Adam Albright
Dept. of Linguistics, UCSC
The representation of regular morphological processes has been the subject of much controversy, particularly in the debate between connectionist models and the dual mechanism model of morphology. In this paper, I present a model of morphological rule induction that seeks to infer the productivity of morphological rules by comparing their reliability in different phonological environments. This model predicts that learners should notice environments that are especially reliable for the regular process, as well as for irregulars. I present results from a nonce-probe (``wug'') experiment in Italian confirming this prediction. Finally, I discuss previous, apparently contradictory results from English (Prasada & Pinker 1993), and argue that phonological environment effects can be detected even for regular English past tenses.
Paper
Related files:
Links