quantity insensitive stress

7/17/2019 Quantity Insensitive Stress

http://slidepdf.com/reader/full/quantity-insensitive-stress 1/63

Springer is collaborating with JSTOR to digitize, preserve and extend access to Natural Language & Linguistic Theory.

http://www.jstor.org

A Factorial Typology of Quantity-Insensitive StressAuthor(s): Matthew GordonSource: Natural Language & Linguistic Theory, Vol. 20, No. 3 (Aug., 2002), pp. 491-552

Published by: SpringerStable URL: http://www.jstor.org/stable/4048055Accessed: 27-09-2015 12:05 UTC

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/ info/about/policies/terms.jsp

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of contentin a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship.For more information about JSTOR, please contact [email protected].

This content downloaded from 193.54.174.3 on Sun, 27 Sep 2015 12:05:54 UTCAll use subject to JSTOR Terms and Conditions

http://www.jstor.org/

http://www.jstor.org/action/showPublisher?publisherCode=springer

http://www.jstor.org/stable/4048055

http://www.jstor.org/page/info/about/policies/terms.jsp






http://www.jstor.org/stable/4048055

http://www.jstor.org/action/showPublisher?publisherCode=springer

http://www.jstor.org/



MATITHEW

ORDON

A

FACTORIAL

TYPOLOGY

OF

QUANTITY-INSENSITIVE

STRESS

*

ABSTRACT.

This

paper

presents

an

Optimality-theoretic

Prince

and

Smolensky 1993)

analysis

of

quantity-insensitive

tress.

A

set

of

grid-based

onstraintss shown

by

means

of

a

computer-generated

actorial

ypology

to

provide

a

relativelytight

fit to the full

range

of

stress

systems

attested

n

an extensive

survey

of


tress

patterns,

many

of which have not

been

previously

discussed

in the theoretical

iterature.

1.

INTRODUCTION

One of

the

major

endeavors

n

theoretical

phonology

over

the

past

quarter

century has been

to

develop

a

metrical stress

theory

which

offers ad-

equate empirical coverage of

attested stress

systems

with a

minimal

amountof overgeneration f unattestedpatterns.As ourdatabaseon stress

has enlarged

during

this time, the

challenge of

constructing

a

restrict-

ive yet

sufficiently

rich

metrical

theory

has

increased.

Several theories

have

emergedin

response

to this growing

challenge, both

in

rule-based

frameworks

e.g.,

Liberman

and Prince

1977; Prince

1983; Hayes 1980,

1995;

Halle

and

Vergnaud

1987, etc.) and, more

recently,

n

a constraint-

based

paradigm

(Crowhurstand Hewitt

1994;

Kager 1999; Kenstowicz

1997;

Walker

1996;

Alber

1997;Eisner

1997;

Bakovic

1996,

1998;

Elen-

baas 1999; Elenbaas and Kager 1999). OptimalityTheory (Prince and

Smolensky 1993),

with

its

hierarchically

ranked

constraints

goveming

phonological

well-formedness,has provided

fertilegroundfor

evaluating

metrical

tress

theories.

By

permuting he

constraint

ankings, t is

possible

to

constructa

factorial

ypology of

stress

whoseempirical

coverage

can be

compared

against

cross-linguistic

stress

data.

This

paper

belongs to the

research

program

employing

factorialtypo-

logies as a means

of

assessing the

predictive

powerof

metrical

constraints

(cf. Eisner 1997; Elenbaasand Kager 1999). The proposedaccount tests

the

stress

patterns

generatedby a

set of

constraints

against those

found

*

The

author

gratefully

acknowledges he

insightful

discussion and

technicalassistance

of Bruce

Hayes,

as

well as

comments from

Michael

Kenstowicz

and two anonymous

reviewers, which

have

greatly

improved

the paper.Thanks

also to

Donca Steriade and

audiences

at

UCSB and the

2001

SWOT

meeting for their

feedback on

thisresearch.

A

Natural

Language &

Linguistic

Theory 20:

491-552, 2002.

?

)

2002

Kluwer

Academic

Publishers.

Printed

n

the

Netherlands.






492

MATTHEW

ORDON

in what

is intended

to

be

a

fairly

exhaustive

survey

of


stresssystems, i.e.,

those

in which

syllable weight

does not influencestress

placement.

This

survey

revealsseveralstress

patterns hat,

to the best of

my

knowledge,

have not been

previously reported

n the theoretical

iterature

and

turnout to be

of interest

or

metrical

stress

typology.

An

area

in which the

present

work differs from other works

on

stress

within Optimality Theory

is

in

terms

of

its

representations

of

stress.

A

theory

of stress is

developed

that

does not

appeal

to the metrical

foot;

rather, he proposed

constraints

refer

directly

to the well-formedness of

different

stress

configurations, ollowing

earlier work

by

Prince

(1983)

and Selkirk (1984) within a derivational ramework,and more recently

by

Walker

1996)

in a

constraint-based

aradigm.

While

many

of the

ad-

opted

constraints

are familiar from

the

literature,

ome are novel either

in their method of evaluation

or their

formulation, ncluding

an

expanded

set

of

anti-lapseconstraints

and a constraint

which

requires

hat

stress be

aligned

with

both

edges

of

a word.

Although

the

goal

of the

present

work

is not to consider the entire body of evidence

for

metrical

constituency

(cf.

Kenstowicz

1994

and

Hayes

1995

for discussion of such

evidence,

e.g., segmentalrules,minimalwordrequirements, rosodicmorphology),

it

is

shown that constraints

referringdirectly

to stress without the inter-

mediary

of

foot structure

are sufficient to account for the full

range

of


tress

systems,

while

suffering

from

a modicum of

overgeneration

f a

typically non-pathologic

nature.

1.1. Representations f Stress

Althoughtheproposedtheoryshareswith muchcurrentwork its adoption

of a

constraint-based

aradigm,

t differs

from

most

contemporarywork

in

its

grid-basedratherthan foot-based representations f stress, follow-

ing

Prince

(1983) and Selkirk (1984). Following this work, the

present

analysis

assumes that the

level of stress associated with a syllable is a

function

of

the number of

levels

of

grid

marks above

a given

syllable.

Thus,

an

unstressedsyllable has a single grid mark above it (abstracting

away

from

weight-sensitivestress),

a

secondarystressed syllable has two,

anda syllable withprimarystress has threelevels of grid marks above it.

Adopting these representations, binarystress system with primary

tress

on

the initial

syllable and secondary tress on odd-numbered yllablesafter

the first

would

display

the

following associationsbetween grid marksand

syllables

for

words with

five and six syllables (1).






A FACTORIAL

TYPOLOGYOF


TRESS

493

Primary

tress

x

x

Secondarystress x x x x x x

Syllable

xxxxx xxxxxx

ccrCYacaxu

aa

uocxxa

Similar representations

re assumed

throughout

he rest of the

paper

with unstressed,primary,

and secondary

stressed

syllables differing

from

each other

in the

number

of

dominatinggrid

marks.

However,

o conserve

space, full representations f metricalgrids will be often be omitted;in-

stead, following

convention,secondary

stress

will

be

markedwith a

grave

accent

and

primary

tress with

an acute accent.

1.2.

A

Typologyof Quantity-insensitive

tress

As

a

startingpoint

in the

investigation,

an extensive

survey

of

quantity-

insensitive stress was conducted using

a combination

of

grammars

and

existing typologies

of


tress, including

those

in works

by Hyman (1977), Hayes(1980, 1995), andHalleandVergnaud 1987). A

total

of

262 languages,

all

languages

with

predictable


stress

patterns i.e., predominantly

non-phonemic

with a

clearly

dominant

pattern)

or

which I

could find data, were

included n the

presentsurvey.'

In

analyzingthe

results of the survey, t is useful to sort stress patterns

into three basic groups:

anguages

with

fixed stress

(either a single stress

or two

stresses), languages

with binary alternating tress, and those

with

ternaryalternating tress.

The presentation

n

this paper

will center around

these

threecore stress types.

For each class of stresssystems, results of the

cross-linguisticsurvey

will

be presented, ollowed

by the constraints elev-

ant for

yielding

the attestedpatterns,and thefactorial ypology of patterns

arising throughpermutation

of

the

rankings

of the proposed constraints.

The

coarse

taxonomyof

quantity-insensitive tress

systems is summarized

in

TableI, along with the section

in

which each system

is covered.

2.

FIXED STRESS

The most

straightforward

ystem of phonologically

predictablestress is

one

in

which

there is a

single stress per word falling

a fixed distance from

the

word

edge. Interestingly,

as pointed out by Hyman

(1977), of the set

1

Many

of the stress

systems

in

the survey are sensitive to

morphology o some degree,

a

factor that is not

treated n

the presentpaper.






494

MATTHEWGORDON

TABLE

I

Coarsetaxonomy of stress

systems

Stresssystem Section

Fixed 2

Binary

3

Ternary 4

of

logically

possible docking

sites

for

stress in

quantity-insensitiveingle

stress

systems, only

five are attested:

nitial

stress,peninitial

stress

(second

from

the

left),

final

stress, penultimate

tress

(second

from the

right),

and

antepenultimate

tress

(third

rom

the

right).

Both

Hyman's

extensive

survey

of

stress

systems

and

the

present

one

indicate

asymmetries

in

the relative

frequency

with

which each

of

these stress sites

is attested.2

Languages

with initial

stress are

abundant

in both

surveys

and

include, among

many others,

Arabela

(Rich 1963),

Chitimacha

Swadesh

1946),

and

Nenets

(Decsy

1966).

Similarly,many

languages

have final

stress, e.g., Atayal

(Egerod

1966), Moghol (Weiers

1972), Mazatec (Pike and

Pike

1947).

Also

very

common are

languages

with

penultimate

stress, e.g.,

Mohawk

(Chafe

1977),

Albanian

(Hetzer

1978),

and

Jaqaru(Hardman-de-Bautista

966).

Considerably

rarer

are

the

peninitial

pattern,

e.g.,

Lakota

(Boas

and Deloria

1933, 1941),

Koryak

(Zhukova1972), and the antepenultimate attern,e.g., Macedonian Lunt

1952),

Wappo (Radin 1929).

The numberof

languages

in

Hyman's

sur-

vey and the

current

one

displaying initial,

peninitial,

antepenultimate,

penultimate,

and final

stress

are

summarized

n

Table II. A

list

of lan-

guages (and

their

genetic

affiliation)

instantiating

hese

patterns

appears

in

Appendix

1.

References for the

languages

are

available

at

the author's

website:

http://www.linguistics.ucsb.edu/faculty/gordon/pubs.

he

relative

rarityof peninitial and

antepenultimate tress

will

be

discussed

further n

the context of the factorial ypologyin section 2.2.

2

Hyman's survey differs from

the present one in

its inclusion of

languages with

quantity-sensitive

tress

systems. Thefigures rom

Hyman's

survey

n

Table

II

also include

languages

with

binary stress.






A

FACTORLAL YPOLOGYOF

QUANTITY-INSENSITIVE TRESS 495

TABLE

II

Number

of

single stress

languages

Hyman(1977) Presentsurvey

Numberof

Igs.

%3

Numberof

Igs.

%4

Initial

114

37.3 57

30.2

Penultimate 77

25.2

53.55 28.8

Final

97

31.7

59.5 32.0

Antepenultimate

6

2.0

7

3.7

Peninitial 12

3.9 10

5.3

Total 306 187

TABLEIII

Dual

stress

systems

Primary

______

Initial Peninitial

Antepenultimnate

Penultimate

Final

Initial

-

1

67

36

Peninitia-

-

Antpenultimate_

Penultimate

3

-

- _

-

Final

_ -

___,.

3

Note

that

percentagesdo

not

addup

to

100,

as

figures

are

rounded.

4

Percentage

reflects

percentage

of

single

stress

languages n

the

survey.

5

The

.5

languages

with

penultimate

and

final stress

are

attributed o

Gurage

(Polotsky

1951;

Leslau

1992), which

displays

penultimate

tress in

nouns and final

stress in

verbs.

6

Georgian

(Zhgenti

1964;

Aronson

1991).

7

Anyula

(Kirton

1967),

Awtuw

(Feldman

1986),

Chimalapa

Zoque

(Knudson

1975),

Sibutu

Sama

(Allison

1979;

Elenbaas

and

Kager

1999),

Sanuma

(Borgman

1989), Murut

(Prentice

1971).

8 CanadianFrench (Gendron1966), Udihe (Kormushin1998), and most varieties of

Armenian,

unless

the

initial

syllable

contains a

high

vowel,

whichdoes

not

carry

secondary

stress

(Vaux

1998).

9

Walmatjari

Hudson

1978).

An

additional

complication n

Walmatjari

s

that

the

sec-

ondary

stress

falls

on the

penult in

words

with

four

syllables

(see

section

2.1.5),

where

antepenultimate

tress

would entail

a

stress

clash,

and

optionally

on the

penultrather

han

the

antepenult

n

words

with

greater

han

four

syllables:

thus,

dada

but

drad'aa

r

6onaaa.

10

Lower

Sorbian

Janas

1984),

Watjarri

Douglas

1981),

Gugu

Yalanji

Oates

andOates

1964).






496

MATTHEW ORDON

Another type of fixed stress

system

involves

two fixed stresses

per

word.

The present

survey

ncludes

14 such 'dual

stress'

languages,

a

small

number n

comparison

o the 167

single

stress

languages.

Table III

sum-

marizes

the numberof

languages

(language

names

given

in

footnotes)

in

the

survey

displaying

each

subtype

of stress

system

involving

two

fixed

stresses

(the

location of

primary

stress

falls on the

x-axis

and

secondary

stresson the

y-axis;

unattested

patterns

are indicated

by

a

dash;

logically

impossible patterns

are

shaded).

Interestingly,

of the 20

possible

combinations

(factoring

in

patterns

which

differ

only

in

which of the stresses

is

primary

and

which is sec-

ondary),only five areexploited.The numerousgaps in the set of attested

fixed

stresssystems with two stresses are

explored

urther n

the discussion

of the factorial

typology analysis

in section 2.2. All of

the

attested dual

stress

patterns,

with

the

exception

of two

of

the three

initial

plus

final

patterns,

Canadian

French and

Armenian,

do not allow stress

clashes;

in

clash

contexts,

for

example,

in

trisyllabic

words

in

languages

with

initial

and

penultimate

stress,

the

syllable

carrying

primary

stress in

non-clash

contexts takes the

only

stress.

2.1. A

ConstraintSet

for

Fixed

Stress

and

Resulting

Patterns

The set of

fixed stress

patterns

s accounted or

in

straightforwardashion

by

a

relatively

standard et of

Optimality-theoretic

onstraints,

certain of

which are formulated

n

slightly

novel

ways.

These

constraints

will

be

in-

troduced

as

we

consider he variousstress

systems

which

provideevidence

for them.

2.1.1. The

ALiGN

ConstraintFamily

Followingprevious

work

n

OT,

the

attraction f

stress

by edges is

modeled

by

constraints

equiring

stress to be

aligned

with

wordedges.

Two of

the

ALIGN

constraintswhich

play an important ole

in the

theory developed

here

serve

the

same function

as

constraintsalready amiliar

rom the

theor-

etical

literature,

.g., McCarthy

and

Prince

(1993),

Crowhurst nd

Hewitt

(1994), among

others. One of

these constraints

requires hat

stressedsyl-

lables be

aligned

with

the left

edge of a prosodic word, while the other

demandsthat

stressed

syllables

be

aligned with the

right

edge. Given the

grid-based

representationsassumed

here, the ALIGN

constraintscan

be

formalized

n

terms of

alignment

of

grid marksto

the metrical grid.

The

relevant

constraints ake the

form in

(2) andrequire hat

every grid

mark

be

aligned

with

either the left

or right

edge of the

immediately ower

level

of

grid

marks.






A FACTORIALYPOLOGY

F


TRESS

497

(2) ALIGN

(Xjevein+j,

tR/L},

level

n,

PrWd):Every grid

mark

of

level

n

+ 1

is

aligned

with the

I

right,

left

I

edge

of level n of

grid marks n a prosodic word.

The generalstress constraints

not

specific

to

primary

stress

emerge

if n

is

0,

in

which case the constraint

s violated

for

every grid

mark on level

1

which

does not

align

with the relevant

edge

of level

0

of

grid

marks.

ALIGN

(Xlevel

,

L, level 0, PrWd)requires hat evel

1

grid

marks

align

with the left

edge

of

level

0,

while ALIGN

(Xlevel

,

R,

level

0, PrWd) requires

that level

1

grid marksalign with the rightedge of level 0. Followingearlierwork,it is

assumedhere thatALIGN onstraintviolationsaregradient, uch thatthey

are calculatedfor each stress and

summed

together.

For

example,

a word

with

stress

on

both

the second and the third

syllables

incurs three total

violations of ALIGN

(xieveii,

L, level 0, PrWd), two for

the stress on the

third

syllable

and one for the stress

on the

second

syllable (see (5)

below

for an example tableau).

Another ALIGN

constraintwhich is not familiarfrom the

literature s

a broad constraint ALIGN

(EDGES,

level

0, PrWd, xlevel

),

which

requires

that the edges of level 0 of a prosodic word, i.e., both the initial and

the final

syllable,

be

aligned

with a

level

1

grid

mark. Violations are

calculated

n a

simple

fashion: one violation is incurred

f

eitherthe initial

or the final

syllable

does not

carry

a

level

1

grid mark,

and two

violations

are incurred

f

both the initial and

the

final

syllables

do not have a level

1

grid mark. ALIGN (EDGES, level 0, PrWd, xlevel

)

is thus in some sense an

apodic conglomeration

of the

constraints ALIGN-LEFT

and ALIGN-RIGHT

(Kager 1999; Elenbaas and Kager 1999), which require hat words begin

and end, respectively, with a foot. ALIGN (EDGES, level 0, PrWd, Xlevel)

will

be shownto play an important ole in languages with both initial and

final stress; these include CanadianFrench (Gendron 1966), Armenian

(Vaux 1998), and Udihe (Kormushin1998), which are discussed below,

and

Tauya (MacDonald 1990), which is consideredin section 3.3. This

adoptionof ALIGN (EDGES,level 0, PRWD,

X1evel

1)

is adoptedrather han

an atomistic

alternativenvolving separateconstraints or each edge, e.g.,

grid-basedversions of ALIGN-LEFTandALIGN-RIGHT,since the atom-

istic approach esultsin increasedovergenerationn the factorial ypology

discussed

in

sections

2.2, 3.3,

4.1.

The atomistic approachresults in 16

additionalunattested

patterns multipliedby two if one factors n variation

in

which of the stresses is primary)not generatedgiven the formulation

adopted

here:

four

additional dual stress patterns,three binary patterns,

and

nine

ternarypatterns see sections 2.2 and 3.3 for furtherdiscussion).






498

MATTHEW

ORDON

The three

ALIGN onstraints

are formalized n

(3) 1

(3)

Constraints

modeling

the attraction f stress to

edges

a. ALIGN

(Xievel1,

L,

level

0, PrWd): Every grid

mark of level

1

is

aligned

with the left

edge

of level

0

of

grid

marks n a

prosodic

word.

b. ALIGN

(xlevel ,

R,

level

0, PrWd):Every grid

markof level

1

is

aligned

with the

rightedge

of level

0

of

grid

marks n a

prosodic

word.

c.

ALIGN

(EDGES,

level

0, PrWd, xlevel

1):

The

edges

of level

0

of

grid

marks n a

prosodic

word are

aligned

with level

1

grid

marks.

There are

also ALIGNconstraints

which

require

that level

2

grid

marks

align

with the

edge

of level 1

of

grid

marks. These

constraints,

ALIGN

(Xlevel2,L,

level

1, PrWd)

and ALIGN

(Xlevel2,R,

level

1, PrWd),

thus refer

specifically

to

primary

tress.

They

are formulated n

(4).

(4)

Constraints

modeling

the attraction f

primary

tress to

edges

a.

ALIGN

xlevel2,

,

level

1, PrWd):Every grid

markof level 2 is

aligned

with the left

edge

of level

1

of

grid

marks n a

prosodic

word.

b.

ALIGN

x1,ve12,

,

level

1, PrWd):Every grid

mark

of level

2

is

alignedwith therightedge of level 1of gridmarks n aprosodic

word.

Violationsof

ALIGN

(Xievel

,

L,

level

1, PrWd)

and

ALIGN(Xlevel , R, level

1,

PrWd)

are calculated

by -counting

he

number

of

secondary stressed

syllables separating

he

primary

stressed

syllable

from

the relevantedge.

Thus,

for

example,

the seven

syllable

form

uacra'aoancurs two viola-

tions of ALIGN

(Xievel

,

L,

level

1, PrWd),

since two

secondary stressed

syllables intervenebetween the left edge of the word and the primary

stressed

syllable,

and one violation of

ALIGN

(Xlevel

,

R, level 1, PrWd),

since one

secondary

stressed

syllable separates

he

primary tress from the

right edge. A tableau llustrating he evaluationof all the proposedALIGN

constraints or a schematic

six-syllablewordappearsas (5). Where there

11

The

ALIGN

constraints

are often abbreviated hroughout he paper by eliminating

the final

two

arguments, .e.,

'level

x'

and

'PrWd',and the subscripted level' in the first

argument.

ALIGN

(EDGES,

level

0, PrWd,

xievel

1)

is abbreviated

s ALIGN

EDGES.






A

FACTORIALYPOLOGY

F

QUANTITY-INSENSITIVETRESS

499

are

several

violations

of a constraintcommitted

by

a

form,

the number

of

violations appears

n

parentheses.

Adoptingthe ALIGN x2, X, 1, PrWd)constraintsas opposedto con-

straints

which count absolute distance of the

primary

stress

from an

edge

(cf.

McCarthy

andPrince

1993)

has the

empirical

advantage

of

creating

a more constrained actorial

typology

of

stress

systems12

as well

as

the

formal

advantage in

a

grid-based

heory

of

representation)

f

being

more

principled,assuming

that all

grid

marks above

level

0 must dominate

a

lower level

grid mark

(Prince's(1983) ContinuousColumn

Constraint).

(5) Evaluationof the ALIGN onstraints

CaCaCaaa

AUGN

AUGN ALIGN

ALIGN

AUGN

EDGEs

(xl,

L)

(x1, R)

(X2, L) (X2, R)

2

x

**

*****(5)

****(5) *

o

xx xx

xx

Cy a6 a

F___Cy

2 X ******

(6)

*********

(9)

**

1 X

x

o

xxx

x

xx

2

x

* *********

(9) ******

(6)

**

I x x x

Oxx

xx x x

2

X ** ****

(4)

******

(6)

*

Ilx

'x

o

xx x

x

xx

ad

C6

aCy C

2

X *

******

(6)

*********

(9)

**

Ilx

x x

O

xx

x

xx

x

2

X *

******(6) *********

(9)

lx

x x

ox

x x

xx x

2

x

******

(6) ********* (9) .

-

lx

x

x

Oxx

xx

x

x

d

a

da da

.__

_ _ _ _ _ _

_ _ _ _

_

_-__

_ _ _ _

Before

proceeding with

analyses

employing the

ALIGN

constraints, it

should

be

noted

that, although

the ALIGN

constraints

discussed in this

paper

will

make

reference o the

word as the

stressdomain, it is

assumed

that

other

members

of

the ALIGN

constraint amily

sensitive to

different

stress domains, such as the root and different phrasallevels, also exist.

These

ALIGN

constraintsplay an

important

role in

characterizingmor-

phologically

sensitivestress (see

Alderete

1999 for

morphologicalstress

12

The ALIGN

X2,

{R/L),

1, PrWd)

constraints

adoptedheregenerate only

79 distinct

stress

systems

as

opposed to

93 generated

by theirhypothetical

counterparts

which count

absolute

distance of the

main

stress from an edge.

The extra

patterns,none of which are

attested,

all

under he

class of fixed stress

systems

displaying wo stressesper

domain (see

section 2.2

for

the factorial

ypology of fixed stress).






500

MATTHEW

GORDON

within Optimality

Theory)

as

well as

phrase-level

stress

(see

Gordon

in

preparation

or discussion

of

phrase-level

ALIGN constraints).

2.1.2.

Initial Stress

and

Final

Stress

Given the

ALIGN

constraints,

we are

in a

position

to

generate

three of the

fixed stress patterns.

The

general

ranking

schema

which

generates

single

fixed stress

systems

involves

one

general

ALIGN

constraintdominating

other

ALIGN constraints

and

*LAPSE.

Following

standard

procedure

n

OptimalityTheory,

variation

n the

edge

referred o

by

the

highly

ranked

ALIGN constraintyields

differences

between

languages

in

the

edge

or

edges towardwhich stressis attracted.Stresssystems with a single stress

on the initial syllable reflect

an

undominated

ALIGN

(x1, L)

ranked

above

ALIGN

(x1,

R), and

ALIGN

EDGES.

In

stress

systems displaying

a

single

stress

on the final

syllable,

ALIGN

(x1,

R)

is inviolable and rankedabove

ALIGN (x1, L)

and

ALIGN

EDGES.

Finally,

ALIGN

EDGES

plays

a crucial

role

in

the

CanadianFrenchpatternreported

by Gendron (1966),

which

involves primary

tress on

the final syllable

and

secondarystress

on the

ini-

tial

syllable,

even

in

clash

contexts.13

Vaux

(1998)

reports

a similar

pattern

involving primarystresson the ultima andsecondary stress on the initial

syllable

in most varieties

of

Armenian;

he

secondary

stress on an initial

syllable containing

a

high

vowel

is

suppressed.

Certainvarietiesof Udihe

also

place primary

tress on the

ultimaand

secondary

on the

initial

syllable

(Kormushin

1998),

with Nikolaeva

and

Tolskaya (2001) suggesting

that

the secondary

stresson the

initial

syllable is

either completely suppressed

or

very

weak

in

disyllables.14

To capture hese

initial

plus

final

stress

patterns,

ALIGN EDGES ensures

thatsyllablesatboth edges of the word are stressed.The promotionof the

final stress to primarystatus

in CanadianFrench,

Armenian, and Udihe

reflects the rankingof

ALIGN (X2,R) over

ALIGN (X2,L). The schematic

example

in

(6) provides

a closer look at the rankings yielding primary

stress

on

the final syllable and

secondarystresson the initial syllable.

13 In his studyof CanadianFrench,Gendron 1966,

p.

150) contrastsvarietiesof

French

spoken

in

Canada,which place secondarystress

on

the initial

syllable,

even in

disyllables,

with

ParisianFrench,

which lacks this secondarystress. He

observes that

vowels in initial

and final

syllables are

longer

and articulatedmore

strongly

than

vowels in

word-medial

syllables, which, in the case of

the high

vowels lil

and

lul, may devoice

or delete word-

medially. Other

accountsof Canadian

French

I

have located

only mention the

primary

stress on the

final syllable.

14

Nikolaevaand Tolskaya

(2001)

also

discuss

a

varietyof Udihe

with

quantity-sensitive

stress.






A

FACTORIAL

TYPOLOGYOF QUANTITY-INSENSITIVETRESS

501

(6) Initial and

finalstress

UUT tALIGN A,GN

ALIGN ALIGN ALIGN

EDxFs

(x2R)

(xl)

(x1,

R)

I(X,

L

~oa

The third

and fourth candidates

both fail

because each lacks a stress

on one of the edge syllables

in

violation

of

ALIGN EDGES. The second

candidate fatally

violates

ALIGN

(X2, R)

because the

rightmost

stressed

syllable

does not

bear

primary

tress.

If

one

assumes

thatthe

initial

syllable

in

Udihe

is not stressed

n

the

clash context arising

n

disyllables,

*CLASH

must outrank

ALIGN EDGES

which

in

turn

s rankedabove

ALIGN

(x1, R).

ALIGN

(xl,

R) is rankedhigher

than

ALIGN

(x1, L) therebyaccounting

or

the

survivalof the stress

on the finalsyllable

in

disyllables.

2.1.3.

NONFINALITY

nd Penultimate

Stress

Repulsion of stress from the right edge is captured n OptimalityThe-

ory by

NONFINALITY

e.g.,

Prince and

Smolensky

1993;

Walker

1996).

For

purposes

of

the

present

grid-basedanalysis,

NONFINALITY

simply

prohibitsa

level

1

grid mark

on the final syllable(7).

(7)

NONFINALITY:

tress

does not

fall on the final syllable. (A final

syllable

does not have a level

1

grid mark.)15

NONFINALITY

laysa roleincapturinganguageswith a singlefixed stress

on the

penult.

In

such

languages,

NONFINALITYs

ranked

above

ALIGN

(x1, R).16

This rankingproducesthe 'rightmost

without being

final' stress

patterncharacterizing anguages

with penultimate

tress.17

2.2. The*L4PSE

Constraints

and Peninitialand Antepenultimate

tress

The *LAPSE

constraints Prince

1983; Selkirk 1984; Green

and Kenstow-

icz 1996;

Elenbaas 1999;

Elenbaas and Kager 1999) ban

sequences of

unstressed

syllables. Two non-position-specific

*LAPSE constraintsand

15

It is

assumed that the

ContinuousColumn Constraint

Prince 1983) rules out a final

syllable carrying

a level

2

grid mark

without a level

1

grid

mark

also.

16

In

languages with penultimate

stress in

disyllabic and longer

words but which

have stressed

monosyllabic

words,

NONFINALITY

is

violated in

order to ensure that

monosyllables are

stressed.

17

Thanksto a reviewerfor

pointing

out that

NONFINALITYallows for an

alternativeet

of

rankings

or

generating

nitial

stress:

NONFINALITY

>>

ALIGN EDGES.






502

MATTHEW

ORDON

three sensitive

to

lapses

at word

edges

are assumed here.

The two

gen-

eral

LAPSE constraintswill

play

a critical role in the

binary

and

ternary

stress

pattems

to be discussed in sections 3 and

4, respectively.

The

first

of these

constraints,

*LAPSE,

bans

adjacent

stressless

syllables

and is

im-

portant

in

binary

stress

systems (section 3).

Violations of

*LAPSE

are

assessed

gradiently;thus,

a

string

of three

adjacent

stressless

syllables

incurs two violations of

*LAPSE,

while a

string

of four

adjacent

stress-

less syllables

incurs three violations of

*LAPSE.

Similarly,

a word

with

two

non-contiguous sequences

of two unstressed

syllables

suffers two

violations

of

*LAPSE.

Thesecondconstraint,

EXTENDEDLAPSE,

is decisivein ternary tress

systems (section

4);

it bans a

sequence

of morethantwo

adjacent

tressless

syllables.

Violationsof

*EXTENDEDLAPSEare

also

gradient; hus,

a

string

of four

adjacent

stressless

syllables

incurs two violations

of

*EXTENDED

LAPSE

(and also threeviolationsof

*LAPSE),

as do two

non-contiguous e-

quences

of three unstressed

yllables.

The

two

general

*LAPSE

constraints

are formalized

n

(8).

(8)

*LAPSE

constraints

a. *LAPSE:

A

string

of more than one

consecutive

stressless

syl-

lable may not

occur.(A

sequence

of

more than one

consecutive

syllable

lacking

a level

1

grid

mark

s

banned.)

b.

*EXTENDEDLAPSE:

A

string

of more

than two

consecutive

stressless

syllables

may

not occur.

(A sequence of

more than

two consecutive

syllables

lacking

a level

1

gridmark

s banned.)

It

should be noted that the

formulationof

*LAPSE

is

different

from other

anti-lapse

constraints n

the

literature,

ncludingversions

referring o met-

rical feet,

e.g., Kager's (1994)

PARSE-2and

Green and Kenstowicz's

(1996)

*LAPSE,

as

well as versions

referring

to the

metrical

grid, e.g.,

Elenbaas

(1999)

and

Elenbaasand

Kager (1999).18

The members of

the *LAPSE

constraint

family crucial for

capturing

the remainingfixed stress patterns are specific to stress lapses at do-

main

edges, following work

by Alderete (1999).

Avoidance

of stress

lapses

at

word

edges is

perhaps

most obvious in

quantity-sensitive lan-

guages

with

stress

windows, e.g.,

Piraha

(Everett

and Everett

1984; Everett

1986, 1988),

Kobon

(Davies

1980

in

Kenstowicz 1997),

Chukchi (Skorik

18

Elenbaas

and

Kager's *LAPSE

allows

sequences of two

consecutive

stressless

syl-

lables

at a

wordedge,

unlikethe

*LAPSE

onstraint

adoptedhere

(see

Elenbaas

and

Kager

1999 for

comparisonof

their

grid-based

anti-lapseconstraint

with

foot-basedones).






A FACTORIAL

YPOLOGYFQUANTITY-INSENSITIVE

TRESS 503

1961), though

apse

avoidance

at

edges

also

figures

prominently

n

certain

quantity-insensitiveatterns,

as we will see.

The constraints esponsible

or

creating

stress

windows are members

of

the

*LAPSEEDGE

sub-family

of constraints.

A

two-syllable

stress window

at the

right

edge, e.g., Kobon,

is the result of a

highly

ranked

*LAPSE

RIGHT

constraint

which bans

more than one stressless

syllable

separating

the rightmost

stress from

the

right

edge

of a word. A

two-syllable

stress

window at

the left

edge, e.g.,

Chukchi,

is attributed o a

highly

ranked

*LAPSE

LEFT banning

more

than one

stressless

syllable

separating

he

leftmost

stress

from the

left

edge.

A

three-syllable

window at the

right

edge, e.g., Piraha,is attributed o a highly ranked

*EXTENDEDLAPSE

RIGHT

constraintbanning

more than two

stressless

syllables

separating

the

rightmost

tress from the

right edge.

The

*LAPSE EDGE

constraints

are assumed

here

to be

categorical;

the forms

acrou and aocscyac

hus both

incur a

single

violation

of

*LAPSE

RIGHT. The latter form also violates

*EXTENDED

LAPSE RIGHT.

The

*LAPSE

EDGE

constraints

are

defined

n

(9).

(9)

*LAPSEEDGE constraints

a.

*LAPSE

RIGHT:

A maximum of one unstressed

syllable

separ-

ates

the

rightmost

stress

from the right edge

of a stress domain.

(No

more than one

syllable

separates

the rightmost syllable

with a level

1

grid markfrom the right

edge.)

b.

*LAPSELEFT:

A

maximumof one unstressed

yllable separates

the

leftmost stress from

the left edge

of a stress domain. (No

more than

one syllable

separates

the leftmost syllable

with a

level

1

grid

markfrom the left edge.)

C.

*EXTENDED

LAPSE RIGHT:

A

maximumof two

unstressed

syllables

separates the

rightmost stress from the

right edge

of a stress

domain. (No

more

than

two syllables

separate the

rightmost yllablewith a

level 1 grid mark rom the

rightedge.)

We

are now

in

a

position

to analyze

two additionalfixed stress

patterns:

peninitial

stress

and antepenultimate tress. Both

of these stress

patterns

are the

result

of an ALIGN constraintpulling stress

toward one

edge and

an

overriding

*LAPSE EDGE

constraint

pulling stresstoward he

opposite

edge.

The result of

this

struggle is

demonstrated n (10) which

illustrates

the

antepenultimate

tress patternresulting

from the ranking

*EXTENDED

LAPSE RIGHT

>>

ALIGN

(X1, L)

>>

ALIGN (X1, R). This ranking

ensures






504

MATTHEW

GORDON

that stress

falls

as far to the

left as

possible

without

falling

to the left

of

the

antepenult.19

(10) Antepenultimate

tress

caaCa3 *ExT

LAPSE

ALIGN X1, L)

ALIGN

x

I

R)

RIGHT

edaaa

aaaad

..-.

t.

.

Fixed peninitial stress is attributed to an analogous ranking except for

switching the

reference

edge

of

the

ALIGN

and *LAPSEEDGE constraints.

In

peninitial

systems,

*LAPSELEFT

is

ranked

above ALIGN

(X1, R) which

in turn akes precedence

over

ALIGN

X1, L).

The effect of this

ranking

s

to force

stress as far

to the rightas possible

without falling to the rightof

the

peninitial

syllable,

as shown

in

(11).

(I 1) Peninitial

stress

Goaaa *L"SE

LLF

ALIGN

x1, R) ALIGN

(Xi,

L)

Interestingly,

I am not aware of any

stress system

with a three syllable

win-

dow at the left

edge

which would

necessitateadoptionof

the hypothetical

left edge counterparto

*EXTENDEDLAPSERIGHT.20

19

Note

that penultimate stress also

follows from

essentially

identical

rankings

to

those

driving

antepenultimate

stress

except

for

replacing

*EXTENDED

APSE

RIGHT

with

*LAPSE RIGHT.

Thus,

fixed

penultimate stress reflects two

possible

rankings:

NONFINALITY

>

ALIGN

X1, R)

>>

ALIGN

X1,

L)

or *LAPSERIGHT

?>

ALIGN

X1,

L)

>>

ALIGN

X1, R).

Despite

this functional

overlap,

both

NONFINALITYnd

*LAPSE

RIGHTmust be

maintained

as

separate

onstraints,

ince

they serveindependent

unctions

in

many cases. For

example, NONFINALITY

lays

a

crucial role

in

binarystress

patterns

withfinal stress

avoidance

(section3.2.1), whereas

*LAPSE

RIGHT

aptures tress window

effectsandplays a decisiverole in dual stresssystems(section2.1.5), andin certainbinary

plus

clash (section 3.3) and

binarypluslapse stress

systems (section

3.2.2).

20

Thanks o an anonymous

reviewer or pointing out

the stress

patternof Terena,

which

potentially provides evidence

for

a

*LAPSEEDGEconstraint

prohibiting

more

thantwo

consecutive stressless

syllables at the left

edge

of a

word.

According to Harden(1946),

stress

n

Terena s

phonemic andmayfall on

any one

of

the

first

three

syllables of

the

word,

provided he

stressed syllableis non-final. n

his

detaileddescriptionof Terena

tress,how-

ever,Bendor-Samuel

1963) shows

thatstress is phonemic within

a two-syllable window

at

the left

edge,

with

third syllable stress

arisingundercertain

phrase-level

morphological






A

FACTORIAL

TYPOLOGYOF

QUANTITY-INSENSITIVE TRESS

505

2.2.1.

Dual

StressSystems

Thusfar, we have accountedfor

all of the

single

fixed stress

patterns:

ni-

tial

stress, peninitialstress, antepenultimatetress, penultimate tress,and

final

stress.

In

addition,we

have

analyzed

the initial

plus

final

dual

stress

pattern.This leaves

only

two

types

of dual fixed stress

systems

to

account

for:

ones with both initial

and

penultimate

tress

(e.g.,

Sibutu

Sama,

Lower

Sorbian)

and

those

with

initial and

antepenultimate

tress

(e.g., Georgian).

Since both of

these stress

patterns

result from

very

similar

rankings,

we

will

only analyze one

subtype of the

more common

pattern

n

detail,

the

initialplus

penultimate

pattern.Differences

between the

rankings

underly-

ing this patternandthose behindclosely relatedones will be pointedout

later.

Our

exemplar

of the

initial

plus

penultimate

pattern

s Sibutu

Sama

(Allison

1979; Elenbaas

and

Kager

1999),

in

which

primary

tress

falls on

the

penultimate

yllable and

secondary

stress

on the initial

syllable,2'

with

the added

provision

thatonly the

penult s

stressed

n

trisyllabic

words;

his

pattern s

exemplified n

(12)

(examples from

Elenbaasand

Kager

1999).

(12) SibutuSamastress

a.

bissala

'talk'

b.

bissalahan

'persuading'

c.

bissalahainna

'he

is

persuading'

d.

bissalahankaimi

'we

are

persuading'

Before

developingan

analysis

of Sibutu

Sama

stress,one

additionalcon-

straintmust

be

introducedto

account for

the

fact that

adjacent

stresses

are

prohibited n

trisyllabic

words in

Sibutu

Sama.This ban

on

clashes is

indicativeof

the

high ranking

of a

constraint

banning

stress

clashes.

This

constraint,

which

was

introducedby

Prince

(1983) as a

rhythmic

principle

prohibiting

adjacent

tresses(cf.

Kager

1994;Alber

1997;

Elenbaas1999;

Elenbaasand

Kager

1999for

*CLASH

in

Optimality

Theory), s

formulated

in

(13).

and

syntactic

conditions

which

would

fall

under the

rubric

of

intonational

phenomena

in

other

languages,

e.g.,

marking

contrastive

focus,

negative

imperatives,

nominal non-

specificity.

Thus,

although

there

may be

a

window

effect in

Terena,

t

seems

unlikelythat

third

syllable

prominence n

Terena

results

from

the

same

class of

*LAPSE

EDGE

stress

constraints

elevant

or

cases

discussed in

the text.

21

This

penultimate

plus

initial

stress

pattern

s

reported or

unprefixed

words

(see

Kager

1997

for

discussionand

analysis

of

prefixed

orms

which

display

additional

complexities).






506 MATTHEW

ORDON

(13)

*CLASH:

A stress domain does not contain

adjacent

stressed

syllables. (Adjacent syllables

carrying

a

level 1 grid

mark are

banned.)

A

string

of

two

adjacent

stresses

incurs one violation of *CLASH.An

ad-

ditional

violation

is

incurredfor each additional

stressed

syllable

in

the

sequence.

A

sequence

of three consecutivestresses thus

violates *CLASH

twice,

as does

a word

containing

two

stress clashes

separatedby

one

or

more unstressed yllables.

We now consider

one set of

rankings

which

generate

he Sibutu Sama

stress facts. As the forms in (12) indicate, in orderfor both stresses to

be

realized,

a word

must have at least

four

syllables;

this indicates that

*CLASH

s undominated.

t

is

crucially

rankedabove

ALIGNEDGESas

trisyllabicwords have

a

single

stress on the

penult

in

violation of ALIGN

EDGES: bissala,

not *bissala.

*LAPSE RIGHT

s also

undominated and is

instrumental

n

accounting

for the stress on

the

penult;

t is

rankedabove

ALIGNEDGES,

as evidenced

by

the stress on

the

penult

ratherthan the

initial syllable

in

trisyllabic

words:

bissala,

not

*bissdla.

*LAPSERIGHT

s

also ranked above both ALIGN (X1, R) and ALIGN(X1, L), since words of

at least four

syllables

stress the

penult,

even

though

this entails additional

violations of

ALIGN constraints:

bissalahanna,

not

*bssalahanna. ALIGN

EDGES s

ranked

above ALIGN X1, R),

as

all

words with at least

four

syllables

have a second stress on the initial

syllable

in

addition

to the

one on the

penult: bissalihan,

not

*bissldhan. NONFINALITYs

ranked

above ALIGN

EDGES,

as the final

syllable

is

never

stressed: bissaldhan,

not *bissalahdn.

Finally,

ALIGN

(X2, R)

is ranked above

ALIGN (X2, L)

thereby ensuringthat the penulttakesprimarystressin forms containing

two stress:

bissalhhan,

not

*bissalcihan.

The

rankings

or

Sibutu Sama are

summarized

n

(14).22

A tableau

demonstrating

he

rankings for Sibutu

Sama

appears

n

(15).

22

Note that the

rankings

n

(14) are not the

only ones which generatethe Sibutu Sama

stress

system.

If

ALIGN

(X1, L)

is ranked above

ALIGN (X1, R), but below

*LAPSE

RIGHT,

NONFINALITY

eases to

play

a

decisive role

in yielding the stresson the penult.This altern-

ate set

of

rankings

s

nearly identical to the rankingswhich generate the

antepenultimate

plus

initial stress

pattern ound in Georgian(see

below).






A FACTORIALTYPOLOGYOFQUANTITY-INSENSITIVE TRESS 507

(14) Constraint

ankings

or SibutuSama

*CLASH,

NONFINALITY,

*LAPSE

RIGHT,

ALIGN

(X2, R)

ALIGN EDGES

4

ALIGN

(X1, R),

ALIGN

(X2, L),

ALIGN

(X1,

L)

(15)

bissala

*CLSH ON-

*LAPsE ALIGN

ALIGN ALIGN

ALIGN

ALIGN

FINAL Ritr'r

(x,.

R,

1)

Eixim

(x1, R)

(x

2

(XL

I)

gE

bissla

_

l5ssasla

_ _ _ _ _ _

bissala

l.

jE_E__

bissaldi

_m

b'{ssalE

bissalahan *CLSH NON- *LAPSE ALICN ALIGN ALIGN ALIGN ALIGN

FINAL

RIGHT

(x2.

R,

1)

EDGES

(XI,

R)

(X2.

L?

(X1

)

bw

ssalahan

,.

;:

'b'sssal ihan

bissalahan

-

**

_

*

bissalahani

_ -

I _

"

1

'

blssalahanl

-

-

---------

btssslahan

_ _ E

bi-ssalahin .

_____

The

Lower Sorbianpattern Janas

1984),

in

which the initial stress rather

than

the

penult

is

preserved

n

trisyllabic

words, e.g.

w5sitsojska

'father-

land', psiijasiel

'friend',

and the

primary

stress is on the initial

syllable

rather han the

penult

in

wordsin which

both the penult

and the initial syl-

lable are stressed

see section 2.2.2 for

discussion of the

link between clash

resolution and primary stress placement), results from some minor re-

rankings

relative to those

observed n

SibutuSama.

First,

FLAPSERIGHT

is

rankedbelow ALIGN

EDGES,

thereby accountingfor

the stress on the

initial

syllable rather

he penult in trisyllabicwords.*LAPSE RIGHT

must

nevertheless still

be ranked

above ALIGN (X1, L) as

the penult

carries

stress

in

all

words over

three

syllables. Finally,ALIGN

(X2,

L) is ranked

above

ALIGN

(X2, R), reflecting the

fact that the stress

on the initial syl-

lable rather

han the

stress on the

penult is promoted o

primary

stress in






508 MATTHEW

ORDON

words with two stresses.

The

rankings

for

Lower Sorbian are summarized

in (16).

(16) Constraint rankings

for Lower Sorbian

*CLASH,

NONFINALITY,

ALIGN

(X2, L)

4

ALIGN

EDGES

4

*LAPSE RIGHT

4

ALIGN (X1, L), ALIGN(X2, R), ALIGN(X1, R)

Georgian (Zhgenti 1964;

Aronson

1991) displays

a

slight

variation on the

Sibutu Sama

pattern.

In

Georgian,

stress falls on both the initial

and

the

antepenultimate syllables

in

words over four

syllables long

and

only

on

the

antepenult

in

tetrasyllabic

words.23

The stress on

the

antepenult

re-

flects an undominated

*EXTENDEDLAPSE RIGHT

(rather

than

*LAPSE

RIGHT) crucially

ranked

above

ALIGN

(X1, L).

The

secondary stress on

the initial syllable

in words of at least five

syllables reflects the ranking

of *LAPSE LEFT

>>

ALIGN

(X1, R).

The fact that

the initial

syllable

rather than the

peninitial syllable

is selected for stress within

the window

defined

by

*LAPSE

LEFT

is ascribed to the

ranking

of ALIGN

(X2, L) over

ALIGN

(X2, R). Finally,

ALIGN

(X2, R)

is ranked

above ALIGN (X2, L),

thereby accounting

for the

promotion

of the stress on

the antepenult to

primary stress in words with two stresses. The rankings for Georgian are

summarized in

(17).

(17) Constraint rankings for Georgian

*CLASH,

NONFINALITY, *LAPSE LEFT, *EXT LAPSE RIGHT,

ALIGN (X2, R)

ALIGN (X1, L)

4

ALIGN

(Xl,

R),

ALIGN

EDGES,

ALIGN

(X2, L)

23

The two sources

consultedon

Georgianstress do not

describethe

same

facts,

although

the data in

each are

mutuallycompatiblewith

the other

source.

According to

Zhgenti's

(1964)

phonetic study

of Georgian

prominence,Georgian

words are

stressed

on the

ante-

penultimate yllable.

Aronson's

(1991) grammarmentionsan

additional

tress on

the

initial

syllable

in

words of

at

least five syllables.






A

FACTORIALYPOLOGY F


TRESS

509

Walmatjari

Hudson

1978)

displays

a variantof the

Georgian

pattern,

but with

primary

stress on the initial

syllable

and

secondary

stress

on

the

antepenult,

xcept

in

tetrasyllabic

words which

have

secondary

stress

on the

penult

rather

he

antepenult.

This

pattern

ollows from

essentially

the

same

rankings

as

Georgian,

save

for

the

promotion

of

ALIGN

EDGES

above

the

other

ALIGN

constraints

and

above

*EXT

LAPSE

RIGHT, ensur-

ing stress on

the

initial syllable

rather han the

antepenult

n

words

with

four

syllables.

The

option

of

stressing

the

penult

rather

hanthe

antepenult

in

words

longer

than four

syllables follows from

the

optional

ranking

of

*LAPSERIGHT

above

ALIGN

(X1, L) (but below

*CLASH

nd ALIGN

EDGES, s trisyllabicwords lack a stressfallingwithinthewindow defined

by *LAPSE

RIGHT.)

The

promotion of the

initial stress to main

stress

in

words with

two

stressesreflects

the

ranking

ALIGN

X2, L)

>>

ALIGN

X2,

R).

2.3. A

Factorial

Typology

of Fixed

Stress

We

now

consider the

factorial

typology

of

fixed stress

systemsgenerated

by the

constraints.The

factorial

typology

consisted of

determining

all

the

sets of stresspatternsgenerated

by

permuting he

constraint

ankings

n

all

logically

possible

orders.

All

logically

possible

stress

patterns

or

words

containing

between one

and

eight

syllables were

considered as

candid-

ates. The

only potential

candidate

stress

patterns

not

consideredwere

ones

which

violated

Prince's

(1983)

CULMINATIVITY

ondition,

.e., it was

as-

sumed

that all

words

have

one

syllable

that

is more

prominentthan

all

others.24

The

candidate

forms

consisting of

two

and three

syllables in

(18)

provide a

relatively

small

example of

the set

of

potential

outputforms

being evaluated.

1

is

a

primary

tressed

syllable,

2

is a

secondary

stressed

syllable, and

0 is a

stressless

syllable.)

(18)

Candidate

orms

containing

two and

three

syllables

2

syllables

3

syllables

[01] [100] [102] [120] [122]

[10]

[010]

[012]

[210]

[212]

[12]

[001]

[021]

[201]

[221]

[21]

24

I leave the

issue open

as to

whether

Culminativity s a

(likely

inviolable)

constraint

or

a

universal

property

of Gen

which

limits the

pool of

candidates

available

or

evaluationby

the

constraint

et.






510 MATTHEW

GORDON

The

number of

possible

stress

patterns

increases

commensurately

with

word length

according

to the formula

n

*

2(n-l),

where n =

number of

syllables:

4 in

disyllabic

words,

12

in

trisyllables,

32 in words with

4

syllables,

80 in words with 5

syllables,

192

in words with

6

syllables,

448

in

words with 7

syllables,

1024

in

words with 8

syllables.

The twelve constraints

n the factorial

ypology

are

repeated

as

(19).

(19) Constraints

n

the factorial

ypology

1. ALIGN

(X1, L)

5.

ALIGN

(X2, R)

9.

*LAPSE

RIGHT

2.

ALIGN

(X1, R)

6.

*CLASH

10.

*EXTENDED

LAPSE RIGHT

3. ALIGN EDGES 7. *LAPSE 11. *LAPSE LEFT

4.

ALIGN (X2, L)

8. *EXTENDED

LAPSE

12.

NONFINALITY

The

only

fixed

ranking

assumed involved

constraints

referring

to

primary stress.

It was assumed that either

ALIGN

(X2,

L)

or ALIGN

(X2,

R)

was rankedbelow

all

other stress constraints.This

ranking

reflects the

cross-linguistic

parameterization

f

primary

tress

placement

n

languages

withmultiplestressesin a single word: anguagesareinternallyconsistent

in

either

promoting

he

rightmost

or the

leftmost stress

to

primary

tress,

a

consistency

reflected

n

the

parameterized

pplication

of

either End

Rule

Right

or End Rule Left

in

derivational rameworks

Prince 1983).25

In

terms of

constraints,

t

is thus

hypothesized

hat there

are no

languages

n

which bothALIGN

(X2,

L)

andALIGN

(X2, R)

are

ranked

highly

enough

to be

active,

an

observation

which

is

captured

by

demoting

one of

them

below all otherconstraints.26

25

As

a reviewer

points

out,

the fact that

only

one of the ALIGN

onstraints

eferring

o

main stress

is

highly

ranked

enough

to

influence

the

grammar

n

any given languagecould

be takenas

evidence that here s a

single

ALIGN

X2, X)

constraint

which is

parameterized

on a

language-specific

basis

according

o

which

edge

is relevant.

26

As an

exercise,

the

factorial

typology

was also

computed

without

fixing

the

ranking

of either ALIGN

X2, L)

or ALIGN

X2, R)

below other stress

constraints.This

resulted

in a

proliferation

of

generated

stress

systems

(includingbinary

and

ternary

systems, as

well

as fixed stress

systems): 302,

as

opposed to 152 with the

a priori

rankingassumed,

includingboth those with an

undominated

ALIGN

X2, L)

and

those with an

undominated

ALIGN X2, R). The extrageneratedpatternsarise whenbothALIGN

onstraints eferring

to

primary

tress are ranked

above one or both

of the

general *LAPSE

onstraints,*LAPSE

and *EXTENDED

APSE,

which

themselves are

ranked above

the ALIGN

constraintsre-

ferring

to

secondary

stress. This set of

rankings

has two

effects.

First, it limits

the

number

of

stresses,

since

increasing

the number of

stresses

necessarily increases

violations of

at

least one

of the ALIGN

constraintsfor main

stress.

Second,

it

positions those

stresses

in

such a

way

as to minimize

violations of

the highly

ranked *LAPSE

onstraint(s).The

overall

effect is to shift the

location of the

stress(es)

as a

function of numberof

syllables in

the

word,

a

mechanism which

appears o

be absent

in

attested

anguages. Although

space






A FACTORIALTYPOLOGYOF


511

The task of

calculating

a factorial

typology

for a set of twelve

con-

straints

s

appreciable.

Permuting

he

ranking

of twelve

constraintsn

all

possible ways yields an a prioritotal

of 12 or

479,001,600 logically pos-

sible rankings.

Given the tremendous

analytic complexity

of

calculating

the

output sets generated

for all these

rankings,

the task was

deleg-

ated to

Hayes

et al.

(2000)

OTSoft software

package,

which

computes

factorial ypologies

for a set

of

constraints

and candidates

using

Tesar

and

Smolensky's

(1993)

Constraint

Demotion

algorithm.

The

input

iles for

the

OT softwareconsist of

a set of

constraints

and a

set of candidate orms

and

theirconstraintviolations.Toexpeditethe

process,

the

candidatesand

their

violations were also generatedby computer.27

Each

output

set

generatedby

the

typology

consists of a series

of

eight

forms,

each one

corresponding

o

a word with a differentnumber

of

syl-

lables. The set of

forms in (20) is one

sample outputset,

corresponding

to

the

stress

pattern

found

in

Sibutu Sama, generated

by

the

typology:

primary stress

on

the penultimate syllable and

secondary

stress on the

initial syllable in

words of at least four

syllables.

(20) Sample outputset (SibutuSama)

[1], [10], [010],

[2010], [20010],

[200010], [2000010],

[20000010]

Overall

results of the

factorial

typology indicated

a

strikingly good

fit

between

the

survey

of

attested stress

patterns

and the

proposed

con-

straint set. Of the

479,001,600

logically possible rankings and the

10,823,318,000,000 logically possible output sets given a set of candid-

ates

consisting

of

between one and

eight syllables,28a

mere 152 of these

limitationsprohibit

detailedexaminationof these patterns,we

may considerone such

case,

involving

final

stress in

di- and trisyllabic

words but

penultimatestress in longer words.

This

system results from the ranking

ALIGN (X2, L), ALIGN

(X2, R), *LAPSE RIGHT

>>

*EXT LAPSE

>>

ALIGN

(X1, R)

?>

ALIGN (X1, L), ALIGN

EDGES. The inviolable

status

of both ALIGN

onstraints

eferring

o main stress

ensures only

one stressper word,

which

must fall on

one

of

the final

two syllables due to highly ranked

*LAPSERIGHT.

Because

*EXT

LAPSE

cannot be

violated

in

disyllables and trisyllables,

the stress is free to fall on

the finalsyllable,thereby satisfying the highest rankedALIGN onstraint,ALIGN X1, R).

However,

n

longer

words, where*EXTLAPSE iolations

become an issue, the stress

shifts

to

the

penult

in

orderto

minimize violations

of

*EXT

LAPSE.

27

Thanksto Bruce

Hayes for providing nvaluableassistance

n this endeavor.

28

This

figure

of

10,823,318,000,000

reflects the

cross-classificationof the number of

stress

patterns

or words

ranging rom one to eight syllables.

It

is

arrivedat by

multiplying

the

numberof

possible stresspatterns or a

word with n

syllables by the numberof stress

patterns

or

a

word with n+l

syllables by thenumberof stress

patterns or a word

with n+2

syllables, etc.,

where the

numberof

syllables ranges from one to

eight.






TABLE

IV

Single

stress languages generatedby

factorialtypolo

1.

Initial:Chitimacha

AL(X1 L)

>>

AL(XI

,R),

2. Peninitial:Lakota

*LAPSELEFT

>>

AL(X1,

3.

Peninitial

plus

nonfinality:

No

q.i.,

Hopi (q.s.)

*LAPSE

LEFT,

NONFINA

4. Antepenultimate:

Macedonian

*EXT

LAPSERIGHT

>

5. Penultimate:

Nahuatl

NONFINALITY

>

AL(X

6.

Final:

Atayal

AL(X1,R)

>>

AL(X1,L),






A

FACTORIAL

TYPOLOGY

OF


TRESS

513

possible

output

sets

were

generated

by

the

proposed

constraints.29

Of

the

152 generated

tress systems,

only

79 were

distinct

stress

patterns

with

dif-

feringlocationsof stress;the other73 of the 152 hadcounterparts mong

the 79

differingonly

in which

of the stressesis the primary

stress.

In

fact,

all stress patterns

nvolving

more than

a

single

stress

per

word have

a

counterpart

ifferingonly

in which

of the stresses

is the

primary

one;

this

symmetry

is attributed

o differences

in which of the

ALIGN

constraints

referring o

main stress

is highly

ranked.

All of the stress

patterns

n the survey

of stress

systems

were gener-

ated,

demonstrating

hat empirical

coveragewas

sufficient.

While

certain

unattestedpatternswere generated, hese unattestedpatternsappear o re-

flect for

the most part

non-pathologic

deviations

from attested patterns.

This point

will

become

clearer

as

we consider

the results

in

greater

de-

tail,

considering

stress patterns

in

turn

by

their taxonomic

description.

While

the focus

in examining

the

coverage

of proposed

constraints

s

on


tress

systems,

stress

patterns

n

words

consisting

of

only light

syllables

in

quantity-sensitive

anguages

were also

considered.

The

justification

for

inclusion

of such

patterns

s

the assumption

hatthe

same constraintsare operativein both quantity-sensitiveand quantity-

insensitive

systems

with quantity-sensitivity

resulting

from

additional

weight-sensitive

constraints

which are

highly

ranked n

quantity-sensitive

but not

in


ystems.30

2.3.1. Factorial

Typology

of

SingleStress

Systems

The single

fixed

stress

systems

generated

by

the

typology

are considered

in

Table

IV, which

contains

a

verbaldescription

of each

stress

pattern,

one

set of rankingswhich generateseach pattern,and a quantity-insensitive

language

(q.i.)

instantiating

he pattern.

n case a

pattern

s not found

in a


ystem

but does

occur

in words

containing

only

light

syllables

in

a

quantity-sensitive

tress system

(q.s.), this

is also

indicated.

Six

single

stress

patterns

were generated

by the factorial

typology,

all

of

which are

attested,

including

all five of

the

attested


29

A complete list of stress pattems and stratifiedconstraintrankingsbehind

these

pat-

tems are availableat the author'swebsite:

http://www.linguistics.ucsb.edu/faculty/gordon/pubs.

30

It should be noted that certainhighlyrankedconstraints n quantity-sensitive ystems

have the effect of banningstress on light syllables. For this reason, some quantity-sensitive

systemshave stresspattems

n

wordsconsistingof only light syllablesthat are notcaptured

by the constraintsdiscussed in this paper, but are capturedby other constraints,such as

ones which repel stress from peripheralsyllables not sonorous enough to supportboth

stress and otherprosodicpropertiesassociatedwith peripheral yllables (see Gordon2001,

in preparation or discussion).






514

MATTHEW

GORDON

single

stress

patterns

initial, peninitial,

antepenultimate, enultimate,

and

final stress),plus

a sixth

pattern,

n whichthe

peninitialsyllable

is

stressed

unless

it is final.

This

pattern

s not attestedin

any quantity-insensitive

systems

I have

found,

but is

found n

Hopi (Jeanne

1982)

in words n

which

the

firsttwo

syllables

are light.

2.3.2. Dual Stress

Systems

Generated

by

Factorial

Typology

Dual stress

results from

the low

ranking

of

the

general

*LAPSE

constraints,

coupled

with the

relativelyhigh ranking

of one or more constraints

which

requirestresses

at

or

near

edges:

ALIGN EDGES

and/or

a

*LAPSE

EDGE

constraint(s). briefly

summarize

he

general

ranking

schemas

generating

the

core

types

of

dual stress

systems

in terms of their stress locations

(see

Table

V for the

rankings

which

generate

the

subpatterns).

As

we saw

in

section

2.1.2,

a

highly

ranked

ALIGN

EDGES

s

responsible

for dual stress

patterns

with initial and

final stress. If

a *LAPSEEDGEconstraint

oriented

toward

the

right edge

is ranked

highly

in

conjunction

with

the

ranking

NONFINALITY

>>

ALIGN

EDGES,

dual

stress

systems involving

stress

on

the

initial

syllable

plus

either the

penult (high

ranked

*LAPSE

RIGHT)

or

the antepenult highranked

*EXTENDEDLAPSERIGHT)

result.Peninitial

plus final stress

reflects

a

highly

ranked

*LAPSE

LEFT

and either a highly

ranked *LAPSERIGHT

or

*EXTENDEDLAPSE RIGHT

n

conjunction

with

the ranking

of ALIGN

(X1, R)

above bothNONFINALITY

nd

ALIGN(X1,

L).

In

such systems,

the stress on the final syllable is

in

response

to a

*LAPSE EDGE

constraint

pulling

stress toward the

right edge.

Because

ALIGN

(X1, R)

is rankedabove

NONFINALITY

and ALIGN

(X1,

L),

the

rightmost tress

falls

on

the final

syllable

within

the window definedby the

LAPSE EDGE

constraint.Peninitial plus penultimatestress follows from

the

ranking

*LAPSE

LEFT,

NONFINALITY

?>

ALIGN

(X1, R)

>>

ALIGN

(X1, L).

The

stress

on the

peninitial syllable

follows from

the

rankingof

LAPSE LEFT above ALIGN

(X1, R),

while the

penultimatestress

results

from the

ranking

of

NONFINALITY bove

ALIGN

(X1, R).

Because

all the

peninitial plus penultimatepatterns

nvolve

an

inviolableNONFINALITY,

all have

penultimate

tress rather

han

peninitial

stress in disyllables.

Hypothetical systems

involving peninitial

plus antepenultimate tress

are not generated,as both peninitialand antepenultimate tress are the

result of highly ranked

*LAPSE EDGE

constraints

acting antagonistically

toward an ALIGN constraint

pulling

toward the opposite edge. Because

one

of

the

constraintsALIGN (X1, L)

or

ALIGN (X1, R) must be ranked

above its sister

constraint pulling

in

the opposite direction,

rankings

will never

give

rise to the

bi-directionalpull

requiredfor peninitial plus

antepenultimate

tress.






A FACTORIAL

TYPOLOGY

OF


515

All

of the

dual stress

systems

come

in

variants

differing

n

which of

the

stresses is the main

one,

according

to whether

ALIGN

X2,

L)

or

ALIGN

(X2, R) is

highly ranked,

and

differing

in

whether stress clashes are

al-

lowed

or

not,

depending

on the

ranking

of *CLASH.n

case clashes are

not

allowed,

there

are different

strategies

for

resolving

the

clash,

depending

on the

relative

ranking

of the

constraints

producing

the stresses

oriented

toward

each

edge

of

the

word.

Table

V

summarizesthe dual

fixed stress

systems

generated

by

the

typology.

The

shaded stress

patterns

are discussed

later.

'1'

represents

both

primary

stressed and

secondary

stressed

syllables

in

Table

V

and

subsequent ables,wheredifferences n whichof the stresses s theprimary

one are

considered a

separate dimension from

the distinction

between

stressed

and

unstressed

yllables. Stress

patterns

which

are not

predictable

from

the verbal

descriptionare indicated in

bracketswhere '1'

indicates

a

stressed

syllable and

'0'

indicates

a

stressless

syllable.

For

example,

[10]

following a

stress

patterndescribedas initial

and final

means

that

the

stress on the

initial

syllable

survives in

the

potentialclash

context

arising

in

disyllabic

words.

Seventeendistinctpatternswere generated,each of which occurs in

two variants

differing

n

which.of the

stresses is

the

primary

one, depend-

ing

on the

relative

ranking of

ALIGN

(X2, L)

and

ALIGN

(X2,

R).

All

of

the dual

stress

patterns

ound

in

the

survey

were

generated.

Other novel

generated

patterns

nvolved

combinationsof

elements

attested n

isolation

but

not

in

all

combinations.

Thus,

unattestedpatterns

deviate

from oth-

ers

in

their

treatment

of stress

clashes,

their

combinationof

stress

sites,

and

their

location

of

primarystress. For

example,

pattern2

differs from

pattern3 only in resolving stress clashes in disyllables in favor of the

initial

rather

hanthe

final

syllable.

Pattern

6 is

analogousto the

attested

initial

plus

penultimate

patterns

4

(Lower

Sorbian)and 5

(Sanuma)

except

for

its

tolerance

of

stress

clashes in

trisyllabic

words. Two

of

the more

interesting

patterns

are 7

and 11.

Pattern7

displays

initialplus

penultimate

stress

wherever

this will

not trigger

a

stressclash,

but

surprisingly

acks

the stress

on

the

initial

syllable in

four

syllable

words

where

there is

no

potential for a

stress clash

with

the

penultimate

syllable.

This

system has

the interesting property of having *EXTENDEDLAPSE be highly ranked

enough

to

influence the

stress

system, but

yet

not highly

enough to be

inviolable, as it

is in

binaryand

ternary

ystems.

The

asymmetry

between

words

with

four

syllables

which lack

stress

on the

initial

syllable

and

longer words

which stress

the

initial

syllable

is

attributed

o

differences

in

the

possibility

of

satisfying

*EXTENDED

LAPSE. For

every

word over

four

syllables,

positioninga

stress

on the

initial

syllable

results

in fewer






516

MATTHEW

ORDON

TABLE

V

Dual

stresssystems

generated

by

the factorial

ypology

.1.

Initia-

&

final,

clash OK

]AL EDGr s >> *CLASI.

Pmaryon initial:

Nonie

Prinman

n

final.

Caadian

Fiench

2. Initial

&

final,

no

clash,

[10]

*CL,Asii

>> AL,

EDGL

`> AL 1x

L)

or

NONFINALITY

t

>>XAl,

Rt

Primar;

n initial:

No.-ne

3 Ilntial&

final,

no

clash,

[0 ])

*CLASH

>

ALIEI

R)1

R

>

A.t,(x

_L)

'ta~ ~~~~~

~

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~OR

~

-nmron

finalUie

4. Initial &

penult,

n1oclash,

[100]

*C'LASt

NONtINALIT

>>

AL

EDwN

S

>>

*LAISE RIGIIr

>>

AL(X1

L)

AL

(XI,

R)

Primair

on inittal Lower

Sorbiai

i

N

5. IntAal&

pentult,

no

clash,

[0101

*CLASH1,

ONFI\ALrry

>> *LA E1.IiGIT

AL EDGE

>>

AI

k(x1,L),

Ai

ttx,

R)

s

~~~~~~~~~Pnnmary

n

penuilt.

Sanuma

6. Initial&

peniul.t. laVsh

OK.

No\taS,i\T-Y, *LAPSERiI:T

>>

Ai.,E.Gol S >>

AL(X

L)

AL,

(Xi,

R)

.................................

................

..............

...... .....................................................

.......

..........................................................

Primarv

n initial:

None

I

Primary

n

penilt:

Nonie

7.

Initial &

penult,

no

clash,

[010],

[0001

. *CL

ASH,

*LAXPSiE

iti'r>>

Ai.tx

L,(

>>

*EXT

LTYS.I

-->AL R)>-

AL EDGES

3_

P

rim r,

on

pnul

None

8. Initia &

antepentit,

lash

OK

NONFINALITY,

EXTL.APSRIIT

>>

AL

EDGE

S,

A(X1,L

>)

AL

X

,

R)

Primar

on

iniitial:

Nonie

.Primal,

on

antepenult:

None

9. Initial

&

antepenutlt

no

clash,

[0100]

*CLASH,

No,NFINA.LD.LnY,

EXTLASE

RIGH(1T,

ULAPs

LUF

T>

>

A

.(x

1,L.)

>>

Ai. EDGES

Al 1

R)

____________________________________

Pnimar

on

antepenlt:

Gieorgian

10. Initial&

antepenult, o

clash, 1

0

1

0]

*CLAs1,.

NoN

NzA

ITY,

*EXT

L.PSE

RI(GiT

Al;>

i,

EDGE S

>>

.AiL)?

>>

AL

x R)

Primary

n

initial.

Wqlmahjari

[10010]

posiblel

. .

Ii.

Initial&

antepenult,

io

Ilash,

1010,]

[00100]

*CiLASH,

Exs

LAPSE iGITr

A

I(x1,L)

>>

.LvsE

>>

ALh?.

R?

>

Ai

EDGE

S

P2mr

o

nepnult:

None

12.

Peninitial

& penult,clsh

OK,

[I01

NaNTFINALITYGirTAPSERGI

,

*LAPSE

LEFT

AL

(XI,

RI

>

i>

A.,(x

1,L

>A>

t.

Eixiws

_ _

i;

-~~~~~~~~~~Pni

ryonpenultimate:

None

13. Peninitial& penult,

no clash,

[10], [1010]

*CLASH,

ONFINALITY,

L.AisE

RioIT,

*LAIp,s EFT>>

AL

(X1,

R)>>

L(X1,L),

ALEDGE

:

_s~isiW

:.

Primary

n

penultiniate:

None

14. PeninitiaI

&

penult,

no clatsh,

10],

[0100]

.

*CIA.SE

*LAPSE

LEFT

NONINA.SX.

TY,

EXTLAPSFE

R

iau

>>

AL

X1,

R

>>

AL'(x1,L),

LEDGE

..........

...........

..............

..................................

..........................

...........................

............. ... ......... . ........ ....... ........ - -.................. . ....

Pnmary

on

veninitial:

None

Primr

on

penultitnate:

None

I1S.

Peninitial &

fin

al,

cash OK

*LA. iL

LEFT,

*LAPSE

RGHT

->

A,

(XI,

Ri).>-

AlIX

,L)

N

FeiiINAL iY

Primiarv

n

peniinitial:

N'one

Pn

mary

o

n

finmalN

-n

16. Peninitial

& final

no

cilash,

[0o1] [0O10]

*CLASII

*LAPSL

LEFT,

*EXTLAPSERicHT

>>

AL

(XI,

._

__

RR

>?>

Ai.,(x

L)

AI

Ec;TE;s

rim

on

peniniitial:

None

y.

17. Penii tial

&

final,

nio

lsh

[010]

*C

AsH

*L

APSERGHT,

*LAPSF EFT

>

Ai

(X1.

R)

>>

aryonAt

4,

NONFINALITY

Prmayonpniiia:NoneN

violations

of *EXTENDED

LAPSE

than

another candidate

without

stress

on the first

syllable;

thus, the candidate

five syllable

form

[00010]

fails

because it

violates *EXTENDED

LAPSE

while the

winner [10010]

does

not.

31

Udihe instantiates

pattern1,

however,

f the initial

syllable actuallycarries

secondary

stress in

disyllables; Nikolaeva and

Tolskaya(2001) express

uncertainly

about

whether t

is completely

stressless or carriesweak stress.






A FACTORIALYPOLOGY

F


517

Crucially,

he winner

incurs

no moreviolations

of

higher

ranked

ALIGN

(X1,

L) than

the failed

candidate

which outranks

EXTENDED

LAPSE.

The

winning

candidate ncurs

more

violations of ALIGN

(X1,

R)

than

the

win-

ner,

but this

is

irrelevant,

as

ALIGN

(X1, R)

is rankedbelow

*EXTENDED

LAPSE.

Similarly,

candidates for

words

longer

than five

syllables

incur

fewer

violations of *EXTENDED LAPSE if

they

stress the initial

syllable

than

if

they

do

not;

hence,

the stress on

the initial

syllable

in

words

longer

than five

syllables.

In

four-

syllable

(and

shorter)

words,

in

contrast,

here

is

no need to

position

a stress

on

the

initial

syllable,

since

doing

so does

not result in

any

improvement

n

the

satisfactionof

*EXTENDED

LAPSE;

thus the winner[0010] satisfies

*EXTENDED LAPSE

as well as the failed

candidate

1010]

but has the

advantage

of

better

honoring

ALIGN

(X1, R).

Pattern

11

is

nearly identical

to pattern7

except

that

*EXTENDED LAPSE

RIGHT

is

ranked

highly

rather han

*LAPSE RIGHT.

Patterns

7 and 11 are

unusual n

lacking an expected

stress

in

non-clashcontexts

and

appear

not

to

have

analogsin real

languages.32

A

relevant

observation s

that the

proposed

constraints

only generate

dual

stress

systems

involving combinations

of

stresses

which

are attested

in isolation. This observation is not trivial, since the set of logically

possible two stress

systems

is

quite

substantial;

hese

include

systems

with

combinations of

stresses not

found

in

isolation,

e.g., initial

plus

preantepenultimate,

ostpeninitial

plus

antepenultimate,

tc., as

well as

combinations

differingas a

function

of number

of

syllables in

non-clash

contexts,

e.g.,

initial

plus

final stress in

three

syllable

words and

initial

plus

penultimate

stressin

four-syllable

words. It is

also a

desirable

prop-

erty

of

the

proposed

constraints,

hough

the

possibility

of

accidental

gaps

precludesestablishingthis with confidence,thatthey fail to generateun-

attested

wo

stress

systems

in

which

both

stresses

fall

nearthe

same

edge,

e.g., initial

plus

peninitial

stress,

antepenultimate lus

penultimate

stress,

penultimate

plus final stress.

Turning

o

the

matterof

overgeneration,

hesmall

numberof dual

stress

languages

(14

in

the

survey)

makes it

difficult to

assess

whether the

un-

attested

dual

stress

systems

generatedby the

constraints

eflect

accidental

gaps

in

the set of

attesteddual

stress

systems

or

whether hey

are

genuinely

pathologicaland thusdestinednever to occurregardlessof the size of the

sample. The

situation with

two

stress

systems

thus

differs from

the

case

of

single stress

systems, where

the

relatively

arge

number

of

single stress

systems makes it

easier

to

distinguish

between

accidental

gaps

and ill-

formed

patterns.

There

are a

few

points,

however, n

favorof

the

accidental

gap

interpretation

or

the

absence

in thereal

world of many

logicallypos-

32

Pattern

7

is

generated

by

foot-based

constraintsas

well

(see

Gordon

n

preparation).






518

MATTHEW

GORDON

sible

dual

stress

patterns.

The

first of these

is that dual stress

languages

as a

whole

are

relatively

rare relative

to

single

stress

languages.

While

we

cannot

be sure

of the

reasons

for the

rarity

of dual

stress in

general,

we may speculate

that

their

rarity

s related to the fact

that a stress

at

or

near

a single

edge

is sufficient

to serve

the demarcative unction

of

stress

suggested

by Hyman

(1977).

Pursuing

this line of reasoning,

adding

an

additional

tress

at the

opposite

end of the wordwouldprovide

little if any

benefit

from

a

parsing

standpoint,

perhaps

even

complicating

the

online

parse by

introducing

a second stress

for

the listenerto

track.

The second

argument

n favor of

the accidental

gap

interpretation

s

thatmost of the unattesteddual stresspatterns,with the exceptionof the

unattested

nitial

plus

final

system

(pattern

2),

involve elements

which

are

independently

rare.

The first

of these

rare

properties

are stress

clashes,

which are

avoidedby

most


anguages

regardless

of

theirtype of

stress

system.

Thus,

all but

two

of the dual stress

languages

(12

of 14) found

in the survey,

he

exceptional

ones

being

Canadian

French

and Armenian,

do not

tolerate

stress

clashes. This

observation s

particu-

larly

obvious

in

the case

of initial

plus penultimate

tress,

the most

widely

attested ypeof dual stresssystem.All ninelanguageswith initialpluspen-

ultimate

stress either

position

stress

on the

penult

or the initial

syllable,

but

not both,

in

trisyllabic

words.

Avoidanceof

stress clash is also apparent

n

binarysystems (section

3)

and

ternary

tress

systems(section

4).33

Given

the

rarity

of stress

clashes

in

general,

it is

plausible

that the unattested

but

generatedpatterns

nvolving

stress

clashes

fail

to

emerge

becausethey

violate an

anti-clash constraint.

The

fact that clash avoidance

is more

widely

attested

han

apse

avoidancesuggests

that t is more than ust sheer

rhythmicprincipleswhichmilitateagainstclashes. Onemayspeculatethat

clash

avoidance effects

are

particularly

trong

because

adjacent

stressed

syllables

detract

rom the relativeprominence

enjoyed

by syllables

in non-

clash

contexts.

Nevertheless,

although

clash avoidance appears

to be an

important

actor

in

creatingmany

of the

gaps

in

the factorial

ypology

of

dual stress, there are two unattested

but generated

patterns

which

stand

out from others

in

lacking

an

expected

stress even

in

non-clash contexts:

patterns

7 and 11. These would appear o

be the most

likely candidates

or

unnatural tatusamongthe dual stress systemsgenerated,although here s

a

possibility

thattheir absence

is an accidentattributed

o the

overall

rarity

of dual stress

systems.

33

It is interestingto note

that clash

avoidance is

another

area in which quantity-

insensitiveand quantity-sensitive

tresssystems

appear o

differ fromeach

other, as

stress

clashes

involving

at least one

heavy syllable

are

commonin quantity-sensitive

anguages.






A FACTORIALYPOLOGYF


TRESS

519

Another

rarely

attested

property

shared

by many

of the

unattested

systems

is

their

positioning

of stress on

syllables

which tend to not

be

stress-attractingn single stresssystems. Severalof the generatedbutun-

attested

patterns

nvolve

peninitial

or

antepenultimate

tress

in

addition

to

anotherstress

at or near the

opposite edge.

Peninitialand

antepenultimate

stress are

independently

rare even

in

single

stress

languages (see Hyman

1977 for possible explanations

or

the

rarity

of these stress

systems).

Given

the

independentrarity

of

peninitial

and

antepenultimate

tress

in

single

stress

systems,

it

is not

surprising

that

systems

displaying

one

of

these

stress locations

in

combination

with anotherstress do

not show

up

in

any

language.

In

summary, here

are

three

factors which

plausibly conspire

to create

many

if

not

all of

the

gaps

in the

match

between

generated

and attested

patterns: he

rarity

of dual stress

systems

in

general,

he avoidance

of

stress

clash, and the

rarity

of

peninitial

and

antepenultimate

tress. Because

all

of the unattested

patternsgenerated

by

the

factorial

ypology

share

at

least

one of these

properties,

and most

display

more than

one,

it is reasonable o

assume

that

their likelihood of

surfacing

s

substantially

reduced relative

to otherstresssystems displayingmore favorable haracteristics.

In

discussing overgeneration,t may be noted thatan

atomisticalternat-

ive to

ALIGN EDGES

involving

separateconstraints,

ach

requiringstress

at a different

edge

of

the

word,

has the

undesirableeffect

of

generating

an

additional our unattested

dual stress

patternsnot generatedby ALIGN

EDGES.

One of these patterns

nvolves stress on only the final syllable

in

words of two

and three

syllables, but both initial and final stress in

longer

words, .e., [01], [001], [1001], [10001],

[100001], etc. This pattern

reflects an undominated onstraint equiringa stress at the right edge. The

stress on the initial

syllable in

words of at least four syllables is due to the

rankingALIGN (X1, L)

>>

*EXTENDEDLAPSE

>>

ALIGN (X1, R).

The

initial stress

in

tetrasyllabicand longer words results

in better satisfaction

of

*EXTENDED APSE

without any additional

violations of ALIGN X1,

L).

An

additional

unattestedpattern generatedby an

undominatedcon-

straint

requiring tress

at

the right edge involves final

stress in disyllables,

initial and finalstress n

trisyllables,andpeninitialand

final stress in longer

words, e.g., [01], [101], [0101], [01001], [010001], etc. The unexpected

stress on the

initial

ratherthan the peninitial

syllable in trisyllables re-

flects the work

of an

undominated CLASHcombined

with an undominated

*LAPSELEFT

and*LAPSERIGHT.

The thirdunattesteddual stress pattern

involves

initial stress in disyllabic and trisyllabic

words, and initial and

final

stress

in

longer words:

[10], [100], [1001], [10001], [100001], etc.

In

this

case,

a

constraintrequiring stress on the

leftmost syllable is un-






520

MATTHEW

ORDON

dominated as is *EXTENDEDLAPSE

RIGHT,

which comes

into

play only

in

words of at least four

syllables.

The

ranking

ALIGN

X1,

L)

>>

ALIGN

(X1, R)

ensures that the

final

syllable

within the

three

syllable

window

at the

right edge

is

stressed,

while ALIGN

X1,

L)

is

ranked

above

the

constraint

requiring

stress on the

rightmost

syllable.

The final

unattested

dual stress

pattern

generated

by

the atomistic

versions of

ALIGN

EDGES

involves initial and

penultimate

stress in

words of at

least four

syllables,

and initialand final

stress

n

trisyllabic

words:

[10], [101],

[1010],

[10010],

[100010],

etc.

The constraint

requiring

hat stress fall

on

the

leftmost

syl-

lable is

undominated,

as are *LAPSERIGHTand

*CLASH,

which

together

yield stress on the final rather hanthepenultimate yllablein trisyllables.

Within he two

syllable

windows at the left and

right

edges

the

location of

stress is determined

by the

ranking

ALIGN

X1, L)

?>

ALIGN

X1, R).

2.3.3.

Uniformity

f

PrimaryStress

Placement

Turning

briefly

to the shaded

sub-patterns

n

Table

V,

all

of which

are

unattested,

here

is

a

property

hat

differentiates hem

from

those

which

are unshaded. n

the shaded

patterns,

primary

tress

fails to fall

a

consistent

distancefrom a wordedge across words of different engths.Forexample,

in

the

unattested

ubtype

of

pattern

5

(primary

tress on

the

initial

syllable

and

secondary

stress

on the

penult

in

words with

two

stresses),

primary

stress falls

on the

penult

in

words with two or

three

syllables,

whereas

primary tress falls on the

initial

syllable

in

wordsof at

least

four

syllables.

Thus,

the location

of the

primarystress relative to

the

word

edge

in

the

unattestedvariant

of

pattern

5

differs as a function of

numberof

syllables

in

the

word. This

contrasts

minimally

with the

attested

variantof

5,

in

whichprimarystressconsistentlyfalls on the penultin wordsof varying

lengths.Only

in

words that are not

long enoughto

provide

a

docking site

for

primary

tress which is

consistentwith

thatfound in

longer

words,e.g.,

in

monosyllabic

words, is the

preference or

uniformityof

primarystress

placement

relativeto an

edge

violated.

The absence

of

stress

systems

such

as the

unattested

variantof 5

reflects a

strong

cross-linguistic

endencyfor

stress

systems

to

be consistent in

their

location of

primary

stress

across

different

words

lengths,

as observed

by

van der

Hulst

(1984)

and Ham-

mond (1985). As Hammondpoints out, this has consequences for stress

clash resolution n

shorterwords.

The

survivingstress in

clash

contexts is

characteristically

he

one

carrying

primarystress in

longer

words where

multiple stresses are

present.

Thus,

in

the attested

variantof

pattern

5, the

stress on the

penult rather

han the

stress on

the

initial

syllable

is

preserved

in

trisyllabic words

where

stressing both the

initial and

the

penultimate

syllables would

entail a

stress

clash.






A

FACTORIAL

TYPOLOGYOF


TRESS

521

While

we

cannot be sure

of

the motivations

behind the

preference

or

uniformity

of

primary

tress

placement,

one

might speculate

hat

position-

ing primarystress the same distance from a wordedge across words of

different

engths

is

functionally

motivated

n

that it

provides

a

consistent

and reliable means for

determining

word

boundaries.34Given

a

string

of

speech

consisting

of several

words,

it is

thus

possible

to use the

primary

stress,

the

perceptually

most

prominent

tress,

to

parse

the

string

nto indi-

vidual

words,

therebyaidingin lexical

access.

Hyman(1977)

has made

a

similar

suggestion

regarding

he

demarcative unctionof

positioning

stress

near word

boundaries;hence the

popularity

of

languageswith stress at or

near wordedges.

Whatever he

ultimate

reason(s)

for

the cross-linguistic

preference or

uniformity

of

primary

stress

placement,

we

can attribute he absence of

the

shaded stress

patterns,which

comprise a

largepercentage

of

the unat-

tested fixed

stress patterns

generated

by the typology, to

their nconsistent

positioning

of

primary tress

across

words of different

engths.

Ultimately,

the

principle

that

excludes

the shaded

candidatesshould be

codified as a

formal

constraint

blocking

their

generation.

However,

heformulationand

implementationof such a constraintgoes beyond the scope of the present

paper, since it would

appearto

require complex

formal

mechanisms not

currentlyavailable

n

Optimality

Theory, n particular,

utput-output or-

respondenceconstraints

evaluated

across forms

differing in

number of

syllables

rather han

forms

which are

paradigmatically

elated o

the same

root.35

3. BINARY

STRESS

Another

widely attested

ype of

stresspatternplaces

stress on every

second

syllable.

These

'binary' stress

patterns are

rhythmicin the

sense that

stresses

are

separatedby a

single

unstressed yllable.

Binarystress

patterns

34

The

preference for

positioning

primary

stress a

uniform

distance

from a word

edge

is

stronger n

quantity-insensitive tress

systems

than in

quantity-sensitive

ystems,

where

it is

routinelyviolated

(see the

discussion

of

binary

stress

in

section

3).

The

issue of

why


and

quantity-sensitive

tress

systems

should

behave

differently

with

respect

to

primary tress

placement s an

interesting

one

that

must

await

further

esearch.

35

See

Myers

(1999) for discussion

of

output-output

orrespondence

onstraints

equir-

ing

identity

between

forms which

are

not

related

o

the same

rootbut

which

are

similar

with

respect

to

one or

more

phonological

properties.

Myers's

constraints

account

or

analogical

extension of

phonological

altemationsfrom

one

paradigm

to

other

paradigms

contain-

ing

forms

bearing

similar

properties,e.g.,

extension

of

velar

softening

altemationsand

irregular

verb

paradigms

n

English.






522

MATTHEWGORDON

TABLE VI

Binary

stress

patterns

Pattern

Schematic forms

Example Lgs. #

of

igs.

1. Odd-numbered

yllables 6a6a&, 6acr&a6a

Czech

(Kucera1961),

from L to R

MaranungkuTryon 1970)

2.

Even-numbered yllables

oo&oo,

o6&cyo&

Araucanian

Echeverra

nd

3

from

L

to

R Contreras

1965),

Sirenikski

(Menovshchikov1975)

3. Even-numbered yllables

oraor6,

&aor6r Cavinefia

Key 1968),

1237

from R

to L

Warao

Osbom 1966)

4. Odd-numbered

yllables

Ka&a6,

&u&a6

Chulupi Stell 1972),

5

fromR to L UrubuiKaapor Kakumasu1986)

differ as a function

of whether the

alternatingpatternorigins

at

the

right

or

the left edge

of the word

and whether he

patternbegins

with

a stressed

or an unstressed

syllable

(Hayes 1980;

Prince

1983).

There are thus

four

logically possible pure binary

stress

patterns.

The

simple binary

stress

pat-

tems are summarized

n Table

VI, along

with

the number

of

languages

in

the survey nstantiating achpattern.A list of languageswithbinarystress

is found

in

Appendix

1.

One of the more

nteresting

observationsaboutTableVI

is the existence

of three

languages

with stress on even-numbered

yllables counting

from

the left

(pattern )

and

five with

stress

on

odd-numbered

yllables

counting

from the right(pattern ). These patternsare nteresting, ince they

involve,

in

foot-based

terms, iambs;

thus,

(ad)(ad)(ad).

The

existence

of

such

pat-

terns,thoughthey

are

relatively

rare

compared

o

patterns

1

and

3, is rather

unexpectedgiven

the

hypothesized

confinementof

iambic-type

stress

pat-

terns to

quantity-sensitive ystems

(Hayes 1985, 1995).

This

hypothesis

has spurreda body of literature

attempting

o

explain the apparent

asym-

metrical occurrenceof patterns

2

and

4

in

Table VI in

quantity-sensitive

but not


ystems (e.g., Kager 1993;

Eisner

1997).

An

important eature of such analyses is theirexclusion of patterns2

and 4

in


ystems. The attestationof all logically

possible

binarystresspatterns n the present survey, however, suggests a greater

degree

of

symmetry

n

binary

stress

systems than such theoriesallow for.

36

Threeof these

languages

display optionalfinal

stress

avoidance

n

odd

parity words:

Burum,

Selepet and Northern

Sami (see Appendix

1).

This

pattern s

observed in

roots

in

Waorani

(countedamong

the 18

pattern

1

systems

in

Table

VI); suffixes

stress

even

numbered yllables

counting

from the right.

37

An

additional

anguage,

Waorani Auca]

(Pike

1964),

stresses

even

numbered uffixal

syllablescounting

from the

right.






A FACTORIAL

YPOLOGY F


TRESS

523

One favorable

feature

of the

proposed

analysis,

shared with stresstheor-

ies positing symmetrical

oot inventories

e.g.,

Hayes 1980;

Prince

1983;

Halle and Vergnaud 1987; Elenbaas and Kager 1999), is its admittance

of all logically possible binary patterns,

as will become

apparent

n

the

discussion

that follows.

3.1. Pure Binary Stress

A

feature

of all

binary systems

is the undominated

tatus of both *LAPSE

and

*CLASH,

which

together produce

the

alternating

tressed-unstressed

pattern.The differentsubtypesof purebinarystresssystems emergefrom

differences

in

the ranking

of

ALIGN EDGES

and

the

ALIGN

(X1,

IL/RI,

0),

constraints.

As a detailed

analysis

of a

language

with a

simple binary

stress system,

consider

a

language

which

stresses

odd-numbered

yllables

beginning

from

the

left

edge

of

a

word,

a

pattern

which is instantiated

by

the forms

in

(21)

from

Maranungku Tryon1970).

(21) Maranungku tress

a. tiralk 'saliva'

b.

mxraxpaet

'beard'

c.

m

utl1iikin 'salt-water urtle'

d.

jaeijarm

ta

the Pleiades'

e.

Da

ltiritiri

'tongue'

In

Maranungku, he highest ranked

ALIGN

constraint s

ALIGN EDGES

which, nevertheless, s crucially rankedbelow

both*LAPSEand*CLASH.

If ALIGN

EDGES

were ranked above

*LAPSE,

we would get both initial

and final stress in even parity words:

*jdyarrnata

nstead of

jdaVarmaita.

f

ALIGN

EDGES

were ranked

above

*CLASH,

we would get stress clashes

in

even parity words:

*jdJacrmatc

nsteadof

jadrarnauta.

LAPSE

is also

crucially

ranked

above

NONFINALITY,

as the

final syllable is stressed in

words with an odd numberof syllables:mdrwepdvtather han*merwpwt.

ALIGN EDGESis

rankedabove

ALIGN (X1,

L),

as evidenced by the stress

pattern

n

odd parity words:mcerwpd?tather

han *mwr4wut.

ALIGN

(X1,

L) is

in

turnrankedhigherthan

ALIGN

(X1,

R); otherwise,stresswould fall

on

even

numberedsyllables counting from

the left in even parity words:

*jajafrmatta

ather han

the attested

adVarmacta.

inally,

ALIGN

(X2, L) is

rankedabove

ALIGN

(X2, R), therebyaccounting

or the promotionof the

first stress to primary tatus:

mWera?pdt

ather

han

*mircep(ft.






524

MATTHEW

ORDON

The rankings

for

Maranungku

are summarized

in

(22).

A

tableau

illustrating

he

rankings

appears

n

(23).

(22)

Constraint

ankings

or

Maranungku

*CLASH,

*LAPSE,

ALIGN

(X2, L)

ALIGN EDGES

ALIGN (X1, L)

4,1

NONFINALITY,

ALIGN

(X1, R),

ALIGN

(X2, R)

(23) mLerspet

*CLASH

ALITN

*LAPSr ALIGN

ALIGN NoNFINAL

AL

IGN

ALIGN

(X;2,

L)

l

.;XfS

(XI,

> w

(X2,

w I

(XI,

R)

iw

nMrrpt

-

___Emm

E...

,

,( L )22 E

m : epa-L

-

-

ja

'(rn1ata *CLks:A LIGNt}tT*LAPSE ALIGnN

A

I(.N NUN.L}INATALIGN

3ALIG,N

w?

j4

arma~ta___

iL'1armbLtA

1___

~j

jaiy

rmatAXI

j4qlannatil

____

____

Other

simple binary

stress

patterns

differ

minimally

from

Maranungku

n

their

ranking

of

the

ALIGN

constraints, ollowing

the

spirit

of foot-based

analysesof binarity e.g., Crowhurst nd Hewitt 1994).TheUrubuKaapor

(Kakumasu1986) pattern pattern

4 in Table

VI),

involving

stress

falling

on odd-numbered

yllables

counting

from the

right

ratherthan the

left,

resultsfromALIGN

(X1, R)

being

ranked

above

ALIGN

(X1,

L).

The

effects

of this

reranking

re seen in even

parity

words:

ocra

in

Urubu

Kaapor,

but

6ucua

n

Maranungku.

The location

of

primary

stress

on

the last

stressed

syllable

in

Urubu

Kaapor

is

the

result

of

ALIGN

(X2,

R)

being

ranked

above

ALIGN

(X2, L).

The Sirenikski (Menovshchikov1975) stress pattern (pattern2), in

which stress falls on even-numbered

yllables counting

from the

left,

res-

ults from

the

ranking

of ALIGN

(X1, R)

above

both

ALIGN

(X1,

L)

and

ALIGN EDGES.

Ranking

ALIGN

(X1, R)

above ALIGN

(X1,

L)

ensures

that

stress

skips

over

the initial

syllable

in

even

parity

words:

u6oo

rather

than

*6a&a.The

ranking

of ALIGN

(X1, R)

overALIGN

EDGES is

decisive

in

avoiding both

initial

and

final

stress

in

odd

parity

words:a6o

and not

*

ciu

,

CF






A

FACTORIAL

YPOLOGY F


TRESS

525

The

Cavinenia

Key

1968) pattern

pattern

3),

involving

stress on

even-

numbered yllables

counting

from the

right,

differs

minimally (beside

the

reranking

of the ALIGN

constraints

referring

o

primarystress) from the

Sirenikski

pattern

due to

ALIGN

X1, L)

being

ranked

above ALIGN

X1,

R). Owing to

this

ranking,

stress

skips

over the ultima

and falls on

the

penultinstead n

even

parity

words:

ououc

and

not

*ucsu6.

3.2.

Binary Plus

Lapse

Systems

3.2.1.

Binary Plus

External

Lapse

Many

languages

display

a

basic

binary pattern

with

stress

lapses

arising

undercertainconditions.Onesubtypeof these 'binaryplus lapse' systems

is

found in

languages

n

which stress

falls on

odd-numbered

yllables

from

left to

right

except

for

final

syllables,

e.g.,

6a'aaa,

6aocxyxcxc.

inal

stress

avoidance results in

a

stress

lapse

at the

right

edge

of

a word

containing

an odd

number

of

syllables.

Languages

displaying

this

type

of stress

lapse

at the

right

periphery

nclude

Pintupi(Hansen

and

Hansen

1969,

1978)

and

Karelian

Leskinen

1984).

I

have

located

14

languages

(listed

in

Ap-

pendix

1) with

this

pattern.

This

pattern s the

result

of

nearlythe

same

constraint ankingswhichgenerate he strictbinarypatternnvolvingstress

on

odd-numbered

syllables

from

left

to

right,

the

important

difference

being

that

NONFHNALITY

s

undominated

or

nearly

so,

subordinate

only

to

CULMINATIVITY

n

languages with

monosyllabic

words)

in

the

binary

plus

lapse

system.

This

ranking

produces

an

external

apse in

odd

parity

words,as

the

lapse

occurs at a

word

edge.

Crucially,

because

there is no

counterpart o

NONFINALITY

t

the left

edge,

there is

no

mirror

mage to

the

Pintupi

pattern

displayinga

lapse at the

left

edge

of a

word.In

fact, such

patterns

appear o

be

unattested,

a

point to

whichwe

returnn

section3.4.

We

now

look

more

closely at

the

constraint-based

nalysisof

the

binary

plus external

apse

system

found in

Pintupi:

stress on

odd-numbered

on-

final

syllables

counting

from the

left

(Hansen

and

Hansen

1969,

1978).

Examples of

stress in

Pintupi

appear n

(24).

(24)

Pintupi

stress

a.

tJutaja

'many'

b.

puiijkalatJu

'we

(sat)

on the

hill'

c.

tkamulimpatiujku

'our

relation'

d.

tftii

uulampatiu

'the

fire

for our

benefit

flaredup'

e.

kiranJiulullmpatJuqa

'the

first

one

(who is)

our

relation'

The

crucial

constraint

reranking

which

differentiates

Pintupi

from

Maranungku

s

NONFINALITY

above

*LAPSE.

This

ranking

yields

final






526

MATTHEWORDON

stress avoidance

n odd

parity

words,

even

though

his creates

a

stress

apse

in final

position:

in

Pintupi,

t1utaja

instead of

*tjtap.

NONFINALITY

is also ranked above

ALIGN

EDGES,

as

indicated

by

the

lack

of

stress

on final

syllables.

ALIGN EDGES

takes

precedence

over

*LAPSE,

as

odd

parity words stress

the initial

syllable

even

though

this

entails

a

stress

lapse later in the word:

tJu'taja

ot

*tJutaja.

*LAPSE

is

still ranked

above

ALIGN

(xl,

L),

since stress

apses

are avoided

n

non-final

contexts even if

this costs

additionalviolations

of ALIGN

(xl,

L):

t1admuli'mpat'

rjku

not

*tJamulimpjaitJ'uku.

inally,

the

ranking

of

ALIGN

(X2, L)

over ALIGN

(x2, R)

correctlypromotes

the first stress to

primary

status

in

words

with

multiplestresses:tdamulimpatJi.ykuot

*ti

iamulimpatJ'4ku.Therankings

of

the

stress

constraintsfor

Pintupi

are summarized

n

(25).

A

tableau

demonstrating

hese

rankings

appears

n

(26).

(25) Constraint

ankings

or

Pintupi

NONFHNALITY,

*CLASH,

ALIGN

(X2, L)

ALIGN

EDGES

*LAPSE

ALIGN

(xl,

L)

ALIGN

(xl,

R),

ALIGN

(X2, R)

(26)

puti3kalatiu

No'QFINAIA

L

AL(N

*CLASIH

ALIGN *LAPSE

ALIGN

ALIGN

ALIGN

(X2}

L)

EDGES

(X1,

L)

(XI,

R)

(X2.

R)

pii~ik~iau _____ ____:__~

P'iLjkaiilatiio

________ _______

p

u li3k

1&

ti

------_-----_----_--------

p4tiIklatju

.. . '.***

tiamuulimpatiuqku

ONFINAL

ALIGN *CLASH ALIGN

*LAPSE

AUGN

ALIGN

ALIGN

(x2L) EDGES

(X1,

L)

(X1,

R)

(X2,

R)

tiJmuflmipatjuqku

I _

________

__l

-

----X--*k






A FACTORIALYPOLOGY F


TRESS

527

3.2.2. BinaryPlus Internal

Lapse

Another

type

of

binary

plus lapse

stress

system

involves an

internal

rather

than

an

externallapse.

A

general

feature of these

systems (termed"bi-

directional"

by Elenbaas and Kager

1999)

is their

positioning

of

a

fixed

stress at

one

edge

and a

binary pattern

nitiating

at the other

edge.

The

fixed

stress at one edge is

the

result of a

highly

rankedconstraint

equiring

stress to

fall

at or near an

edge;

constraints

which

have this effect

are

the

*LAPSE

EDGE constraints and ALIGN EDGES.

The

origination point

of

the

basic

binarypattern

ound in

binary

plus lapse systems

is a functionof

the

ranking

of the ALIGN

(x1, {L/R},

0)

constraints and ALIGN EDGES.

An example of a binary plus internallapse system is provided by

Garawa

(Furby 1974),

in

which stress falls on even-numbered

yllables

counting

from

right

to

left but

skips

over

an

even-numbered

peninitial

syllable

in favor of

stressing

the initial

syllable.

The result

is a

stress

lapse

following

the initial

syllable

(which

carriesmain

stress)

in words with

an

odd

number

of

syllables.38

Examples

of Garawa

tress

appear

n

(27).

(27) Garawa

tress

a. jami 'eye'

b.

puinJ

la

'white'

c.

watJimpaIJu

'armpit'

d.

naTneinmufkunJinamira

at

your

own

many'

The

proposed

analysis

of

the

differencebetween

GarawaandPintupifol-

lows the

basic

insight of

McCarthyand

Prince's (1993) work within a

foot-based ramework nd nvolvesa differencebetween the two languages

in

the relative

rankingof

ALIGN

X1, L) and ALIGN X1, R).

In Garawa,

unlike in

Pintupi,

ALIGN

X1, R) is

rankedabove ALIGN X1,

L), thereby

accounting

for

the

fact that

the stress

lapse

in

odd parity

words occurs

immediatelyafter the

initial stress

rather han immediately

before the fi-

nal

stress;

thus, nadriginmukun

namira, not

narivlnmukin'

namira. The

Garawa

rankings

are

summarized

n

(28),

followed by a

representative

tableau n

(29).

38

Stresses on

syllablesotherthan

he

initial one and thepenultare

reported o be weaker

than those

on the

initial and

penultimate

syllable in

Garawa, a fact not captured n the

presentanalysis, which

assumes two levels

of stress.






528

MATTHEW

GORDON

(28)

Constraint ankings

or

Garawa

*CLASH,

NONFINALITY,

ALIGN (X2, L)

ALIGN

EDGES

*LAPSE

ALIGN (X1, R)

ALIGN (X1, L),

ALIGN

(X2, R)

(29)

punJala

NOFIN

ALIGN

*CLSH

ALIGN

*LPSE

ALICGN

ALIGN

ALIGN

' (X2, L)

EDGES (X1, R)

(X1,L)

(X22

R)

pdiinala

___

puniala

nariqiinmukunhnamira

NONFIN

ALIGN *CLS11 ALIGN *LPsE ALIGN ALIGN ALIGN

_

(x

2L) EDGES

(X,

)

x

,Lx

,

(x2

R)

ndri.nmaW=manimira

_

'__8_

nArixjinmuiikun1inkimira

__"_ _1__*"*

I3rlin

niniamira

___

***

* I~

1

ndriginirukniinamira

___

- -- 19

*

Garawa

s

not

the

only

binaryplus

lapse

stress

system

withinternalstress

lapses, though such systems appear o be fairlyrare.A virtually dentical

system exists

as

an

option

to strict

binarity

n

Polish

(Hayes

and Puppel

1985), as well

as

in

Spanish

(Harris1983)

and Indonesian(Cohn

1989),

with

Spanish

and Indonesiandiffering

from Polish

and

Garawa

n

being

sensitive to weight

in

locating primary

stress.

Piro

(Matteson

1965) also

displays

a binary plus

lapse stress

system,

in which

stress

falls on the

penult,

on the initial

syllable,

and on odd-numbered

yllables

following the

initial

syllable;

Piro thusdiffers from

Garawa

n

commencingthe

alternat-

ing stress pattern

at the

left edge rather han

the right

edge.

The

Spanish,

Indonesian, Polish,

and

Piro patternsare all

generated

by the proposed

constraints,

as

we will

see

in section 3.4.3.

3.3. Binary Plus Clash

Systems

Another

variantof the binary

stress

pattern s found in

languages

n which

stress follows a basic

binarypattern

but falls on adjacent

syllables

under






A FACTORIAL

YPOLOGY FQUANTITY-INSENSITIVETRESS

529

certain

circumstances.These

'binaryplus

clash'

systems

are the result

of

nearly

dentical

rankings o those relevant n

binaryplus lapse

systems,

ex-

cept

for the low

ranking

of

*CLASH nd the

undominated tatus

of

*LAPSE

in

binaryplus

clash

systems.

Like

binaryplus lapse

systems, binaryplus

clash systems involve a

fixed

stress

at

or

near one

edge,

due to the

high

rankingof one of the *LAPSE EDGE

constraints

or

ALIGN

EDGES,

and

a

binarycount

originatingat the

opposite

edge

of the word.

Binaryplus

clash

patternsappear o be

rare

n


ystems,

but

are

found

in

Tauya

(MacDonald

1990), Biangai

(Dubert

and Dubert

1973),

Southern

Paiute

(Sapir

1930,

Harms

1966),

and

Gosiute Shoshone

(Miller

1996).39

Toillustrate he rankingswhichgeneratebinaryplus clash systems,let

us consider n

detail the

analysisof the

binaryplus

clash

system

in

Tauya,

a Trans

New Guinean

anguagespoken

n

Papua

New

Guinea

(MacDonald

1990).

In

Tauya,

primarystress

falls

on

the final syllable, and

secondary

stress

falls on

the initial

syllable

and

on

alternatingyllables

counting

back

fromthe

final one. The

requirement

hat

the initial syllable

be stressed

n

Tauya

creates

a stress clash

between the

first two

syllables

in

words with

an even

numberof

syllables;

forms

illustrating

stress in

Tauya appear

n

(30).

(30) Tauya

stress

a.

nono

'child'

b.

Tuneta'

'mat'

c.

momunepa

'X

sat and

X...'

d. japatijWf6 'myhand'40

The

constraint

rankings

at work in Tauya

closely

resemble those

operat-

ive in

Maranungku.

One

important

differencebetween the

two

languages,

however,

lies

in

the

relative

rankingof ALIGN

EDGES and

*CLASH. The

presence of stress

on both the

initial and

the final

syllables is

indicative

of

ALIGN EDGES

being rankedabove

*CLASH

in Tauya,as a stressclash

regularly

arises in

words with an

even number

of

syllables. The Tauya

39

It

shouldbe

noted

thatthe

Southern

Paiuteand

Gosiute

Shoshonestress

patterns ould

be

treatedas

displaying

a

binary

pattern

at the

level of

the

mora,

and

not

necessarily

at

the

level

of

the

syllable

(see

Sapir

1930;

Harms

1966;

Hayes

1995 for

description

and

discussion).

40

Note

that

unstressed

vowels

optionally

reduce to

schwa.






530

MATTHEW

GORDON

constraint

rankings

are summarized

n

(31),

followed

by

a tableau n

(32)

illustrating

he relevant

rankings.

(31) Constraint

ankings

or

Tauya

ALIGN

EDGES,

*LAPSE,

ALIGN

(X2, R)

4

ALIGN (X1,

L)

4

NONFINALITY,

*CLASH,

ALIGN

(X1, R),

ALIGN

(X2,

L)

(32)

?uneta

ALIGN ALIGN

*tLPSE

ALIGN

NONFINIAL'

CLASH

ALIGN

1

ALIGN

EDGES

(x14L)(XL)X1R

momunepa

ALIGN

;ALIGN

*LPSE ALIGN

NONFINAL, 4CLAS1H

ALIGN

IALIGN

EDGES

(X24R)

tX(X2,) t(X-,

(XI

R)

momnunepa

A-IGN

ALIGN

*LPsE

ALIGN

NONFINAL

*CLSH

'

A

LIG

ALIGN

mfOnulp

4 '' '6

1

:

*

4*E

momrunepa

__ __ ~;

4

in6rnZ~p,

,______ ***** 4

_4*

4*

3.4.

A

Factorial

Typologyof

Binary

Stress

The

binary

stress

patterns

generatedby

the factorial

typology appear

n

Table

VII,along

with the

rankings

generating

ach

pattern

and an

example

of a

language,

if

any

exist, displaying

each

pattern.

Each

system

includes

a schematic

pattern

for an

even and

odd

parity

word

(1

=

stressed

syl-

lable,

0

=

stressless

syllable),

as well as one for shorter

words

whose

stress

patterns

are not

recoverable rom the

description,

e.g.,

in

potential

stress

clash

contexts

in

systems

not

tolerating

clashes.

Shaded

systems

fail

to

positionprimary tress a uniformdistancefrom an edge, followingearlier

discussion

in

the context of

dual stress

systems

in

section

2.2.3.

3.4.1.

Pure

Binary Systems

Generated

by

the

Factorial

Typology

A

total of

23

distinct

binary

stress

systems

were

generated

by

the

factorial

typology,

including simple

binary,binaryplus

lapse,

and

binaryplus

clash

systems.

All

23

patterns

were

generated

n

pairs

differing

in

which

of the

stresses is

primary

and which is

secondary.

Looking

at the

subclasses

of






A

FACTORIAL

TYPOLOGYOF


531

TABLE VII

Dual

stress systems

Simple BinarT

1.Even-numberedromL:

11010 [01010101J

(issl

-.

Ar

\

RI'

*i,li

most

is

prima:Arau ian

bri oat o

r

:

Odd-nLiumberedrom

R:

[CLOASIIA

[0E1)0101

(IAH

Ls>>

ALEr

rs's

AL(x1,Ry>,-

0.4

0,1

hspim'No1

"Resm

1

s

prima:

UtrubuKaapo-

3.

.ven-numb'drom

R

[9J,

01010101]

*CIAMS

*LAPSi'

A

LI

-A

_

Leflm

_ _ _ _ _

_ _ _ _ _ _ _ _ _ _ _ _ _

_ _ _ _ _ _ _

_ _ __R__h_

__C_vi_ef_

prima

.

lalalo

1 o

iPso

(

a,)

Rig------h_mot

s

primain

sv

4

Odd-numb5dfromL:

1

01010],

[1

,

I010

"(I Ii

*LAPSi

.

wAIE

"rs>>

AL.(XL)

>I>

Lefmostis

p)rimary:

aramm'ku

.

O

noi

s

r i

~ ~

r(;

Binary plus

clash

5.

Even-numb'd

rom

L

&

final,

clash

O1K:

*LAPSE

>'

AL.(.x

R)

>AL

EXL$Gs"-ALSX

,L),

*CLAsii

.........

t

(....l0-.....O-IO--01

.................

Lefimost S

primary:

Noite

Rightmost

S

primaiy:

No

qi1.

Central

A

l1

an

Yupik

___tterance-nSeliaal

hI.L)

6.

Eve-numb'dfroinL

&

peitult,

lash OK:

I.NASs

*L.Ant

RwAcr

l (

R)N1issli3A

1

R<)>>

I9LPJIQIO19)1g A

rS.1^L, * All$ 1L1 Cssi

1-

El

.

Ltmost

is

primary:

otitlem Paiute

_

RigInmost

s

prnmary

Non

7,

Odd-numb'drom L

&

final,

clashOK

f

s

ii,4

Extu's

A

ix

O

>>

A

.:xix1L),

*CLsiil

Lefimost

s

prismary:

osiute

Shoshone

Rig1100

is

prim

None

8.

Odd-nuib'd froit

I

&

penolt,

clash

OK:

LAS

,nt'

N\sINA.ITY

*LAeSt

RiGHT

>>, Eww

s>

[10101 101,

IOIOIOIOL

At

Lx

RI

g,x

Ii

*CiASi

Left

ost

is

.rimarv:

None

RiLsltmost

s

prinmaryNone.

9.

Eves-numb'd

ont

R

&

iniitial,

lash

OK:

NT.INH, US Al

1>:

LI,

AL

EDGE.ss"

[1)10ll19

[[l].[l01.1AU.X RI

5

AtSIlSA

, ,,,,,,,,,,,,,,,, ,,v,,,,,,,,,,,,,,,,,,,,,,,,,CLA,

.......................... ... .

,............f+?,

Lefimosi s

primaar:

one

Rightmosr

s

primar

at:

'ii

I

0.

Odd-numb'd from R & initial,

clash OK: Lsis Al

0

is.

AI

L)

>Ax1iRI,

GAISI

Lefimst

is

primary:

None

Rigitmnsot

s

primary

Tanya

BinAri

ohis lapse

I

1.

Odd-niwhb'drom

L

&

penult

no

claslit

[010]

.CLs,,tsH. oNos

NALITY1LAPE

RiGHirn>>

L

Eis""S'

[10100101i,

[10101019

j,

, L PSEis

>

.....................-.

i..x

I.......

R

1

12.

Odd-nunib'd rost L &

penult,nto laslir 1

001 *CLArS

NeNrisso\tNIzT

>

5,' ElYi-5s>>L5

LPSI

[19]09 M101010101.I*Oplj'

vLsRiltrir Ar

.............

(R 1> ) >""A

A1(X

Lfimost

s

primar.

None_____________________

13.

Odd-numb'd

from

L.

minus

final:

.

CLAS;ii,

No.NF'1A1X5

1>

At

.c,rs

""

AP,

>

[1010tO

2.1.9]9.91

.3

A

(t~L"

(A>jtk

RI

;

-~~~~~~~~~~~~~~t

ftntnst

s

primary':

intopi

' ois'

a''N

14.

Odd-nunb'd from

L &

[fina1:

I0

*CLAsH,

*LAPSERI'HTs>>"AL EDS

>"

..[l..)91)1.1010 i0031]

..

*Ls'.s...

>>AS(N

L-

..

i

..

h

.

Lentmst inprima':

. .'m

i

15.

Odd-nuntb'd front

L

&

penult,

clash

in

trisyllables:

NONIFtNAlIStY

LAPsts RGTOiT

>

AL

IXSts

*CLASiI

_[Tt

[P)199

............

""s ( L)"

mo,st

Lis

rimary:

None

Rig'tmtost

ti

primar

None

16.

Odrd-numb'd

fromL

& final,

clash s

disyllales: Ai,EDGES

'

>

AIt- >

LAPSE

>>

Al,(XL):

[101010)1,

[10101001].~~~AL(X1j,R3

,[,,,1.

.,,

,,,.....

e.i

-,o

.

........

.. .

........

........... ..... .......

...

Lt4mast

s

primar:

None

Rightmost

sprimary:

'one

17,

Even-numb'd from R & initial,

no

clash:

[010]

'CrLASi

NON-FINLiYi AL

EDGEs>>

*LAP'S'

i>

[It) ]0i0].[i

101010]

,

.

.

'

sa~~~~~~~~~~~~~~~~~~~~~~~~~~.........h...."...

..........L...

=Le

s

=isprim

'iN

n

,

'

Rightmo

is

primary.

ndonesian

18,

Eves-oiumh'diont R &

iiiitial,clash

ii

trisyllable

.

NNCFs1INAtirs

>Al

T

f

0s>

>

Cl,ASH>>

*LA

PS.>>

Letimost

sprintary:

None

Ri

h

ntmo

i

prinmary

oie

19.

Even-numib'd

roisO

&

initial, Io tasb:

1 CLSIH

o.'?NALITY

>>

,iEi or

" L

s'L> 5*Ltsrs

>>

[101010], [10101010]1

4W

xg RI. [.(A1,L)

Lftmios

s

primas':

Garawa

a

imryN

20.

Evren-noins'd

from

Rplus peitinitial:

[1010]

*C.LA'Hi Nos'TirNA LTYsL.

LEs'E

?

AL1,OhRI"";-

[Q(919 PL[9.(99]9[9]

.

~is

A

LI 5L.elsFrAl

DOt'.

==00IgjiolilA:

8

Left a

Lprimary:No.

..

,

g~~~~hnmost

s

simar_t

one

21

.

Odd-nuntb"d

from

R

&

initial no clash:

[01 ]

CiJAs

>H"Ai,

EIF

'

LAsi'S

"-"

AL

1,Ri

>>

22. O

id-n lunsb

l'om

Oh

& initial,

e4shitn

dilyllables

.At.

' -

,s

..L....

........

...........

o~~~~~~~~~~~~~~~~~~s

_

'

pfs ar

Leftnmost

s

prmary

None

Rightintsitss

ritnarsv:

one

23. Odd-numbered

rom

R

&

initial:

[10]

*C.LASl

>"

AlE..

.ws

>>, *LAPSg

NoNFiNALirry">>

.[ t).19191LL9

..........tO.'..RI:.........L.

Leftmoat

s,

prima

Nose

OsIms






532 MATTHEW

ORDON

binary

stress

systems

in

detail,

all four of the

simplebinarypatterns

pat-

terns

1-4)

are attested:

even-numbered

rom left to

right,

even-numbered

from

right

o

left,

odd-numbered

rom eft to

right,

andodd-numbered

rom

rightto left. These

are

differentiated

romeach other

by

the relative

ranking

of the ALIGN

onstraints,

ollowing

discussion

in section

3.1.

.

3.4.2. BinaryPlus

Clash

Systems

Generated

by

the Factorial

Typology

Of the six

binary plus

clash

systems,

five are attested:

patterns

5,

6,

7,

9,

10, with one of the

patterns

5),

however,

ound

only

in

all-light

words

in

a

quantity-sensitive

ystem.

The unattested

pattern 8)

differs from

Southern

Paiute (pattern6) in stressingodd-numbered ather haneven-numbered

syllables;

the

directionality

of the stress count

is

the same

in

both

pattern

6

and

pattern

8. The difference between

patterns

6

and 8 is attributed

to

differences

in

the

ranking

of

the ALIGNconstraintsrelative to

each

other. In

pattern

6,

ALIGN

X1, R),

is rankedabove

both ALIGN

X1, L),

and

ALIGN

EDGES,

whereas

in

pattern 8,

ALIGN

(X1,

R),

is

sandwiched

between

higher

ranked ALIGN

EDGES,

and lower

ranked

ALIGN

(X1, L).

An

important

eature of the

binaryplus

clash

systems

is

that the fixed

stressoppositetheoriginationpointof thebinarypatterns always situated

at or near

the

periphery:

n the initial

syllable,

the

penultimate

yllable,

or

the final

syllable.

A

fixed stress on

either the initial or

final

syllable is

attributedo a

highly

ranked

though

not

necessarily

undominated)

ALIGN

EDGESranked

above the ALIGN

(X1, {L,

RI)

constraint

pulling stress

toward the

opposite

edge,

as

in

patterns5,

7, 9,

and

10. The

result is

an

additional stress at the

opposite

periphery

rom the

origination

point

of

the

binary

pattern.

A

fixed stress

on

the penult

(patterns6 and 8) is

attributedo an undominated*LAPSERIGHTn conjunctionwithundom-

inated

NONFINALITY.

his ranking

orces stress on the

penult, even

if the

antepenult

s stressed.

A

key feature of the

proposed

constraints

s

that they only

generate

binary plus

clash

systems

with

a fixed stress on

the

initial syllable, the

final

syllable,

or

the

penult.

This

confinement

of clashes to the

periphery

captures

he

confinement,

n

foot-based

terms,

of

degenerate eet to peri-

pheral

positions

(see Kager 1989;

Hayes 1995

for discussion of

degenerate

feet). Theonly threeclashcontextsgeneratedn thepresent heoryinvolve

the

initial and

peninitial syllables,

the

antepenultimate nd

penultimate

syllables,

and the

penultimateand final

syllables; n fact,

these are

the only

three clash

environments

attested n

actual


ystems.

At first

glance,

it

might appear

hatthe

proposedconstraints

would pre-

dict

additional

clash contexts, such

as

peninitialplus postpeninitial

stress

or

preantepenultimate

lus

antepenultimate

tress, given

the existence of






A

FACTORIAL

YPOLOGY F


TRESS

533

*LAPSE LEFT and *EXTENDED

LAPSE

RIGHT constraints.

The

rankings,

however, ail to

generate

hese

hypothetical

clash

types,

as,

for

every

con-

straint, here is at least one other candidate ackingthese types of clashes

that commits fewer violations.

Candidates

with

peninitial

stress

and

right-

ward terationof

stress

starting

with the

postpenitial

yllable,

i.e.

'6du6cr6,

requirethe

ranking

of ALIGN

(X1, L)

above

ALIGN

(X1, R).

Under

this

ranking,

however,

here s

another

candidate,

6ac6uFa6,

which

better

satis-

fies

ALIGN

(X1, L)

and

fares

no

worse

with

respect

o

constraints

other

han

lowly rankedALIGN

(X1, R).

Candidateswith

antepenultimate

tress

and

leftward

terationof

stress

starting

with the

preantepenult,.e.,

oduuauu,

requirethe rankingof

ALIGN

(X1, R) above

ALIGN

(X1, L). Under this

ranking,

though, there is another

candidate,

a6csc6cca6,

hich

better

satis-

fies ALIGN

(X1, R)

and fares no

worse with

respect

to constraints

other

than

lowly rankedALIGN

(X1, L).

3.4.3.

Binary

Plus

Lapse

SystemsGenerated

by

the Factorial

Yypology

Turning

to the

binary plus

lapse

systems,

of

the

13

generated

patterns,

four are

attested.

Pattern

11, odd-numbered rom

left to

right

plus

penult,

is found in Piro (Matteson1965). Pattern13, odd-numbered rom left to

right

except

final

syllables,

is

widely

attested,as, for

example,

in

Pintupi

(Hansen and

Hansen

1969,

1978). It is

worth

noting that the

proposed

constraints ail

to

generate the mirror

mage

counterpart

f

Pintupi,

odd-

numbered

from right

to left

except the

initial

syllable.

This

patternis

not

generated

because there is

no

left-edge

counterpart

o

NONFINALITY

which

bans

stress on

the

initial

syllable.

The

failureto

generate he

mirror

image

counterpart

o the

Pintupi

pattern

s a

virtue of

the

constraint et,

as

it is unattestedo the best of my knowledge.

Pattern17,

even-numbered

rom right

to

left plus

initial, occurs

in

languages

with

the

'initial-

dactyl'

effect

(see

Prince

1983;

Hayes

1995

for

discussion),

e.g.,

Indonesian,

Spanish.

This

pattern

differs

minimally

from

pattern

19,

which

is

also

attested

(e.g.,

in

Garawa), n

preserving

the

stress

on the

penult rather

than the

initial

syllable in

clash

contexts

in

three-syllable

words.

Most of the

unattested

patterns

closely

resemble

attested

patterns,

differing

minimally

along

a small

set of

dimensions:

he

directionality f stressassignment, heanchoring ite forthestressfixedat

the

opposite

edge from

the

originationpoint

of

the

alternating

tress

count,

and

the

resolution

of

stress

clashes.

It

is

interesting to

note

that some

binary plus

lapse

systems

display

stress

clashes in

relatively

short

words.For

example,

stress

patterns16

and

22

have

stress

clashes in

disyllabic

words

and

patterns

15 and 18

display

clashes

in

trisyllabic

words.The

clash

in

disyllabic

words in

patterns16






534

MATTHEWORDON

and22 arisewhen

ALIGN EDGES is ranked above *CLASH

which in

turn

is rankedabove *LAPSE.

This

rankinggives

rise to

stress clashes but

only

in

disyllabic words where ALIGN EDGES

can otherwise not be

satisfied.

A similar situation can arise

in

trisyllabic

words in a

language (patterns

15

and 18)

in which *LAPSE

RIGHT

andNONFINALITY are

undominated

and crucially

rankedabove

ALIGN

EDGES

which in

turn

s

ranked

above

*CLASH. Under

this

scenario,

the initial

syllable

and the

penult

will

al-

ways be stressed

even

if

this

entails a stress

clash,

as

arises

in

trisyllabic

words. This subset of

binary plus lapse systems

involving

stress clashes

in

either disyllabic or

trisyllabic

words

might

be

regarded

as

pathologic,

sincethey represent n unattestedypeof hybridsystemwitheitherclashes

and lapses depending

on the

length

of

the word. Similar

ternarysystems

are also

generated

by

the

proposed

constraints

see

section

4.1).

There

is,

however, at least

a

possibility

that

this

gap

is an

accidental one

related

to the

independentrarity

of


tress

systems

which

are

tolerant of

clashes,

following

discussion

in

section 2.2.2.

I leave this an

open matter

or futureresearch.

In all of the

binaryplus lapse

systems featuring

an

internal

apse,

the

fixed stress situated at the oppositeedge of originationpoint of the bin-

ary

count

falls on one

of

four

syllables:

the initial

syllable,

the

peninitial

syllable,

the final

syllable,

and

the

penult.

Of these

four

docking

sites

gen-

erated

in

binary plus lapse

systems,

two are

attested:the initial

syllable

(patterns17, 19)

and the

penultimate

syllable (1 1).

Whether he

absence

of

binary plus lapse

systems

with

a

fixed stress on the final

or

peninitial

syllable

reflects an accidental

gap

or

not is

unclear,

hough

the

accidental

gap interpretation

eems reasonable

given

the small

number

of

binary

plus

lapse systems observedcross-linguistically.

If the

atomistic alternative

o ALIGN

EDGES

were

adopted,

an

addi-

tional three unattested

binaryplus

lapse patterns

would

be

generated.

One

of

these

patterns

s

nearly

identical to

pattern 14, differing only in

the

preservation

f the

stress

on

the final

syllable

rather han

the

initialsyllable

in

disyllables:

[01], [101], [1001],

[10101], [101001], [1010101].

This

pattern

results from

a

combinationof an

undominated

constraint

requir-

ing

stress at the

right

edge

in

conjunction

with an

undominated CLASH.

The ranking

ALIGN

(X1, L) >>

ALIGN

(X1, R) produces a stress lapse

immediatelypreceding

the final

stress

in

even

parity words. This pattern

is

not

generatedby

the

unitary

ALIGN

EDGES

constraint,

which

neces-

sarily

conflicts

with

*CLASH

in

disyllables. If ALIGN EDGES

outranks

*CLASH,

both

syllables

are stressed

in

disyllabic

words, yielding

pattern

16 in

Table VII.

If

*CLASH

outranks

ALIGN

EDGES (assuming

ALIGN

(X1,

L)

?>

ALIGN

(X1, R),

pattern

14

results.

Two

additional

unattested






A

FACTORIALYPOLOGY

F


TRESS

535

binaryplus lapse

systems

generated

by

separating

STRESS

EDGES

nto

in-

dividualconstraints

eferring

o different

edges

display

stress

on the

initial

andpenultimatesyllablewith a binary patternoriginating rom eitherthe

initial or

penultimate

yllable,

depending

on the

relative

ranking

of

ALIGN

(XI,

R) and ALIGN

Xl,

L).

These

patterns

differ from

patterns

12

and

19

(attested in

Garawa), generated

by

STRESS

EDGES,

in

having initial

and

final

stress

in

trisyllabic

words rather

han

only

initial

stress,

i.e.

[101]

rather han

[100]. The two novel

systems involve

an

undominated CLASH

and

*LAPSE

RIGHT,

in

addition to an

undominated

decomposed version

of ALIGN

EDGES

requiring

stress

on the

leftmost

syllable.

NONFINALITY

is also highlyrankedbut is violated in trisyllabicwords in order o satisfy

both

*CLASH and

*LAPSERIGHT.

3.4.4.

Primary

Stress

Placement in

Binary

Systems

Before

concludingdiscussion

of

binary

stress, there

are a

couple

of

in-

terestingpoints

about

primary

stress

which are

apparent n

Table VII.

First,

parallel to

dual

stress

systems,

in

virtually all

binary

stress lan-

guages,primary tressfalls a

consistent

distance

romthe

wordedge

across

wordsdifferingin numberof syllables.41A second observationgermane

to

primary

stress is that

there exist

several

patterns n

which the

loca-

tion of

primary

stress in

words

with

heavy

syllables is

inconsistent with

the location of

primary

stress

in

words

containingonly

light

syllables.

It

may

be noted

that

many subtypes of

purely

quantity-sensitive

tress

are

inherently

ncapableof

positioning

primary tress

a

uniform

distance

from

a

word

edge

while

simultaneously

stressing

heavy

syllables

regardless

of

their

position.

Thus, the

quantity-sensitive

analog

to the

Araucanian

patternattested in Creek (Haas 1977; Tyhurst1987), but with the final

stress

as

primary,

positions

its leftmost

stress

on the

initial

syllable if it

is

heavy,

otherwiseon

the

peninitial

syllable.

Thus,

promoting

the

leftmost

stress

to

primary stress

would

not ensure

consistency in the

location of

primary

stress across

different words,

since

words

with an initial

heavy

syllable

would have

primary

tress on

the initial

syllable

while

those with

an

initial

light

would place

primarystress

on the

second

syllable:

io...

41

Malakmalak

Birk

1976)

appears to

be

an

atypical


anguage

in

which

primary

stress

does

not

fall a

consistent

distance

from a

word

edge

across

words

differing n

length.

Stress is

reported

o fall on

even

numbered

yllables

counting

from the

right,

with the

leftmost

stress

being

the

main

one, e.g.,

[102020],

[0102020].

Given

the

preference

for

uniformity

n

stress

placement,one

would

expect the

rightmost

stress,

the

one

falling

closest

to

the

edge at

which

the

binary

pattern

originates

(i.e.,

the one

on the

penult)

to be

the

primary

stress.

Malakmalak

appears o be

the

only


stress

system

which

violates

the

preference

for

positioning

primary

stress

in

the

same

positionacross

the

entire

vocabulary

of

words with

phonologically

predictable

tress.






536

MATTHEW

GORDON

but .6...

A

similar

situationobtains in

quantity-sensitive

ystems

which

stress

heavy syllables

and even-numbered

yllables

going

from

right-to-

left.

It is only

in

quantity-sensitive

ystems

which stress

odd-numbered

syllables

from

either

left-to-right

or

right-to-left, .e.,

'peak

first'

systems

in

Prince's

(1983) grid-

based

framework,

r those

with an

additional ixed

stress at

a

word

edge,

that it is

possible

to

position

primary

stress a

uni-

form

distancefrom

a

word

edge

across

the

entire

vocabulary.

nterestingly,

peak-first

quantity-sensitive ystems

which do not

position

primary

tress

a uniform distance from a word

edge appear

to be

rare; possible

cases

include

Egyptian

Radio Arabic

(Harrell

1960)

and

PalestinianArabic

(see

Hayes 1995andreferencescitedtherein).Giventhepaucityof suchcases it

is difficult o

state at

the

present

whether


nd

quantity-

sensitive

systems truly

differ

in their

strength

of

preference

for

placing

primarystress a

consistent

distance from the word

edge

across

words of

varying

lengths.Hopefully,furtherresearch

will

shed

more

light

on

this

issue.

4. TERNARY STRESS

There

are a

very

small number of

languages

with

ternarystress

patterns,

in

which stress

falls

on

every

third

syllable.

Probably

he

clearest

example

of a


ernary

ystem

comes

from

Cayuvava

Key

1961,

1967),

in which

primary

tress

falls on

the

antepenult

and

secondary

stress

falls

on

every third

syllable

before the

antepenult.Examplesof

Cayuvava

stress

appear

n

(33). (Note

that

sequences

of

vowels, including identical

vowels, areheterosyllabic.)

(33) Cayuvava tress

a.

e.pje

'tail'

b.

fi.ka.he

'stomach'

c.

ki.hi.be.re 'I ran'

d.

a.ri.ui.u.tfa

'he came

already'

e.

d3L.hi.ra.ri.a.ma 'I must

do'

f.

ma.ra.ha.ha.e.i.ki 'their

blankets'

g.

i.ki.ta.pa.re.re.pe.ha

'the

water s clean'

h.

tiA.a.di.r6.bo.3u.ru'.tfe

'ninety-nine

firstdigit)'

i.

me.d'a.ru.tfe.tfe.i.ro.hi.i.jie

fifteen

each (second

digit)'

The

analysis of

the Cayuvava

pattern, and

more generally,

the analysis

of

ternaritypursued

here

follows

the basic

insight of

Elenbaasand Kager






A FACTORIAL

YPOLOGY

F


TRESS

537

(1999), who

attribute

ernary

intervals to the

ranking

of an

anti-lapse

constraint

requiring

that

every

stressless

syllable

be

adjacent

to

either

a

stressed

syllable or

a

word

edge

above

a

foot-based

alignmentcon-

straint,

which in

turn

s

rankedabove a constraint

requiring

hat

syllables

be

parsed into

feet.

In

the

present

work,

ternary

stress

results

from

a

lowly

ranked

*LAPSE

but an

undominated*EXTENDEDLAPSE. The

low

rankingof *LAPSE

ensures that

stress

lapses

entailing

two

consecutive

stressless

syllables

are

freely

tolerated, yet

the

undominated

status of

*EXTENDED

LAPSE

guarantees hat

stress

lapses do not exceed

two

syl-

lables.

*EXTENDED

LAPSE is

ranked

above all the

ALIGN

constraints

referring o level 1 gridmarks,as anypressure o push stressestowardan

edge

is

superseded

by

the

prohibition

against stress

lapses

of

greater

han

two

syllables. Variation n

the

ranking

of

the ALIGN

constraints

accounts

for

different

sub-types

of

ternary

stress.

In

Cayuvava,

n

which

the tern-

ary pattern

nitiates with

the

antepenult,

we

have

the

ranking

ALIGN

(X1,

L)

>>

ALIGN

(X1,

R)

>>

ALIGN

EDGES.

The

rankingof

ALIGN

(X1, R)

aboveALIGN

EDGES

is

decisive in

words in

which the

initial

syllable

goes

unstressed;

hus,

ariuutfa,

rather

than

*ariuiutfa.

The

ranking

of

ALIGN

(X1,L) above

ALIGN

(X1,R) plays a crucialrole in determininghe loca-

tion

of the

stress in

candidates

ncurring

qual

violations of

*LAPSE;

thus,

kihi'bere, ather

than

*kihibere.The

promotion

of

the

rightmost

stress to

primary

stress is

the

result

of ALIGN

(X2, R)

being

rankedabove

ALIGN

(X2, L).

Summary

rankingsfor

Cayuvava

appear n

(34),

followed

by a

representative

ableau n

(35).

(34)

Constraint

ankings

or

Cayuvava

*EXTENDED LAPSE, *CLASH, NONFINALITY, ALIGN (X2, R)

4

ALIGN

(X1,

L)

4

ALIGN

(Xi,

R)

4

ALIGN

EDGES,

*LAPSE,

ALIGN

(X2, L)






538

MATTHE-W

ORDON

(35)

_ _ _ _

_ _

kihibere

ET

NON-:

ALIGN *CLSH ALIGN

ALIGN

ALIGN *LpsE:

ALIGN

Lpsr-

FINAL:

(X2,

R)

(x1,L)

(x1

I,

)

EDGEs:

:

X2,

L)

sw

kihibere

____

*

*

kihibe're

*

*

0WIhbere

__ __

**

kihibere

**

*

kThiber6

.

keihi-be

.....N....**.**.

ariuutfa

*ETM-ALiGN

*CLSH~

ALIGN

ALIGN ALIGN

*LPSEI

ALIGN

LPSE

FINAL:

(X2

)

(X1,44

(xl,

R)

EIxEs

(x,4

a.ri.d.u.tfa__

___

__

marahahaeiki

*EXT: NON-:

ALIGN

*CLSH

ALIGN

ALIGN.

ALIGN *LPSE

ALIGN

LFSE

:FINAL:

X2,R):

(X1,L)

(x

1,R)

EDGES

(X2,

L4

ow

marAhaha.61iki

___ _ __ _ _

7__

mnakahaha.6.i.ki

4 _________

8

mnarAhaha.,.i.ki

*

7

*

4M,

The

description

of

Joway-Oto

n Whitman

(1947)

also

suggests

a tern

ary stress system following the primarystress, though the author does

not

provide examples

illustrating

he

pattern.

There are

a

few

quantity-

sensitive stress

systems

which also

display ternarity

under

certain

weight

and/or

morphological

conditions,

e.g.,

Pacific

Yupik

(Leer

1985;

Rice

1992),

Sentani

Elenbaas

1999;

Elenbaas

and

Kager

1999),42

Finnish

(Sadeniemi

1949;

Hansonand

Kiparsky

1996),

the

variety

of

Estoniande-

scribed

by

Hint

(1973),43

and

Winnebago

(see

Hayes

1995 and references

cited therein or

discussion).

4.

1. A Factorial

Typology f Ternary

Stress

We now turn o the

ternary ystems generatedby

the

factorial

ypology.

Be-

cause of the

rarity

of

ternary

tress,

the

systems generated

by

the

proposed

constraints

will

be

only briefly

summarized

here.

42

Sentani

is

sensitive

to a

three-way

weight

distinction

with

closed

syllables

and

syl-

lables

containingdiphthongs

being

heaviest,

schwa in

open

syllables

being

lightest,

and

non-schwa vowels in open syllables being intermediate n weight

(see Elenbaas

1999

for discussion

and

analysis).

The

Sentani

stress

patternfound

in

words not

containing

heavy

syllables

(i.e.

words

lacking

diphthongs

and

closed

syllables)

is

not

generated

by

the

quantity-insensitiveonstraintsdiscussed

in

this

paper;

rather

t

requiresan

additional

context-sensitive

weight constraint

banning

stress on

light

(i.e.

insufficiently

sonorous)

initial

syllables

due to

tonal

crowding avoidance

(see Gordon

2001,

in

preparation

or

weightconstraints

operativeat

theperiphery).

43

It should also be

noted thatthe

Estonianstress

pattern

described

by

Hint

is the subject

of

controversy;

Eek

(1982) reports

a

different

pattern

not

involvingternarity.






A FACTORIALTYPOLOGY

OF


TRESS

539

A total of 33 ternary ystemsweregenerated,all

of which are

generated

in

pairs differing

in

the

location

of

primary

stress.44

Given the

extreme

rarity of ternarystress

in

general,

it

is

not

surprising

hat

only

a

small

minority

of the

generated

patternsactually

exists.

The

only

clear

quantity-

insensitive ternarypattern,

the

Cayuvava system,

was

generated.

The

Ioway-Otopattern

nferred rom

descriptions

was

also

generated,

as

were

three quantity-sensitive

tress

patterns

observed

in words

consisting

of

only light syllables:

Estonian,

Pacific

Yupik,

and

Winnebago.

This leaves

a total of 28

generated

but

unattested

patterns (31

if

quantity-sensitive

patterns

are

excluded).

Fourpure ternarysystems were generated: ernaryfrom left starting

with the third

syllable, ternary

from left

starting

with

the

peninitial syl-

lable, ternary romright

starting

with the

antepenult,

nd

ternary

rom

right

starting

with the

penult.45

A

total of nine

ternaryplus

clash

systems

were

also generated.These

systems

are the

ternaryanalogs

of the

binary plus

clash

patterns: ernary

stress iterates

n one

direction

and there is a fixed

stress at or near the

opposite edge

of the word. In certainword

shapes,

the

combinationof

ternary

stress

and the

fixed stress

trigger

a

stress

clash.

While I am not aware of any ternaryplus clash systems in actual lan-

guages,

this mismatchbetween

generated

and attested

patterns

s

plausibly

an

accidentalgap attributed o the confluence of two independentlyrare

phenomena: tresssystems with clashes andternary tressin general.

The

final type of

ternary

stress

system generatedby

the factorial

ty-

pology entails

ternarity

mixed

with

binary

stress intervals

n

certainword

shapes.

All of

these

systems

have the

property

of

displaying

more thanone

ternary nterval

n

at least one wordtype of sufficientlengthto diagnose

iterativeternarity.These systems thus differ from the binary plus lapse

systems

which have

a

maximumof

one ternary nterval

n

any one word.

The mixed

ternarysystems are similar to the ternaryplus clash systems

in

displaying ternarity

n conjunctionwith a fixed stress at the opposite

44

The

number

of

generatedpatterns ncreases to

42 if

ALIGN EDGES

is broken into

separate

constraints ach

referring

o

a different

edge.

Of the nine

additionalpatterns,

wo

involve pureternarity

tarting

rom a

peripheral yllable (left-to-right ernarity tartingwith

the initial

syllable andright-to-left ternarity

tarting

with the final syllable) and 7 involve

ternaritymixed withbinary stress intervals n certaincontexts.

45

Pure

ternarypatterns

differing

from

those

generated

n

initiating

a

ternarycount from

an

edge syllable, e.g., a hypotheticalpattern nvolving ternary tress terating rom the right

edge

and

beginning

with

the final

syllable, are not generated, ince, in certainword shapes,

pure ternary

andidateswith

stresses distanced romthe peripheryhave fewer stresses and

thus

better

satisfy

ALIGN

(X1, L)

andALIGN

(X1, R); for example,

aaeacda'

incurs fewer

violations of both

ALIGN (X1, L) and ALIGN (X1, R) thana'cravcr6. Crucially,because

ALIGN

EDGES is

lowly ranked

n

pure ternary ystems,

it

is unable to ensure that an edge

syllable is stressed.






540

MATTHEW

ORDON

edge

of the word

from the

origination

point

of the

ternary

count.

Three

of these

patterns

are the

ternary

analogs

to similar

binary plus

lapse

pat-

terns nvolvingclashesin shortwords(see section3.4.3). Inaddition, hree

left-to-right

ernarypatterns

originate

their

stress counts

closer

to

the left

edge

of

the word

in

relatively

short words

in

order to

avoid

violations

of

undominated

NONFINALITY. Such

patterns

are

analogous

to

patterns

een

in

single stress

languages

(e.g.,

Hopi

=

pattern

3 in

Table

IV)

and

binary

stress

languages

(e.g.,

SouthernPaiute

=

pattern

6

in

Table

VII).

A

total of

20 mixed

ternarysystems

were

generated,

including

the

systems found in Pacific

Yupik

(ternary

rom

left

starting

with

peninitial

syllableplus a binary ntervalat therightedge), Winnebago ternaryrom

left

startingwith

thirdsyllable

plus

a

binary

nterval

at the

right

edge),

and

Estonian

(ternary rom left

plus

penultimate

tress in

non-clash

contexts).

5.

FACTORIAL

TYPOLOGY:

A

SUMMARY

In summary, he factorialtypology generatedby the proposedconstraint

set offers

sufficient

empirical

coverage

of the


ingle

stress,

dual

stress,

binary

and

ternary

tress

systems

found in

an

extensive

survey.While the

thoroughnessof the

empirical

coverage

of

the

proposed

theorycomes at the

price

of

a certain

degree

of

overgeneration, he

amount

of

overgeneration s

relatively

small,

at

least

relative to

the

logically pos-

sible set

of

stress

systems.

Furthermore,

he

overgeneration

ommitted

by

the

present

analysis,

with

some

possible

exceptions

(e.g.,

mixed

clash

and

lapse systems; see discussionin section 3.4.3), is plausibly attributedo

accidental

gaps

resultingfrom

the overall

rarityof

certain

types of

stress

systems.

This

is

particularly pparent

n

the

case of

ternary

ystems

which

are a

vanishingly

rare

breed

cross-linguistically.

Thus,

virtually

any

ternary

stress

system

generated

by

the

theory

is certain

not

to

occur

just

given

the

paucity

of

ternary

stress in

general. Similar

cases of

overgeneration

which

could

plausiblybe

attributed

o

accidental

gapsare

observed n

dual

stress and

binary stress

systems,

where

most of

the

unattestedbut

gener-

atedpatterns nvolvedelementswhichareattested n isolationor in certain

combinations

but

happen

not to exist in

all

of the

combinations

generated

by

the

factorial

typology.

Although

the

present

work

can

neitherbe

re-

gardedas a

complete

metrical

stress

theory,as

such

a

theory

would

require

constraints

designed to

also

handle

quantity-sensitive

tress

(see

Gordon

to

appear,

n

preparation

or

analysis

of

quantity-sensitive

tress), it

never-

theless

suggests that

Optimality

Theory

offers a

promising

framework

or

expression

of

grid-based

metrical

heories.






A

FACTORIALYPOLOGY

F


TRESS

541

APPENDIX:

LIST OF LANGUAGES

WITH


STRESS

Stress

pattern

Language

Family46

Initial

Afrikaans

Indo-European

Initial

Arabana-Wangkanguru

ustralian

Initial

Arabela

Zaparoan

Initial

Arawak

Arawakan

Initial

Chepang

Sino-Tibetan

Initial

Chitimacha

Gulf

Initial

Chutiya

Sino-Tibetan

Initial Comox Salish

Initial

Danish

Indo-European

Initial

Diola

Niger-Congo

Initial

Dizi

Afro-Asiatic

Initial

DjapuYolngu Australian

Initial

Enets

Uralic

Initial

Even

(Lamut)

Altaic

Initial

Gilyak

Isolate

Initial

Gondi

Dravidian

Initial GosiuteShoshone Uto-Aztecan

Initial

Gurung

Sino-Tibetan

Initial

Hanty

Uralic

Initial

Hewa

Sepik-Ramu

Initial

Hualapai

Hokan

Initial

Huitoto

Witotoan

Initial

Irish

Indo-European

Initial

Jemez

Kiowa-Tanoan

Initial

Kalkatungu

Australian

Initial

Kambera

Austronesian

Initial

Kanauri

Sino-Tibetan

Initial

Ket

Yenesei

Ostyak

Initial

Kolami

Dravidian

Initial

Korafe

Tran-New

Guinea

Initial

Kota

Dravidian

Initial

Latvian

Indo-European

Initial

Naga

Creole

Initial

Nenets

Uralic

Initial

Olo

Austronesian

Initial(of root) Papago Uto-Aztecan

Initial

ParintintinTenharim

Tupi

Initial

Pomo,Eastern

Hokan

Initial

Sami,

Eastern

Uralic

Initial

Sango

Creole

Initial

Santali

Austro-Asiatic

46

Genetic

affiliations are

according

to

the 14th

edition

(2001) of

the

SIL Ethnologue

(CD

version)

edited

by

Barbara

F.

Grimes.






542

MATTHEWORDON

Stress pattern

Language

Family

Initial

Senoufo

Niger-Congo

Initial Siona Tucanoan

Initial

Sumbanese Austronesian

Initial

Swedish

Indo-European

Initial

Tigak

Austronesian

Initial

Tinrin

Austronesian

Initial

Tunica Gulf

Initial

Wembawemba

Australian

Initial Yurok

Algic

Initial (mora)

Nama

Khoisan

Initial (nouns), final of root

(verbs),

Dumi Sino-Tibetan

Initial

(nouns),

verbs-??

Pawnee Caddoan

Initial (root)

Coreguaje

Tucanoan

Initial

(root)

Kung(Zu1'Haasi)

Khoisan

Initial (root)

Mixe,

Totontepec

Mixe-Zoque

Initial

(root)

Tewa

Kiowa-Tanoan

Peninitial Assiniboine Siouan

Peninitial

Basque

Isolate

Peninitial

Ignaciano Arawakan

Peninitial

Koryak

Chukotko-Kamchatkan

Peninitial Lakota Siouan

Peninitial

Lezgian North

Caucasian

Peninitial

Paraujano

Arawakan

Peninitial Zan South

Caucasian

Peninitial

of

root)

Siroi

Trans-New

Guinea

Peninitial Tolai

Austronesian

Antepenult

Cayubaba

Isolate

Antepenult Cora

Uto-Aztecan

Antepenult Kela

Austronesian

Antepenult Macedonian Indo-European

Antepenult Mae

Austronesian

Antepenult Parnkalla

Australian

Antepenult

Wappo Yuki

Penult

Alawa

Australian

Penult

Albanian

Indo-European

Penult

Amara

Austronesian

Penult

Andamanese

Andamanese

Penult

Anem

East Papuan

Penult Arapesh Torricelli

Penult

Atchin

Austronesian

Penult

Bukiyip

Torricelli

Penult

Chamorro

Austronesian

Penult

Chickaranga

Niger-Congo

Penult

Chumash,

Hokan

Barbareno

Penult

Cocama

Tupi

Penult

Cofan

Chibchan






A FACTORIAL

TYPOLOGYOF


543

Stress

pattern

Language

Family

Penult

Dayak

Austronesian

Penult Djingili Australian

Penult

Jaqaru

Aymara

Penult

Kaliai-Kove Austronesian

Penult

Kola

Austronesian

Penult

Kutenai Isolate

Penult

Kwaio

Austronesian

Penult

Labu

Austronesian

Penult

Lamba

Niger-Congo

Penult

Laz

South Caucasian

Penult

Lingala

Niger-Congo

Penult

Lusi

Austronesian

Penult

Mohawk

Iroquoian

Penult

Monumba

Niger-Congo

Penult

Movima

Unclassified

Penult

Mussau

Austronesian

Penult

Nahuatl

Uto-Aztecan

Penult

Onondaga

Iroquoian

Penult

Paez

Paezan

Penult

Pagu

West

Papuan

Penult(of phrase) Paiwan Austronesian

Penult

Quicha

Quechan

Penult

Quileute

Chimakuan

Penult

Rapanui

Austronesian

Penult

Shona

Niger-Congo

Penult

Siriono

Tupi

Penult

Solor

Austronesian

Penult

Southern

Sotho

Niger-Congo

Penult

Tacana

Tacanan

Penult Temate WestPapuan

Penult

Tetun

Austronesian

Penult

Tiwi

Australian

Penult

Tolo

Austronesian

Penult

Tolojolabal

Mayan

Penult

Tonkawa

Coahuiltecan

Penult

Tswana

Niger-Congo

Penult

Tuscarora

Iroquoian

Penult

Usan

TransNew

Guinea

Penult Wardaman Australian

Penult

Wikchamni

Penutian

Penult

(nouns), final

(verbs)

Gurage

Afro-Asiatic

Final

Abun

WestPapuan

Final

Alabama

Muskogean

Final

Apinaye

Macro-Ge

Final

Aramaic

Afro-Asiatic

Final

Atayal

Austronesian

Final

Azerbaijani

Altaic






544

MATTHEW

ORDON

Stress

pattern

Language

Family

Final Bashkir Altaic

Final Cakchiquel Mayan

Final

Canela-Krah6

Macro-Ge

Final

Dani TransNew

Guinea

Final

Gagauz

Altaic

Final

Greenlandic

nuktitut

Eskimro-Aleut

Final

Guarani

Tupi

Final Haitian

Creole

Creole

Final Hebrew

Afro-Asiatic

Final

Than

Austronesian

Final

Ivatan

Austronesian

Final

(of root)

Kabardian North

Caucasian

Final Kavalan Austronesian

Final

Kayabi

Tupi

Final

Kayap6

Macro-Ge

Final Kazakh Altaic

Final

Konkani

Indo-European

Final

Kurdish

Indo-European

Final

(of

root)

Lango

Nilo-Saharan

Final

Maricopa

Hokan

Final Mazatec Oto-Manguean

Final

Moghol

Altaic

Final

Nanai

Altaic

Final

Pemon

Carib

Final

Persian

Indo-European

Final

Sa'ban

Austronesian

Final

Semai

Austro-Asiatic

Final

Stieng

Austro-Asiatic

Final

Tajiki

Indo-European

Final TamazightBerber Afro-Asiatic

Final

Tatar

Altaic

Final

Temiar

Austro-Asiatic

Final

Thai

Daic

Final

Tolowa

Na

Dene

Final

Tsotsil

Mayan

Final

Turkmen

Altaic

Final

Tzutujil

Penutian

Final

Udmurt

Uralic

Final Uighur Altaic

Final

Uzbek

Altaic

Final

Waiwai Carib

Final

Yagua

Peba-Yaguan

Final

Yakut Altaic

Final

Yuchi

Isolate

Final

(of

phrase)

French,European

Indo-European

Final

(of

phrase)

Wari'

Chapacura-Wanham

Final

Diegueno Hokan






A

FACTORIAL

TYPOLOGY

OF


TRESS

545

Stresspattern

Language

Family

Final (of

root)

Paipai Hokan

Final (of root) Shilha Afro-Asiatic

Final (of

root)

Waskia TransNew

Guinea

Final (of

root)

Yavapai Hokan

Final (of root)

Zapotec,

Mitla

Oto-Manguean

Final Zazaki

Indo-European

Final (lary) & initial

(secondary) Armenian

Indo-European


(secondary)

French,

Canadian

Indo-European


(secondary)

Udihe

Altaic

Antepenult

lary) &

initial

(secondary)

Georgian South Caucasian

Initial

(lary)

&

antepenult

secondary) Walmatjari

Australian

Initial

(lary)

&

penult

(secondary)

Gugu

Yalanji

Australian

Initial

(lary)

&

penult

(secondary)

Lower

Sorbian

Indo-European

Initial

(lary) & penult

(secondary)

Watjarri Australian

Penult(lary) & initial

(secondary)

Anyula

Australian

Penult

(lary) &

initial

(secondary)

Awtuw

Sepik-Ramu

Penult

(lary) & initial

(secondary)

ChimalapaZoque

Mixe-Zoque

Penult(lary) &

initial

(secondary) Murut

Austronesian

Penult

(lary)

&

initial

(secondary) Sanuma

Yanomam

Penult

(lary) & initial

(secondary)

Sibutu Sama

Austronesian

Evennumbered romL Araucanian Araucanian

Even-numbered rom

L

Hatam

WestPapuan

Even

numbered rom

L

Yupik,

Sirenik

Eskimo-Aleut

Even

numbered rom

R

Anejom

Austronesian

Even

numbered rom R

Berbice

Creole

Even

numbered rom

R

Cavinefna

Tacanan

Even

numbered

rom

R

Ese

Ejja

Tacanan

Even-numbered rom

R

(root)

Larike

Austronesian

Even

numbered rom R

Malakmalak

Australian

Evennumbered romR Nengone Austronesian

Even

numbered

rom

R

Orokolo

TransNew Guinea

Even

numbered

rom

R

To'aba'ita

Austronesian

Even

numbered

rom

R

Tukang

Besi

Austronesian

Even-numbered rom R

Ura

Austronesian

Even

numbered rom

R

Warao

Isolate

Odd

numbered

rom R

Asmat

Trans-New

Guinea

Odd-numbered rom

R

Chulupi

Mataco-Guaicuru

Odd

numbered rom

R

Kamayur

Tupi

Odd-numberedrom

R

UrubuKaapor Tupi

Odd-numbered rom R

Weri

Trans-NewGuinea

Odd-numberedrom L

Bagandji

Australian

Odd

numbered rom

L

Burum

Trans-New

Guinea

(with

optional

nonfinality)

Odd

numbered rom

L

Czech

Indo-European

Odd

numbered

rom

L

Hungarian

Uralic

Odd

numbered rom L

Icelandic

Indo-European

Odd

numbered rom L

Livonian

Uralic






546 MATTHEW

ORDON

Stresspattem

Language

Family

Odd

numbered

rom L

Mansi

Uralic

Oddnumbered romL Maranungku Australian

Odd numbered rom L

Murinbata

Australian

Odd numbered

rom

L

Ningil Torricelli

Odd numbered rom L

Ono

Trans-New

Guiena

Odd-numberedrom

L

Panamint

Uto-Aztecan

Odd-numberedrom L

Sami,

Northem

Uralic

(with

optional

nonfinality)

Odd-numbered

rom L

Selepet Trans-New

Guinea

(with

optional

nonfinality)

Odd numbered rom L

Sinaugoro Austronesian

Odd

numbered

rom

L

Timucua

Arawakan

Odd

numbered rom L

Votic

Uralic

Odd numbered rom L

(roots);

even

Waorani

Unclassified

numbered

rom

R

(suffixes)

Odd numbered

rom

L

minus

final

Anguthimri

Australian

Odd-numberedrom L

minus

final

Badimaya

Australian

Odd

numbered rom L

minus final

Bidyara/Gungabula Australian

Odd

numbered

rom

L

minus

final

Dalabon

Australian

Odd

numbered

rom

L

minus

final Dehu

Austronesian

Oddnumbered romL minus final Diyari Australian

Odd numbered rom L

minus final

Karelian

Uralic

Odd-numbered

rom

L

minus

final

Kate

Trans-New

Guinea

Odd-numbered

rom

L

minus

final

Pintupi

Australian

Odd

numbered rom L

minus

final

Pitta

Pitta

Australian

Odd

numbered from L

minus final

(of

TenangoOtomi

Oto-Manguean

root)

Odd

numbered rom

L

minus

final

Wangkumara

Australian

Odd

numbered rom

L

minus

final

Wirangu

Australian

Oddnumbered romL minus final Yingkarta Australian

Evennumbered

rom L &

penult

Southem

Paiute

Uto-Aztecan

Evenlodd

numbered

lexical)

from

Biangai

Trans-New

Guinea

R

& initial

Even-numbered

rom

R

&

initial

Garawa

Australian

Odd

numbered rom L

&

penult

Piro

Arawakan

Odd

numbered rom L

& final

Gosiute

Shoshone

Uto-Aztecan

Odd-numberedrom R

& initial

Tauya

Trans

New

Guinea

Penult

&

odd-numbered rom

R/L

Polish

Polish

Temary romR, Rmost on antepenult Cayuvava Isolate

Temary

rom

L

starting

with

loway-Oto

Siouan

initial/peninitial






A

FACTORIAL

YPOLOGY

F


TRESS

547

REFERENCES

Alber, Birgit.

1997.

'Quantity

Sensitivity

as the Result of

Constraint

nteraction',

n

Geert

Booij and

Jeroen van

de

Weijer (eds.), Phonology

in

Progress-Progress

n

Phonology,

HIL

Phonology Papers III, HollandAcademic

Graphics,

The

Hague, pp.

1-45.

Alderete, John. 1999.

Morphologically

Governed

Accent in

Optimality

Theory, unpub-

lished

Ph.D.

dissertation,

University

of

Massachusetts,

Amherst.

Allison, E.

J.

1979. 'The

Phonology

of Sibutu Sama:

A

Language

of

the

Southern

Philip-

pines', in

C. Edrial-Luzares nd A. Hale

(eds.),

Studies

in

Philippine

Linguistics,

Vol.

3,

No.

2, Linguistic Society

of

the

Philippines

and SummerInstitute

of

Linguistics,pp.

63-104.

Aronson,Howard. 1991. Georgian:A ReadingGrammar,Slavica, Columbus,OH.

Bakovic, Eric. 1996. 'Foot

Harmony

and

Quantitative

Adjustments',manuscript,Rutgers

University.Availableon

RutgersOptimality

Archive,ROA-168,

http://ruccs.rutgers.edu/roa.html.

Bakovic,

Eric. 1998. 'UnboundedStress and Factorial

Typology',

manuscript,Rutgers

University.

Availableon

RutgersOptimality

Archive,ROA-244,


Bendor-Samuel,

J. T.

1963.

'Stress in

Terena',

Transactions

of

the

Philological Society

1962, 105-123.

Birk, D. B. W. 1976. TheMalakmalakLanguage, Daly River (WesternArnhemLand),

Australian

National

University,

Canberra.

Boas, Franz and Ella

Deloria. 1933. 'Notes on the

Dakota, Teton

Dialect',

International

Journal of American

Linguistics

7, 97-121.

Boas,

Franzand Ella Deloria. 1941.

Dakota Grammar

Memoirsof

the

National

Academy

of

Sciences

33], United States Government

PrintingOffice,Washington,

D.C.

Borgman,

Donald. 1989.

'Sanuma',

n

Desmond

Derbyshire

and

Geoffrey

Pullum

(eds.),

Handbook

of Amazonian

Languages,

Vol.

2, Mouton,

New

York,pp.

15-248.

Chafe,

Wallace. 1977. 'Accent and

Related Phenomena n

the Five Nations

Iroquois

Lan-

guages', in Larry Hyman (ed.), Studies in Stress and Accent, Universityof Southern

California,

Department

f

Linguistics,

Los

Angeles, pp. 169-181.

Cohn, Abigail.

1989. 'Stress n

Indonesianand

Bracketing

Paradoxes',

Natural

Language

and

LinguisticTheory7, 167-216.

Crowhurst,

Megan

and Mark

Hewitt.

1994. 'Directional

Footing, Degeneracy,

and

Align-

ment', manuscript,

University of North

Carolina and University of British

Columbia.

Available

on

Rutgers

OptimalityArchive,

ROA-65,


Davies,

H.

J.

1980.

Kobon

Phonology, Australian

National

University,Canberra.

Ddcsy, Gyula.

1966.

Yurak

Chrestomathy,

ndianaUniversityPress,

Bloomington.

Douglas, Wilfrid.

1981.

'Watjarri',

n

Robert

Dixon and BarryBlake

(eds.),

Handbookof

AustralianLanguages,Vol.2, John

Benjamins,Amsterdam,pp. 196-272.

Dubert,Raymondand Marorie

Dubert. 1973.

'Biangai

Phonemes',

Phonologies of Three

Languages

of Papua

New

Guinea, Summer

Institute of

Linguistics,Ukarumpa,Papua

New

Guinea,pp. 5-36.

Echeverria,

Max

and Heles

Contreras. 1965.

'Araucanian

Phonemics', International

Journal

of

American

Linguistics

31,

132-135.

Eek,

Arvo. 1982.

'Stress and

Associated

Phenomena:

A

Survey with

Examples

from Esto-

nian',

Estonian

Papers

in

Phonetics

1980-1981, Academyof Sciences of

the Estonian

SSR,

Instituteof

Language

and

Literatures,

Tallinn,pp. 20-58.






548

MATTHEW ORDON

Egerod, S0ren. 1966.

'A Statement

on

Atayal Phonology',

ArtibusAsiae

Supplementum

XXIII Felicitationvolume or the 75th birthdayof Professor

G. H.

Luce),

Artibus

Asiae,

Ascona, Switzerland,pp.

120-130.

Eisner, Jason. 1997. 'FootFormDecomposed: Using Primitive Constraints n OT', MIT

Working Papers

in

Linguistics.

Available

on

Rutgers Optimality Archive, ROA-205,


Elenbaas,

Nine.

1999.

A

Unified

Account

of Binary

and

TernaryStress,

Netherlands

Graduate

School

of

Linguistics,

Utrecht.

Elenbaas,

Nine

and Rene Kager.

1999.

'Temary Rhythm

and the

Lapse Constraint',

Phonology 16, 273-329.

Everett,

Daniel.

1986.

'Piraha',

in

Desmond

Derbyshire

and

Geoffrey

Pullum

(eds.),

Handbookof AmazonianLanguages,

Vol.

1, Mouton,

New

York,pp.

200-325.

Everett,Daniel. 1988. 'On Metrical ConstituentStructure n Piraha',Natural Language

and

LinguisticTheory6,

207-246.

Everett,

Daniel and

Keren

Everett. 1984.

'Syllable

Onsets and Stress Placement

n

Piraha',

Proceedings of the

WestCoast

Conference

on

Formal

Linguistics3,

105-116.

Feldman,Harry.1986.

A

Grammarof Awtuw,

AustralianNational

University,

Canberra.

Furby, Christine. 1974.

Garawa

Phonology, Canberra,

Australian

National

University,

Canberra.

Gendron,

Jean

Denis.

1966.

Tendances

Phonetiques

Francais

Parle

au

Canada,

Les

Presses de L'universiteLaval, Qudbec.

Gordon,

Matthew.2001. 'The Tonal Basis

of Final

Weight Criteria',Chicago Linguistics

Society 36,

141-156.

Gordon,

Matthew. To

appear. 'Syllable Weight',

in

Bruce

Hayes,

Robert

Kirchner,

and Donca Steriade

(eds.),

Phonetic Bases

for Phonological Markedness,Cambridge

University

Press.

Gordon,

Matthew. n

preparation.

A Grid-based

Optimality-TheoreticConstraintSet for

Stress', manuscript,University

of

California,

SantaBarbara.

Green, Thomas

and

Michael

Kenstowicz. 1996.

'The

Lapse Constraint',Proceedings of

the6th AnnualMeeting of the FormalLinguisticsSociety of theMidwest,pp. 1-15.

Haas, Mary. 1977.

'TonalAccent

in

Creek',

in

LarryHyman (ed.), Studies

in

Stress and

Accent[SouthernCaliforniaOccasional PapersinLinguistics4], Universityof Southern

California,Department

of

Linguistics,

Los

Angeles, pp.

195-208.

Halle,

Morrisand

Jean-RogerVergnaud.1987.

An

Essay

on

Stress,

MIT

Press, Cambridge,

MA.

Hammond,

Michael. 1985.

'Main

Stress and Parallel Metrical

Planes', Berkeley Lin-

guistics Society

11:

Parasession

on

Poetics, Metrics,

and

Prosody,Berkeley Linguistics

Society, Berkeley,pp.

417-428.

Hansen,

Kenneth and L. E.

Hansen. 1969. 'Pintupi Phonology', Oceanic Linguistics 8,

153-170.

Hansen, Kennethand L. E. Hansen. 1978. The Core of Pintupi Grammar,Institute for

AboriginalDevelopment,Alice Springs,Australia.

Hanson,

Kristin and Paul

Kiparsky. 1996.

'A

Parametric Theory of Poetic Meter',

Language 72, 287-335.

Harden,Margaret.1946. 'Syllable Structure f Terena', nternational ournalof American

Linguistics12,

60-63.

Hardman-de-Bautista,MarthaJames. 1966. Jaqaru:Outlineof Phonological and Morpho-

logical Structure,Mouton,

The

Hague.






A

FACTORIALYPOLOGY F


TRESS

549

Harms,

Robert. 1966.

'Stress,Voice,

and

Length

n

Southern

Paiute',

International

ournal

of

AmericanLinguistics

32, 228-235.

Harrell,Richard. 1960. 'A

Linguistic

Analysis

of

Egyptian

Radio

Arabic',

in

CharlesFer-

guson (ed.), Contributions o ArabicLinguistics,HarvardUniversityPress,Cambridge,

MA, pp.

3-77.

Harris, James. 1983.

Syllable

Structure

and

Stress in

Spanish:

A Nonlinear

Analysis

[LinguisticInquiryMonograph8], MIr

Press,

Cambridge,

MA.

Hayes,

Bruce. 1980.

A

Metrical

Theoryof

Stress

Rules,

Ph.D.

dissertation,

MIT.

Published

in

1985 by Garland

Press,

New York.

Hayes, Bruce. 1985. 'Iambic and Trochaic

Rhythm

n

Stress

Rules', Berkeley

Linguistics

Society

11:

Parasession

on

Poetics,Metrics,

and

Prosody,

Berkeley LinguisticsSociety,

Berkeley,pp. 429-446.

Hayes, Bruce. 1995. Metrical Stress Theory:Principles and Case Studies,Universityof

Chicago Press,Chicago.

Hayes, Bruce and Stanislaw

Puppel. 1985.

'On

the

Rhythm

Rule

in

Polish',

in

Harry

van

der Hulst and Norval Smith

(eds.),

Advances

n

Nonlinear

Phonology, Foris,

Dordrecht,

pp. 59-8 1.

Hayes,

Bruce, Bruce

Tesar,

and Kie

Zuraw.

2000.

OTSoft,

oftware

package.

Available at

http://www.humnet.ucla.edu/humnet/linguistics/people/hayes/otsoft/index.htm.

Hetzer,

Armin. 1978. Lehrbuchder

vereinheitlichten lbanischen

Schriftsprache

mit einem

deutsch-albanischen

Worterbuch,Helmut Buske

Verlag,Hamburg.

Hint,Mati. 1973. Eesti KeeleSonafonoloogia I, EestiNSV TeadusteAkadeemia,Tallinn.

Hudson,

Joyce.

1978. The

Coreof

Walmatjari

Grammar,

Australian nstituteof

Aboriginal

Studies,

Canberra.

Hyman,Larry.

1977. 'On the

Nature of

LinguisticStress',

in

Larry

Hyman (ed.), Studies

in Stress and

Accent,University of Southern

California,pp. 37-82.

Janas,

Petr. 1984.

NiedersorbischeGrammatikflrden

Schul-gebrauch,

Domowina-Verlag,

Bautzen.

Jeanne, Laverne. 1982. 'Some

Phonological

Rules of

Hopi',

International Joumal of

AmericanLinguistics

48, 245-270.

Kager,Rene. 1989.

A

Metrical

Theory of Stress and Destressing

in English and

Dutch,

Foris,Dordrecht.

Kager,

Rend.

1993. 'Alternatives o the

IAMB-TROCHEEaw',

Natural

Language

and

Linguistic

Theory11, 381-432.

Kager,

Rend.

1994.

'Ternary

Rhythm

in

Alignment

Theory',

manuscript,

Utrecht

University. Available on

Rutgers

Optimality Archive,

ROA-35-1094;


Kager, Rend. 1997.

'Generalized

Alignment and

MorphologicalParsing',

Rivista di

Linguistica

9,

245-282.

Kager,Rend. 1999.

OptimalityTheory,

Cambridge

UniversityPress,Cambridge.

Kakumasu,James. 1986. 'Urubu-Kaapor',n Desmond Derbyshireand GeoffreyPullum

(eds.),

Handbook

of

Amazonian

Languages,

Vol.

1, Mouton,New

York,pp.

326-406.

Kenstowicz,

Michael. 1994.

'Evidence

for Metrical

Constituency',in K. Hale

and S. J.

Keyser

(eds.), The View rom

Building 20: Essays

in Linguisticsin Honor of

Sylvain

Bromberger,

MIT

Press, Cambridge,MA, pp.

257-273.

Kenstowicz,Michael.

1997.

'Quality-SensitiveStress', Rivistadi

Linguistica9,

157-188.

Key,

Harold.

1961. 'Phonotactics of

Cayuvava', International

Journal of

American

Linguistics27, 143-150.

Key,

Harold.

1967.

Morphologyof

Cayuvava,Mouton,The Hague.






550

MATTHEW

GORDON

Key,

Mary. 1968.

Comparative

Tacanan

Phonologywith CavinefiaPhonology

and

Notes

on Pano-Tacanan

Relationship,

Mouton,

The

Hague.

Kirton,

Jean.

1967. 'Anyula

Phonology',

in D. Glasgow,K. Glasgow,

J. Kirtonand W.

J.

Oates (eds.), Papersin AustralianLinguistics,AustralianNationalUniversity,Canberra,

pp.

15-28.

Knudson,Lyle.

1975.

'A NaturalPhonology

and Morphophonemics

f Chimalapa

Zoque',

Papersin

Linguistics

8, 283-346.

Kormushin, gor

Valentinovich.

1998.

Udykheiskii

Udegeiskii)Iazyk,

Nauka,

Moscow.

Kucera,

Henry.

1961.

The Phonology of

Czech,Mouton,

's-Gravenhage.

Leer,

Jeff. 1985.

'Prosody

n

Alutiiq',

in Michael

Krauss

ed.),

Yupik

EskimoProsodic

Sys-

tems: Descriptive

and

Comparative

Studies,

Alaska Native

Language

Center,

Fairbanks,

pp.

77-134.

Leskinen,Heikki.1984. 'Uberdie Phonemsystemder KarelischenSprache', n PeterHajdu'

and Laszlo

Honti

(eds.),

Studien

zurPhonologischen

Beschreibung

uralischer

Sprachen,

Akademiai

Kiad6,Budapest,pp.

247-257.

Leslau, Wolf.

1992. 'Outline

of

Gurage

Phonology',

in Wolf Leslau

(ed.),

Gurage

Studies:

CollectedArticles,

Otto Harrassowitz,

Wiesbaden,pp.

1-116.

Liberman,Mark

and Alan

Prince.

1977.

'On

Stress and Linguistic

Rhythm',

Linguistic

Inquiry8,

249-336.

Lunt,Horace.

1952.

A Grammar f

the Macedonian

LiteraryLanguage,

n.p.,

Skopje.

Lynch,

John. 1974.

Lenakel

Phonology,

unpublished

Ph.D.

dissertation,

University

of

Hawaii.

Lynch,

John. 1978.

A

Grammar f

Lenakel,

Australian

National

University,

Canberra.

MacDonald,

Loma.

1990.

A

Grammar f

Tauya,

Mouton

de

Gruyter,

New York.

Matteson,

Esther.

1965.

The

Piro

(Arawakan)

Language,

Universityof

CaliforniaPress,

Berkeley.

McCarthy,

John and

Alan Prince. 1993.

'GeneralizedAlignment',

manuscript,

University

of Massachusetts

and

Rutgers

University.

Available on

RutgersOptimality

Archive,

ROA-7, http://ruccs.rutgers.edu/roa.html.

Menovshchikov,

G.

A.

1975. Iazyk

NaukanskikhEskimosov,Nauka,

Leningrad.

Miller,

Wick. 1996. 'Sketchof

Shoshone,

a Uto-Aztecan

Language',

n Ives Goddard ed.),

Handbookof AmericanIndianLanguages, Vol. 17 Languages,SmithsonianInstitute,

Washington,

D.C., pp.

693-720.

Myers,

James. 1999.

'Lexical

Phonology

and the Lexicon',

manuscript,

Graduate nsti-

tute of

Linguistics,

National

Chung Cheng

University,

Taiwan. Available on

Rutgers

Optimality

Archive,

ROA-330,http://ruccs.rutgers.edu/roa.html.

Nikolaeva,

Irna

and Maria

Tolskaya.

2001.

A

Grammar

of

Udihe,Mouton,New

York.

Oates,

Williamand

Lynette

Oates. 1964. Gugu-Yalanji

nd

Wik-Munkan

anguage

Studies,

Australian nstituteof

Aboriginal

Studies,Canberra.

Osborn,Henry.

1966. 'Warao

:Phonologyand

Morphophonemics',

nternational

ournal

ofAmericanLinguistics32, 108-123.

Pike,

Kenneth. 1964.

'Stress

Trains

n

Auca',

in

David

Abercrombie,

Dennis Fry,

P. A. D.

MacCarthy,

N.

C. Scott, and

J. L. M.

Trim

(eds.),

In Honour of Daniel

Jones,

Longmans,

London, pp.

425-431.

Pike,

Kenneth and Eunice

Pike. 1947.

'Immediate

Constituentsof

Mazateco

Syllables',

IntemationalJournal

of American

Linguistics13,

78-91.

Polotsky,Hans Jakob.

1951. Notes

on Gurage Grammar,

Israel

OrientalSociety,

Jerus-

alem.






A FACTORIAL

YPOLOGY

F


551

Prentice,

D.

J.

1971. The

Murut

Languages

of Sabah,

Australian National

University

Canberra.

Prince,

Alan. 1980.

'A

Metrical

Theory

for

Estonian

Quantity',

Linguistic

Inquiry11,

51 1-

562.

Prince,

Alan. 1983.

'Relating

o the

Grid',Linguistic

Inquiry

14, 19-100.

Prince, Alan

and Paul

Smolensky. 1993.

'Optimality Theory:

ConstraintInteraction n

Generative

Grammar',

manuscript,Rutgers

University

and

University

of Colorado

at

Boulder.

Radin,

Paul. 1929.

A

Grammar

of

the

WappoLanguage,

University

of

California

Press,

Berkeley.

Rice, Curtis. 1992.

Binarity

and

Ternarity

n Metrical

Theory:

Parametric

Extensions,

unpublished

Ph.D.

dissertation,University

of

Texas,

Austin.

Rich, Fume. 1963. 'ArabelaPhonemes and High-level Phonology', in BenjaminElson

(ed.),

Studies in

Peruvian Indian

Languages,

Vol.

I,

Summer Institute

of

Linguistics,

Norman,Oklahoma,

pp. 193-206.

Sadeniemi,Matti.

1949.

Metriikkamme

erusteet,

SKST,

Helsinki.

Saelzer,

Meinke.

1976. 'Fonologia

Provis6riada

Lingua

Kamayura',

Serie Lingidstica

5,

131-170.

Sapir,

Edward.

1930. SouthernPaiute: A

Shoshonean

Language,

American

Academy

of

Arts

and

Sciences,

Cambridge,

MA.

Selkirk, Elisabeth.

1984.

Phonology and

Syntax:

The Relation

between Sound and

Structure,MITPress,Cambridge,MA.

Skorik,

Petr.

1961. Grammatika

Chukotskogo

azyka, Izdatelstvo Akademii

Nauk,

Mo-

scow.

Stell, Nelida

Noemi. 1972.

Fonologia

de

la

Lengua

Alulaj,

Universidadde

Buenos

Aires,

Buenos Aires.

Swadesh, Morris.

1946.

'Chitimacha',

n

CorneliusOsgood

(ed.),

LinguisticStructuresof

Native

America, VikingFund

Publications

n Anthropology,

New

York,pp. 312-336.

Tesar,

Bruce and

Paul

Smolensky. 1993. 'The

Learnabilityof

Optimality

Theory: An

Algorithm

and

Some Basic

Complexity

Results', Technical

Report

CU-CS-678-93,

University

of

Colorado,

Boulder.

Available on

Rutgers

OptimalityArchive,

ROA-2,


Tryon,

Darrell.

1970. An

Introduction o

Maranungku,

AustralianNational

University,

Canberra.

Tyhurst,James.

1987. 'Accent Shift in

Seminole Nouns', in

P. Munro

(ed.), Muskogean

Linguistics[UCLIA

Occasional

Papers in

Linguistics6],

UCLALinguistics

Department,

Los

Angeles,

pp.

161-170.

van

der

Hulst,

Harry.1984. Syllable

Structure

andStress in

Dutch, Foris,

Dordrecht.

Vaux, Bert.

1998.

The

Phonology

ofArmenian,

Clarendon

Press, Oxford.

Walker, Rachel.

1996.

'Prominence-Driven

Stress',

manuscript,

University of

California, Santa Cruz. Available on Rutgers Optimality Archive, ROA-172,


Weiers, Michael. 1972.

Die

Sprache der

Moghol

der Provinz Herat

in

Afghanistan,

Westdeutscher

Verlag,Opladen.

Whitman,

William.

1947.

'Descriptive

Grammarof

loway-Oto',

IntemationalJournal of

American

Linguistics

13(4),

233-248.

Zhgenti, Sergi.

1964.

'The Problem of

Rhythmical

Stress and

Intonation

Structureof

the

Georgian

Language',

Zeitschrift ur

Phonetik,

Sprachwissenschaft

ndKommunika-

tionsforschung

17, 357-368.






552

MATTHEW

ORDON

Zhukova,Alevtina

Nikodimovna. 1972. GrammatikaKoriakskogo

azy/ka:

Fonetika,Mor-

fologiia, lzdatelstvoNauka,Leningrad.

Received23 August 2000

Revised

12

October2001

Department f

Linguistics

Universityof California,

Santa

Barbara,

CA

93106

USA

<mgordon@

inguistics.ucsb.edu>

quantity insensitive stress

Documents