Two studies investigated 3-to 5-year-olds' trust in a reliable informant when judging novel labels and novel plural and past tense forms. In Study 1, children (N=24) endorsed the names of new objects given by an informant who had earlier labeled familiar objects correctly over the names given by an informant who had labeled the same objects incorrectly. In Study 2, children (N = 24) endorsed novel names given by an informant who had earlier expressed the plural of familiar nouns correctly over one who had expressed the plural incorrectly. But children overwhelmingly endorsed the regular plural and past tense forms of new words provided by the formerly unreliable labeler (Study 1) or morphologist (Study 2) rather than irregular forms of those words provided by the formerly reliable informant.From a very early age, English-speaking children can easily generate the regular plural and past tense forms of novel nouns and verbs. For example, in a classic study, Berko (1958) showed 4-to 7-year-olds, a picture of a novel creature and referred to it as a wug. They were then shown two of the same kind of creature and were prompted to provide the plural: "Now there are two -." Most children responded by saying wugs. When they were shown a picture of a novel action called ricking (e.g., a man swinging a spiky ball around) and prompted to provide the past tense, most children responded by saying ricked. Although some words are easier for children to regularize than others (e.g., wugs is easier than nizzes, ricked is easier than motted), the basic phenomenon is quite robust: Children can create the regular plural or past tense forms of words they have never heard before.The vast majority of English nouns and verbs fit this regular paradigm (i.e., add -s for plural, add -ed for past tense). But 2% of noun plural types and 14% of the 1,000 most common verb past tense types are actually irregular (Marcus, 1996). For example, the plural of mouse is mice, not mouses; the past tense of go is went, not goed; and so on. This raises an important question: Given children's robust expectations about regularity, how do they deal with input that suggests that a particular word takes an irregular form?On the one hand, we know that they do not automatically dismiss this input because of course, they do learn that some words are irregular. On the other hand, we also know that they do not automatically accept any input that conflicts with their expectations. For Naigles, Gleitman, and Gleitman (1993) found that 3-and 4-year-olds who heard a familiar intransitive verb used transitively (e.g., "The lion comes the giraffe") tended to treat that unexpected usage as if it were an error (e.g., as if the speaker meant to say "The lion comes to the giraffe" or even "The lion combs the giraffe"). The issue of how children reconcile input about irregular nouns and verbs with their expectations about regularity is complicated further by the fact that they sometimes hear overregularized tokens. Their peers, for example, sometimes say breaked i...

No abstract

Overgeneralization is a phenomenon in which language learners—first, second, or additional—apply a rule or a pattern in a situation where it does not apply in the target language, resulting in a nonconventional form. It is most obvious in morphology, where it is often referred to as overregularization, but can also be found in the domains of argument structure, syntax, and phonology. Overgeneralization provides evidence of rule application, analogical reasoning, or use of a template, as opposed to rote memorization. As such, it demonstrates evidence of acquisition of a language as a system and of a language learner's ability to extract regularities and apply them.