How acoustically reduced forms activate the lexicon : evidence from eye-tracking

Abstract

Most research on spoken word comprehension has focused on carefully articulated speech that is read aloud by selected speakers (Cutler, 1998). But the type of speech we most often encounter is spontaneous speech, in which no attention is paid to careful pronunciation. The production of a word shorter than its citation form is called reduction, which is highly frequent in casual speech (Ernestus, 2000; Johnson, 2004). The challenge for models of word comprehension is to explain how listeners recognize reduced forms such as [pjutǝr] which deviate drastically from their canonical counterpart [kɔmpjutǝr] 'computer'.