Optimality Theory and Prosody
Michael Hammond
University of Arizona

A. Overview

What is linguistics, phonology, and prosody?
Syllables
Constraints and rules
How to do syllables with OT?
Stress
Why should you care?

B. Background

GENERATIVE LINGUISTICS attempts to characterize and account for the acquisition of the unconscious knowledge a speaker has of his/her language

brick an actual word of English
bnick a nonoccurring and impossible word of English
blick a nonoccurring but POSSIBLE word of English

Phonology is the study of sound systems. What sounds can a language have? Are there limits on the distribution of sounds in a language.

A phonological analysis of some phenomenon is a model of what a speaker knows (unconsciously) about his/her langauge.

A theory of phonology is a proposal about the innate predisposition a language learner has. This innate predisposition is specific enough so that only "possible" languages can be learned, yet general enough so that any of those languages can be learned.

Prosody ~ syllables and stress.

C. Syllables

Syllables: consonants and vowels grouped into peaks of intrinsic loudness (SONORITY). In English, there is usually a vowel in the middle.

graphically:

impressionistically:
 1 4 3 1 4 2
[s a r d i n]
   sardine

Are these simply acoustic patterns or are they reflective of the syllable as a unit of cognitive organization?

Argument #1: words are made up of syllables (LICENSING). 
                 a  ardins        431421        
                 b  *rdinsa       314214        
                 c  dinsar        142143        
                 d  insard        421431        
                 e  *nsardi       214314

Argument #2: EXPLETIVE INFIXATION 
                 a  *sf-ardin       1f-43142        
                 b  *saf-rdin       14f-3142        
                 c  sarf-din        143f-142        
                 d  *sardf-in       1431f-42        
                 e  *sardif-n       14314f-2

Conclusion: syllables are a unit of cognitive organization.

*sardf-in vs. sarf-din, but dinsar and insard.

Assume that syllabification is subject to two principles:
a. FAITH: pronounce everything.
b. ONSET: syllables want a consonant on the left.

Priority (RANKING): That words in English can begin with vowels shows that the pressure for syllables to have onsets is outranked by the need to pronounce what you have (FAITH >> ONSET).

maltreat 2431341: mælf-trit vs. *mæltf-rit

Another principle: NOCODA: syllables don't want consonants on the right.

FAITH >> {ONSET,NOCODA}

D. Typological consequences

There are four possible rankings of these principles which predict 4 kinds of languages. 
             a      (O)V(C)    FAITH >> {ONSET,NOCODA}
             b      OV         {ONSET,NOCODA} >> FAITH
             c      OV(C)      ONSET >> FAITH >> NOCODA
             d      (O)V       NOCODA >> FAITH >> ONSET

Examples of each type: 
		(O)V(C)		English
		OV		Senufo (Guinea)
		OV(C)		Yawelmani (California)
		(O)V		Hawaiian

None of the following are possible languages: 
                                 onset
                          Ø      req'd     opt.
               Ø          V      -         -
         coda  req'd      VC     OVC       (O)VC
               opt.      V(C)    -         -

This approach thus accounts for the array of possible languages with the same descriptive machinery required to treat one language (English).

E. Constraints and Rules

The analysis above was in terms of the CONSTRAINT-based OPTIMALITY THEORY. What's the alternative?

RULES. Syllables are constructed in a step-by-step fashion, gradually converting unsyllabified strings into syllabified strings.

A rule-based approach: 
	a.	Vowel Rule (#1):	V -> [V]
	b.	Onset Rule (#2):	c[V] -> [OV]
	c.	Coda Rule (#3):		...V]c -> ...VC]

English (sardine, Joey) 
            unsyllabified     sardin      joi
                  Rule #1   s[a]rd[i]n    j[o][i]
                  Rule #2   [sa]r[di]n    [jo][i]
                  Rule #3   [sar][din]    [jo][i]

Notice that rules need not "succeed" if their environment isn't met.

The rule-based typology 
          language type      req'd rules                req'd success?
                OV           vowel, onset               onset
             (O)V(C)         vowel, onset, coda
               (O)V          vowel, onset
              OV(C)          vowel, onset, coda         onset

Constraints are better than rules 
	a.	"the rule must succeed" != "the rule exists"
	b.	"The rule must succeed" ~ a constraint

F. Stress

Stress: syllables are organized into patterns of alternating prominence (METRICAL FEET). In English, each foot contains a stressed syllable on the left and at most one stressless syllable on the right .

graphically:

impressionistically:
 2  0 1 0
Minnesota

Are these simply acoustic patterns or are they reflective of the foot as a unit of cognitive organization?

Argument #1: words are made up of feet (LICENSING):



Argument #2: EXPLETIVE INFIXATION 
                 a     *Mif-nnesota     [2f-0][10]
                 b     Minnef-sota      [20]f-[10]
                 c     *Minnesof-ta     [20][1f-0]

G. Constraints and Rules Again

Why do stress with constraints instead of rules?

Under certain conditions, a stressless vowel can go away in fast speech (SYNCOPE)
  1. a. at the beginning of words:
    • paráde -> práde
    • Torónto -> Trónto
    • Canádian -> Cnádian
  2. b. before a stressless syllable
    • ópera -> ópra
    • géneral -> génral
    • chócolate -> chóclate
  3. c. before a stressless syllable before a stressed syllable
    • réspiratòry -> réspratòry
    • glòrificátion -> glòrficátion

Compare:

A rule-based account: Remove a stressless vowel if it precedes a stressless syllable or if it is word-initial.

Problem: the syncope rule misses the generalization that vowels syncopate only when an optimal (=disyllabic) foot would result. The two environments above can only be reduced to a single environment when the output is considered.

A constraint-based account:
		a.	STRESS: pronounce stressed vowels.
		b.	FOOTLESS: avoid unfooted syllables.
		c.	STRESSLESS: pronounce unstressed vowels.

STRESS >> FOOTLESS >> STRESSLESS

H. Why should you care?

Why should you care about prosody generally?

Syllables and feet figure in lots of other domains

Why should you care whether linguists characterize syllables and feet in terms of constraints instead of rules?

It suggests that the mind is structured in a different way.

It provides a constraint-based model that might be profitably exported to other domains.

Constraint-based analysis make different empirical claims and it thus behoves anybody interested in the role of syllables and feet in any domain to consider these different claims.

References

Bagemihl, B. (1988) Alternate Phonologies and Morphologies, UBC doctoral dissertation. [language games]

Chomsky, N. and M. Halle (1968) The Sound Pattern of English, Harper and Row, New York. [the classic rule-based theory of phonology]

Cutler, A., Mehler, J., Norris, D., & Segui, J. (1986) "The syllable's differing role in the segmentation of French and English", JML 25, 385-400. [classic paper on the role of the syllable in speech perception]

Gerken, L.A. (1994) "Young children's representation of prosodic structure: Evidence from English-speakers' weak syllable omissions", JML 33, 19-38. [children's language production uses feet]

Hammond, M. (1991) "Poetic meter and the arboreal grid", Language 67, 240-259. [metrical theory applied to poetic meter and a recent discussion of Expletive Infixation]

Hammond, M. (1994) "Syllable Parsing in English and French", ms., U. of Arizona. [OT applied to speech perception etc.]

Hammond, M. (1995) "Metrical theory", Annual Review of Anthropology 24, 313-342. [an overview of metrical phonology]

Hayes, B. (1981) A Metrical Theory of Stress, 1980 MIT doctoral dissertation, distributed by IULC, published by Garland Press 1985. [a classic rule-based approach to stress]

Jakobson, R. (1962) Selected Writings: Phonological Studies, Mouton, The Hague. [the presentation of the syllable typology claim]

Kenstowicz, M. (1994) Phonology in Generative Grammar, Blackwell, Cambridge. [pre OT phonology text with a problem on English syncope]

McCarthy, J. & A. Prince (1993) Prosodic Morphology I, ms., U. Mass. and Rutgers U. [the first real empirical elaboration of OT]

Meador, D. (in prep.) UA dissertation. [soon to be classic dissertation on the syllable in lexical access!]

Mester, R. A. (1994) "The quantitative trochee in Latin", NLLT 12, 1-62. [constraint-based footing]

Pérez, P. (1992) "Gradient Sonority and Harmonic Foot Repair in English Syncope", Coyote Papers 8, 118-142. [an early OT-esque treatment of syncope in English]

Prince, A. (1990) "Quantitative consequences of rhythmic organization", CLS 26. [early paper on optimizing foot structure]

Prince, A. & P. Smolensky (1993) Optimality Theory, ms., Rutgers U. and U. of Colorado. [the first OT work]

Steriade, D. (1982) Greek Prosodies and the Nature of Syllabification, MIT doctoral dissertation. [classic rule-based syllabification]

Zwicky, A. (1972) "Note on a Phonological Hierarchy in English", in R. Stockwell and R. Macaulay, eds., Linguistic Change and Generative Theory: Essays from the UCLA Conference on Historical Linguistics in the Perspective of Transformational Theory, IULC, 275-301. [a rule-based treatment of syncope in English]