Optimality theory offers a brilliant explanation of children's consonant harmony by setting “REPEAT” as a structural constraint and determining an appropriate ranking order with relation to such faithfulness constraints as Faith(Dor) and Faith(Cor). No matter how delicate it may be, however, it does not provide any plausible reason why their ranking order should be so in children's phonological development. This paper attempts to give an account of it by referring to children's preference for a more marked prosodic structure and also to the characteristics of an acoustic feature ‘grave’.
2. Children's canonical prosodic structure
3. The structural unmarkedness of binary branching
4. The consonant harmony