Questions: Constraint Ranking and Typology in Optimality Theory

Question 1

Language A ranks MAX >> DEP; Language B ranks DEP >> MAX. Both languages encounter an input with a final consonant cluster that violates a high-ranked NOCODA constraint. What is the predicted difference in their repair strategies?

• A. Language A deletes the extra consonant; Language B inserts a vowel to break up the cluster
• B. Language A inserts a vowel to preserve the consonants; Language B deletes a consonant rather than insert
• C. Both languages insert a vowel, because NOCODA dominates both MAX and DEP
• D. Both languages delete the consonant, because MAX and DEP rank lower than NOCODA in both cases

Answer: (B)

MAX penalizes deletion of input material; DEP penalizes insertion. When NOCODA is high-ranked and forces a repair, the choice of repair depends on which of MAX and DEP ranks higher. In Language A (MAX >> DEP), preserving consonants matters more than avoiding insertion, so a vowel is epenthesized. In Language B (DEP >> MAX), avoiding new material matters more, so the consonant is deleted. A single ranking reversal produces a genuinely different phonological behavior — neither grammar is defective, they simply have different priorities.

Question 2

An OT linguist finds a phonological pattern that does not appear in any attested human language. What does factorial typology predict about this pattern?

• A. The pattern is a coincidental gap — it might appear in an unsampled language
• B. The pattern is impossible if it is not generated by any ranking of the universal constraint set
• C. The pattern is impossible only if it violates a constraint that is undominated in all languages
• D. Nothing definitive — OT cannot make predictions about unattested patterns

Answer: (B)

Factorial typology makes a strong prediction: any pattern not generated by any possible ranking of universal constraints is predicted to be impossible, not merely unattested. This distinguishes OT from earlier frameworks where gaps could be written off as accidental absences from the sample. If a pattern falls outside the factorial typology, OT predicts it will never be found — a genuinely falsifiable claim that finding such a pattern in any language would challenge.

Question 3

Different languages have different phonological systems because they evolved different sets of constraints — some languages developed constraints others lack.

• T. True
• F. False

Answer: False

OT's central claim is that all languages share the same universal constraint set — no language has a constraint unavailable to others. What differs is the *ranking* of these shared constraints. Linguistic diversity emerges not from different constraint inventories but from different priority orderings of the same universal set. This is a radical hypothesis: a typology built on constraint inventories would make weaker predictions than one based on universal rankings.

Question 4

In Optimality Theory, strict domination means a single violation of a higher-ranked constraint makes a candidate worse than any number of violations of lower-ranked constraints combined.

• T. True
• F. False

Answer: True

Strict domination is the defining formal property of OT ranking. Unlike weighted constraint systems (where many small violations might outweigh one large one), OT uses pure rank order: the highest-ranked constraint that distinguishes two candidates determines the winner, regardless of what happens lower in the hierarchy. This is what makes OT's typological predictions crisp and its factorial typology well-defined.

Question 5

How does OT constraint ranking explain harmonic typologies — the observation that certain phonological properties tend to co-occur across languages?

Model answer: Harmonic typologies arise because the same high-ranked constraints that penalize one feature also penalize related features. A language that ranks NoCoda very high will tend to lack both syllable codas and complex onsets, because the same constraints that eliminate codas interact with consonant cluster constraints to disfavor complex onsets as well. These co-occurrences are not stipulated as separate universals — they emerge automatically from the constraint interaction logic. OT makes such correlations explanatory rather than merely descriptive.

This unification is one of OT's theoretical strengths. Earlier rule-based phonology had to stipulate each co-occurrence pattern as a separate universal tendency. OT derives the co-occurrence from shared ranking: the property of being a 'simple syllable structure language' is not two independent facts (no codas AND no complex onsets) but one fact — the constraint ranking that produces one produces the other as a consequence of the same evaluation logic.