Chapter 7: Speech Perception: Auditory Processing and Language Sounds

Speech Perception: Auditory Processing and Language Sounds begins with an overview of acoustic principles, explaining how temporal sequences of air pressure fluctuations are transformed by the cochlea into neural representations similar to frequency-based spectra. A primary challenge in this area is that the physical stimulus—represented visually through three-dimensional sound spectrograms—does not contain discrete, easily identifiable boundaries between words or sounds. The text examines the "cocktail party phenomenon," illustrating how humans use binaural localization cues to isolate specific voices amidst background noise. Central to the discussion is the role of linguistic units, ranging from the binary distinctive features proposed by Jakobson and Halle to larger structures like phonemes, syllables, and grammatical constituents. Research using superimposed clicks demonstrates that listeners segment speech according to these deep grammatical rules rather than just physical pauses. The narrative compares various theoretical frameworks, including template-matching correlation and parallel filtering, but ultimately advocates for the "analysis-by-synthesis" model. This theory posits that listening is an active, constructive process where the brain generates internal hypotheses about incoming messages and matches them against the auditory input. This perspective incorporates the motor theory of speech perception, which suggests that we understand spoken language by referencing the internal articulatory movements required to produce it. The chapter concludes by examining how these constructive mechanisms account for phenomena such as auditory imagery, hallucinations, and the verbal-transformation effect, where repeated words appear to shift in meaning or sound, proving that speech perception is a complex interaction between passive preliminary processing and active mental synthesis.