Children's judgments of regular and irregular novel past-tense forms: new data on the English past-tense debate

Is language governed by formal rules or by analogy to stored exemplars? The acquisition of the English past tense has long played a central role in this debate. In the present study, children rated the acceptability of a regular and an irregular past-tense form of each of 40 novel verbs (e.g., fleeped, flept) using a 5-point scale. The novel verbs were chosen to vary continuously along the orthogonal dimensions of (a) similarity to existing regular forms and (b) similarity to existing irregular forms. A developmental progression was observed whereby the acceptability of novel regulars was shown to increase as a function of similarity to existing regulars, with the magnitude of this effect increasing with age. The acceptability of novel irregulars was shown to increase as a function of similarity to existing irregulars, with no developmental changes observed. These findings are discussed in the light of 3 current models of past-tense acquisition: the single-route model (e.g., Bybee & Moder, 1983), the dual-route model (e.g., Prasada & Pinker, 1993), and the multiple-rules model (e.g., Albright & Hayes, 2003).