A replication of a test of the metrical segmentation strategy in spoken word recognition

Abstract

Norris et al. (1995) tested the Metrical Segmentation Strategy (MSS; Cutler and Norris, 1988) as part of the spoken-word recognition model Shortlist. We replicate their study in a different dialect of English, with a different population and items. Norris et al. used a word-spotting task, in which listeners had to spot words within speech (e.g., "stamp" in [stæmpɪdʒ]). Target words were CVCC like "champ" or CVC like "done," and were followed by a full vowel ("champ" in/tʃæmpoʊʃ/, "done" in /dʌnʤeɪb/) or a reduced vowel ("champ" in /tʃæmpəʃ/, "done" in /dʌnʤəb/). The original study found different behavior for CVCC versus CVC targets, with the results suggesting that listeners hypothesize a word onset at the start of a full-vowel strong syllable (the MSS). Doing so makes it harder to detect "champ" when it is followed by a full vowel than a weak vowel because the full vowel leads the listener to think the /p/ is the onset of the following word, while the following vowel has little influence for "done," where the equivalent consonant is not part of the word. The results for the current study (underway) will show whether these effects generalize across English dialects, listener populations, and words.