Vocalizations in the cat: behavioral methodology and spectrographic analysis

Abstract

Attempts to understand the neural mechanisms underlying mammalian vocal behaviors, including speech, require study of the neural activity and anatomy of vocalization-controlling brain structures. Such studies necessitate the application of invasive neurobiological techniques in animal models. In the current study, cats are used in the development of an animal model of vocal tract control. The animals are instrumentally conditioned to vocalize for food reward. Acquisition of this task can occur within a few minutes, although additional training generally is required to solidly establish the behavior and to train subjects to produce consistently high rates of vocalization for prolonged periods of time. Following training, animals can generally sustain a rate of two calls per minute for a period of over two hours. Optimal task performance is partly dependent on motivation level. Although there is considerable variation between animals, the vocalizations produced have an average duration of 600 ms and a fundamental frequency of around 500 Hz. In addition, during a typical vocalization, there are dynamic variations of about 150 Hz for fundamental frequency and 17 dB for sound intensity. These variations provide opportunities for relating neural and muscular activity to different aspects of the vocal behavior they control. Based on a number of considerations, the model and techniques discussed here probably are most applicable to studying the neurobiology of sub-cortical nuclei subserving vocal control. Similar mechanisms might well be present in other species, including humans. Thus, data obtained from study of this model may be applicable to understanding the processes underlying vocal tract control during human speech.