CAS LX 500 Topics in Linguistics Language Universals Fall 2000 October 19, 2000 Paul Hagstrom Week 7b: Prosody
Word stress and Optimality Theory
Kager (1995). The metrical theory of word stress. In Goldsmith, J. (ed.), Handbook of phonological theory. Oxford: Blackwell
Kager (1999). Optimality Theory. Cambridge: Cambridge University Press.
(1) Alabama (Ala)(bama) *
(2) x x x x x x x Alabama
(3) word qp ws 33 swsw A la bam a
Word stress—crosslinguistic preferences
(4) Stressability: Words have to be big enough to carry a stress.
Minimal word requirement.
(5) a. íky4s /k-y4-s/ ‘I put it’ Mohawk b. íktats /k-tat-s/ ‘I offer it’ c. í:keks /k-ek-s/ ‘I eat’ d. íkya÷ks /k-ya÷k-s/ ‘I cut it’ (6) a. s÷aal/s÷al/ ‘ask (masc. sg.)’ Levantine Arabic b. idrus /drus/ ‘study’ Iraqi Arabic
(7) Demarcation: Words should have stress near the edge.
Crosslinguistically popular: Initial, Penultimate, Final. (Primary stress) (8) a. púli˜kalatju ‘we (sat) on the hill’ Pintupi b. yapurukitaneháse ‘verily to climb’ Warao c. ak¨netepál ‘times’ Weri
(9) Rhythm: Bounded. Avoiding lapses and clashes (alternating stress)
(10) tjiÆiri˜ulampatju Pintupi (tjí.Æi).(rì.˜u).(làm.pa).tju ‘the fire for our benefit flared up.’
(11) a. yapurukitanehase Warao (yà.pu).(rù.ki).(tà.ne).(há.se) ‘verily to climb’ b. enahoroahakutai e.(nà.ho).(rò.a).(hà.ku).(tá.i) ‘the one who caused him to eat’
Hungarian: main stress initial, secondary stress on odd-numbered syllables.
(12) a. bóldog ‘happy’ b. bóldogsà:g ‘happiness’ c. bóldogtàlan ‘unhappy’ d. bóldogtàlansà:g ‘unhappiness’ e. légesˇlègmegèngestèlhetètlenèbbeknèk ‘to the most irreconcilable ones’
(13) a. σ« σ b. σ« σ σ` c. σ« σ σ` σ d. σ« σ σ` σ σ` e. σ« σ σ` σ σ` σ σ` σ σ` σ σ` σ σ`
Pintupi: like Hungarian, except no stress on the final syllable
(14) a. pá˜a ‘earth’ b. tjúˇaya ‘many’ c. máÆawàna ‘through from behind’ d. púÆi˜kàlatju ‘we (sat) on the hill’ e. tjámulìmpatjù˜ku ‘our relation’ f. ˇíÆirì˜ulàmpatju ‘the fire for our benefit flared up’ g. kúranjùlulìmpatjùïa ‘the first one (who is) our relation’
(15) a. σ« σ b. σ« σ σ c. σ« σ σ` σ d. σ« σ σ` σ σ e. σ« σ σ` σ σ` σ f. σ« σ σ` σ σ` σ σ g. σ« σ σ` σ σ` σ σ` σ (16) Foot dominance: The side of the foot where the head is (the strong element). left-dominant feet are trochees (X .), bounded right-headed feed are iambs ( . X)
A way to think about this is as building trochees or iambs from one side or the other. So, for Hungarian,
(17) légesˇlègmegèngestèlhetètlenèbbeknèk ‘to the most irreconcilable ones’ (σ« σ) (σ« σ) (σ« σ) (σ« σ) (σ« σ) (σ« σ) … (σ« σ) (σ« σ) (σ« σ) (σ« σ)(σ« σ) (σ« σ) (σ« σ) (σ« σ) (σ« σ) (σ« σ)(σ« σ)(σ« σ) (σ«)
When we get to the last syllable, we can’t build a trochee, so we just make a one-syllable foot. But in Pintupi, we can’t build a trochee, so we leave the syllable out.
(18) ˇíÆirì˜ulàmpatju ‘the fire for our benefit flared up’ (σ« σ) (σ« σ)(σ« σ) (σ« σ)(σ« σ)(σ« σ) (σ« σ)(σ« σ)(σ« σ)σ
In both cases, the primary stress is the leftmost stress.
Left-dominant, right-to left…
Warao: main stress penultimate, secondary on even syllables counting from end
(19) a. yà.pu.rù.ki.tà.ne.há.se ‘verily to climb’ b. e.nà.ho.rò.a.hà.ku.tá.i ‘one who caused him to eat’
Right-dominant, left-to-right.
Araucanian: main stress on the second syllable, secondary on following even
(20) a. e.lú.a.è.new ‘he will give me’ b. ki.mú.fa.lù.wu.lày ‘he pretended not to know’ Right-dominant, right-to-left…
Weri: main stress on final syllable, secondary on preceding odd
(21) a. ulùamít ‘mist’ b. àkunètepál ‘times’
Unbounded. One stress per word, allow long lapses (more rare)
(22) Quantity sensitivity: “Heavy” syllables attract stress. Some languages care about heavy syllables, some don’t.
Seminole/Creek
(23) a. cokó ‘house’ b. osána ‘otter’ c. pomosaná ‘our otter’ d. anokicíta ‘to love’ e. amanokicitá ‘to love mine’ f. amanokicakíta ‘to love mine (pl.subj.)’
(24) a. σ σ« b. σ σ« σ c. σ σ σ σ« d. σ σ σ σ« σ e. σ σ σ σ σ σ« f. σ σ σ σ σ σ« σ
Only the primary stress realized, but still counts from the left in iambs.
If there is a heavy syllable, the counting starts over
(25) a. ta…skitá ‘to jump (sg. subj.)’ b. ta…shokíta ‘to jump (dual subj.)’ c. toko˚hokíta ‘to run (dual subj.)’ d. nihá… ‘lard’ e. hitotí… ‘snow’ (26) a. σ σ σ« b. σ σ σ« σ c. σ σ σ σ« σ d. σ σ« e. σ σ σ«
(27) IAMB: Form (σ σ« ) if possible, otherwise form (σ« ). Constructing a system to predict and account for stress.
(28) word qp ws 33 swsw A la bam a
(29) a. word b. word 31 FS FW F 11 3 σσ σσ cón tèst tém pest
So, there are (prosodic) words, feet, and syllables (so far).
Distinctions in syllable weight which are encoded by moras.
But what counts as heavy differs a little bit from language to language:
(30) CV light CVC light or heavy CVV heavy VV heavy
The idea is that heavy syllables have two of something that light syllables only have one of. The unit of weight is the mora, and it is represented in the prosodic structures. All vowels have a distinct mora, onsets do not (the presence of onsets never seems to contribute to a syllable’s weight class), codas do sometimes:
(31) σσ qgp 3 C µµ C µ t 11 t 1 VV V aa a
(32) σσ qgp 3 C µµ C µ t 11 t 3 VC VC ap ap The intuition about stress in words is that it indicates groupings of syllables (the feet), each group having a prominent syllable (a head).
Prosodic structure: PrWd 1 Foot 1 σ 1 µ 1 x
PrWd Ft Ft σ σ σ µµµ µ p í s k %« lan ‘It rains so hard that it is dark or hard to see’ (Passamaquoddy)
Hixkaryana
(33) a. to.ró….no ‘small bird’ b. ne.mó….ko.tó….no ‘it fell’
IAMBIC A foot is like this: (LH) or (H) if it has to be. PARSE-SYLL Syllables are contained in feet. WSP Heavy syllables are stressed, stressed syllables should be heavy. ALL-FEET-L Every foot is at the left edge of the PrWd.
We can tell: IAMBIC >> PARSE-SYLL It’s more important to have feet of (LH) shape than to foot all syllables.
(34) a.tßó….wo.wo ‘wind’
What’s wrong here? Why didn’t it come out as *a.tßó….wo.wó…?
NONFINALITY No foot is final in PrWd.
So, it seems: NONFINALITY >> PARSE-SYLL
What’s wrong with *a.tßó….wó….wo? This is getting hard to keep track of... /a.tßo.wo.wo/ NONFINALITY IAMBIC PARSE-SYLL (a.tßó…).(wó…).wo *! * ☞ (a.tßó…).wo.wo ** (a.tßó…).(wo.wó…)*!
(35) kwá….ja ‘red and green macaw’
/kwa.ja/ NONFINALITY IAMBIC PARSE-SYLL (kwa.já…)*! ☞ (kwá…).ja * *
NONFINALITY >> IAMBIC >> PARSE-SYLL
(36) a. ák.ma.tá….rπ ‘branch’ b. tóh.ku.rjé….ho.na ‘to Tohkurye’ c. tóh.ku.rjé….ho.ná….ha.ßá….ha ‘finally to Tohkurye’
These seem to be stressing odd syllables… why? Oh, and the vowels aren’t lengthening.
(37) a. nák. ˜óh.yátß.ke.ná….no ‘they were burning it.’ b. kha.ná….n-íh.no ‘I taught it’ c. mπ.há….na.n-íh.no ‘You taught him’
Demuth (1995). Markedness and the development of prosodic structure. NELS 25.
(38) Stage I: Core syllables: CV Stage II: Minimal words/binary feet a. Core syllables: CVCV b. Closed syllables: CVC c. Vowel length distinction: CVV Stage III: Prosodic words (larger than a binary foot) Stave IV: Prosodic words (target)
(Demuth does this with three constraints. NO-CODA, ALIGN-PRWD, NO-DEL). /o:li:fant/ [fa] I [«ho:ta] IIa [faut] IIb [«o:fa «fan] III [«o:li: »fant] IV
(39) Stage I Stage IIa Stage IIb,c Stage III PrWd 1 Ft/PrWd Ft/PrWd Ft 111 σσσσ 11 µµ
Stage I: NO-CODA >> ALIGN-PRWD … >> … NO-DEL Stage IIa: NO-CODA, ALIGN-PRWD … >> … NO-DEL Stage IIb: ALIGN-PRWD … >> … NO-DEL … >> … NO-CODA Stage III *COMPLEX … >> … NO-DEL, ALIGN-PRWD … >> … NO-CODA Stage IV NO-DEL, ALIGN-PRWD … >> … NO-CODA, *COMPLEX CAS LX 500 Language Universals HOMEWORK #7 DUE TUESDAY OCTOBER 24
Wargamay (Australia) (i) báda ‘dog’ (ii) gagára ‘dilly bag’ (iii) gídawùlu ‘freshwater jewfish’ (iv) durágaymìri ‘Niagara Vale-from’
1. State the generalizations about the position of primary («) and secondary (`) stress.
2. Suppose we have the following constraints in universal grammar: TROCHEE: Feet look like (σ« σ) IAMB: Feet look like (σ σ«) ALL-FEET-R All feet should be at the right edge of the PrWd. ALL-FEET-L All feet should be at the left edge of the PrWd.
First, determine the ranking (for Wargamay) of TROCHEE with respect to IAMB, and ALL- FEET-R with respect to ALL-FEET-L. (Is it more important to have trochees or iambs in Wargamay? Is it more important for the feet to be lined up at the left edge of the word or the right edge of the word in Wargamay).
3. Now, take just the top ranked constraint of each pair above (for simplicity, assume the lower ranked constraints are ranked so low that they don’t have any effect). So, for example, TROCHEE and ALL-FEET-L (though your pair may be different). Consider, in addition, the following constraint. Assuming stress always indicates feet (that is, every stress has a foot):where is PARSE-SYLL ranked with respect to TROCHEE/IAMB?
PARSE-SYLL All syllables must be included in a foot
4. ALL-FEET-L counts violations by how far each foot is from the left edge of the word, in syllables. So: #(σσ)(σσ) has two violations because the second foot is two syllables away from the left edge. By comparison, #(σ)(σσ)σ has only one violation because the second foot is only one syllable away from the left edge. The same goes for ALL-FEET-R except with respect to the right edge.
Is ALL-FEET-R/L (whichever one you chose in #2) ranked higher or lower than IAMB/TROCHEE in Wargamay? Is ALL-FEET-R/L ranked higher or lower than PARSE-SYLL?
5. Write out the total ranking of the three constraints (IAMB/TROCHEE, ALL-FEET-R/L, PARSE-SYLL).