Spoken word recognition involves selecting one word out of many competing candidate word-forms. However, there is disagreement as to what candidates actually compete with each other given a particular signal, though similarity is commonly invoked. We investigate whether the word similarity literature, drawing from explicit similarity judgments of word forms, can speak to this current state of confusion. Many word similarity models emphasize the primacy of featural overlap. With a single set of stimuli, we find that featural overlap and overlap position influence a non-time-pressured lexical learning task as well as explicit similarity judgments. The implications for appropriate metrics of similarity in word recognition and models of word similarity are discussed.
Cite as: Creel, S.C., Dahan, D., Swingley, D. (2006) Effects of featural similarity and overlap position on lexical confusions and overt similarity judgments. Proc. Interspeech 2006, paper 1298-Wed1A3O.1, doi: 10.21437/Interspeech.2006-301
@inproceedings{creel06_interspeech,
author={Sarah C. Creel and Delphine Dahan and Daniel Swingley},
title={{Effects of featural similarity and overlap position on lexical confusions and overt similarity judgments}},
year=2006,
booktitle={Proc. Interspeech 2006},
pages={paper 1298-Wed1A3O.1},
doi={10.21437/Interspeech.2006-301}
}