22.04.2013 Views

Morphology and prosody in Huave phonological domains

Morphology and prosody in Huave phonological domains

Morphology and prosody in Huave phonological domains

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

<strong>Morphology</strong> <strong>and</strong> <strong>prosody</strong> <strong>in</strong> <strong>Huave</strong> <strong>phonological</strong> doma<strong>in</strong>s<br />

Yuni Kim (yuni.kim@manchester.ac.uk)<br />

University of Manchester<br />

CASTL Colloquium, Tromsø * May 6, 2010<br />

(1) <strong>Huave</strong>: a language isolate of Oaxaca State, Mexico. All data here are from fieldwork on the San Francisco<br />

del Mar dialect (


(6) Phonological: With<strong>in</strong> these morphological constra<strong>in</strong>ts, prefixal vs. suffixal realization is conditioned<br />

largely by <strong>phonological</strong> properties of the base to which the affix attaches (Kim, <strong>in</strong> press). Mobile affixes are<br />

suffixes by default, but prefix to V-<strong>in</strong>itial stems if this will avert the need for epenthesis.<br />

Crucial separation of “affix hierarchy” (abstract constituent structure) from “affix placement” (factors<br />

locat<strong>in</strong>g a mobile affix on one or another side of the stem; Stump 1993, Noyer 1997).<br />

(7) Do the <strong>in</strong>dependently motivated morphological constituents co<strong>in</strong>cide with <strong>phonological</strong> doma<strong>in</strong>s?<br />

(8) Outl<strong>in</strong>e of talk<br />

• Motivat<strong>in</strong>g the layer model of <strong>Huave</strong> verb structure<br />

• Phonology that is sensitive to “<strong>in</strong>ner” (Stem, L1) <strong>and</strong> “outer” (L3+) layers<br />

• The stem + suffix doma<strong>in</strong>: phonology that excludes prefixes --> not a morphological constituent<br />

• Is the stem + suffix a Phonological Word? If not, what other options exist?<br />

Hierarchical constituent structure <strong>in</strong> the <strong>Huave</strong> verb<br />

(9) This section: Despite affix mobility, it can be shown that affixes attach <strong>in</strong> a fixed order. The only th<strong>in</strong>g<br />

that varies is which side of the verb the mobile affixes surface on.<br />

(10) In addition to mobile affixes, <strong>Huave</strong> has prefixes <strong>and</strong> suffixes of the normal k<strong>in</strong>d.<br />

a. Future (FUT) i- i-m-a-rang ‘s/he will do (it)’<br />

FUT-SUB-TV-do<br />

b. “Reflexive” (RFL) -e a-rang-e ‘s/he does (it) for self’<br />

TV-do-RFL<br />

c. 1 st & 2 nd plural (PL) -n i-rang-an ‘you (pl.) do (it)’<br />

2-do-PL<br />

(11) Two k<strong>in</strong>ds of stems<br />

• “Prefix<strong>in</strong>g” (default, all trans. & some <strong>in</strong>trans.) - form stems with a valence-related theme vowel (TV)<br />

• “Suffix<strong>in</strong>g” (rema<strong>in</strong><strong>in</strong>g <strong>in</strong>transitives) - if C-f<strong>in</strong>al, form stems with a suffix vowel (V) of predicatble<br />

harmonic quality <strong>and</strong> debatable morphological status<br />

• (Terms are based on placement of earliest-attach<strong>in</strong>g mobile affixes, <strong>and</strong> do not mean that a stem<br />

conditions only one or the other type of affix placement)<br />

Prefix<strong>in</strong>g stems Suffix<strong>in</strong>g stems<br />

a. a-jaw ‘see, know (it)’ d. mojk-o ‘lie face down’<br />

TV-see<br />

face.down-V<br />

b. a-chiot ‘break (it)’ e. chiot-a ‘break’, itr.<br />

TV-break<br />

break-V<br />

c. u-ndiak ‘speak’, itr. f. tye- ‘hang’, itr.<br />

TV-speak<br />

hang<br />

(12) “Layers” <strong>in</strong> <strong>Huave</strong> verb structure: alternat<strong>in</strong>g zones of mobile <strong>and</strong> non-mobile affixes. Mobile affixes<br />

are found <strong>in</strong> Layers 1 <strong>and</strong> 3. Layer 2 consists of prefixes/suffixes occurr<strong>in</strong>g between Layers 1 <strong>and</strong> 3, <strong>and</strong> Layer 4<br />

consists of prefixes/suffixes outside Layer 3. With<strong>in</strong> each layer there is still a fixed order of attachment.<br />

[L4 [L3 [L2 [L1 [Stem] L1] L2] L3] L4]<br />

| | | | | | | |<br />

Prefix Mobile Prefix Mobile Mobile Suffix Mobile Suffix<br />

(13) Constituent structure is not obvious when affixes are on opposite sides of the root.<br />

t a rang as OR t a rang as ?<br />

CPL-TV-do-1 CPL-TV-do-1<br />

But relative order<strong>in</strong>g is observable when two mobile affixes are on the same side of the root.<br />

2


(14) 2 nd -<strong>in</strong>transitive -r- (2I) is <strong>in</strong>side (closer to the root than) 2 nd person & everyth<strong>in</strong>g outside that<br />

a. Nonapparent: m-e-wit-ior ‘that you (sg.) get up’ t-e-wit-ior ‘you (sg.) got up’<br />

SUB-2-rise-2I CPL-2-rise-2I<br />

b. More apparent: m-e-r-u-ty ‘that you (sg.) eat’ t-e-r-u-ty ‘you (sg.) ate’<br />

SUB-2-2I-TV-eat CPL-2-2I-TV-eat<br />

(15) Completive -t- <strong>and</strong> 1 st -subord<strong>in</strong>ate -n- are <strong>in</strong>side first-person -s-/-x-:<br />

a. Nonapparent: t-a-rang-as ‘I did (it)’ x-i-wit-io-n ‘I will get up’<br />

CPL-TV-do-1 1-FUT-rise-V-1SB<br />

b. More apparent: wit-io-t-u-s ‘I got up’ x-i-n-a-rang ‘I will do (it)’<br />

rise-V-CPL-ITR-1 1-FUT-1SB-TV-do<br />

c. And: ndil-i-t-ey-as ‘I turned my head’ ndil-i-n-ey-as ‘that I turn my head’<br />

turn-V-CPL-RFL-1 turn-V-1SB-RFL-1<br />

(16) Note <strong>in</strong> (14b) that first-person -s- is outside FUT on the prefix side...<br />

... while <strong>in</strong> (14c), it is outside RFL on the suffix side. Verb structure, schematically:<br />

[ s FUT ..... STEM .... RFL s ]<br />

L3 L2 L1 L1 L2 L3<br />

(17) Strategy: use the “fixed po<strong>in</strong>ts” FUT <strong>and</strong> RFL to divide the verb <strong>in</strong>to zones, <strong>and</strong> group affixes.<br />

a. Layer 1 - affixes that are <strong>in</strong>side FUT <strong>and</strong> RFL<br />

b. Layer 2 - fixed po<strong>in</strong>ts FUT <strong>and</strong> RFL<br />

c. Layer 3 - affixes that are outside FUT <strong>and</strong> RFL<br />

(18) Justify<strong>in</strong>g Layer 1: All mobile affixes that are <strong>in</strong>side FUT when prefixal are also <strong>in</strong>side RFL when<br />

suffixal. CPL -t- <strong>and</strong> FUT never cooccur, but we <strong>in</strong>fer Layer 1 as CPL is <strong>in</strong>side RFL (17g).<br />

[L3 [L2 [L1 [Stem] L1] L2] L3]<br />

a. x i n a-uñ e ‘I will go shopp<strong>in</strong>g’<br />

1 FUT 1SB TV-buy RFL<br />

b. x i ndil -i n e ‘I will turn around’<br />

1 FUT turn 1SB RFL<br />

c. i m a-uñ e ‘s/he will buy (it) for self)<br />

FUT SUB TV-buy RFL<br />

d. i ndil -i m e ‘s/he will turn around’<br />

FUT turn SUB RFL<br />

e. i m-e-r u-ty ‘you (sg.) will eat’<br />

FUT SUB-2-2I-TV-eat<br />

f. i m-e ndil -i r e ‘you (sg.) will turn around’<br />

FUT SUB-2 turn 2I RFL<br />

g. ndil -i t ey as ‘I turned around’<br />

turn CPL RFL 1<br />

3


(19) Layer 3 consists just of the mobile 1 st Person affix -s-. It occurs outside all L1 <strong>and</strong> L2 affixes.<br />

Even where it does not cooccur with other affixes, we can now diagnose it as occupy<strong>in</strong>g an outer<br />

morphological slot.<br />

[L3 [L2 [L1 [Stem] L1] L2] L3]<br />

a. x- [i- n- a-jch] ‘I will give (it)’<br />

1 FUT 1SUB TV-give<br />

b. s- [a-jch] ‘I give (it)’<br />

1 TV-give<br />

c. x- [i- chut-u -n] ‘I will sit’<br />

1 FUT sit-V 1SUB<br />

d. [t- a-jch] -ius ‘I gave (it)’<br />

CPL TV-give 1<br />

e. [t- a- xot -ey] as ‘I hid myself’<br />

CPL TV-hide RFL 1<br />

(20) Layer 4 conta<strong>in</strong>s the non-cooccurr<strong>in</strong>g plural-mark<strong>in</strong>g suffixes...<br />

o -n (default)<br />

o -f/-w/-h/-Ø (3 rd person; allomorphs due to productive <strong>phonological</strong> rules)<br />

o -r <strong>and</strong> -ts (1 st <strong>in</strong>clusive, <strong>in</strong>terchangeable)<br />

... plus the progressive (PROG) prefix ndyu-/dyu- (phonetically variable prenasalization)<br />

(21) [L4 [L3 [L2 [L1 [Stem] L1] L2] L3] L4]<br />

a. [t- a-ndiak -as] -an ‘We (excl) spoke’<br />

CPL TV-speak 1 PL<br />

b. [t- a-xum -us] -uØ ‘I found them’<br />

CPL TV-f<strong>in</strong>d 1 3PL<br />

c. [i- m- a-jaw -ey] -ar ‘We’ll see e.o.’<br />

FUT SUB TV-see RFL INCL<br />

d. dyu- s- a-t -ion ‘We (excl) are eat<strong>in</strong>g’<br />

PROG 1 TV-eat PL<br />

(22) Summary: Affixes occur <strong>in</strong> strict layer-order; distance from the root is morphologically specified.<br />

L0 1 L1 L2 L3 L4(a/b)<br />

Prefixes -- i-/e- 2 i- FUT -- ndyu- PROG<br />

Suffixes -(j)ch CAUS [+rd] ITR -e RFL -- -rVn PASS -n PL<br />

-f 3PL<br />

-r INC<br />

Mobile -- t CPL<br />

m SB<br />

n 1SB<br />

n ST<br />

r 2I<br />

-- s 1 --<br />

4<br />

-(j)ts INC<br />

(23) Semantic or other pr<strong>in</strong>ciples beh<strong>in</strong>d the morphologically specified order await future research. But<br />

several aspects seem problematic for scopal or mirror-pr<strong>in</strong>ciple approaches, for example:<br />

o 2 nd person <strong>in</strong> Layer 1 but 1 st person <strong>in</strong> Layer 3<br />

o 2 nd person <strong>in</strong>flection <strong>in</strong> Layer 1 but Reflexive derivation <strong>in</strong> Layer 2<br />

1 The number<strong>in</strong>g of the first layer as “0” <strong>in</strong>dicates the nonproductive status of the causative/transitiviz<strong>in</strong>g affix.


Morphologically conditioned phonology<br />

(24) Vowel epenthesis with consonantal suffixes is common, s<strong>in</strong>ce consonant clusters are not allowed.<br />

Epenthesis after V-f<strong>in</strong>al roots: never with Layer 1 suffixes, always with Layer 3 <strong>and</strong> 4 suffixes<br />

(25) L1 L3<br />

a. tye-m b. *tye-am c. t-a-tyey-as d. *t-a-tye-s<br />

hang-SUB CPL-TV-hang-1<br />

Stem-L1 L1-Stem-L3<br />

‘that it hangs’<br />

L1 L4<br />

e. rra-t f. *rra-at g. a-sa-af h. *a-sa-f<br />

dawn-CPL TV-tell-3PL<br />

Stem-L1 Stem-L4<br />

‘It dawned’ ‘they say, tell’<br />

(26) Overapplication of diphthongization: variably with L3/L4 suffixes, (almost) never with L1 suffixes.<br />

(27) Basic consonant <strong>in</strong>ventory:<br />

Labial Coronal Velar Labiovelar Glottal<br />

Stops p mb t nd k ng[ŋg] kw ngw[ŋgw]<br />

Affricates ts nts<br />

Fricatives f[ ] s h<br />

Nasals m<br />

n<br />

Liquids l r[ ] rr[r]<br />

Glides w j<br />

(28) All consonants come <strong>in</strong> pla<strong>in</strong> <strong>and</strong> palatalized versions (/C bk /, /C pal /)<br />

• Palatalization as secondary/abstract (no POA change): All non-coronals <strong>and</strong> rhotics<br />

• Palatalization <strong>in</strong> <strong>in</strong>herent place: Non-rhotic coronals<br />

Pla<strong>in</strong> t nd l ts nts s n<br />

Palatal ty [c] ndy [ ] ly [ ] ch [ ] nch [ ] x [ , ] ñ [ ]<br />

(29) Surface vowels: monophthongs [i e a o u] <strong>and</strong> diphthongs (from underly<strong>in</strong>g monophthongs)<br />

Diphthongs are observed where VC]σ conflict <strong>in</strong> frontness/backness; the frontness/backness of the coda<br />

consonant is realized on the second half of the vowel nucleus.<br />

Front monophthongs [i e] Ris<strong>in</strong>g diphthongs io [jəә, jo] <strong>and</strong> ia [ja]<br />

a. /xix/ xix 'mosquito' c. /mik bk / miok 'bat'<br />

b. /pek pal / pek 'shoulder’ d. /pets bk / piats 'tortilla'<br />

Central/Back monophthongs [a o u] Fall<strong>in</strong>g diphthongs ai oi ui [aj oj uj]<br />

e. /pang bk / pang 'chair' h. /-lak pal / u-laik 'his/her tooth'<br />

f. /ndok bk / ndok 'fish<strong>in</strong>g net' i. /-long pal / a-lo<strong>in</strong>g 'hang, 3sg.'<br />

g. /xur bk / xur 'pot' j. /puk pal / puik 'feather; fur'<br />

(30) S<strong>in</strong>ce <strong>Huave</strong> disallows consonant clusters (e.g. *CVCCVC), codas only occur word-f<strong>in</strong>ally.<br />

Normally when suffixes are added, coda-C front/back features can be realized on a follow<strong>in</strong>g vowel,<br />

the underly<strong>in</strong>g monophthong surfaces.<br />

a. /i-lak pal / i-laik ‘your tooth’ b. /i-lak pal -<strong>in</strong>/ i-lak-ion ‘your (pl.) teeth’<br />

2-tooth 2-tooth-PL<br />

c. /a-jir bk / a-jior ‘has’ d. /a-jir bk -af/ a-jir-af ‘they have’<br />

TV-have TV-have-3PL<br />

5


(31) Optional overapplication:<br />

a. /t-a-lang pal -is/ t-a-la<strong>in</strong>g-ius ~ t-a-lang-ius<br />

CP-TV-cross-1 CP-TV-cross-1<br />

‘I crossed (the road)’<br />

b. /t-a-chit-as/ t-a-chiot-as ~ t-a-chit-as<br />

CP-TV-break-1 CP-TV-break-1<br />

‘I broke (it)’<br />

c. /t-a-jimb pal -is-an/ t-a-jimb-ios-an ~ t-a-jimb-is-an<br />

CP-TV-sweep-1-PL CP-TV-sweep-1-PL<br />

‘We (excl.) swept’<br />

(32) Can be seen as a cyclic effect: the environment obta<strong>in</strong>s at <strong>in</strong>ner levels (Stem/L1, L3).<br />

At each level, there is an option to apply diphthongization, if the conditions are met; variation arises<br />

from whether diphthongization applies “earlier,” or at the level of the whole word.<br />

Root UR Stem L1 L3 L4 Post-cyclic/Word level<br />

a. /-lang pal / a-la(i)ng t-a-la(i)ng t-a-la(i)ng-i(u)s -- t-a-la(i)ng-ius<br />

cross add TV add CPL add 1<br />

b. /-jimb pal / a-jimb t-a-jimb t-a-jimb-i(o)s t-a-jimb-i(o)s-an t-a-jimb-i(o)s-an<br />

sweep add TV add CPL add 1 add PL<br />

Such variation <strong>in</strong> application doma<strong>in</strong> is predicted by the natural diachronic “life cycle” of<br />

phonolological processes (see e.g. Zec 1993).<br />

(33) Diphthongization never(?) overapplies <strong>in</strong> suffix<strong>in</strong>g stems: even at stem level, conditions are not met.<br />

However, there is the possibility for diphthongization at other pre-word levels.<br />

(More data needed to confirm exactly how “bad” diphthongization is <strong>in</strong> this environment.)<br />

Root UR Stem L1 Post-cyclic/Word level<br />

a. /chup pal / chup-i chup-i(o)-m chup-io-m */? chuip-io-m<br />

fill up add V add SUB<br />

(34) Summary: Vowel epenthesis <strong>and</strong> diphthongization show that phonology can be sensitive to<br />

morphological constituency; <strong>in</strong>dependently established morphological doma<strong>in</strong>s also seem to be referred to<br />

<strong>in</strong> morpho<strong>phonological</strong> processes.<br />

The Stem + Suffix Doma<strong>in</strong><br />

(35) The processes discussed until now have only obta<strong>in</strong>ed <strong>in</strong> suffixal contexts, but their application cannot<br />

really be tested <strong>in</strong> the morphologically correspond<strong>in</strong>g prefixal doma<strong>in</strong>s.<br />

• Vowel epenthesis: not attested with prefixes (though San Mateo del Mar dialect may have this)<br />

• Diphthongization: no C-f<strong>in</strong>al prefixes, so no possibility for word-medial coda consonants<br />

(36) Two additional <strong>phonological</strong> processes show a more dramatic limitation to the stem + suffix doma<strong>in</strong>,<br />

where prefixes clearly provide the <strong>phonological</strong> environment - which however is ignored.<br />

(37) Vowel harmony: basic pattern<br />

Base -VC Epenthetic V Base -VC Epenthetic V<br />

iC bk a iC pal i<br />

eC bk a eC pal e<br />

aC bk a aC pal i<br />

oC bk o oC pal i<br />

uC bk u uC pal i<br />

6


(38) Generalizations:<br />

a. Epenthetic V always matches the <strong>in</strong>terven<strong>in</strong>g C for [±back] – phonotactically driven<br />

b. Preced<strong>in</strong>g vowel matters (=is copied) only if it also matches the <strong>in</strong>terven<strong>in</strong>g consonant for [±back]<br />

c. Otherwise, default epenthetic vowel is /a/ after C, <strong>and</strong> /i/ after C’.<br />

“Sour grapes” (Padgett 1995): If you can’t copy, due to the phonotactics of [±back], pretend like you never<br />

wanted to propagate any features <strong>in</strong> the first place. Insert default vowel.<br />

(39) Vowel harmony illustrations: Epenthetic vowel quality (1 st person completive forms: tcpl-atv-ROOT-Vs1)<br />

CVC root (Pla<strong>in</strong> C2) CVC’ root (Palatalized C2)<br />

i a. |t-a-mit-as| ‘I buried (it)’ f. |t-a-xijp pal -is| ‘I bathed (it)’<br />

e b. |t-a-ndek-as| ‘I spoke (it)’ g. |t-a-rrejk pal -es| ‘I touched (it)’<br />

u c. |t-a-mut-us| ‘I wrote (it)’ h. |t-a-un pal -is| ‘I bought (it)’<br />

o d. |t-a-ndok-os| ‘I fished (it)’ i. |t-a-kots pal -is| ‘I scratched (it)’<br />

a e. |t-a-pal-as| ‘I closed (it)’ j. |t-a-kal pal -is| ‘I entered’<br />

(40) So: epenthetic vowel quality depends on the preced<strong>in</strong>g VC sequence.<br />

However, <strong>in</strong> the case of s<strong>in</strong>gle-consonant roots, the preced<strong>in</strong>g V-C will span a prefix-root boundary.<br />

2 nd person plural forms (abstract<strong>in</strong>g away from diphthongization):<br />

a. /mb bk / ‘go’ t-e-mb-an Expected default [bk] after VfrontC bk<br />

CPL-2-go-PL<br />

b. /jts pal / ‘give’ m-e-jch-<strong>in</strong> *m-e-jch-en<br />

SUB-2-give-PL<br />

c. /w pal / ‘borrow’ t-e-w-<strong>in</strong> *t-e-w-en<br />

CPL-2-borrow-PL<br />

(41) In (40bc), default front vowels are <strong>in</strong>serted, even though the environment for full vowel copy is met.<br />

It appears that vowel harmony cannot see leftward past the root.<br />

(42) Laryngeal dissimilation: the second of two [+spread glottis] with<strong>in</strong> a one-syllable w<strong>in</strong>dow is deleted.<br />

All voiceless fricatives are [+spread glottis]: j [h], s, x [ɕ], f [ɸ]<br />

In these examples, passive alternations are produced by add<strong>in</strong>g pre-root aspiration <strong>and</strong> depalatalizat<strong>in</strong>g<br />

the root coda. The pre-root aspiration causes a vowel aspiration <strong>in</strong> the root to disappear.<br />

a. a-naijp<br />

TV-sell<br />

‘S/he sells it’<br />

c. a-paj<br />

TV-call<br />

‘S/he calls, yells’<br />

b. a-j.nap<br />

TV-be.sold<br />

‘It is sold’<br />

d. t-a-j.pa<br />

CP-TV-be.called<br />

‘S/he was called’<br />

(43) The follow<strong>in</strong>g examples show how the 3 rd pl. suffix -f loses [+s.g.], but reta<strong>in</strong>s labiality, when<br />

follow<strong>in</strong>g a root with [+s.g.].<br />

a. t-a-kaj-aw<br />

CP-TV-look.for-3PL<br />

‘they looked for (it)’ (3;41)<br />

c. a-j.ch-iow<br />

TV-give-3PL<br />

‘they give’ (1;174)<br />

e. t-a-xijp-iow<br />

CP-TV-bathe-3PL<br />

‘they bathed’ (3;41)<br />

g. u-lajk-aw<br />

POS1-ear-3PL<br />

‘their ears’ (1;173)<br />

b. t-a-kojñ-iow<br />

CP-TV-fold-3PL<br />

‘they folded (it)’ (3;46)<br />

d. mi-kius-aw<br />

POS-dog-3PL<br />

‘their dogs’ (1;173)<br />

f. dyu-m-a-lox-iow<br />

PROG-SB-TV-throw-3PL<br />

‘they are throw<strong>in</strong>g (it)’ (3;42)<br />

h. u-wix-iow<br />

POS1-h<strong>and</strong>-3PL<br />

‘their h<strong>and</strong>s’ (1;173)<br />

7


(44) Sibilants s <strong>and</strong> x trigger, as shown above, but do not undergo dissimilation.<br />

a. t-a-j.tyux b. t-a-najp-ius<br />

CP-TV-rot CP-TV-sell-1<br />

‘it rotted’ ‘I sold it’<br />

c. piaj-u-s d. ty-u-paj-as<br />

lie-ITR-1 CP-TV-yell-1<br />

‘I lie down' ‘I yelled’<br />

(45) However, when s- is a prefix, it never triggers [+s.g.] deletion <strong>in</strong> the root.<br />

This cannot be a simple root-affix asymmetry, s<strong>in</strong>ce suffixal s (L3) triggers dissimilation <strong>in</strong> the 3PL -f.<br />

a. s-a-jch b. s-a-j.mba<br />

1-TV-give 1-TV-break<br />

‘I give it’ ‘I break (it)’<br />

c. t-a-xum-us-uØ<br />

CPL-TV-1-3PL<br />

‘I met them’<br />

(46) Summary: Vowel harmony <strong>and</strong> laryngeal dissimilation cannot see past the stem to prefixes.<br />

Therefore the stem + suffixes constitute some k<strong>in</strong>d of doma<strong>in</strong> for the application of phonology.<br />

Discussion<br />

(47) How does the stem + suffix doma<strong>in</strong> fit with the morphological picture developed earlier?<br />

[L4 [L3 [L2 [L1 [Stem] L1] L2] L3] L4]<br />

It doesn’t: there is no morphological constituent boundary between the prefixes <strong>and</strong> stem.<br />

The <strong>phonological</strong> doma<strong>in</strong> is def<strong>in</strong>ed l<strong>in</strong>early, but morphological constituents are symmetrical <strong>and</strong><br />

hierarchical.<br />

(48) Could the stem + suffixes be a Phonological Word?<br />

Problematic: neither prefixes nor stems always form <strong>in</strong>dependently valid <strong>phonological</strong> units.<br />

a. [s]-[a-jch] b. [m-e]-[jch]<br />

1-TV-give SUB-2-give<br />

L3-Stem L1-L1-Stem<br />

(49) Possibilities: morphological symmetry has a diachronic basis but is synchronically an illusion (i.e.<br />

prefixes are morphologically separate doma<strong>in</strong>); m<strong>in</strong>imality on (48b) is assessed with a prefixal theme vowel<br />

that is later deleted; phonology can see morphology without be<strong>in</strong>g mediated by the Phonological Word.<br />

Acknowledgements<br />

I would like to thank the <strong>Huave</strong> speakers <strong>in</strong> San Francisco del Mar who have patiently <strong>and</strong> generously hosted me <strong>and</strong> helped<br />

me learn about their language. Fieldwork was supported by Richard Diebold’s Salus Mundi Foundation via the <strong>Huave</strong><br />

Language <strong>and</strong> Culture Project grant to Bill Hanks <strong>and</strong> Maurizio Gnerre; the research was also supported by a National<br />

Science Foundation graduate fellowship.<br />

References<br />

Kim, Yuni. 2008. Topics <strong>in</strong> the phonology <strong>and</strong> morphology of San Francisco del Mar <strong>Huave</strong>. PhD dissertation, UC<br />

Berkeley.<br />

Kim, Yuni. In press. Phonological <strong>and</strong> morphological conditions on affix order <strong>in</strong> <strong>Huave</strong>. <strong>Morphology</strong> 20(1).<br />

Noyer, Rolf. 1993. Mobile affixes <strong>in</strong> <strong>Huave</strong>: optimality <strong>and</strong> morphological well-formedness. In E. Duncan, D. Farkas <strong>and</strong> P.<br />

Spaelti, eds., Proceed<strong>in</strong>gs of the Twelfth West Coast Conference on Formal L<strong>in</strong>guistics, 67-82. Stanford: CSLI.<br />

Noyer, Rolf. 1997. Features, positions <strong>and</strong> affixes <strong>in</strong> autonomous morphological structure. New York: Garl<strong>and</strong>.<br />

Stump, Gregory. 1992. Position classes <strong>and</strong> morphological theory. In G. Booij <strong>and</strong> J. van Marle, eds., Yearbook of<br />

<strong>Morphology</strong> 1992, 129-180. Dordrecht: Kluwer.<br />

Zec, Draga. 1993. Rule doma<strong>in</strong>s <strong>and</strong> <strong>phonological</strong> change. In S. Hargus <strong>and</strong> E. Kaisse, eds., Studies <strong>in</strong> Lexical Phonology,<br />

365-405. San Diego, CA: Academic Press.<br />

8

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!