SENSITIVITY OF CHILDRENS INFLECTION TO GRAMMATICAL STRUCTURE

What is the input to the mental system that computes inflected forms l ike walked, came, dogs, and men? Recent connectionist models feed a wo rd's phonological features into a single network, allowing it to gener alize both regular and irregular phonological patterns, like stop-stop ped, step-stepped and fling-flung, cling-clung. But for adults, phonol ogical input is insufficient: verbs derived from nouns like ring the c ity always have regular past tense forms (ringed), even if they are ph onologically identical to irregular verbs (ring the bell. Similarly, n ouns based on names, like two Mickey Mouses, and compounds based on po ssessing rather than being their root morpheme, such as two saber-toot hs, take regular plurals, even when they are homophonous with irregula r nouns like mice and teeth. In four experiments, testing 70 three- to ten-year-old children, we found that children are sensitive to such n onphonological information: they were more likely to produce regular i nflected forms for forms like to ring ('to put a ring on') and snaggle tooth (a kind of animal doll with big teeth) than for their homophonou s irregular counterparts, even when these counterparts were also exten ded in meaning. Children's inflectional systems thus seem to be like a dults': irregular forms are tied, to the lexicon but regular forms are computed by a default rule, and words are represented as morphologica l tree structures reflecting their derivation from basic-words are roo ts. Such structures, which determine how novel complex words are deriv ed and interpreted, also govern whether words with irregular sound pat terns will be regularized: a word can be irregular only if its structu re contains an irregular root in, 'head' position, allowing the lexica lly stored irregular information to percolate up to apply to the word as a whole. In all other cases, the inflected form is computed by a de fault regular rule. This proposal fits the facts better than alternati ves' appealing to ambiguity reduction or semantic similarity to a word 's central sense. The results, together with an analysis of adult spee ch to children, suggest that morphological structure and a distinction between mechanisms for regular and irregular inflection may be inhere nt to the design of children's language systems.