Manifestations of Ergativity in Quiché Grammar

440 62 17MB

English Pages [578] Year 1988

Report DMCA / Copyright

DOWNLOAD FILE

Polecaj historie

Manifestations of Ergativity in Quiché Grammar

Citation preview

UC Berkeley Dissertations, Department of Linguistics Title Manifestations of Ergativity in Quiché Grammar

Permalink https://escholarship.org/uc/item/22p292kr

Author Larsen, Thomas

Publication Date 1988-01-01

eScholarship.org

Powered by the California Digital Library University of California

Manifestations of Ergativity in Quichd Grammar

By Thomas Walter Larsen B.S. (University of Colorado) 1969 M.A. (University of Colorado) 1973 C.Phil. (University of California) 1983

DISSERTATION Submitted in partial satisfaction of the requirements for the degree of DOCTOR OF PHILOSOPHY in Linguistics

in the GRADUATE DIVISION OF THE UNIVERSITY OF CALIFORNIA, BERKELEY

Date

Chairman

DOCTORAL DEGREE CONFERRED .....

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Manifestations of Ergativity in Quich£ Grammar Copyright c 1988 Thomas Walter Larsen

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

i

Acknowledgements There direct

are so many people that I wish to thank for

and

indirect

help they have given me

that

the

it

is

difficult to know where to begin.

Prominent among these are

the

Wallace

members

Fillmore,

of

my

committee:

and Lily Wong Fillmore.

Chafe,

Charles

They have helped me not

only in the preparation of this dissertation but also in

my

overall development as a linguist,

and they have been

ever

kind in putting up with my whims.

There are so many

other

ways

in

which

I

am also indebted to

them

that

I

will

probably never be able to fully repay their kindness. I

am

especially indebted to

the

numerous

Quichd

speakers who have helped me over the years in learning about their

language.

Obviously this dissertation

would

never

have been possible without their assistance.

These include

the

with

Quichd-speaking

Proyecto

linguists that I worked

Linguistico

Guatemala:

Francisco

Miguel Bulux Vicente,

Ixchop Poroj,

Miguel Lux Carrillo,

Gaspar Pu Tzunux,

Marroquin

at

the

(PLFM)

in

Juan Cac Tzunux,

Ernesto

Jacobo Baltazar Pu Lux,

and Pedro Sanic Chanchavac.

These seven

people have taught me most of what I know (or think I about the Quichd language. at

the

Ajpacajd,

PLFM

have also

Manuel

Chox,

know)

Other Quichd-speaking linguists been

of

Diego

Guarchaj, and Lucas Tepaz Raxuleu.

great

help:

Guarchaj,

Florentino

Juan

Rodrigo

During the year in which

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

ii

I taught at the University o£ Iowa, of

I had the good

meeting and working with Luis Enrique

Sam

fortune

Colop.

His

colaboration has been crucial in the formation of much of my thinking

in

California,

this I

dissertation.

have

this

my

also had the pleasure of

working with Martin Poz P6zez. eventually

During

time

in

knowing

and

He provided the data

which

led to some of the major conclusions reached

work,

and

lifesaving.

his

friendship has

at

times

I also wish to thank the PLFM,

even

in been

especially the

past and present directors Narciso Cojti and Martin Chacach, for

allowing me to participate in the unique program

which

permitted me to learn so much about Mayan languages. Among the other linguists who have helped me in one way or

another,

Kaufman,

who

everyone

else

Norman, field

special

mention

must

be

made

of

has taught me much of what I know knows) about Mayan

languages,

Terrence (and

and

what

William

who had already blazed a rather large trail in of

Quichean

entering

it.

include

Allen

Nichols,

linguistics before I

Others

who

Bell,

Leanne

even

have helped me Hinton,

Schein, Donca Steriade, and Alan Taylor.

in

Paul

David Rood (my first linguistics

thought

the of

my

studies

Kay,

Johanna

teacher),

Barry

In addition I have

spent many enjoyable and productive years discussing Quich&, Mayan languages, linguistics, and numerous other topics with Glenn Ayres, Judie

John Dayley,

Maxwell.

Jack Du Bois,

Nora England, and

I have also benefitted from^^onver sat ions

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

ii i with

Parcell Ackerman,

Campbell,

Colette

Margaret Datz, Quizar,

Craig,

Amy

Knud Lambrecht,

Jim Mondloch,

Smith-Stark,

Judith Aissen,

Linda

Dahlstrom,

Brown, Karen

Martha Maori,

Linda Munson,

and Stephen Stewart.

Lyle Dakin,

Robin

Cathy O'Connor,

Mader Thom

It is unlikely that all

of these people will approve of what I have done

here,

but

their help and friendship is much appreciated nevertheless. I wish to thank my parents, G.

Larsen,

not least,

Walter H. Larsen and Doris

for supporting me in my studies.

And last, but

I thank Julia, Oscar, and Adelina for putting up

with so much while I worked on this dissertation. this

is

over,

they can all breathe a sigh of

Now that relief

least until the next thing comes along).

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

(at

iv TABLE OP CONTENTS Acknowledgements

i

1.

Introduction

1

2.

Phonology

2.1

^

Phonemes and Orthography

11 11

2.1.1

Vowel Phonemes

11

2.1.2

The Phoneme /h/

16

2.1.3

The Prevelar Phonemes

26

2.1.4

The Phoneme /q'/

33

Phonological Rules

34

2.2 2.2.1

Underlying PhonologicalRepresentations

34

2.2.2

Phonetic Representations

44

2.3

Other Orthographies

57

Notes

81

3.

97

Morphology I

3.1

Morphological Processes

97

3.2

Root and Stem Classes

97

3.3

Personal Pronouns

3.4

Nouns

3.4.1

Inflection

100 101 101

3.4.1.1

Possessor Agreement

101

3.4.1.2

Number

103

3.4.1.3

Subject Agreement

105

3.4.2

Noun Classes

108

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

V

3.4.2.1

Class 1

1GS

3.4.2.2

Class la

109

3.4.2.3

Class 2

110

3.4.2.4

Class 3

113

3.4.2.5

Class 4

117

3.4.2.6

Class Y

118

3.4.2.7

Class Z

119

3.4.3

Complex Nouns

* 119

3.4.4

Noun Subtypes

125

3.4.4.1

Enumeratives

125

3.4.4.2

Relational Nouns

127

3.4.5 3.5

Derivation

Adjectives

3.5.1

Inflection

130 133 134

3.5.1.1

The Attributive and Degree Suffixes

134

3.5.1.2

Subject Agreement

136

3.5.1.3

Number Agreement

137

3.5.2

Derivation

141

Notes

145

4. Morphology II

152

4.1

Verbs

152

4.2

Intransitive Verbs

152

4.2.1

Inflection

153

4.2.1.1

Subject Agreement

153

4.2.1.2

Tense/Aspect/Mood

161

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Vi 4.2.1.3

Phrase-FinalSuffix

176

4.2.1.4

IncorporatedMovement

180

4.2.2

Derivation

184

Notes

197

5.

211

Morphology III

5.1

Transitive Verbs

211

5.2

Inflection

212

5.2.1

Subject and Object Agreement

212

5.2.2

Tense/Aspect/Mood

223

5.2.3

Phrase Final and Other Status Suffixes'

229

5.2.4

Incorporated Movement

243

5.3

Derivation

247

Notes

274

6.

Morphology IV

288

6.1

Positionals

288

6.2

Adverbs andParticles

302

6.2.1

Directionals

302

6.2.2

Demonstratives and Articles

309

6.2.3

Prepositions

315

6.2.4

Interrogatives

321

Notes

326

7. Syntax I: Word Order

327

Notes

384

8. Syntax II

389

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

vil 8.1

Complex Sentences

8.1.1

Complement Clauses

389 389

8.1.1.1

Pinite Complements

389

8.1.1.2

Non-finite Complements

395

8.1.2

Purpose Clauses

415

8.1.3

Temporal Clauses

424

8.1.4

Conditional Clauses

429

8.1.5

Other Types of Subordinate Clauses

431

8.1.5.1

Reason Clauses

431

8.1.5.2

Manner Adverbial Clauses

432

8.2

Negation

433

8.3

Reflexives and Reciprocals

441

8.4

Voice

459

8.4.1

Passive

460

8.4.2

Antipassive

467

8.4.3

Instrumental Voice

472

Notes

490

9.

497

Syntax III: WH-movement

Notes 10.

536 Conclusion: Ergativlty inQuichd Syntax

540

Notes

555

References

558

1

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

1

CHAPTER 1 INTRODUCTION This work is about Quichd,

a Mayan language spoken

approximately

900,000 persons (Grimes 1984:59,

highlands

Guatemala.

of

This

language is

61) in part

of

by the the

Eastern Mayan Division of the Mayan Family (see table 1; cf. Kaufman

1976a:11-13

and Kaufman 1976b:85) and

related to the neighboring languages Sacapultec, of

and Sipacapa.

Cakchiquel,

spoken Rabinal.

in

1966)

distinguish

the towns of San Some

linguists

closely Tzutujil,

Some linguists, especially those

the Summer Institute of Linguistics

Neuenswander

is

(e.

another

Miguel

Chicaj,

However,

Shaw

language,

also include the

Andrds SajcabajA and Joyabaj.

g.,

Achi,

Cubulco, towns

and

of

and San

Campbell (1977:31)

argues that there is no strong reason to distinguish Achi as a language separate from Quichd. all

of

Sipacapa

Kaufman (1976a:59) classes

these as dialects of Quichd.1 languages,

The

spoken in the towns of

Sacapultec

and

Sacapulas

and

Sipacapa, respectively, have also been considered by some to be dialects of Quichd; they

however, Kaufman (1976b) argues that

are separate languages (see also

Campbell

1977:15-20

and Du Bois 1981). At the time of the arrival of the Spanish in 1524, Quichds

controlled a rather extensive tribute state in

Guatemalan highlands,

the the

which was destroyed that same year by

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

2

TABLE 1 The Mayan Family I. Huastecan Group 1. Huastec 2. Chicomuceltec (Coxoj) II. Mayan Group 3. Yucatec

16.

Jacaltec

17. b) Cotoque (Moch6)

Eastern Rivision V. Greater Mamean Branch A. Mamean Proper Group

4. Lacand6n

18. Teco

5. ItzA

19. Mam

6. Mop&n

B. Ixilan Group

Western Blxlaian

20. Aguacatec

III. Greater Tzeltalan Branch

21. Ixil

A. Cholan

Group

V I . Greater Quichean Branch

7. Chorti

A. 22. Uspantec

8. Choi

B. Quichean Proper Group

9. Chontal B. Tzeltalan Proper Group

23. Sipacapa 24. Sacapultec

10. Tzotzil

25. Quichd

11. Tzeltal

26. Tzutujil

IV. Greater Kanjobalan Branch A. Chujean Group

27. Cakchiquel C. Pocom Group

12. Tojolabal

28. Pocomam

13. Chuj

29. Pocomchi

B. Kanjobalan Proper Group

D. 30. Kekchi

14. a) Kanjobal 15.

Acatec

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

3

the invaders. be

To this day, however, the Quiches continue to

one of the largest and most important ethnic

Guatemala.

groups

in

Their language is the language of the Popol Vuh

(Poopol WuuJ or Poop(o) WuuJ), one of the greatest monuments of

aboriginal American

other

things,

mythical

the

literature,

which

relates,

history of the Quichd people

among

from

the

creation of the world to a time shortly after

the

Spanish conquest. Soon begun

after the conquest linguistic work on Quichd

by Spanish priests who were interested in

language

in

using

the

Quichd

to

One of the most important early grammars

of

their

Christianity.

attempts

to

convert

Quichd was the Arte de las tres lenguas;

the

Cakchiquel, Quiche

y Tzutuhil written by Francisco Ximdnez in 1734. portions by

was

The Quichd

of thi3 grammar were eventually published in

the French priest Charles Etienne Brasseur de

1862

Bourbourg

under his own name with added commentary in French (Brasseur 1862). In more recent times there has continued to be interest in

the

language

published

work

by

linguists

on Quichd

things,

Alvarado

Ch&vez

(no date),

and

grammar

Ldpez (1975), Edmonson

others. includes,

Carbonell

(1967),

Fox

The

recent

among

other

Pastor

(1973),

(1965),

Kaufman

(1986a), Mondloch (1978a, 1978b, 1981), Norman (1978, 1980), Sam

Colop

(to appear),

Schultze

Jena

(1933),

Wick

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

and

4

Cochojil-Gonzdlez (1966-9).

The quality of this work

the gamut from excellent to abysmal. Ldpez,

Chdvez,

The work of

runs

Alvarado

and Sam Colop is interesting because it has

been written by native speakers; and Sam Colop's grammatical sketch is actually quite good. one

of

Mondloch (1978a) is probably

the best modern works on

pedagogical

in

approach

Quichd;

(as are Fox

however,

1965

and

it

Wick

is and

Cochojil-Gonz&lez 1966-9) and rather limited in scope (as is all

of the work listed above).

thorough

Mondloch (1981) is a

very

treatment of voice in Quichd and also includes

extensive

treatment

however,

vowel

of verb inflection.

length

does

not

seem

In to

some be

an

cases,

indicated

perfectly although it is certainly better than in most other works. from

Kaufman (1986a), is also quite good, treating Quichd a comparative Mayan point of view.

Mondloch

(1978b)

and the two articles by Norman are excellent,

but they

are

all

grammar.

In

short papers on limited topics in Quichd

spite

of all of the linguistic work that has been

Quichd

in

recent

comprehensive

times,

reference

there has grammar of

not the

yet

done

on

appeared

language

a

written

according to modern linguistic principles. The present work is also not a comprehensive grammar;

however,

it

is

a

rather

large

step

reference in

that

direction in that I attempt to cover a relatively wide range \ of topics as accurately as I am able. The principal purpose of this work is to investigate the phenomenon of

ergativity

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

5

and

the

various ways in which it is manifested

grammar.

Quichd

morphologically

is

ergative

ergative/absolutive

the

ways

discourse

in

of

as

virtue

system,

Quichd a

of

its

which

is

In Chapter 7 there is a discussion

which ergativity

the

by

agreement

pragmatics of Quichd.

discussion

recognized

language

verb

discussed in Chapter 5. of

generally

in

is

manifested

the

In Chapter 10 there is

question of whether or

syntactically ergative.

in

not

Quichd

a is

Some of the more recent work on the

nature of syntactic ergativity has been conducted within the framework

of

derivatives

Chomsky's

Government-Binding

of that theory (see,

Marantz 1984).

e.

g.,

theory

Levin

and

1987

and

With this in mind, I present an analysis of

some of the features of Quichd syntax within this

framework

m

in

order

does

to determine the extent to which Quichd

not

conform

results

of

interesting

this

to such a

investigation

theoretically.

fact syntactically ergative, been

recognized

manifested intransitive

in

theory

in

turn

the

verbs.

VP

out

to

The

be

quite is

in

but in a manner which has

syntax

not

studies:

ergativity

is

in

"subjects"

of

that

verbs are not dominated by the S node

the subjects of transitive verbs, by

or

ergativity.

I conclude that Quichd

previous

Quichd

of

does

node as are the direct

but rather are objects

of

as

are

dominated transitive

Since intransitive "subjects" and transitive direct

objects are treated alike sytactically, and both are treated

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

6

differently

from

transitive

ergative pattern.

subjects,

this

an

This pattern is shown to have some rather

interesting consequences in Quichd syntax. concluded

defines

Thus, it can be

that ergative phenomena in Quichd are

manifested

in morphology, syntax, and discourse.

For

The

language

treated in this work is

the

most part I do not deal with older

modern

Quichd. of

the

language as preserved in writings such as the Popol Vuh

and

the

older linguistic work.

variety of dialects, on

forms

Modern Quichd is spoken

in

and most of the modern published works

Quichd deal with a particular dialect and

present

that

whole.

One

dialect as representative of the language as a

difference between this work and many of the other works Quichd

is

a

that I try to give some idea

of

the

on

dialectal

diversity of the language. Most of my work on Quichd was done while I was as

an asesoz tdcnico at the Proyecto

Marroquin in Guatemala. group

of

Santa

Maria Chiquimula,

directing

investigating

from

teaching them

in the compilation of

the

basic a

and

linguistics,

dictionary,

language.

and the

linguistics courses that I taught I also had a student

from

speaking

grammar

of

Momostenango

a

In

Nahuald.

the

Francisco

From 1976 to 1978 I worked with

Quichd speaking students

them

Linuistico

working

In addition I had contact with some other linguists

at

the PLFM

from

Nahuald

and

Quichd Santa

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

7

Catarina Ixtahuac&n. California, with

a

while I was in

I had the good fortune of being able to consult

speaker

country.

In more recent years,

from Zunil who is

While

now

resident

in

teaching at the University of Iowa

this

during

the 1985-86 academic year, I had the further good fortune of being

able

to consult with a speaker from Cantel

studying linguistics at the University. Quichd well

I incorporate data from all of as

further

representing Ixtahuac&n,

data

the

from

dialects

Cantel,

of

the

who

was

In my discussion of these

dialects,

published

Nahuald,

Chichicastenango,„

as

literature

Santa

Catarina

San

Cristdbal

and

Totonicap&n among others. The and

differences between these dialects is

they

differences lexical. and

even

discussion extensive

are

all mutually

are

phonological

intelligible. and,

to

a

not Most

great, of

lesser

the

extent,

There are relatively few morphological differences fewer of

syntactic

differences.

Thus,

phonology in Chapter 2, there

is

in a

the

rather

coverage of the dialectal differences of which

am aware.

In the discussion of syntax in Chapters7-10, on

the

hand,

other

most of the examples given

are spelled

according to the pronunciation of the Momostenango which

is

chapters

the dialect I am most familiar dialectal

with.

differences are pointed out

they are known to exist and are relevant to the

I

dialect, In

those

only

when

discussion.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

8

Data

which

is cited from published

sources

is

generally

spelled according to the pronunciation of the dialect treated

in

respelled

the in

source

although

such

examples

being

will

the orthography that I have adopted

in

be this

work. A

number of Quiches have criticized certain

linguists

for what these Quiches consider to be an overemphasis of the dialectal

differences of Quichd and other Mayan

especially people

in

the creation of alphabets.

languages,

Some

of

feel that this is a deliberate attempt by

these

linguists

and others to keep them divided (the primary loyalty of many Quichds

is

to their home town rather than

to

the

Quichd

nation as a whole) and, thus, thwart their attempts to unite for the purpose of overcoming the oppression which they have suffered the

for over 460 years and continue to suffer from

present

discussion

time. of

I

want

to

make

it

clear

dialectal differences should in

at

that

my

way

be

no

construed as an attempt to undermine the unity of the Quichd people.

It

differences

should be clear from my discussion that are

relatively

minor and that

dialects are clearly varieties of a single is

not

to negate the fact,

however,

that

all

of

these

language.

This

the

dialectal

differences are interesting in their own right. that

the

language in

these

best way to demonstrate the unity of

I

believe

the

Quichd

is to try to present it as accurately as

all of its various forms (or in this case at

possible least,

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

as

9

many

of its forms as I am able) rather than presenting

one

particular variety and pretending that everything else

that

is Quichd simply does not exist. Chapter

2

presents

fairly

traditional

major

phonological

informally,

and

orthographies Chapters classes.

there

is have

treatment

extensive

derivational limited,

rules

are a

also

used

morphology

of

and on

of

to

the

write

inflectional complete. the

other

from of

presented

discussion

been

Some

contain a discussion of morphology

The

relatively

Quichd

phonemic point of view.

which

3-6

the phonology of

the

rather various Quichd. and

word

morphology Treatment

hand

is

a

is of

somewhat

for the most part presenting only some of the more

important derivational morphemes, especially those which are relevant to later sections.

Chapters 7 is a discussion

word order and its role in discourse. a

of

Chapters 8-9 contain

discussion of some of the major syntactic

constructions.

Most of the syntax is presented rather informally; there is a more formal treatment of WH-movement and

however, certain

other phenomena from the point of view of Government-Binding (GB) theory (Chomsky 1981, 1982).

Chapter 10 concludes with

a discussion of syntactic ergativity in Quichd.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

10

NOTES 1. Although

"Achi" dialects may not differ markedly

from

other Quichd dialects, it does seem to be the case that at

least

some people from the "Achi"

speaking

towns

consider themselves to be ethnically distinct from Quichd

people.

This division has apparently

since

pre-columbian times according to the

(see,

e.

people

g.,

are

Tedlock

1985:357),

mentioned as one of

the

existed

Popol

where the thirteen

the

Vuh

Rabinal tribes

allied with the Quichds and arriving with them from the east.

Part

of the Rabinal area was conquered by

Quichds in the fifteenth century.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

the

11

CHAPTER 2 PHONOLOGY 2.1. Phonemes and Orthography The

phonemes

symbols

of Quichd are shown

in

used in this chart are those of

Phonetic

Alphabet (IPA).

parentheses

in

table

the

The

International

Those phonemes which

table 1 are not found in all

1.

appear

in

dialects

and

will be discussed further below. Throughout Quichd

work,

be written in IPA,

phonemes

rather

of

written according to the practical orthography presented

in

This alphabet was developed by Terrence

and others (Kaufman the

but

the

be

2.

not

rest of this

will

table

will

the

work

Kaufman

1976a)1, and has been used primarily in

of the Proyecto Linguistico

Francisco

Marroquin

(PLFM) in Guatemala. 2.1.1

Vowel Phonemes

Some

dialects of

Quichd

(e.g.,

Momostenango,

Santa

Maria Chiquimula, Nahuald, Santa Catarina Ixtahuacdn, Zunil) have

ten

vowel phonemes:

five short vowels and

corresponding long vowels (see table 1). the

the

In such dialects,

short vowels are written as single vowels and the

vowels dialects

are (e.

written g.,

doubled (see

table

Cantel) have only six

five

2).

Some

vowels,

long other

phonemic

vowel length having been lost in all cases except on the low

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

12

TABLE 1 Phonemic Inventory Consonants:

voiceless stops

b i 1 a b i a 1

e o 1 a r

P

t

voiceless affricates

a 1 V

ts

9 lottalized stops

t

a 1

ts'

voiceless fricatives

s m

P r e

u

g l

V

V

V

o

e 1 a r

e 1 a r

u 1 a r

t t

a 1

(K)

k

q

?

(K’>

k*

tl

P1

glottalized affricates

voiced nasals

a 1 v e o 1 P a 1 a

(q*)

tl (h)

n

voiced lateral voiced trill voiced semivowels

w

Vowels: front high mid low

i

short central

back u

front

long central

back

(i-)

(u«)

(e ■)

(o-)

(a*)

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

TABLE 2

Orthographic Symbols Consonants:

voiceless stops

b i 1 a b i a 1

e o 1 a r

P

t

voiceless affricates

a 1 V

tz

glottalized stops

b'

a 1 v e o 1 P a 1 a t a 1

P r e

t* ch’

voiceless fricatives

s

x

1

voiced trill

r w

Vowels: front

V

e 1 a r

e 1 a r

u 1 a r

l o t t a 1

(ky)

k

q

7

(q' )

j

(h)

n

voiced lateral

voiced semivowels

V

( k y ' ) k'

tz'

m

V

ch

glottalized affricates

voiced nasals

q

u

y short central

back

front

long central

back

high

i

u

(ii)

(uu)

mid

e

o

(ee)

(oo)

low

a/a

(aa/a)

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

14

central

vowel.

In

written as , while the single long vowel is

(see table 2).

For some

purposes

to re£er to some Quichd dialects as

it

"ten

dialects'9 and to others as "six vowel dialects".

is

vowel

It should

be understood, however, that such a classification is purely typological genetic

and

in no way corresponds

classification

1977:14-20,

Edmonson

1976a:55-9).3

of Quichd 1965:viii,

to

dialects Fox

any

recognized

(cf.

Campbell

1968:190-1,

Kaufman

At least one dialect, that spoken in the town

of Cubulco, is a "five vowel dialect", phonemic vowel length apparently having disappeared completely (Kaufman 1976a:103; Campbell 1977:15). The Proto-Quichean vowel length contrasts are preserved in

modern

Quichd only in

word-final

syllables

(Campbell

1977:38).

Proto-Quichean long vowels in non-final syllables

have

become short in modern Quichd.

all

dialects,

ten

vowel

however, do exhibit vowel length contrasts in all

syllables. morphemes

Many

This of

has

come

about

because

Proto-Quichean

the form *(C)VhC have come to have

the

form

( C ) W C in Quichd (Campbell 1977:38,42,44-5), and long vowels from this source have not been shortened in many modern vowel

dialects of Quichd.3

Some Quichd examples are

ten shown

in table 3, together with cognates from Tzutujil, a language which

better preserves some of the Proto-Quichean

morpheme

shapes.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

15

TABLE 3 Some Proto-Quichean Morpheme Shapes in Quichd and Tzutujil

ErotQ.-Quichean_ _ _ _ _ _ _ _ flui.ch.6_ _ _ _ _ _ _ _ *sootz' 'bat'

sootz1

sootz*

*iichaaj ‘herb'

ichaaj

iichaaj

*pohp 'mat'

poop

pojp

*b'ahlam ‘jaguar'

b'aalam

b'ajlam

Not all ten vowel dialects have phonemic in non-£inal syllables, to

have

ten

however.

vowel

length

For example, Zunil seems

contrasting vowels;

but the

one

speaker

I

consulted from that town seemed to only have short vowels in non-final vowels that

syllables

even in those words

in some other ten vowel dialects.

have

appears

Zunil has lost phonemic vowel length from all

sources

in final syllables.

The situation with

dialects is somewhat less clear.

which

accompany

Cochojil-Gonz&lez Quichd

my

vowel

When I listened to

the language lessons

(1966-9),

six

The one six vowel dialect

that I am familiar with is Cantel.

impression

speaker (Cochojil-Gonzdlez,

the

in

Wick

and

was

that

the

from Cantel)

preserved

the contrast between /a/ and /a/ in all syllables. the

long

Thus it

except

tapes

which

However,

one Cantel speaker with which I have worked first

contrasted these two vowels only in final syllables.

hand I

not

certain whether this means I was mistaken about what

had

heard on the tapes,* or if it means that there is

kind

of

dialectal

difference (perhaps

age

based)

am I

some among

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

16

Cantel speakers. 2.1.2 As

The Phoneme /h/

noted above,

disappeared trace

(in

the Proto-Quichean phoneme

in morphemes of the form *(C)VhC,

*/h/

has

leaving

its

at least some cases) in the lengthening

preceding vowel.

of

the

This phoneme also underwent other changes

in most other environments, remaining in Quichd only in word final position. position

The way in which /h/ is manifested in this

differs

however;

and

considerably

in

different

in fact its status as a phoneme

dialects, is

somewhat

questionable in all dialects. The

most

dialects

straightforward situation is

like

Ixtahuac&n.

those

of

Nahuald

and

that Santa

seen

in

Catarina

In these dialects the phoneme /h/ appears as a

clearly audible [h] in word final position when the word spoken of

in isolation or appears at the end of certain

syntactic

definition

phrases

("phrase-final

is

types

position").

The

of the term "phrase final position" can here

be

taken to mean "at the end of a noun phrase or at the end

of

a

clause".

The

vowel preceding the /h/ is

always

short

regardless of its length in Proto-Quichean,® as seen in

the

examples in (1). (1)

(a)

wah

'food

(made

of

corn

dough)'


I-back] / ___

1 +syll I I -syll ~"| |_-round _| j -high I |_+back _|

(The details of this rule may differ somewhat from

language

to language and dialect to dialect.)

Campbell

(1977:116-8) Quichean If

this

rule

has

According to

diffused

into

some

languages from the Mamean languages to

the

distribution of the prevelar sounds

entirely

by

allophones

a

rule

like

(10),

then

of the velar stop phonemes.

the

is

they

of

the west.

determined

are

However,

clearly in

some

dialects of Quichd other historical changes have taken place which

may

render

prevelar stops.

unpredictable the

distribution

of

the

In such dialects, then, the prevelar stops

may have to be treated as separate phonemes.

According

to

Kaufman (1976a:103) such dialects exist, but he does not say which ones nor give any data. The this

only

point

rather

dialect for which I have extensive

is Cantel.

complex,

and

In this dialect the

rule

things,

the

Thus,

situation

is

must take

into

stops

First of all

in (10) is not really correct for this

correct

on

the status of the prevelar

separate phonemes seems rather marginal. rule

data

the

dialect.

consideration,

as

The

among

other

position of the stop within the root or

stem.

in Cantel the prevelar sounds are found:

(a) at

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

the

28

beginning

of

morpheme

a

root

morpheme if followed

which ends in /j/ (however,

by

/e/

in

a

the prevelar sound is

in free variation with the velar sound if the root

morpheme

is unstressed, i. e., not the final syllable); (b) in a stem when followed by /a/ or /a/ which in turn is followed by /q/ or

/q'/;

(c) medially in a stem when preceded by

followed

by

/i/

/a/ or /a/ which in turn is followed

by

and /j/.

Some examples of (a) are seen in (11). (11) (a) kyej

'horse'


keeb';

a glottal

about

stop

as

cannot

appear before a glottalized consonant in Quichd unless there is a syllable break between them.

It would not be too

far

fetched to suppose that the Cantel form kyeb* also came from a form like ke7eb* by means of contraction:

ke7eb* > kye7b*

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

31

> kyeb’.

This would be similar to the process seen in words

like tyo7j %fat' < *ti7-ooj.

The form tyo7J results from a

rule which is roughly of the form /CVi7Va/ » CyVa7 where is I-back -low -long] and Va is l-high]. really

a regular phonological

seems

to apply only to some,

rule,

meet the structural description. Cantel)

g.,

it

which

In some dialects (e.

Nahuali).

g.,

consistently

In any case, if the form

does result from the application of this

rule,

one could argue that the in (15c) is really a of

not

because

of the words

the rule seems to apply somewhat more

than in others (e. kyeb'

This rule is

however,

not all,

Vi

/k/ followed by /y/ and is not a

prevelar

then

sequence

stop.

This

would then effectively eliminate the possibility that

there

could be a phoneme /ky/ in Cantel. word in (15), CK1

then,

by rule (b) ;

The only

would be (15d). however,

"unexplained"

This word should have

it does not.

One

might

be

inclined to think, especially considering the variation seen in (11c),

that rule (b) fails to apply in (15d) because the

/k/ is too far from the stressed

syllable.

Unfortunately,

though, the prevelar sound does show up equally far from the stressed syllable in (12c) and even farther in (16). (16) kyaqarinaq 'reddened' It is possible, analogy derived.

with

however, (12a),

from

that (12c) and (16) have which

they

are

both

(KI

by

clearly

The etymology of (15d), on the other hand, is not

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

32

at all certain. In

any case,

it can be seen that there is

relatively

little evidence for treating the prevelar sounds as separate phonemes

in

Cantel

Quichd.x*

Nevertheless,

looking

at

things from the point of view of a practical orthography, it appears that the prevelar sounds must be written differently from the corresponding velar sounds. though from

the initial sounds in (15a-c) can be shown sequences

pronounced 13).

This is because

of velar stop followed

by

/y/,

even

to

come

they

are

identically to the prevelar sounds seen in

If one were going to insist on writing some

(11-

prevelar

sounds as or

-b'isooj 'to

be

sad about

something' (b) keej 'horse' /-keej+e+ej/ » -kejeej *to mount a horse' (c) -paam 'stomach' (possessed form) /paam+aaj/ > pamaaj 'stomach* (unpossessed) (d) -joloom 'head' (possessed form) /joloom+aaj/ > jolomaaj 'head* (unpossessed) (e) suutz' 'cloud' /-suutz'+ir/ » -sutz'ir 'to get cloudy' There

are,

however,

some

long vowels which

shorten when they appear in non-final syllables.

do

not

These are

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

36

the

long vowels which have arisen from the loss

Quichean preconsonantal */h/.

of

Proto-

In order to distinguish these

long vowels from those that undergo the shortening rule, one might want to posit an underlying /h/ in such words and then write a rule which converts the /h/ to vowel length.1®

Some

examples are seen in (19). (19) (a) /-mahk/ > -maak 'sin' (possessed form) /mahk+aaj/ > maakaaj 'sin' (unpossessed) (b) /ch'ihch'/ > ch'iich1 'smokey smelling1 (adjective) /-ch*ihch1+ool/ » -ch'iich'ool ‘smokey smell1 (noun) The

presence of this abstract

distinguishes been

what

those long vowels which do not

hypothesized

evidence.

underlying

largely

on

the

basis

/h/, shorten,

of

has

historical

There is only very shaky synchronic evidence that

distinguishes these vowels has anything to do

phoneme

which

/h/;

however,

some

rather

indirect

with

a

synchronic

evidence can be seen in the NahualA forms of the. transitive verb root /-tah/ *to hear'.

Unfortunately,

the /h/ at the

end of this root is never heard.

When this root appears in

phrase-final

the

presumably

position,

it

has

form

which

results from a contraction of the root with

phrase-final suffix for transitive verbs -oh: -toh.

-toh,

/-tah+oh/

When the root appears in non phrase-final

the »

position,

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

37

it has the form -taa. non

phrase-final

Evidence that this latter form is the

form of a morpheme ending in /h/

can

seen from the fact that all transitive verb roots in

be

Quich6

have the form CVC; and if we assume that the final consonant is /h/,

we can explain the form -taa as resulting from

the

NahualA

sandhi

the

preceding

rule

vowel

in

which drops non

/h/

phrase-final

and

lengthens

position.

Further

evidence that the final consonant of the root is /h/ can

be

seen from the fact that the verbal noun which can be derived from

this

tayik.

root

As

by means of the suffix -ik

will be seen later,

when it appears between vowels.

has

the

form

/h/ regularly becomes

/y/

If one agrees that the root

indeed has the form /-tah/, then it should be noted that the Completive Passive stem which can be derived from this by means of the suffix -taj has the form -taataj. vowel

in

explained an

the

first syllable of this stem

The long

can

be

if we assume that there is a rule which

underlying

/h/

before a consonant

to

root

length

easily converts on

the

preceding vowel.17 In addition to the rule discussed above which

shortens

underlying long vowels in non-final syllables, there is also a

sandhi

underlying they

rule

in

ten

vowel

dialects

which

shortens

long vowels in word final closed syllables

appear in non phrase-final position.

An

when

example

seen in (20).

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

is

38

(20) (a) xuukunaaj (b) juun

's/he cured him/her'

'one'

(c) xuukunaj jun ixoq

's/he cured a woman'

Both of the words in (20a) and (20b) have long vowels in the final closed syllable. appear

in

However,

when both of these

the sentence in (20c),

their final

words

vowels

shortened because they appear in non phrase-final

are

position.

This sandhi rule does not apply in the six vowel dialect Cantel.

of

Actually, even in Cantel the vowels /a, e, i, o, u/

sound phonetically shorter in non phrase-final position than they do in phrase-final position. /a/

still contrasts with the short vowel /a/

differ in quality. to

However,

Note that the vowels /e,

the long vowel because

they

i, o, u/

seem

undergo this phonetic shortening non phrase-finally

in

Cantel regardless of whether they were historically long

or

short. As noted before, /y/

between

vowel-initial

vowels.

there is a rule which changes /h/ This rule generally applies

to

when

suffix is added to a stem which ends in

a

/h/.

Some examples are seen in (21).ta (21) (a) /k'ah/ > k'aa ‘bitter' /-k'ah+iil/ > -k'ayiil 'bitterness' (b) /k'ih/ > k'ii ‘many' /-k'ih+aal/ » -k'iyaal 'multiplicity'

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

39

(c) /xehpuh/ > xeepuu 'fat* (noun)

(


xinw&rtis&(k )j

*1

put

him/her to sleep' (32) (a) /r+uxlaab'/ » rdxl&(&)b'

‘his/her breath'

(b) /x#0#uxlaab'+ik/ > xdxldb'ik (33) (a) /x#0#in+taq+o/ > xlnt&qd

‘s/he breathed'

*1 ordered him/her'

(b) /x#0#in+taq##b'i+ik/ > xintdq b'i(l)k

‘I sent him/her'

Terrence Kaufman (personal communication) also noticed these pitch checked

differences for

the

in a dialect survey for PLFM

in

the

early

Zunil

1970's.

that

he

However,

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

54

Kaufman's informant also had clear vowel length distinctions in addition to the pitch differences. Monosyllabic pronounced

with

Polysyllabic initial

vowel-initial words in Quich6 are an

initial glottal

vowel-initial

words are

glottal stop when they occur

stop

when

stressed.

pronounced in

always

with

an

utterance-initial

position and also when they are preceded by a word ending in a vowel.

In other environments such words have no

glottal stop.

initial

In any case these initial glottal stops

never written.

are

The glottal stop is always written in other

positions, however, since it is phonemic in these positions. When /7/ appears preconsonantally (i. the

form (C)V7C ),

e.,

in morphemes

there is generally a very

vowel following the /7/.

short

echo

This echo vowel may be voiceless.

There may also be an optional echo vowel after a /I/.

of

word-final

In this case the echo vowel is always voiceless. There is a sandhi rule which applies to word-final

which

is similar to the sandhi rule which applies

final /h/ discussed in section 2.1.2.

/7/

to

word

In NahualA and Santa

Catarina Ixtahuac&n a word final glottal stop is dropped

in

non

is

phrase-final

lengthened.

position and a preceding short

Thisrule

applies

quite regularly

vowel in

these

dialects to most word-final glottal stops which form part of a

suffix

however,

or which are the final consonant of

a

it generally does not apply to word final

particle; glottal

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

55

stops

which

particle.

are

part of a root morpheme which

in

not

a

The only exception to this that I am aware of is

the transitive verb root -ya7, -yaa

is

which generally has the form

non phrase-final position.

An

example

of

this

process can be seen in (34). (34) (a) ke7ntija7 (b) ke7ntijaa This

rule

seems

Momostenango

*1 am going to drink it' nah

'I have to go drink it*

to be optional

in

other

dialects. seems

to

apply more often than not to the Dependent Status suffix

on

root

and Santa Maria Chiquimula the rule

In

transitive

3.2.4.3.1.3),

verbs

-a7

~

-o7

~

-u7

(see

section

but often does not apply to other word

final

glottal stops. There NahualA

is and

an additional complication to this

rule

in

Santa

not

in

Catarina

Ixtahuac&n

Momostenango or Santa Maria Chiquimula. (34b),

but

As illustrated

the glottal stop at the end of the Dependent

suffix -a7 is dropped in non phrase-final position, short

vowel

However,

preceding

this glottal

when a verb like the one

direct object noun phrase, as seen is even

stop

is

in

Status and the

lengthened.

in (34) is followed by in (35),the glottal

dropped but the vowel is not lengthened. if other constituents intervene

This is

between the

a

stop true

verb

the noun phrase, as seen in (35b).

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

and

56

(35) (a) ke7ntija

jun serwees

I .go.to.dzink.it

one beer

'I am going to drink a beer' (b) ke7ntija I .go.to.drink.it

na

lee

tzaam

NECESSITATIVE

the

liquor

'I have to go and drink the liquor' In Momostenango and Santa Maria Chiquimula the vowel of suffix

-a7

dropped.

is always lengthened when the glottal

Thus,

the

stop

the verb in both of the sentences in

is (35)

would have the form ki7ntijaa in these dialects. At rule or

normal conversational speed there is a

which glottal

often

applies

stop-final word is followed by

vowel-initial word. Maria

when an unstressed

Chiquimula

In NahualA,

a

Momostenango,

the contraction rule

deletes

contraction vowel-final polysyllabic and

Santa

the

final

vowel of the first word (and also the final glottal stop, if present) and inserts a glottal stop after the initial of

the

V=7CV(V).

second

word;

that

is,

vowel

Vt(Vx)(7)##Va (V= )CV(V)

Some examples are seen in (36).

(36) (a) ri the

achii > ra7chii man

'the man'

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

»

(b) ma

xaak'am

not

ta

uloq

you.carried.it IRREALIS

hither

> ma xaak'am tu71oq 'You didn't bring it.' In Santa Catarina Ixtahuac&n the contraction rule is similar except

that

initial

rather

vowel

than inserting

a

glottal

of the second word is simply

stop,

the

lengthened

as

contraction rule in the six vowel dialect of Cantel

is

shown in (37). (37) rii achih » raachih The like

that

course,

seen in Santa

there

is

'the man'

Catarina

Ixtahuac&n

except,

no vowel lengthening since there

vowel length contrast in non-final syllables.

of

is

no

An example is

seen in (38). (38) le

ixoq

the

>

lixoq

woman

'the woman' 2.3

Other Orthographies

The

PLFM orthography used here is only one of

orthographies that have been used for Quich6. chart is

several

A conversion

for the principal orthographies which have been

shown in table 5.

orthography (1976a).

is

based

Basically

The principles upon which are explained

in

full

the in

used PLFM

Kaufman

this orthography is supposed to be

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

an

58

unambiguous

phonemic

orthography using

only

which may be found on a Spanish typewriter. attempt possible,

is

made

to

especially

avoid using those

backspace on the typewriter.

which

symbols

Furthermore, an

diacritics would

the

as

require

much

as

one

to

The letters of the orthography

TABLE 5 Comparison of Orthographies PLFM

Colonial

Officialized

ChAvez

a/a aa/a b' ch ch' e ee h n i ii j k k' ky ky1 1 m n o oo P

a a b ch *h e e

a a b ch ch’ e e

a/— a b ch * e e

a a b ch ch' e e

i/y/j i/y/j h/h c/qu * — — 1 m n o 0 P k t r Z/g t tt tz */ u/v u/v u/v/uh i/y/j

I i j c/qu c'/q'u qui q'ui 1 m n o o P k k* r s t t' tz tz' u u w/u Y/i '

i i j k V ki Pi 1 m n o o P k 0 r s t 4 tz T u u w/u y/i

1 i j k k' ki k'i 1 m n 5 o P

q q'

r s t t• tz tz1 u uu w y 7

W)

❖/$

Salaza

q q’

r s t d tz tz5 u u w/u y/i »

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

59 aze to have as much as possible the values that they have in Spanish;

however,

since Mayan languages have sounds

ace

not found in Spanish,

are

used

certain combinations of

for such sounds that would not be

orthography

used for Spanish.

One of the

symbols

found most

which

in

the

important

principles is that each symbol should represent one and only one phoneme, and

and each phoneme should be represented by

only one symbol.

For this reason,

one

the symbol

is

used to represent the phoneme /k/, in spite of the fact that

is

used

quite rarely for

this

phoneme

in

Spanish,

because the usual Spanish convention of writing /k/ as before

front vowels and as elsewhere would violate

principle of one symbol for each phoneme.

the

same language,

and even

the

There is also

attempt made to emphasize the similarities between of

between

an

dialects

different

Mayan

languages, rather than exaggerate the differences. One

of

orthography It

the

more controversial aspects

has been the use of for the

of

glottal

was decided not to use the symbol for this

as is done in some orthographies,

the

PLFM stop.

phoneme,

because this would

cause

confusion in some languages which have a distinction between glottalized consonants, written as C', and sequences of nonglottalized consonant followed by /7/. Nahaul&

dialect

companion'. distinct

of Quich6,

there is a

For example, in the word

wach7i2

'my

The sound of the sequence /ch7/ in this word is

from

the sound of /ch'/ in a

word

like

ach'i7y

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

60

'tough (like rubber or meat)'. were written as < ’>,

However, if the glottal stop

it would be difficult to indicate

difference between /ch7/ and /ch'/. symbol the

Therefore, a different

is needed to write the glottal stop,

and I

suppose

numeral 7 was chosen because it resembles the

phonetic symbol for this sound.

the

Nevertheless,

standard

many

people

seem to object to this because they feel uncomfortable

with

the

seen

idea of using a numeral as a letter.

below,

however,

As will be

there is a precedent for using numerals in

this way. The

first

orthography

used

to

write

presumably the native hieroglyphic writing great

strides

hieroglyphics hieroglyphic the Quichd

have in

been

recent

made

in

years,

Quichd

system.

Though

deciphering

none

of

these

the

texts has been demonstrated to be language.

was

extant

written

in

None of the hieroglyphic books which

are said to have once existed among the Quiches is known have survived,

to

and stone monuments with carved hieroglyphic

texts were apparently not erected in the Quichd area as they were in some other Mayan areas. After Spanish

the

priests

Quiches were conquered by introduced the writing of

Latin

alphabet as part of their

began

in

developed

the early 1540's.aes

missionary The

Spain

in

Quich6

1524, in

effort,

orthography

was used by literate native speakers as

the

which

that

they

well

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

as

61

foreign

missionaries until the

orthography,

shown

"Colonial",

in

table

not found in Spanish,

4h,

4,,

century.

5

column

in

the

This labeled

basically follows Spanish orthographic practice

of the sixteeenth century.

represent

nineteenth

these tz,

some special symbols were invented to

sounds.

tt >,

However, since Quich6 has sounds

These special symbols are < £, 4,

which

represent the phonemes /q', k',

ch', tz', tz, t'/ respectively. The first four of these letters are usually referred to as

the "Parra letters" because they are said to

have

been

invented by the Spanish Franciscan priest Fray Francisco la

Parra in 1545 (Edmonson 1971:vii).

These four

de

letters

are called tresillo 'little three', cuatzillo 'little four', cuatrillo con hache ’little four with h', coma the

'little four with comma1. possible

and cuatzillo con

It is curious that of

symbols that could have

been

chosen,

Parra

chose to base these letters on the numerals 3 and 4. even

more curious that the 3,

backwards.

Historians

explanation

tresillo.

It

discussed

for

at

above,

the

written to

have

I think there is

least

the

form

should first of all be remembered phoneme / q ’/ is a

number of Quich£ dialects. for

however,

It is

was

and linguists do not seem

wondered too much about this; reasonable

but not the 4,

all

a

of

the

that,

as

pharyngeal

in

a

The Parra letters were also used

writing the Cakchiquel language,

and it is probably

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

a

62

relevant £act that in the Cakchiquel dialects spoken La

Ciudad de Santiago de los Caballeros de Goathemala

old

colonial capital,

site its

both at its former location

(the

at

the

of the present day Ciudad Vieja from 1527 to 1543

and

later location at the site of the present

Guatemala

from

day

1543 until 1773) the phoneme / q 1/

all

it

should be noticed that

the

Antigua is

pronounced as a pharyngeal in word-initial position. of

around

tresillo

also Second

bears

a

striking resemblance, especially in its handwritten form, to the

Arabic

letter

"?ain",

pharyngeal fricative /?/. to

suppose that Parra,

least

which

represents

It would not be too far or some associate of

passing familiarity with Arabic.

spoken

in Spain for some 800 years,

stronghold

in

Spaniards

until

"discovered"

the

same

his,

and the

year

last

had

at

widely Moorish

the

Christian

that

Columbus

America and just 36 years before the

of the Quiches.

voiced fetched

Arabic was

Spain was not reconquered by 1492,

the

conquest

Arabic continued to be spoken in parts

of

Spain for some time after 1492 as evidenced by the fact that Pedro

de

AlcalA

published

a

grammar,

vocabulary,

catechism (Alcald 1505) in the Arabic dialect of Granada an aid in the effort to convert the Arabic speaking to Christianity. Valencia

in

Another Arabic vocabulary was

the early

1600's.

Furthermore,

and as

Muslims

produced in Antonio

de

Nebrija, in his Gram&tica castellana published in 1492 after the

reconquest of Granada but before

Columbus'

discovery,

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

63

makes frequent comparisons of the sounds of Spanish to those of Latin, Greek, Hebrew, and Arabic.

Nebrija's book was the

first grammar of a Romance vernacular written in Europe served

as

a

model for the were

grammarsof

later

writtenin

Native

and

American

languages

that

colonial

Latin

America.**

If we assume, then, that Parra may have had some

familiarity with Arabic, and if we note that some Guatemalan languages had a pharyngeal sound similar to the Arabic including the language spoken in the vicinity of the and political center of the country, surprising

/?/, social

then it should not

be

that it would have occurred to Parra to use

the

Arabic letter •'Tain1', that is, , to write this sound. The cuatrillo is not so easily explained; logic

behind it could conceivably have been something

this: not

however, the like

a letter is needed for the sound Ik']; this sound is found

time

so

in other languages that were well known a

language;

letter

could

not

be

borrowed

at

from

the

another

the / q ’/ would be written with the Arabic

letter

, which looks somewhat like a 3; therefore, one might use the

next number,

number

begins

4,

to write /k'/.

The fact

that

with the sound [k] in Spanish may

some

influence

on the choice of this numeral

some

other like 2.

In any case,

have

rather

once the letter

this had than was

chosen for the /k'/ for whatever reason, the letters for the phonemes phoneme

/ch1/ /ch1/

and

/tz'/

follow

quite

logically.

sounds somewhat like the sound of

/ch/

The but

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

64

with a difference which is similar to the difference between /k/ and /k'/.

If /ch/ is written as ,

that is, by

(which has the sound [k] in Spanish except when before front vowels)

followed by ,

then /ch1/ should be

written

as

,

that is by (which has the sound Ik*]) followed by

.

My explanation of is somewhat more

but perhaps reasonable nevertheless.

speculative,

In Nebrija's time

in

the 15th century, the Spanish letters (or optionally before front vowels) and were pronounced [ts] and respectively. had

[dzl,

By Parra's time in the 16th century, Spanish

undergone some rather extensive

phonological

changes.

Among these changes was one which changed the sounds of both and to Is], for

and thus both of these letters are used

the phoneme /s/ in the colonial

Quich&

orthography.2' 7'

However, it is possible that Parra had some knowledge of the pronunciation of ;

former

perhaps there were even

people who still pronounced it that way.

If so,

logic

behind may have been something

sound

of /tz'/ is similar to the sound

/tz'/

differs from /tz/ in the same way that

from

/k/;

like

of

then

the

this:

the

/tz/;

however,

/ k 1/

since the sound of the Quich£ phoneme

formerly written in Spanish (that is, pronounced

differs /tz/

as ,

Ik] except before front vowels,

some

was

which is

with a

cedilla

written underneith), then it would make sense to write /tz'/ as (which has the sound Ik*]) with a cedilla underneith; however,

it

is

difficult

to write a

cedilla

under

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

65

because

it

therefore,

already has a tail which goes below one

the

line;

could write the cedilla to the side of

the

cuatrillo (). As

can be seen,

the phoneme /q/.

the letter was used to

represent

One might consider the possibility of

Arabic connection for this convention too since there

seems

to be a history of transliterating the Arabic letter as in Spanish:

note,

other facts,

however,

less than compelling. colloquial /q/.

Thus,

(both

of

"qaaf*

for example, that the name of the

country Iraq is always spelled in Spanish. some

an

There are

which make this idea

For one thing,

somewhat

it is known that the

dialects of Arabic spoken in Spain did not

have

Pedro de AlcalA consistently wrote or which

represent

the

sound

[k])

in

his

transcriptions of Granadan Arabic words which would have had /q/ in Classical Arabic.

Furthermore,

the letter

was

used for the sound Ik'] in the orthography developed for the Yucatec Maya language,

which does not have /q/.

I think it

is probably more likely that there was simply a tradition of using the letter to transcribe any foreign k-like sound. In Quichd (as in Arabic) that ’'foreign k-like’1 sound out

to

be [q] while in Yucatec it turned out to

Note also that the prevelar sounds were never

be

turned [k’l.

distinguished

in the colonial orthography, which probably means that these sounds did not exist in Colonial Quich6.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

66

The Parra

colonial

letters

reasonably without

orthography,

and other special symbols,

workable

its

with the addition proved to

orthography for Quichd.

problems,

however.

One

of

of

It the

the be

was

a not

principal

problems was that while some writers used all of the special symbols

with extreme accuracy,

sloppily.

many others used them

Some writers would only use the special

very

letters

sporadically in some words while other writers might use any of

the

various

"strange"

Parra

sound.

accurately,

letters

Even

however,

when

indiscriminately the

symbols

for

were

there still were potential

any used

problems.

For one thing,

vowel length was never written nor was there

any

for writing /h/.

provision

apparently

did

distinctions

have

The

Spanish

some awareness of

the

grammarians vowel

length

since they were sporadically noted in some

the early dictionaries.

Thus,

of

if two words which formed a

minimal pair differing only in vowel length fell together in the dictionary listing, they would be given separate entries with identical spellings. have

However, one of the entries would

the annotation "breve" ('short') and the

have the annotation "larga" ('long'). was

no minimal pair like this,

other

would

In cases where there

there was no annotation

of

vowel length. There were three methods of indicating the glottal stop which

were used sporadically in the

colonial

orthography.

Either a grave or circumflex accent was used over the

vowel

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

67

preceding glottal

the stop

glottal stop, might

or the

be doubled.

vowel

The

preceding

latter

the

method

was

probably a recognition of the echo vowel which sometimes phonetically inserted after a glottal stop. chee7 'tree' might be written as , Most commonly, as ; at all. since

Thus, the word

,

or

.

however, such a word would have been written

that is, the glottal stop was usually not written Furthermore, the doubled vowel method was ambiguous

two consecutive vowels could also represent

sequence.

Thus,

both

reeta7aam ‘s/he knows it'

a

/V7V/

(phrase-

final) and reeta7m ‘s/he knows it' (non phrase-final) have

is

could

been written (if indeed the second form

existed in Colonial Quich6; also been written ,

if it did exist, ,

even

it could have

or though I am

not sure that it ever was). Another problem was that /u/,

/uu/, and /w/ were never

distinguished in the colonial orthography since any of these three phonemes could have been written as or as

was

common

the

practice

in European languages at the

the

though

might appear as the

however;

,

to

and any subsequent one

same word was written as .

always strictly adhered to, ‘paper1

Generally

was that the first one of these three phonemes

appear in a word was written as , in

time.

This rule

was

and a word like

,

,

or

first was probably the most

common

way.

not wuuj

,

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

The

68

voiceless .

allophone of /w/ was also frequently

written

as

Thus the word uleew 'land' was frequently written as

.

Note

that this means that at least some

of

the

above spellings of wuuj could also be interpreted as /u(u)w/ although no such word seems to exist. these

spellings

could

be

taken

Note also that any of

to

represent

/u7j/

or

/u(u)7u(u)j/. A

similar

situation existed with the

phonemes

/i/,

/ii/, and /y/, any of which could be written as , , or .

Usually the first such phoneme in a word was

as ,

and any subsequent one in the same word was written

as .

Thus,

written

as

the word iwiir ‘yesterday1 could have ,

possibilities .5ZS used at all,

though

there

were

The was rather rare,

also and if

written

Thus, .

Although

/y//

like

/w/,

it

was

voiceless

/r/ was never

claims

that

a

may

be

this was never written as

as might be suspected from the use of .

(1977:121)

other

a word like ti7iij ‘meat1 might have

voiceless in word final position,

the

been

it was usually used as the second member of

doubled vowel. been

written

written

was used

as for

Similarly

.

Campbell

the

voiceless

allophone of /!/, but I have never actually encountered this myself. During the 19th century Brasseur de Bourbourg published a

grammar of Quich6 (Brasseur de Bourbourg 1862) which

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

was

69

largely an annotated edition of a manuscript grammar written a century earlier by Francisco had

used

the

colonial

orthography;

introduced some changes in his

was

substituted

substituted for .

Ximdnez.

for

however,

and

the

letter

was

thus, he did not distinguish These new conventions

since been followed by others in printed Quichd

Furthermore, Brasseur generally wrote

/ch'/ from /ch/ nor /tz'/ from /tz/.

colonial

Brasseur

published version: the letter

,

as and as ;

have

Xim£nez, of course,

manuscripts

such

as

editions

those

found,

of for

example, in Recin6s (1957). The

orthography labeled "Officialized11 in table

probably

the most commonly used alphabet for Quich6

5

is

today.

This is not to say that it is commonly used since Quich£ seldom

written by anybody in any

alphabet.

Nevertheless,

the officialized alphabet is the one which is most taught

and the one which has the most published

(mostly

Bible

primers,

and

conference practical This

translations, the like).

is

healthpamphlets,

commonly literature

spelling

This alphabet was approved at

a

held in 1949 for the purpose of deciding upon

a

alphabet for writing the languages of

conference

was

attended largely by

Guatemala.

members

of

the

Summer Institute of Linguistics (or SIL, affiliated with the Wycliff

Bible Translators) and other foreign linguists

anthropologists. adopted

The

alphabet that was decided

by the Summer Institute of Linguistics for

upon

and was

use

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

in

70

its work with the Quichd language. the

Instituto

Indigenista

Institute or U N ) ,

Nacional

as

the

"official"

government adopting this

has

never

It is not,

alphabet enacted

however, referred

because any

kind

the

Guatemalan

of

legislation

this alphabet as "official" in any

sense.

also been used by a number of other entities

those

Since UN,

of

Spanish. read

teaching Indians

to

speak,

read,

for and

officialized

write

their own

alphabet

language

first

before beginning to

taught

using

learn

Proyecto Nacional de Educacidn Bilingue (National

dissent

Project by

publication

some

or

PRONEBI)

of the

describing

though

not

participants. this

orthography

recent

Bilingual

without

The is

the

Spanish.

officialized alphabet has also been used by the

Education

the write

In these courses the students are usually and

it

including

which conduct courses in "castellanizacidn"

purpose

The

Indianist

alphabet has been adopted by the SIL and the

has

to

(National

by

and for this reason it is referred to as

the "officialized" alphabet. to

It was also adopted

most a

some recent

booklet

published by the Instituto Indigenista Nacional (1977). As owes

can be seen in table 5,

quite a bit to the colonial

the officialized orthography;

alphabet

however,

number of changes have been made to make it conform more modern Spanish orthographic practice. rather than is used for /j/.

Thus,

for

a to

example,

Also the Parra letters

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

71

have been eliminated in favor of symbols which can be on

a

typewriter and which more

nature of the phonemic system. that

consistent

vowel

length

consistently

display

the

The most important change is

symbols have been adopted

and

found

the glottal

stop.

for

The

indicating

result

is

an

alphabet which is more or less phonemic in that all phonemic distinctions /h/) It

are made (with the exception of the

and most symbols represent one and only is

not

the

represented this

case,

however,

that

"phoneme**

one

each

phoneme.

phoneme

by one and only one symbol principally

alphabet

continues

to

transport

into

is

because

Quich6

the

traditional Spanish convention of writing /k/ as before front vowels and as elsewhere. It should also be noted that the prevelar phonemes represented as and .

peculiarities, In

some

were

a

This

however,

alphabet

has

other

which are definately

similar

unaesthetic.

cases the phoneme /y/ is written as as vowel

consonant.

when,

for example,

it

is

for.

stop since,

as we have seen,

this

makes

it

difficult to distinguish between /C7/ and /C'/. The fourth column in table 5 presents an alphabet which was invented by Adri&n In6s Ch&vez. the

four

orthographies

developed

entirely

This is the only one of

shown in table 5

which

by a Quiche speaker.

has

ChAvez

been

was

the

founder of the Academia de la Lengua Maya-Quich6 (Academy of the

Maya-Quich6

Language)

in

1959.

He

presented

alphabet to the educational authorities in 1963 as a orthography

for

communities,

but

officialized

alphabet

1983:29).

alphabetization

in

Quich£

more

"scientific"

better speaking

it was rejected on the grounds was

that (Sam

ChAvez's orthography is presented and

a book (Ch&vez no date) which was produced with the

of

a

which

funded by a

German

The

typewriter, serious

ignored alphabet

which

obstacle

alphabet.

fact was to

has

the

the

general

is

only

great

of

his

one

expense,

adoption

aid

of

such is

a

Ch&vez's

ChAvez's work has been pretty much

scholarly

found

there

produced at

Furthermore, by

that

Colop

foundation,

allows Ch&vez to type the special characters

orthography.

the

explained

in

special typewriter,

his

community.

favor with

some

Nevertheless, native

speakers

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

his of

74

Quich£,

who

use

it

occasionally for

writing

letters and other types of documents in Quichd. who use this alphabet tend to be some of the

(by

hand)

The Quichds

younger,

more

educated ones,

many of whom have recently taken an interest

in maintaining

traditional Quichd culture. Such persons are

pround of the fact that one of their own people was able develop a viable writing system for their language and tend

to also

to resent the fact that the officialized alphabet

been

imposed

Guatemalans.

upon This

them

by

foreigners and

has

non-Indian

imposition has been seen as

a

further

example of the same kind of paternalism that they ascribe to the

UN

and the program of

castellanizacidn,

which

have

generally looked upon the native peoples as a "problem" Guatemala

which

"incorporate"

had

them

to

into

be the

alleviated

by

"mainstream"

for

trying of

to

Guatemalan

society (cf. Sam Colop 1983:32-4). Chdvez’s

orthography employs the special letters
for the phonemes / b ’, c h ’, k', q, q', t ’, tz'/, is

respectively.

ch'umiil

The letter resembles a star, which

in Quichd,

a word begining with

the

phoneme

/ch'/.

The letter resembles the sun, which is q'iij in

Quichd,

a

letter

word which begins with the

is borrowed from Scandinavian

represented the sound [kl. this

letter

phoneme

/ q 1/.

runes

The

where

it

Chdvez claims that the sound

of

was similar to the sound of the

phoneme

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

/k'/

75

which

this

date:68).

letter represents in his

alphabet

(ChAvez

no

This alphabet also uses the letter , which is

not shown in table 5.

Chavez (no date:68) claims that this

letter is used in only one word: = /sanyab'/ 'sand1. Note

that in Santa Maria Chiquimula this word has the

£orm

/sanayib'/. In addition to the special letters, employs a number of accent marks. vowel

Chavez's

The grave accent over

represents a word-final glottal stop,

/chee7/ 'tree1.

phoneme

ChAvez

because

consonant. rapid

as in

uses he

as in = /b'o7j/

these two

does

not

symbols

recognize

for the

/7/

pronunciation

of

the vowel and

True

vowel

indicated in his alphabet, San

spoken. do

same as

a

He describes the grave accent as representing a

date:69).

from

=

'pot1.

the

the

circumflex

representing a prolonged pronunciation of the vowel no

a

A circumflex accent over a vowel represents

a preconsonatal glottal stop, Apparently

alphabet

length however.

is

not

(ChAvez

consistently

ChAvez is

Francisco el Alto where a six

as

vowel

originally dialect

is

He claims that there are some words in QuichA which

not have vowels,

Apparently

the

for example,

=

phoneme /a/ has a shwa-like

/chaj/ sound

'pine'. in

his

dialect, and he does not seem to consider [a] to be a vowel. ChAvez always writes the long vowel /a/ as as in = /chaj/ 'ashes'.

Thus,

in some cases he does have a way,

though a rather strange one,

of distinguishing /a/ and /a/.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

76

However,

he is not consistent about this because he

in numerous words which should have /a/,

=

/ab'aj/ 'rock*.

Note in table 5

writes

for

example,

that

Chavez's

alphabet is similar to the officialized alphabet in the

way

it

represents the prevelar phonemes and in its use

of

and in certain environments for writing /y/ and /w/. Ch&vez This

also uses an acute accent to

indicate

stress.

accent is employed according to the same rules

Spanish

orthography.

In

Spanish it is assumed

as

in

that

the

stress falls on the penultimate syllable if the word ends in a vowel, , or , but it is assumed to fall on the final syllable given

of

a word ending in any other

word is stressed in conformity with this

no accent mark is written. a

consonant.

syllable

assumption,

such

that

If

assumption,

However, if the stress falls on it

does

not

conform

with

then an acute accent is written over the

of that syllable.

a

the vowel

Chdvez follows this exact same rule

using the acute accent in Quichd.

for

Note, however, that this

practice is completely absurd in a language like Quiche.

In

Spanish stress is phonemic: termino *1 finish', tdrmino 'end (noun)1,

termind 's/he finished’.

phonemic

in Quichd.

It is entirely

However,

stress is not

predictable,

falling

regularly on the final syllable; therefore, there is no need to write it at all. The final orthography shown in table 5 is one that

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

was

77

originally

adopted by Manuel Salazar for use in his

thesis

on Cakchiquel literature (Salazar Tetzaguic 1978).

He does

not say in any detail how he arrived at this alphabet than

to say (Salazar Tetzaguic 1978:4) that after

the

various

existing

alphabets,

and

having

studying had

experience in trying to transcribe and translate decided to use the alphabet shown.

a

Mayan

language;

some

texts,

he

Thus, this is the second

alphabet in table 5 which was developed by a native of

other

speaker

however, unlike ChAvez*s, Salazar's

appears to be a derivative of the others.

It appears to be

most like the PLFM alphabet since it uses for the phonemes /k, controversial

k', q, q'/.

in

officialized alphabet. to

favor

However, Salazar abandons the of

,

of

principles

which are identical to those advocated by the PLFM and in

the

used

is

the

In this case it was decided

as in the

officialized

alphabet,

rather than the controversial formerly advocated by

the

PLFM.

has

While

generally gone

the

Summer

Institute

of

Linguistics

been opposed to this development,

along

with

the proposals of

the

the

Academia

PLFM

has

and

has

officially adopted < ’> as its symbol for the glottal stop in place of . On

November

23,

1987,

Linguistics Newsletter (vol. president of Guatemala, a

presidential

as XIV,

reported no.

in

the

Mayan

2, April 1988), the

Marco Vinicio Cerezo Arevalo signed

decree (acuerdo

gubernativo

no.

1046-87)

which made official the alphabets which had been adopted the

Academia

de las Lenguas

Mayas.

While

I

by

personally

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

80 approve of and support this development, I have retained the use

work.21*

The

principal reason for this is that the major portion of

this

work

of

was

for

the glottal stop

in

this

completed before I became aware of

status of the new alphabet,

the

official

and I find it too difficult

at

this point to convert all of my glottal stops to < ’> without simultaneously

converting all of my numeral sevens also

< ’>.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

to

81

NOTES 1. Note

that

there have been some

minor

modifications,

largely made at the suggestion of native

speakers,

in

the practical orthographies used by the PLFM since publication

of

modification

which

doubled

Kaufman

vowels

represent

applies

(i.

long

(1976a).

e.

) that was

As of 1987,

to use the alphabet adopted by the

such

the oo,

e:,i:, o:,

advocated in Kaufman (1976a).

las

only

vowels (see table 2) in place

vowel + colon (i.

begun

The

the

the PLFM has Academia

which is discussed at the

de

end

of

this chapter. 2. The "six vowel" system is probably an areal phenomenon. It

is also found in some

contrast

with

Cakchiquel. six at

other

Cakchiquel "nine

dialects,

vowel

dialects"

of

There is some evidence that in Quichd the

vowel system is spreading among younger least in some areas.

communication; older

which

speakers

cf.

Terrence

Kaufman

Kaufman

speakers, (personal

1976a:106-7)reports

in TotonicapAn have ten

vowels

that while

younger speakers have six. 3. Proto-Quichean morphemes of the form *(C)V7C have become ( C)WC in QuichA, that

these

also

and Campbell (1977:38) claims

long vowels too remain long

in

non-final

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

82

syllables; however, he presents no examples which would demonstrate

this to be true.

In any

case,

there is plenty of evidence that it is not example,

from

true.

incidently, (C)V7C

Quichean

source

the

syllable.

that modern Quichd morphemes of the

are reconstructed as *(C)V7VC

(Campbell

according

in can

which

root vowel is clearly short in the non-final

form

For

according to Campbell 1977:44) one

a noun stem -tewaal 'cold(ness)1 in

Note,

think

the adjective teew 'cold' ( < *te7w

Proto-Quichean derive

I

1977:44).

In

to Campbell (1977:14),

in

Cubulco,

even /I/

Protohowever,

from

has disappeared (without a trace since

this

Cubulco

also has lost all vowel length contrasts). 4. One

can

not

resolve

transcriptions marking

of

reliable.

I

the

vowel

to

by

looking

tapes in the

is

to

It should be further noted

contrast

between /a/ and /a/ is not always

even

final syllables in Cantel:

in

neutralized

not

able

before

the

the

since

the tapes are not at

me so I have not been

heard.

at

text

length in the text

Furthermore,

available what

of

this

the

always present recheck

that

the

maintained contrast

/I/ with only the long

vowel

is /a/

appearing in that environment. 5. There

is

Ixtahuac&n,

one exception to but

not

in

this: Nahual&,

in the

Santa word

Catarina meaning

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

83

'house' (Proto-Quichean *jaah) appears as rather word,

than the

jaa.

shortening the vowel before /h/ vowel

length is retained and

Thus, in

the

this

/h/

is

dropped. 6. Certain particles such as wih and nah are exceptions to this.

The

particle

wih,

which is

placed

after

predicate when a locative phrase is fronted, lengthen its vowel when it appears in non position.

does

a not

phrase-final

This is illustrated in (i).

(i) (a) chwa in.front.of

jaa

k'oo

wih

house

be.someplace

PARTICLE

'At home is where s/he is.1 (b) chwa

jaa

k'oo

wi

chwe7q

PARTICLE

tomorrow

*At home is where s/he will be tomorrow.' (Similarly,

the

particle

vowel

the

/h/ is dropped in

when

position.) suggested

Kaufman that

nah does not

lengthen

non

its

phrase

final

(personal communication)

has

this is because the vowel

underlyingly

short whereas

underlyingly

long.

There

the vowel is a problem

in

in

wih

k'ih

with

is is

this,

however, in that the vowel in k'ih was originally short in Proto-Quichean (cf. Du Bois 1985:68). there

Nevertheless,

is a further difference between words like

k'ih

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

84

and

words like wih.

Proto-quichean;

Words like k'ih ended in /h/

in

however, there is comparative evidence

(see Du Bois 1985) that some words,

including wih

and

nah,

ended in vowels in Proto-Quichean in spite of the

fact

that

they end in /h/ in modern

Quich6.

It

is

possible that the different treatment of the vowels may be

related

underlying words

to this in

some

way;

for example,

vowel length may be relevant only in

which originally ended in vowels.

the those

It is

also

possible, however, that the length of these vowels

has

more

are

to

do

with

the fact

that

the

particles

unstressed in sentences like (ib) (sometimes the is dropped non phrase-finally,

vowel

as well as the /h/,

in

wih and it is pronounced [u]) whereas the quantifier in (2b) has at least some degree of stress. 7. Again, that

particles like wiih and naah are exceptions their

position

as

vowels

are

short

illustrated

in

in (i)

non

in

phrase-final

below.

(Cf.

the

NahualA/Santa Catarina IxtahuacAn forms in footnote 6.) (i) (a) cho in.front.of

jaah

k'oo

wiih

house

be.someplace

PARTICLE

'It was at home that s/he was.' (b) cho

jaah

k'oo

wi‘

chwe7q

PARTICLE

tomorrow

*It is at home that s/he will be tomorrow.'

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

85 Note

that

forms.

the two sentences

above

are

Momostenango

The Santa Maria Chiquimula forms are identical

except that in each case the first word is chi. 8. The

possible exceptions that I am aware of are

‘tooth’,

ak'wa(a)h

'tamal

de elote (tamale

fresh corn)1, and naab'eeh ‘first1. a

compound

tooth1.

of -wa(ch) ‘face’ and

-wareh made

of

The stem -wareh is r-eeh

‘his/her/its

The last morpheme, eeh ‘tooth’ now appears in

isolation however,

only

as a day name in the

Mayan

calendar;

it originally had a long vowel (cf.

Tzutujil

eey), which would lead one to expect the final vowel of -wareh

to be long also.

It is also curious that

the

unpossessed form of this stem,

which requires a suffix

-a(a)j,

shows

uncharacteristically

inserted between the root and the The

a

glottal

suffix:

stop

ware7a(a)j.

second morpheme of ak'wa(a)h is the same

morpheme

wa(a)h ‘food’ that we have already seen; therefore, one would

expect

kaxlanwah

this

discussed

word below

to

pattern in

the

like text.

the

word

For

some

speakers this is true; but some speakers seem to have a long vowel in both the possessed and unpossessed

forms

while others seem to have a short vowel in both The

word

naab'eeh has the expected

Tzutujil

najb'eey).

position

the

though

However,

in

vowel often strikes me

long non as

forms.

vowel

(cf.

phrase-final being

I have to admit that for some reason I

short have

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

a

86

hard time hearing the length of the final vowel in this word. 9. Actually, (8d) is regular in NahualA and Santa Catarina Ixtahuac&n,

where this word has the form aanimah.

It

may be that the vowel does not lengthen in (8c) because of

the

vowel

consonant cluster. does

Santa

It also may be

not lengthen in (8d) in

Maria

Chiguimula

antepenultimate

because

syllable.

that

the

Momostenango

and

it

is

in

Since I do not have

examples of borrowed words with these

the other

characteristics,

however, I am unable to prove these hypotheses. 10. In

NahualA and Santa Catarina IxtahuacAn all words

this class follow the same rules as other words in /h/:

the vowel is short before /h/,

of

ending

and the /h/ is

dropped and the final vowel lengthened when the word is in non phrase-final position. two dialects,

many,

Furthermore,

in

perhaps all, of the nouns in this

class are of the type which lengthen their final when

possessed.

these

As

usual

when

such

a

vowel

vowel

is

lengthened, the /h/ is dropped. 11. Terrence Kaufman (personal communication) says that one motivation for writing /h/ in dialects like NahualA and Santa

Catarina

assignment.

Ixtahuac&n

Stress

has

to

do

with

stress

in Quich£ is regularly placed

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

on

87

the last syllable of a word in isolation. least

one

exception

to

this,

however.

adjective appears before a noun that it adjective

takes a suffix -V.

There is at When

modifies,

An example is

an the

seen

in

Though this suffix forms part of the final syllable

of

the

is

(i) . (i)

(a) nim 'big' (b) nim-a

jaa

big-ATTRIBUTIVE

house

'big house'

adjective under such conditions,

never stressed. syllable

than noun. for final

although

this

suffix

which

receives

this stress is significantly

the stress on the final syllable of the

weaker modified

those /h/-final words which appear in non position,

phrase-

in which case the /h/ is dropped

and

Thus, a more accurate version of

the stress rule could be stated as follows:

stress the

syllable of a word which ends in a consonant

in a long vowel.

Note that such a formulation of

stress rule may also relate to the presence of a vowel

the

Quichd words generally end in consonants except

the vowel lengthened.

final

suffix

Indeed, it often sounds like it is the

preceding

stress,

this

in

the

particles wih

and

nah,

or the

short

discussed

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

in

footnote 6, their

which loose /h/, but which do not lengthen

vowels,

appear

in non

and

which are

phrase-final

unstressed, position.

when

they

Actually,

of

course, an accurate treatment of Quich6 phonology would probably

have

to be stated in

terms

of

lengthening

stressed final vowels rather than in terms of assigning stress to final long vowels; however, the point is that by writing /h/ when it occurs in the NahualA and Catarina

Ixtahuac&n

determine

stress placement in written

however, useful

that

dialects

writing

one

can

Santa

accurately

Quichd.

/h/ would not serve

purpose in dialects like Momostengo

Note,

any and

such Santa

Maria Chiquimula.

In these dialects the vowel is

not

always

when /h/

for

lengthened

is

dropped.

Thus,

example, in sentence (7a) in the text, the final /h/ is dropped

in the word kaxnwah because it appears in

phrase-final position.

non

Furthermore, the final vowel in

this word is not lengthened;

nevertheless,

the

final

vowel in this word is stressed. 12. Henne (1980) lists the word sik'ai ‘apazote (a type herb)';

however,

my informant insists that this

of word

should be pronounced siky'aj. 13. The Proto-Quichean word *ha7 ‘water1 has become ja7 modern

Quichd;

however,

*/h/ regularly

becomes

in /y/

between vowels.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

89

14. Even

if

one

prevelars

were

were

to

decide

that

the

separate phonemes from

velars

a

and

traditional

point of view, they would still have to be related from the point of view of generative phonology. discussed in section 3.2.3.2, which

can

As will be

there is a suffix

be added to an adjective to

-Ctoj

derive

adjective meaning 'somewhat (adjective)'.

a

new

Thus,

from

the adjective kaq 'red' (Momostenango dialect), one can derive the adjective kaqkoj 'reddish*. latter

word has the form kyaqkoj.

involves

reduplication,

In Cantel this

Since

the

this would seem

to

suffix indicate

that the underlying form of this word in Cantel must be the

same as the surface form in Momostenango and

that

there must be a phonological rule which changes /k/ (I?) that argue

in the appropriate environment. such

a rule would be needed

that

many,

if not

all,

to

Given, therefore, anyway,

one

prevelars

could

should

be

derived from underlying velars by the appropriate rule. 15. Actually the situation is slightly different for speakers maintain syllables.

those

of six vowel dialects like Cantel Quiche no

vowel

length

contrasts

in

who

non-final

For such speakers the contrast between /a/

and

/a/ is maintained in final syllables that

any

consonant

other than /I/ or

/h/.

environments the contrast is neutralized,

But

end in

in

other

and any

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

low

90

central

vowel which appears in such an environment

pronounced although

either

as

/a/ or /a/

in

free

it is always written as .

is

variation

Some

examples

are seen in (i). (i) (a) chak 'work' (noun) /-chak+un/ > -chakun *to work’ (b) kyaq 'red1 /-kyaq+ar/ > -kyaqar *to turn red' None

of this discussion is relevant to the other

vowels of Cantel Quich6 since there is no vowel

four length

contrast with these. 16.

Of

course such a rule would have to be ordered

after

the vowel shortening rule discussed previously. 17. This example is not as compelling as it could be some

of

discussed forms

the

forms

of

this

verb

above) are irregular and since some

are different in other dialects.

Henne (1980),

form

Completive

-ta7 in phrase-final Passive

has

the

as

of

the to

phrase-

In Cantel, it has

position, form

-toh,

According

this verb has the form -ta7o in

final position in Chichicastenango. the

(e.g.,

since

and

the

-ta7taj.

In

Momostenango and Santa Maria Chiquimula the verbal noun has the form ta7ik.

All of these forms suggest a final

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

91

/7/ in the root rather than a final /h/.

However,

of

phrase-final

these

position;

dialects have -ta(a)

in

and the

dialects

latter

two

non

agree

all

with

NahualA in having -too in phrase-final position.

There

are also some forms in Cantel which do not exhibit /7/.

the

Another example of the phenomenon being discussed

here, which may be somewhat clearer, can be seen in the adjective tzah 'salty' (in Momostenango, (tsa'fi]).

As

will be discussed in

section

such adjectives can take a suffix -Cioj, a new adjective meaning 'somewhat this

suffix is added to tzah,

'somewhat syllable

salty'.

tzaa,

i. e.,

3.2.3.2,

which derives

(adjective)1.

the result is

The long vowel in the

When

tzaatzoj

penultimate

can be explained if we assume that

the

root

ends in /h/ (which it clearly does) and that there is a rule

which

changes

consonants.

Such

/h/

to

vowel

length

before

examples seem to be pretty few

and

far between, however. 18. Historically underwent example, two

there

seem to be some cases

different there

changes

between

where

*/h/

vowels.

are cases where */h/ appeared

For between

vowels and resulted in a single long vowel in

modern language.

One example is paar 'skunk' < Proto-

Quichean *pahar (Campbell 1977:48). happened

This may be

with the phrase-final form of the

discussed above:

the

/-tah+oh/ > -tooh.

verb

what -tab

This latter form

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

92

would then become -toh, as

discussed

before

as observed, in NahualA since,

previously,

/h/ in this dialect.

which

only

short

vowels

appear

There are other

suggest that */h/ sometimes became

examples

/w/

between

vowels if one of the flanking vowels was [+round]. example is k(a)woq,

a day name of the Mayan

The Pocomchi form of this word,

kahoq,

One

calendar.

suggests

the Proto-Quichean form must have been *kahoq.

that

Neither

of these rules appears to be synchronically productive, however. 19. Word

initial

generally

*/h/

become

in Proto-Quichean /w/ before rounded

has

in

vowels

Quichd

and

/j/

elsewhere though there are some exceptions to this. 20. In

word-final

between

position the

/p/ and /b'/,

native

speaker,

Cantel,

where

is /p/

most

salient

difference

at least to the ears of a the aspiration of

is generally

the

non­

/p/.

unaspirated,

In

it

is

rather difficult to hear the difference between /p/ and /b’/ in word-final position. 21. According to Campbell (1977:15), /n/

/___# in the dialect of

there is a rule /m/ »

Joyabaj.

In

the

dialects with which I am familiar it often sounds /n/ becomes /m/ before bilabials; speakers

I

however,

have worked with unanimously

other like

the native reject

this

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

93

claim

and

insist on writing

for

underlying

/n/

before bilabials. 22. I

will

observe

this

orthographic

convention

here;

however, it should be noted that native speakers have a very strong tendency to ignore the convention. of

no

case

quotation

where the presenceor

absence

mark would lead to ambiguity by

distinguish

between

technically

a

two native words although

possibility.

speakers,

uul"

however,

of

There

'rubber' (
zlin-taat).

Note

that the use of

in- rather than nu- on a limited set of

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

147

words

is

also found in some dialects

which

suggests

Cakchlquel,

that this may be a feature

Quichean

which is being lost in some

than

innovation

an

of

in

NahualA

of

Proto-

dialects

rather

and

Santa

Catarina

Ixtahuac&n. 5. It is possible that the forms in (11) and (12) are

not

simple noun phrases but actually relative clauses.

If

so,

then (11) would literally be %the (=those who) are

guides',

and (12) would literally be 'the (=those who)

are men*. 6. Actually,

-xaaq

is

not a class 1

noun

either.

As

discussed below, it belongs to class y: nouns which are always possessed. 7. Teeth (see example 25) are frequently lost, knocked

out;

thought

of

but even so, as

being

one's

they are

pulled, or

still

teeth.

typically

They

are

not

typically bought, sold, or used for other purposes. 8 . Note

that

-achajiil 'husband' appears to

from *-achiij 'man' ( > achii in modern

be

derived

Quichd;

achij

means 'male (of animals)' in modern Cakchiquel) + -ill, the

suffix

nouns.

We

used see

on the possessed forms the

w-ixoq-ill 'my wife',

same

thing

ixoq-il-oom

in

of

class

ixog

'wife'.

2

'woman', Thus,

appears that at least some of the class 3 nouns

it

taking

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

148

the absolutive sufix -oom are originally class 2 nouns: the relationship between husband and wife is not one of prototypical

possession.

unpossessed the

Note,

simple suffixless unpossessed form. to

suffix -a(a)j

note,

is

plural suffix, here,

suffix

consonants

given

that

the

has the same form as

and

seen

in

and

in

these

also

-oom,

seen

irregular

(10b).

(The

forms

were

It is possible that the

suffix -atz seen in alib'atz

law' is not really a suffix.

'daughter-in-

It may be that

actually a compound of the noun -allib'

with

from

absolutive

the

(9b)

/t/ seen

originally part of the stems.)

in-law,

is

the

sometimes translated as if it were a

-oom

/b'/

absolutive

It

that the absolutive suffix

apparently

plural

that

form in -oom has a different meaning

interesting

is

however,

woman's parent-in-law, woman's

allb'atz 'daughter-

sister-in-law'

the noun -atz 'older sibling of the

same

sex'.

It is not clear how likely this would be, however. 9. The

word

b'aq'ach - b'aq'wach seen in (54a)

pronounced b*oq,och, It

is

b'oq'wach,

waq'ach,

not clear whether or not the latter

have the same etymology as the others. also

be used as a Class 3

noun,

e.

is

also

or vog'och. two

forms

This noun g.,

may

b ’aq'achaaj

'eye', nub'aq'ach 'my eye’. 10. Numbers

are

considered true nouns in

Quichd

because

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

149 they

can

be possessed.

possessed

forms

of

ordinal numbers. singular

The

person

the numbers above

Thus,

possessed

third

singular

two

form

the

for example, the third person

form of kai,

the root

of kaj-ib'

'four', is ukaai 'fourth' (lit., 'its four'). number root begins with a vowel,

however,

When the

the ordinal

number is formed by prefixing two third person singular set A prefixes in modern Quichd. the

root

of

ox-ib*

u-r-oox 'third'. of

the

For example, from ox,

'three',

one

forms the ordinal

This was not the case in older forms

language,

however.

The

attested

forms

in

Colonial Quichd only have one prefix: . 11. Historically, actually suffix

the

the

of

enumeratives

been a numeral classifier -V(V)b*

plural things. still

system

as a general

numeral

have

with

the

classifier

for

Some Mayan languages (e. g., Jacaltec)

have a system of numeral Quichd

system,

may

system

is no

classifiers;

longer

considered

however, a

true

numeral classifier system for the reasons stated in the text.

Note

further

that

the

suffix

-V(V)b*

is

obligatory even when counting (i. e., reciting the list of

numbers

respect,

from one to ten or

this

morpheme

whatever).

does not behave like

In

this

a

true

numeral classifier in modern Quichd. 12. The

word

r-ach'll 'and,

together with'

in

(60)

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

is

150

itself

a

relational

noun whose

'his/her/its companion'. form for

r-achi71,

meaning

is

In Cantel this word has

the

but a shorter form chil is often

the relational noun.

used

literal

This shorter form

used

is

also

in Nahuald. while the non-relational noun has

form z-ach7il in that dialect. sentence

the

the

Note also that in this

masculine name Xwaan is

preceded

by

a

particle a while the feminine name Xwa7n is preceded by a

particle al.

first

These particles are used

before

the

names of people whom one would address with

the

second person familiar pronouns. whom

one would address with the second

pronouns

either

optionally

feminine

have

preceded

masculine

names)

no preceding by the word

or

names).

masculine

the In

familiar

feminine

familiar

particle

a

while

The names of

word

Santa

person

particle

taat naan

Maria

formal or

are

*father'

(for

'mother*

(for

Chiguimula

the

particle has the form al particle has

people

and

the

form

all.

- al is derived from the

word

ala

the particle al - all is derived from

the The 'boy*

the

word

all ‘girl'. 13. Ch&vez (no date:42), the

attributive

appearing

with

who speaks a dialect which

suffix, the

claims

attributive

that

suffix

adjectives in

colonial

Quichd

manuscripts are erroneous and that these

appear

only

because the manuscripts were

lacks

forms

written

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

by

151

Spaniards who tended to hispanize Quichd pronunciation. Clearly, however, such an assertion is ridiculous given that

this suffix is found in many modern dialects

and

given that its use is clearly unlike anything found Spanish.

Clearly,

if

a

Spaniard

had

in

difficulty

pronouncing the word g'an in a construction like

(85),

he would have had equal difficulty pronouncing it in construction colonial

like

(88).

manuscripts

adjective

have

Nevertheless,

nor

in modern

a final vowel added

neither

Quichd in

a in

does

an

constructions

like (88). 14. One

might

want to analyze the attributive

suffix

-V

followed by the suffix -laj as a simple suffix -Vlaj in all

dialects.

However,

in

those

regularly use the attributive suffix, be

stipulated

identical

to

dialects

it would have to

that the vowel of the suffix the

vowel

of

the

which

-Vlaj

attributive

is

suffix

(although that vowel may undergo vowel harmony in

some

dialects).

It would also be possible to analyze

-laj

as an adverbial particle rather than as a suffix.

This

is

in

fact

Tzutujil.

what I

am

Dayley

(1985:196,

268)

unaware of any evidence

argue for one analysis over the other;

then

one

would

probably

have

that

for would

however, if one

chose to treat laj as an adverbial particle, does,

does

to

as Dayley treat

the

preceding vowel as the attributive suffix.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

152

CHAPTER 4 MORPHOLOGY II 4.1 Verbs In Quichd,

verbs can be distinguished from other types

of (non-verbal) predicates (e.g., predicate nouns, predicate adjectives) by the fact that the former, but not the latter, have inflections for tense,

aspect,

verbs come in two basic types: Furthermore, root

transitive

and/or

mood.

Quichd

transitive and intransitive.

verbs are of two different

(or non-derived) and derived.

These three

types:

different

types of verbs can be distinguished by their inflections and by their derivational possibilities.

This chapter will deal

with the morphology of intransitive verbs. 4.2 Intransitive Verbs Intransitive verbs are those which have inflections for subject agreement but not for direct object

agreement.

In

the simplest cases, the structure of an intransitive verb is as shown in (1). (1) tense/aspect/mood+agreement+stem(-phrase.final.suffix) Derived

intransitive

polysyllabic

verbs

(generally

stems) are treated identically to

those

with

non-derived

intransitive verbs (generally those with monosyllabic stems) with respect to morphology.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

153

4.2.1 Inflection 4.2.1.1 Subject Agreement Subject

agreement on intransitive verbs is

marked

by

means of* the set B clitics shown in table 3 of Chapter 3.

A

sample paradigm of an intransitive verb is shown in (2).* (2) (a) x+in+war-ik (b)

x+at+war-ik ‘you (sg.) slept'

(c)

x+war

laa

(d) x+0+war-ik (e) x+oj+war-ik

written

‘you (sg. formal) slept' ‘s/he slept' ‘we slept'

(f)

x+ix+war-ik ‘you (pi.) slept'

(g)

x+war alag

(h) x+e7+war-ik The

*1 slept'

set

B

‘you (pi. formal) slept* ‘they slept'

clitics seen in (2)

together with the verb,

have

generally

been

as shown, ; in most of

the

orthographies that have been used for Quichd, including that of the PLFM. used by set used

B

(Two exceptions to this are the orthographies

Brasseur (1862) and Ch&vez (no date).) morphemes have often been considered

on

1981:16).

verbs

(e.

g.,

by Mondloch

Indeed, the

prefixes

1978b:3

and

when Dayley

However, Kaufman (1986b:24) claims that the set B

morphemes in Proto-Mayan were enclitic to the first word the

predicate.

Looking at (1) one could perhaps say

of that

the set B morphemes are still enclitic to the first word

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

of

154

the

predicate in modern Quichd,

predicate"

would

tense/aspect/mood however, Quichd

that

to

predicate

interpreted

particle.

I

to am

mean

the

inclined

the

set

B morphemes do

anything when used with

adjectives.

not

to

more

think,

sense

modern

seem

predicate

to

nouns

In some Mayan languages the

to say that in modern Quichd

the

initial

clitics appear after the stem in such situations. make

of

this really is not quite correct for

because

enclitic

be

where "first word

set

be and B

It might the

set

B

morphemes have become proclitics. Nevertheless, enclitics, believe

proclitics,

that

criteria

or

their exact status may whatever,

be

as

reason

to

prefixes

in

I see no

the set B morphemes have become

modern Quichd. six

whatever

Zwicky and Pullum (1983) offer the following for

distinguishing between

clear

cases

of

clitics and inflectional affixes: A.

Clitics can exhibit a low degree of selection with respect high

to their hosts,

degree

while affixes exhibit

of selection with respect

to

a

their

stems. B.

Arbitrary gaps in the set of combinations are more characteristic

of

affixed words than

of

clitic

groups.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

155

C.

Morphophonological characteristic

idiosyncracies

of

are

affixed words than

of

more clitic

groups. D.

Semantic idiosyncracies are more characteristic of affixed words than of clitic groups.

E.

Syntactic

rules

can affect

affixed words,

but

cannot affect clitic groups. P.

Clitics can attach to material already

containing

clitics, but affixes cannot. If we compare the set A morphemes, clear

cases

of

inflectional

which seem to be

prefixes,

fairly

with theset

B

morphemes on each of the above six criteria, it would appear that

the

set

B

morphemes

behave

somewhat more

like

clitics:3 A.

Set

A

prefixes

transitive verbs. nouns

and

adjectives,

can only

appear

on

nouns

and

Set B morphemes can appear

transitive

verbs

as

well

as

on on

intransitive verbs, and certain minor

word classes. B.

I am not aware of any arbitrary gaps in the set of Set B + Predicate combinations.

A possible gap in

the set of Set A - Noun combinations would be with those noun stems of class Z (nouns which are never

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

156

possessed, section 3.4.2.7). C.

There

are

few

(if

any) cases

Predicate combinations exhibit idiosyncracies. such

However,

idiosyncracies

combinations.

where

Set

B

morphophonological

there are a number

with

+

Set

A

-

of Stem

These include the unexpected use of

in- rather than nu- on nouns (section 3.4.1.1) and the

use of the preconsonantal prefixes

initial 5.2.1, forms

words.

on

As will be discussed in

section

there are also some idiosyncracies in of the first person singular set

A

used on vowel initial transitive verbs. also

vowel

the

prefix

One might

include here the rather idiosyncratic

forms

of- the various allomorphs of the different set prefixes.

There

is

little

or

no

A

allomorphic

variation in the forms of the set B morphemes. D.

I

am unaware of any semantic idiosyncracies

with

either the set A prefixes or the set B morphemes. E.

Set

A - Stem combinations are always

units by syntactic operations. that

the

same

is

true of

treated

as

However, it seems Set

B

+

Predicate

combinations (assuming that set B morphemes are in fact proclitics as I have suggested).

Thus, set A

prefixes and set B morphemes do not differ on this point.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

157

F.

As

will be discussed below,

set B morphemes

can

attach to material already containing clitics (the incorporated movement morphemes). prefixes

can

only be attached

However, set A directly

to

the

stem. Thus,

while set A and set B morphemes do not seem to differ

with

respect

to

criteria D and E,

they

seem

to

differ

significantly with respect to the other four criteria. set A morphemes seem to be fairly prototypical affixes with respect to all six criteria. assume

throughout

the

The

inflectional

Therefore, I will

rest of this work that

the

morphemes are not prefixes but rather are clitic

set

particles.

This conclusion makes sense in another respect also: that

in

most

cases the forms of the

set

B

namely

clitics

identical to the forms of the Independent pronouns

I

consider

particles,

the

however,

orthographic

set I

tradition

B

will

morphemes continue

to to

be

Even clitic

follow

of writing them together

are

while in

many cases the forms of the set A prefixes are unique. though

B

with

the the

verb word. When

the

intransitive lengthened

set verb,

except

B clitics the in

appear

initial the

third

on

vowel person

a of

vowel-initial the

stem

plural.®

lengthening does not take place in this latter case

is The

because

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

158

the

third

person plural clitic and the

stem

undergo

the

contraction rule described in section 2.2.2 (in spite of the fact

that the third person plural clitic ends in a

stop

in

point

may

perhaps be additional evidence that the set B morphemes

are

clitics

many of these dialects).

Since

this

boundaries.

rule

This latter

glottal

generally

vowel

across

word

A sample paradigm is shown in (3) using

intransitive verb root -ok *to enter, short

applies

go in, come in'.

of this root shows up,

for

example,

the The

in

the

action nominalization okeem.* (3) (a) x+in+ook-ik

'I entered*

(b) x+at+ook-ik

‘you (sg.) entered'

(c) x+ook

'you (sg. formal) entered*

laa

(d) x+0+ook-ik

'he/she/it entered'

(e) x+oj+ook-ik

*we entered'

(f) x+ix+ook-ik

'you (pi.) entered'

(g) x+ook

'you (pi. formal) entered*

alaq

(h) x+o7k-ik There

are

'they entered' (
PERFV+2PL.B+go+bathe-DEPENDENT ‘we went to take a bath' (f) x+e7+b’e+chakun-oq PERPV+3PL.B+go+work-DEPENDENT ‘they went to work' (g) x+0+ee+k'aam-a

rii teem

PERFV+3SG.B+go+bring/take.PASS-DEP the bench ‘Someone went to get the bench.'

(NahualA

dialect, Mondloch 1981:147) In

when

the

incorporated movement clitic appears on a passive stem,

the

clitic

(29d)

and

(29g)

it can be

unexpressed.

shown

that

indicates the movement of the agent or

not the patient or theme, is

seen

and

not

Thus, as

experiencer,

even if the agent or (29d) can only be

‘would

that

he

go

experiencer

interpreted to

be

as

called'.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

183

Similarly,

(29g)

can

bench' (lit., 'the

only mean 'someone went to

bench was gone to be gotten*).

be interpreted as 'the bench examples

of

Chapter 3.

get

this

It cannot

wentto be gotten'.

were seen previously

the

Two other

in example(55)

of

In that example x+n+ee+k'aam-a chi+uloq can only

mean 'someone went to bring me here' (lit., I was gone to be brought

again hither'),

Similarly, went

not 'I went to be

x+n+ee+yaak-a

to get me up' (lit.,

brought

chi+uloq can only

mean

'I was gone to be

here*. 'someone

raised

again

hither'), not 'I went to be raised up.' Notice in (29e) that when the clitic e7 appears a

vowel-initial verb stem,

This

could

Movement than this, the

be

taken as

the contraction evidence

that

rule

the

before applies.

Incorporated

morphemes are clitics rather than affixes.

Other

there is not a whole lot of evidence one way

other for treating the Incorporated Movement

as clitics or affixes.

or

morphemes

About the best than can be said

is

that there are no arbitrary gaps in the set of

combinations

and

or

that

there

idiosyncracies. significantly

are These

from

no

morphophonological

morphemes

do not

inflectional affixes

Zwicky and Pullum's other three criteria.

seem with

semantic to

differ

respect

to

Nevertheless, the

the fact that the Incorporated Movement morphemes originated as intransitive verbs is quite transparent,

and this

would

not ordinarily be the case with inflectional affixes.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

184

In

any

case,

Movement morphemes must

also

if one decides are clitics,

that

the Incorpocated

then the set B

be clitics since they precede

Movement

clitics.

the

morphemes

Incorporated

An additional possible bit of

evidence

for treating the set B morphemes as clitics is that

glottal

stop

to

insertion (section 4.2.1.1) sometimes applies

Incorporated

Movement

Chiguimula.

Thus, for example, the verb x+ln+il+ch'ajan-oq

'I

came

to

morpheme ul (~ il)

wash*

has

x+in+i71+ch'ajan-oq. glottal

stop

been

Since,

recorded

in

Santa

the

with

Maria

the

form

as discussed previously, this

insertion is otherwise only

known

to

occur

across word boundaries, this could be taken as evidence that the

set B morpheme must be cliticized to

the

Movement morpheme rather than prefixed to it. given

that

originally

the

Incorporated

Movement

intransitive verbs,

favor

the suggestion that was made in

the

applies to these may

origin

of

were

the fact that glottal

sometimes

about

Furthermore,

morphemes

insertion of

Incorporated

initial

vowel

be

stop

evidence

section

in

4.2.1.1

lengthening

in

intransitive verbs. 4.2.2 Derivation In

this

affixes

which

These

are

section can

we will look

be added to

-inag

‘LOCATIVE/INSTRUMENTAL

four derivational

intransitive verb

‘PERFECT NOUN'

at

-eel

STATUS',

stems. -ib'al

‘AGENT NOMINALIZATION',

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

185

-eem

ik

'ACTION NOMINALIZATION',

and -isa

'CAUSATIVE1.

As mentioned in section 4.2.1.3, the suffix -inaq marks one

of

Status, -inaq The

the three status

categories/

on

verbs.

intransitive

namely

the

Perfect

Specifically, the

suffix

derives Perfect Participles from intransitive

verbs.

Perfect

predicate

Participle

which,

like

can any

function other

as

a

non-verbal

non-verbal

predicate,

exhibits subject agreement by means of the set B clitics but does not have a tense/aspect/mood clitic.

When the Perfect

Participle is used as a predicate, it can be translated into English

as

a

verb in the Perfect

English Perfect construction,

aspect.1*9

however,

Unlike

the Quichd

the

Perfect

Participle is not further marked for tense and therefore can be Past the

translated in English into either the Perfect (Pluperfect), context.

A

paradigm

Present

Perfect,

or Future Perfect according is

given

in

(30)

using

to the

intransitive verb stem -b'iln 'walk'.1® (30) (a) in b'iininaq (b) at b'iininaq

'I have/had/will have walked* 'you (sg.)

have/had/will have walked'

(c) b'iininaq laa 'you (sg. formal) have/had/will have walked' (d) (are7) b'iininaq *(he/she/it)

has/had/will have walked'

(e) oj b'iininaq

*we have/had/will have walked'

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

186

(£) ix b'iininaq

'you (pi.)

have/had/will

have

walked' (g) b'iininaq alaq

'you (pi. formal)

have/had/

will have walked' (h) e7 b'iininaq

'they

have/had/will

have walked'

In

addition

to indicating the

Perfect Participles,

Perfect

aspect,

some

in particular those derived from verbs

which assign patient-like semantic roles to their can also function as stative non-verbal

subjects,

predicates.

examples are seen in (31) using the intransitive

verb

Some stem

-kam ‘die'. (31) (a) at kaminaq 'you have died; you are dead'

In

some

translated

(b) e7 kaminaq 'they have died; theyare

dead*

cases these stative Perfect Participles

are

into

English in the Progressive

best

aspect.1’'

An

example is seen in (32). (32) war-inaq

ri

sleep-PERFECT

ak'aal

the child

'The child is sleeping' Example

(32)

can

also be translated

slept',

'the

child has fallen asleep',

as

'the

child

and 'the child

has is

asleep', all of which indicate a state which persists at the present

time and which resulted from a change of

state

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

in

187

the past.

Thus,

all of these translations are

with the notion of perfect aspect. can

Another example of this

be seen with the Perfect Participle

when

used

as a predicate,

consistent

is often

peet-inag,

translated

which,

's/he

is

coming'

or 's/he is on his/her way' rather than

come'.

This makes sense considering that in English, 's/he

has come* generally means that s/he has arrived Quichd Intransitive verb root -pee(t), English

verb

direction1, here

is

Quichd.

not 'to arrive here'.

expressed

here.

however,

can only mean 'to start

The

unlike the

out

The notion of

by the intransitive verb

has

root

in

this

arriving -ul

in

It should be easy to see how from a Perfect form of

~pee(t), out

come,

's/he

whose

literal meaning would be 's/he has

in this direction',

started

it could be inferred that s/he

is

coming or that s/he is on his/her way. The

stative

Perfect Participles can also be

attributive adjectives. never

The Perfect

Participle,

takes the Attributive suffix nor the

Degree

used

as

however, suffix.

An example is seen in (33). (33) jun one

kam-inaq

tz'i7

die-PERFECT

dog

'a dead dog' Even

though

the

Perfect

attributive adjective,

Pariciple

can

be

used

as

an

it is different from ordinary Qulchd

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

188

adjective stems in that,

with a few exceptions,

dexive

abstract nouns and versive intransitive

them.

Note,

derived

in any case,

from

a

Perfect

one cannot stems

from

that if a versive stem could be Participle,

its

meaning

would

probably be the same as the meaning of the intransitive stem from

which

Stative

the

Perfect

Participle

was

derived

Perfect Participles can also be used as

anyway.

nouns,

as

seen in (34). (34) iwiir yesterday

xmuug

ri

kaminaq

s/he.was.buried

the

dead(.person)

'Yesterday the dead person was buried' The Locative/Instrumental suffix -ib'al derives a

noun

which denotes either an instrument for performing the action denoted by the verb stem or a place where the event

denoted

by

of

the

have

the

the

suffix

verb stem takes place. is

The initial

vowel

often dropped in those dialects which

vowel dropping rule, and it often undergoes vowel harmony in those

dialects

which have the vowel

harmony

rule.

Some

examples are seen in (35). (35) (a) atin-ib'al bathe-LOC/INST 'bath, place for bathing'

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

189

(b) war-ib'al sleep-LOC/INST 'bedroom, dormitory' (c) el-eb'al leave-LOC/INST 'exit' (d) ok-ob'al enter-LOC/INST 'entrance' (e) oq'-ib'al cry-LOC/INST 'something lamentable' Most

such

nouns are o£ class la since the final

lenthened in the possessed form.

vowel

is

A few such nouns, such as

the ones seen in (36), have idiomatic meanings. (36) (a) u-peet-ib'aal 3SG.A-come-LOC/INST 'his/her arrival* (b) u-b'iin-ib'aal 3SG.B-walk-LOC/INST 'his/her way of walking' There are also a few such nouns which have irregular

forms,

such as the ones in (37), both of which are derived from the

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

190

irregular intransitive verb final

form

-wo7q

~

-va7

-wa71k ~ -wi7k, -wo7qoq,

'to eat*

Dependent

Dependent

non

(Plain

phrase-

phrase-final

form

phrase-final form

-wa 7ga). (37) (a) weeb'al 'instrument for eating, plate' (b) wo7gib'al 'place for eating, dining room' The Agentive Nominalizatlon suffix -eel derives a

noun

which denotes the "doer1* of the action described by the verb stem.

Actually

the name Agentive Nominalization is not

very apt name since this suffix can also derive nouns

a

which

denote patients or themes rather than agents when it appears on

intransitive verb stems which assign patient- or

like semantic roles to their subjects. when

this

-Jb'iin

suffix

'walk',

traveller,

is

added to

it derives

passenger'.

example,

in

proverb

the intransitive verb stem 'walker,

However,

to

one who dies'. the

for example,

an agent noun b ’iineel

intransitive verb stem -kam 'die', kameel 'dier,

Thus,

theme­

when

added

it formsa patient

the noun

This latter word is used, for

shown

in (38)

(from Mondloch

1978a:153, NahualA dialect).

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

191

(38) xaa just

uj

kam-eel,

xaa

uj

ok'ow-eel

1PL.B

die-AGENT

just

1PL.B

pass-AGENT

pa

wee

u-wach

uleew

in

the 3SG.A-face

earth

'We

are just diers,

we are just passersby in this

world' Colonial grammars of Quichd (e. often

treated

these Agentive

g.,

Brasseur

1862:78-117)

Nominalizations

as

"future

participles" such that a form like kameel was taken to 'one who has to die'. (38),

these

predicates

As can be seen in the first clause of

agentive

nouns

clause could be translated as

an

inference

predicate

and

interpretation second

can be

used

as

non-verbal

with a future tense interpretation:

going to die'.

mean

*we are just

this

first

(ones who

are)

However, this future tense interpretation is derived real

from world

the

knowledge.

is not obligatory,

clause of (38),

literal

meaning The

of

the

future

tense

as can be seen from

which clearly is intended to

that we ARE just passersby in this world,

the

mean

not that we

WILL

BE passersby in this world. There

are a few Agentive Nominalizations which

either

optionally or obligatorily take the prefix aj- (see

section

3.4.5).

derived

from

the

For example, intransitive

ajt’ilso7maneel 'tailor,

the Agentive Nominalizatlon verb

stem

seamstress,

-t'iiso7man

‘sew'

one who sews'.

is There

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

192

are

also

a

few non-productive Agentive Nominalizations in

-own, e. g. elaq'oom 'thief' (< -elag' *to steal'). The suffix -eem derives the Action Nominalizatlon intranstive verb stems.

from

Some examples are seen in (39).

(39) (a) b'iin-eera walk-NOM 'walking, a walk, trip* (b) eetz'an-eem play-NOM 'playing, game' (c) atin-eem bathe-NOM 'bathing, bath* In a few exceptional cases the suffix has a vowel other than /ee/.

All

such examples that I am aware of are

shown

(40). (40) (a) b'een-aam go-NOM ’going' (b) war-aam sleep-NOM 'sleeping, sleepiness*

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

in

193

(c) wi7-m

(NahualA, wa7-iim; Cantel, wa7-im)

eat-NOM 'eating' Note that in (40a), the intransitive verb root meaning 'go', which

usually

-e(e)(7)), -aam.

has

has

the

form -b'ee

form

b'een-

(or

in some dialects,

when it takes the suffix

One might also note that the intransitive verb

meaning 'come', peet

the

when

it

root

which often has the form -pee, has the form takes

any

suffix,

including

the

Action

Nominalization suffix, as seen in (41). (41) (a) ka+0+pee IMPERF+3SG.B+come

chwe7q tomorrow

'S/he is coming tomorrow.' (b) ka+0+peet-ik IMPERF+3SG.B+corae-PHRASE.FINAL 'S/he comes.' (c) peet-inaq come-PERFECT 'S/he is coming, s/he is on his/her way' (d) peet-eem come-NOM 'coming' There

is another suffix,

-ik,

which can also

derive

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

194

Action Nominalizations from intransitive verb intransitive

stems (including the majority

stems. of

Some

non-derived

intransitive stems and many derived intransitive stems) take the suffix -eem while others take the suffix -ik. while

Note that

the Action Nominalizatlon suffix -ik is identical

form

to

there

the Plain Status suffix

for

intransitive

in

verbs,

is a difference in that the Plain Status suffix is

phrase

final

suffix

suffix is not.

whereas

Some

the

Action

a

Nominalizatlon

examples of Action Nominalizations in

-ik are shown in (42). (42) (a) jiiq'-ik choke-NOH 'choking' (b) q'abar-ik get.drunk-NOM 'getting drunk' There

may

be some dialectal variation as to

take -eem and which take -ik. in

his

vocabulary

as the

Action

Nominalizatlon

verb

root -kam 'die' in the

whereas

usual

form in Momostenango

Chiguimula

stems

Mondloch (1978a) lists kameem

intransitive the

which

seems to be kamik.ie>

NahualA and

of

the

dialect

Santa

Maria

There are also some

verbs

which have both forms, e. g., peeteem - peetik 'coming*. is

not

clear whether there is any

difference

in

It

meaning

between the two forms.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

195

At least one intransitive verb root, -og' 'to cry', has an irregular Action Nominalizatlon: og'eej 'crying'. The suffix -isa derives transitive causative verb bases from intransitive verb stems.1* added

to any Intransitive stem which assigns a patient-

theme-like

semantic

productively assigns but

This suffix can probably be

be

role

added

to

to

its

any

subject.

It

intransitive

can

stem

an agent- or experiencer-type role to its

it can be added to a few such

stems.

The

or not

which subject,

transitive

bases derived by this suffix must undergo further derivation by means of STATUS',

suffixes such as

-x 'PASSIVE',

As usual,

-Vj

'TRANSITIVE

NON-PERFECT

-n 'INTRANSITIVE/ANTIPASSIVE', etc.

the subject of the intransitive stem

corresponds

to the direct object of the transitive causative form the

causer

shows

causative form.

up

as the

subject

of

the

while

transitive

Some examples are seen in (43-48).

(43) (a) -kam

'to die'

(b) -kam-isa-

'to kill'

(44) (a) -kow-ir hard-VERSIVE 'become hard'(b) -kowir-isa- 'make hard'

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

196

(45) (a) -poq'ow

*to boil' (intransitive)

(b) -poq'ow-isa-

*to boil (transitive),

to make boil1

(46) (a) -xojow

'to dance'

(b) -xojow-isa(47) (a) -atin

'to bathe* (intransitive)

(b) -atin-isa(48) (a) -b'iin

'to bathe' (transitive)

*to walk'

(b) -b'iin-isaAt

least

one

causative suffix. (49) (a) -war

*to make dance*

'to make walk, to drive'

verb takes an

irregular

form

of

the

This is shown in (49). *to sleep'

(b) -war-tisa-

*to put to sleep'

At least two intransitve roots have irregular forms when the causative suffix is added. (50) (a) -el

These are shown in (50-1).

'leave, go/come out*

(b) -ee-sa-

'to make

leave,

take/bring

out,

extract' (51) (a) -qaaj

*to fall, descend, go/come down*

(b) -qaa-sa-

*to lower, make descend, bring/take down1

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

197

NOTES 1. In NahualA, Zunil

Santa Catarina Ixtahuac&n,

the first person plural clitic in (2e)

form uj. plural

has

the

In Santa Catarina Ixtahuac&n the third person

clitic in (2h) has the form ee.

has the form e. of

and Cantel and

words,

In Cantel

it

When citing the morphological analysis

as in (102),

I will use the symbol "+"

represent

clitic

boundaries

represent

affix boundaries.

and the

symbol

Since laa and

to to

alaq

are

clitics,

the symbol "t" should be used with these too;

however,

since these particles are written

separately

in the orthography, this has not been done in (2). 2. In the following discussion I ignore the formal person

morphemes laa and alaq.

clearly enclitic particles.

These

second

morphemes

Furthermore,

are

since their

forms are identical whether they are functioning as set A forms or as set B forms,

and since their position is

different from that of the other set A and set B forms, they really fall outside the set A/set B dichotomy. 3. This variation in vowel length, of course, is not found in six vowel dialects. intransitive NahualA, third (e.

this

verbs

With monosyllabic vowel-initial

in Santa Catarina

vowel

Ixtahuac&n

lengthening occurs only

person singular form in the Imperfective g.,

k+0+ook-ik).

in

and the

aspect

In all other cases (e. g., the

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

198

forms

shown

below in 103) the vowel

these dialects.

stays

short

With polysyllabic vowel-initial stems

the vowel is always lengthened in the finite forms.

inflected

The Santa Maria Chiguimula dialect is generally

like Momostenango in that the vowel lengthening in

all

finite

inflected

intransitive verbs. is

in

irregular in

singular.

Thus,

of

vowel-initial

However, the stem -atin 'to bathe'

in

lengthened

forms

occurs

that

any

the

initial

finite form in for example,

vowel

the

is

third

not person

Santa Maria Chiguimula

has x+0+atin-ik 's/he bathed', whereas Momostenango and Santa Catarina Ixtahuac&n have x+0+aatin-ik.

All three

dialects have x+in+aatin-ik 'I bathed'. 4. In

Cantel

does

the third person plural form seen

in

(3h)

In very

slow

following

the

not undergo the contraction rule.

speech

a

clitic

e;

glottal

glottal

stop is

however,

stop.

pronounced

in fast speech there is no

Thus,

(3h)

is pronounced

such

xe7okik

~

xeokik in this dialect. Another example, using the stem -atin 'to bathe', is xe7atinik - xeatinik.

Futhermore,

with vowel-initial intransitive stems, the third person plural Cantel.

clitic

may

optionally have the

form

eb'

Thus, one can also say xeb'okik. The form eb*

seems to be obligatory with stems which begin with vowel

in

/e/.

For

some reason the stem

-atin

the

'bathe'

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

199

seems

to be irregular,

accept

the

xeatinlk set

B

however,

form eb'.

in that it does

The existence

of

forms

may constitute additional evidence morphemes

because

diphthongs

are

clitics

rather

like /ea/ do not

not like

that

than

the

prefixes

otherwise

exist

word internally in any dialect of Quich6. 5. As

noted in footnote 4,

this is

the only

form which

undergoes such vowel lengthening in monosyllabic in

Santa

particular

Catarina IxtahuacAn.

form seems to involve greater

lengthening in and

The reason

Momostenango,

the

may

be

underlying form of theImperfective aspect

particle ends in a vowel: length

normal

Chiguimula,

Santa Catarina Ixtahuac&n (among others?)

that

why this

than

Santa Maria

stems

/ka/.

Thus,

probably involves some kind of

the increased contraction

of

the vowels. 6 . Additional

evidence that at least the Past

Perfective

morpheme x is a clitic rather than a prefix can be seen from

the fact that this morpheme

word classes besides verbs. on

can appear

For example, it can appear

numbers when these are used in past time

expresions.

on other

adverbial

An example is shown in (i) (from the Santa

Maria Chiguimula dialect).

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

200 (i) jee

taq

wa7

x+lujuj

like D1STRIBUTIVE.p l this

junaab'

PERFV+ten

year

'ten years ago* 7.

I

will follow the convention of Comrie (1976) and

terms

with

initial

capital

Past

Perfective,

Imperfective, language

particular grammatical

letters etc.)

(e.

(presumably

g.,

to

indicate

categories.

Similar

terms spelled with initial lower case letters imperfective,

use

(e.

g.,

perfective, past, etc.) will be used for universal)

semantic

notions

of

tense,

the Cantel dialect the Imperfective aspect

clitic

always has the form /k/ and never has a vowel.

Thus,

aspect, mood, etc. 8 . In

example (6d) would have the form k+0+b'in-ik *he/she/it walks'.

In

consonant

cluster,

aspirated. always

forms like k b ’inik which have

an

intial

the initial /k/ is rather

heavily

In the officialized alphabet such forms are

written

Imperfective

as

aspect

where

clitic.

This

is

would

the

seem

to

indicate

that this morpheme does have a vowel in

dialect;

however, the one Cantel speaker that I worked

with

very clearly did not have a vowel in

(unless voiceless

one

were

vowel).

to interpret the It

such

aspiration

should also be noted

this

forms as

a

that

Colonial Quich6 the Imperfective aspect clitic had

in the

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

201

form q- before the first person plural set B clitic oj, as noted in colonial grammars such as that published by Brasseur (1862:21). Cakchiquel morpheme marks

and

This can also be found in

Tzutujil

(in

which

cognate with the Quich6 the

Imperative;

see

modern

languages

the

Imperfective

Dayley

clitic

1985:87-8

for

Tzutujil); however, as seen in (6e), the paradigm seems to

have

been

Nevertheless, have

regularized

in

modern

Quich6.

in the Santa Maria Chiquimula dialect

seen a text,

consisting of a

formal

I

"dialogue"

between two men about the history of the town, in which the q- allomorph of the Imperfective clitic seems to be consistently used before the first person plural set clitic.

B

Thus it may be the case that this allomorph is

still used in at least some dialects in certain

formal

styles. 9. The

Progressive

progressivity

particle and

imperfective reading.

does

tajin not

actually

just

force

indicates a

simple

Thus, in translating Quich£ into

Spanish, a form like kachakunik in the past tense would generally

be

imperfective) working;

translated into the in Spanish (i.

Imperfect

(=

e., trabajaba 's/he

past was

s/he used to work*) whereas a form like tajin

kachakunik

in

translated

into

the the

past

tense

Imperfect

would

generally

Progressive

(=

be past

progressive) in Spanish (i. e., estaba trabajando ‘s/he

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

202

was working'). in English,

Kote further that like the Progressive

the Quichd Progressive cannot generally be

used with statives, as illustrated in (i). (i) *tajin PROG

ka+0+r-aaj IMPERP+3SG.B+3SG.A-waat

**S/he is wanting it.' According

to Comrie (1976:32-40),

this is one of

the

defining characteristics of progress!vity. 10.

Sam

Colop

particles

(to appear) which can

inference.

mentions

also

force

two a

One such particle is w»7>

other

modal

future

tense

which he

says

indicates an action which should be executed very soon, a hope or conformity that something will take place that it is very likely to happen.

or

One of the examples

he gives (in the Cantel dialect) is shown in (i). (i) k+in+e7

wa7

IMPERF+3SG.B+go

PARTICLE

wr should go now.r He

also says that this particle can be

used

together

with the Necessitative particle, as seen in (ii). (ii) kine7 na I.go

wa7

NEC PARTICLE

‘Well, r have to go.'

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

203

The

other particle he mentions is ri7,

indicates

an

immediate

action

which he

which

executed as a response to a demand.

is

says

generally

The one example he

gives is shown in (iii). (iii) k+0+in-b'an-o

ri7

IMPERF+3SG.B+1SG.A-do-PHRASE.FINAL PARTICLE 'I will do it (right away).' I

know

nothing about these two particles

what Sam Colop says about them. them

in

although sentences Mondloch

given

do

show up in a in

(1978a:34}

Henne's

than

I have not encountered

any other dialect that I they

other

am few

(1980)

familiar of

the

with, example

dictionary,

gives the example shown

in

and (iv)

(NahualA dialect) without comment. (iv) k+in+b'ee-k,

ri7

IMPERF+lSG.B+go-PHRASE.FINAL

PARTICLE

'I'm going right away.' The forms of these particles are identical to those two of the demonstrative pronouns (see section however,

their

use seems to be quite

those of the demonstratives.

6.2.2);

different

In particular,

of

from

note

in

(iii) and (iv) that the phrase final suffix appears

on

the verbs (see sections 4.2.1.3 and 5.2.3) in spite

of

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

204

the

fact

that the verbs are immediately

followed

by

r i 7 . T h i s would hot be the case with the demonstrative pronoun ri7.

The comma in Mondloch's example suggests

that ri7 may be some kind of parenthetical which

would

be consistent with the

expression,

presence

of

the

phrase final suffixes. 11. In

the Cantel dialect the Imperative/Potential

never

has

a

vowel,

consonants.

appearing

According

to

as

ch

Mondloch

even

form is also used in certain

to

one action to another.

He refers

form

as

Imperative aspect".

From

in

this

Mondloch's

context

before

(1981:84)

Imperative relate

clitic

the

constructions

the

to

the

"relative

examples this use

of

the

imperative form is used in main clauses associated with a

temporal clause such that the time reference of

main

clause is a relative future with respect

time

reference

of the

temporal

clause.

to

Such

clauses are often introduced by k 'aa te7 'then, afterward,

just*

constructions,

or

maja7

however,

the

'not

yet’.

In

introductory

particles take an enclitic particle oq ~

the the main

later, these

adverbial

(o)qa.

Some

examples are shown in (i) and (ii) (NahualA dialect).

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

205

(i) k'aa to7+qa jU S t

0+saq-ir-oq 3SG .B+Wh ite-VERSIVE-PHRASE.FINAL

aree taq

k+uj+wa71ij-ik

when

IMPERF+1PL.B+ar ise-PHRASE.FINAL

'It will have just gotten light when we arise.1 (ii) majo7+qa ch+0+qa-wok-oo not.yet

IMP+3SG.B+1PL.A-buiId-DEPENDENT.STATUS

lee

jun

jah

aree taq

x+0+alax

the

one

house

when

PERFV+3SG.B+be.born

x*

q-alk'u7aal 1PL.A-child 'We had not yet built the house when our child was born.1 Note that in (i) the Imperative/Potential clitic chi is absent in

under conditions which will be discussed

the

text.

transitive. next

verb

form

is

Transitive verbs will be discussed in the

chapter.

(1980:112),

In (ii) the Imperative

below

The

appears

example to

in

(iii),

be an example

of

from this

Henne same

construction in the Cantel dialect.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

206

(iii)

majo+q

ch+0+ul

ri

a

not.yet

IMP/POT+3SG.B+arrive.here

the

MALE

Xwan

are taq

x+0+pil

John

when

PERFV+3SG.B+butcher.PASSIVE

ri

aq

the

pig

’Before John came, the pig was butchered.1 12. It is possible that the Potential forms do not exist in the Cantel dialect.

I checked forms like those in (19-

21) with a speaker of that dialect,

and he refused

to

accept any of these examples,

prefering instead to use

the

clitic

k

clitic

ch

imperfective

Imperative/Potential According

rather in

the Mondloch (1981:82),

than

such

the

sentences.

the Potential

form

(which he calls "affirmative mood") was more common Colonial Quichd than it is now. in

He also says that even

the cases in which it is still used in

language,

it

Imperfective discussed

is

always

form.

the

interchangeable

In addition to

in the text,

in

the

modern

with

the

environments

Mondloch (1981:83)

also

says

that the Potential form may be used following k'aa ’then,

later,

constructions"

afterwards, using

the

just1 and Irrealis

in

te7

"optative

Particle

taj.

Examples are shown in (i) and (ii) (NahualA dialect).

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

207

(i) k'aa te7 later

ch+in+wa7-ik POT+1SG.B+eat-PHRASE.FINAL

*111 eat later.' (11) ch+pee

ta lah

POT+COme

IRR

2SG.FORMAL

'May you (sg. formal) come.1 It should be noted/ can

only

however,

be used in these

that the Potential environments

to

form

indicate

future tense. 13. The

verb

-kuun

in (26)

is

pronounced

-k(o)win

NahualA, Santa Catarina IxtahuacAn, and Cantel. construction

illustrated

in

(26),

the

in

In the

tense/aspect

clitic on the two verbs must obligatorily be the

same.

Another characteristic of this construction is that the two verbs must have the same subject. 14. In NahualA,

Santa Catarina Ixtahuacdn,

preposition

che(e)

and Cantel the

would be used in place of

chi

in

(27). 15. The

analysis of -inag as a Status suffix is

(1986a).

A

similar

analysis

(1985:76-85) for Tzutujil.

is

given

Kaufman's by

Dayley

Personally, however, I have

a little difficulty with the idea of the Perfect suffix being a status suffix on the same order as -ik and -og. My qualms about this not only have to do with the

fact

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

208

that

the

Plain

phrase-final

and

Dependent

status

suffixes

are

while the Perfect suffix is not, but also

with the fact that the Plain and Dependent suffixes are inflectional

while

derivational.

the

Perfect

suffix

is

clearly

The Plain and Dependent status suffixes

do not change the

meaning nor the grammatical class

of

the stem they are

attatched

indicate

(in

conjunction

categories

such

dependency. indicate

The

with as

other mood

Perfect aspect,

is

not

inflectional special

normal verb. Perfect below,

morphemes)

and/or

inflectional

certain

and aspect is

kinds

of

is used

to

generally

an

However, the Perfect aspect in

marked

morphology

kind

Rather they

Perfect status suffix

inflectional category. Quich&

to.

by

any

special

kind

but rather by the

of non-verbal

predicate

of

use

of

a

instead

of

a

This special non-verbal predicate is the

Participle,

which,

as will be

clearly

is a type of deverbal adjective,

seen

DERIVED from

an intransitive verb stem by means of the suffix -inaq. Thus,

the

grammatical therefore,

suffix

-inaq clearly

does

change

class of the stem it is attached

the

to

and,

must be considered a derivational affix.

would rather have the notion of Status be reserved true

verbs

derivational

and

treat

affix,

the

Perfect

suffix

as

which has nothing to do with

system of status inflections.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

I for a the

209

16. In many dialects, dropping

rule

especially those in which the

vowel

applies, the first vowel of the

-inaq is generally dropped.

suffix

In those dialects in which

the vowel harmony rule applies, this same vowel usually undergoes vowel harmony. 17. Chafe

(1980) notes a similar phenomenon in Seneca

Japanese.

He

languages Present

argues

that

those

which are best translated into Perfect

are those which denote

have "perceptible consequences". best are

translated into the English

Quich6

This

also.

predicate

event

in

English

events

Those forms which are Present

that a

Perfect

like e7 kaminaq 'they have

Progressive have

for

such

and

which has

the

those beings

subject of this predicate.

had

on

in

Participial

died,

they

are

a

past

denotes a state which has resulted from

consequences

which

is probably what is going

Notice

(dying),

these

the

those which denote events which do not

consequences.

dead'

forms

and

very denoted

serious by

Also note that while

the this

predicate can be translated as a Perfect, it can NOT be translated as a Progressive: **they are dying'. other hand,

On the

the predicate e7 warinag can be translated

as a Progressive (‘they are sleeping'); seen

that

very

serious

and it can

here the past change of state has consequences for those

denoted

be

not

had

by

the

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

210

subject. 18. The this

form kamik can also mean meaning

the

suffix: JcamiJcaal. form

word often has

the

however,

with

abstract

noun

In some dialects (e. g., Cantel) the

kamikaal seems to be the only form possible

the meaning 'death'. fact

'death';

with

This could possibly be due to the

that in a six vowel dialect the word kamik

would

be indistinguishable from the word kamiifc 'today'. 19. In many dialects, dropping

rule

especially those in which the

applies,

-isa is often dropped.

vowel

the first vowel of the suffix In those dialects in which the

vowel harmony rule applies,

the initial vowel of

-isa

often undergoes vowel harmony.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

211

CHAPTER 5 MORPHOLOGY III 5.1 Transitive Verbs Transitive verbs differ from intransitive verbs in that they

not

object

only exhibit subject agreement

agreement.

Transitive

verbs

but

also

also

direct

differ

from

intransitive verbs in that only Intransitive verbs take phrase-final suffixes -ik and -og. take

a

Transitive verbs either

different phrase-final suffix

suffix

at

all.

verbs:

non-derived

transitive verbs.

the

or

no

There are two basic types

phrase-final of

transitive

(or root) transitive verbs and

derived

The basic structure of a root transitive

verb is shown in (1). (1) tns./asp./mood+obj.agr.+subj.agr.-root-phrase.f inal Transitive verb roots are all monosyllabic morphemes of form CVC. are

Derived transitive verb stems, on the other hand,

mostly

polysyllabic.

As

seen

in

(2),

the

structure of a dervied transitive verb differs from a

root

the

transitive verb in that there

is

no

basic that of

phrase-final

suffix. (2) tns./asp./mood+obj.agr.+subj.agr.-stem Derived

transitive

verb stems are of

three

basic

types:

those ending in /j/, those ending in the suffix -V*b*a7, and

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

2 12

other

steins ending in /7/.

The differences between

these

will be discussed below. 5.2 Inflection 5.2.1 Subject and Object Agreement Transitive

verbs do not mark subject agreement in

same way as intransitive verbs: agreement

the

rather than marking subject

by means of the set B clitics,

transitive

verbs

use the set A prefixes seen in table 2 of Chapter 3 for this purpose. but

Transitive verbs do also take the set B

these

are

used to indicate

rather

than subject agreement.

Quich6

has

The

set

direct

object

Thus it can be

an ergative/absolutive verb

clitics, agreement seen

agreement

B clitics mark subject agreement

on

that

system.

intransitive

verbs and direct object agreement on transitive verbs

while

subject agreement on transitive verbs is marked by the set A prefixes.

Note,

however,

ergative/absolutive case,

that the verb agreement is

in the formal second person.

intransitive subjects,

transitive

subjects

are all three marked by means

both

a

though,

of

because one can never

formal second person subject and

a

reflexives

and

reciprocals,

which

the

This does not

formal

person direct object in the same clause (except in the of

this

transitive direct objects and

enclitic particles laa (sg.) and alaq (pi.). lead to any confusion,

In

not

are

have second case marked

differently).

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

213

These are some slight differences between the forms the first person singular set A prefixes used on

of

transitive

verbs and those of the first person singular set A prefixes, shown in table 2 of Chapter 3, discussed prefix

in

used

section 3.4.1.1,

which are used on nouns. the first

on consonant-initial nouns

person is

As

singular

generally

nu-

although on a few nouns in some dialects this prefix has the form in-.

On vowel-initial nouns the first person singular

prefix

has the form w-.

On

consonant-initial

transitive

verbs,

however, the usual form of the first person singular

set A prefix is in-.

On vowel-initial transitive verbs the

first person singular

prefix

or

the

form inw-.

familiar,

may

either have the form

In the dialects with which I

am

most

it seems to be the case that the form w- is often

used on the vowel-initial derived transitive verb stem 'want'

w-

but that the form inw- is most common on

vowel-intial

transitive

verbs.

all

Nevertheless,

-aaj other

native

speakers seem to agree that either form can in principle used

on any vowel-initial transitive

form

in-

is the one which is practically

consonant-initial transitive verbs, the

Nahuala

dialect

Movement clitic Quichd language

verb.

testify the

when the

verb

(see section 5.2.4). to the fact that in form

nu-

was

used

Although always

be the

used

on

the form nu- is used in has

an

Incorporated

Colonial grammars older more

forms

of

of

the

extensively

on

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

214

consonant-initial transitive verbs.

According to

(1862:44,

apparently

47)

the

form

nti-

was

Brasseur obligatory

following the Imperfective aspect clitic, and it was in free variation with in- following the Past Perfective clitic

and

the Imperative/Potential clitic. It

might be noted that the allomorph in- of the

person

singular

corresponding

set A prefix is identical in form set

B

ergative/absolutive

clitic. systems

ergativity" (Silverstein 1976,

It

is

also

generally Dixon 1979).

first to

the

true

that

display

"split

One variety of

split ergativity that has be found in some languages is

one

in

the

which

there is a nominative/accusative

system

in

first and/or second person and an ergative/absolutive system elsewhere.

Thus,

one might consider the possibility

Quichd has such a split ergative verb agreement am inclined to think, the

that

system.

however, that the identity of form in

first person singular morphemes is fortuitous and

not mean that the verb agreement is split ergative. thing,

note

that

I

the

verb

agreement

does

For one is

not

nominative/accusative in the first person singular since the form

agreement

with

transitive subjects and intransitive subjects but also

with

direct

in

appears

objects.

not only as a marker

of

It can also be seen that there are

other

of the first person singular set A prefix

which

are identical to the forms used with nouns (or in any

case,

allomorphs

different from the set B clitic).

It would be unusual,

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

to

215

say the least,

to find a split ergative system in which the

split occured with only some first person singular forms but not

others.

allomorph the

Furthermore,

it can also be seen that

of the set A prefix is used depends

morphophonological

entirely

environment and has nothing

with either syntax or semantics. much

to

It does not seem to

sense to me to say that the verb agreement

ergative/absolutive

which

on first person singular

on do

make

system

is

vowel-initial

transitive verbs and consonant initial transitive verbs with an

Incorporated

Movement

ergative/absolutive

clitic,

but

on consonant-initial

without

Incorporated Movement clitics.

one

the

of

normally recognized

that

it

is

transitive

verbs

This would not

environments

for

not

be

"split

ergativity" (Dixon 1979J.1 A paradigm of a consonant-initial transitive verb third

person singular direct object is given in

(3)

with using

the transitive verb root -b'an 'do, make'. (3) (a) x+0+in-b'an-o

'I did it1

(b) x+0+aa-b'an-o

'you (sg.) did it1

(c) x+0+b'an laa

'you (sg. formal) did it'

(d) x+0+uu-b'an-o

's/he did it'

(e) x+0+qa-b'an-o

'we did it1

(f) x+0+il-b'an-o

'you (pi.) did it*

(g) x+0+b'an alag

'you (pi. formal) did it'

(h) x+0+ki+b'an-o

'they did it'

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

216

As can be seen in (3b,

d,

and £) the set A prefixes

which

have the form V(w)- when used on nouns (as shown in table

2

of Chapter 3) have long vowels when used on the verb in (3). In

Momostenango

lengthening appear

on

and

occurs a

Santa

Maria

whenever

Chiquimula

these three

transitive verb following

this

set the

A

vowel

prefixes

third

person

singular set B clitic. This

is also true in Santa

Catarina

Ixtahuac&n,

only when the Imperfective aspect clitic is Norman

William

(personal communication) says that NahualA

Santa Catarina Ixtahuac&n; in

used.

is

with

tense/aspect/mood clitic as in Momostenango and Santa The

like

however, the Nahuald forms given

Mondloch (1978a) seem to show the lengthening

Chiquimula.

but

one NahualA speaker that I

any Maria

consulted

on

this matter seemed to always have short vowels in the set prefixes

used on transitive verbs when the prefix

three or more syllables from the end of the if

A

appeared

word;

however,

one of the set A prefixes of the form V(w)~ appeared

in

the penultimate syllable of a transitive verb, the vowel was always lengthened following the third person singular set clitic

regardless

present. did it', a-

of which

tense/aspect/mood

be

was

Thus, for example, in the verb x+0+a-b'an-oh ‘you this speaker had a short vowel because the

is three syllables from the end of the word.

would

clitic

B

true when the verb has

the

The

Imperfective

prefix same aspect

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

217

clitic.

However, when this verb appears in non phrase-final

position,

the phrase-final suffix is dropped, and the set A

prefix is in the penultimate syllable. the Nahuald speaker I consulted the

set

A

these

In

threeset

preceded

vowel

of

did

it

all four of these dialects thevowels

of

prefix:

yesterday'.

In this case, then,

lengthened

x+0+aa-b’an

A prefixes are short

iwiir 'you

whenever

by any non-zero set B clitic.

's/he hits

(c) k+oj+i-ch'ay-o

'you

-ch*ay 'hit'.

you (sg.)'

(pi.) hit us'

the third person plural set B clitic

one

immediately

of

the

before

vowel a

initial

vowel

contraction rule applies.

are

(sg.) hit m e ’

(b) k+at+u-ch'ay-o

When before

'you

they are

Some examples

shown in (4) using the transitiveverb root (4) (a) k+in+a-ch'ay-o

the

set

initial

A verb

e7

appears

prefixes, stem,

or the

A paradigm is given in (5) using

the vowel-initial transitive verb root -il *see'.a (5) (a) xi7nwilo

(< /x#e7#inw+il+o/) 'I saw them'

~ x+e7+w-il-o (b) xa7wilo

(< /x#e7#aw+il+o/) 'you

(sg.)

saw them'

(c) xi71

laa

(
3PL > 3SG Some

additional

examples illustrating this

are

(12 ). (12) (a) aree

ri

achy-aab' x+e7+ch'ay-ow-ik

FOCUS the man-PL

PERFV+3PL.B+hit-F0C-PLAIN

'It was the men who hit him.'

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

507

(b) aree

ri

achy-aab1 x+in+ch'ay-ow-ik

FOCUS the man-PL

PERFV+lSG.B+hit-FOC-PLAIN

‘It was the men who hit me.' This

kind of situation makes it somewhat difficult

which NP,

to

say

the underlying A or the underlying 0, is actually

in S function in the Focus Antipassive construction. I of

will now turn to a more detailed syntactic

these phenomena.

It is a fairly simple matter to

thatQuich6 has a VP constituent the NP in 0 function.

analysis

This

show

which includes the verb and

can be done by

considering the

"weak crossover" phenomenon as illustrated in (13). (13) jachin* who

x+0+uu-ch'ay

ri

PERFV+3SG.B+3SG.A-hit

the

rj-achalaal 3SG.A-relative ‘Who* did his/herj relative hit.1 In this Quich6 sentence,

as in its English translation, the

WH-phrase cannot be coreferential with the possessor of head of the NP in A function.

the

This is easily accounted for

if we assume that the structure of this sentence is as shown in (14). (14) [ [ jachini ] [ [ INFL [ [ x+0+uu-ch*ay ] S 1 COMP SI1 VP V I e± ] ] ] I I ri NP NP DET

] [r-achalaal N

] [ ej ] ] ]] NP

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

508

The

empty

NP

dominated by VP is

the

trace

movement of the WH-phrase jachin into COMP. are

variables

and,

hence,

subject

of the

Such WH-traces

to clause

C

of the

"binding conditions" in (15) (cf. Chomsky 1982:20). (15) A. An anaphor is locally A-bound in its governing category. B. A

pronominal

is

locally

A-free

in

its

governing category. C. An R-expression (i. e., a name or a variable) is locally A-free. As discussed previously,

"bound" means coindexed with a

commanding constituent.

The term "free" means 'not bound'.

A

constituent is said to be "A-bound" if it is bound

constituent

in an "argument (or "A-") position".

"locally bound" to Y, also

binds

defined

X

to

as a category which is

Y.

"locally

A

a

If X

then there is no constituent Z

and is itself bound to

c-

is

which

variable

A-bound",

is

i.

e.

locally bound by an operator in a non-argument position such as COMP. in

The notion of govern: g category has been defined

a number of ways;

simple

definition

however,

given

for present

in (16)

(Chomsky

purposes, 1982:20)

the will

suffice. (16) The governing category of X is the minimal S

or

NP which contains X and a governor of X.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

509

The In

notion of government was defined in (67) of Chapter (14) the WH-trace,

which is governed

by

locally A-free in its governing category,

the

verb,

8. is

namely, S; and it

is locally A-bound by its antecedent, the WH-phrase in COMP. Therefore, such

as

the

WH-trace is a variable.

Empty

WH-trace are also subject to the

Principle"

(ECP),

categories

must

which states that be properly

categories

"Empty

Category

non-pronominal

governed

(Chomsky

empty

1986:17).

Proper government is defined in (17) (Chomsky 1986:17). (17) a

properly

governs

B

iff

a

0-governs

or

antecedent-governs B . Antecedent-government

holds

where

The chain relation

a

governs 6.

coindexing.

The

of a link (a,

notion of ©-government

B) of

is

a

chain

expressed

is defined

by

as in

(18) (Chomsky 1986:19). (18) a ©-governs B iff a is a zero-level category e.

(i.

a zero-bar category of X-bar theory: V, N, A,

P, etc.) that ©-marks 8 and a, 8 are sisters. The WH-trace in (14) is properly governed by the verb the

verb

both governs and assigns a 0-role to

the

since direct

object position. Like drop" position

Spanish,

(or in

and unlike English,

"null-subject")

language

Quich6 is a since

a tensed clause can be occupied

the by

"pro­ subject

an

empty

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

510

category. Chomsky

This empty category is usually taken, (1982:78-89),

to

be a pronominal

following

empty

category

called "pro", subject to clause B of the binding conditions. The

presence

languages of

of

pro in subject position

is

in

like Spanish and Quichd by a relatively rich

subject agreement (or SAGR) features in a

Since

licensed

English

features,

does

not

have such

a

tensed

rich

pro is not licensed in subject

set

INFL.

of

SAGR

position.

English does not have null subjects in tensed

set

Thus,

clauses.

In

Quichd the SAGR features show up at s-structure as the set A prefix

on

a transitive verb.

Since transitive verbs in Quichd also agree with the NP in 0 function by means of the set B clitics,

and since

the

"pro-drop” phenomenon also applies to NP's in 0 function Quich6,

I assume that pro may also appear in direct

position

and that the INFL node includes

features

(OAGR).

assigns

SAGR

governs the subject

case to that position.

OAGR,

govern the direct object position. assumed

will be seen presently, assign

assign

case to the direct

case

agreement

however,

and

does

not

In many languages it is

However,

in

the

direct

Quichd,

it seems that the verb itself

following Safir (1985:172-269), may

object

position

that a transitive verb assigns case to

object position that it governs.

not

object

in

object

position.

the

verb

does

Rather,

I assume that OAGR in

to a set B clitic on

as

if

INFL INFL

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

511

governs the VP. word,

it

Since the set B clitic is part of the verb

governs the direct object position and

its case to that position.

transmits

This is quite similar to Safir's

analysis of the free inversion phenomenon in certain Romance languages in which a subject clitic appears on the verb when the

subject NP undergoes inversion and is adjoined

verb phrase. may

the

Safir argues that if INFL governs the VP,

assign case to a subject clitic base generated

verb.

to

it

on

the

This subject clitic can then transmit its case to the

subject NP adjoined to the VP. The

other

possessor This

empty category in (14)

of the head of the NP in A

the

pronominal

function,

zachalaal.

empty category must also be pro:

its

governing

category

(here,

therefore it must be pronominal; since, is

is

the

it must be free NP

in

however,

A

in

function),

it cannot be PRO

as can be seen by the set A prefix on zachalaal,

case marked

previously, assigned

I

case

and,

therefore,

governed.

assume that a possessor NP is by

possessor

agreement

As

it

discussed

governed

(PAGR),

and

which

manifested at s-structure by the set A prefixed on the

is head

noun. Suppose now that the two empty categories in (14) coindexed. by

were

In this case they would both be locally A-bound

the WH-phrase in COMP.

But such a situation

seems to be ruled out in languages of the world.

generally One way of

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

512 talking the

about this has been to say that this would

"bijection

which

stipulates

operators traces)

a

one-to-one

and

Sportisch

1982),

correspondence

between

(such as WH-phrases) and variables (such and,

phenomenon.* any

principle" (Koopman

violate

other

thus,

accounts

for

the

weak

that is,

if the

WH-

crossover

If the two empty categories in (14) configuration,

as

were

WH-trace

in were

immediately dominated by S, and if the NP in A function were either immediately dominated by S or dominated by there

would

be

no

weak crossover

categories were coindexed: other

then

the

empty

the WH-trace would c-command the locally

Thus, this situation would not be ruled out by the

bijection principle.

(13)

if

empty category so only the WH-trace would be

A-bound.

and

effect

VP,

the

However,

since in fact the WH-phrase

possessor of the head of the NP in A

must be disjoint in reference,

function

in

the structure must

be

that shown in (14). I now want to consider the structure of sentences (4) and (10). unusual

like

As noted previously, this type of sentence is

in a number of ways.

Such sentences appear to

transitive,

but

the

Furthermore,

this

intransitive verb agrees with the NP

underlying O function;

verb

however,

underlying

A

clearly

intransitive. in

thus it is this NP which appears

be in derived S function. WH-phrase,

is

function,

to

The grammatical function of

which

would

be

appear to be the

is not overtly indicated

the

NP by

in verb

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

513

agreement, fact

preposition,

should

or relational noun.

be noticed about the

Focus

One

further

Antipassive

verb

form: it can only be used in WH-questions like (4) and (10), relative (6d)

clauses like (5d),

and

simple

(7).

and Focus

constructions

This verb form can never be

declarative

used

transtive or intransitive

like

in

main

any

clause

(except in the special case to be discussed below).

All of

these facts will have to be accounted for in some way. The fact that the verb in (10) does not agree with WH-phrase

calls

to mind the formation of

WH-questions

Palauan as discussed by Georgopolous (1985). Quichd,

is a VOA language.

in Quich6,

the in

Palauan, like

Transitive verbs in Palauan, as

show both subject and object agreement; however,

in Paulauan, when a WH-phrase that appears in a preverbal Aposition is the subject of a transitive verb, the verb lacks a subject agreement prefix. why this is the case, this

but let's assume for the moment

type of WH-question in Qulch6 has a structure

to that seen in Palauan. however,

unlike

some

distinction

between

instransitive

verbs.

agreement appears then,

Georgopolous does not consider

prefix to

assume

off

that

similar

Quich6 and other Mayan languages, languages,

have

morphologically

a

very

transitive

One cannot simply leave the of a transitive verb

the done in Palauan.

rigid

I will for

in

and subject

Quich6 the

as

moment,

that the structure of sentences like (4)

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

and

514

(10)

is

exactly what would be expected

for

a

transitive

sentence whose subject had undergone WH-movement except that there is no set A prefix on the verb, which entails that the verb is morphologically intransitive.

This is shown in (19)

(cf. 4). (19) [ jachin± [ [ INFL ( x+0+paq'-ow S* S I' VP [ ri sii7 ] ] ] [ et ] ] 1 NP NP I

assume

SAGR.

that INFL in (19) includes OAGR features

no

OAGR assigns case to the set B clitic, which in turn

transmits

its

case to the NP in

direct

object

Since there is no SAGR there is no set A prefix; verb form is intransitive.

hence

though

the

however,

intransitive,

assign two 0-roles like a transitive verb:

type role to the

position.

We do have to assume,

that the Focus Antipassive verb form, does

but

an

agent-

subject position and a patient/theme

role to the direct object position. be a violation of the 0-criterion.

type

Otherwise there

would

We have already seen

similar situation with Passives in section 8.4.1.

The empty

category in (19) is properly governed by the conindexed phrase in COMP;

thus there is no ECP

violation.

same

WH-

However,

since there is no SAGR, the subject is not casemarked. latter

This

point has one very desirable consequence and at time

disastrous

another

consequence

which

would

for the hypothesis being considered

a

seem

to

here.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

the be The

515

desirable Focus

consequence

Antipassive

involving

is that this could explain

verb

form only

WH-movement:

if

a

appears

non-empty

in NP

why

the

structures

appeared

in

subject position, it would not be casemarked, which would be a

violation of the Case Filter;

and if any empty

besides WH-trace appeared in subject position, be

properly governed.

category

it would not

Thus there would be a violation

the ECP.

The disastrous consequence,

structure

(19)

of

however, is that the

cannot possibly be correct

since

WH-trace

must be casemarked (Chomsky 1981:175-83). There however.

is

a

way of getting

around

this

difficulty,

First'of all, let's reconsider Focus constructions

like the ones seen in (6d) and (7).

At first glance it

might be assumed that focussed constituents are simply moved into

COMP

as

are

WH-phrases.

However,

there

is

some

evidence that the clefted constituent is in fact in a higher clause.

To see this,

we have to reconsider negation.

was discussed in section 8.2,

simple negation in Quich6

indicated by the Negative Particle before the predicate the Irrealis Particle taj after the predicate.

As is and

We have also

seen that the Irrealis Particle loses its final consonant in non phrase-final position.

Now consider Focus constructions

like the one in (20).

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

516

(20) are7 s/he

x+0+ch'ay-ow.

ri

achli

PERFV+3SG.B+hit-FOCUS

the

man

'S/he was the one who-hlt the man.1 In such a sentence the focussed constituent may be and

when

it

is,

the Irrealis

Particle,

if

directly before the verb as it does in (21),

negated;

it

appears

must have

its

phrase-final form taj. (21) ma

are7

NEG s/he

taj/*ta IRR

x+0+ch’ay-ow

ri

achii

PERFV++3SG.B-hit-FOCUS the man

'S/he was not the one who hit the man.1 This

suggests that the Focussed constituent is

non-verbal

predicate

and that there is a

actually

clause

a

boundary

immediately before the Focus Antipassive verb form.3 Thus, the structure of Focus constructions in Quichd is similar to that of relative clauses like (5d).

It is

not unlike the structure of cleft constructions in I

would like to suggest that the structure of

like

(10)

is

also like that of

relative clauses; not

simply

that is

moved to COMP

Focus

English.

WH-questions

constructions

to say, as was

also

and

the WH-phrase suggested

in (10) is

in (19),

but

rather is in a higher clause as is the Focussed (or clefted) constituent

in (21).

I would also suggest that the

Antipassive

construction still

involves

Focus

WH-movement,

but

that it is PRO that undergoes WH-movement rather than a non­

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

517

empty

WH-phrase.

Chomsky

has

argued

that

certain

constructions in English, such as the non-finite relative in (22)

involve

the movement of PRO

to

COMP

(see,

e.

g.,

Chomsky 1981:64-66). (22) I bought a booki [PROi [PRO to give e± to Mary 1] S' S In

(22)

PRO

structure. satisfy

both

ungoverned

PRO

give -in

clause It

can

positions;

A

and

only

do

thus,

clause

B

of

the

this

is

by

appearing

to move to COMP,

binding

PRO is exempt from

the direct object position is governed by has

d-

However, PRO is a pronominal anaphor which must

conditions.

Since

appears as the direct object to

where it is

the

ungoverned

coindexed with book by principles of control. in (22) is in a casemarked position;

the

in ECP.

verb, and

is

The WH-trace

however, it seems that

we have to allow for the possibility that traces of movement of

PRO

to

COMP can appear

in

non-casemarked

positions.

Consider, for example, (23). (23) The book [ PROi [ e* to be given et to Mary ] ] S' S is missing. In

this sentence neither of the traces of the

PRO is in a casemarked position: cannot

movement

the past participle

assign case to the NP position to which

it

of

given assigns

the theme role, and the infinitive to be given cannot assign

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

518

case to the subject position.

Thus,

let's assume that the

structure of (10) is not (19),

but rather something

more

like (24). (24) [jachin* t [INFL [x+at+ch'ay-ow [ ej ]]][ e± ]]] S' S I' VP NP NP PRO

cannot stay in subject position in (24) because

governed by a tensed INFL, it

it

is

so

it

must move

into COMP

is ungoverned and where it

is

coindexed

with jachinby

principles properly since

of control.

governed by the coindexed PRO

there

casemarked

The trace in subject

is

no SAGR in INFL,

COMP;

subject

is

however, is

not

a can

occupy this position is a trace left by the movement of

PRO

This,

Therefore,

position

the only thing that

to COMP.

position.

the

in

where

then, accounts for the fact that the Focus

Antipassive verb form can only appear in the three types

of

constructions we have seen which involve WH-movement:

since

the

be

subject position is not casemarked,

violation

of

the Case Filter

if

there would

anything

a

other than the

trace of a WH-moved PRO occupied that position. There still seems to be a analysis

illustrated

in

problem,

(24).

however, with

This

can

be

the

seen

considering the sentence in (25). (25) jachin* x+0+ch'ay-ow who *WhOi

ri

r j-a chalaal

PERFV+3SG.B+hit-FOCUS the 3SG.A-relative hit his/nerj relative.'

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

by

519

Here

the

WH-phrase

possessor (cf.

be

coreferential

of the head of the NP in direct

13).

given

cannot

But

with

object

position

it is not clear why this should be

the structure in (24),

the

the WH-trace would

since,

c-command

the possessor NP, so coreference should be possible as it is in

English.

One might consider the possibility

that

the

N P 1s in subject and direct object position in (24) should be reversed will

so that neither the WH-trace nor the possessor

c-command

coreference problem: assign

the other.

problem,

the

Focus

Though this

would

solve

the

one

new

have

to

it would introduce at least Antipassive verb form

would

an agent/experiencer-type role to the direct

position position.

and

a

This

patient/theme-type is

role

to

NP

object

the

the opposite of what we have

subject in

(13)

where the transitive verb assigns the agent/experiencer-type role to the subject position and the patient/theme type role to the direct object position.

But there is evidence

the Focus Antipassive verb form must assign a

theme/patient

role to a sister NP just like a transitive verb does. can

be

seen by considering again the kind of

that

Verb

This Phrase

Compounds that were discussed in sections 3.4.2.4 and 3.4.3, an example of which is shown in (26). (26) ( b'an-ow NP do-FOCUS

sii7 ] firewood

‘firewood making1

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

520

The

Focus

Antipassive verb stem in (26) clearly

assigned a patient role to sii7; sister

to the verb stem.

"subject"

position

ungoverned

since

construction.

If

the

patient/theme-type

role

is

if

be

and sii7 clearly must be a

It could not be in any

because there

must

it were,

no

set

verbstem

A

in

it

to a sister NP,

of

would

prefix (26)

kind

be

in

this

assigns

then it

a

must

do

likewise in (24). There is a way around this dilemma, all,

let’s

reconsider the idiomatic

however.

First of

construction

xuuk’am

awach 'you got accustomed’ that was discussed in section 8.3 (see

example 74).

As discussed previously,

a-wach

'your

face* is the direct object of the transitive verb -k'am bring/take’

in

nonreferential English,

this construction while the

subject

is

empty pronominal much like expletive

which

is

a

nonreferential

lexical

'to

it

a in

pronominal.

Expletive it in English can only show up as the subject of a verb like seem or of "weather verbs” like rain which

assign

case but do not assign a 0-role to subject position.

It can

be

Quich6

seen that the empty expletive pronominal in

idiom

the

is casemarked as indicated by the third person set

prefix

uu- on the verb.

It would also seem reasonable

assume that this empty expletive does not receive a However, transitive transitive

this verb

is

somewhat

-k'am

in

problematical this

idiom

is

to

0-role.

because an

A

the

ordinary

verb which usually assigns an agent role to

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

its

521

subject

and a theme role to its direct object,

as seen

in

(27). (27) x+0+uu-k'am

uloq

PERFV+3SG.B+3SG.A-bring/take

hither

'S/he brought it.' The empty expletive subject in the idiom, on the other hand, is clearly not an agent NP; it

and since it is nonreferential,

seems that it could not have any 0-role

might

lead

assigns

one

to think that the

at

transitive

all.

This

verb

-k ’am

two 9-roles in some constructions but only

others.

one

in

It might make more sense, however, to say that the

verb in the idiom does assign the usual two 0-roles but that while

the patient role is

direct

object

possessor

of

-wach,

assigned,

as expected,

the agent role is

assigned

-wach which is the one empty NP

this idiomatic construction.

the

to

the

position

in

Thus, I would suggest that the

structure of example (74) of Chapter

8 is as shown in (28).

(28) [

[ [INFL [x+0+uu-k'am [ [a-wach] S ’ S I* VP NPx N ’

NP*,

to

Ce 1111 [e]]J NPss N P 3

which functions as the possessor of -wach, is assigned

case by PAGR in NPx (which is manifested as the set A prefix on

-wach)

and is assigned a 0-role by the

verb.

NPx

direct object position receives the verb's other 0-role

in and

is assigned case by OAGR in INFL via the set B clitic on the

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

522 verb.

NPs in the subject position is assigned case by SAGR

but receives no 8-role. The unusual;

assignment of 8-roles in (28) however,

I

would

like

may

to

seemsomewhat

propose

that all

transitive verbs assign both of their 8-roles within the in a way similar to that seen in (28),

the difference being

that in ordinary transitive sentences the position to the

verb assigns the agent/experiencer-type role is

casemarked position.

VP

which not

Thus the agent/experiencer NP has

a to

move to subject position in order to be casemarked and avoid a

violation

of the Case Filter.

This is

illustrated

in

(29). (29) [ [ [INFL [ [x+0+uu-paq1[e±]] [ri sii7]]] S ' S I' VP VP NP NP [ri achiili]] NP In (29) the direct object is assigned its 8-role by the verb and

is

assigned case by OAGR via the set

subject NP, the

VP

on the other hand,

B

clitic.

receives its 8-role

and then moves to subject

position,

The inside

where

it

is

assigned case by SAGR. There is in fact some evidence for a NP position is

assigned an agent/experiencer-type role within

which

the

VP.

It will be recalled from section 5.2.1 that there are a pair of

enclitic

particles laa and alaq which mark

the

formal

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

523

second

person

singular and plural

discussed in that section, things,

respectively.

these clitics can,

As

among

appear after a transitive verb to mark a

with a formal second person NP in A function.

was other

agreement

An example of

this is shown in (30). (30) [ [ [ INFL [ [ x+0+paq’+laa [ e ] ] S' S I' VP VP NP PERFV+3SG.B+split+2SG.FORMAL [ ri NP the

sii7 1 ] ] [ e ] ] ] NP firewood

'You (sg. formal) split the firewood.' As

was discussed previously,

sentences like (30). function

is

appears

in

Furthermore,

there is no set A

special

position

clitic after

laa the

verb.

no non-empty NP can appear in subject position A way

account for this would be to say that in sentences

(30),

A

which

in such a sentence, not even the independent pronoun. to

in

The person and number of the NP in

indicated soley by the the its

prefix

like

the person, number, and case features in SAGR are not

assigned to the subject position, but rather are assigned to the clitic laa.

As discussed previously,

AGR features

in

INFL can be assigned to a clitic on the verb if INFL governs the

VP.

For this reason SAGR does not surface as a set

prefix.

Furthermore,

assigned

to

the

since the case features of SAGR

clitic,

the

subject

position

is

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

A are not

524

casemarked.

Thus,

position.

no

non-empty

NP can

appear

in

that

Since the subject position which does appear

the subject position is not assigned case or a

0-role,

in but

is governed by a tensed INFL, it is presumably the expletive empty category (EXE) discussed by Safir (1985). clitic

laa,

however,

The subject

transmits its case features

to

the

empty NP position within the VP.

Thus there is no violation

of the

this

CaseFilter.

properly

And since

governed by the verb,

empty

category

there is no EC?

is

violation.

We have to assume that the NP position to which the

subject

clitic

because

transmits

case

is

in

the

VP

as

shown

otherwise the position would not be governed by the verb and the clitic could not transmit its case features to it. It is now possible to resolve the dilemma we were faced with in the analysis of the Focus Antipassive

construction.

Givenwhat

structure

has

been concluded

transitive sentences in Quichd, that

the

about

the

of

I would now like to propose

structure of sentences like (10) and (25)

is

as

shown in (31). (31)

[jachin* [ PRO* [ [ INFL t [x+0+ch'ay-ow [e*] ] S' S' SI' VP VP NP [ ri r-achalaal [ ej ] 1 ] NP NP

Here

PRO moves out of the leftmost NP

][ e ] 1 ] ] NP

position,

which

is

governed by the verb and which is assigned the agent 0-role, into COMP,

where it is ungoverned and where it is coindexed

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

525

with jachin in the higher clause. INFL,

Since there is no SAGR in

the verb is intransitive (there is no set A

prefix);

and the subject position is not casemarked, so no lexical NP can appear there.

However, the subject position is governed

by a tensed INFL,

so the subject position is not an

for movement of PRO. clitic

to

lexical its

Since there is no casemarked

transfer case to the leftmost

NP can appear there either.

case to the set B clitic,

direct object position. in

(31)

head

NP

position,

two

in (25).

NP's

which transmits this to

The important new feature to

cannot

no

assigns

is that the leftmost NP and the possessor

these

observed

subject

OAGR in INFL

of the direct object NP do not c-command

Thus,

option

be

were

note

of

each

coreferential,

If these two N P ’s

the

the

other. as

was

coreferential,

they would both be locally A-bound by the same operator (PRO in

COMP),

which

would

be a violation

of

the

Bijection

Principle. One feature of the Focus Antipassive construction has

yet to be accounted for is the variable verb

that

agreement

pattern.

It was claimed above that at least one of the two

arguments

in

the Focus Antipassive

construction

must

be

third person and that the verb agrees with that NP which

is

higher on the hierarchy shown in (11).+ this is not quite correct. third

Actually,

however,

It is possible to have two non-

person NP's if one of them is a formal second

person

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

526

NP.

An example of this Is shown in (32). (32) laal

x+in+ch'ay-ow+laa

you.SG.FORMAL

PERFV+1SG.B+hit-F0CUS+2SG.FORMAL

'You (sg. formal) were the one who hit me.1 This demonstrates that in the Focus Antipassive construction there are two clitic positions just as there is in an Active transitive clause. that

There is a problem

here,

however,

we have to assume that the postverbal clitic

in

somehow does not assign case to the VP-internal NP that

is

expect,

assigned the agent role.

did,

position we

would

However, if the clitic laa did transmit case

to the NP position it governs, case features from SAGR in INFL. in

it would have to receive the It has been claimed

the Focus Antipassive construction does

SAGR features; (32)

it

(32)

contrary to fact, that a lexical NP could appear in

this position.

INFL

If

in

and

no

have

therefore, case cannot be assigned to laa in case features can be transmitted

position governed by laa. explaining

not

that

to

the

NP

But there is still a problem

in

the presence of laa in a sentence like

(32)

if

there

is no SAGR.

I will assume that in fact INFL in

the

Focus

Antipassive construction does contain the person

and

number of SAGR.

features of SAGR but merely lacks the case

features

If this is true, then we can still account for the

fact that there is no set A prefix in this construction

and

for

the

the

fact that the NP position which

is

assigned

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

527

agent/experiencer role cannot contain a lexical NP. also

We can

account for the presence of the clitic laa in (32)

assuming

that

this is licensed by the

features of SAGR.

person

and

by

number

This may make sense given that the Focus

Antipassive verb form assigns two 6-roles. With this in mind/ for

the

we are now

vferb agreement pattern of

construction.

We have seen that,

the

Focus Antipassive verb

the

normal preverbal set Bposition

position

in a position to account

which

the Focus

Antipassive

like a transitive

form hastwo

cliticpositions:

and another

may be occupied by laa

or

postverbal

alag.

I

further assume that the third person singular set B which is phonologically null, two clitic positions.® construction

both

is

will

clitic,

may appear in either of these

Since INFL in the Focus

has case features only in OAGR,

Antipassive not in

only one NP position in the VP can be casemarked. it

verb,

possible to mark the person and number of the VP-internal NP positions.

SAGR,

However,

features

Thus in a

of

sentence

like (6d), OAGR assigns case only to the direct object NP ri achii via the null postverbal clitic. two NP positions here:

Both

of

includes

there

are

the direct object position, which is

assigned the patientrole, occupied by theWH-trace,

However,

and the which

non-casemarked

position

is assigned the agent role.

these positions are governed by the two clitic positions.

Thus,

the

verb,

the two

positions also govern both of the NP positions.

The

which clitic direct

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

528

object

position Is occupied by a third person singular

and the clitic agreeing with its person and number (and

features

which transmits its case) is in the postverbal

position.

NP,

clitic

Since the WH-trace is coindexed with PRO,

which

is coindexed with the second person singular pronoun in

the

higher clause,

and

number

the clitic which agrees with its person

features

appears

in the

preverbal

set

B

clitic

position. The

non-null

set

B clitics can only

preverbal clitic position. the

second

person

set

Thus, B

clitic

appear

in

the

in a sentence like (10), still

appears

in

the

preverbal set B position even though in this case the set

B

clitic marks agreement with the patient/theme NP rather than with

the agent/experiencer NP like it does in (6d).

sentence

like

position

is occupied by the first person

this

only

(12b) it can be seen that the set

third

position; person

B

clitic

leaves the postverbal position available NP.

person plural set B clitic may not only

formal

sentence

(12b),

agreeing

with

the

may

appear

here.

with a null third person the

third

person

plural

for

a

However,

occupy

third person singular and

clitics

a

singular clitic,

clitic agreeing with the third person plural the

In

the

this second

Nevertheless, singular clitic NP,

is

still

acceptable since plurality is clearly marked in the NP, plural verb agreement is often optional in the third

and

person

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

529

anyway.

When a third person plural NP cooccurs with a third

person

singular

NP as in (12a) the third

person

clitic

may appear in the postverbal position

singular

thus

leaving

the preverbal position available for the third person plural set B clitic. \ Thus

the

verb

agreement

pattern

of

the

Focus

Antipassive construction follows automatically from the kind of

structure

Furthermore,

I

have

argued

for

in

this

construction.

we can now account for the fact that at

least

one of the NPs in the Focus Antipassive construction must be either third person or formal second person. quite

simply that if both of the arguments

person and non-formal,

were

non-third

then both of them would have to show

non-null set B agreement. position

The reason is

However, there is only one clitic

that can be occupied by a non-null set

B

clitic.

It would be impossible to have two non-null set B clitics in the same clause.

Thus, two non-third person non-formal NP's

would be impossible in the Focus Antipassive construction. The

Focus Antipassive verb form is used in

construction

that has not yet been discussed.

one This

other is

a

noun incorporation construction in which a simple unmodified patient

or

Antipasssive follows agent

theme verb

noun form,

immediately and an agent

follows or

a

Focus

experiencer

the incorporated patient or theme noun (unless or

experiencer NP is put into

the

preverbal

NP the

Topic

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

530

Position).

The

only examples of this construction I

have

seen in the dialects that I have worked on are shown in third

and

fourth

lines of the

example (16) of Chapter 3.

ritual

oration

the

shown

in

However, Mondloch (1981:248-258)

says that this noun incorporation construction is productive when itis preceded

by a manner adverbial.

Some examples

of

this are seen in (33) (NahualA dialect). (33) (a) utz good

k+in+paj-ow

tzaam

IMPERF+lSG.B+measure.out-FOC

liquor

'I liquor-measure well.' (b) utz good

ka+0+b'iin-isa-n

ch'iich'

IMPERF+3SG.B+walk-CAUS-FOC

lee

achih

the

man

metal

'The man car-drives well.' (c)

utz k+ee+pil-ow

waakax lee

good IMPERF+3PL.B+butcher-FOC cow

achih

the man

‘The man cow-butchers well.1 (d)

k'ax paint ful)

k+ee+nik1-o-n

alaq

IHPERF+3PL.B+7-TH.V-FOC2PL.FORMAL

uk’ louse 'You (pi.,

formal) lice-pick painfully.'

or

'You (pi., formal) hurt when you pick lice.'

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

531

(e) junaam same

k+ee+ket-ow

ajiij

IMPERF+3PL.Breut-FQC

sugar.cane

la7

lee

oxib'

achij-aab'

that

the

three

man-PL

‘Those three men sugar cane-cut in

the

same

way.' According to Mondloch (1981:254), also (34).

the incorporated noun can

be the Reflexive Relational Noun,

as

illustrated

in

This is the only case in which the incorporated noun

can be possessed. (34) b'alaj very

utz

k+at+ch'aj-ow

aw-iib'

good

IMPERF+2SG.B+wash-FOC

2SG.A-REFLEX

‘You self-wash very well.' The example in (35), from Henne (1980:17) is presumably from the Cantel dialect.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

532

(35) ri the

al

Chi71

xa

tajin

FEMALE

Cecilia

just

PROG

k+0+qa-sa-n

q'i j

IMPERF+3SG.B+descend-CAUS-ANTIPASS

day

r-uk'

ri

u-mya71

ri

3SG.A-with

the

3SG.A-man's.daughter

the

a

Xwan

ch+u-chi7

ri

pila

MALE

John

at+3SG.A-mouth

the

wash.basin

Cecilia is just passing time with John's daughter near the wash basin.' This

appears

construction

to

be

which

a

case

of

a

noun

has become lexicalized

incorporation as

a

compound

intransitive verb (cf. Mondloch 1981:256-7). From the examples in (33) it can be seen that the

verb

agreement pattern in the noun incorporation construction the

same

as

construction.

that The

observed

in

the

Focus

is

Antipassive

incorporated noun will always be

third

person, thus satisfying the requirement that at least one of the

arguments of the Focus Antipassive verb form

person.

The

verb will then agree with a non-third

argument if there is one, third plural plural, in

person

argument,

agreement

as in (33a). the verb will

the

third person

If there is no nonshow

third

person

with whichever argument is

third

person

as illustrated in (33c) and (33e).

(33d),

be

verb will also agree with a

As illustrated second

person

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

533

formal argument by means of a second person formal

enclitic

in addition to the usual set B agreement. In

spite

Antipassive

these

similarities

construction,

construction from

of

the

appears Focus

the

with

noun

superficially to be

Antipassive

the

Focus

incorporation quite

different

construction.

The

main

difference is that the agent or experiencer NP does not have to undergo WH-movement. Baker

(1988:76-145)

incorporation

within

GB

develops

a

theory

theory according

incorporation is a syntactic operation. an

incorporated

position

noun

must appear

in

noun

Baker argues

that

normal

in D structure in order to be assigned

However,

in

s-structure

the

noun

which

a

to

of

incorporated

argument a

8-role.

noun

appears

adjoined to a zero-bar constituent which governs it;

and

coindexed trace appears in its d-structure position. (1988:106-124) not

be

also argues that an incorporated

assigned

case.

constituent

governs

incorporated

noun,

As

the

long

as

d-structure

the

a

Baker

noun

need

incorporating

position

of

the

there will be no violation of the

Case

Filter. With this in mind,

it can be easily demonstrated

the

analysis of the Focus Antipassive

for

above

construction

is entirely consistent with the

incorporation

in

Quich6.

Consider a

noun

facts

that argued

of

noun

incorporation

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

534

construction

like

(33b).

The

relevant

structure of such a sentenceare shown

parts

of

in (36).

(36) ( ( ( (ka+0+b'iin-isa-n VP VP V V

]( ch'iich1 ]± ] N

[ laachii ] ] [ [ e* NP NP N The

agent

and theme 0-roles are assigned to

internal NP positions as usual. the

theme

NP

coindexed

the

] ] )

two

VP-

The head noun ch'iich'

is adjoined to the

trace.

the

verb

leaving

of

behind

Since the NP position which is

a

assigned

the theme role is governed by the verb with the incorporated noun,

this position does not need to be assigned case.

features clitic then

of SAGR in INFL are transmitted to one of the positions on the verb.

to

transmit

in

having

Antipassive

position

undergo

construction,

in

which

is

free

which is

this

WH-movement. as

was seen

position is not a casemarked position. of

would

to also

This, then, explains why a lexical NP

the agent to

NP,

the clitic is

the case features to the agent NP,

appear

without

trace

features

two

However, since the theme NP does not

be assigned case in (36),

governed by the verb. can

These case

ordinarily be transmitted to the theme

governed by the verb. need

The

PRO to COMP may appear in that

construction In

the

Focus

above,

the

agent

Therefore, only the position

in

Focus Antipassive constuction.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

the

535

It

should

incorporating

also the

be

noted

Reflexive

that

the

possibility

Relational

accounted for by this same analysis.

noun

is

also

The Focus Antipassive

constuction can not be used when the patient or theme NP the Reflexive Relational Noun. possessor

of

the

Reflexive



is

In a Reflexive sentence the Relational

Noun

coreferential with the agent or experiencer NP.

must

be

However, in

the Focus Antipassive construction such a sentence would

be

ruled out in the same way that a sentence

like (25) would be

ruled

patient

out

if

coreferential

the possessor

of

with the agent NP.

the Thus,

NP

when an agent

experiencer NP undergoes WH-movement in a reflexive the

than

incorporation

the

Focus

Antipassive

construction,

form.

however,

Relational Noun is adjoined to the verb. possessor

of

the

Reflexive

possible (indeed, necessary

anaphor)

used

a

noun

Reflexive

In this

Relational

commanded by the agent or experiencer NP. is

In the

or

clause,

normal active form of the transitive verb must be

rather

the

were

position

Noun

is

c-

Thus, coreference

since the possessor

between the possessor of the Reflexive

is

an

Relational

Noun and the agent or experiencer NP.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

536

MOTES 1. According to Mondloch use

(1981:223-5),

it is possible

the Focus Antipassive construction with

two

third person NP's if the patient or theme MP is by the Genitive Relational Noun. the

to non-

marked

Thus, sentences like

one shown in (i) are said to be possible

(NahualA

dialect). (i) in I

x+in+ch'ay-ow

aw-eh

PERFV+lSG.B+hit-FOCUS

2SG.A-possession

'I was the one who hit you.' This

construction can also be used even when one of the

NP's

is third person, as illustrated in (il) and (iii). (ii) aree x+0+ch'ay-ow

q-eh

s/he PERFV+3SG.B+hit-FOCUS 1PL.A-possession ‘S/he was the one who hit us.' (iii) ix

x+ix+ch’ay-ow

r-ee

you.PL PERFV+2PL.B+hit-F0CUS 3SG.A-possession lee

achih

the

man

'You (pi.) were the ones who hit the man.' It can be seen that the construction illustrated in (iiii), appears

involving

the

Focus

Antipassive

verb

form,

to be a true antipassive construction with

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

an

537

intransitive verb form, a

patient

NP

encountered

in

an

an agent NP in S function, and obliquecase.

constructions

speakers

have

(EnriqueSam Colop,

I

however,

related

any

to

encountered them

in

the

(1862:73). may

some such

closely

and they have also

also some Quichd examples of this type of

construction

the

personal communication).

Cakchiquel language,

Brasseur

of

reject

reported in Tzutujil (Dayley 1985:350-351).

in

never

and at least

been reported

sentences have,

have

like these in

Quichd dialects I have worked with, Cantel

I

Isuspect

There are construction

that

be a feature of older

Quiche which has disappeared in many,

been

this

varieties

of

but perhaps

not

all, modern dialects. 2. The

Bijection Principle is not without

See,

for example Safir (1984).

its

However,

problems. it seems to

work well

enough in simple cases such as the one

being

discussed

here, so such problems need not concern us at

this point. 3. Another similar the examples in

argument can be seen in sentences

like

footnote 7 of Chapter 2 (repeated below

in i). (i) (a) cho in.front.of

jaah house

k'oo be.someplace

wiih LOC.FOC

'It was at home that s/he was.'

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

538

(b) cho in.front.of

jaah

k'oo

wi

house

be.someplace

LOC.FOC

chwe7q tomorrow 'It is at home that s/he will be tomorrow.' As

was discussed in section 2.1.2,

dropped

word-final /h/

in non clause-final position

and Santa Maria Chiguimula.

in

is

Momostenango

However, as shown above in

(i), the /h/ is present in the word jaa (i. e., /jaah/) when

the

word

constituent. there

appears Thus,

at

this

the

end

of

too seems to

is in general a clause boundary

a

focussed

suggest after

that

focussed

constituents. 4. In

those

dialects

which

discussed in footnote 1, Antipassive manner

of

the

construction

this restriction on the Focus

construction speaking,

permit

can be

circumvented,

by having the

second

in

a

non-third

person argument appear as the possessor of the Genitive Relational Noun. to

the

These constructions actually conform

restriction

arguments,

however.

against

two

non-third

Since the Relational Noun phrase

itself is a third person singular noun phrase, appear

in

Antipassive

direct

person

object

construction

position in

in

accordance

it

the

can

Focus

with

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

the

539

restriction even when the agent/experiencer NP is tv -

third person.

,’

5. Alternatively, there

one could consider the possibility that

is no third person singular clitic and that

features

are

singular

NP

could

non-

' *

be

transmitted directly to a in some way.

done,

though;

third

It is not clear and I will not

AGR

person

how

this

pursue

this

possibility here.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

540

CHAPTER 10 CONCLUSION: ERGATIVITY IN QUICHE SYNTAX We

have

seen

morphologically

in

section

ergative

ergative/absolutive

5.2.1

language

that

Quichd is

since

verb agreement system.

it

has

an

Both NPs in

function and NPs in S function are crossreferenced on by means of set B clitics.

a

0

verbs

NPs in A function, on the other

hand are crossreferenced by means of set A prefixes.

Thus,

as far as verb agreement is concerned, NPs in 0 function are treated

like NPs in S function while NPs in A function

treated

differently.

This

then

is

a

are

manifestation

of

ergativity on the morphological level. We

also

ergativity observed the

saw in Chapter 7 some of the

is

manifested in

Quichd

the patterns of „overt vs.

ways

in which

discourse.

When

empty NPs and

patterns of the ordering of overt NPs with

observed

respect

the verb,

it was found that NPs in S function pattern

more like

NPs

function.

This again is an ergative/absolutive pattern.

was

argued,

in 0 function than they do like

following

DuBois

(1987),

that

we

NPs

to much

in

this is

A It a

reflection of universal patterns in discourse pragmatics. In this chapter I want to consider whether or not there are will

ergative patterns In Quichd syntactic be

shown,

based

on the

results

structure.1 of

the

It

preceding

section, that ergativity is manifested in Quichd syntax in a

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

541

very

fundamental

recognized.

way

that

has

not

been

previously

First, however, X will review some o£ the major

theories of syntactic ergativity that have been proposed. Dixon, 1979},

in his comprehensive study o£ ergativity (Dixon

introduces the notions o£ A,

syntactic-semantic

primitives.

0,

and S as universal

In deep

structure,

claims that syntactic rules operate in terms o£ a syntactic Dixon

category

recognizes

called

another

"Subject", level

called "shallow structure", the

operation

(Chomsky shallow

1957)

of



defined

optional

"singulary

the

structure,

and

"antipassive".

sense of Chomsky

surface structure.

1957)

In most languages,

which are morphologically ergative, continue Such

to

are

such

as

derivation

including

some

these latter operations

equate A and S as is done in

languages

At

("generalized

coordination and subordination take place in the of

by

transformations"

structure other syntactic operations in

(A,S>.

derived from deep structure

such as "passive"

transformations"

universal as

syntactic

Dixon

deep

said to operate in terms

structure. of

an

"S/A

pivot".

However, some languages operate in terms of an "S/0

pivot",

which

is

to say,

syntactic

shallow structures equate S and 0. to

be syntactically ergative.

passive

rules

operating

on

Such languages are said

The function of rules

like

and antipassive is often to put NPs which are in

A

(or 0) function in deep structure into derived S function in

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

542

shallow

structure

so

that they are

accessible

operating in terms of the S/0 (or S/A) then,

may

surface

pivot.

be manifested at two levels:

morphology

structure.

to

Ergativity,

at the

and at the sytactic

rules

level

level of

of

shallow

All languages which are syntactically

ergative

exhibit at least partial morphological ergativity.

However,

not

ergative

all

languages with ergative morphology

have

syntax. A

rather different notion of syntactic

presented

by

grammatical Binding

Marantz

(1984),

who proposes

relations within a development

theory.

syntactic

Marantz

structure:

structure1*,

"syntactic

structure".

L-s

logico-semantic

ergativity

recognizes

a of

three

"logico-semantic

structure relations

is among

a

or and

and

the

which

it

relations between a verb and the assigns

representation

of

semantic the

"surface

roles.

S

grammatical

of

constituents

sentence such as the relation between subject and

of

"l-s")

representation the

the

of

a

predicate

constituents structure

relations

of

Governmentlevels

(

(or "s") structure",

theory

is

is

among

to a the

constituents of a sentence such as the relation between

the

verb

its

phrase

object.

and

its subject or between

principle

verb

The mapping between l-s structure and s

is determined not by transformations, properties

a

1984:52-57).

structure

but rather by lexical

of the constituents and a very (Marantz

and

general

Surface structure

mapping is

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

a

543

representation configurations mapping

o£ of

the

surface

the constituents

phrase of

a

structure

sentence.

between s structure and surface structure

constrained

by a general principle which

The

is

states,

also

roughly,

that every s structure relation must have its counterpart in surface structure. structure

The surface structure counterparts of s

relations may be manifested as either

relations,

structural

casemarking, or agreement phenomena depending on

the language. In his discussion of l-s structure, 35) notes that in languages like English,

Marantz

(1984:32-

semantic roles of

the agent type are generally assigned by predicates to their logical

subjects,

patient

types

objects.

being or

are

semantic roles of

assigned

However,

arbitrary. language

while

this

by

verbs

the

to

theme

their

generalization

is

logical entirely

There is no a priori reason to suppose that makes the opposite generalization of

no

agent

roles

assigned by verbs to their logical objects and

theme

patient

roles

logical subjects.

being assigned by

predicates

to

there

are

languages, called

languages,

which

there

other languages, called

are

their

Marantz (1984:196-221) argues that there

are languages which do make the latter generalization. is,

or

That

"nominative-accusative"

choose the generalization

in

"ergative"

(la),

and

languages,

which choose the generalization in (lb).

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

544

(1) (a) Nominative-accusative languages agent

role

-

assigned

by

predicate

to

logical subject theme/patient

role - assigned

by

verb

to

logical object (b) Ergative languages agent

role

- assigned by

verb

to

logical

object theme/patient role - assigned by predicate to logical subject There

are some languages which have been

languages These

but which in fact are not by Marantz*s

are the

ergative,

languages which

which

Marantz

are

only

calls

ergative

generalization ergativity

is

languages in

are

(lb).

morphologically

that

only

choose

Marantz's

the

notion

a kind of syntactic ergativity since

based on the way semantic roles are paired with relations.

criteria.

For Marantz the

those

Thus,

ergative

nominative-accusative

languages with "type B" casemarking. real

called

it

of is

grammatical

However, it should be noted that this conception

of (syntactic) ergativity is not exactly the same as Dixon's notion of syntactic ergativity. Marantz's

theory

predicts

some

crucial

differences

between ergative languages (by his criteria) and nominativeaccusative

languages.

One

prediction

Involves

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

the

545

distribution of PRO in control constructions.

Since the

s

structure subject in a control construction must be PRO, PRO will

be

assigned the agent role in

a

transitive

control

construction in a nomninative accusative language, as can be seen in the English example (2) (Marantz 1984:199). (2)

Elmer persuaded Hortensei [ PRO± to buy a green porcupine I

In

an ergative language the theory predicts that PRO

in

transitive control construction must be assigned a theme patient

role.

Another

involves dative shift, role

other

that

themes

made

by

the

and In

or

theory

where an argument bearing a semantic

than theme or patient is expressed in

language. that

prediction

a

patients are

usually

the

way

in

the

expressed

a nominative-accusative language this

a non-theme/patient appears as an s structure

means object,

as illustrated for English in (3). (3) John gave Mary a book. In

(3) the constituent bearing the semantic role

of

namely Mary,

appears as the object of the verb giv:,

is

theme/patients

the

English.

way

are

ordinarily

goal, vhich

expressed

In an ergative language the theory predicts

in that

the non-theme/patient argument will appear as an s structure subject since that is the way theme/patients are expessed

in such languages.

A third

prediction

ordinarily involves

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

546

lexical

reflexives.

Marantz (1984:152-65) shows

how

theory predicts homophony between the reflexive and

the

passive

forms of a verb.

Thus, if a nominative-accusative language

has

reflexive verb form

a

lexically

himself', is

meaning

%he

washes

then the same form should be interpretable as 'he

washed1.

However,

in an ergative language it

is

the

agent of a transitive verb which shows up as the subject the passive form.

Therefore, the theory predicts that in an

ergative

language

himself'

should

(something)'. last

point

of

a lexical reflexive meaning also

It that

be

interpretable

as

'he

washes

*he

washes

should be noted in connection with for Marantz the passive

in

an

this

ergative

language is equivalent to what other people have called antipassive.

Similarly what Marantz calls the

the

antipassive

is, in an ergative language, equivalent to what other people would

call the passive.

In this dissertation I have

been

following the "traditional" usage of the terms 'passive* and 'antipassive' rather than Marantz's. Marantz

is

constructions

unable in

any

(syntactically) ergative;

to find

clear

cases

language

which

he

therefore,

of

suspects

he Is unable to

his prediction about the distribution of PRO. argues

Eskimo.

are This

borne

out in Dyirbal

demonstrates,

he

and

claims,

is test

However,

that his predictions about dative shift and

reflexives

control

Central that

he

lexical Arctic

these

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

two

547

languages

are

indeed

(syntactically)

ergative

(by

his

criteria) and, therefore, that such languages do exist. Turning now to Quichd, claim,

based

it would appear that one

on what was seen in the

last

Quichd is syntactically ergative by Dixon's syntactic rules which form WH-questions, and

cleft

(or

straightforward way

chapter,

that

criteria.

The

relative

Focus) constructions

used.

clauses,

operate

on NPs in S and 0 function,

NPs in A function.In order for such rules to NP in A function,

could

in

a

but not

on

apply to

an

the Focus Antipassive constuction must be

Since the Focus Antipassive construction Involves an

intransitive verb,

one could argue that the function of the

Focus Antipassive construction is to put an NP in underlying A

function

accessible clauses, operate

into

derived

S function so

to the rules which form and clefts.

that

it

WH-questions,

will

be

relative

Thus, it might appear that such rules

on an S/O pivot in Quichd,

which is

the

defining

characteristic of syntactically ergative languages according to Dixon. However,

it

is

not entirely clear

really do operate on an S/O pivot.

that

such

rules

It was seen in examples

like (8) and (9) that it is not always necessary to use Focus

Antipassive construction

clauses,

and

clefts.

in

Furthermore,

WH-questions, even when

the

relative the

Focus

antipassive construction is used, it is not absolutely clear

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

548

that the underlying A NP is put into derived S function.

It

was seen that the set B clitic on the Focus Antipassive verb form

sometimes

agent/experiencer

indicated NP

and other times

with the patient/theme NP. to

say

Also,

which

agreement

with

indicated

the

agreement

This makes it somewhat difficult

of the two NPs is actually

in

S

function.

given the kind of structure that l argued for in

Focus Antipassive construction,

the

it can be seen that neither

of the two NPs ever actually occupies the subject

position.

This

NP

might

lead

one to claim that there is

no

in

S

function in this construction if S is taken to mean 'subject of an intransitive verb'.

Thus, it is not absolutely clear

that Quichd is syntactically ergative by Dixon's criteria. It

is also not clear whether Quichd

ergative lexical

by

Marantz's

reflexives.

criteria. As

was

Quichd

mentioned

is

syntactically

does in

not

5.3,

have a

few

Absolutive Antipassive verb forms have a middle or reflexive meaning.

This is not,

phenomenon not

lexical

however,

an example of the kind of

Marantz is referring to. reflexives with

These verb forms

alternative

are

antipassive

"passive" in Marantz's terminology) interpretations.

(or

Rather

they are intransitive verbs with an agent or experiencer in

S

function,

alternative

only

small

reflexive or middle

seen in section 8.3, form

a

in Qulchd.

minority

of

which

interpretations.

NP have

As

was

true reflexives have a quite different

There is also no rule of dative shift

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

in

549

Quich*.

Thus,

predictions

it

about

is

impossible

complements

test

These and

are

non-finite

all

of

lexically

these

the

kinds

purpose

intransitive

intransitivized (i. of

involve

transitive

constructions

of

or

control

non-finite

clauses

that

However,

nominalizations

verbs

were as

of

of

then,

these

was

either

nominalizations

There

Quichd;

thus

are

no

transitive

of

there is no

control

way

to

Marantz's predictions about the distribution of PRO. be seen,

in

e., passivized or antipassivized) forms

verbs. in

shift

Quichd does have

discussed in sections 8.1.1.2 and 8.1.2. seen

Marantz's

lexical reflexives and dative

Quichd since these do not exist. constructions.

to

test It can

that it is impossible to determine by means

tests whether or not

Quichd

is

(syntactically)

ergative in Marantz's sense. Given chapter,

the analysis that was proposed in however,

it

the

is not clear that Quich6

previous is

either

nominative-accusative or ergative by Marantz's criteria. was

argued

Antipassive type

that

a

transitive

verb

verb form) assigns both

(and an

also

a

If this is true,

Focus

agent/experiencer-

role and a patient/theroe-type role to NPs in the

phrase.

It

verb

one could claim that Quichd does

not conform to either (la) or (lb).

On the other hand, the

fact

the

that

the NP which is assigned

agent/experiencer

role moves into the subject position in a simple

transitive

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

550

sentence could be taken to mean that Quich6 in £act conforms to (la) and, therefore, is nominative-accusative. But

if

correct, it.

a

I

and

analysis

preceding

chapter

type roles are assigned by verbs

within the VP in

that

Antipassive verb,

the

rather unexpected conclusion can be drawn

patient/theme

argued

of

Quich6.

constuctlon,

which

involves

of an

in

INFL which could assign case

position in this construction.

to

I

NP have

the

Focus

intransitive

follow from the fact that there are no case

SAGR

roles to

Furthermore,

some of the unusual properties

is from

have argued that both agent/experiencer type

positions

of

the

features

the

subject

One consequence of this was

claimed to be the fact that there is no set A prefix on Focus Antipassive verb form. suggests

But if all of this is true, it

that in Quich6 all NPs in S function,

say all "intransitive subjects",

are not in fact

that is

for example,

to

syntactic

subjects because they are dominated by VP rather than by Consider,

the

S.

the simple intransitive sentence in

(4). (4) [ [ (INFL ( x+0+b’iin [ ri achii 1 1 1 ( e ) ] ] S' S I* VP NP NP PERFV+ 3SG.B+walk the man 'The man walked.1 The VP.

verb in (4) assigns an agent role to the NP inside OAGR in INFL assigns case to the set B

clitic,

the which

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

551

transmits its case to the agent NP.

Thus the NP rl achil is

in a casemarked position inside the VP and there will be violation o£ the Case Filter in this position. is

no set A prefix,

SAOR in INFL. a

intransitive

Quichd

Thus,

impersonal

thus

cannot

in

S

no

move

to

NP

function"

subject

VP

may

in

an

at

s-

position.

in (4) for intransitive

similar to the structure of

The

clauses the

construction with a dummy subject seen

other languages. (4),

"NP

no lexical

like (4) is dominated by

illustrated

is

therefore,

the

sentence It

structure

is

Hence, the subject position in (4) cannot be

there.

structure;

Since there

it must be the case that there

casemarked position and,

appear

no

in

kind in

of some

However, as can be seen by the example in

this structure also applies to intransitive

sentences

in Qulchd which have agent (or experiencer) "subjects".

In

those languages which have impersonal constructions the non­ expletive argument of the verb can generally only be of

the

theme/patient type. The in

only

intransitive

manifested

not

way

Focus

however,

be

to

assume

sentences there was SAGR but that as

transitive sentence, would,

around this would

set

A

agreement

as

but rather as set B

it

explanation

it

would

was in

agreement.

section.

Since

that

analysis

a

This

effectively dismantle the analysis of

Antipassive construction that was argued for

previous

that

the

in

the

afforded

an

for all of the unusual characteristics

of

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

the

552

construction, that would be a pretty heavy price to pay. In

any case,

for

the

in

(4).

In

8.3we discussed idiomatic constructions

like

the

analysis

there is additional

evidence

of intransitive clauses illustrated

section

one in (7.2), which is repeated below as (5). (5) ( I I lk+0+uu-kam [r-iib' [ki-vach te]])]]Ie])] S' 3 I' VP NP NP NP NP 'They are congenial with each other.' The

subject

expletive

position is

which

occupied by

an

emptypronominal

is casemarked by SAGR (hence,

prefix) but does not receive a 8-role.

the

set

OAGR assigns case to

the NP in direct object position (riib' kivach) via the B clitic. object the

A

set

The verb assigns a theme-type role to the direct

NP.

The agent-type role is presumably assigned

possessor

Chapter 9.

of -vach just as it was in example

(20)

to of

It was claimed in section 8.3 that the Reflexive

Relational Noun could have a lexical NP as possessor just in case the NP in subject position was an expletive,

i.

e. an

NP which is not assigned a 8-role. Now consider the sentence in (6).3 (6) x+0+ch'aj-taj PERFV-t-3SG.B-t-wash-C.PASS 'The man washed himself.*

r-iib'

ri

achih

3SG. A-REF LEX

the

roan

(lit., 'The man's self

got washed.')

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

553

Note first of all that the man is understood to be the agent in this sentence. clause

to

One would expect the agent of a

show

up

as

the

Relational Noun -umaal, However,

it

possessor

of

the

Agentive

as was discussed in section

can be seen in (6) that there is

Relational ^Noun;

passive

and furthermore it is

no

8.4.1. Agentive

ungrammatical

for

there to be:one, as shown in (7).a (7) *xch'ajtaj riib' rumal ri achii. Thus,

we

have

assigned

to

directly

assume that the agent role to

ri achll as

Reflexive Relational Noun. in (5).

However,

Relational

in

possessor

if

of

is the

in (5) it was claimed that the Reflexive

Noun could only have a lexical NP

(6)

(6)

This is similar to what was seen

the subject NP were expletive. have

the

in

the

analysis

illustrated in (4) is correct.

possessor

if

But this is exactly what we of

intransitive

clauses

It must be the case that the

"NP in 8 function" riib' ri achii in (6) is dominated by and

receives the patient role from the verb.

casemarked by OAGR via the set B clitic. SAGR in INPL, the

position

NP

also

Since there is no

position is can appear

does

not

there.

casemarked.

Therefore,

Furthermore,

not receive a 8-role.

The

the

agent

assigned to the possessor of the Relational Noun. possessor

is

there can be no set A prefix on the verb; and

subject

lexical

It

VP

no

subject role

is

Since the

is in a casemarked position (as evidenced by

the

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

554 set A prefix on the Relational Noun), of not

the Case Filter. receive

a

there is no violation

Also since the subject position

0-role,

this licenses the

presence

does of

a

lexical NP possessor of the Reflexive Relational Noun. This,

then,

correctness argued,

constitutes

of (4).

(4)

does

Note,

addtional evidence however,

represent

the

intransitive sentences in QuichA,

that if, correct

has

ergativity: like

not

been previously

as I

then it can be

recognized

NPs in S function in QuichA are

the have

analysis

that QuichA is in fact syntactically ergative, that

for

concluded

but in a way

in

studies

of

syntactically

NPs in 0 function for the simple reason that they

both dominated by VP at s-structure.

of

are

NPs in A function are

treated differently syntactically because they are dominated by^'S- at s-structure.

This defines an

pattern in QuichA clause structure, be

ergative/absolutive

which could conceivably

another reflection of the universal ergative

patterning

in discourse pragmatics of NPs in A, S, and 0 function.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

555

NOTES 1. Some

of the ideas discussed in this chapter have

been

discussed previously in Larsen (1987). 2. This

sentence

was

elicited by. Glenn

Ayres

from

a

Speaker from Santa Catarina IxtahuacAn and is discussed in

Ayres

(1980:56).

consulted

did

(1981:141,

A speaker from

not accept

148)

such

Cantel

that

sentences.

I

Mondloch

claims that reflexive simple

sentences like that shown in (i) also exist,

passive in

which

the Reflexive Relational Noun is unpossessed. (i) utz good

ka+0+loq'-o-x

lib’

IMPERF+3SG.B+?-TH.V-PASS

REFLEX

'It is good that they (indefinite) love

each

other; it is good to love one another.' In (i) it appears that the finite passive clause is the "subject" passive Nouns

of

the

adjectival

predicate

utz.

clauses with unpossessed Reflexive apparently

correspond

to

those

Such

Relational

non-reflexive

passives where the agent 0-role is not assigned (i. e., simple passive clauses with no -umaal phrase). 3. Example Santa

(7) was judged ungrammatical by Ayres' Catarina

(1981:162),

Ixtahuac&n

however,

claims

informant.

(1980) Mondloch

that sentences like

below are grammatical in the NahualA dialect

(i)

(although

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

556

my Cantel informant rejected these also). (i) ka+0+ch'aj-taj

w-iib'

w-umaal

IMPERF+3SG.B+wash-C.PASS 1SG.A-REFLEX lSG.A-by 'I will finish washing myself.' I am not sure why this discrepency exists;

however, it

is interesting to note that Ayres' (1980:141) involve

a third person agent while all

(1981:162) agents. like

examples I

(i)

involve first

of

or



be analyzed as

in

Mondloch's

second

would also like to suggest that

should

examples

(ii),

person

sentences which

consistent with the analysis presented of (5) and consistent

is

also

with the analysis of (6) which I argue

for

below in the text. (ii) I ( C ( ka+0+ch'aj-taj S' S I' VP

[ w-iib' I w-umaal NP NP ( e NP

])])] ( e ]]] NP

The agent 6-role is assigned to the NP (w-umaal

Cel 11,

which is the possessor of w-iib'.

6-role

is (

assigned

the

entire

w-iib' ( w-umaal [ e 1 ] ].

about

this example,

Relational rather But

to

Relational What is

Phrase

problematical Reflexive

Noun shows first person singular

agreement

expected

is that

Noun

the

than the

however,

The patient

third

it should be rememberedthat my

person Cantel

singular. informant

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

557

preferred Reflexive

third

person

Relational

Noun

plural in

agreement (5).

It

on would

interesting to know what kind of agreement there be

in

sentences

like (5)

with

a

the

non-third

be would

person

possessor of -vach.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

REFERENCES

Alcal&,

Pedro

de.

1505.

Arte para llgeramente

saber

la

lengua arauiga. Granada. Aissen,

Judith L.

1987. Tzotzil clause structure. (Studies

in Natural Language and Linguistic Theory.)

Dordrecht:

D. Reidel Publishing Company. Alvarado Ldpez, Bspafiol.

Dr.

Miguel.

1975.

(Publicacidn

Ldxico Mddico

especial

del

QuichdInstituto

Indigenista Nacional.) Guatemala. Ayres,

Glenn.

1980. A note on Mayan reflexives. Journal of

Mayan linguistics 2, no. 1.53-9. Baker,

Mark C. 1988. Incorporation, a theory of grammatical

function changing.

Chicago:

The University of Chicago

Press. Bentivogllo, Paola. 1983. Topic continuity and discontinuity in discourse: a study of spoken Latin-American Spanish, in Giv6n 1983b:255-311. Braaseur de Bourbourg, livre

Charles fttienne. 1861. Popol vuh: le

sacrt et les mythes de

1 'antiquits

amdricaine.

(Collection de Documents dans les Langues Indigenes l'Amdrique Ancienne, 1.) Paris: Arthus Bertrand.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

de

559

Brasseur de Bourbourg,

Charles fttienne.

la Langue QulchSe.

1862. Graauoaire de

(Collection de Documents dans

Langues Indigenes de l'Amdrique Ancienne,

2,

pt.

les 1.)

Paris: Arthus Bertrand. Brody,

Jill.

1984. Some problems with the concept o£ basic

word order. Linguistics 22.711-36. Campbell,

Lyle.

1977.

Quichean

linguistic

(University o£ California Publications in

prehistory. Linguistics,

vol. 81.) Berkeley: University o£ California Press. Carbonell

Pastor,

Fernando.

(Publicacidn

especial

1973. del

Gram&tica

Instituto

Quichd.

Indigenista

Nacional.) Guatemala. Chafe,

Wallace

L.

de£initeness,

1976. subjects,

Subject and topic,

ed.

Givenness,

contrastiveness,

topics, and point o£ view, in by Charles N.

Li,

25-55. New

York: Academic Press. Chafe,

Wallace L. 1980. Consequential verbs in the Northern

Iroquoian languages and elsewhere. Indoeuropean Beeler,

studies,

American Indian and

paperB in honor of

Madison

3.

ed. by K. Klar, M. Langdon, and S. Silver, 43-

49. The Hague: Mouton Publishers. Ch&vez,

Adri&n Inds. (no date) Ki-chd ftib, escritura ki-chd

y otros temas.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

560

Chomsky,

Noam.

1957.

Syntactic

structures.

The

Hague:

Mouton. Chomsky,

Noam. 1970. Remarks on nominalization. in Readings

in English transformational grammar, ed. by Roderick A. Jacobs and Peter S.

Rosenbaum, 184-221.

Waltham: Ginn

and Company. Chomsky,

Noam.

by P.

1977. On Mh-movement, in Formal syntax, ed.

Culicover, T. Wasow, and A. Akmajlan.

New York:

Academic Press. Chomsky,

Noam.

1981.

Lectures on government and

binding.

Dordrecht: Foris. Chomsky,

Noam.

1982. Some concepts and consequences of the

theory of government and binding.

(Linguistic

Inquiry

monograph 6.) Cambridge: MIT Press. Chomsky, Noam. 1986. Barriers. (Linguistic Inquiry Monograph 13.) Cambridge: The MIT Press. Comrie, of

Bernard. 1976. Aspect, an introduction to the study verbal

Textbooks

aspect in

University Press. CoztAzar,

Julio.

and

related

Linguistics.)

problems.

(Cambridge

Cambridge:

Cambridge

[Reprinted with corrections, 1978).

1982. 62 / modelo para armar.

Barcelona:

Editorial Bruguera, S. A. (First published in 1968.)

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

561

Croft,

William,

no date.

Cross-reference,

catagoriality in Quichd. Davies,

William.

universal grammar Rosen,

ergativity and

Unpublished MS.

1984. Antipassive: Choctaw evidence for a

characterization, 2,

ed.

331-76.

in Studies in

by David M. Chicago:

relational

Perlmutter and Carol

The

University of

G.

Chicago

Press. Dayley,

Jon

P.

languages. Dayley,

Jon

1981.

and

ergativity

in

Mayan

Journal of Mayan linguistics 2, no.2.3-82.

P.

California

Voice

1985.

Tzutujll

Publications

in

grammar.

(University

Linguistics,

vol.

of

107.)

Berkeley: University of California Press. Dixon, R. M. W. 1979. Ergativity. Language 55:59-138. Dixon,

R. M. W. 1987, ed. Studies in ergativity. Amsterdam:

North-Holland. [Also published as Lingua 71.] Du Bois,

John W.

University

1981. The Sacapultec language. of

California.

Unpublished

Berkeley: Ph.

D.

Mayan morpheme structure and

the

dissertation. Du

Bois, final

John W. vowel

linguistics

1985.

question in Quichean.

Journal

of

Mayan

5, no. 1.1-71.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

562

Du Bois,

John W.

1987.

The discourse basis of ergativity.

Language 63.805-55. Edmonson,

Nunro.

American

1965. QuichA-Bnglish dictionary.

Research

Institute,

Publication

(Middle 30.)

New

Orleans: Tulane University. Edmonson,

Munro.

1967.

Classical QuichA.

Middle American Indiana, 5,

Linguistics,

ed.

ed.

in Handbook

of

by Robert Wauchope, vol.

by Norman

A.

McQuown,

249-67.

Austin: Univerlsty of Texas Press. Edmonson, of

Munro S. 1971. The book of counseli the

Research

QuichA Mayaof Institute,

Guatemala. Publication

the popol vuh

(Middle 35.)

American

New Orleans:

Tulane University. v

England,

Nora C.

1983. A grammar of Mam, a Mayan language.

Austin: University of Texas Press. England,

Nora C.

and Stephen R.

Elliot, to appear, (eds.)

Lecturaa sobre la linguistics mays. Fox,

David.1965.

Lecclones

elementales

en

gulchA.

Guatemala: Mlnisterio de Educacidn Pdblica. Fox,

David.

1968.

Review of QuichA-English dictionary, by

Munro Edmonson. Language

44.190-6.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

563

Georgopolous,

Carol.

1985.

Variables

in Palauan

syntax.

Natural language and linguistic theory 3.59-94. Givdn,

T.

1983a.

introduction, Giv6n,

T.

Topic

continuity

in

discourse:

an

in Giv6n 1983b:5-41.

1983b.

quantitative

Topic

continuity

cross-languaqe

in

discourse:

study.

a

(Typological

Studies in Language, Vol. 3.) Amsterdam: John Benjamins Publishing Company. Grimes,

Barbara

F.

Ethnologue,

1984,

10th

ed.

edition.

Languages Dallas:

of

the

world,

Wycliffe

Bible

Translators. Henne,

David. 1980. Diccionario Qulcht-Espaffol.

Guatemala:

Instituto Linguistico de Verano. Instituto

Indlgenlsta

Nacional.

1977.

Alfabeto

de

las

lenguas mayances. Guatemala: Editorial Universitaria. Kaufman,

Terrence.

1971. Tzeltal phonology and morphology.

(University of California Publications in

Linguistics,

vol. 61.) Berkeley: University of California Press. Kaufman, de

Terrence. 1974. Idlomas de mesoamdrica. (Seminario Integraci6n Social

Guatemalteca,

Publicacidn

no.

33.) Guatemala: Ministerio de Educaci6n, Editorial Jos6 de Pineda Ibarra.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

564 Kaufman,

Terrence.

1976a.

oztogzafias

para

(Ediciones

idiomas

Pzoyecto

esczlblz

las

de

alfabetoa

lenguas

guatemaltecos.)

y

mayances. Guatemala:

Ministerio de Educaci6n. Kaufman, Terrence.

1976b. New Mayan languages In Guatemala:

Sacapultec, Sipacapa, and others, in Mayan linguistics, volume I,

ed. by Marlys McClaran, 67-89.

Los Angeles:

UCLA American Indian Studies Center. Kaufman, Terrence. 1986a. Some structural characteristics of Mayan

languages

Unpublished

MS.

with especial

reference

to

Quichd.

(to appear in Spanish translation

in

England and Elliot.) Kaufman,

Terrence.

1986b.

Outline

of

grammar I: morphology and particles. Koopman,

H.

and

D.

Sportiche.

1982.

comparative

Mayan

Unpublished MS. Variables and

the

bisection principle. The linguistic review 2.135-70. Larsen,

Thomas W. 1981. Functional correlates of ergativity

in Aguacatec. BLS 7.136-53. Larsen,

Thomas W.

1987. The syntactic status of ergativity

in Quich6. in Dixon 1987:33-59. Levin,

Beth.

1987. The middle construction and ergativity.

in Dixon 1987:17-31.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

565

Marantz,

Alec

P.

1984.

relations.

On

the

(Linguistic

nature

Inquiry

of

grammatical

Monograph

10.)

Cambridge: The MIT Press. Mithun,

Marianne.

1987.

Is basic word order universal? in

Coherence and grounding in discourse, ed. by Russell S. Tomlin,

(Typological Studies in

Language,

vol.

11,)

281-328. Amsterdam: John Benjamins Publishing Company. Mondloch, £or

James L. 1978a. Basic Qulcht grammar. Mesoamerican Studies Publication No.

(Institute 2.)

Albany:

State University of New York at Albany. Mondloch,

James

L.

objects in Quichd.

1978b.

Disambiguating

subjects

and

Journal of Mayan linguistics 1, no.

1.3-19. Mondloch, James L. 1981. Voice in Qulchd-Maya. Albany: State University

of

New

York.

Unpublished

Ph.

D.

dissertation. Nebrija,

Elio

Antonio de.

1523.

De

lltteris

hebralcis.

AlcalA. Norman,

William. 1973. Positional and transitive verb roots

in Quichd.

Unpublished paper presented at the

annual

meeting of the American Anthropological Association.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

566

Norman, William. 1976. QuichA text, Louanna

Purbee-Losee.

American

Linguistics,

Vol.

No.

I,

in Mayan text I, ed. by

(International

Journal

Native American

1,) 40-60.

Chicago:

Texts

of

Series,

The University of

Chicago Press. Norman,

William. 1978. Advancement

change:

rules

and

the loss of instrumental voice in

syntactic Mayan.

BLS

4.458-476. Norman,

William. 1980. Grammatical parallelism in

QuichA

ritual language. BLS 6. 387-399. Postal,

Paul

M.

1977.

The

antipassive

in

French,

Proceedings of the seventh annual meeting of the

in

North

East Linguistic Society. Cambridge: M. I. T. (Reprinted in Linguisticae Investigationes 1:333-74. Recinos,

Adri&n. 1957. Cr6nlcas

Indigenas de

Guatemala.

Guatemala: Editorial Universitaria. Rivero, Maria-Luisa. 1978. Topicallzation and Wh movement in Spanish. Linguistic Inquiry 9.513-17. Rivero,

Maria-Luisa.

1980.

topicallzation in Spanish.

On

left-dislocation

Linguistic Inquiry

and

11.363-

95.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

567 Robertson,

John

S.

1980.

The

structure

of

incorporation In the Mayan verbal complex.

pronoun

Mew

York:

Garland. Safir,

Kenneth

J.

1984.

Multiple

variable

binding.

A

Linguistic Inquiry 15.603-38. Safir, Kenneth J. 1985. Syntactic chains. (Cambridge Studies in

Linguistics 40.)

Cambridge:

Cambridge

University

Press. Salazar Tetzagfiic, Manuel de Jes&s. 1978. Caracterlstlcas de la Llteratura indigene cakchiquel. the

Pacultad

de

Humanidades,

Thesis presented to Universidad

Rafael

Landivar, Guatemala. Sam Colop, Luis

Enrique.

1983. Hacia una propuesta de ley de

educacidn bilingSe. Thesis presented to the Facultad de Ciencias

Juridicas

y

Sociales,

Universidad

Rafael

Landivar, Guatemala. Sam Colop, Luis

Enrique,

to appear. Bosquejo

de la gram&tica

Quichd. inEngland and Elliot. Schultze Jena, der

Leonhard.

Quich4

1933.

von Guatemala.

Leben, Glaube und Sprache (Indiana

I.)

Verlag

Gustav Fischer in Jena.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

von

568

Silverstein,

Michael.

ergativity. languages,

in

1976.

Hierarchy

of

Grammatical categories

ed by R.

M.

W.

Dixon,

features in

and

Australian

112-71. Canberra:

Australian Institute of Aboriginal Studies. Smith-Stark, facts ed.

Thomas C.

1978.

and fictions. by

Nora C.

Linguistics Miscellaneous

The Mayan antipassive:

In Papers In Mayan

England,

no.

2.

169-187.

Publications

in

linguistics,

Studies in

(University

some

of

Mayan

Missouri

Anthropology

no.

6.)

Columbia: University of Missouri. Tedlock, the

Dennis. 1985. Popcl Vuh, the definitive edition of Mayan book of the dawn of life and the glories

of

gods and kings. Hew York: Simon and Schuster. Torrego,

Esther.

its effects. Vargas Llosa,

1984. On inversion in Spanish and some of Linguistic Inquiry 15.103-29.

Mario.

1983.

La tla Julia y el

escribidor.

Barcelona: Editorial Seix Barral, S. A. Vick,

Stanley

A.

and Remigio

Spoken Quichd (Maya).

Cochojil-Gonz&lez.

3 vols.

Chicago:

1966-9.

Department of

Anthropology, University of Chicago. Zwicky,

Arnold

M.

Cliticization vs.

and

Geoffrey

inflection:

K.

Pullum.

English n't.

1.983. Language

59.502-13.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.