| ROA: | 1046 |
| Title: | Modelling the formation of phonotactic restrictions across the mental lexicon |
| Authors: | Silke Hamann, Diana Apoussidou, Paul Boersma |
| Comment: | CLS 45 |
| Length: | 13 |
| Abstract: | The goal of this paper is to model the experimental data found by Hamann & Ernestus who showed that adults learnt a phonotactic restriction better when presented with many word types than when presented with a smaller set of word types that were repeated. The combined result of virtually learning the connections between four levels of representation (meaning, underlying form, surface form and auditory form) with a single learning algorithm is a sensitivity to type frequency: virtual learners trained on 20 word types (presented once) ended up with stronger phonotactic constraints than those trained on 10 word types (presented twice), predicting that the former group is better at generalizing the learned phonotactic restrictions. |
| Type: | Paper/tech report |
| Area/Keywords: | Phonology,Computation,Learnability,Language Acquisition |
| Article: | Version 1
|
