ABSTRACT
The phenomenon of affect is widespread in the animal kingdom. It plays a role in the control of behavior at many levels, ranging from individual maintenance of internal physiological stability to the emergence of patterns of social interaction, as they are influenced by the temperament of individual members (Scherer, 1982). There is a wide spectrum of comparative variation on these themes, and, by rights, research on affect in animals should be a rich source of insights into its nature and functional significance. Yet, with a few exceptions, this has not been a productive approach (see Candland, 1977, p. 54). One obvious reason for the failure to exploit comparative approaches to research on affect is the virtual impossibility of identifying emotional states unequivocally, and the difficulty of distinguishing them in a reliable way from states that are primarily based not on activities of the autonomic nervous system but on cognitive activities of the brain.
