Intelligibility of one‐ and two‐syllable words in speech‐modulated noise

Abstract

A classical finding in the auditory word recognition literature is that two‐syllable words presented in white noise are identified more accurately than one‐syllable words [see Licklider and Miller in Handbook of Experimental Psychology, edited by S. S. Stevens (1951)]. An explanation commonly given for this phenomenon is that two‐syllable words have greater redundancy. However, the effect may have been due to inadequate methods for matching the S/N ratio of the two word sets. We randomly selected one‐ and two‐syllable words of low, medium, and high frequency of occurrence in English. We used speech‐modulated noise in which the amplitude of the noise was correlated with the amplitude of the speech signal. Words were presented at a 0 dB S/N ratio to 24 listeners for identification. There was no significant difference in proportion correct between one‐ and two‐syllable words (0.42 and 0.40, respectively). However, there was a reliable frequency effect; more frequent words were more often identified correctly.