Speech perception by the chinchilla: Identification functions for synthetic VOT stimuli

In an attempt to clearly differentiate perceptual effects that are attributable to “auditory” and “phonetic” levels of processing in speech perception we have undertaken a series of experiments with animal listeners. Four chinchillas (Chinchilla laniger) were trained to respond differently to the “endpoints” of a synthetic alveolar speech continuum (0 ms VOT and +80 ms VOT) and were then tested in a generalization paradigm with the VOT stimuli between these endpoints. The resulting identification functions were nearly identical to those obtained with adult English-speaking listeners. To test the generality of this agreement, the animals were then tested with synthetic stimuli that had labial and velar places of articulation. As a whole, the functions produced by the two species were very similar; the same relative locations of the phonetic boundaries, with lowest VOT boundaries for labial stimuli and highest for velar stimuli, were obtained for each animal and human subject. No significant differences between species on the absolute values of the phonetic boundaries were obtained, but chinchillas produced identification functions that were slightly, but significantly, less steep. These results are discussed with regard to theories of speech perception, the evolution of a speech-sound repertoire, and current interpretations of the human infant's perceptual proclivities with regard to speech-sound perception.