that the entity thought had the role R in fact did
not.(ibid. 269)
Before moving on to the discussion of topicalization, I will first examine the
structures I have so far called ‘focus’ over the course of the dissertation. Consider the
following example in which my primary consultant and the speaker of the second
text used in the quantitative analysis attempt to specify what type of grasshopper was
involved in a previous agricultural crisis.
(119)
AL bishe lagaha?
grasshopper leaf
‘The leaf grasshopper?’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4 0 8
(119 cont.)
P bi n-ak
no stat-be
‘That’s not it.’
AL bishe xa gashe na?
grasshopper clothes yellow demdist
‘The yellow grasshopper?’
P be yelhen xo
clan big old
‘the big, old ones’
AL dx-ago-shka benh yelh
cont-eat-emph genan milpa
‘It sure does eat up the milpa.’
P leka dx-ago=ba’=x
much cont-eat=3an=well
‘It surely does.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4 0 9
(119 cont.)
AL leka dx-om=ba’ che=to lashe ni
much cont-do=3an of=lplexcl field demprox
‘It does that to our (stuff) in the field here.’
dx-ago=ba yelh=en’
cont-eat=3an milpa=det
‘It eats the milpa.’
P lheba lheba b-e-gan
3sgan 3sgan comp-freq-stay
‘that’s it, that’s the one that stayed.’
Examples like (119) are what led me to consider these constructions to be focus
constructions. In (119), the speaker is clarifying which entity, out of a set of
potentially contrastive entities, was the correct one. Similarly, in (120), the speaker is
attempting to clarify a misunderstanding.
(120) bi zegh to-z=e’
neg stat.go one-emph=3f
‘She didn’t go by herself.’
chupe zghaak
two stat.pl.go
‘Two went.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
410
In those cases which have a fronted object with no resumptive pronoun,
analogous to the subject focus constructions, similar examples are found.
(121) AL bizx n-ak=en go-k chi=o’ kate
how stat-be=3inan comp-be of=2sg when
zh=o’, ‘bibi yeen xogh chi=a’ de’
cont.say=2sg, ‘neg plate sauce of=Tsg cop
‘What is it that happened to you when you said “I don’t have a plate for
salsa.’”
‘na b-en=a’ yeghxogh’ yegh bin
‘and comp-make=T sg stone sauce’ yegh gensm
zh=o’
cont.say=2sg
“ ‘and I made a mortar and pestle” a stone is what you said’
bin zh=o’ dx-ak=da’
gensm cont.say=2sg cont-think=lsgexp
‘That’s what I think you said’
P uhum yegh
affirmative stone
‘yes, stone’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4 1 1
(121 cont.)
to yegh=dao g-cheen=a’ g-cheen=a’ na’
one stone=dim comp-carve=lsg comp-carve=lsg and
b-en=a=n yegh xogh
comp-make=lsg=3inan stone salsa
‘ A little stone, I carved it and carved it and made it into a mortar and pestle.’
What I have been calling ‘topicalization’ is much more difficult to define. It is
structurally a left-dislocated argument, with a resumptive pronoun following the
verb. Often the left-dislocated element is a pronoun as in the following.
(122) na neda zhagalaw=a’ zghe-la=a’ kafe che benhe
and lsg hurry=lsg stat.and-clean=lsg coffee of people
‘and I am rushing to clean people’s coffee’
Sometimes it is a full NP, as in (123).
(123) kelio kabi dx-e-la=be’ ke?
Aquelio neg cont-freq=arrive=3inf really
‘Aquelio doesn’t come, right?’
In general, constructions such as these are used to emphasize that a participant
behaved in a certain way. Even in a relatively pragmatically unmarked example such
as (124), the preverbal placement of benhe with the following resumptive pronoun is
used in such a way, indicating what it is that people give for payment for work.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4 1 2
(124) benhe go-nh=e’ no medxo
person pot-give=3inf indef money
‘ People might give some money.’
benhe go-nh=e’ panelh
person pot-give=3inf su g arlo af
‘People might give sugar loaf.’
A better term for this construction might be to call it a ‘highlighting’ construction, as
used by Ann Cooreman when discussing Chamorro (Cooreman 1992).
8.7 Conclusions
The hypothesis stated above is restated below:
Hypothesis: The length of subjects in VSO structures w ill be shorter than or
equal to the length o f objects, in order to facilitate the recognition of the VP.
This has been strongly confirmed. There is an additional wrinkle in the theory
however which has been unearthed during the analysis of the data described above.
Consider Du Bois’ Table 9 ‘Dimensions and constraints of Preferred Argument
Structure’ (ibid. 829), repeated here as Table 28.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 8.28 Du Bois’ Table 9 ‘Dimensions and constraints of Preferred
Argument Structure’ (ibid. 829)
GRAM M AR PRAGM ATICS
QUANTITY One Lexical Argument
Constraint
One New
Argument
Constraint
ROLE Non-lexical A Constraint Given A
Constraint
In table 8.28, there are four inter-related constraints. The first constraint, the
‘One Lexical Argument Constraint’ states that it is preferential to the grammar for
there to only be one lexical argument per clause. The ‘Non-lexical A Constraint’
states that this argument is not likely to be an agent of a transitive clause. The ‘Given
A Constraint’ states that the agent of a transitive clause is likely to be given. Finally,
the ‘One New Argument Constraint’ states that it is likely that there is only one new
argument per clause. These constraints have been verified for many languages,
SBZZ among them. Consider Tables 22 and 23 which confirm the ‘One Lexical
Argument Constraint’, Table 25 which confirms the ‘Non-lexical A Constraint’, and
Table 26 which confirms the ‘Given A Constraint’. The ‘Given A Constraint’
directly leads to the ‘One New Argument Constraint’; as, if the A is given and not
new, there can be at most one other new argument per clause. These constraints arc
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4 1 4
given solely in terms of pragmatics and grammar. There is, however, one other
aspect which should be considered: processing.
Accessibility theory, as argued for by Ariel (1990) is basically concerned
with the fact that ‘the choice o f a referring expression is dependent on the
Accessibility status the mental representation of the referent is assumed to have for
the addressee at the current stage of the discourse’ (ibid. 69). She arrives at the
following scale.
Table 8.29 Accessibility M arking Scale (Ariel 1990;73)
Low Accessibility
Full name=modifier
Full (‘namy’) name
Long definite description
Short definite description
Last name
First name
Distal demonstrative=modifier
Proximal demonstrative=modifer
Stressed pronoun ^gesture
Stressed pronoun
Unstressed pronoun
Cliticized pronoun
Extremely High Accessibility Markers (gaps, including
pro, PRO and wh traces, reflexives, and Agreement...)
High Accessibility
In general, if one impressionistically considers the scale above in terms of the
length of the expression, it is, for the most part, clear that the more accessible an item
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4 1 5
is, the shorter it is. Similarly, less accessible markers are, by and large, longer. This
can be related to the fact that given references are generally shorter than new
references. Du Bois’s Preferred Argument Structure Constraints above could
potentially be considered the result of processing constraints as well. The ‘Avoid
Lexical A Constraint’ in a verb initial language (like SBZZ, Yagua, and Sacapultec)
could easily be viewed as being a processing constraint, as a corollary to the
hypothesis stated above in that, if the A is lexical, there is more intervening material
between the verb and the object and the constituent recognition domain is not as
maximized as it is when the A is non-lexical. The ‘Given A Constraint’ is even more
strongly motivated by such a processing approach, as a New A is more likely to be
longer, and if the A is given it is more likely to maximize the constituent recognition
domain. The ‘Given A Constraint’ leads to the ‘One New Argument Constraint’ and
similarly the ‘One Lexical Argument Constraint’ could potentially be related to the
‘Avoid Lexical A Constraint’.1 5
One final issue to contemplate, as promised, is the issue of the status of the
clitic pronouns. I feel that there are arguments for considering them to be both affixal
agreement and clitic pronouns. Given the hypothesis which was confirmed the
current chapter (that Subjects would be shorter than Objects in sentences with VSO
ordering) and given the discourse motivations which Du Bois (1987) discovered and
which were discussed above and put forth as Table 27, there is a great pressure for
the subject to be pronominal, and reduced. In fact, over 73% of all of the clauses
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
416
which were counted in the corpus used in the current chapter had a clitic pronoun
subject.
The question remains however, at what point do these pronominal clitics
become agreement affixes on verbs. I believe that we are seeing a morphological
class in transition in SBZZ. In addition to the pragmatically marked structures with
repeated clitics such as (125)-(128), there are also cases like (129) and (130).
(125) per to-z=e’ dx-egh=e’
but one-emph=3f cont-go=3f
‘But she goes alone...’
(126) b-lei=dx=da’=n
comp-see=emph=T sgexp=3inan
neda’
lsg
‘I still saw it.’
(127) pur dizha
only language
dxop=dxo
two=lplincl
(128) nhe ka
neg demadv
zxon
zapotec
na ne=dxo
now stat.speak=lplincl
gu-ne-z=be’ da liaventura
comp-speak-emph=3inf deceased Buenaventura
‘The late Buenaventura didn’t say that.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4 ! 7
(129) n-ak=ba to kabayw shish
stat-be=3an one horse white
‘There is a white horse.’
(130) tonhe-dxgwa n-ak=e ’ jef=en kleka benhe
yeto
tall-int cont-be= 3f boss=det than person
other
‘The boss is taller than the other person.’
Examples (129) and (130) indicate the initial stages of the grammaticalization o f
agreement markers. Note that both o f these examples have copular verbs, which
would be a perfect place for such a change to begin.
In this chapter, I hope to have presented some idea o f the typology of verb
initial languages. In addition, I have confirmed a hypothesis concerning the
processing of VSO structures and have also by-and-large confirmed the findings of
other textual studies of verb initial languages. This study is unique in that it is the
only study of word-order variation in a verb initial language that I know of that nol
only looks at the information status of the individual noun phrases, but also looks at
their weight, and as such represents a major contribution to both processing theory
(which has not previously had such a textual study conducted on a verb initial
language) and to the theory of word order variation in verb initial languages.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 See 5.6 for m ore inform ation.
2 I will discuss her im portant claim s in the next section o f the current chapter.
3 N ote that all o f that w hich is italicized and follow s in the Payne discussion is quoted verbatim from (Payne
1990: pp 11-15), the only changes w hich have been m ade are the italicization and quotation m arks put around
sentences w hich w ere italicized in the original.
4 See 5.4.3 for a discussion o f ditransitive constructions in Z oogocho Zapotec.
5 Taking ‘oblique’ to m ean a non-direct object.
6 N om inative=A & S m arked the same, A ccusative= 0 m arking, A bsolutive= S& 0 m arked the sam e. E rgative-A
m arking.
7 By gram m atical, Du Bois refers to instances, like we have seen in SBZZ, where, for exam ple, a Spanish verb is
borrow ed as a verbal argum ent, and show s up as the com plem ent o f a transitive verb (such as do).
8 In these cases mS and mO represent that S and O are constructed with their Im m ediate C onstituent constructed
with the lexical item on the left periphery.
9 N ote that these tw enty m entions included stray N P ’s and other such things which were not encoded in this
version.
1 0 W hile this is potentially a problem for the overall w eight o f verb phrases, it should be noted that there were a
statistically insignificant num ber for w hich this w ould be a problem . There w ere also few er than 10 incidents
where tw o verbs (-ak ‘to becom e’ and -zoalao- ‘to begin’) occurred with no argum ents w hatsoever, contrary to
w hat one w ould norm ally expect. See H uang (2000) for a discussion o f the increased likelihood o f auxiliary verbs
to allow null subjects. See 6.5.4 for m ore inform ation.
1 1 From this point further, I am labeling the subject o f an intransitive with an S, the subject o f a transitive or
ditransitive with an A, and an object with an O. For ditransitive clauses, 1 assign the labels O l and 0 2 based on
w hich one occurred first in the clause. I am using lower-cased letters to represent clitic pronouns and upper ease
letters to represent lexical and full pronom inal argum ents.
1 2 For the purpose o f this individual chart, I included the reflexive o f possessor construction and the AOVI ,()(’a
construction in the calculation. From this point onw ards, 1 will not include either o f those constructions in the
generalized charts.
1 3 N ote that SBZZ, as a VO language w ould follow the sam e predictions.
1 4 For S(ubject o f intransitive) the accessible w ere 5% o f the total and for A(gent o f transitive), .5%.
1 5 Both o f these im plications follow from Du B o is’s theory and have nothing to say about processing.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
419
Chapter 9: Conclusion
This dissertation, a descriptive grammar of San Bartolome Zoogocho
Zapotec, is now concluded. This typological grammar is the first of its kind for a
Zapotecan language. I have striven to include not only grammatical judgements, but
also to use textual performance data whenever possible. The initial six chapters
describe the major grammatical features of the language, while the final two examine
two major current theoretical issues: parts-of-speech and word order.
The first six chapters provide descriptions of the ethnographic and
sociolinguistic situations of the Zoogocho Zapotec community, the sounds and
orthography of the language, the pronominal system, the morphology, and the syntax
of the language. While no particular theoretical framework was used, the inspiration
for much of the descriptions comes from the typological universal grammar research
program.
San Bartolome Zoogocho Zapotec is a tonal language which can be quite
complex phonologically. It is an agglutinative, slightly fusional language. As a
prototypical VSO language, it has prepositions, NAdj, NDem, NGen, and NRel
orders. Various means of combining clauses exist; including complementation,
coordination, and relativization. While this is the most complete grammatical
description of SBZZ to date, there is still much work to be done, especially in the
areas o f phonology and phonetics.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Chapter Seven examines the lexical classes present in SBZZ. While 1 try to
define necessary and sufficient conditions for each lexical class, it can be difficult to
find conditions which are both necessary and sufficient. Therefore, 1 have to rely on
multiple definitions and a multifactorial approach to the idea of lexical class. 1 try to
use definitions which, while being informed by a variety of cross-linguistic data, arc
based on and presented by the SBZZ grammar. I devote much of the discussion to
the grammaticalization of relational nouns, a topic which has received a great deal of
discussion in the literature, both specifically for Zapotecan and Otomanguean
languages and more generally. I conclude that relational nouns are indeed a separate
category from prepositions and regular nouns in SBZZ, sharing characteristics with
both. Principled comparative work on the historical development of lexical
categories in Zapotecan, while out of the scope of the present work, remains an
interesting topic which will shed much light on the processes of grammaticalization
both within the Otomanguean family and more generally crosslinguistically.
Chapter Eight is an examination of verb initial word order, and both places
SBZZ in two typologies of verb initial languages and compares a study of word
order in two SBZZ texts with other textual studies of word order in other verb initial
languages. This study is the first of its kind for a verb initial language to take into
account phrasal length, and as such is a major contribution not only to textual studies
of word order, but also to processing theory, as it confirms a hypothesis about the
processing of VSO languages which comes from the processing theory of John A.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Hawkins. This chapter concludes with a discussion of pronominalization in San
Bartolome Zoogocho Zapotec based on the quantitative textual performance data,
central issue not only for SBZZ, but also for Zapotecan languages in general.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
422
B ib l io g r a p h y
A ik h en v a ld, A lexandra Y. 2002. Typological parameters for the study o f clitics, with special
reference to Tariana. Word: A Cross-linguistic Typology, ed. by Dixon, R.M.W. and Alexandra Y.
Aikhenvald, 42-78. Cambridge: Cambridge University Press.
A issf. n , Judith L. 1987. Tzotzil Clause Structure. Dordrecht: Reidel.
A lS S E N , J U D IT H . 1999. External Possessor and Logical Subject in Tz’utujil. External Possession, ed.
by Doris L. Payne and Immanuel Barshi,167-193. Amseterdam: Benjamins.
A N D E R S O N , S te p h e n R. 1992. A-Morphous Morphology. Cambridge: Cambridge University Press.
A riel, M ir a. 1990. Accessing Noun-Phrase Antecedents. London and N ew York: Routledge.
A v elin o, H er ib er to. 2004. Topics in Yalalag Zapotec Grammar, Dictionary and Texts. Los
Angeles: University o f California, Los Angeles dissertation.
Av elin o, H eriberto, John O. For em a n, Pam ela M u n r o, and Aaron H uey Sonnensctihin.
2004. Covert Subjects in Zapotecan. Paper presented at the annual meeting o f the SSSILA,
Boston.
Ba r th o l o m ew , Do r is. 1983. Gramatica zapoteca. Diccionario zapoteco: Zapoteco de Juarez, by
Neil N ellis and Jane Nellis. M exico City: Institute Linguistico de Verano.
Bla ck, C heryl A. 1994. Quiegolani Zapotec syntax. Santa Cruz, CA: University o f California. Santa
Cruz dissertation.
B L A C K , C h e r y l A. 1996. A Backwards Binding Construction in Zapotec. 1996 Work Papers o f the
Summer Institute o f Linguistics, 73-87. Bismark: University of North Dakota.
Bla ck, C heryl A. 2000. Quiegolani Zapotec Syntax: A Principles and Parameters Account.
Arlington: SIL International and University o f Texas at Arlington.
Beals, Ralph L eon. 1975. The Peasant Marketing System O f Oaxaca, Mexico. Berkeley: University
O f California Press.
BeltrAn Mo r a les, Filem o n. 1982. Medicina tradicional en la comunidad zapoteca de Zoogocho,
Oaxaca. Mexico: Mexico City: Institute Nacional lndigenista- Secretarfa de Educacion Publica-
Centro de Investigaciones Superiores en Antropologia Social.
B e n v e n is t e , E m il e . 1968. Mutations o f linguistic categories. Directions for Historical Linguistics,
ed. by W.P. Lehmann and Y. Malkiel, 85-94. Austin: University o f Texas Press.
Berg, Richard Lew is, Jr . 1974. El impacto de la economta moderna sobre la economia tradicional
de Zoogocho, Oaxaca y su area circundante. M exico : Institute Nacional lndigenista.
Ber in stein, A va. 1985. Evidence for Multiattachment in K'ekchi Mayan. New York: Garland.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4 2 3
BOHNEMEYER, J u e r g e n . 1998. Time relations in discourse. Evidence from a comparative a p p ro a c h to
Yucatek Maya. Brabant: Katholieke Universiteit Brabant dissertation.
B r iggs, E lin o r.1961. Mitla Zapotec Grammar. M exico : Instituto Lingiiistico de Verano.
Br u g m a n, C laudia and M onica Ma c a u le y. 1986. Interacting Semantic System s: Mixtec
Expressions o f Location. Proceedings o f the Twelfth Annual Meeting o f the Berkeley Linguistics
Society, ed. by Vassiliki Nikiforidou, Mary Van Clay, Mary Niepokuj, and Deborah Feder, 3 15-
327. Berkeley: Berkeley Linguistics Society.
Bu tler, Inez M. 1976. Reflexive Constructions o f Yatzachi Zapotec. International Journal o f
American Linguistics 42. 331-337.
Bu tler, In e z M. 1980. Gramatica zapoteca : zapoteco de Yatzachi, El Bajo, Serie de Gramaticas de
Lenguas Indigenas de M exico Num. 4. Mexico: Instituto Lingiiistico de Verano
Butler, Inez M. 1985. Event prominence in Zoogocho Zapotec narrative discourse. S1L Mexico
Workpapers 7. 16-60.
By b ee, Jo a n. L. 1985. Morphology. Amsterdam :John Benjamins Publishing Company.
Bybee, Joan L.W illiam P agliuca, and Revere D. Perkin s. 1990. On the asymmetries in the
affixation o f grammatical material. Studies in Typology and Diachrony, Papers presented to
Joseph H. Greenberg on his 75th birthday, ed. by William Croft, Suzanne Kemmer, and Keith
Denning, 1-43. John Benjamins: Amsterdam.
By b ee, Joan L., Revere D. Perkins, and W illiam Pa g liu c a. 1994. The evolution o f grammar:
tense, aspect, and modality in the languages o f the world. Chicago: University o f Chicago Press.
Ca m pbe ll, L y le. 1991. The explanation o f syntactic change: a historical perspective. Historical
Linguistics 1991 Papers from the 10th International Conference on Historical Linguistics, ed. by
Jaap Van Marie, 49-70. Amsterdam/Philadelphia: John Benjamins.
C am pbell, Lyle, T errence K aufm an and T homas Sm ith-St a r k. 1986. Mesoamerica as a
linguistic area. Language 62. 530-570.
Ca r sta ir s, A n d r ew . 1981. Notes on affixes, clitics, and paradigms. Bloomington:Indiana University
Linguistics Club.
Castellan os M artin ez, Ja vier, et a l . 1995. Diccionario Zapoteco-Espanol: Reglas para el
entendimiento de las variantes dialectales de la sierra. Solaga Oaxaca: Zanhe Xbab.
C h a f e , W a l l a c e L ., R o y F r e e d l e , and J o h n W. D u B o is . 1980. The Pear Stories: Cognitive.
Cultural, and Linguistic Aspects o f Narrative Production. Norwood, NJ : Ablex.
Ch u n g, S. 1998. The Design o f Agreement: evidence from Chamorro. Chicago: University ol
Chicago Press.
Co m r ie, B er n a r d. 1989. Language Universals and Linguistic Typology. 2nd Edition. Oxford: Basil
Blackwell.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
424
Co o r em a n, An n e. 1992 The Pragmatics o f Word Order Variation in Chamorro Narrative Text.
Pragmatics o f Word Order Flexibility, ed. by Doris L Payne, 243-263. Amsterdam: Benjamins.
Co r d o v a, Fr . Ju a n d e . 1987 [1578a], Vocabulario en lengua ?apoteca. Mexico: Ediciones Toledo
(Instituto Nacional de Antropologia e Historia).
Co r d o v a, F r. J uan d e. 1886 [1578b], Arte del idioma zapoteco. Morelia: Imprenta del Ciobierno.
C r o f t , W il l ia m . 1985. Indirect object Towering’. The Proceedings o f the Eleventh Annual M e e tin g
o f the Berkeley Linguistic Society, ed. by Mary Niepokuj, Mary Van Clay, Vassiliki Nikiforidou,
and Deborah Feder, 39-51. Berkeley: Berkeley Linguistics Society.
C r o f t , W illia m . 2000. Parts o f speech as language universals and as language-particular categories.
Approaches to the Typology o f Word Classes, ed. by Petra M. and Bernard Comrie, 65-103.
Berlin/New York: Mouton de Gruyter.
D a y l e y , JON P. 1985 Tzutujil Grammar. Berkeley : University o f California Press.
D e A N g u l o , J a i m e . 1926a. The Linguistic tangle o f Oaxaca. Language 1. 96-102.
De A n g u lo, Ja im e. 1926b. The Development o f affixes in a group o f monosyllabic languages o f
Oaxaca. Language 2. 46-61,119-33.
D iessel, H o lg er. 1999. Demonstratives: Form, Function, and Grammaticalization. Amsterdam: John
Benjamins Publishing Company.
D i x o n , R.M.W. 1972. The Dyirbal language o f North Queensland. Cambridge: Cambridge
University Press.
D ixon R.M .W . and A lexandra Y. A ik h en v a ld. 2002. Word: A Cross-linguistic typology.
Cambridge University Press: Cambridge.
D ixon R.M.W. and A lexandra Y. A ik h en v a ld. 2002. Word: a typological framework. Word: A
Cross-linguistic typology, ed. by R.M.W. Dixon and Alexandra Y. Aikhenvald. Cambridge:
Cambridge University Press.
D o n a h u e, M a r k. 1995. The Tukang Besi Language. Canberra: Australian National University
dissertation.
D r y er, M.S. 1997. On the six-way word order typology. Studies in Language 21. 69-103.
Du B ois, J o h n W. 1987. The discourse basis for ergativity. Language 63. 805-855.
FernAn d e z de M iranda, M a . T er esa. 1995. El protozapoteco. Mexico : Colegio de Mexico-
Instituto Nacional de Antropologia e Historia.
Fo r e m a n , J o h n . In progress (2004). The Morphosyntax o f Subjects in Macuiltianguis Zapotec. Los
Angeles: University o f California, Los Angeles Dissertation.
FORIS, D a v id. 1994. A Grammar o f Sochiapan Chinantec. Auckland, N ew Zealand: University o f
Auckland dissertation.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Fuen te, J ulio de la. 1977. Yalalag : una villa zapoteca serrana. Mexico : Instituto Nacional
lndigenista.
G iv o n , T a l m y . 1977. The Drift from VSO to SVO in Biblical Hebrew. Mechanisms o f Syntactic
Change, ed. by Charles N. Li, 181 -254. Austin and London: University o f Texas Press.
G ivon, T a l m y . 1983. Topic Continuity in Discourse: A Quantitative Cross-language Study.
Amsterdam: John Benjamins Publishing Company.
G iv on, t a l m y . 1984. Syntax. A functional-typological introduction. Volume 1.
Amsterdam/Philadelphia: John Benjamins Publishing Company.
G iv on, T alm y. 1991. The evolution o f dependent clause morpho-syntax in Biblical Hebrew.
Approaches to Grammaticalization Vol. ii. ed. by Elizabeth Closs Traugott, and Bernd Heine, 2 5 7 -
310. Amsterdam: John Benjamins.
G r een b er g, J.H. 1963. Some universals o f grammar with particular reference to the order o f meaning!
elements. Universals o f language, ed. by J.H. Greenberg, 73-113. Cambridge, Mass.: MIT Press.
G r im es, Barbara F. and Joseph E. G rim es, (eds). 2000. Ethnologue. Volume 1: languages o f the
world; volume 2: maps and indexes. Dallas: SIL International.
Ha eg em a n, L iliane. 1993. Introduction to Government and Binding Theory. Cambridge: Blackwell
Pulbishers.
Ha im a n, Jo h n. 1983. Iconic and Economic Motivation. Language 59. 781-819.
H a im a n , JOHN. 1985. Natural syntax: iconicity and erosion. Cambridge/New York : Cambridge
University Press.
Ha w k in s, Jo h n.A. 1983. Word Order Universals. Academic Press. New York.
Ha w k in s, Jo h n.A. 1988. Explaining language universals. Oxford: Blackwell.
Ha w k in s, John A .1994. A performance theory o f order and constituency. Cambridge, New York
Cambridge University Press.
H a w k in s , J o h n A. 1999. Processing complexity and filler-gap dependencies across grammars .
Language 75. 244-285.
H a w k in s , John A. 2001. Why are categories adjacent?. Journal o f Linguistics 37. 1-34.
H a w k in s , J o h n A . 2 0 0 2 a . S y m m e trie s a n d A sy m m e trie s: T h e ir G ra m m a r, T y p o lo g y a n d P a rsin g .
Theoretical Linguistics 28. 95-150.
H a w k in s , J o h n A. 2002b. Issues at the Performance-Grammar Interface: Some Comments on the
Commentaries. Theoretical Linguistics 28. 211-226.
H a w k in s , J o h n A. 2003. Efficiency and Complexity in Grammars: Three General Principles. T h e
Nature o f Explanation in Linguistic Theory, ed. by John Moore and Maria Polinsky. Stanford:
CSLI Publications.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4 2 6
H eine B ernd , U lrike C laudi, and Friederik e H u en n em ey er. 1991. Grammaticalization A
Conceptual Framework, Chicago: The University o f Chicago Press.
H ein e, B er n d. 1997. Possession: cognitive sources, forces, and grammaticalization. Cambridge/New
York: Cambridge University Press.
H en g ev eld, K ees. 1992. Parts o f Speech. Layered structure and reference in a functional
perspective, ed. by Michale Fortescue, Peter Harder & Lars Kristofferson, 29-56. Amsterdam:
Benjamins.
Ho llen b a c h, Barbara E. 1990. Semantic and syntactic extensions o f Copala Trique body-parl
nouns. Homenaje a Jorge Suarez, ed. by Beatriz Garza Cuaron and Paulette Levy, 275-296.
Mexico: Colegio de Mexico.
Ho llen b a c h, B arbara E. 1995. Semantic and Syntactic Extensions o f Body Part terms in
Mixtecan: The Case of'Face’ and ’Foot’. International Journal o f American Linguistics 61.168-
190.
Ho m b er t, Jean-M a r ie. 1978. Consonant Types, Vowel Quality, and Tone. Tone: A Linguistic
Survey, ed. by Victoria A. Fromkin, 77-111. N ew York: Academic Press.
Ho ppe r, Pa u l. 1987. Emergent Grammar. Berkeley Linguistics Society: Proceedings o f Hie
Thirteenth Annual Meeting, ed. by Jon Aske, Natasha Beery, Laura Michaelis, and Hana Filip.
139-157. Berkeley: Berkeley Linguistics Society.
Ho ppe r, P.J. and T hom pson, S.A. 1984. The discourse basis for lexical categories in Universal
Grammar.Language 60. 703-752.
HOPPER, Paul and Elizabeth Closs T r a u g o tt. 1993. Grammaticalization. Cambridge: Cam bridge
University Press.
HUANG, Y an. 2000. Anaphora: A cross-linguistic study. Oxford: Oxford University Press.
http:7www.ini.gob.m x/indica2000/mpo/oax83.htm.
JAEGER, J er i. 1983. The Fortis/lenis Question: Evidence from Zapotec and Jawon. Journal o f
Phonetics 11. 177-189.
Je n se n d e L o p e z , K ristin e . 2002. Baskets and Body-Parts: A Cross-cultural and Crosslinguistic
investigation o f Children's development o f Spatial Cognition and Language. Aarhus: Aarhus
University dissertation.
J e s p e r s e n , O t t o . 1924. L a n g u a g e; its n a tu re , d e v e lo p m e n t, a n d o rig in . N e w Y o rk : H . H olt.
Ka u fm a n, T er re n ce. 1994. Proto-Zapotec Reconstructions, ms.
Ka u fm a n, T errence, 2004. ms.
Ka u fm a n, T erren ce, n.d. The Phonology and Morphology o f Zapotec Verbs, ms.
K eena n, Ed w a r d. 1977. Summary o f word order typology, ms. UCLA.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
427
K e e n a n , E d w a r d . 1979. Word order typologies: The verb initial typology, ms. UCLA.
K e e n a n , E d w a r d . Relative Clauses. Language Typology and Syntactic Description ,Volume 11, ed.
by Timothy Shopen, 141-170. Cambridge: Cambridge University Press.
K e e n a n , E d w a r d and B e r n a r d C o m r ie . 1977. Noun-phrase accessibility and universal grammar.
Linguistic Inquiry 8. 63-99.
K o p t j e v s k a j a -T a m m , M a r ia . 1993. Nominalizations. London: Routledge.
L a d e f o g e d , P e t e r and Ia n M a d d ie s o n . 1996. The sounds o f the world’s languages. Oxford:
Blackwell.
L a k o f f , G e o r g e and M a r k J o h n s o n . 1985. Metaphors we live by. Chicago: University o f C h ic a g o
Press.
L a r s o n , R.K. 1988. On the double object construction. Linguistic Inquiry 19. 335-391.
L e a l , M a r y . 1950. Patterns o f tone substitution in Zapotec morphology. International Journal o f
American Linguistics 16. 132-136.
L e a l , M a r y and O t is L e a l . 1954. Noun possession in Villa Alta Zapotec. International Journal o f
American Linguistics 2 0 . 2 1 5 -1 6 .
L e e , F e l ic ia A. 2003. Anaphoric R-Expressions as Bound Variables. Syntax. 6. 84-114.
LEE, P e n n y . 1996. The Whorf Theory Complex: A critical reconstruction. Amsterdam: John
Benjamins Publishing Company.
L e h m a n n , C h r is t ia n . 1982. Thoughts on Grammaticalization: A Programmatic Sketch. Vol. I
(Arbeiten des Koelner Univesalien-Projekts 48). Koeln:Universitat zu Koln. Institut fuer
Sprachwissenshaft.
L e h m a n n , C h r is t ia n . 1999. Possession in Yucatec Maya. Structures, functions, typology. M u n ich :
Lincom Europa.
L e v in s o n , S t e p h e n C . 1996. Language and Space. Annual Review o f Anthropology 2 5 . 3 5 3 -3 8 2 .
LlLLEHAUGEN, B r o o k . 2 0 0 3 . The Categorial Status o f Body Part Prepositions in Valley Z a p o te c
Languages. Los Angeles: UCLA MA Thesis.
LONG, R e b e c c a . 1985. Topicalization in Zoogocho Zapotec expository discourse. S IL Mexico
Workpapers 7. 61-100.
LONG C ., R e b e c c a and S o f r o n io C r u z M. 1999. Diccionario Zapoteco de San Bartolome Z o o g o c h o ,
Oaxaca. Serie de vocabularios y diccionarios indigenas "Mariano Silva y Aceves" 38. Mexico.
Instituto Lingiiistico de Verano.
L o p e z L., F il e m o n and R o n a l d o N e w b e r g . 1990. La conjugation del verbo zapoteco : zapoteco dc
Yalalag. M exico : Instituto Lingiiistico de Verano Mexico.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4 2 8
L y o n s , J o h n . 1971. Introduction to Theoretical Linguistics. Cambridge : Cambridge University
Press.
M a c a u l a y , M o n ic a . 1987. Cliticization and the morphosyntax o f Mixtec, International Journal o f
American Linguistics 53. 1 19-135.
M a c a u l a y , M o n ic a . 1996. A Grammar o f Chalcatongo Mixtec. London: University o f California
Press.
M a c LAURY, R o b e r t E. 1989. Zapotec Body-part Locatives: Prototypes and Metaphoric E x te n sio n s,
International Journal o f American Linguistics 55. 119-154.
M a r l e t t , S t e p h e n A. 1985. Some aspects ofZapotec clausal syntax.. Workpapers o f the Summer
Institute o f Linguistics. University o f North Dakota.
M a r l e t t , S t e p h e n A. 1993. Zapotec Pronoun Classification. International Journal o f American
Linguistics 59. 82-101.
M a r l e t t , S t e p h e n and V e l m a P ic k e t t 199 9 El pronombre inaudible en el zapoteco del istmo. m s.
Instituto Lingiiistico de Verano.
M a r l e t t , S t e p h e n A. and V e l m a B. P ic k e t t . 2001. Pluralization in Zapotec Languages
(http://www.sil.org/mexico/zapoteca/G025a-ZapotecPlurals.htm)
M a t t h e w s , P.H. 1991. Morphology: An introduction to the theory o f word-structure. Cambridge.
Cambridge University Press.
M a t t h e w s , P.H.. 2002. What can we conclude?. Word: A Cross-linguistic typology, ed. by R.M.W.
Dixon and Alexandra Y. Aikhenvald. Cambridge: Cambridge University Press.
M it h u n , M a r r ia n e . 2000. Noun and verb in Iroquoian languages: Multicategorisation from m u ltip le
criteria. Approaches to the Typology o f Word Classes, ed. by Petra M. and Bernard Comrie, 65-
103. Berlin/NewYork: Mouton de Gruyter.
M o l in a C r u z , M a r io . 1996. Volcan de petalos : ya'byalhje xtak yeje. Mexico: Letras Indigenas
Contemporaneas.
M uN R O , PAMELA 19 9 7 C o u rse n o te s a n d p e rso n a l c o m m u n ic a tio n .
M ij n r o , P a m e l a , and F e l ip e H . L o p e z , with O l iv ia V . M e n d e z [M a r i In e z ], R o d r ig o G a r c ia ,
and M ic h a e l R. G a l a n t . 1999. Di'csyonaary X:tee'n Dii'zh Sah Sann Lu'uc (San Lucas Q u ia v in i
Zapotec Dictionary / Diccionario Zapoteco de San Lucas Quiavini). Los Angeles: (UCLA)
Chicano Studies ResearchCenter Publications.
M u n r o , P a m e l a . 1995. Coreference Phenomena in San Lucas Quiavini Zapotec. Ms.
M u n r o , P a m e l a . 2 0 0 2 . From Parts o f Speech to the Grammar, m s. presented at the S y m p o siu m on
Grammatical Description o f Undocumented Languages. Dallas : SIL. 19 October 2002.
N e l l is , N e il and J a n e N e l l is . 1983. Diccionario zapoteco: Zapoteco de Juarez. M exico City:
Instituto Lingiiistico de Verano.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
429
N ic h o l s , J o h a n n a . 1986. Head-marking and dependent-marking grammar. Language 6 2 : 5 6 - 119.
N ic h o l s , J o h a n n a . 1992. Linguistic diversity in space and time. Chicago:University o f Chicago
Press.
PAYNE, D o r is L. 1990. The Pragmatics o f Word Order: Typological Dimensions o f Verb Initial
Languages. Berlin: Mouton de Gruyter.
P a y n e , T h o m a s . 1997. Describing Morphosyntax. Cambridge: Cambridge University Press.
P ik e , E u n ic e V ic t o r ia . 1948. Problems in Zapotec tone analysis. International Journal o f American
Linguistics 14. 161-170.
P ik e , K e n n e t h . 1944. Analysis o f a Mixteco text. International Journal o f American Linguistics
10:113-138.
P o l in s k y , M. 1997. Dominance in precedence: SO/OS languages. Papers from the 33rd Regional
Meeting o f the Chicago Linguistic Society, ed. by Kora Singer, Randall Eggert, and Gregory
Anderson, 254-269. Chicago: University o f Chicago Press.
PRIDE, K it t y . 1965. Chatino Syntax. Norman, OK: Summer Institute o f Linguistics.
R a m o s P io q u in t o , D o n a t o . 1991. Migracion y Cambios Socioeconomicos En La Comunidad De
Zoogocho,Oaxaca. Estudios Demograficos y Urbanos, 6. 313-345.
R e in h a r t , T . 1981. Pragmatics and Linguistics: An Analysis o f Sentence Topics. Philosophic;! 27.
5 3 -9 4 .
Rios M o r a l e s , M a n u e l . 1994a. Zapotecos de la Sierra de Norte de Oaxaca. Sint. Ma. T e re sa
Ruiz.Mexico: Instituto Nacional lndigenista - Secretaria de Desarrollo Social.
R io s M o r a l e s , M a n u e l . 1994b. Los zapotecos de la Sierra Norte de Oaxaca : antologia e tn o g ra llc a .
Mexico: Centro de Investigaciones Superiores de Antropologia Social.
SADDOCK, J e r r o l d . 1991. Auto-lexical Syntax, Chicago: The University o f Chicago Press.
S a p ir , E d w a r d . 1921. Language: An introduction to the study o f speech. New York: Harcourt B race
& World.
SASSE, H a n s -J u e r g e n . 1988. Der Irokesische Sprachtyp, Arbeitspapiere des Insituts fu e r
Sprachwissenschaft-Universitaet zu Koeln 9.
S c h a c h t e r , P a u l . 1985. Parts-of-speech systems. Language Typology and Syntactic Description,
Vol. I, ed.by Timothy Shopen, 3-61. Cambridge: Cambridge University Press.
S e il e r , H a n s j a k o b .1 9 8 2 . Possession : as an operational dimension o f language. Tubingen
[Germany] : G. Narr.
S o n n e n s c h e in , A a r o n , pending. Attributive possession in Zoogocho Zapotec. Proceedings o f the
International Linguistics in the Northeast Conference. Hermosillo, Mexico: Universidad de
Sonora.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
430
S u a r e z , J o r g e A. 1985. The Mesoamerican Indian languages. Cambridge: New York. Cambridge
University Press.
T h o m p s o n , S a n d r a A. 1988. A discourse approach to the cross-linguistic category ‘adjective’.
Explaining language universals. ed. by John A.Hawkins, 167-185. Oxford: Blackwell.
To m lin, R ussell S. 1986. Basic Word Order: Functional Principles. London: Routledge (Croom
Helm).
Vo g e l, Pe i ra M. and Bernard C o m r ie. 2000. Approaches to the Typology o f Word Classes.
Berlin/New York: Mouton de Gruyter.
W h orf, Benjam in L ee, and G eorge L.T r a g er. 1938. Report on linguistic research in the
department o f Anthropology o f Yale Univ. for the term Sept. 1937 to June 1938.The Whorf
Theory Complex: A critical reconstruction, Penny Lee. Amsterdam: John Benjamins Publishing
Company.
W h orf, B enjam in Le e. 1945. Grammatical Categories. Language 21. 1-11.
W it t g e n s t e in , L u d w ig . 1958. Philosophical Investigations. New York: The Macmillan Company.
Za v a la, Ro b er to. 2000. Inversion and Other Topics in the Grammar o f Olutec (Mixean). Eugene,
Oregon: University o f Oregon dissertation
Zw iC K Y , A.M.. 1977. On clitics. Bloomington: Indiana University Linguistics Club.
Z w ick y, A rnold M. and G eoffrey K. Pu llu m . 1983, Cliticization vs. inflection:English N'T.
Language 59. 502-513.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Appendix I: Text
‘One of my mother’s stories’
by Alberta Marcial Martinez
Transcribed and translated to Spanish by Alberta Marcial Martinez and Aaron l luev
Sonnenschein, and translated to English and glossed by Aaron Huey Sonnenschein
1) Benhe golh tao=a’ na
person old grandmother=lsg demdist
gu-yegh=e’ w-e-zoa yinh=na’
comp-go=3f inf-freq-plant chile=det
‘My grandmother went to plant chile.’
2) Kate’ b-zhinh=e’ ganh dee yoo.
when comp-arrive where exist land
ganh gha-zo=e’ yinh=na’,
where pot.and-plant=3f chile=det
na zoa’ tio chi=e’.
demdist stand uncle[Sp.] of=3f
‘When she got to the piece of land where she was going to plant chiles, there
was her uncle.’
3) Naa sh-cheen=e’ ganh gu-zo=e’ yinh-na’.
then hab-dig=3f where pot-plant chile=det
‘She was digging where she was going to plant the chiles.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4) Na ghe-dxak=de’ chigo lhbaha
demdist comp.and-feel=3fexp bush vine
go-
sweet_potato
‘She found a bush of sweet potato vines’
5) Dx-ak=de’ sh-cheen=e’ go.
hab-feel/think=3fexp cont-dig=3f sweetjpotato
‘She thought that she was digging up potatoes. ’
6) Lhoo lhbaha go na
in vine sweet_potato demdist
dx-ak=de’ sh-cheen=a’
cont-feel/think=3 fexp cont-dig= 1 sg
‘She dug in the roots of the sweet potatoes, thinking she was digging
sweet potatoes.’
7) Kate golhe’ pur medxo
when at the time pure money
b-dxax=en.
comp-come_out=3 inan
‘When she saw that it was money coming out.’
sh-cheen=e’,
cont-dig=3f
go-
sw cetpotato
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
8) N a’ b-olhwizh=e’ tio chi=e’
and comp-call_to=3f uncle of^3f
dx-e=e’ lhee
cont-say=3f 3f
‘She called out to here uncle and told him:’
9) Daka ni tio ga yoo=dxgwa
imp.come demprox uncle where stat.be_inserted=emph
medxo.
money
‘Come here uncle, where there’s a lot of money.’
10) Kate’ b-eyozh b-een=e’ lhoo yoo,
when comp-finish comp-check=3f inside ground
bitbi medxo yoo.
neg money stat.beinserted
‘When he looked in the ground, the money wasn’t there anymore.’
11) Tio chi=e’ gage swert dx-e-le’i=de’ medxo.
tio of=3f neg luck cont-freq-see=3fexp money
‘It wasn’t her uncle’s luck that he was going to see the money.’
12) Swert chi=e’ lhee’.
luck of=3f 3f
‘It was her luck.’
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
434
13) A lhe’i=de’ b-le’i=de’ medxo.
excl pot.see=3fexp comp-see=3fexp money
‘She saw, that’s why she found the money.’
14) Na gage’ bi medxo chi=e’
and neg neg money of=3f
dele go-k=e’ benhe yashe’.
because comp-be person poor
‘She didn’t have money because she was poor.’
15) Na b-zhelh x-medxo=e’,
demdist comp-find poss-money=3f
sheka bi b-olwizh tio chi=e’.
if not com p-callto uncle of=3f
‘She would have found money there if she hadn’t called out to her uncle.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Discourse functional units: A re-examination of discourse markers with particular reference to Spanish

PDF

Borrowing and borrowability

PDF

Grammaticalization and the development of functional categories in Chinese

PDF

Beyond words and phrases: A unified theory of predicate composition

PDF

Children's scope of indefinite objects

PDF

Form and meaning: Negation and question in Chinese

PDF

Ellipsis constructions in Chinese

PDF

Correspondence and faithfulness constraints in optimality theory: A study of Korean consonantal phonology

PDF

Effects of email on learners' compositional ability: A quantitative and qualitative study

PDF

"Master of many tongues": The Russian Academy Dictionary (1789--1794) as a socio -historical document

PDF

Asymmetry of scope taking in wh -questions

PDF

A culture of sociability: Popular speech in ancient Rome

PDF

Early Analytic Philosophy: The history of an illusion

PDF

Connectionist phonology

PDF

Impostors, antichrists, and satanists: Textual strategies of the central and eastern European fin de siecle

PDF

Causativity in Korean: Syntactic causative, control, and purpose

PDF

Gossip, letters, phones: The scandal of female networks in film and literature

PDF

Generals and particulars in Thucydides

PDF

Acquisition of Spanish verb morphology by bilingual children. A longitudinal study between the ages of 2;9 and 3;3

PDF

A longitudinal study into the learning style preferences of university ESL students

Asset Metadata

Creator Sonnenschein, Aaron Huey (author)

Core Title A descriptive grammar of San Bartolome Zoogocho Zapotec

School Graduate School

Degree Doctor of Philosophy

Degree Program Linguistics

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag language, linguistics,OAI-PMH Harvest

Language English

Contributor Digitized by ProQuest (provenance)

Advisor Comrie, Bernard (committee member), Hawkins, John A. (committee member), Munro, Pamela (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c16-326245

Unique identifier UC11335691

Identifier 3155481.pdf (filename),usctheses-c16-326245 (legacy record id)

Legacy Identifier 3155481.pdf

Dmrecord 326245

Document Type Dissertation

Rights Sonnenschein, Aaron Huey

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA