Animal Cognition
Home Page




Stimulus Control

Space & Time


Operant Conditioning


General Links

Dr. Robert Cook's Home Page


Selective Attention and Related
Cognitive Processes in Pigeons


Donald A. Riley & H.L. Roitblat

University of California, Berkeley


It is common observation that when we notice, think about, or observe some stimulus, it may be difficult to notice, think about or observe other stimuli. In other words, perception and other cognitive processes are often selective. One explanation for these observations is that in order to notice some stimulus, we must attend to it and we can only attend a limited number of stimuli at a time. Often, other explanations are available. For example, it may be difficult to notice two things because they are spatially separated, and we can only look in one direction at a time. Such peripheral explanations are usually suggested as alternatives to an attentional explanation. Attention can, then, be viewed as a central adaptation to information overload. There is more information present in the stimulus array than is possible to process so it is necessary to selectively attend to some stimuli at the expense of others.

Selective attention has also been proposed as an explanation for the performance of animals in discrimination problems (e.g., Sutherland and Mackintosh, 1971; Riley and Leith, 1976). Every time a selective interpretation has been proposed by some investigator, however, other experimenters have been quick to demonstrate other more peripheral and perhaps simpler mechanisms to account for the apparent attentional effects. After almost 50 years it is still a matter of controversy whether an attentional hypothesis is at all necessary to explain animal discrimination performance (Mackintosh, 1975).

This lack of resolution may be true, in part, because experimenters attempting to investigate attention in animals have usually failed to structure their experiments to provide any overload of the animal's ability to take in all the available information within the allowed time. It is our view that when this condition is met, that is, when the animal's information processing ability or capacity is taxed, the likelihood of demonstrating selective attention will increase. The modern study of attention in humans, beginning with Broadbent's book Perception and Communication (1958) has been sensitive to this position. Broadbent (1961) also suggested that such a strategy would be useful in studying animal attention.