chumash_ono1996_o
TRANSCRIPT
INFORMATION TO USERS
This manuscript has been reproduced from the microfilm master. UMI
films the text directly from the originid or copy submitted. Thus, some
thesis and on copies are in typewriter face, while others may be
from any type of computer printer.
The quality of this reproduction is dependent upon the quality of the
copy submitted. Broken or indistinct print, colored or poor quality
illustrations and ph print bleedthrough, substandard margins,
and improper alignment can adversely
In the unlikely event that the author did not send UMI a complete
manuscript and there are missing pages, these will be noted. Also, if
unauthorized copyright materid had to be removed, a note will indicate
the deletion.
Oversize materials (e.g., maps, drawings, charts) are reproduced by
sectioning the original, at the upper left-hand corner and
continuing from left to right in equal sections with small overlaps. Each
original is also photographed in one exposure and is included in reduced
form at the back of the book
Photographs included in the o manuscript have been reproduced
xerographically in this copy. Higher quality 6" x 9" black and white
photographic prints are available for any photo hs or illustrations
appearing in this copy for an additional charge. Contact UMI directly to
order.
A Bell & Howell Information Company 300 North Zecb Road. Ann Alfaor N1 48106-1346 USA
3 13/76 1-47OO 800/52 1-0600
UNIVERSITY OF CALIFORNIA Santa Barbara
Information Flow and Grammatical Structure in Barbarefio Churnash
A Dissertation submitted in partial satisfaction of the requirements for the degree of
Doctor of Philosophy
in
Linguistics
by
Tsuyoshi Ono
Committee in charge:
Professor Marianne Mithun
Professor Wallace Chafe
Professor Sandra A. Thompson
Professor William Ashby
Professor W. Randall Garr
August 1996
UMI Number: 9807454
UMI Microform 9807454 Copyright 1997, by UMI Company. All rights reserved.
This microform edition is protected against unauthorized copying under Title 17, United States Code.
300 North Zeeb Road Ann Arbor, ME 48103
The dissertation of Tsuyoshi Ono i s approved
Committee Chairperson
August 1996
Copyright by
Tsuyoshi Ono
1996
iii
ACKNOWLEDGEMENTS
There are countless people who have made the
present work possible and who have provided various
types of support throughout my graduate education at UC
Santa Barbara. Rather than try to list them all and
r i sk revealing how poor my memory is, I l ist here only
the people who were most d i rec t ly involved in this work.
I would l i k e to thank Marianne Mithun, Wally Chafe,
Sandy Thompson, B i l l Ashby, Randy Garr , Suzanne Wash,
and Pat Mayes. I am especially grate fu l to Marianne
Mithun for her constant help and encouragement, and to
Suzanne Wash, who generously provided me with the
Barbareno discourse data.
M y work on Barbareno Chumash has been made p o s s i b l e
by grant BNS90-U018 from t h e National Science
Foundation (Marianne Mithun, Principal Investigator).
1 am also grateful to the Santa Barbara Museum of
Natural History for kindly providing the microfilms of
J. P. Harrington's manuscript materials as w e l l as M. S .
Beeler's or ig ina l manuscripts, thereby making this study
possible.
VITA
1982 B.A., Konan University
1985-1987 Teaching Assistant, Department of East Asian Languages and Literatures, University of Oregon
1987 M.A., University of Oregon
1992 Winter Visiting Instructor, Department of East Asian Languages and Literatures, University of Oregon
1987-1994 Research Assistant and Teaching Assistant, Department of Linguistics and Department of Germanic, Oriental, and Slavic Languages and Literatures, University of California, Santa Barbara
1994-1996 Adjunct Instructor, Department of East Asian Studies, University of Arizona
SELECTED PUBLICATIONS
1990 &g, L, and a Clauses in Japanese Recipes : A Quantitative Study. Studies in Language 14:73-92.
1992 With R. Suzuki, Word Order Variability in Japanese Conversation: Motivations and Graimnaticization. Text 12:429-445.
1992 With R. Suzuki, The Development of a Marker of Speaker's Attitude: The Pragmatic Use of the Japanese Grmaticized Verb shhau in Conversation. Proceedings of the 18th Annual Meeting of the Berkeley Linguistics Society. 204- 213.
1992 The Grmaticizatian of the Japanese Verbs oku and shimau. Cognitive Linguistics 3:367-390.
With P. Mayes, The ~ c c p i s i t i o r i of t h e Japanese Subject Marker m and Its Theoret ica l Impl icat ions . In P. Clancy, ed*, JapaneseIKorean L ingu i s t i c8 Vol. 2. Stanford: Center for t h e Study of Language and Infarmation, Stanford Universi ty. 239-247-
With M. H t h u n and S* Wash, The S h i f t i n g S t a t u s of I n i t i a l G l o t t a l Stop i n Barbarefia Chumash. Proceedings of t h e 1993 Hakan-Penutian Workshop. 199-207
With S. Wash and M. Kithun, F ina l G l o t t a l i z a t i a n i n Barbarefiu Chumash and Its Neighbors. Proceedings of t h e 1993 Hokan-Penutian Workshop. 208-21'7
With S. Thompson, Unattached W s in English Conversation. Praceedinqs of t h e 20th Annual Meeting of t h e ~ e r k e l e ~ - ~ i n ~ i s t i c s Society. 402- 419.
With S. Thompsmn, What Can Conversation T e l l U s about Syntax? I n P. Davis, ed., Descr ipt ive and Theore t ica l M d e s i n t h e New L i n p i s t i c s . Amsterdam: Benjamins. 213-271.
With S * Thompson, The Dynamic Nature of Conceptual Stricture Building; Evidence from Conversation. I n A* Goldberg, ed., Cunceptual S t ruc tu re , Discourse and Language. Chicago: Center f o r the Study of Language and Information, Univers i ty of Chicago Press . 391-399.
With E. Yoshida, A Study of CO-Canstmctian i n Japanese; W e Don't -Finish Each Othe r f s SentencesOm I n N. Akatsuka & S * Iwasaki, eds., Japanese/Korean Lingu i s t i c s 5. Chicago: Center f o r t h e Study of Language and Information, Univers i ty of Chicago P r e s s m 115-129.
To Appear With S. Th~mpson, I n t e r a c t i o n and Syntax i n t h e S t r u c t u r e of Conversational Discourse. In E. Hovy & Dm Scot t , eds., Diecourse Processing: An I n t e r d i s c i p l i n a q Perspective. Heidelberg: Springer-Verlag-
To Appear With S * C&ngF Ad HOC Hierarchy: Lexical S t ruc tu re s f o r Reference i n Cansumer Reports Articles. I n B* FoxF e d m F Studies i n Anaphosa. Amsterdam: Benjamins.
To Appear With S * Cmming, Discaurse Syntax* I n T. A* van Dijk, ed., Discourse: A M u l t i d i s c i p l i n a q Int roduct ion. London: Sage.
FIELDS OF STUDY
Major Field: Discourse and G r a m m a r
Studies i n Discourse and Grammar ; Gr-aticization Professors Wallace Chafe, P a t r i c i a Clancy, Susanna Cmming, John Du B o i s , Marianne Mthun and Sandra Thampson
Studies i n Native American Languages; Descr ipt ive L ingu i s t i c s
Professors Wallace Chafe and Marianne Mithun
Studies i n Japanese Linguistics Professor P a t r i c i a Clancy
Studies i n Phonetics Professor Ha S* Gopal
vii
Infoz=mation Flow and G r a m m a t i c a l S t r u c t u r e i n Barbarefio Chumash
by
Tsuyashi Ono
This d i s s e ~ a t i o n p re sen t s a d e s c r i p t i o n of some of
the f u n c t i o n a l a spec t s of the grammar of Barbarefio
Chumash, a C a l i f o r n i a language, by c l o s e l y examining
a c t u a l d i s c o u r s e d a t a wi th r e s p e c t t o one area of
research called ' i n f o m a t i a n f low8.
I begin w i t h a b r i e f grammatical ske t ch of
Barbareiio, and t h e n discuss t h r e e d i f fexer i t b u t t i g h t l y
connected areas of t he grammar. 1 f i r s t demonstrate t h e
importance of t h e cure-oblique d i s t i n c t i o n in t h e
argument structure of t h e language by exmin ing s e v e r a l
different aspec t s2 pronominal a f f i x e s , e a n s t i t u e n t
o rder , and core a r w e n t - b u i l d i n g devicese
1 then d e s c r i b e t h e var ious r e f e r e n t i a l forms and
their r e s p e c t i v e discouxse functionse These inc lude
v i i i
pronominal affixes, full NPs, independent pronouns, and
the demonstrative be?. 1 suggest that pronominal
affixes are the default choice because they are
grmatically recpixed and are most commonly used for
the least cognitively demanding task of keeping track af
e s t h l i s h e d pmic ipants in the discourse. In contrast,
f u l l W s are used for the more cognitively demanding
tasks of introducing/reintrod~cing pa*icipants and
clarifying potentially ~ i g u o u s referents in the
discaurse. Independent pronouns and the demonstrative
he2 are used for such tasks as making a contrast and - referxinq ta pa~icipants that are referentially complex
or whose exact identities have been le f t unspecified.
Finally, I describe different constituent order
types and t h e i ~ respective functions. I suggest that
the basic constituent order in Barbare50 is a verb
perhaps fallawed by either the intransitive subject or
the transitive direct object, and perhaps further
followed by an ablique, and that NPs occurring befare
the verb are associated with such marked functions as
t o p i c and focus.
Acknowledgments ........................W....... iv ............................................ V i t a v
Abstract ...m.................................... viii Table of Contents ............................m.e X
List of Uhreviations ..............ern........... xiii
l . Intxoduction ..............m..m............... l L 1 . Language and people ................W....m 2 1.2. Previous work .............m.............. 3 1.3. Theoretical foundation .............W..... 4
1.4. Data and methodology ....................m 6
1.5. Organization .....................W....... 8
Hates for chapter l ........................... 10
2 . Grammatical sketch .....................~..m.. l1
2.L Distinctive sounds ............W.......... 11 2 - 2 0 Basic mo~pho~ogical structure
and lexical categories .W......e.......... 14
2 * 3 m Argument structure and pronominal
affixes .................................. 20 2.4. Constituent arder ........................ 25
4 . Referential choice . a . e e e e o e e e . e e . e . o e m . . . e . . e 72
4.1. Pronominal affixes . . e o e e * e o . o e . . e . . o e e e m . 72
4 . 2 * Full noun phases . e . e e o . e . o e o . . . . e . e . e . e . 79
4 * 3 . Independent pronouns o . . e e e e e e e e e . e . . e e . . . 94
4.4. Daonetrative he2- . e e . . . e . e . . e . e e e e o e e . . . . 103
4 - 5 * Summary e o e e . e o . e . . . o . e . e . e e o e e e o o ~ e e o . . . . 113
Bibliography a o e e m e e e * w . * * w m . e e m * a e . m m m * a m . a o e * a . 168
xii
1
l0 =
1FL =
2
20 =
2PL =
3 - - 3PO =
AGT =
A2 =
ART =
CA =
COB! =
DAT
DIR =
DIS =
DP =
DU =
D r n =
EM =
EVT =
first person
f i r s t person direct object
f i r s t person pluxal
second person
seeand person direct abject
second person plural
third person
third person plural direct object
agentive
Alienable possessiun
article
causative
comitative
dative
directional
distal
dependent
dual
dusat ive
emphatic
future
habitual
inde f in i t e
ins t m e n t a l
locative
negative
nominalizer
noun past
plural
proximate
past
reduplication
resultative
reflexive
remote
repetitive
superlative
transitive
xiv
1. Introduction
Baxbarefio Chumash is a North h e r i c a n Indian
language indigenaus to Santa Barbara, California. This
d i s se r t a t ion descxibes some funct ional aspects of the
grammar of t h e Barbzueiio Chumash by c l a se ly examining
ac tua l ckiscourse data. Specif ical lyg t h i s w i l l be done
based on the findings and the asswtptions i n one area of
research ca l led 5nfonnation flowF (Chafe 1987, 1994}g
which suggests that t h e use of d i f f e r en t grammatical
foms is motivated by the changing cogni t ive s t a t e of
the information i n t he speaker% and t h e hearer*^ minds.
Aspects of t h e grammu which will be dealt with in t h i s
disse-ation are axgument structure g r e f e x e n t i a l choice,
and consti tuent ozder. These have been chosen because
they axe the areas where the in t e r face of grammar and
discourse function is most evident.
Since s tudies of l i ngu i s t i c farms based on
discourse data have been xather scarce, p a r t l y because
of a general lack of tex tua l materialg a8 w e l l as the
dominance of uconstmcted sentencem methodolagy, this
d i s se r t a t ion w i l l make an impoz=txint new contxibution t o
t h e f i e l d . F ~ i c u l a r l y because o f its r i c h morphology,
the s t ruc ture of Baxbarefio Chumash is r a the r d i f fe ren t
from inany well-docmented languages* For t h i s reasonp
the present diseertation will provide the f i e l d with a
much-needed database which should help to genexate new
ideas and to evaluate already exis t ing theor ies of
language.
1.1. Language and people
The Chuniash languages w e r e o r ig ina l ly spoken by
between S a n L u i s Obispo and Malibup at the western edge
of the San Joaquin Valley? and on the i s l ands i n t h e
S a n t a Barbara Chanriel. The family i nc ludes at l e a s t six
dif ferent languages, now referred to by the names aÂ
~ s s i ~ n s around which t h e i r speakers l ived*
Ohispeiio Central Chumash
Ventureiio Barbareiio Ynesefio Puris he50
Island Chumash
Barbazefio Chumash belongs t o t he Central Chumash
subfamily and was spoken u n t i l 1965. The las t s p e d c e ~ ~
Masy Yee? who died in 1965, w a s a l so the last speaker of
any Chumash language* The genetic re la t ionship of t he
Chumash languages t o any larger f d ~ y ~ s t o c k has not
been es tab l i shed-
1.2- Pzevious work
There has been little published woxk on t h e Chumash
languages in general, and on Bazbazefio i n pa r t i cu la r .
There is a sho r t Barbarefio Chuxnash grammatical sketch
(Beeler 19761, a Barbaxeiia narrat ive with a s m a l l word
list of a p p r o x a a t e l y 1,000 words (Beeler 19781, and a
longer version of t h e same narra t ive ( B e e l e r and
Whistler 1980)- There is a r a t h e r extensive grammar of
Ynesefio Chumash (Applegate 1972)- There are some works
on redupl icat ion in Ynesefio and Basbarefia (Applegate
1976; Wash 1995aIt on g l o t t a l i z a t i o n i n Barbarefio
(fithun et al. 1994; Ono et a l . 1994a1, and on a
phenamenan called F s ib i lant harmony in Barbarefio and
Ventusefio ( B e e l e r 1970b; Harrington 1974; Puser 1982;
Mithun t o appear)- There is also some comparative waxk
on t h e Chuxnash family ( K l u 1973, 1977# 1981).
Additionally thexe ie an unpublished wo~king dict ionary
of Barbareiio Chumash which contains a p p s o x h t e l y 2,000
wards (Whistler 1980) T h i s i t e m has been widely
c i rculated amang scholars who are intezested in Chumash.
There is alsa an unpublished manuscript containing a
grammatical sketch, a word list, and a na r ra t ive (Eeeler
19'70a) which was the b a s h f o r some of Beelerrs
publications listed above.
1*3* Theoretical foundatian
Thie d i s s e e a t % o n is based on an approach o f t en
ca l l ed funct ional l i ngu i s t i c s . The c h a a c t e r i s t i c s of
t h i s approach can be summarized as f ~ l l o w s : it focuses
on the funct ional aspects of language; it examines t h e
re la t ionsh ip between par t icu la r forms and t h e i r
functions. The functions propased by t h i s approach are
of ten psycholugi~ally andior socio-culturallymutivated.
This tendency is partly due to the working a s s m p t i ~ n s
t h a t language i s a pa r t of t he human cogni t ive
mechanism, and t h a t it is used foz s o c i a l functions in
a large sacia-cuXtura1 context*
As a na tu ra l consequence of these warking
assumptions, f u n c t i o n a l ~ y ariented l i n g u i s t s very often
use natural discourse as a database r a t h e r than
constructed data, which are the standaxd f o r many o the r
schools af l i n g u i s t i c s ~ When epaken discourse is
examined by func t i ana l ly oriented l i n g u i s t s , f o r
example, audiotapes or videotapes are made af naturally
aecuxring discourse. These are later transcrfied, of ten
using a de t a i l ed transcription system (@.g*, nu B o i s et
al. 1993)* Using these data, eithex very r i c h
q u a l i t a t i v e analyses (emgm, s t u d i e s i n Chafe 1 9 8 0 ) a r
q a n t i t a t i v e analyses axe pe r f amed ( e * g m , s t u ~ e s i n
Giv6n 1983 ; Du B a i s 1987 1 . Sometimes, even e q e r h e n t a l
method01aq-y is emplayed t o eva lua te pre-exist ing
t heo r i e s (e .g . , Tomlin 1987)- The care taken by
funct ional ly a r i en t ed l i n g u i s t s with t h e data as w e l l as
t h e i r methodological considera t ion8 may be seen as a
natural progression of t h e f i e l d .
There ase two othex c h u a c t e r i s t i c s of t h i s
appxoacb which cut across i ts emphasis an p s y c h a l ~ g i c a l
and sacio-cul tura l factors m One c h m a c t e r i s t i c is an
i n t e r e s t i n diachronic aspects of language* T h i s can be
seen i n t h e notion af g r m a t i c i z a t i o n , which has played
a major role i n t h e work of numy func t iona l ly o r i en t ed
linguists (@.gm, s tud i e s in m a u g o t t and Heine 1991;
Hopper and Traagott 1993). It is based on t h e idea t h a t
language is f l u i d and c~ns tant ly changing; tkough
h i s t o r i c a l praeesses, new grammatical c o n s t m c t i o n s are
cans tan t ly evolving t o f u l f i l 1 var ious func t iona l
purposes. The other c h a x a c t e r i s t i c is an intexest in
examining not only one language but a n&ez of
languages i n the woxld and in learning from t h e
diffexences mci s ~ l ~ i t i e s among them ( e. g g , Hopper
and Thompson 1980: Given 1984, 1990; Mithun 1987; C o m r i e
1989) g
Among many different approaches and areas of
invest igat ion i n functional 1 i n ~ i s t i . c ~ , t h i s
dissedation w i l l focus on one broad area called
finfoxmation flowf (Chafe 1987 , 1994) which suggests
that t h e use of particular l i n g u i s t i c forms is motivated
by t h e changing cognitive s t a t e of the information i n
the speaker's and the hearer's minds.
1 .4 . Data and methodology
There are t w o main sources of data for the study of
Baxbaxefia Chumash. F i s s t , enunnous amounts of data were
collected by the l a t e John Harringtan, an
anthopological l i n g u i s t for the Sndtheunian
InstitutionF from the early 1910's u n t i l h i s death in
1961. During this period, he worked with several
epeakers, including the last Chumaeh speaker, M a r y Y e e .
Barrington's data exist in the farm o f micrafibs of
50,000 to 100,000 pages of his original f i e l c i n ~ t e s , slip
files and texts (Haxrington 1912-1961). The textual
materials, which are most relevant to this dissertation,
include many different genres such as traditional tales,
ethnographic accounts, personal narratives, historical
narratives, myths, and religious texts. Second, the
late Madison Beeler, a Berkeley linguist, worked with
Mary Yee during the 1950's and 1960's and left
fieidnotes and texts, as well as audio recordings of
some of these materials (Beeler 1954-1961). Both
Harrinqt.on8s and Heeler's data are in the collections of
the Santa Barbara Museum of Natural History, from which
all of the data for "this dissertation come.
Approximately 250 texts from microfilm reel # S 9 of
Barrington's data have been given an interlinear,
morpheine-by-morpheme analysis by Suzanne Wash for the
Chumash Language Project headed by Dr. Marianne
Mithun. These texts are primarily ethnographic
accounts, all from Mary Tee. I will make my functional
analyses of Barbareno grammar based on these textual
data. Though these texts will be the primary source for
this dissertation, I will be complementing then with
examples from BeelerOs data as well as other published
sources .
Analyses of each aspect of Barbareno Chmnash w i l l
be based on detailed examinations of t h e discourse
context where the linguistic forms i n quest ion occur
( i . e . , particular cons t i tuen t order variation,
particular r e f e r e n t i a l forms, etc.). Thus my analyses
w i l l pr imari ly be qual i tat ive i n nature. Based on. these
analyses, I w i l l i d e n t i f y t h e funct ional categories
which are associa ted with linguistic forms i n question.
I w i l l present the results as funct ional descr ip t ions
and I w i l l a l s o seek their t heo re t i ca l implications by
p lac ing t h e c h a r a c t e r i s t i c s of Barbareno Chumash i n t h e
context of t h e existing funct ional theories of
linguistics i n general and the information f l o w
l i t e r a t u r e i n pa r t i cu l a r .
1.5. Organization
T h i s d i s s e r t a t i o n i s organized in t h e following
way. In chapter 2, I w i l l give a grammatical sketch of
Barbareno Chumash s o t h a t readers w i l l have enough f a c t s
about Barbareiio t o follow t h e rest of the d i s se r t a t ion .
I n chapter 3, I w i l l describe the bas i c argument
s t ruc tu re of Barbareno and the prominent morphological
and syntac t ic f ea tu re s associated with it. I w i l l also
discuss other devices which are closely associated with
the basic argument structure. In chapter 4, I will
discuss factors involved in using different referential
forms. In chapter 5 , I w i l l describe the structures of
different constituent order types and the functions
associated with them. In chapter 6 , I w i l l summarize
the findings of this dissertation and their theoretical
implications. I will also discuss problems of this
investigation and suggest directions for future
research.
N o t e s for Chapter 1
1. Suzanne Hash (p.c. ) states that the analyses of these
t e x t s were done several years ago and not all of them
reflect her current analysis.
2. Grammatical sketch
This chapter gives a brief overview of the grammar
of Barbarefio Chumash in order to provide readers with
basic facts of the language. Included are distinctive
sounds, basic morphological structure and lexical
categories, argument structure and pronominal affixes,
constituent order, and clause-combining. It should be
noted that many of the analyses given in this chapter
are based on work by Beeler (particularly l970a, 1976;
Beeler and Whistler l980 ) .
2.1. Distinctive sounds
Barbarefia Chumash contains the following vowels:
Table Is Barbarena vowels
i a: U
e 0
a
The front and back vowels M, /e/, /U/, and /o/
tend to be low. This conclusion is based on the
following two observations. First, often these sounds
are written with symbols [I], [ g ] , [v], and [ a ] , respectively, in Beeler s f ieldnotes . Second, the
audio-recordings of Mary Tee's speech clearly support
this. In general, the vowels seem to be higher in open *
syllables and lower in closed syllables.
Barbarego Chumash contains the following
consonants :
Table 2: Barbareno consonants
P t k q 2
k A
S S X h
S X
c c
6 &
W Y
W 9 Barbarefio's relatively large set of consonants consists
of plain consonants and their glottalized counterparts.
The sole exception for this reqular-glottalized pairing
is [ S ] for which no [h 1 counterpart is attested in the
existing data. Glottalized sonorants do not occur word
in i t ia l ly .
Some previous studies of Chumash (Applegate 1972;
Beder 1976) have treated aspirated consonants as
dis t inct ive sounds. However, following K l a r (1977) and
Wash ( 1995a) , I w i l l treat them as a combination of a
regular consonant plus an /h/; most of the instances of
aspirated consonants resul t f r o m a morphological process
i l lus trated i n the following example (a
abbreviations is provided on page x i i i ) :
(1) JH59.128L-l28R, t x t 11' line 004 59.128L
&^at t&tty. s-sa?-tuhuy 3-FUT~r&in 'it is g. to rain. '
h The symbol [ ] indicates aspiration.
list of
Aspirated
consonants are found when two identical (or similar)
consonants are placed next to each other. Remaining
aspirated consonants i n a f e w lexical items are
interpreted as a combination of a regular consonant and
an /h/ (e.g., [sha] 'toothp as /*ha/).
Barbareno is known for a phenomenon called sibilant
harmony i n which s ib i lant s within words harmonize w i t h
t h e last s ib i lant of the word in terms o f tongue
posit ion. Thus words contain e i ther only apical (or
alveolar) s ib i lant s [ S ] , [ & l , [c], and [&l, or only
ldnal/blade (or palata l ) s ib i lant s [g 1, [c], and [ c ]
(For details, see Beeler l970b; Harrington 1974.)
Examine the following examples:
(2) JH59.24&241R, txt 26, line 006 59.243X
a e ~ i ~ a ~ ~ i - u i - &S h i ~ e t g t i d e w , &e=s-iy-aqniwili8 hiçAe=fitani EM=3-PI.--think DP=EM=little 'they just do a little thinking.r
(3) JH59.407R-409&, txt 58, line 004 59.408L
k5.m kaGyw6fi. k b ka =B-iy-wefi and thens3-PL-go.to.bed 'then they would go to bed'
N o t i c e that in these examples the third person
pronodnal prefix is realized as [ S ] and [ S ] , respectively, depending on the final sibilant of each
word (for ( 3 ) , itself is the last sibilant ) It
should be noted, however, that this harmony is very
often not observed in the data I looked at (cf. Mithun
to appeal: ) .
2.2. Basic morphological structure and l e x i c a l categories
As could be seen i n the examples above, another
interesting characteristic of Barbare60 Chumash is
polysynthesis. Examine the following examples:
(4) JH59.299L-300L, txt 55, line 003 59.2991,
?alauksukepwon holS?lfi? iwn ?&l-R-su-kep -won ha -1 -win NU -B-CA-bxthe-3P0 DIS-ART-chlld-PL 'She's bathing the childrenF
(6) JES9.7R-8R, txt 5, line 001 59.81.
Barbareno words generally contain multiple morphemes.
As the examples show, most structures axe agglutinative,
and segmentation of words into morphemes is relatively
straightforward.
The two main lexical categories in Barbareno are
nouns and verbs. Here are some examples of nouns:
(7) JH59.205L-205R, txt 21, line 002 59.205L
iheitub&+ hi=he?=l -tubu<r DP=PRX=ART=r ain "rain'
(8) JH59.665Rw666R, txt 69, line 002 59.665R
( 9 ) JH!59.656R-657Lt txt SO, Line OOlb 59.656R
m i c a l l y , nouns are marked with t h e dependent p x o c l i t i c
hi=, as shown i n ( 9 l f and (IO).~ Nouns are a l s o
often marked w i t h one of the three p r o e l i t i c s he?=
*proximate* ( ( 7 ) ) , h02= FdistalF ( ( g ) ) , and hu= @remote0
( ( 8 } ) . These clitics can CO-accur with hi= an nouns, a s
shown i n ( 7 ) and ( g ) * The sound of hi=, heZ=, ha?=,
and hu= is o f t en e l ided, which is seen i n ( 7 ) m The
p r o c l i t i c i n ( 7 ) and ( 8 ) seems to be bes t
c h a a c t e r i z e d as a d e f i n i t e a r t i c l e . I n ( 9 ) and (XO),
possession i s ind ica ted by prefixes and g= on t h e
nouns . (See below f o r more d e t a i l s about these
pref ixes . ) I n p l u r a l i t y is ind ica ted by
redupl icat ion accompanied by lengthening and
g l o t t a l i z a t i o n of t h e final sy l l ab le (Wash 1995a). In
( 10 ) , t h e suffix -iwaS indicates t h e defunct status of
t h e referent ind ica ted by the noun. I n t h i s p a r t i c u l a r
examplef it ind ica t e s a rib of a dead whale.
Here are some examples of vexbs:
(13) J'H59*3OOR-301R0 txt 56, line 002 59.300R
( 13 1 and ,&-. in (14 ) are all pronouns. (See below
for mare details 1 These prefixes are the same as those
used for nouns to indicate possession* In ( I1 ) & is
a ciurative pref In ( 12 ) , sa2- is a future tense
aspect* In (Id), is a negative pre f ix , dli- is a
habitual prefix, and -wag is a past tense suffix*
Notice that both in tenns of form and functionf the
suffix -was is very t 3h i la r to -iwaGf which is used for
nouns to indicate the defunct status of the referent*
It suggests that these suffixes may be related
histaricalLy. Finally, as can be seen in ( 7 ) and (12),
often the same (or a similar) moqheme is used either as
a noun ox a vezb ( 9 a i n F and tuhuv %a rainp)* In
this particular case, glattalization is functioning as
absence of an adjective category (cf. Dhon 1977;
Schachter 1985) . The pxedicating functian of adjectives
in other languages seems ta be served by verbs in
As can be seen in these examples, the morphemes ,hotF#
@goadt, @healthye, and @tastyFi which would be
adjectives in some languages, are serving a predicating
function while being treated m o ~ h o ~ o g i c a l l y as vezbs:
notice the presence of pronou~e S - ~ V - ~ and k, the
negative prefix e, and the past tense suf f ix -wag.
Similarlyi the attributive function of adjectives
in other languages appears to be gemred by nouns in
Barbarefio. Examine the following examples:
(19) JHS9.262R-263L, t x t 32, l ine 001 59.262R
(211 m59-658R-659L, tart 63, l ine 0 0 % ~ 59-658R
( 2 2 ) -59-656R-657Lr txt 50, line 004 59.656R
As can be seen in the above examplesf t h e morphemes
tmanyt, %nall~, t s t rongr f and # o t h e r t , which would be
adjec t ives in some languages, are sesving an a t t r i b u t i v e
function whi le being t r e a t e d m o ~ h o l ~ ~ i c a l ~ y as nouns:
no t ice t h e presence of p r o c l i t i c s bi= and &. In (221,
the noun s t a t u s of the a t t r i b u t i v e element t o the r? is
indicated f u r t h e r by t h e redupl icat ion accompanied by
q l o t t a l i z a t i o n and LenGhening on t h e f i n a l sy l l ab l e
(seen above)m
2m3o & p e n t s t r u c t u r e and pronominal a f f i x e s
I n terns of argument s t ruc tu re , Barbare50 shows
nodnative-accusative pa t teming . Examine t h e following
examples :
(23) JE59.47R-57R, txt 170# l ine 036 59.S2R
( 2 5 ) JE59.633L-635R, txt 128, line 0 0 1 59.633L
naq?fiw b i M - d h i t i t f i , na q-?uv his3 whens3-eat. D P à ˆ A R T gg DPçdo *When a dog ate eggs, p
The sole argument of the intransitive clauses in (23 )
and (24) is marked by the prefix on the verb,
regardless of Its agency (agent in the former and
patient in the latter). Similarly in ( 2 5 ) , the agent of
a tzansitive clause is marked by the same prefix =. The verb does not show any marking for the patient.
This shows that Barbarefio exhibits nominative-accusative
patterning in terms of verbal marking, and can thus be
said to recognize subject and direct object categories.
One thing we should keep in mind is that these
verbal affixes are not "agreement markersw, which are
simply copies of features of independent NPs. They are
full referential pronouns in the form of affixes. For
this reason I will refer t o then as pronominal affixes.
The best evidence for this analysis is that Barbareno
utterances can be complete without full NPs. So
examples (23)-(25) would still be complete utterances
without the noun phrases after the verb. Examples ( 3 )
and ( 4 ) above illustrate such cases. Consider (4) which
is repeated here as (26):
( 26 ) JH59.299L-300L, txt 55, line 003 59,2991.
siyalisoKo3,R6 - in, -iv-~i-aoqolitoy-in 3-K.-DO&-aurroand-20 @they are surrounding your
The subjec t ' they* and t h e d i r e c t object @your are
expressed by pref ixes 8-iv- and a s u f f i x - in
respectively. There are no full NPs i n t h e utterance,
but it s t i l l cons t i tu t e s a complete predication,
suggesting t h a t these a f f ixes are not agreement markers
but full-fledged pronouns. The question of when t o use
t h i s kind of ut terance rather than an ut terance with a
f u l l HP w i l l be addressed in chapter 4.
Since it is crucial to have a good understanding of
how these pronominal a f f ixes work i n order to follow any
Barbareno examples, I w i l l present the spec i f i c s of t h e
Barbarefio pronominal system i n the remainder of t h i s
section. The following t ab le lists t h e subject
prefixes :
Table 3: Barbareno subject pronouns
SG DU PL IDF
I k- k-is- Ic-iy-
3 S s-is- s-iy- s-am-
Subject marking i nd ica tes person and number: lst, Znd,
and 3rd persons; singular, dual, and plural . That is,
the pref ixes br and indicate lst, 2nd, and 3rd
persons respectively. The prefixes i S k - and &v- indica te
dual and plural respectively, and t he lack of these
prefixes ind ica tes singular. The indefinite 9-am-
indicates a nonspecific r e fe ren t 'someone * or ' they8 . For the 3rd person dual, plura l , and indef bite , t h e
pxefix is of ten not used:
(27) JEf59.662R-663R, txt 67, line 001 59.662R
?awaly6nt'e kalaina~liil~r& P S huln6 - 1%~' ?awalyente ka=l m? -am -awl-wag-g? hu=l asaioloq brandy KA=ARTçSM-IDF-&ink-PST-E RM=ART=long.ago 'Brandy was the f avorite drink, '
The l i teral t rans la t ion of t h i s example would be
something l i k e 'Brandy is what they drank long ago'.
Only the i nde f i n i t e am- without t he t h i r d person E is
used to ind ica te t h e nonspecific re ferent 'they'. It
should also be noted that t h e third person g= i s of ten
replaced with t h e nominalizer Pal-:
(28) JH59.27Lg29R, tast 172, l ine 013 59.28R
( 2 9 ) JH59.27L-29R, t x t 172, line 012 59.28L
?ik%i ?alKut6-wnKi. ?ik&u U--qutuwuqs but NM -whistle *but he whistles.'
These examples are taken f r o m the same text. In ( 2 8 ) ,
a fox is referred to by the third person prefix g=, and
in (291, it is refesred to by the nadna l i zer 2al-
instead of g=. 2
The following table lists the direct object
pronominal suff ixes:
Table 4: Barbareno direct object pronouns
SG NON-SG
Object pronouns make fewer distinctions: lst, Znd, and
3rd persons; singulax and nonsingular. The s u f f i x - i w w
marks both 1st and 2nd person nonsingular. 3rd person
singular i s unmarked, as i l lus trated i n ( 2 5 ) above.
When the 3rd person direct object i s plural , it may be
marked by the non-singular suffix -wn. Hawever? as the
following examples suggest, speakers seem to have had
t h e option of using -WUQ or not:
(31) JH59.299L-300L, txt 55, line 002 59.2991,
indicated by the reduplication on the ward 'eyer, and
the verb btiy 'to see@ is marked by the third person
plural suffix -mn. In (321, the dixect object 'many
houses@ is plural as indicated by the reduplication on
the word rhouser; however the same verb bti 'to seer
does not take the plural suf f ix - a n (kutiv and m are
two variants of the Bme verb).
2.4. Constituent order
Many of the above examples may suggest that
Barbarefia is a verb initial language in which the
arguments follow the verb. Examinatian of the data
shows that t h i s is in fact the most common pattern;
however, other types of canst i tuent order are also
found. Examine the follawing examples:
(34) JH59.662Rw663R, ?xt 67, l ine Q O X 59.662R
These examples are taken front the 8-e text. In (33)
pelu ePedxo which appears to be the subject , is placed before the verb. In (341, discussed above as (271, the
morpheme am- in the second word is another pronominal
prefix indicating a nonspecific referent @someoner or
%heyf (called rindef&nite8]. The literal translation
of thie example would be ~umething like @Brandy is what
they drank long agar. As c m be seen, in this example
2awalvente rbrandyr, which appeaxs to be the direct
object, is expressed before the verb (see chapter 5 for
the syntactic status af preverbal NPs).
Even when the axcpments follow the verb, their
order does not seem to be rigid$
nas?aw ilkti na ==S-?uw =h when=3-bite DP=ART=pexaon DP=mWrattleanee 'When a pexson gets bitten by a rattlesnd~e~~
'When the lightning struck the house an 3aley Street in Santa Bazbaxa, *
In ( 3 5 ) the d i r e c t object pxecedes the subject whereas
in (36) the subject precedes the direct object. Unlike
preverbal NPs i n ( 3 3 ) and (341, postverbal HPs axe
typically muked with the dependent proclit ic hi=, a8
shown i n ( 3 5 ) and (36)m
In chapters 3 and 5 , I will discuss in more detail
the statistical skewedness of these different
constituent orders and the motivations behind c h o i c e s
among them in discauxse.
Xn terns of the ordexing of elements w i t h i n noun
phrases , though the modifier o f t e n follows the head, the
apposite order [ i. em, the madif ier + the head) is not uncommon. The following examples illustrate these t w a
types :
( 3 7 ) JH59.345L-347Lt txt 59, line 001 59.345R
(38) JE59.20R-2lR, t%t 4, line 001 59.2X5
We saw above that pzonodnal prefixes mark the subject
on verbs. As seen in (9) arid (IQ), they are also used
to maxk possessors on nouns. In (37) the third person
singular pronominal prefix marks the possessor of
#fat0# 'sheeptm As can be seen' in ( 3 7 ) the modifier
sheep * follows the bead @ fat F and in ( 38 ) the modifier
'long ago0 precedes the head @Indiansg (the modifiers
are underlined).
This section presents a brief overview of clause-
caxnbining in Barbarefio Chumash. Included axe a relative
clause constructiont a sexial verb construction,
sentential subjects and sentential camplements,
conditioaal and time adverbial clau~es~ reason and
purpose clauses, and sequential clausesm
In relative clause constmctiona, the relative
clause follows the head:
?<a house which they lived in at the ai~sion>~
In (39) the head is the subject o f the relative clause,
and in (40) the head is the direct object af the
relative clause. In (41) the head appears to be the
locative oblique of the relat ive clause. In a l l of
these cases, the relative clause is marked by a
dependent p s o c l i t i c hi=, which, as I showed above,
appears on nouns. In ( 39 ) , there is no third pexson
subject prefix g= on the verb marking the head gpersonO
who ta lks . This lack of the prefix is frequently
found in relative clauses.
Two verbs me often juxtaposed, and when t h i s
happens, the first verb seems to indicate the manner or
the aepect in which the event expressed by the second
verb takes place:
( 4 3 ) J.3590275R-276L, txt 31, a line O O X S9.27SR
gyuxl6 - &K hisn&& h&, g - m l e l e k bi*m-R-~+nah 3-tiptoe DP=3-R-R+gu 'he goes t i p t o e i n g (raising h e e l s and going an balls ef feet). @
( 4 4 ) JE5gO525R-S2'7Lr t x t 100, Line 004
lcid~asiyn6 -W& h i ~ i y a n 8 i . n ~ kint+ka =S-iy-nowon h%=s-iv-angin adith%n=3-PL-stap D-3-PL-8at.a.meal *<and then they would step mating>'
As can be seen i n each of these examples, the t w o vexbs
take the same pronoun (S-iv- in (42) and ( 4 4 1 , and in
( 43 ) ) . The secand vexb is marked by a dependent
Sentential subjects and aentential camplements
follow the main c l a w e :
(48) JH59-209L2O9Rf t x t 20, line 003 59.2095
rThey have no word in the Indian lainguage meaning "to lockmf
In each of these examples, the main clause is followed
by the sentential subject or the eentential complement.
In this type of c o n s t ~ c t i o n , the sentential
subject /complement is typically marked by a dependent
proc l i t i c U, which is seen in examples (451- (48) . In
which is a direct quote.
Below we will examine seve ra l repxesentative
condi t ional and t h e adverbial c ~ a u s e s i n Barbareiia.
These tmes of clause^ appear e i t h e r before or after the
main clause, thaugh it h much more common to f ind them
before t h e main clause:
siYeq8t6 hinab&? h h a ~ t u h t u h & y S-iv-erne8 hi==noho+-? hi==na =s-R-tuhay 3-PL-sing DP==much+EH D-hens3-R-rain
H U B T m E Vl%ey sing a l o t when it is raining.'
As can be seen in these examples, the proclitic na=
means e i ther if OS @when m In ( 50 the p=-marked
c l a u s e precedes t h e main clause, and in (51) it follows
the main clause. I n t h e farmex, the main clause is
marked by a prac1iti.c 2i=, sad i n the liittez the
dependent c lause is marked by a dependent p r o c l i t i c hi.=.
This is a t y p i c a l pa t t e rn associated with condit ional
and t h e adverbial clauses: when the dependent clause
precedes t h e main clau8e, t h e main clause is marked by
t h e p r a c l i t i c 3i-; and when the dependent clause follows
the w i n clause, t h e dependent c lause is marked by t h e
dependent proc l i t i c &. Here are nore examples:
- UmD
*If a conet leeks red or ha8 a xeddish color,
cam if they have thoughts.'
a-iv-R+anSin -
-? he?=l =~+?invu - - 7 when 3-PL-R+eat.a.meal-? PRX=ART=R+Indian-EM
TInE #-en Indims are eating a mealr>
- ?i=s-iy-e-tieawi& ?X=3-PL-N-talk
mmi they don't talk.'
The particles ham and may mark conditional and
time adverbial clauses respectively. As can be seen i n
( 52 1 - ( 55 ) , they can occux befoxe or after the main verb.
The following examples i l lu8 tra te the use of reason
purpose clauses:
lmxle 'We &ink the ocean
to vomit a t both enda m
The reason clause is mukeci by a part ic le b y k & and the -
puxpose clause is maxked by a p r o e l i t i c Eukanu. Shorter
foxm @uk4, m, and are often used to mark purpose
clauses. As can be seen i n these examples, these types
of clauses follow the main clause.
Finally, the following example illustrates haw
sequential events are typically expressed in Barbarefio:
( 5 8 ) JH59.36R-40L, txt 175, line 015 59.38R
*The next a.m. very early the rooster crowed. His crowing sounded very sad.
h i k a s h - han isiw?I-won Jrii+kq==s-sdan hi=$-R+iwon DP+so~3-continue-to DP=3-R+aound And he kept on crowing
<and theo ld man got sad at him,>'
Each sequential clause is marked by a combination of
2 . 6 . Summary
I have presented an overview of the grammar of
Barbareno Chumash so that readers may have enough fac t s
to follow the rest of the dissertation. Spec i f ica l ly we
examined dis t inc t ive sounds, basic morphological
structure and lexical categories, argument structure and
pronominal affixes, constituent order, and clause-
35
combining.
Motes for Chapter 2
1. The examples i n t h i s dissertation are given i n the
format illustrated in the following example:
1 JE59.3L-SR. t a c t 166, GN line 013 59.5L \\
2 hilled h i l t ipA -wil hilr?kiyaKllir, 3 h i d =h hi-l çtipawi hi=he?*k-iy-?aqli& 4 DPçARTçpera DPÈARTçtal DPÇPRXçl-PL-langua 5 a person who talks our <1anguaqe>.
Line 1 gives the text reference information. In the
above example, 'JH59.3L-5R' indicates that the example
is taken out of the text which appears from the Left
page of frame #3 to the r ight page of frame t 5 of reel
# 5 9 of the Earrington microfilms. (Each frame contains
t w o or ig ina l pages. ) 'Tact 166 * refers to the number
assigned t o the t e x t and ' l i n e 013 refers to the line
number within the text, both assigned by Suzanne Wash.
' G N P indicates t h a t the example i s taken from the
grammatical notes found with the text, which Harrington
often gave. Finally, the last item on the reference
l i n e '59.5Lr indicates the exact location where the
example is found: the left page of frame #S of reel #59.
Simi lary 'MB, FN2, p74' would mean that the example is
taken from page 74 of Heelerrs f i e ld notebook # 2 . Line
2 gives the way t h e original data w e r e recorded. Some
substitutions in Harringtonss original orthography have
6 - h 9 been made. They are as follows: g -> g, -> -, - -> 2.e
Glottalized consonants as written by Harrinqton (e.g., ^\ S, G, g, S, ff, 2) are written here as &, U, g, &, A,
L. Line 3 and line 4 are the morphological analysis and
the morpheme-by-morpheme gloss respectively, given by
Suzanne Wash or myself. A hyphen ( - ) and an equal sign
(=) on those lines indicate an affix and a clitic
boundary respectively. Though not illustrated in the
above example, a plus sign (+) there would indicate a
morpheme boundary in a lexicalized expression. Line 5
gives the English translation for the example. If the
translation is given in c>, it means that it is provided
by Suzanne Wash or myself because either it was not
provided by the original recorder or it was given in
Spanish. Often, the original translation does not
exactly cover the meanings of all the words and the
morphemes found in each example; it represents only what
the relevant example basically means. I chose to
represent exactly the translations of the original
recorders in order to stay as close as possible to what
is intended not only by them but also by the speaker
M a r y Yee. This may sometimes cause readers difficulty
in matching up the word-by-word gloss and the
translation* I will t ry to explain
2 . The tern @dependent was first
(1995b) to describe bi*.
when necessaq.
introduced by Wash
3 * Certain verbs, such as eha @goodp, wil 'be' and
univiw znecessaryt, seem to prefer the pref ix 2 a L ,
e spec ia l l y when they have a sentent ia l subject as in the
following example:
( 3 0 ) JB59-327L-332&, tx t 138@ line 021 59.331R
3. Argument s t ruc ture
The puqpase of t h i s chaptex is t o describe the
argument s t ruc ture of Ba.rbaxefio Chumash* By examining
t h e pronominal system of Barbarefio, I will first suggest
t h a t certain arguments of verbs are t r ea ted as integxal
parts of t h e vexb because they are expressed i n t h e form
of gr-tically required psonaminal a f f ixes . I w i l l
take this as c r i t e r i a 1 far t h e claim t h a t they are core
arguments m Next I w i l l show t h a t f a c t s about
constituent order corxelate with t h i s observation: when
both core arguments and obliques are expressed as f u l l
NPs i n clauses, the core arguments d i r ec t l y follow t h e
vex% and are folluwed by t h e obliqnes. Then I w i l l
describe t h e m o ~ h ~ l o g i c a l , syntactic, and pragmatic
c h ~ a c t e r i s t i c s associated with the two verbal suffixes
-us and -in* I w i l l suggest t h a t these suffixes are
core =meat-bui ld ing devices, devices umd t o create
d i r e c t objects from argwnents t h a t might atherwise be
c a s t as obliques. The presence of these devices i n the
grammar i t s e l f fur ther suggests t h e importance of the
core-oblique d i s t inc t ion i n Barbaxeiio* Finally, I w i l l
b r i e f l y discuss the gsamauatical s ta tus of pat ients
occurring with verbs maxked by ane uf the core argument-
building devices*
nodnative-accusative pattenling i n which t h e agent
argument of trans i t ive clauses and the s ingle argument
of intrans i t ive clauses are treated in one w a y and the
patient argument of t z a n ~ i t i v e clauses is treated in
another way. W e observed this patteming with re lat ion
to the pzanuminal marking on the verb. L e t us go back
to the examples in 2*3. which i l lu s tra ted this
In these examplea, the subjectss me marked w i t h the
prefix cm the verb regardless of t h e i z agency or the
t r m s i t i v i t y of t h e verb, and t h e r e is no mzufing for
the ciixect abject* Thus the pronominal marking on t h e
verb exhibits nodnative-accusative pa t teming .
I n 2*3., w e further examined the spec i f i c s of this
pronominal marking, and observed t h a t it makes fairly
e labora te d i e t ine t ions for both subjects and dixect
ob jec t s i n tern of pexson and number (see Tables 3 and
4 in chapter 2 ) . What seems to be most s i g n i f i c a n t
&out t h i s pronominal system is t h a t i t s e l abora t e
marking is qrmatically xequired and only certain
arguments of verbs are marked by it. Xn terms of the
argument s t r u c t u r e of t h e verb, t h i 8 f a c t seems t o
suggest that the arguments expressed by t h e pronouns may
have a gpecia l s t a t u s i n t h e g r m x of Barbaxefia
Chumash: they are core arguments of t h e verb (Crof t
1991; Thompson t o appear )*
The present hypothesis n ice ly agrees with t h e
moqhologicd, syntactic, and p~agmat ic c h a r a c t e r i s t i c s
associa ted with t h e t w a verbal su f f ixes -us and -in
discussed i n 3 . 3 * : t h e s e me devices which give coxe
s t a t u s (as d ixec t ob jec t s ) t o arguments t h a t would
otherwise be oblique* I n o ther wards, t h e behaviar of
these suf f h e s farthex demonstrates that the distinction
between coxe azquments and obliques plays an important
role in the grammar of Bubarefio. Befare we examine
these suffixes, howeve=, I will discuss another area of
BazbaxeEo grammar which displays the impartancze af this
distinction: constituent order.
3 . 2 . Constituent order
1x1 accordance w i t h the pronominal m~king on verbs,
facts of constituent order seem to correlate with the
core-obli-e distinction made for t h e arguments of
verbs* Specifically, when both core arguments and
obliques appeax as Cull SIPS in a clause, nomally the
core arguments follaw the verb and are then followed by
the obliques. X t has been found cross-lin~istically
that the subjects of transitive clauses are rarely
expressed as full NPs in discourse (Du Bois 1987 ) . I
found this to be the case with Barbaxefio* (Far more
detail, see chaptez 5 1 In the unusual instances in
which transitive eubjects are expressed as full NPs, it
is extremely rare to find them with obliques, which are
=extra argumentsa. For this reason, below we will only
look at the order of intransitive subjects and
transitive dimet objects in relation t o abliques
[ 4 ) JHS9.6L-?S, 3, line 002 59e6R
( 1 0 ) JH59*96Z-97R8 t%t 18S8 l i n e 001 59-96L
h- - ii k i & - k ~ ~ k t t ~ ~ ~ ~ hifa? hiho~wdUd$, hi=mafi k-i&-R-IcupuiS us1 =?o? &&,Q- D h h e n l-DW-R-dip DBARTwater DP-DIS=ART'-bucket
v Do/ PM! SuURcE 'when we are dipping water out of a bucket,'
(11) JH59*676R-681L8 txt ?S8 l i n e 005 59.679R
(12) JH59*474L-475R8 txt 83, line 005 59.474L
(13) m59.518R-S20G862%G629L8 ts& 9T8 line 016 59.628% \l
?apS kigeKw&l *i, ?apZ k-i5-eqwel =I =a let 1-DV-maJce D-3-seat DP=ART=IDF
W m/Pl4T =C * L e t s make a seat for 8amebody,#
(14) JE59.466L-467R8 txt 798 l i n e 007 59.466L
v m / P ~ msTR 'she had her face c o v e r e d with a cloth8
Notice that there is no formal marker which indicates
the semantic r e l a t i o n s h i p of these obliques to the verb
(Beeler and Whistler 1980). In each of these examples,
a core argument precedes an oblique in a clause
regardless of the semantic type of the oblique.
S p e c i f i c a l l y , i n ( 4 ) - ( 8 ) , the subject precedes the
oblique i n btrmsitive clauses and in ( 9 ) - ( 14 ) , the direct object precede8 the abLique in transitive
clauses. This particulax ordering can be interpreted as
i conica l ly motivated: aquments semantically more
closely t i ed to the verb are placed closer t o it than
less c ~ u s e l y t i e d anes. This further suggests that the
grammar of Barbmefio attends t o t h e ~ s t i n c t i o n between
core arguments and obliquese
As we wiL1 see i n the next section, certain
argument8 which might otherwise be oblique are cast as
core arguments when one of the core u m e n t - b u i l d i n g
s u f f ~ e ~ i s attached. Xn such cases , these arguments
are normally expressed immediately a f t e r the verb,
ref lect ing t h e status change. This order further
supports the present analysis.
3.3* Care arwent-bui ld ing device3 -us and -in
E3axbaxefio has two verbal suffixes, -us and -in,
which change the =guntent stricture of the verb.
Specif ical ly , the w e of these su f f ixes indicates t h a t
certain arguments which would otherwise be obliques are
direct objects of the verb, that ist cure arguments. l
This type of device ia found i n many other languages and
1988; Croft 1991; Mithun 1994; F a h e r 1994; Ichihashi-
Nakayama 1996; Shibatani 1996; Austin to appear).
Belowp we w i l l discuss the mo~holagical, syntactic, and
pragmatic eh~ac tex i s t i c s assoc iatedwiththe use of -us
and -in. These c h e a c t e r i s t i c s w i l l indicate t h a t
relevant arguments have direct abject status ( L e , core
argument status 1 . W e will f i x s t discuss the suffix -us,
which gives direct object status t o datives, and then
discuse t h e saffix --in, which gives direct object status
3.3.1. The dative -us
The vexbal s u f f i x -us CO-occuxs with a certain type
of NP i l lu s tra ted the following examples:
(15) JH59.563R-57lR, t a c t 107, line 020 59*566L
(17) JHS9.4OR-44L, txt 176, line 008 59.41R
( 2 0 ) JH59.63lR-632R, t x t 127, line 008 59-632s
?and have the me who is to be a, mother <&ink its'
Note that, in ( X 6 ) , the patient 'itr refers ta pitch or
~ l k e e d - j u i c e , which is not expressed as a full HP.
Readers may xemember that, in & ~ ~ b a r e f i a ~ f u l l W s axe
not requixed fox a clause to be grammatical, and that
the third person singular direct object has no mulcing
on the verbe Note also that, in (201, the verb @drinkr
appears w i t h the causative prefix su-, and the @mother8
is t h e person who i~ made to &inke The patient @it8
refexs ta tea, which is not expreesed as a full NP.
Naw, in all of the above examples, the suffix -us
is found at the end of the verb and i.8 followed by a
f u l l HP expressing an e n t i t y which seems to be b e s t
characterized as t h e t a r g e t of an a c t i v i t y or emotion,
o r an e n t i t y i n d i r e c t l y affected (Blake 1994 ) . Further, t h e r e fe ren t of t h e NP co-occurring with -ua is a
sen t i en t being, a na tura l self -moving e n t i t y like 'wavef
i n (19 ) o r an e n t i t y subjected to a re l ig ious a c t i o n
because it is associated with a supernatural being such
as 'a rock p lace which was blessed by a p r i e s t because
t h e Devil was appearing the re* in (18) . Similar sets of
NPs are found to be l i n g u i s t i c a l l y t r ea t ed as a category
i n d i f f e r e n t languages and are of ten called 'dativef
(Blake 1994; Palmer 1994). I w i l l adopt t h i s term to
refer t o t h e type of NPs i l l u s t r a t e d i n t h e above
examples.
The use of the dative suffix -us seems to be
lexically determined t o a l a r g e extent. There are
several observations which suggest this analysis.
F i r s t , only c e r t a i n verbs appeax with us. I n examining
t h e data , paying a t t en t ion t o t h e cases i n which -us
appeaxs, one not ices t h a t t h e verbs that occur with -us
are r a t h e r limited i n number. Second, some verbs seem
always ta occur with t h e s u f f i x -us; 1 have not seen
them without -ua i n my data. In te res t ing ly , t h e s e verbs
together with -U% a ~ e tzeated a8 independent l e x i c a l
i t e m s by earlier researchers including Whistler (1980),
and fo- without -us are not found in these suuxces.
One such example is t h e verb yikus # t o give * , seen i n
( 1 6 ) above. Notice t h a t t h e event described by t h i s
verb always involves a r ec ip i en t , which may be
responsible f o r the l ex ica l i za t ion of t h i s verb
involving t h e d a t i v e recipient . Third, t h e r e a r e o ther
verbs which contain -us whenever a da t ive argument is
present i n t h e ut terance; speakers do not seem t o have
had a choice . One such example i s t h e verb a ' t o sayp
seen i n ( 15 ) above. Whenever 'a person who is ta lked
t u g is discussed by t h e speaker, ? i ~ appears with the
suffix -us. Fourth, of ten t h e semantics indicated by
the use of -us is r a t h e r id iosyncrat ic ; each verb seems
to have had a p a r t i c u l a r meaning associa ted with t h e use
of --uso That is, t h i s suffix sometimes ind ica tes t h e
r ec ip i en t of t h e event, o ther times t h e beneficiary of
t h e event, depending on t h e verb; o the r meanings do not
seem t o be possible. So, f o r example, t h e eu f f ix -us on
t h e verb alwaye indicates the person who is ta lked
t o , not t he benef ic iary of t h e event, which should
c e r t a i n l y be a p o s s i b i l i t y i n a sentence l i k e 'I s a i d
that f o r himr. These observations seem to suggest that
the uses af -ua are laxgely lexically determined*
The addition of -us to the verb seems ta change its
argument structure. Spec i f ica l ly , the dative acquires
core argument status by becoming the direct object. One
indication of this may be seen i n the f a c t that the
marker -12s is expressed as a part of the verb. There
are several other observations which support this
analysis* F i r s t , appearing after -us an the verb, the
3rd person plural suffix -wun marks the plura l i ty of the
dative, not of the patient. Examine the following
examples :
(21) JH59.70R-76R, txt 181, line 021 59.76R
' they bring food to the Indians*Â
( 2 2 } JS59.214L-214R, txt 222, line 002 59.214L
In (21), it is 'the Indians which is plural , not
'food', and the verb takes the th i rd person plural
markex -wun a f tex the dative marker -us. Similarly, in
( 2 2 ) , it is ' the people' which is plural , not ,it ( a
whale) c , and t h e wezb again takes t h e same suf f ixes -us-
wun. W e have t hus seen t h a t the dat ives appearing with - -us a r e treated moqhologically like d i r e c t o b j e c t s a
Second, a f a c t about cons t i tuent oxder a l s o
supports the psesent analysie. In t h e above examplesf
the da t ive NI? directly follaws t h e verb i n a clause.
How, when both the dative and the pat ien t are expressed
as f u l l Ws in a clausef t h e da t ive typically follows
the verb and i s followed i n turn by the p a t i e n t , as
shown in ( 2 1 ) and in t h e follawing example:
I n these examples, t h e verb is masked by t h e s u f f i x -us,
and t h e d a t i v e W precedes t h e p a t i e n t i n t h e clause.
This order is exact ly the apposite of the ane shown by
t h e p a t i e n t direct object and the oblique i n regular
t r a n s i t i v e c lauses , discussed i n 3 - 2 . We have thus seen
that when d a t i v e NPe appear with the suffix -us on the
vezb, they are treated sptactically l i k e direct
ab jects . Finally, when we examine t h e s e m a t i c e of datives
as well as the discourse cantexts i n which they appear,
w e f i n d t h a t they are cen t r a l p m i c i p a n t s i n the event
described by t h e verb, e~pecially compared t o t h e
pa t ien t . It was noted above t h a t t h e r e are Borne verbs
which a l w a y s accur w i t h t he suffix -us and there are
ather verbs which occxu with the suffix -us whenever a
da t ive is present in the ut terance* 1 would like t o
hypothesize t h a t t h i s may be related to the often-
propcased euqgestian t h a t humans like to t a l k about
themselves {Giv6n 1983 Thampson 1990 ) . That is,
because af t h e inherent semantics of dat ives ( L e * ,
humanness), it i~ l i k e l y t h a t they aze viewed by t h e
speakexxi as c e n t r a l pa*icipants i n an event*
Especia l ly w h e n thexe are both a d a t i v e and a patient i n
the uttezance, it is l ike ly t h a t t h e speaker views the
da t ive as a more c e n t r a l pa r t i c ipan t to t h e event than
the p a t i e n t because the fomer is normally human but the
lattez is not m It seems that far this reason datives
are t r e a t e d l i n ~ i s t i c a l l y as an i n t e g r a l part of t h e
verb (i.e., by t h e verbal 8 ~ f f i . 3 ~ -us), just l i k e
subjects and direct objects in regular transitive
clauses. In othex wordsf I am suggesting here that
since datives are viewed as central p e i c i p m t s i n the
event, t h e i r pxesence itself normally guaxantees that
their involvement is expressed as a coxe iiqtzment ( i. e.,
t h e direct objec t ) and accordingly they appeaz with the
However, there are a f e w cases in which the
presence of a dative doe5 not seem to be enough to
result i n t h e use of -us. One such example is seen
regarding t h e use of the verb euwel ,makeg, which
typically appears with -us when the dative i s present i n
the event:
(24 ) JH59.27L49R, tact 172, GN l i n e 004 59.29L
We have already seen one exception to t h i s tendency
in (13) abovef reproduced here with the en t i re t e x t as
Twa little g i r l s w e d ta play making a seat with by pasping each athests arms. One Little gixL would with her zight hand gxaap he= own left a m neaz her hand, and the other I L t t L e girl wauld do likewise. Then with their left hands they wauld grasp each othexxe m, amking a square seat, on which a t h k d little gLr1 would sit. They wd say:
In this example, the vexb appeaxe without the
dative suffix -us and the dative rec ip ient HE! tsameane F
is expressed aftex the patient direct object NI? @a
seat? As can be seen, the dative 8someoner is non-
referential and nan-identifidle* The text ie about a
game of making a seat using hands* In the context in
which this example appearst the speakex is saying that
when they didnet have anything else to do, they played
t h i s game of making a eeat* Thus neither the
referentiality nor the identifidility of t h e dative
seems to really matter here. Probably for that reason,
it is the patient seatF which is treated as a core
argument { i . e , the direct object ) , not the dative in this example*
So these examples suggest that datives, unless they
are playing a minor role in the discauxse as seen i n
( 25 ) , tend to be viewed as central peicipants in events because of their inherent semantics (Le.,
humanness)* Accnrdingly, datives are treated as central
peicip-ts also in the language by being expressed as
an integral part of the verb. In other waxds, 1 am
suggesting that speakers axe indicating the centrality
of dative peicipats by expressing their involvement
as core arguments.
Thus the m a ~ h a ~ a g i c a ~ , syntactic, and pxagmatic
facts all seem to suggest that when the verb is marked
with -us, the dative becames the direct object, and this
suggests that the suffix -us is a device for building
care arguments. Interestingly, Barbarefio has another
device which turns semantic instruments i n t o direct
abjects. This is the t op i c of the next section.
3 . 3 . 2 . The instrumental -in 3
The instmental marker -in seems tu be another
core arvent-building device: it allaws speakers to
cast semantic instruments as direct objects. There are
several ch~actezistics associated with this suffix
which are similar to the ones which we saw regarding the
dative marker -us in the last section. Examine the
following examples$
(27 ) JH59.551L-552R, txt 105, line 002
(28 ) JH59.535L-538R, txt 102, line 015
*There w e r e 2 kinds of money made of <olivella>*
In these examples, the suff ix -in is attached to the end
of the verb and indicates the instrumental function of
the associated HP. In ( 2 7 ) , the patient direct object 'itr refers to a child, which is not expressed as a full
NP. The reader nay remember that in Barbareno, full NPs
axe not required for a clause to be grammatical, and
that the third person singular direct object has no
marking on the verb. As discussed in 2.3., the s u f f i x
-an, which appears between the verb a d the suffix -in
i n (281, indicates the plurality of the patient direct
object 'two kinds of moneye.
One thing which should be pointed out here is that
the term 'instrumental' is used not only for roles like
that i n ( 2 7 ) , which contains a canonical example of an
instrument, but a l s o for roles l i k e that in ( 2 8 ) , in
which 'olivella' is clearly not an instrument for making
money. A better descriptive term for the latter may be
something like Omaterial* or 'source8. It seems that
both of these types of objects represent what the agent
manipulates in order to make the event take place. I am
using the tern 'instrumentalr for both types partly to
account for the fact that these two somewhat different
roles are indicated by the sane marker in Barbarefio.
There are several phenomena which suggest that
instruments, like the datives discussed in t h e last
section, are dizect ub jects when they occur with -in.
First of a l l , when the instruments are plural, their
plurality may be indicated by the third person plural
marker -wun which i s suff5xed af tez the instrumental
marker -in:
( 2 9 ) JH59.313L-316R, t x t 163, l i n e 005 59.314R
'The Indians had many kinds of seeds which they used as food.
They used to make pinole out of them or boil them as mash.'
The suffix -WUQ marks the plurality of the instrumerit
@seedsr which are used to make 'pinole*. This is
similar t o the use of to maxk the plurality of the
patient direct object i n a regular transitive clause.
W e have thus seen that the instruments appearing w i t h
- i ~ are treated moqhologically l i k e direct objects.
Notice in the second line of (29) that the same referent
'seedsr i s kept track of as the patient direct object of
the transitive verbs 'to boil' and ' t o mush' by the
plural suffix -wun.
Constituent order patterns also support the present
analysis. In the above examples, the instrumental HP
directly follows the verb in a clause. Interestingly,
when the patient is also a full HP, it typically f allows
the instrument, as shown in (28) and the following
examples :
(30) JH59.18lL-l82L, txt 17, line 001 59.1811,
(31) JH59.367R-38E, tar t 142, GÃ line 004 59.370R
gKili?~~w6 - i$n h i ~ p 6 h h+lxÈ6 hilstok, S-qili-eqwel-?-h Jii-S-wft ha .4 =%go? bi=i =stuk 3-H&B -sake -?-=S DPm3-wood DPmARTç8yca3~or DP=ARTÈbow
V IMSTR PAT 'he always made a bowl of sycamore wood. '
Notice t h i s order is exactly the opposite of the one
taken by the patient direct object and the oblique i n
regular transitive clauses, examples of which were
discussed in 3.2. We have thus seen that when
instruments occux with the suffix -in on the verb, they
are treated syntactically like direct objects.
Finally, when we examine the roles of these
instrumental arguments in the discourse contexts where
they appear, we find that they are treated as more
central participants than the patients at the level of
discourse. Examine the following example:
(32) JH59.165L-166L, t x t 202, line 00lab 59.1651.
*The people long ago used to be in a hurry to get ready for war , they would take deer-arrows and break each in t w o , maJu.ng two arrows out of one,
hiiaba? hilkuhku ? . hi=l =?&h&? hi=l =R+ku -? DP=ART*aany DP=ARTsR+person-EM
PAT
and w i t h the short arrows "they killed many people.'
This is the entire text. Though it is untitled, one can
tell from the content that the text is about the arrows
(instrument) not about the people (patient). As can be
seen, the instrument 'short, arrows' occurs with the
suffix -in on the verb, and in terns of constituent
order, the instrument precedes the patient.
Examine ( 3 3 ) , which is a reproduction of ( 2 9 ) with
more context:
(33) Title: Native Seeds Do N o t B l o a t One
JH59.313L-316R8 txt 163, line 001 59.313R
Â¥<Thei seeds that they used to grow here and there, after they would eat them,> they don't bloat you. Only the brought-in beans produced bloatedness. The Indiana had many kinds of seeds which they used as food.
a) siyKwelinvun ilpin6 - li s-iy-eqvel-in -win hi=l q i n o l i 3-PL-make -IHS-3P0 DP=ART=pinole
PAS They used t o make pinole out of them
b) kehafcu aiya&lnwun siysumoSwfin. k e h h s-iy-?a*in-wun a-iy-su-moS -wun or 3-PL-boil -3P0 3-PL-CA-mush-3P0 or boil them as mush.
None of these seeds bloated a person, as beans do. The Indians had to learn to stand eating beans as Mexicans do.'
As the content as w e l l as the title suggest, t h i s t e x t
is about Chumash native seeds. The seeds are more
central to the text than 'pinole8, which is mentioned in
passing. As shown in a), the instrument 'seedsp appears
with the suffix ?in an the verb. In the follawing
clauses given i n b ) , the referent 'seeds8 is sustained
as the patient dizect object by the suffix -wn.
Cases such as these seem to suggest that because of
the centrality of some participants in the discourse,
they are treated as such also i n the language by being
expressed as an integral part: of the verb, much as are
subjects and direct objects in regular transitive
clauses. In other words, I am suggesting here that
speakers are indicating the centrality of those
participants by expressing their involvement as core
arguments.
Supporting the above observations, instruments
expressed as obliques do not s e e m t o play a centra l role
in the discourse. One such example i s ( 14 ) above, which
i s reproduced here with t h e entire tex t as ( 3 4 ) :
( 3 4 ) JH59.466Lv467R, txt 79, line 001 59.466L
'There were some Japanese wood-choppers cutting chaparral at San Rogue. It was chaparral 'that they were cutting. It was a single family, consisting of a man and a woman and some children. The woman was cutting wood the same as the man --
-> kesiKm&ywaS iho?stek iio^row i l a a x i - & S , Re =S-iqmay-wag bi=ho?=s-tier -? hi=l =?o&ow h i = l =inaxaki& and=3-cover-PST DP=DIS=3-face-? DP-ARIawhite DP=ARTÈclot
DO/PAT IHSTR she had her face covered with a cloth
She said that it is so I donet get tanned by the heat. '
In the utterance marked by the arrow, the verb is not
suffixed by -iq. Further, the patient 'her face'
directly follows the verb and is followed by the
instrument 'white c loth' . These facts suggest that 'her
face* is the direct object and 'white cloth* is an
oblique argument of the verb. The text is about a
Japanese woman trying to keep her face from getting
tannede4 Thus #white cloth' i s probably not as central
ta the s t a r y as %er face' itself. That fact seems ta
be responsible for the fact that t h e former is expzessed
as an oblique and the latter the dizect abject.
Ta sum up, moqhological, syntactic, and pzagmatic
patterns a l l seem to indicate that when the verb is
marked w i t h -in, the instrument has core argument status
as a direct object, which suggests tha t the euffix -in
is a core awent -bu i l&ng device; it turns instruments
inta direct objects.
3 * 4 * The status of the patient
One thing we need to discuss befaze we c lose t h i s
chapter is the grammatical status of t h e pat ient
occurring w i t h a verb marked by one o f the core
arment-building devices. The different areas of t h e
language which we examined above do not seem t o give us
a clear picture. In terms af murphology, the patient
can st i l l be marked by -wn, as can be seen in (28)
above, reproduced hexe as ( 35 ) :
The p l u r a l i t y of t h e pa t i en t @ 2 kinds of money8 As
marked by --Q which appears befaze the instrumental
suffix ? in on the verb. Obliques cannot be xnaxked with
-mn, which suggeste t h a t t h e pa t i en t i n (35 ) is t r ea t ed
a5 the direct object. However, s y n t a c t i c a l l y this type
of p a t i e n t is not t r e a t e d as the direct object: it
occupies t h e syntac t ic pus i t ion i n which obliques are
normally expzessed, as i l l u s t r a t e d in (35) . Finally, as
w e have discussed above, p r a m a t i c a l l y t h i s type of
p a t i e n t does not play a c e n t r a l role i n t h e d i s ~ o u r s e ,
compared t o t h e da t ive o r t h e instrument. Thus
examination uf these areas of the g r m a x suggests t h a t
the patient a c c u ~ ~ i n g with a verb marked by one of t h e
core a r m e n t - b u i l d i n g devices exh ib i t s only one of t h e
f ea tu res aseociateci with ciixect objects . This euggests
it may i n s t e a d be a rSecandary Object* i n t h e sense of
Dryer (1986).
3.5. S--
W e have looked at the ~ ~ e n t s t ~ c t u r e of
Barbarefiu Cht~mash. W e fkst saw that in terms of verbal
m o q h o l o ~ i t e x h 3 i t s n o h a t i v e - a c c u s a t i v e pat terning:
the agent argument of ii txansitive clause and the s i n g l e
argument of an i n t r a n s i t i v e c l a u s e are t r e a t e d i n one
way ( n a d n a t i v e ) : and the p a t i e n t argument of a
txansitive c l ause is t r e a t e d i n another way
( accusa t ive ) . Spec i f ica l ly , the farmer is expressed by
verbal pref ixes , and t h e lat ter by verbal suf f ixes .
These p a r t i c u l a r ug tments are thus treated as integral
parts of the verb by being expressed in t h e form of
g r m a t i c a l l y required a f f ixes . This fact seems to
suggest that they have a s p e c i a l s t a t u s with relation to
t he verbz they are core arguments. In te res t ing ly , there
is a syntactic phenomenon which carrelatee with t h i s .
Core arguments tend t o be expressed before ab l iepes i n
clauses: they are placed claser to the verb than are
ubliques. W e then examined 8evera l c h m a c t e z i s t i c s
associated with two vexbal suffhea -us and -in. These
c h ~ a c t e r i s t i c s 8uggest t h a t -us and -in axe core
= w e n t - b u i l d i n g devices that cast datives and
instruments as direct objects.' The pxesence of t h i s
type of device itself further suggests that the coze-
oblique distinction play8 an important role in the
grammar of Baxbarek Chumash* Finally, we saw that the
patient occurring w i t h t h e verb marked by one of t h e
core = v e n t - b u i l d i n g devices exhibits only one of the
characteristics associated with direct objects*
Notes for Chapter 3
l. Beeler and Whistler (1980:94, Footnote 13) first
suggested this analysis for -U@.
2. Earlier studies have suggested that -us is a pronoun:
in essence it has been analyzed as a dative pronoun
(Beeler 1970a, 1976; Beeler and Whistler 1980) . The
dative pronoun analysis of -us has certain appeal.
First, -us accupies the peaition in the verb where
direct object suffixes appear. Second, -us is not used
for the first and the second persons$
Unlike (19) above, the involvement of the second person
is indicated not by -us but by the direct object
pronominal suffix A. If -us is a core argument-
building device as the present study suggests, this is
a puzzling phenomenon: -us should also appeax in the
above example. Based on this type of data, the dative
pronoun analysis would present a pranount&le including
-us, as shown belaw:
Table l: Baxbaxefio d ixec t object pronouns
SG (PAT) SG (DAT) MOH-SG (PAT) NON-SG (DAT)
l -it -it - iyuw -iyuw
2 - in - in -iyuw -iyuw
3 -us -wun -us -wun
how eve^, t h e r e is a problem a l s o far t h e da t ive pronoun
analysis: why does only t h e t h i r d person make a
d i s t i n c t i o n between p a t i e n t and dative?
The present d i s s e ~ a t i o n instead analyzes -us a s a
core a r m e n t - b u i l d i n g device (appl ica t ive) f o r t h e
following reasons* ( B e e l e r and Whistler (1980) a l s o
suggest t h i s p o s s f i i l i t y . ) F i r s t , the highly
lex ica l ized nature of -us suggests it is a der iva t iona l
morpheme: t h e function of 'core = v e n t - b u i l d i n g p is
much more l i k e l y t o be expressed by der iva t iona l
morphology than are pronouns. Second, more importantly,
the dat ive pxanoun ana lys i s would need t o account f o r
t h e moqholagical , syntact ic , and pragmatic
c h m a c t e r i s t i c s associated with -us, which all suggest
t h a t it i s a core = w e n t - b u i l ~ n g device. Pronouns
a r e not known f o r 8uch cha rac te r i s t i c s* Third, under
t h e da t ive pronoun analysis, m e would a l s o have t o set
up a category @ i n s t m e n t a l pronoung f o r anather suffix
-in, which, as w e w i l l see later in this chapter,
instruments.
3 . Many of the idea8 expressed in this section were
first presented in Ono et ale (1994b].
4. It should be mentioned that for a woman to keep hex
face from getting tanned (that is, keeping it "whiteu}
is traditionally valued in Japanese c u ~ t u ~ e .
5 . There axe t w o other s u f f i x e s -pi and -mu2 which may
also be care =vent-building devices* These suffixes
are used mostly as locative nodnal izers as in$
(36) ~59.518R-S20L,628&-629&, tx t 97, l i n e 002[2]59-518R
(37 ) J859.458R-464, t a c t 158, line 024 59.463R
@ the storelaouses @
However, 1 have found a few interesting examples:
tThere in the willows was where the bad women hung around far men.
In ( 3 8 ) , the locative - ~ i appears as part of the verb
and s e a s to indicate the location where the moon was
shining: on the mussels. Interestingly, - ~ i is further
followed by the third pexson singular direct object
suffix - a n a In (391, the location pwillowsr where the
bad women hung around for men is expressed preverbally
(see chapter 5 regarding this constituent order), and
the locative -mu2 is suffixed to the verb and again
ioll~wed by - w n a I have not found enaugh examples in
my data to determine whether -mi. and -mu2 are core
= v e n t - b u i l d i n g devices. The rare occurrences of the
type of examples illustrated in ( 3 @ ] and (391 could be
precisely because locations are usually m t viewed by
the speaker as central t a the event, and thus are not
expreeseci as direct objects. Further investigation of
these morphemes is necessary.
4 . Refez-ential choice
Like a the r languages, Barbare60 has several
d i f fe ren t ways of r e fe r r ing t o a given part icipant
(e. g. , diffexent f amus Feitb, t h a t wf and can r e f e r
t o exact ly t h e same man i n English) (Clancy 1980; Giv6n
1983; Chafe 1987; Fox 1987, 1996). In t h i e section, I
w i l l d i s c u ~ e r e f e r en t i a l cho i ce i n Barbareiio and examine
factors which motivate t h e use of d i f f e r en t forms i n
discourse. W e w i l l f i r s t l a ak a t the two most cwunonly
used r e fexen t i a l f o m i n Barbarego discourse:
pronominal a f f i xe s and f u l l noun phxases. Then w e will
look at two minor forms: independent pronouns and the
proximate demonstrative he?. W e w i l l see t h a t the use
of each form is associated w i t h specific factors which
are manifested i n discourse con t ex t s*
4 . L Pronadna l a f f ixes
I n the preceding chapters, w e saw t h a t verbs i n
Basbarefio are marked by pronominal affixes for core
anpmentsf and t h i s is a grammatical requirement f o r
Barbare60 verbs. This requirement has been
g r m t i c i z e d to such an ex ten t t h a t the re axe even
cases of pronominal a f f ixes which seem ta be present
only t o satisfy it. This is particularly evident in
subject pronouns, illustrated in the following examples:
(1) JH59.667I.-668R, txt 71, line 024 59.668R
"then they have to throw the <bowl> awayt
( 2 ) JH59.327L-332I,, txt 138, line 021 59.331R
?a l e t ih6 hisaakaywun iltoht6? ?al-e-Eho hi=s-am -kuy -W h i 4 =R+to -? MM -8-good DPÈ3-IDF-talce-3P DP=ART=R+mussei-EM '<It is not good to gather mussels>'
As discussed in chapter 2, the prefix 2al- in (2 ) is
used as a replacement for the thizci person prefix e.
The use of t h e subject pronouns i n these examples seems
to be motivated by grammar: their main function i s t o
refer to the s i tua t ion expressed in the embedded clause
(similar to the 'it - thatR construction in English). The following set of data further Illustrates the
highly grammaticized nature of Barbareno subject
pronouns :
( 3 ) JH59.128L-l28R, txt 11, line 004 59.1281.
s h a ~ t ~ h t t ~ . 8-8 a? -tuhuy 3-FUT-rain 'it is g. to rain.?
ijAi noho saxtatax; ikhu noho &-&&tax but vezy.much 3-cold * B u t i f s very ccold;
'11 first make a fire.'
( 5 ) JH59.181L-182L. tact 17, line 003.8 59.1811.
tguma?ii p-t £ -mai l g-naxyit and=when 3-be. morning 'in the morningf
In ( 3 ) - ( 5 ) , none of t h e sub jec t pronouns is r e f e r e n t i a l .
They are t h e r e simply because Barbareno grammar requires
them (i.e., they can be thought of as 'dummy subjects').
Thus, t h e above examples suggest t h a t Barbareno
pronominal affixes are highly grananaticized. (Unlike
other c h a r a c t e r i s t i c s , Barbareno is similar to English
at least i n t h i s one respec t . ]
However highly graamiaticized t h e uses of t h e
pronominal a f f i x e s may be, it seems t h a t t h e i r main
funct ion st i l l is t o i n d i c a t e t h e involvement of some
p a r t i c i p a n t s i n t h e even t described by t h e verb. They
a r e always on t h e verb, and that is what is minimally
requ i red by the grammar. This suggests t h a t , i n terms
of t h e r e f e r e n t tracking system i n Barbareno, pronominal
affixes may be t h e d e f a u l t referential choice.
And i n fact when one looks at Barbareno d a t a with
this hypothesis in mind, one finds that the use of
pronominal affixes alone Is by far the most common w a y
of keeping track of referents in discourse. The
following examples are typical discourse segments in the
data. They are indicative of the high frequency of
reference by pronominal, affixes alone:
(6) JH59.407R-409L, txt 58, line 001 59.408L
h6?€ilfk niiiisiy-dutapin~ai , ho =?alikon maii =S-iv-ahtapin -wag DIS=In<iian.Orchard afte-3-PL-eat.supper-PST '<At Indian Orchard, after they ate supper,>
iroe=tSi siyRili?alas&l, ?i=&eEi S-iy-qili-?alaa&l ? I=always 3-PLHAB -pray <they used to pray,:
t ~ d i sifraldalwag %as*-saKutlhan, Eu ==BM&. S-iv-?alas&l-wag Uika=s-iv-saqt*m but=after 3-PL-pray -PST DPiso=3-PL-tell.a.story <but after they prayed> they told bed-time stories
kXffl kasiyw6n. k h ka =s-i~-weh and then*-PL-go.to.bed then they vd go to bed'
( 7 ) JH59.88R-89R, txt 9, l i n e 001 59.89L
m&- ii si$v&ahSgh he?li~in@: ?, mail S-iy-R+anBin -? he?=l =R+?inyu - - ? when 3-PL-R+eat.a.meal-? PRX=ARTçR+Indian-E ^When Indians axe eating a meal,>
i-siye-tip& - wil . ?i=s-iv-e-tipawil ?Id-PL-N-talk they don't talk.
? iy6 -he p s ~ h - ~ h a ~ h a l a l & - -nus ttn, ? i y e h 2-R+saqhalalm-us -man though 2-?%+holler -PAT-3P0 the you holler at them
they don't hear you.'
In ( 6 ) , the participant 'they8 is kept track of by
several instances o f the 3rd person plural subject
prefixes S-iv-. In ( 7 1 , the participant ' Indians is
kept track of again by the same prefixes as well as the
3rd person plural dixect object suffix m-n, and
similarly the generic 'you' is kept track of by both the
2nd person singular subject prefix g= and the 2nd person
singular direct object suffix -in. These examples thus
illustrate that much of the referential work in
Barbarefio is performed by pronominal affixes alone.
Now the amount of information which these affixes
encode is rather small ( o n l y person, number, and case).
For this reason, they seem to be used in situations in
which that much information is sufficient for clear
identification of the referents. That is, these affixes
are used to talk about participants whose referents have
already been fully established in the discourse ('given
information' in Chafe's term (1976, 1987, 1994)).
Example (7) above, which is taken fromthe beginning of
the text, illustrates this point. In (71, the
participant 'Indians' is established at the beginning by
being explicitly introduced by the pronominal affixes -
and a full HP. The participant 'you' is a generic
reference which does not require an explicit
introduction. (Example (6) is discussed below.)
Here is another example taken from the beginning of
a text,:
(8) JH59.205L-205R, txt 21, line 001 59.2051,
a) nasiyexpftS ihe? lt&&* ?i?alsa?tuh&y . na =s3-iy4qeE w=he?=l =&U? ?i=?al-sa?-tuhuy when=3-PL-sing DP*PRX=ART=bird.sp ?I=NM -FUT-rain 'When the huitacoches sing, it is going to rain.
I \\ ?iyalth& -&in hi?alsa?aktinlr 2: hiiuw6-&U, ly-?al-&akin hi=u-sa?-akti-nuna him1 =?uwthu
likely PL-MM -know DP=NM -EWT-come-bring DP=ART==food they know it is going to bring food,
e ) siyexp8t5 hinoh67 hina~tuhtubfl~ fi-iv-expeE hi=naho+-? hi-a =a-R-tuhuy 3-PL-sing DPçnuch+E DPawhenm3-R-rain They sing a lot when it is raining. '
Again bu2 F huit-acoches p is explicitly introduced in
line a) with the third person plural subject prefixes
iv- and a full W . After this introductionf it i8 kept
subject prefix &= ( ~ s c ~ ~ s e c i in 2.3.). Xn line d), the
participant 'rain, F explicitly introduced in line b) , is
r e fe r r ed to by t h e prefix Jal-.
Example ( 6 ) above, which is a l s o taken from t h e
beginning of t h e text, e x h i b i t s a s l i g h t l y d i f f e r e n t way
i n which par t ic ipants are establ ished i n discourse. In
t h i s example, t h e p a r t i c i p a n t 'they' is introduced by
t h e subjec t pronouns S-iv- together with t h e place name
' Indian Orchardr, and t h a t seems t o be s u f f i c i e n t t o
establish the referent i n t h e discourse: apparently
' theyr are ' the people a t Indian OrchardF. That is, i n
t h i s example, though the in t roduct ion of t h e pa r t i c ipan t
i n t h e discourse is not done as e x p l i c i t l y as i n ( 7 ) and
( 8 1 , t h e re fe ren t s t i l l s e e m s t o be c l e a r enough t h a t it
can be referred back t o by t h e subject pronouns i n t h e
rest of t he discourse.
Among different . a c t i v i t i e s deal ing with
pa r t i c ipan t s i n discourse, it seems t h a t maintaining
es tab l i shed re fe ren ts is no t a task which demands much
cogni t ive effort.. The verbs appearing with pronominal
a f f i x e s alone ind ica t e t h a t t h e same par t ic ipants are
s t i l l being ta lked about. W e s a w above t h a t pronominal
affixes a r e p a r t of what is minimally required by t h e
grammar. That is, what w e have here i s t h e following:
t h e smallest l i n g u i s t i c device, minimally required by
the graimnar, is reserved for a task requiring probably
the least amount of cognitive effort in handling
discourse participants. In other words, this pairing of
form and function in Barbareno is iconically motivated,
as has been found cross-linquistically (Clancy 1980;
Given 1983). Interestingly, we noted above that this
grammatically least marked referential choice is most
prevalent in Barbarefio discourse in terms of its
frequency. We will examine in the next section the
other extreme of this iconic pairing in which a
linguistically larger device than pronominal affixes is
used for tasks which seem to require more cognitive
effort than maintaining established participants.
Thus we have seen that pronominal affixes are part
of what is minimally required by the grammar. They are
used when the referents have already been fully
established in the discourse, and they are the most
common referential choice h Barbareh discourse.
4.2. Full noun phrases
In the last section, we saw some instances of
pronominal affixes co-occurring with a full noun phrase.
Specifically, we saw that full NPs are used along with
pronominal a f f i x e s when t h e pa r t i c ipan t s are introduced
i n t h e discourse f o r t h e f i r s t time. Unlike t h e task of
maintaining a l r eady established par t i c ipan t s i n t h e
discourse, t h e in t roduc t ion of pa r t i c ipan t s can be
argued t o be a more cogni t ively demanding t a sk ,
requi r ing a m o r e e x p l i c i t form than pronominal affixes.
Thus for this t a s k no t only t h e pronominal a f f i x e s , t h e
d e f a u l t marking requi red by t h e grammar, bu t also f u l l
Nfs are used. As f o r obliques, they are not marked an
t h e verb by a f f i x e s , so only full NPs are used t o
express them. I n this sect ion, we will examine
s i t u a t i o n s in which full NPs are used for d i f f e r e n t
tasks which seem to r equ i re more cogni t ive effort than
maintaining r e f e r e n t s i n the discourse.
As jus t noted, one obvious s i t u a t i o n i n which full
NPs are used i s t h e in t roduct ion of pa r t i c ipan t s into
t h e discourse . Examples ( 7 1 and ( 8 ) are two such cases.
Here is another one:
( 9 ) JH59.656R-657L, t x t 5 0 , line OOla 59.656R
?iaiyXilihlk hilati~wit~'!? B hiho? siyy&?, ?i=a-iy-qili-hik a s 1 S ?at- * W-5 ' =ho?=s-iv-va? ?l=3-PL-H&B -apply DPÈARTçpois DP*DIS=3-PL-arrow ethey used to apply poiaon to their arrows,>
siykfty iho?smol6 - &n h i ~ c o ~ k 8 ~ , S-iy-kuy 3-PL-take DP=DIS=3=saliva DPÈARTçfrog. <tailcing the saliva of a frog,>
A siyuÈ? -tap hilt$ofi$oÈn - ? hiUOhkiki - 7 , a-iy-uS?itap hi-l =R-&ovni--? his1 =~-Icxlb> -
- W ? 3-PL-ffiix DPçART=R-other- DP=ARTçR-thing- They mix it with other things'
This is the beginning of the text. As can be seen,
several di f ferent participants (underlined) are
introduced f o r the f i r s t time i n the discourse. Full
noun phrases (along w i t h pronominal prefixes f o r t h e
subject) are used f o r t h i s purpose. A f t e r the
introduction, 'the Tularefios', 'a frog', and ' the
s a l i v a F are kept track of by pranominal a f f i x e s S-iv-,
, and non-marking respect ively . (Remember that the
third person singular direct object i s unmarked.;
I t should be noted tha t the introduction of
participants by one explicit mention does not always
establ ish them i n the discourse. There are cases i n
which participants axe repeatedly mentioned by f u l l NPs
at the introductions
(10) JH59.583L-584L, txt 49, line 001 59.5831.
a) mal&*&e ki& anan&-&, '?i& s i i a i k - t i ' l t6, mal* MU. snanahu ?ikhu ai iaAe5i t o any what kind but mostly mussel 'any kind of shell, but mostly mussel shell
b) ftaykg he?gtu?l-w&f hilt6 ftayk6 )Ã e?=(tu?iwa hi=l =&a because PRX=shell DP=ARTaaaussel because ethe shell of a mussel,>
They put. "the mussel-shell on the hot coals, \ i
e ) hikaaa- ii s?aluipaw& -tun, hi+ka=maii S-?aluSpawatun DP+so=when 3-bum.to.ashes and when it is reduced to ashes
hihulSBw, hi=ho =l =Sow DP=DIs=ART-wild. tobacco
they take it and mix it with the Sow8
This is at the beginning of t h e text. The discourse
participant 'mussel s h e l l 0 is explicitly mentioned t w i c e
with a full ftP along with a pronominal affix, and after
that it is kept. track of only by the pronominal a f f i x e s .
Notice i n l i n e a), at the very beginning of t h i s
sequence, 8mussel' itself is introduced with a f u l l NP
f irst . It shows that sometimes it takes more than a
single explicit mention t o es tabl i sh participants i n the
discourse.
Here i s another example:
(11) JH59.334L-344R, txt 137, line 001-1 59.334R
'The old-time Indians had only f i r e l i g h t in the house. <They would make a. f i r e in. the middle of the house. So that as the flame burned, the whole inside of the house was lit up. The old-time Indiana had only f ire l ight in the house.
When the Spaniards came they brought t a l l o w candles, from Mexico, - v\
b) 6 he?lvelweiS-? ?i-At6 -wits higiyapl- ?ikhu he?=l =R+wela -? ?i=s-SutowiE hi.==s-iy-apiqen but ?RX=ARTÈR+candle-E ?I=3-be.quick DP=3-P&-bum
but candles burned down q&Ac3.y and w e r e expensive.'
Again, the participant 'tallow candlesr is expressed
t w i c e w i t h a f u l l HP. N o t i c e here that it is expressed
for the first time a s 'tallow candlese and then
'candlese suggesting that the second mention may not
have t o be as explicit as the first mention (Given
1983). These examples show that sometimes f u l l NPs are
repeated i n order t o es tabl i sh the participants fully i n
the discourse. We have thus examined participant
introduction in discourse, for which full NPs seem to be
the primary device. As we saw in the last section, once
participants are fully established in the discourse,
they are kept track of by pronominal affixes.
Interestingly, however, even after participants are
fully established, they are sometimes still expressed by
f u l l NPs. Several factors seem t o be responsible for
this. One obvious factor i s the presence of other
semantically compatible participants in the discourse
context (Clancy 1980; Giv6n 1983; Fox 1987). Examine
the following example:
(12) Title: ?anafianfcAy
JH59.9L-15L0153L, txt 167, line 001 59.9R
*<It is called> ?aiia&unfi&y, <but> in Spanish la llorona. The Indians believed much in him. ?ana&a~fi&y <cries out like a recently born baby>. <When it cries out near a house>, <it means that some person is going to dies.
rt ?ik% he ?al i iat&o hikl- lkL ? m u &e ?*l-i8a -?&xi hiam- but EH UK -seat-different DPçthin but somewhat different.
A v\ suxwalkA tu his?eKw6- lefi. s-uxwal 4 a t u himas-? -eqwel-Vfi 3-resemble-cat DP=3-NM-make -RES It looks l ike a cat.
The above segment is taken from a text entitled
?anabamb&w, a mythical figure. As can be seenr
is discussed as the main topic in the segment
leading up to line d) . In line d}, Jana&mb&~ is
expressed with a full NI? probably because there are t w o
p e i c i p m t s being contrasted in the utterance. Unless
they are expressed explicitlyr it would be difficult to
ident i fy the referents correctly.
Here is another example:
*They say that the t e c o l o t e t a l k s just like tha XnciiansD Take notice when you hear an -1 aound! H e says it very plain: d4 good person will die.% They used ta be afzaid of ?Ae tecolote- They say that when he sounds near a house, it is a bad amen.
<The coyote3 comes right to the house
e l ?i-siw-?S -w8n w6? we? we? w6? P ?i=s-Riiwon w6? we? we? we? ?1=3-R+sound woh w a h w o h woh and he starts woh woh w o h wah.'
This example comes f r o m the beginning of a tex t . As can
be seen, several animals aire introduced in this segment:
the tecolote (a type o f owl) , the coyote, and the fox.
The coyote is explicitly mentioned i n line d ) probably
because otherwise it would be difficult to h o w which
animal is being referred t o . In particular, the coyote
has just been introduced in line b), and as we s a w
above, that may be part of the reason why it is repeated
with a full B?P in line d ) . Eowevez, more importantly,
in line b) the fox is also introduced fox the first time
in the discourse, and that 3eems t o be the pxbary
reason for the use of a full HP to refer to the coyote
i n l ine d ) . If a f u l l NP w e z e not used there, it would
be very d i f f i c u l t t o identify the referent correctly.
Thus we have seen that the presence of semmtically
compatible p w i c i p a n t s in the discourse cantext
motivates the use of full noun phrases.
WO related factors which seem to be respons31e
far the use of full NPs after the participants have
become ful ly established are thematic break and
perspective change (Clancy 1980; Fox L987; Tondin 1987).
[14) 3HS9.287L-280R, t a c t 34 , line OOla 59.287L
'They were a l l numb with fear when that 1ewLew came in. They say it is samething terrible, it is just like a little child but it stxikes yau numb. In spite c i f the lewlew being there, Jade xeached evex and tcmk a &ink of pispibata and L t gave hin strength.
Jabe was the bravest of all the Indians
He was ~cultured.> He knew haw ta read Spanish and a little Latin. '
This excerpt is also from the beginning of the text. It
appeare to be divided into t w o parts . The first is a
description of a spec i f i c event$ what Jahe did when a
mythical f igure hwlew came i n . The second, maxked by
the azxow, is a description of Jabees assets
Interestingly, Jahe is expressed w i t h a full NP when
the second part begins even though the segment just
before that describes a series of actions i n which Jabe
is a principal participant ( p Jahe reached over and took
a drink af pispibata and it gave him s t r e n g t h f ) . This
e f f ec t of thematic breaks on referent ia l choice has been
observed c x o s s - l i n ~ i e t i c a l l y . It is almost as i f a f t er
the thematic break the p m i c i p m t s that have been
established i n the discourse are cleared from the
memories of t h e speakers, so they must be re-introduced
into t h e discouxse with e x p l i c i t form.
Here is another example. It is taken from the
beginning of (131, reproduced here as (15):
(15) JH59*289L-292R, txt 159, line 001-112 59-289L
eThey say that the tecolote talks just like the Indians.
nafio? ahaxtawti - gin noha-? S-saxtawasin vexyeEl4 3-say-clearly he says it very plain:
g) @&P namutey bihaiikp hi@,w-?i*won, \F
swam -Tip na =mutey -? =l =?&p hi=s-R+iwen 3-IDF-say when=be.neax-2 D ~ D I S = ~ h o u s e DP=3-R+sound They say that when he sounds near a house,
l.. h ) ?i-?a=l&LyU&-
?i=?al-qilalyiqs ?I =a -bad.omen it is a bad omen- p
Again, @teco1oteR is a type of owl which is introduced
in line a) fo r the fhst t h e in the discourse. The
segment leading to l i n e f) is about t h e Chumash
perception of the tecolote t a l k i n g ju s t l i k e themselves.
Line f) begins a new sequence which describes haw t h e
Chumash were a f r a i d af t h e tecolote. So t he r e seem to
be a minor thematic break after l i n e e), and in l i n e f)
the tecolote is expressed as a f u l l W even though it
has been ta lked about i n t h e pzeceding context . That
is, the use of a f u l l W here coincides wi th a thematic
break. W e have thus seen that a thematic bzeak is
another factor which motivates the use of full noun
phrases for t h e p m i c i p m t s which have a l r eady been
es tab l i shed i n t h e discouse.
Another po in t which should be made regarding (15)
i s t h e use of a full HP for t h e t e c o l o t e i n l i n e c) .
One can s a y t h a t a f u l l HP is repeated t h e r e because t h e
t e c o l o t e was j u s t introduced (see t h e discuss ion above
regarding the r e p e t i t i o n of f u l l NPs for newly
introduced pa r t i c ipan t s ) . However, another p o s s i b i l i t y
is t h a t t h i s is due t o a change i n perspective. That
is , t h i s is quoted material (see t h e exclamation mark
provided by Harrington), which seems t o present the
perspect ive of t h e Chumash, who would say what is i n t h e
quote. The quoted material does not d i r e c t l y belong t o
t h e s t o r y l i n e but t o t h e world of t h e quote, and that
may be why the t eco lo te is expressed as a f u l l HP. In
other words, I am suggest ing t h a t t h e r e is a c e r t a i n
kind of break i n perspective f r o m t h e s t o r y l i n e t o the
quoted material, which is responsible f o r t h e use of a
full HP t o r e f e r to t h e t e c o l o t e i n line c ) , j u s t as t h e
use of full MPS can be motivated by a break i n
H e r e is another example:
(16) JH59.36R-40&, txt 175, line 008 59.37R
'<When the Spaniards first came here, they gave a rooster to an old man who lived on the island across. And he took the rooster,> he said; <I thank you much>. And he took it to the island heading for the island
can be seen, t h e rooster is intxuduced at the
beginning of the d i s c o u s e , and it is re-introduced with
a full ?iIP h line a) after the thematic beak ( i . e . , the
next morning after the old m bzm~ght the raoater back
to the i s l a n d ) . A f t e r that, the rooster is kept track
of with pronominal affixes and seems to be fu l ly
established in the sequence preceding line f). Notice,
howevez, tha t i n line f) the rooster is referred to with
a fuI.1 NP tb i rdro again, what is represented in l i n e f]
is quoted material, which seems t o be responsible for
the use of a f u l l there. This is similar t o the use
of full NP3 motivated by thematic break: in t h e former,
there is a break created by a change i n perspective fram
the s to ry line to the world of the quote and i n t h e
latter the re is a break in t h m a t i c i t y * Thus we have
seen that the use of f u l l NPs far e s t d l i s h e d
p a e i c i p m t s i n the discourse is partly motivated by a
break in themt ic i ty ar a change in perspectivee
Finally, another factor which has of ten been
discussed in the literature regarding t h e use of'
explicit form for e s t a l i s h e d p m i c i p a n t s is the
absence of mention or mention only by i n e x p l i c i t ~ O ~ I U S
for a certain duration of discourse (Clancy 1980; Given
1983 } . That i s , p a ~ i c i p a t s are often e x p l i c i t l y re-
introduced after not being mentioned a r a f t e x being
mentioned only by such i n e x p l i c i t fanus as pranouns fax
a s t r e t c h of d i ~ c a u r s e o It has been suggested t h a t t h i s
happens because t he refezents fade fxom the memoxies of
t h e speakers. However, 1 did not f ind t h i s factar
playing a zole in l3arbarefio discoursee This particular
r e su l t could be because mast of the t ex t s i n my data are
r e l a t i v e l y 8hort; they axe not appzopxiate far examining
th i s factor. But it is a l s o pas s ib l e that t h i e factor
is only an epiphenomenon of such factors as thematic
break and perspective change (Fox 1987 ; Tamlin 1987 ) . That isf it is oftem suggested that the longer a stretch
extends without e x p l i c i t mention of the p a ~ i c i p a t s , the snore likely it w i l l be that the p-icipants w i l l be
referred t o with more e x p l i c i t foms* However, this
could be simply because of the fact that the longer the
stretch, the more likely it i s that there are breaks in
thematicity and changes in perspective in the stretch.
This suggests that the real factor in the use ~f
e x p l i c i t fcmw i n such cases may not be the absence of
explicit mentions but breaks i n thematicity and changes
in perspective* And i n fac t , as X have discussed above,
I found in my data t h a t these l a t t e r two factors play a
significant r o l e i n the se lect ion a f full naun phrases
in Bazbarefio.
Thus we have seen that full naun phrases are used
for such puqases a s to introduce particzipants in the
discourse, to dis t inguish some p a a i c i p a n t s from other
pa&icipmts clearly, and to re-intraduce p e i c i p a n t s
after thematic breaks or when there are changes i n
perspective In the last section, we saw that
pronominal affixesf which axe the smallest referential
devices in BarbareEof are used far probably the least
cognitively demanding task of maintaining established
pa-icipats in the discouxse. In the present section,
w e have seen that f u l l NPs, which are lin~istically
much larger? are used for a variety of possibly much
more cognitively d m ~ ~ n g tasks.
In the past two sections, we saw two referential
forms in Barbarefio: pronominal affixes, which are
gr-atically required on verbs? and full NE%, which are
used along with pronominal affixes fox core ~ ~ e n t s .
Even a glance at Baxbareiio discourse data shows that
these two forms axe responsible for most referential
work in Barbarefio: they axe the t w a most commonly used
referential forms*
4.3. Independent pronouns
In the preceding sections, I have suggested that
most referential wark in Barbareiio is performed by full
NPs and pronominal affixes; however occasional~y one
comes across other referential forms in the data* These
are the independent pranauns and the proximate
daunstrative bet?. Since verbs always take pronominal
affixe~ for the core arguments, if independent pronauns
or the d-anstrative be2 axe used for the core
arguments, they appear along with the pxonominal
affixes . Each of these two less commonly used
referential forms seems to be associated with particular
discaurse function8 and pmicular syntactic
c ~ n ~ t m c t i ~ n s which aze used to expzess certain semantic
content* In the present and the following sections, X
will illustxate the uses of these forms*
Among these two referential f o m , 1 will focus un
the independent pzonouns in this section. The following
are the independent pronouns in Bzubarefio Chumash
(Beelex 197Ua, 1976) :
Table 12 Barbare50 independent pronauns
singular dual plural
1st pexson no2 k i 6 E m U S k i kAyM
2nd person pi2 pig& pisk i p i y u
Two categories of person and three of number are
dietinwished in these pxonouns. There do not seem to
be any grammatical restrictions regarding the use of
these pronouns; a8 the examples below suggeet, it
appears that they can be used fox any grammatical
re la t ions (i.e., subjec t , direct object , obliques). The
readers should keep i n mind the fact that these pronouns
are not like pronouns i n English because they occur very
infrequently f o r r a t h e r spec ia l ized func t ions and i n a
limited set of construct ions , which we w i l l examine
below. I have found only about a dozen cases of
independent pronouns i n the data. In terms of discourse
functions, pronominal a f f i x e s are muchmore l i k e English
pronouns. Further, it should be pointed out again t h a t
independent pronouns appear along with pronominal
af f ixes for subjects and d i r e c t objects . This f a c t
nicely cor re l a t e s wi th the functions of independent
pronouns, which involve more than merely keeping t r a c k
of es tabl ished pa r t i c ipan t s i n discourse.
There are three d i f f e r e n t types of s i t u a t i o n s i n
which these independent pronouns occur. F i r s t of al l ,
they are used when a par t i c ipan t is cont ras ted with
other par t ic ipants i n the discourse. Examine t h e
following example:
( 1 7 ) JH59.302L-308Rr txt 161, line 001 59.302L v\
napkti - ti hiin-& - 6k1 hifahai h i m , na =p-hti bi=s-R+R+nah hi=a-&a6 his1 =ku when=2-çe DP4-R+R+go D-3-spirit DP=ARTçoerso *When you see the s p i r i t of a person walking about,
?ikag?ig%XY hibalia? aKi&n hipl? , ?i-ka=s-?ishfiy hiq-?al-sa?-aqSa hi* ?Imiai=3-b.a.siqn.that D-2-MM -m-die DP=2 it either is a sign that you are going to die,
keh&*& h o ? h hulicagah&i kalsa? a K S h kehalbi ho?=i =h hu=l =ka-s-ahaS ka=l =sa?-aq6an or DIS=ARTsperson Rlt=ARTçKA=3-~piri KAzARTçFUT-di or that the person, whose s p i r i t it is, is going t o die.'
As can be seen, 'youF, the person who sees the spirit,
and the person of the s p i r i t are clearly contrasted.
Here is another example:
(18 ) JH59.22L-24L, txt 170, l ine 001 59.22R
klgkS? inakiyhti - wun he?lkopkop& : f) kivM? ?i-a =k-iy-kutiy-wun he?=l =R+kopk6p IPL ?I çwhen=l-PL-se -3P0 PRX=AKTÈR+toa
'To us ail frogs look alike,
And to us all lizards look a l i k e . But to each other they must look different. '
In t h i s example, we, humans, who see frogs and lizards
in one way, and these animals, who see each other in
another way, are contrasted.
Examine the following example:
( 1 9 ) JH59.293I.-298R, txt 160, line 001 59.2931.
'The Indians say that there is everything in the ocean that there i s on the land. <For instance, the baracuda is the snake of the ocean, but the gopher snake is the snake of the land. Just like the baracuda i s the gopher snake of the ocean, "the gopher snake is "the baracuda of "the land.> The sardine i s the lizard of the ocean. O u t the l izard is the sardine o f the land, and the crayfish is the matavenado (Jerusalem cricket?) of the ocean, and the matavenado is the crayfish of t h i s land.>
a) ?ikaBi9d: ? hi?f*tilSap ?i=ka=ç-is-R-k -? hi=?iti=l -<up
IPX. ?I*KA=3-AP-R-person-= DP=here=ARTçlan We are the people of the land.
b) When we get tick we wash ourselves out w i t h the water of the ocean. For three days we drink no water except the water of the ocean. We drink the ocean t o vomit a t both ends.
The <swordfish> are the people of the sea.'
A t the beginning, the Indians are said t o claim that
there i s everything i n the ocean that there is on the
land. After several examples which i l l u s t r a t e this
claim, people on the land are contrasted with swordfish
i n the sea as equivalent: line a) and l i n e c).
Independent pronouns are sometimes found in
situations in which contrast is only weakly implied.
Examine the following example:
'The -three who were befriended. Hawk and Raven and Coyote, happened t o be walking past the house of skunk, and Coyote just happened to say: *Come, let's go in and see the old man, the dancerIn Coyote said: "You two go infw Coyote wanted Skunk to kill them. And so they (all) went in. "Sit down, my friendsf- And Coyote said to Skunks "Please dance a Little, SO that these fellows are amused." And Skunk said: -Alright, I will. I'm already getting a little old, but at least 1' l1 try to dance. But it ' 8 very cold; I l1 first make a flee. Then Irll try t o dance.- Then he straightened his tail. Then he raised h i s tail, and he started to dance. He kept whirling around and around and making his anus get closer and closer t o the faces of his friends. Skunk cried out, saying: "Get a little closer1*, because he wanted to
squirt , his poison on them. And then Raven t h r e w a hot rock SO that it quickly entered h i s anus. Skunk was running a l l over from suffering so much. And he cried out: "This fellow is a bad person; he already has killed many pe~ple.~ And so Skunk died. And Hawk told Coyote:
hoQi h i iti h&% hi ill coae.here DP this.one DP 2 too 'You come here toolu
Â¥Wh should I asked Coyote. And he was already afraid. "Get a move on, hurry and corner But Coyote didn't obey him. And so Hawk finally caught Coyote and threw him into the fire. And there he burned up. Coyote was a bad holllbre . '
Unlike t h e earlier examples, t w o i tems are not over t ly
contrasted i n t h i s example. The independent pronoun pi2
occurs i n a context i n which another p a r t i c i p a n t has
been discussed. A f t e r Skunk, who almost k i l l e d Hawk and
Raven, died, it is malicious Coyote's t u r n t o die: 'you
come here t o o p . In t h i s particular case t h e role of t h e
re fe ren t of t h e pronoun 'you' as t h e primary a c t o r of
the event described by t h e verb @comer can be seen as
weakly contras ted with Skunk, who has just been k i l l ed .
The following example i l lustrates t h e second type
of s i t u a t i o n i n which independent pronouns are used:
(21) JE59.116R-117R, tx t 16, line 001 59.116R
'Once upon a time Luc. brought a magnet hone and picked up a whole chain of needles. She tried to explain to Luiaa, who was horrified, she wd not pay attention and said,
e f i p balyawliitfi hlpt? . h=k-?ip p-?a&yawluE hi- ==l-think 2-beedevilish? DP=2 <I think you are devilish(?)>. '
In this example, there does not seem to be any element
which is contrasted with the referent of the independent
pronoun pi? 'you8. Instead pi2 here seems to be used to
call special attention to the referent, who is
characterized as having the property 'devilish'. This
use will be called 'emphaticr in the following
discussion.
It is of interest that in all of the above cases
the referents referred to by independent pronouns are
participants which are probably not totally new in the
discourse. They may have been already mentioned in the
prior discourse, they may be present at the scene in the
story, or they may refer to the Chumash. In 4 .1 . , I
suggested that pronominal affixes are used to keep track
of fully established participants in the discourse.
Independent pronouns can be said to be more explicit
than pronominal affixes not only because they are more
independent, but also because they appear along with
pronominal affixes . As we have seen above, they perform such functions as contrast and emphatic, which seem to
be more cognitively demanding than merely keeping track
of established referents. In 4.2., I suggested that
full noun phrases are used to introduce participants for
the first t h e in the discourse and to re-introduce them
after thematic breaks, perspective changes, and when
there axe other semantically sixnilax p m i c i p m t a in the
context.. Independent pronouns are less explicit than
full noun phrases because they are generally shorter and
less informative, and the tasks they perform do not seem
to be as cognitively demanding as introducing
participants for the first time in t h e discourse and re-
introducing them for various reasons. That is, this set
of referential forms lies in the middle in terms of
explicitness, performing tasks that also seem to be in
the middle in terms of cognitive demand.
The first two uses of Barbareno independent
pronouns (i.e., contrast and emphatic) are characterized
by discourse functions. The last one seems to be more
semantic/syntactic. There appear to be certain semantic
content expressed by particular syntactic constructions
which require the use of independent pronouns. Examine
the following example:
(22) JH59.47R-57R, txt 178, line 034 59.52L.52R
&aykg ?alw6t W&-y% ki-fenat& in&? kayk6 pal-wot va9i k d e 6 ki=!n~? because MM -leader too like DP=1 '<For he is also a leader like me. 3'
The use of the independent pronoun in this example seems
to be required by the syn tac t i c const ruct ion which
expresses the semantic content ' l i k e me' ; t h e r e is no
other way of expressing it. This i s c l e a r l y not a case
of emphatic, and there does not seem t o be any clear
sense of con t r a s t which is observable in t h e context.
H e r e is another example:
(23 ) JH59.124R-l26R, txt 194, line 001 59.124R
Â¥I is the big green flies that <give notice about something;-. They have a sound of buzzing when they fly. When they get near you, it is a sign that people ace g. t o come to visit you. When they buzz around at night,
it is a sign t h a t one of ye is g. to dieor
Again, the use of the independent pronoun here appears
t o be required by t h e syn tac t ic const ruct ion which
expresses the semantic content 'one of you*. T h i s
u t t e r ance seems t o be associa ted with some sense of
emphasis ( ' i t ' s one of you who is going t o d i e * ) , which
I think comes not from the use of t h e independent
pronoun itself but f r o m t h e expression 'one of you8.
Thus I have shown t h a t the use of one of the less
frequent r e f e r e n t i a l forms is associated wi th p a r t i c u l a r
d iscourse functions and with p a r t i c u l a r syn tac t i c
constructions used to express certain semantic content.
4.4. Dmonstrative be2
The o the r less frequently used r e f e r e n t i a l form i n
Baxbarefio is t he proximate dmonst ra t ive be2. 1 have
found only about a dozen cases of t h i s use of be? in the
data. Just like independent pronouns, this form is
associated with specific discourse functions and
specific syntactic constructions usedto express certain
semantic content. Below, we w i l l examine examples which
illustrate these functions and constructions.
First of all, the dmonst ra t ive he2 is used to
refer t o such types of information as facts, rumors,
events, and ideas which have already been introduced i n
t h e discourse using several clauses. It seems ra the r
d i f f i c u l t to r e f e r back to them i n o ther ways because
t h e full reference t o t h e m would require mentioning
whole ideas a l l over and there are no lexical items
which express these types of information. Examine t h e
following example:
(24) JH59.115L-l16L, txt 190, line 001 59.115L
'There was one Losenzma who used to turn into a at.-lion
b) iBS.-ii~~a?~~-i&n ilkt-it hi^naii=~-@a?-xonon-? hi=l =kiJli DPs=vhen=3-FUT-8teal-TIt DPsART'asometfing when he wd. go out to steal.
\\ A c) hos?unl-*i ?ialth isaaitikfl -yus ih6?.
ho =B-?uniiri ?i=s-itaq hies-am -tikuyus hi=be? DIS=3-wife ?1=3-hear DP=3-IDF-talk-about DP=PRX His wife heard of this. '
As can be seen in line c}, the dmonskrative refers
to t h e rumor that the wife's husband becomes a lion when
he goes out to steal.
Here is another example:
(25) JB59.116R-117R, t a c t 16, line 001 59.116R \v a) saxip&-kg hisnukfi-~u.
0 hi-gnet hi-lokl6 - aya,
saxipaka hi=s-nukuni hi=l -iaagnet hi=lokleaiya once D-3-bring.hoaie DP=ARTçaiagne DP=Lucrecia *Once upon a time Lac. brought a magnet hone
b ) hikap~pa - fcl - fnAn h i h h ~ ? X his?aw-?awux?a-? hika=s-uSpak -7 -in -win hi=l =?&h&? hims-R-?awuxa--? and =3-pick.up-TR-IKS-3PO DPaART=roany DP=3-R-needle-EM
and picked up a whole chain of needles.
c) k6-f& kanaç?lps hilgwi - ga kibi kaçn -a-?ip-UB hi=luvisa now =-hen-3-say-DAT DPçLuis
hihe? alas6 - 1: hihft?, hidea=?al-?açel himbe? DPÈEMsN -be.horrified DPÇPR
She "tried to explain to Luisa. who was horrified, '
to the full event of Lucrecia picking up a whole chain
of needles w i t h a magnet. This was a horrifying event
to Lucrecia's mother, Luisa, because she apparently had
not seen a magnet before.
Here is another example:
( 2 6 ) JH59.60R-6lR, txt 1, line 001 59.6IL \v
a) na nono pyint i i ?i ?aleti% na n d o p-yin6i ?i ?a%-e-Sho when very 2-beehot ?I MM -H-good
'When you are heated you don't want to drink too much cold water.
b 1 ? iyaltb d n w a i ib6? iholkuhkd? ih6?. iy-? a~-&abh-wa~ h i m h a ? hi=ho?=l = R - h + - ? hi==he? PL-NM -know -PST DP=PRX DPÇDISmABTÈR-person+ DPÇPR The Indians knew that. '
In t h i s example, the dmonstrative be2 sefexs t o t h e
idea that when your body temperature is high it is not
good to drink a lot of cold w a t e r .
Notice that, in a l l of these cases, the information
referred to by the proximate demonstrative has just
been intxaduced into t h e discouxse, and he? is used to
refer t o it f o r the second time. Barbare60 has s eve ra l
r e f e r e n t i a l forms with which t h i s kind of r e f e r e n t i a l
t a sk could be performed. However, none of them s e e m s
appropriate here. F i r s t , though the information has
j u s t been introduced, r e f e r r i n g t o it jus t by a
pronominal a f f i x (o r by zero i f it is a t h i r d person
s ingula r d i r e c t o b j e c t ) seems too radical; it would be
d i f f i c u l t f o r t h e hearer to i d e n t i f y t h e r e fe ren t s i n c e ,
unlike t y p i c a l cases of r e f e r r ing , t h e information
represents something which needs several clauses t o
describe it. Second, there is probably no l e x i c a l i tem
which represents t h e kind of infoxmation which he2
refers to i n t h e above examples. That is, I am
suggesting t h a t s i n c e there are no words f o r t hese
ideas, t h e demonstrative be2 is used ins tead to t a l k
about then. F ina l ly , one could a l s o r e f e r t o t h e
information by repeating the whole t h i n g again, b u t t h a t
seems very uneconomical e spec i a l ly because t h a t is
exact ly how t h e information is presented i n t h e i r f i r s t
introduction.
I n addi t ion t o t h i s f i r s t func t ion of f u l l ideas ,
t h e proximate dmons t ra t ive a l s o has a eecond,
related function. It is used t o refer to participants
and concepts f o r which the use of a fu l l NP does not
seem t o be appropriate, or for which there does not seem
to be an appropriate lexical item. Examine the
following example:
( 27 ) JH59.98L-l00L, txt 186, line 001 59.98L
*The h o r s e s would be in the morning with the hair of the <inane> braided. Florentino said: The horses <go> i n the bsuah in the night, ao their manes are t a n g l e d . But Guisa> said: The D e v i l rides the horses at night,
t B u k a he? ?ikas?aly6nt',Ã hilly& -wl'u. Euka be? ?i=ka=s-?alyenta h i = l =Lyawlu thus PRX ?I=K&=3-reins DP=ARTsdevil those are the <reins of "the Devil>.'
The dmonstrative be2 apparently refers to tangled
manes, which have been introduced i n the discourse only
periphrastically using verbs 'the hair of the mane
braided' and 'their manes are tangledF. Referring to
t h i s referent by just a pronominal a f f ix seems to be t o o
radical , because it has not been introduced as a single
concept using a f u l l HP. However, t h i s concept is not
t o t a l l y new either because as we just noted it has
actually been talked about using verbs, which makes the
use of a f u l l NP unnecessary. These reasons seen to
motivate the use o f the dmonstrative be2 in this
example.
Another point which should be made regarding ( 2 7 )
is t h a t what is intended by the speaker could be
something along t h e lines of 'whatever it is which is
t h e tangled s t a t e of manesc. There is probably no
lexical i t em which represen ts this meaning exactly, and
that could be another reason motivating t h e use of he?.
In other words, I am suggest ing t h a t t h e lack of an
appropr ia te l e x i c a l item nay be responsible for t h e use
of he?. In fact, he2 is often used for referents whose
i d e n t i t i e s are not e n t i r e l y straightforward. The
f o l h w i n q example i@ one such use of he?:
( 2 8 ) JH59.172L-173R, txt 205, line 001 59.172L
'The razor-clan lies in "the sad of the Goleta Slough. To gather t h e m , you use a stick which has been nade with a knob a t its end. You find the hole, and by the bole know where the razor-clam is down in the mud. And you stick the stick down i n the amd. T h e razor-clam i s evident ly agape, and closes on the knob.
4 &y~i& ? d i p : ? a l s i t ~ ~ & *^cl hih6? P feiyafoi ?al-?ip ?al-S-iehaxi hi+he? perhaps maybe KM -think MM -3-enemy DP+PRX Perhaps he thinks it is an enemy.'
The d m o n s t r a t i v e he? in t h i s example may appear t o
r e f e r to the s t i ck . However, I would like to suggest
that the use of t h i s p a r t i c u l a r form is motivated by t h e
fact t h a t to t h e clam it is something ne i ther whose
i d e n t i t y nor whose purpose is very clear. That is, t o
the clam, it is not a stick but something i n front of
him which he th inks is a n enemy. There probably is no
s p e c i f i c lexical i tem which presen t s exac t ly this
meaning. The d m o n s t r a t i v e he2 is apparently used f o r
t h i s kind of r e f e r e n t i a l task.
Here is another example:
(29) OT59.227R-228L, txt 231, line 001 59.227R
' B u t to be lucky in love they used ?ayip, they sprinkled ?ay& in the coffee or on the -taw>
tau kanu mhfiiaw hihft ? Su kanu l =?uw hi=-e? for so-that from=ART-eat DPÇPR
so that <eating it> one would f a l l in love.
The d a o n s t r a t i v e he? seam t o r e f e r to whatever l i v i d
food ?ayip, a love potion, is added to; t h e verb Zuw is
normally used to refer to t h e a c t i o n of eating food. It
seem that t h e r e f e r e n t referred t o by he2 is thus
introduced i n a way i n which its exact identity does no t
really matter . To refer to t h i s r e f e r e n t by r epea t ing
t h e whole t h i n g as ' ea t ing t h e coffee o r the s t e w t would
make t h e identity of t h e r e f e r e n t precise , which is
probably not what t h e speaker in tends here. Further,
for the same reason, probably t h i s r e f e r en t has no t been
es tab l i shed i n t h e d iscourse to t h e ex ten t t h a t it could
be r e f e r r e d t o by a pronominal affix. These reasons
@ e m to mativate the use of t he d-onstrative be2 i n
t h i s exanple.
Thus we have seen that t h e proximate demonstrative
he? is used t o refer t o participants or concepts when
e i t h e r t h e use of a full NP is not appropr ia te o r t h e r e
is no appropr ia te lexical item to refer to them. This
type of be2 is vezy similar t o the first type of he2,
which is used t o r e f e r t o such information as f a c t s ,
rumors, events , and ideas which have j u s t been
introduced i n t h e discourse using s e v e r a l c lauses . I n
both of t h e s e types, t h e use of pronominal a f f i x e s does
not seem t o be poss ib le because t h e r e f e r e n t s have not
fully been e s t ab l i shed i n t h e discourse, and f u l l NPs
are not used because t he re is no appropr ia te lexical
item. I n o t h e r words, these two types of he? seen t o
occur i n s i t u a t i o n s i n which t h e use of more common
forms is not a v iab l e option.
It is of interest t h a t t h e d m o n s t r a t i v e he2 is i n
t h e middle i n terms of t h e exp l i c i t ne s s of t h e form,
compared wi th pronominal affixes and f u l l noun phrases.
That i s , t h e demonstrative bez is genera l ly more
independent than pronominal a f f i x e s bu t shorter and less
informative t h a n f u l l NPs. And it is found t o be used
f o r a t a s k which also seems t o be i n t h e middle i n
cognitive demand: what it does i s to talk about the -
referents which have just been introduced in the
discourse but which are for various reasons difficult to
talk about using more common forms. It seems that this
task is cognitively less demanding than introducing
participants for the first time in the discourse and re-
introducing them after thematic break, perspective
change, and when there are other semantically similar
participants, all of which are performed by full NPs.
On the other hand, it is cognitively more demanding than
merely keeping track of fully established participants,
which is performed by pronominal affixes.
The first two uses of the proximate demonstrative
be2 are characterized in terms of discourse functions.
The last one seems to be more semantic/syntactic. There
appear to be certain types of semantic content expressed
by particular syntactic constructions which require the
use of he?. Examine the following examples:
(30) JH59.260L-262L, txt 239, l i n e 001 59.260L
'There are old-men who indulge this way. Constantly they talk a lot of lies.
napesK6 n\is k& h i l M - n % t ~ hlh67: na =p-esqen-us ku h i s 1 =k&n85 h i + u wben=2-ask -DAT person DPçARTÈlil DP+PRX When you ask such a person:'
Ill
(31) JH59.139&-151R, txt 198, line 010 59.140R
'When one '8 spirit leaves oneF s body and wanders about, it is a sign that the person i s going to die.
napantl? hilnahn~-h'an h i i a b a . ~ ~ . ? i ki*&eti hiha?, na --anti? M.4. =R-R+~& h i 4 ==?aha&i?S k d e E h i + m when=2-meet DPÈARTÇR-R+ DP=ART=ghoat like DP+PRX If one meets such a. spirit walking about,'
There do not seem t o be o ther ways of expressing such
ideas as 'a person l i k e t h i s s and 'a s p i r i t like thisr.
This i s similar t o the use of independent pronouns i n
t h a t the use of t h e proximate d a o n s t r a t i v e he2 is
associated not only with p a r t i c u l a r discourse functions
but also with p a r t i c u l a r syntactic construct ions used to
express c e r t a i n semantic content. Notice, however, t h a t
in ( 3 0 ) and (31) even i f it w e r e grammatically poss ib le
t o use o the r more frequent r e f e r e n t i a l forms, it would
still be d i f f i c u l t to do so on o the r grounds. I n both
of t h e s e examples, the re fe ren t of the d a o n s t r a t i v e he2
is not given a name: 'old men who constant ly l ie ' i n
( 3 0 ) and 'a s p i r i t t h a t l e f t t h e body of a person and is
wandering about i n (31) . That is, these r e f e r e n t s seem
t o be t o o complex t o be r e fe r r ed t o by pronominal
affixes, and t h e r e are probably no appropr ia te lexical
items which r ep resen t these meanings. Thus examples
such as t hese suggest that sometimes both the funct ion
and t h e semantics/syntax are responsible for the use of
be?.
In this section, we have examined one of the t w o
less frequently used referential forms: the proximate
d-onstrative be?. S have shown that the dmonstxative
he2 is used far re ferents whose ~efexential mnipulation
seems to be difficult because they are referentially
complex or their exact identities are Left unspecified.
These referents have just been introduced in the
preceding discourse. However, because of their
referential characteristics, they still have not fully
been established in the discourse, so it is not poss ib le
t o use pronominal affixes t o refer to them. Similarly,
it is not possible t o use full Nfs either because again
due to the referential characteristics of these
referents there are probably no appropriate lexical
items for them. I have also suggested that there are
certa in types of semantic content expressed by
particular syntactic constructions which require the use
o f he2.
4.5. Summary
In this chapter, we f i r s t looked at t h e two
referential forms which axe found commonly in Barbarefio
discourse: pronominal affixes and full noun phrases.
These two forms axe responsible for most referential
work in Barbareno discourse. Pronominal affixes are a
part of what is dmhally required by the grammar, and
these are the most common referential choice. This
least marked form is used for what seems to be the least
cognitively demanding task of keeping track of
established participants in the discourse. The more
marked form, full noun phrases, is used for potentially
more cognitively demanding t a s k s such as introducing
participants for the first time in the discourse, re-
introducing ¥the after thematic break and perspective
change, and when there are other semantically similar
participants in the discourse. When the referents are
core arguments, full NPs are used along with pronominal
affixes. Full MPS are the second most common
referential choice in the data. There are two other
referential forms which are used much less commonly:
independent pronouns and the proximate demonstrative
he?. They are in the middle in terms of eqlicitness of
different referential forms, pronominal affixes the
least explicit and full Nfs the most explicit.
Independent pronouns and the proximate demonstrative he2
are found to be associated with particular discourse
functions and particular syntactic constructions used to
express certain semantic content. Some of the functions
with which these two forms are associated seem to be
more cognitively demamlhg than merely keeping track of
established par-ticipants in the discourse, but less
cognitively demanding than introducing and re-
introducing participants in the discourse.
5. Constituent order
The purpose of t h i s chapter is t o descr ibe t h e
cons t i tuen t order of Barbarego Chmnash while re-
examining some of t h e r e s u l t s obtained i n earlier
chapters. I will f i r s t show t h a t though Barbarefio
allows a number of different cons t i tuen t o rde r types, it
is actually very rare that one finds ut te rances with two
core arguments expressed as full NPs (Du B o i s 1987;
Mithun 1987; Payne 1987, 1990; Duranti and Ochs 1990).
I will suggest t h a t this makes it less f r u i t f u l to talk
about const i tuent order i n t h e t r a d i t i o n a l sense: t h e
order of the subject and the d i r e c t ob jec t wi th relation
to t h e verb (cf. Mithun 1987; Payne 1987, 1990) . I n
t h i s dissertation, cons t i tuen t o rder will be examined in
terms of what is exhibited i n discourse data . W e w i n
see t h a t the most typ ica l order i n Barbareno is a verb
followed maximally by only one core argument full NP.
This f inding nicely c o r r e l a t e s with one structural
c h a r a c t e r i s t i c of t h e language: core arguments are
already marked on t h e verb by t h e grammatically required
pronominal affixes. He w i l l further see t h a t one full
HP following the verb is most t y p i c a l l y e i t h e r a
t r a n s i t i v e d i r e c t ob jec t o r an i n t r a n s i t i v e subject,
supporting the cross-linguistic findings referred to as
Preferred Argument Structure' by Du Bois 1987.
Then, based on the findings in chapter 3, I will
discuss the order of obliques in utterances. I will
show that if both core arguments and obliques are
present as full NPs in utterances, obliques are
typically expressed after core arguments. When
arguments which would otherwise be obliques are encoded
as direct objects of a verb containing one of the core
argument-building devices, relevant noun phrases appear
immediately after the verb because they are now core
arguments.
Finally, I will discuss cases in which full noun
phrases are expressed before the verb. Specifically, I
will discuss two types of preverbal NPs, which co-occur
with two diffexent clitics zi= and ka= respectively. I
will suggest further that these preverbal NPs are not as
common as postverbal NPs in the data, and that they are
associated with rather specialized discourse functions.
5.1. Basic constituent order
We have seen above that Barbarefio allows a number
of different constituent order types. Examine the
following examples:
(1) JH59.332R0333R, txt 2, line 001 59.333L
nas?Qw i l k f i na -8-?uw hi.=& =h
+#&P, &=l =xian
whenç3-bit DPKAftTÇperso DPçARTçrattlesna V DO S
'When a person gets bitten by a rattlesnake,'
(2) JE59.686Ro688R, t x t 76, line 004 59.687L 9
ma? li hugpinti - kay hiho? s~unt&w hi? L i p mail hu=s-pintikay hi=ho?=sountaw bi=l =?an when RM=3-strike DP=DIS=lightning DPÈART=hous
v ll S DO hiho? kayi xeyll hihe~ganta walwa- 2; hi=ho?=kayi xeyli hi=he?=santa valwala DP=DIS=street Haley DP=PRX=Santa Barbara
'When the l ightning struck the house on Haley Street in Santa Barbara,
(3) JH59.662RO663R, txt 67, line 005 59.662R $1 v1
pe - lu ?iAetSi sl~.ltafct&~$& hilawalyfint% pelu ? i d d i S-qili-R-tauy-wag bi=l =?awalvente Pedro ?I=always 3-H&B -R-carry -PST DP=ART=brandy
v DO 'Pedro all the tine carried some brandy,'
(4) JH59.662R-663R, txt 67, line 001 59.662R
Â¥awaly6nt kaima~m.ld- ?S halm6 - l&, ?awalvente ka=l m? -am -&&l-was-D? hu=l a~rooloq brandy KA=ARTasNM-IDF-&ink-PST-EM RM=ART=long.ago
v 'Brandy vas the favorite drink,'
In the first two examples, the subjects and the direct
objects are expressed after the verb, but as can been
seen, the order of the two arguments is reversed in the
t w o examples: V DO S in (1) and V S DO in ( 2 ) . In the
last two examples, some arguments are expressed before
t h e verb. Specifically in ( 3 ) what appears to be the
subject and in (4) what appears t o be the d i r e c t object
axe expressed before the verb (however see 5 .3 . f o r the
syntactic status of preverbal We).
Traditionally, discussions of basic constituent
order have presupposed the presence of two full noun
phrases i n clauses (e.g., Greenberg 1966; Hawkins 1 9 8 3 ) .
In particular, researchers have discussed the order of
the subject and the direct object in re la t ion to the
verb as can be seen i n t h e i r use of abbreviated labels
to talk about constituent order: SOV, SVO, VSO, etc.
However, clauses containing two core argument full NPs,
such as in (1) and ( 2 ) , are relatively rare in my data.
In several hundred pages of text, I have found only 15-
20 transitive clauses with both subject and d i r e c t
object noun phrases. This figure concurs with recent
cross-linguistic f indings (e.g., Du Bois 1987; Hithun
1987; Payne 1987, 1990; D u r a n t i and Ochs 1990). At the
same tine it seems t o make it very d i f f i c u l t to
determine the basic constituent order of Barbareno.
However, the present finding seems to make much
more sense if we consider it i n terms of the general
characteristics of human discourse. That is, there
seems to be a certain cognitive constraint against
presenting too many pieces of information a t a time i n
discourse. Speci f ica l ly , it has been suggested t h a t one
cannot present more than one piece of new information
per intonation unit. An intonation u n i t is a speech
u n i t characterized by its prosodically coherent contour
(Chafe 1987, 1994). One of t h e primary reasons f o r
using f u l l NPs is t o present new information i n
discourse. Since using two full. NFs would often mean
presenting two pieces of new information, one tends not
t o f ind two f u l l MPS with a verb.
I n f ac t , as w e saw i n the last chapter, i n
Barbareno discourse, full NPs and other independent
fo rm (i.e., independent pronouns and the proximate
demonstrative m) are used fox cognitively demanding
tasks which include presenting new information. On t h e
other hand, pronominal af f ixes are used f o r t h e
cognitively much less demanding task of keeping t r ack of
f u l l y es tabl ished pa r t i c ipan t s i n discourse. W e
further saw that pronominal affixes are the most typical
r e fe ren t i a l forms and much more commonly used than
independent r e f e r e n t i a l forms i n Barbareno discourse.
I suggested t h a t a l l these f a c t s ind ica te that
pronominal affixes axe the defaul t choice and t h a t a l l
other referential forms are reserved far marked
ac t iv i t i e s . R e f e r e n - f c i a l choice in Barbareno thus
appears to respond t o the cognitive cons t ra in t
concerning the presentation of new information, and that
seems t o be the primary reason for the rarity of verbs
with two core arguments expressed as full NPs.
Having suggested t h e reason f o r t he r a r i t y of
t r ad i t iona l ly celebrated types of const i tuent order i n
my data, however, we s t i l l want to know about t h e
consti tuent order i n Barbarefio Chuniash. Considering t h e
d i f f i cu l t y of ident i fying t h e basic consti tuent order i n
Barbareno i n t r a d i t i o n a l terms, the approach which
should be taken toward understanding Barbareno
consti tuent order seems t o be t o present a descr ipt ion
of d i f fe ren t const i tuent order types commonly found i n
Barbareno discourse. In other words, I am suggesting
t h a t the first step toward understanding the constituent
order in a language may be to observe what speakers of
the language do instead o f simply s t a r t i n g with
assumptions based on s tudies of other languages.
Looking at t h e question i n t h i s way, what appears
t o be t h e most typ ica l consti tuent order type i n
Barbarefio discourse is a verb followed by at most one
core argument f u l l HP. The following is a typical
sequence from the data:
(5) JH59.205L-205R, txt 21, line 001 59.205L
nas iyexp&ti ihe?ltk&r ? i?çlsa?tuhSy na =S-iy-expei ))i=he?=l =&U? ?i=?al-sa?-tuhuy wben=3-P&-sing DPÇPRX=ART=bird.s ?I== -FUT-rain 'When "the huitacoches sing, it is going to rain.
7iyalaKiwa law iheituhfi* iy-?al-aqfiwalaw bi=he?=l çt.ubu-f PL-MM -like DP=PRX=ARTçrai They do like rain.
kg b'a kasiYexp8tg k±p ka=s-iy-expeS now ~o=3-PL-sing They sing and sing,
PS ^
?iyalti^ -&An hi?alsa?aktin& - n'a hiiuwfl - h&, PS iy-?al-&bin hi=?al-sa?-alcti-nuna hi=l = ? u d u likely PL-KM -know DP=liM -FUT-coffle-bring DPa=ART=food they know it is going to bring food,
siyexp&ti hino&? hina5tuhtuhSy B-iy-expefi hi=nofeo+-? hi=na =a-R-tuhuy 3-EL-sing DP=inuch+EM DPçwhen=3-R-rai They sing a lot when it is raining.'
This is the entire text. There are three core argument
f u l l NPs in t h i s sequence, which are underlined. As can
be seen, all. of them are expressed after the verb. In
all other cases, the core arguments are expressed by
pronominal af f ixe s only.
Here i s another example:
(6) JH59.88R-89R, txt 9, line 001 59.89L
a) m&-iisi<ra(iaASlA he?iihi&: ? , mali S-iy-R+aniin -? be?=l =R+?invu - * ? when 3-PG-R+eat.a.meal-? PRX=ARTÈR+Indian-E *<When Indians are eating a meal,>
i-siye-tip&-wil. ?is#-iy-e-tipawil ? 1=3-PL-M-talk they don't talk.
? iy6 -&m psal^-~haU^alalt -nut*, ?iy& p-R+saqhalalm-us -vim though 2-R+holl~~ +AT-3P0 tho you holler at them
?isiyat?it&-Kin, ?i=s-iy-e-itaq-in ?1=3-PL-N-he~-20 they don't hear you.
nod? ~iyaiaiflw iit - Ain noho iy-?alal-?uw hi=! ==?&Q very PL-ACT -eat DPsARTsmeat <they were real mat-eatera.>
tfiunasamlw& -wan iiwi, fiu çn =s-am -ivawan his1 =?a+i soWthatçwhen=3-IDF-cu DP=ARTÇID <So that when they cut someone,>
? i&'^at6 -wits ish& - las iho? f i w a w t - his ?i=s-SutuwiE hiss-%alas hi=ho?=l =?iwawafu.B ?I=3-be-quick D P Ã § 3 - h e a DP=DIS=ART=cut the cut heals quickly.
nan6ho sa&dw ih&.n, na =noho #-am -?uw b n n when==much 3-IDF-eat DP=AR'T^Beat When they eat so much ineat
i a k t 6 -wit8 h i a b t lag iiiwawi - &is. ?i=s-Sutowi& hiss-su-xai&8 hi=l =?iwawaAiS ?I=3-beequick D-3-CA-heal DPÇARTçc <the cut is soon healed.>'
Again, this is the entire text. As can be seen, there
are several core argument NPs appearing after the verbs.
In a l l other cases, the core arguments are expressed by
pronominal affixes alone.
Examine the following example:
( 7 ) JH59.128I,-128R, txt 11, line OOla 59.1281.
*The frogs must have a certain control over the w a t e r .
tayltS s 2 ipas ilwl- JA kayk6 s-?ip-as himluwisa because 3-say-RP DP=Luisa For Luisa. said:
?itauswun i h e l w a ~ w a f c a -k, itaq-us -win hi=he?=l =R-wa&a& - 0 7 hear-DAT-3PO DP=PRX=ARTÈR-frog.sp-E listen to the frogs,
S ? tuhfly. a-S&?-tuhuy 3-FUT-rain it is g. to rain.
And pretty soon r<inclouds would appear. '
Again, t h i s is the e n t i r e tex t . There are five core
arguments expressed as full M P S . Other core arguments
are expressed by pronominal affixes. As can be seen,
all of these NPs except one appear after the verb. The
one HP which appears before the verb, 'the frogs ', i s
found at the beginning of the text. I will discuss this
type of preverbal NP later i n this chapter. These
examples thus suggest that a verb followed by a t most
one core argument full UP is the most typical
c~~stitue~t order in Barbarefie discourse. In factt 1
have found that the vexb-initial order accounte fur
approxdtely 90 percent of the clause8 in the datae2
ZA chapter 4, X suggested that full RIPS are used
when the referents must be explicitly mentioned for such
xeasons a% intxaducing pa~icipats for the first t h e
in the discaurse, clearly distinmishing some
paeicipmts from other participants and re-introducing
pmicipants after thematic breaks ox when these are
changes in perspective. The f u l l NE% in ( 5 ) - ( 7 ) can be
accaunted for by these factorsm In ( 5 1 , all of the
thee fall =S express referents which are being
introduced for the first time in the discourse. In ( 6 1 , the same obse~ation can be made segarding the first
four NPs; those paeicipants axe introduced for the
first t h e in the discourse* Regarding the last two
NPs, 'meate in line h) and 'cut * in line i) , these items have just been talked aboutt so in a strict sense this
is not their fixst introduction* What seems to be
happening is that it is actually a part of a commonly
observed narrative convention in which basically the
same utterance is repeated. Notice the repetition of
similar utterances in lines e)-i)m Far this reasan, 1
would like to suggest that the two HPs %eat8 in line h)
and gcutg i) are still being intzuduced into the
discourse for the f i r s t t h e a In ( 7 1 , a l l NPa except
#frogs in line c) express referents which are
introduced for the f i r s t time in the &scaursea As for
?frogsg, notice that it appears in a quote. As we
discussed in the last chapter, going from the story line
to a quote creates a change in perspective which seems
to motivate the use of a f u l l M? here.
The above observations lead us to the following
hypothesis regarding the basic constituent order in
Barbareiioz a verb is generally fdlawed by at most one
core argument full KP which has to be present for
various reasons. That is, what we have here seems to be
a verb which expxesses either an event or a state and
possibly one core argument expressed as a full W tu
perfom such cognitively demanding tasks as introducing
and re-introducing pa~icipmts and clearly
distinwishing some peicipats fromother pa~icipants
in the discoursee The suggested basic cunstituent order
is summarized in the following:
v (NI
This may look like a rather mdhentary structure f r o m
the perspective of syntax, but as we saw above, it is
possible paztly because most of the xeferential tracking
activities in Barbare60 are pe~fomeci by the pronominal
affixes on the verb. That is, we axe observing the
moqhological richness of the language c~~zelating with
its syntactic sparseness. Or to put it another way, the
moqbological tightness of the language is correlating
with its syntactic looseness.
Studies have shown that cross-lin~istically
transitive direct ~bjects and intransitive subjects are
the structuxal d o t s in which new information is often
expressed (Du Bois 1987). As we just saw above, in
Barbarefia, full W s typically fallow the verb and they
are present because very often they express new
information. Interestingly, when 1 looked at the
syntactic relation of postvexbal NPs, I faund that they
are mostly transitive direct objects and intransitive
snbjects. In fact, most af the a s in ( S ) - ( ? ) above
seem to have one af these syntactic relations. In (51,
*huitacoches (a type af bixd) in line a) is an
intransitive subject, and @xainP in line b) and *foodF
in line d) axe txansitive direct object8 xespectively.
In ( 6 1 , 'Tndians* in line a) and @cut8 in line g) axe
intransitive subjects, and ' m e a t in lines e) and h) , rsameuner in line f), and OcutF in line i) are
transitive dixect objects. Sn (71, rwaterr in line a)
and ?€ra ' h line c) axe txansitive direct objects,
and 'c1oud8 ' in line e) is an intrmsitive subject. The
NP rfrogsf in line a) may seem to be a case of a
transitive subject, hut as I will show in 5.3, it does
not hold a syntactic relationship with the following
part af line a): it is not a syntactic a ~ ~ e n t of the
verb * ~ l e @ l The NP 'LuisaP in line b) seems to be
another candidate for a transitive subject, but even
this case is suspecte First of all, 2ip 'sayF in b) is
a verb which can take only one nominal (i.e., the
speaker) ; another element taken by the verb is the
speaker's utterance* The verb 2in is thus more like an
intransitive verb, and the 'Luisa' moxe like an
intransitive subject. Further, it should be noted that
in my data I have not seen t h e verb m marked with the plural ciixect abject euffhc -wune3 This fact sugge~ts
that Z ~ D may nut be a transitive verb after all (cf.
Munro 1982)e Regazdless of its actual syntactic status,
it thus seems reasonable to suggest t h a t a 'to say is
in essence like an intransitive verb, and accordingly
the NE! *LuisaO is an intransitive subject.
To sum up, mast of the f u l l NPs i n Barbaxefia which
appeax postverbally seem to be e i t h e r in trans i t i ve
subjects or transitive direct objects* As we saw
earlier, t h e s e NPs are pxesent prhaxi ly because they
express new informatione These facts suppart the crass-
l i n g u i s t i c finding of Du B o i s (1987), which suggests
that trans i t ive direct objects and i n t r a s i t i v e subjects
are the grammatical slots in which new infamat ion is
o f t en presented. More hpartantly, they show tha t , i n
terms of g~ammatical reli itians# the mast t y p i c a l
constituent order i n Barbare60 is a verb perhaps
followed by e i t h e r an in txans i t ive subject o r a
t rans i t i ve direct object. Transit ive eubjects are
normally expressed as pronominal prefixes (cf . Dn B o i s
1987) The fol lowing s ~ i z e s the suggested
constituent order h Barbaxe50 in te- of grammatical
relation:
In this sec t ion , we have thus came up w i t h a
description af the basic constituent oxdez o f Barbare60
based on the examination of discourse data* We have
discussed the func t iona l and t h e s t r u c t u r a l
c h a r a c t e r i s t i c s of this b a s i c cons t i t uen t order. Though
what is presented here may seem r a t h e r d i f f e r e n t from
t r a d i t i o n a l l y celebrated types of c o n s t i t u e n t order
(Greenberg 1966; Hawkins 1983), this i s apparent ly what
Barbareno speakers ope ra t e with when they use the
language in discourse, and thus it can be said to
r ep re sen t what they know about t h e s y n t a c t i c s t r u c t u r e
of t h e language.
5 . 2 . Obliques
Based on the f indings i n chapter 3 and i n the last
section, I would l i k e to discuss i n t h i s s e c t i o n t h e
order of obliques with relation to core arguments in
Barbareno ut terances . As we saw i n t h e l a s t sec t ion , i n
Barbarefio discourse core argument f u l l NPs are t y p i c a l l y
e i t h e r t r a n s i t i v e direct ob jec t s o r i n t r a n s i t i v e
sub j ec t s , while t r a n s i t i v e subjects are t y p i c a l l y
expressed as pronominal p re f ixes .
I n order t o t a l k about the order of c o r e argument
NPs and obl ique NPs within u t t e rances , of course we need
a sufficient number of utterances in which both of these
types of arguments ace overtly expressed. However, as
I pointed out earlier, it is actually very difficult to
find combinations of two full NPs in Barbarefio
utterances. Specifically, it is not common to find both
transitive subjects and obliques overtly expressed in
utterances because in Barbareno discourse transitive
subjects are typically expressed as pronominal prefixes.
Further, obliques axe in some sense "extra argumentsw,
which are only occasionally used in utterances. For
this reason it is not realistic or even useful to
determine the order of obliques in relation to
transitive subjects, leaving us only with the question
of the order of obliques in relation to other types of
core arguments (i.e., intransitive subjects and
transitive direct objects).
We saw in the last section that the basic
constituent order in Barbaz-e5o is a verb followed
perhaps by either a transitive direct object or an
intransitive subject. As we saw in chapter 3, if both
core arguments and obliques are present in utterances,
obliques typically appear after core arguments. The
following examples illustrate the order of obliques in
relation to intransitive subjects:
(8) JH59.6L-7L, txt 3, line 002 59.6R
?iyena siypilikllkw ilpil46-li hjunl-&up, ?iyena S-iy-pililclaw hi-l -oimoli $ i d S u ~ if 3-pl,-f all DP-ART-beans DP=down
v S GOAL 'If beans fell on the floor,#
(9) JH59.658R-659&, tx t 63, line 002 59.658R
hilsa?e~w^l hiskitdn hi=l =sa?-eqwel hi=s-kitwon DP=ARTçFUT-m& D-3-co3~e.out
v hilnohd? s?&i ihca-?is h iho?66k, hi=l ==nofio+-? a-?ua8 hi=l = ? a x u h bi=ho?=n-?tk DP=ART=veq+EM 3-stink DPÈART=bloo DPsDIS=2-nioutA
S SOURCE 'which will snake very stinking blood come out of your mouth. '
(10) JH59.281R-283I., txt 244, line 005 59.283%
sKi.11-wil ilsy6-lpe hiho?sis?6k ika?6?, 8-qili-wil hi=l =sveloe bi=ho?=~-is-?&k hi=ko?o? 3-?iAB -be DPçARTÈserpe DP=DIS=3-AP-mouth DP=Zaca.Lake
v S LOC 'There used to be a <serpent> in Zaca Lake*
(11) JB59.409R-412L, txt 147. line 010 59.412L
8 ~ i l i t ~ ~ 6 kiyantik m6 loK, S-qili-Sho k-iv-antik hu-rooloq 3-HAB -good l-PL-spirit RM-long-ago
v S TIME ?our s p i r i t was always glad in ancient times,'
(12) JH59.485L-494L. txt 87, line 053a 59.493L
swilwaS isiy?ak8y i~ he?ik8skes?i?- ? S-wil-wag hi=s-iv-?akaviS he?=l =R-kesu - * ? 3-be -PST D-3-PL-rack PRXmARTÈR-cheese-E
v S REC 'there was their <rack> for the cheeses'
The verbs are immediately followed by the intransitive
subjects in these examples. The obliques appear after
the intransitive subjects. As can be seen, the meanings
of obliques are not formally marked (Beeler and Whistler
The following examples illustrate the order of
obliques in relation to transitive dixect objects:
(15) Ji359e676R-681Lt tact 75, L i n e 005 59.679R
A s can be seen, the verbs are d ixec t ly fallawed by t h e
t r a n s i t i v e direct objec ts in these examples. The
obliques appear after the t r a n s i t i v e dizect objects.
All of these examples thu8 show t h a t , in term of t h e
co re -ob l iqe d i s t i n c t i o n , the bas ic consti tuent ordex of
Barbare60 seam to be a verb maybe followed by a core
argument and then maybe fu r the r followed by an oblique.
As 1 suggested earlier, t h e one core argument is
typically e i t h e r the i n t r a n s i t i v e subject o r the
transitive d i r e c t abject* These findings regarding
Barbare50 cons t i tuen t order are summarized i n the
following:
It should be noted again t h a t t h i s type of order i n
which NPs appear a f t e r t h e verb amounts to about 90
percent af c lauses i n t h e data.
We saw in chapter 3 t h a t Barbaxefio has a process i n
which certain azguments which would othemise be
abliqyes become direct objec ts by appeaxing with one of
t h e core a r p e n t - b u i l d i n g suff ixes . When this happens,
and both the tikect object and t h e pat ient are expressed
as f u l l NPs, t h e d i r e c t ob jec t i s typically expressed
immediately af ter the verb and is followed by the
patient, reflecting the status change. Spec i f i ca l ly ,
t h i s happens with dat ives and instruments. For datives ,
when this happens the verb takes the dative suf f ix -us
and the plurality of the dative can be marked by the
In both of these examples, the verb takes the euffix
-us, and the dative is now the direct object appeaxinq
r ight after the verb. The patient, l abe l l ed #PATt,
appears after the d i m e t abject. Xn (191, the plurality
of the direct object FIndiansg is marked by t h e direct
object plural auffh reflecting the status change.
can be expressed as direct objects by appearing with an
-in-marked vexbe The new direct object can be marked
with i f the referent is plural, and typically it
directly follows the verb and is fallowed by the patient
i f bath of these arguments are expressed as full NPs.
E x d n e the fallawing examples:
(22) JHS9.18lL-I82Lr t x t 17, line 001 59.18E
1x1 these exwles, instruments are direct objects
appeazing right aftex the verb? which is marked with the
~ n s t m e n t a l suffix -ine The patient, labelled fPATff
follows the direct object .
To sum up, I have 8uggested in t h e last two
section8 t h a t the basic consti tuent order of Barbare50
is a vexb perhaps followed by e i t h e r the t r an s i t i ve
direct object er the i n t r an s i t i ve subject , and pexhaps
fuxther followed by oblique^. Certain types of
arguments which would otherwise be obliques (i.e.,
da t ives and ins t ruments) are cast as direct objec t s of
verb8 containing one of t h e core =ven t -bu i l d ing
suff ixes . When t h a t happens the direct objects
t m i c a l l y a p p e a d i r ec t l y after t h e verb and may be
folluweci by t h e p a t i e n t , reflecting the status change.
5.3. Preverbal elements
In t h e preceding sec t i ans , we have examined the
basic cans t i tuent order of Barbaxeiio Chtamash. In this
sec t i on , w e w i l l examine nonbasic types of cons t i t uen t
order which u e marked i n terms of frequency and
functions. As has been suggested eaxlier, my data
include u t t e r ances i n which full HPs are expressed
before the vexb, as i n ( 2 4 ) and (25 ) :
(251 JH59.662R-663R, t a c t 67 , line 001 59.662R
In ( 24 ) , 'Pedro8 , which appears to be the subject, is
expressed before t h e verb, and i n ( 2 5 ) , *brandy8, which
appears to be the direct object, is expressed before t h e
verb. I ant discuss ing cases such as t h e s e i n t h i s last
s e c t i o n because they are not as f requen t as the bas ic
c o n s t i t u e n t order which we discussed i n t h e last two
sections and f u r t h e r they seem to be assoc ia ted with
marked pragmatic funct ions. The preverbal HP order
amounts t o about 10 percent of c l ause s i n t h e data.
Among ut te rances i n which f u l l HPs occur before the
verb, there appear t o be at l e a s t t w o d i f f e r e n t types.
The first type is charac te r i zed by a p r a c l i t i c ?i= on
t h e c o n s t i t u e n t following t h e preverbal HP, which is
t y p i c a l l y t h e verb. Examine t h e following examples:
(26) JH59.59OR-591L, txt 42, l i n e 001 59.590R
'The Indians did not expect evaryfching they ate to be tender. *
In both o f these examples, there is a full NP occurring
befare the verb, and the verb is marked with a proc l i t i c
2i=. This order is not rare i n d i scoux~e , but it is not
as common as the basic constituent order Ln which f u l l
W s occur after the verb. Further, this order i s
chuacterized by the marked function which the pzeverbal
The pzagmatic function associated w i t h t h i s type af
preverbal NP seems t o be best chaacter ized as the
introduction of a topic (Beexer and Whistler 1980) .
These Ws very often express i d e n t i f i d l e characters
such as ?Indians and 'the old-time peopler,
s tories . Theae kind8 of chaxacte~s /elements are first
set up, then pzedications relevant t o thtm are made i n
the res t of the utterances. (See below for a discussion
of t h e syntactic relationship between this type of
preverbal HP8 and the verbs.]
One piece of evidence supporting the proposed
function of preverbal NPs w i t h 2i= may be seen i n the
fact that they are ch=acteristic of the utterances at
the beginnings of stories. In fact, all of the examples
of these preverbal NPs given in this section come fzom
the openinqs of stories ; mast of them are the very first
utterances. If prevezbal NPs with 2 i = axe associated
with a topic introducing functicm, this d i s t r 3 u t i o n a l
characterist ic makes sense because main p m i c i p a n t s of
the s tor i e s are naturally intzoduced a t the beginning of
the s tor ies . The following examples are typical cases
of preverbal W s with the p r a c l i t i c 3i=:
( 2 8 ) JE59.32OL-322Rr tat i t 165, line 003 59-32UL,32OR
'The early-Spaniards soan after they emer established vineyaxda .
In (281, the HP 'the euly-SpaniardsR is first
presented, and the rest of the utterance talks about
w h a t they did in the Chumash land. In (29 1 , the MP Wwtir seeds that they used to grow here and theret [an
HP ?seeds @ and a relative clau~e modifying- it4) is
first presented, and the rest ~f the utterance is about
their characteristics.
The fallowing example further illustrates the topic
introducing function of pseverbal HPs :
(301 JH59.657R-658L, t x t 61, line 091a 59.657R
t\ b) ?iamsf i - t ip Pihe?
a=s-am -su-tip hi-he? ?X=3-XDF-a-salt DP=PRX they salt it,
The W 'the leaf of angelicat is first presented, and
the rest of the utterance talks about haw it is used as
a remedy fez poison oak. This example is particulaxly
interesting because it shows that more than one
predication caulci be made for the pxeverbal NP: @the
leaf of angelica8 is fisst presented as seen in line a),
and two predicatians regmding angelica leaf aze made,
both of which are marked with the proclitic ?is, as seen
in lines b) and c). Examples such as t h i s furthex
support the proposed function of the preverbal NPs by
showing that the infannation expressed by the preverbal
IWs plays a role at the level af discourse.
It should also be nated that there are often
preverbal elements which are locatgve and temporal
expres S ions :
t h e ranch we always have a goud-ciippexer
*When one &eve a buggy or wagon along one of the old-thie raads @
Both of these examples are the first utterance of the
stories. It seems that theee preverbal elements set up
a frame in which the rest of t h e story is t d d .
XLthough this particular function seems different from
the topic intraducing function which we have examined s o
farf it i s interesting t o see t h a t both of these
functions are served by preverbal elements occuxxing
with the same psoclitic and that they concern the level
Finally, as some of the preceding examples almady
suggest, the elements which accur before the verb in
t h i s t m e of constituent order m y not be the syntactic
azguments of the vezb (Wash 1995b), as in ( 3 3 ) :
(331 JH59.345L-347L, txt 59, line 001 59.345R
ihe? l-vul6-wu
PRX=3-MM -fat-RES DPçPRX=ABTçshe
r-ctallow*- i s the fat o f the mutton, it is a very hard grease. "
A literal translation of this example would be 'sheep's
fa t , its grease is very hardF. The preverbal HP
'sheep's fat0 is not a semantic argument of what is
represented in the la t ter part of the utterance.
Here i s another example:
( 3 4 ) JH59.641R-642R, tact 7 8 , l i n e 005 59.6421.
*Salt. sometiaes cones out white on a person. On a hard working roan, <it comes out> s a l t on h i s back.
\\ ?i n67n6 sutl-SiS h i s son soxyop hiho?lkainl- .
noim S - u t i ~ i i t lti=s-on ~ o x y o p hi.=ho?=~ = Is amiaa ?I very 3-hard DP=3-IDF wash DP=DIS=ARTÈshir
t h a t s a l t coning out makes his sh ir t hard t o wash.'
A l i t e r a l translation of ( 3 4 ) would be 'the coming out
salt, washing the shirt is hard". As the translation
given i n the example suggests, what it means is 'because
of the salt coming out, it is hard t o wash the shirt'.
Again, the preverbal NP "the salt coming out' is not a
semantic argument of what is presented i n the la t ter
part of the uttenince.
W e have seen in earlier chaptexs that, in
&uba.refic#8 verbs themselves can constitute complete
predications. This fact, h conjunction with the above
examples, seems to suggest the following analysis
regaxding the structure of the preverbal constituent
order with the proclitic ?i=. In this particular order,
some topic element and same complete predicatian
relevant to it are juxtaposed on the basis of a rather
loose relationship between the two: the relationship is
not syntactic, but pragmatic.
The present analysis is further supported by the
morphology af the preverbal W s . Preverbal PIPS do not
take the dependent proclitic hi=B which is a
chuacteristic of NPs accu~~ing after the verb (cf.
Beeler 1970aB 1976; Beeler and Whistler 1980; Waeh
1995b). That is, even a glance at the data shows that
typically postvesbal NPs are marked with the proclitic
hi=, but this proclitic does not appear on preverbal
NE%. Fox example, none of the preverbal W s presented
in this section are marked by hi=. So here is a
phenomenon in which the morphology of Barbarefio t x e a t ~
preverbal =S differently from postvecbal WS# which are
cleuer cases of syntactic arguments of the verb. And
this seems to i n d i c a t e further t h a t preverbal NPs are
no t syntactic axguxnents of the verb in the ut terance*
The above analysis concerning the syn tac t i c status
of the poreverbal NPs occurring with seems to accord
nicely with the function of these NPs which 1 suggested
earlier* That is, preverbal NPs have been suggested to
se rve a tapic introducing function, a discourse-level
funct ion, and they are conjoined with ~ o m e pred ica t ians
due to praqnatic relevance.
Thus f a r w e have examined one type of preverbal
c o n s t i t u e n t order* This type is characterized with
preverbal NFs occurring with t h e p r o c l i t i c +?is on the
word f d l o w i n g them, and is not as conman as the basic
c o n s t i t u e n t order in which full NPs follow t h e verb. In
terns of function, it can be characterized as a t o p i c
c o n s t m c t i o n i n which a pzevexbal topic element is
conjoined with some camplete predicat ion based not on
syntax but on pragmatics*
The second type of constituent order in which NPs
occur beeore the verb is c h a a c t e r i z e d by the p r o c l i t i c ,
Jca=f on the verb. Examine t h e f o l l ~ w i n g examples:
(36) -S9.554&562R, txt 106, line 008
#It wa8 grown up people that drank toluache.'
In (35) and (36)# rflowers@ and 'ruoteF which appear to
be the direct objects, are expressed before the verb,
and in (37)? the subject 'grown up people' i s expressed
before the verb. In all of these examples, the verb is
marked with the pxoclitic ka=, and further the
information expressed in the prevexbal NP seems to be
focused, which is part ly observed in the use of such
elements as #rather8 (example (36)) and ,onlyr (example
(37)). Further, ae the examples show, this type of
order is very dten tzanslated by Harringtan with the
cleft construction in Sng1isha M d i t i ~ n a l l y ~ often t h e
infarmation expxessed in these prevexbal NPs may be
contrasted with other information in the utterance, as
seen in the following exaaplez
(38) JH59.58L-GOL, txt 179, line 003 59.59L
'They say t h a t when you see the s p i x i t of someone else,
either you axe g. ta dier or that pes8on is g. to d i e s r
In t h i s example, pthat person @ is contxasted w i t h 'you r ,
and is expressed preverbalzy w i t h the proc1iti.e ka=.
St appears that the preverbal NPs occurring with
ka=, like those with 3 i s f are not syntactic azgtments of
the following element. There are several pieces of
evidence far this analysis.
F i r s t f some ka=-marked words axe actually nouns
rather than verbs:
?It must have been an ep. that chilcken only not gxown-ups e '
In this example, there are two words marked by &a=, bath
of which seem to be nauns because t h e y take the
alienable possessive prefix i s - . Stmctura l ly , what w e
have here in each case is t w a =S jutapased next to
each other. For this reason, though the first IQ in
each cunstmctian (Fchildren* and *gram-ups* } seems to
be the possessor of the following noun epidemic , these
NPs probably axe not its syntactic argument.
The following example further supports t h e present
analysis :
( 4 0 ) JE?59.359R-367&, t x t 141, line 004 59.360R
-> K e b h ho?a;aJiiin kaaiykepkepnu?l-waS ^ysg-iy-R+kep -nu?-iwa5
or DIS-ocean KA=3-PL-R+bathe-LOC-NPST or the ocean was their bathing places. @
The word marked by Jca= is a noun: it i s marked with the
locative nominalizer pm2 and the suffix -iwaS, which
indicates the defunct status of the referent of the
noun. An HP ' a deep place i n a creek, a basin, a slough
or the ocean* is placed before that. Here, there does
not seem to be any syntactic relationship between the
t w o HPs: the f i r s t IIP does not even have a possessive
relationship with the J c a = - ~ k e d m. Such examples seem
to show that utterances containing ka= may be equational
constructions with no syntactic relationship between the
t w o parts.
Second, supporting the above analysis, there are
cases of ka=-marked utterances without an NP preceding
them, as in:
(41) 3H59.139L0151R, txt 198, line 060 59.149L
P% ~eiiia?l'l kaiiy-filihi.kwan, P% Aesnaii kÇ= =? -iy-qili-hik-wun perhaps EMejust KA=ART^NM-PL-HAB -do -3PO *that was probably all that they used to do.'
(42) JH59.428R-432L, txt 150, GM line 001 59.431R e
? ik%?mekaBi~iina&iii -wag ?ikhuÈ&e=Is&=a-iy-?iÈndiÈ-iw but =EM=K&=3-PL-custom -NPST *but that was their custom.*
Agreeing with the e a r l i e r suggestion that ka= may be a
type o f predicate marker (Beeler 1970a, 1976 ; B e e l e r and
Whistler 19801, these examples show t h a t &-marked
ut terances are complete without an HP preceding them.
Notice t h a t kaà appears on a verb i n (41 ) and on a noun
i n (42) : t h e former takes t h e habi tua l p re f ix Uli- and
t h e th ixd person plural d i z e c t ob jec t s u f f i x -mq, and
t h e latter takes t h e defunct nominal s u f f i x -iwas. The
f a c t that ka=-marked ut terances are complete without an
HP preceding them suggests t h a t those NFs appearing
before Jca= may not be real syntactic arguments of what
follows . Final ly , morphology a l s o supports t h e present
analysis. Speci f ica l ly , these preverbal MPS are
typically not marked by the dependent p r o c l i t i c hi=,
which is c h a r a c t e r i s t i c of postverbal NPs, clearer cases
of syntactic arguments of t h e verb.
The f a c t s presented above a l l seen to point t o t h e
conclusion t h a t this second type of preverbal order is
a focus construct ion i n which t h e preverbal HP is
syn tac t i ca l ly independent from what follows and
funct ional ly t h e information which t h e HP presents is
focused. Final ly , in terms of d i s t r ibu t ion , I should
note t h a t , compared t o other types of cons t i tuen t order,
this order seems to be rathex marked; it is used
relatively infxequently in Basbaxefio discoursee
W e have thus discussed two types of prevezbal
orders in Barbaxefio, each of which is characterized by
a particular proc1iti.c CO-occurring with theme
Intereatingl~~ 1 have come across a f e w cases i n the
data which seem to exhibit preverbal NPs aesociated with
both of these two types of orderse Heze is one such
?The Indians l i k e d big game animals rather than amall.?
There are two NPs occurring befare the verbe The first
NP 'Indiansr is followed by the proclitic Z i = , and the
second M? 'big animals * is followed by the proclitic
ka= This utterance seems to be about one
chmacteristic of the IndiansF which ~uggests the topic
status of the first NP. -hext part of the
txanslatian @rather than smallr F which is not in the
Chumash, seems to suggest that the infuxmation expressed
in the secand MP is focused. Finally? though in my data
I have faund only a few cases in which two NPs occur
before the verb, all of them show the ozder shown in
( 43 ) : topic and focus. Thus, these examples not only
show that these t w o types of preverbal NPs can be used
together but further suggest that there are i n fact two
di f ferent s tructural @ l o t s before the verb.
Thus w e have seen Barbareso constituent orders in
which full Ws axe expressed before the verb.
Spec i f ica l ly , I have discussed two types of orders, each
of which i s chaacterized by particulax structural and
functianal characteristicsm It should be noted that the
functians of these two preverbal order types are very
similar to what has been discussed in the l i terature
concerning preverbal orders c r o s s - l i n ~ i s t i c a l l y (Mithun
1987, 1993; Payne 1987; C&ng 1991)e That is, it has
been found i n many languages that preverbal elements are
associated with such prawat ica l ly marked functions as
topic and focus.
5-4- S--
In this chapter, w e have looked at canstituent
order i n Barbarefie. 1 first suggested that the most
typical const i tuent order i n Barbaxe50 i s a verb
followed by at most one core argument full NP . Based on the functions of full Ws given earliex, I then
presented a functional description of Barbare60 basic
constituent order as a vexb which expresses either an
event or a state which may be followed by one core
argument full NP performing such coqnitively demanding
tasks as introducing and re-introducing pmicipants and
clearly distinpishing some pmicipants from other
paeicipmts in the discourse. We then saw that these
f u l l NPs axe typically either intransitive subjects or
transitive direct objects, which suggests a structural
description of Barbareso basic constituent order as a
verb which may be followed by either an intransitive
subject or a transitive direct ob ject . Transitive
subjects are normally expressed as pronominal prefixes.
I further suggested that this rather rudimentary
structure is possible because most refexential tracking
activities in Baxbareiio are performed by the pronominal
affixes an the verb.
Then I talked about the order of obliques in
relation to core arguments. Specifically, 1 suggested
that if utterances have bath a core argument and an
oblique as full NPs, the oblique typically follaws the
core argument* Barbarefio has a pxocess in which c e r t a i n
arguments which would otherwise be ublicpes (da t ives and
instruments) are treated as a e c t objects by appearing
with a verb containing one of t h e core u m e n t - b u i l d i n g
suffixes* 1 discussed t h e consequence i n cons t i tuen t
order of t h i s process t h a t alters g ~ a m m t i c a l xela t ions .
I shmed t h a t when t h i s happens and both t h e d i r e c t
ob jec t and t h e p a t i e n t appear i n t h e ut terance, the
direct object d i r e c t l y fallows t h e verb and is followed
by t h e pa t i en t? r e f l e c t i n g t h e s t a t u s change. To sum
up, B a r b a r e 5 0 basic cons t i tuen t oxder can be summarized
as follows: a verb maybe followed by one core axgtment
f u l l NI? ( e i t h e r t h e t r a n s i t i v e direct objec t or the
i n t r a n s i t i v e s u b j e c t ) ? and maybe f u r t h e r fo lhwed by
obliques.
Finally, w e examined cons t i tuen t o rder types other
than t h e basic a m . Specif ical ly , I discussed cases i n
which NPs a c c w before verbs. These cases seem to be
more marked i n terms of both frequency and functions
than the type that has basic cons t i t uen t ~ x c i e r * I
suggested t h a t there are at least t w o diffexent types of
preverbal orders* I f i r s t d i s c u ~ s e d cases i n which
preverbal NPs co-occur with a p x o c l i t i c Ji= an t h e word
following them* I suggested t h a t these NPs may not have
a syn tac t i c r e l a t i ansh ip with t h e vexb because t h e i r
r e fe ren t s are of ten nut semantic arguments uf what is
represented i n t h e l a t t e r part of t h e ut terance*
Further, these NPs do not take t h e dependent p r o c l i t i c
hi=, which is c h u a c t e r i s t i c of Ws fo lhwinq t h e verb.
1 also suggested t h a t t h e function of these NPs can be
chmacter ized as t h e introduction of a topic ,
representing such i d e n t i f i d l e characters as ' Indiansr
and ' the old-time peopler, m h a t e l h m a n beings, and/ar
main par t ic ipants o f t h e s to r i e s , o r t h e frame i n which
t h e s t o r i e s are to ld . I then discussed cases i n which
pzeverbal elements CO-occur with a pxoc l i t i c ka= on t h e
verb. This is a xe la t ive ly mintx construction which
occurs rather infrequently. I suggested t h a t t h i s
canstruction is used when t h e element accurr5ng before
t h e verb i s focused in t he utterance. I ale0 suggested
that these preverbal NPs xnay not belong syn tac t i ca l ly t o
what follows.
Notes for Chapter 5
1. Even among these cases, the basic constituent order
cannot be easily determined: approximately half of them
had VOS order and the other h a l f VSO. These cases do
not include such examples as ( 3 1 because, as I w i l l
suggest i n 5.3., preverbal NPs do not hold a syntactic
relationship with the verb.
2 . This number includes cases in which oblique NPs
appear after the verb. We w i l l . examine such cases i n
3. Of course, this fact could simply be a consequence of
the fact that quotes are normally singular. As we saw
in czhaptez 3 , the verb Zin @say8 can become a verb that
takes two nominal arguments 'say to/telle by addition of
the dat ive suffix -us, but that i s not what we have i n
( 7 )
4. Suzanne Wash (p.c.) analyzes the material on the
second line in (29) as an appositive clause t o 'their
seeds8 instead of a relative clause modifying it.
6. Conclusion
In this dissertation, we have looked at the
interface between grammatical structure and functional
factors in Barbareno Chmnash, a California language, by
closely examining discourse data. In particular, this
was done based on the findings and the assumptions in
one area of research called 'information f low8 (Chafe
1987, 1994). In this final section, I would like to
review the discussion and summarize the major findings
of the disser'tation.
In chapter 2, as an introduction to Barbarefio
Chumash, I presented its basic structure. First, I
discussed the distinctive sounds in Barbareno. I showed
that it has s i x vowels and a relatively large number of
consonants. Then, I briefly discussed its basic
moqhological structure while presenting the two main
lexical categories (nouns and verbs ) . Next, I discussed
the argument structure of Barbareno; I illustrated the
nominative-accusative patterning of verbal affixes which
indicate the arguments of the verb. I further suggested
that these affixes are pronouns because they are used
without full NPs. I then discussed the specifics of
this pronominal system. Next, I gave a brief discussion
of Barbareno constituent order; I showed that Barbareno
exhibits various constituent order types. Finally, I
described basic clause-combining patterns in Barbarefio.
After examining the basic grammatical structure, we
looked at three tightly related aspects of Barbareno
grammar and their functional correlates. In chapter 3,
we looked at the argument structure of the language. I
first pointed out that certain arguments in Barbareno
are expressed by grammatically required pronominal
affixes (the subjects and the direct objects). That is,
Barbareno treats only certain arguments as integral
parts of the verbs, which can be taken as criteria1 for
core arguments. I then pointed out that facts about
constituent order support this observation. That is,
when these arguments are full NPs, they tend to be
expressed before other NPs ( i.e., obliques ) in a clause.
We then examined the structure and "the function
associated with two verbal suffixes ?us Odativeg and -in
@instrumental0. I suggested that these are devices for
expressing certain arguments as direct objects which
would otherwise be cast as obliques (i.e., datives and
instruments). There are several pieces of evidence
which support this analysis. First, when the referent
of the direct object i s plural , just l i k e other direct
objects it can be marked by the p l u r a l d i r e c t object
su f f ix -mn. Second, when both the dat ive o r
instrumental direct object and the pa t i en t are present
as full NPs i n the clause, the direct object tends to be
expressed before t h e pat ient . Finally, t h e d i r e c t
object seems to represent a par t i c ipan t t h a t is more
cen t ra l t o the event described by the verb and/or more
prominent in the discourse than the pat ient . Thus the
morphological marking, the syn tac t i c pat terning , as well as the semantic or" discourse status a l l seem t o ind ica te
that these arguments are now d i r e c t objects, and t h i s
further suggests t h a t core argumenthood plays an
important role i n the grammatical s t ruc tu re of
Barbareno.
Finally, we briefly examined the grammatical s t a t u s
of the pa t i en t occurring with a verb marked by one of
the core argument-building devices. I suggested t h a t
this type of p a t i e n t does n o t have a l l t h e fea tures
associated with d i r e c t objects.
I n chapter 4, we examined r e f e r e n t i a l choice i n
Barbareno. We first looked at pronominal affixes, which
are the most typical r e f e r e n t i a l forms i n discourse. As
we saw earlier, these affixes have been grannnaticized to
such an extent that subject prefixes are used even when
they do not refer (duanny subject). Thus both frequency
and grammar suggest that pronominal affixes may be the
unmarked or default referential choice for B a r b a d o
speakers.
Pronominal affixes indicate only person, number,
and case, so they seem to be used only when that much
information is enough to keep track of the referent:
more explicit reference is not necessary. That is,
these affixes are used to talk about participants who
have already been fully established in the discourse
(8given information8 in Chafe's term (1976, 1987,
1994) ).
These pronominal affixes often co-occur with full
MPS. I showed that full NPs are used in such s i t u a t i o n s
a s when referents are introduced in the discourse for
the first time or re-introduced after a thematic break
or perspective change, and when there are other
semantically similar referents in the discourse. That
is, these are situations in which reference only by
pronominal affixes would not be explicit enough.
After establishing the uses of these two
r e f e r e n t i a l forms i n discourse, w e looked a t t w o other
fo rm2 independent pronouns and t h e d m o n s t s a t i v e be2.
I suggested t h a t these two forms are not used ve ry o f ten
and t h a t they are associa ted wi th spec ia l i zed funct ions
as w e l l as particular s y n t a c t i c const ruct ions used t o
express c e r t a i n semantic content .
I n chap te r 5, w e looked at the cons t i tuen t o rde r of
Barbarefirt. Barbareiii-i seem t o al low severa l d i f f e r e n t
types of cons t i t uen t orders ; however t r a n s i t i v e verbs
with t h e two core arguments expressed both as f u l l HPs
are extremely rare i n d i scourse (cf. Du Bois 1987) . I
thus suggested t h a t it may no t be use fu l o r even
possible t o t a l k about c o n s t i t u e n t order i n the
t r a d i t i o n a l sense.
W e t hen looked a t the most t y p i c a l cons t i t uen t
order i n discourse: v e r b - i n i t i a l order. I suggested
t h a t i n t h a t order , the verb i s genera l ly followed by a t
most one core argument f u l l NP. As w e saw i n chap te r 4 ,
f u l l WPs are used when t h e r e f e r e n t needs t o be
e x p l i c i t l y mentioned fox va r ious reasons. I thus
suggested t h a t funct ional ly t h e basic cons t i t uen t order
i n Barbareno is a verb represen t ing e i t h e r an even t o r
a state which may be followed by one core argument f u l l
NP performing such cognitively demanding t a s k s as
introducing and re-introducing par t ic ipants and clearly
dis t inguish ing some par t i c ipan t s f r o m o ther pa r t i c ipan t s
i n t h e discourse.
Studies have shown t h a t transitive d i r e c t ob jec t s
and i n t r a n s i t i v e subjects are t h e grammatical slots i n
which new information is o f t en presented (e.g., Du Bois
1987) . I a l s o found that i n Barbareiio postverbal NPs
are mostly transitive direct ob jec t s and i n t r a n s i t i v e
subjects . I thus suggested t h a t s t r u c t u r a l l y the basic
cons t i tuen t order i n Barbareno is a verb perhaps
followed by e i t h e r t h e d i r e c t ob jec t or t h e i n t r a n s i t i v e
subject.
Regarding obliques, we found them appearing after
the core arguments (i.e., t r a n s i t i v e d i r e c t ob jec t s and
i n t r a n s i t i v e sub jec t s ) . As discussed i n chap te r 3,
Barbareno has a process in which certain arguments which
would otherwise be obliques (i.e., datives and
instruments) become direct ob jec t s . When this happens
and both t h e direct object and the pat ien t are expressed
as f u l l MPS, t h e d i r e c t ob jec t t y p i c a l l y appears right
a f t e r t h e verb and is followed by the p a t i e n t ,
r e f l e c t i n g the s t a t u s change.
Finally, we looked at constituent orders in which
full Nfs appear before the verb. Two different types
were discussed. The first type is characterized with
preverbal NPa appearing with a clitic 2i= on the
following word. These preverbal NPs are not strictly
syntactic arguments of the verb. That is, in this type,
preverbal NPs serve a marked function which can be
characterized as the introduction of a topic. Preverbal
NPs and complete predications are conjoined because
there is some type of pragmatic relationship between
them. This particular type of preverbal order is not
uncommon, but it is not as frequently used as the basic
verb initial order. The second type of preverbal order
is characterized with another clitic, ba=, appearing an
the following word. This is a focus construction in
which the preverbal Nfs are the focused elements, and it
is used infrequently in discourse. I suggested that
these preverbal MP8 are not syntactic arguments of w h a t
follows them.
The significance o f this dissertation may be
summarized as the following. First, in spite of the
l a r g e amount of data left by earlier researchers,
Barbareno Chumash has been a little-studied and thus
little-known language. The present dissertation is the
first study on this scale to describe the grammar of
Barbareno Chumash. Particularly because of the rich
morphology of Barbarefio, it provides the field with a
solid database which should help to generate new ideas
and to evaluate already existing theories of language.
Further this dissertation has demonstrated that the
grammar of Barbaxeno Chumash is heavily dependent on the
functional factor's exhibited in discourse. It has thus
firmly established the functional basis of linguistic
structure and has amplified the growing literature in
functional linguistics. Specifically, this was done
with respect to the 'information flow' theory of Chafe
( 1987, 1994 1.
Finally, the investigation for this dissertation
was carried out solely based on the examination of
discourse data. For this reason, the study has also
shown the importance of natural discourse data in the
study of the structure and the functions of a language.
The limitations of this study can be summarized as
the following. First, the research for this
dissertation was executed without consultation with the
specialists of the language, native Barbareno speakers.
T h i s w a s p a r t i c u l a r l y needed t o check t h e formal aspects
of t h e language. Since Barbareno is no longer spoken,
of course t h i s w a s not possible. I only hope t h a t w h a t
I have presented above comes as c lose a s poss ib l e t o
what w a s a c t u a l l y going on i n the language. Second,
related t o t h e first l imi ta t ion , s ince this d i s s e r t a t i o n
was written based solely on data co l lec ted by e a r l i e r
researchers, there were often s i t u a t i o n s i n which it was
d i f f i c u l t to determine what t h e o r ig ina l recorders
intended, be it an unreadable phonetic symbol o r several
d i f f e ren t fonns and/or mult ip le t r ans l a t ions apparently
fo r t h e same i t e m . Harrington of ten made it clear what
each d i f f e r e n t form and translation w a s f o r , bu t there
are still unclear cases which are left for researchers
t o decide. For instance, I o f t e n had to pick one form
or one t r a n s l a t i o n out of several with similar meanings
given by Harrington, since it seemed impract ical t o list
a l l of them. Though I always did my bes t t o represent
what 1 took to be intended by the or ig ina l recorders,
and i f I had t o choose, I chose what appeared t o m e to
be the best a l t e r n a t i v e , some indeterminacy will always
remain i n this kind of research. I only hope my
decisions have not been too far" from the i n t e n t i o n s of
t h e or ig ina l recorders or of the speakers. Third, the
data that I used to examine Barbare60 grammar are mostly
d i f fe ren t types of narratives. For t h i s reason, the
r e su l t s which I have obtained may pertain only to t h i s
genre. Further investigations on other types of data
will be necessary to see how generalizable my r e s u l t s
are. Final ly , t h i s study lacks quantitative
verif icat ion. Same of the f indings and the claims made
i n t h i s d i s s e r t a t i o n would benef i t from quantitative
ver i f ica t ion . These l imi ta t ions can be at least
p a r t i a l l y overcome i n future s tudies .
Overall, despi te t h e l imi ta t ions s tated above, some
interesting results have been obtained. The grammar of
Barbareiio Chumash has been shown t o be heavily dependent
on functional f ac to r s exhibited i n discourse, and each
d i f f e ren t aspect of t h e grammar examined here appears t o
be responsive to s p e c i f i c functional fac tors , a strong
confirmation of the value of a function-based approach
t o grammar.
Bibliography
Applegate, Richard B. 1972. Ineseiio Chumash Grammar. Ph-D. Dissertation, University of California, Berkeley .
Applegate, Richard B. 1976. Reduplication in Chumash. In Margaret Langdon & Shirley Silver, Eds., Hokan Studies: Papers f r o m the First Conference on Hokan Languages, 271-283. The Hague: Mouton.
Austin, P e t e r . To appear. Causatives and Applicatives in Australian Aboriginal Languages. In Peter Austin & B a r r y Blake, E d s . , Theoretical Iiinquistics and Australian Aboriginal Languages.
Beeler, Madison S. 1954-1961. Manuscript Materials and Audio Recordings of Barbareiio Chumash.
Beeler, Madison S. 1970a. Topics in Barbareno Chumash Grammar. Unpublished Manuscript.
Beeler, Madison S. l970b. Sibilant Harmony in Chumash. International Journal of American Linguistics 36:14-17.
Beeler, Madison S. 1976. Barbare50 Churnash Grannnar: A Farrago. In Margaret Langdon & Shirley Silver, E d s . , Hokan Studies: Papers from the First Conference on Bokan Languages, 251-269. The Hague: MOU~OII.
Beeler, Madison S. 1978. Barbareiia Chumash Text. and Lexicon. In Mohaannad A. Jazayery, Edgar C. Polom6, 6 Werner Winter, Eds., Linguistic and Literary Studies in Honor of Archibald A. Hill, 11: Descriptive Linguistics, 171-193. The Hague: Mouton .
Beeler, Madison S. and Whistler, Kenneth W. 1980. Coyote, Hawk, Raven and Skunk (Barbareno Chumash). In Martha B. Kendall, Ed., Coyote Stories 11, 88- 96. International Journal of American Linguistics Native American Texts Series Monograph No. 6. Chicago: University of Chicago Press.
Blake, B u r y J. 1994. Case. Cambridge: Cambridge University P r e s s .
Chafe, Wallace L. 1976. Givenness, Contrastiveness, Definiteness, Subjects, Topics, and Point of View. Charles N. Li, Ed., Subject and Topic, 25-55. New York: Academic Press.
Chafe, Wallace L. (Ed.). 1980. The Pear Stories: Cognitive, Cultural, and Linguistic Aspects of Narrative Production. Norwood, NJ: Ablex.
Chafe, Wallace L. 1987. Cognitive Constraints on Information Flow. Russell S. Tomlin, Ed., Coherence and Grounding in Discourse, 21-51. Amsterdam: John Ben jamins.
Chafe, Wallace L. 1994. Discourse, Consciousness, and Time: The Flow and Displacement of Conscious Experience in Speaking and Writing. Chicago: Chicago University Press.
Clancy, Patricia M. 1980. Referential Choice in English and Japanese Narrative Discourse. In Wallace L. Chafe, Ed., The Pear Stories: Cognitive, Cultural, and Linguistic Aspects of Narrative Production, 127-202. Norwood, NJ: Ablex.
Comrie, Bernard. 1989. Language Oniversals and Linguistic Typology, Second Edition. Chicago: University of Chicago Press.
Croft , William. 199 1. Syntactic Categories and Grammatical Relations: The Cognitive Organization of Information. Chicago: University of Chicago Press.
Gumming, Susanna. 1991. Functional Change: The Case of Malay Constituent Order. Berlin: Mouton de Gruyter.
Dixon, R. M. W. 1977. Where Have All the Adjectives Gone? Studies in Language 1:19-80.
Dryer, Matthew S. 1986. Primary Objects, Secondary Objects, and Antidative. Language 62:808-845.
Du Bois, John W. 1987. The Discourse Basis of Ergat iv i ty . Language 63:805-855.
Du Bois , John W. , Schuetze-Coburn, Stephan. , Cununing, Susanna., and Paolino, Danae. 1993. Out l ine of Discourse Transcript ion. I n Jane A. Edwards & Martin D. Lamport, Eds., Talking Data: Transcript ion and Coding i n Discourse Research, 45- 89. Hillsdale, NJ: Lawrence Erlbaum.
Duranti, Alessandro and Ochs , Elinor. 1990. Genit ive Constructions and Agency i n Samoan Discourse. Studies in Language 14:l-23.
Fox, Barbara A. 1987. Discourse S t ruc ture and Anaphora: Written and Conversational English. Cambridge: Cambridge University Press.
POX* Barbara A. (Ed.). 1996. Studies i n Anaphora. Amsterdam: John Benjamins.
Given, T. (Ed.). 1983. Topic Continuity i n Discourse: A Quan t i t a t i ve Cross-Language Study. Amsterdam: John Ben j amins.
Given, T V 1984 and 1990. Syntax: A Functional- Typological Introduction, Vol. 1-11. Amsterdam: John Ben jamins.
Greenberg, Joseph H. 1966. Some Universals of Grammar with Pa r t i cu l a r Reference t o t h e Order of Meaningful Elements. I n Joseph H. Greenberg, Ed., Universals of Language, Second Edi t ion, 73-113. Cambridge, HA: MIT Press.
Harrington, John P. 1912-1961. Manuscript Materials of Linguistic and Ethnographic Work among t h e Chumash, Kraus Microfilm Edit ion, V o l . 3. Washington, D.C.: National Anthropological Archives, Smithsonian I n s t i t u t i o n .
Harrington, John P. 1974. S i b i l a n t s i n Venturefio. In te rna t iona l Journal of American L ingu i s t i c s 4 0 : 1- 9.
Hawkins I John A. l9 8 3. Word Order Universals : Quantitative Analyses of Linguistic Structure. New York: A c a d d c Press.
Hopperf Paul J. and 7!hompsanf Smdra A* 1980. Txansitivity in Grammar and Discourse. Language 56~251-299.
Hopper? Paul J. and Traugott Elizabeth C* 1993. Grmaticalization. C&ridge:CdridgeUniversity Press.
Ichihashi-Nakayama, Kumiko. 1996. The wApplicatAve* in Bualapai: Its Functions and Meanings. Cagnitive Linguistics 7~227-239.
Khenyi, Alexandre. 1980. A Relational Grammar of Kinyaxwanda. University of California Publications in Linguistics? Vol. 91. Berkeley: University of California Press.
KlarI Kathryn A. 1973. Northern Chumash and the Subgrouping of the Chumash Dialects. M.A. Thesis, Uniweristy of Californiaf Berkeley.
Klar, Kathryn A. 1977. Topics in Eistorical Chumash Grammar. Ph.D. Dissertation, University of California, Berkeley.
Klarf Kathryn A. 1981. Proto-Chumash Person and Number Markers. In Jmes Redden? Ed., Proceedings of the 1980 Hokan Languages Workshop. Southern Illinois Universityf Department of Linguistics Occasional Papers on Linquistcs 9:06-95.
Mithunf Marianne. 1987 . 1s Basic Word Order Universal? In Russell S. T011tlhI Ed. I Caherence and Grounding in Discourse, 281-328. Amsterdam: John Benjamins.
Mithunp U i a n n e . 1993. Prosodic D e t e d n a n t s of Syntac t ic Form: Central Pomo Const i tuent Ordera I n David A. Petersonf Ed.,, Proceedings of the Nineteenth Annual Meeting of t h e Berkeley Linguis t ics Society, F e b m q 12-lS, 1993 : Spec ia l Session on Syntac t ic Issues i n Native American Languages, 86-106. Berkeley: Berkeley Linguis t ics Society.
Mithun, Mariamee 1994. The m p l i c a t i a n s of E r g a t i v i t y fox a Phi l ipp ine Voice System In Baxbara Fox & Paul J. Hopper, Eds., Voice: Farm and Function, 247-277. Amsterdam: John Benjamins.
Mithun, Marimne* To Appear. The Regression of S i b i l a n t Harmony through t h e L i f e of Barbarefio Chumash. I n Jane H i l l & Zyle Campbell, Eds* , a secret f e s t s e h r i f t
Mithun, Mariannemf Ono, Tsuyashie, and Wash, Suzanne. 1994. The S h i f t i n g Sta tus of Initial G l o t t a l Stop i n Barbare50 Chumash. I n Hargaret Gangdon, Ed., Proceedings o f t h e Meeting of t h e Society f a r t h e Study af the Indigenous Languages of the Americas July 2-4 1993 and t h e Hokm-Penutian Workshop July 3 , 1993, 199-207. Survey of Ca l i fo rn i a and Other Indian Languages Report 8. Bezkeley: Department of Linguis t ics , Universi ty of Ca l i forn ia , Berkeley.
Hunro, Pamela. 1982. On t h e T r m s i t i v i t y of rSayF Verbs. In Paul J. Eeppez & Sandra a. Thompson, E d s : , Syntax and Semantics, Vole 15: Studies zn Trans i t i v i ty , 301-318. New York: Academic Press.
Ono, Tsuyoshi., Wash, Suzanne., and f i t h u n , Marianne. 1994a . Fina l G lo t t a l i za t ion i n Barbare60 Chumash and Its Heiqhbars. In Margaxet &angdon, Ed., Proceedings of t h e Meeting of t h e Society f o r t h e Study of t h e Indiqenoua Languages of t h e Americas J u l y 2-4 1993 and t h e Hokan-Penutian Workshop J u l y 3, 1993, 208-217. Survey of Cal i fo rn i a and Other Indian Languages Report 8 . Bexkeley: Department of L i n ~ i s t i c s , Universi ty of Ca l i fo rn i ap Berkeley.
Ono F Tsuyoshi *, Wash, Suzame*, and Mithun, Marimne 1994be The M o q h o a p t a c t i c and Pragmatic C a r ~ e l a t e s of t h e I n s t m e n t a l Suff ix i n Barbauefio Chumash* Paper presented at t h e Hoka-Penutian Workshop, Univezsity of Oregon*
P a h e r , F. R. 1994- G r a m m a t i c a l Roles and Relations. Cambridge: Cambridge University Presse
Paynet Doris L. 1990- The Pragmatics of Word Order: !Pypological Dimensions of Verb I n i t i a l Languages. Berlin: Mouton de Gruyter*
Poserf W i l l i a m * 1982m Phonological Representat ion and Action-at-a-Distmce. I n Harq van der Hulst & Nomal Smithf Eds.@ The Structure of P h ~ n o l o g i c a l Representations, P a r t IIf 121-158* Dordrecht: Faris *
Schachter, Paula 1985. Pes-of -Speech Sys tems . 1x1 Timothy Shopen, Ed*, Language Typology and Syntactic Deecription, V d m 1: Clause Struc ture , 3- 6 L Cambridge: Cambridge University Press.
Shibatani, Masayashi. 1996 . Applicat ives and Benefactives: A Cognitive Accounta In Masayoshi Shibatani & Sandra A. Thompson, Zds . G r a m m a t i c a l Constmctions: Their Form and Meaning, 157-194* Oxford: Oxford University Press*
TamlinR Russell S. 1987 m Linguis t ic Ref lec t ions of Cognitive Events Russell S T o d i n , Ed . Coherence and Grounding i n DiscourseR 4S5-47ge Amstercim: John Benjamins
Thompsonr Sandra Am 1990. Information F l o w and Dative S h i f t i n English Diecouxse* I n Jerold A* E&ondsonF Crawf ord Feagin? & Peter MGhlhgualer, Eds , Develapsnent and Diversityz Lingu i s t i c Variat ion across T h e and Space* A Fes t sch r i f t f o r Charles- Jmes N. Bailey, 239-253 * Dallas: Summer I n s t i t u t e of L inguis t icse
Thompson, Sandra A. Ta appear* Discourse Motivations far the Core-Oblique Distinctian a8 a Language UnivexsaL In Akio K d o , Ed., ~nctionalism in Linguistics. Berlin: Mouton de Gru*er.
Txaugott, Zlizabeth C. and seine, Bern& (Ed8.)* 1991e Approaches to G&-ticalization, Vol. I-XI* hstexdaxn: John Ben jamins
T ~ g g y , 0 a ~ i . d ~ 1988 . H&huatl Cau~ative/Applicatives in Cognitive Grammar In Brygida Rudzka-Ostp, Ed * , Topics in Cognitive Linguistics, 587-618. Amsterdam: John Benjamins.
Wash, Suzanne M. 199Sae Productive Reduplication in Barbare60 Chumash. M . I L Thesis, University of California, Santa Barbaxa.
Waeh, Suzmne M. 1995b- Dependency and Constituency in Barbare50 Chumashz A Laok at the Connectives hi- and Zi-* Paper presented at Native hesican Indian Languages Meeting, Department of Linguistics, University of California, Santa Barbara.
Whistler, Kenneth We 1980e An Interim B a x b a r e f i o Churnash Dictionary. Unpublished Manuscript*