Abstract

Peter Ladefoged’s outstanding qualities include meticulous description of differences between ‘‘the same’’ phonetic categories in different languages, and a tendency to explore new research areas, not least by investigating things that others judge better left alone. This paper pays tribute to these qualities by discussing the phonetic category as a theoretical construct in speech perception. By demonstrating that F1 of a preceding phrase influences whether a word is heard as ‘‘bet’’ or ‘‘bit,’’ Ladefoged and Broadbent (1957) stressed that phonetic categorization is plastic: sense depends on context. Many other contextual influences on sound categorizaton have been identified, yet are poorly integrated into theory, especially when they extend over long durations, or indicate something other than lexical contrast. I argue from (human) perceptual and (animal) neurophysiological and behavioral data that category identification depends on context and perceived relevance—hence on potential meaning. Human data confirm the centrality of context and meaning to categorization. Animal data show that a sound which is behaviorally significant produces long-lasting changes to response characteristics of single neurons in the primary auditory cortex, and probably earlier. Such data encourage radical re-evaluation of how sensory information is used in understanding speech. [Funded by the Leverhulme Trust.]

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call