Number Representation In Animals

Why might nonhuman animals need to represent number? In nature, many stimulus attributes correlate with number. When a set of items is simultaneously present, the total surface area, brightness, density, and cumulative perimeter of the objects are available as cues to describe the set. If the to-be-counted items are presented sequentially, then duration, rate, or rhythmic patterns may covary with number. Similarly, a food patch with a greater number of discrete items will have higher hedonic value, as measured by the volume of food ingested, the feeling of fullness, or perhaps the amount of saliva that the eating animal produces. When animals are required to discriminate endogenous stimuli, such as the number of their own responses on an operandum, the total effort expended or the duration of responding is easily confounded with number. Perhaps animals do not need to represent number to maximize food intake or function adaptively in other domains. Animals may only represent number as a last-resort strategy when there is no other way to solve a given problem (Davis and Perusse, 1988).

An alternative possibility is that animals routinely represent number and that such processing is a fundamental part of the perceptual and cognitive systems of most vertebrate taxa (Capaldi, 1993; Gallistel, 1990). In foraging decisions, many animals behave as if they calculate the rate of return in a given food patch in order to decide whether to stay or go (Krebs and Davies, 1993). In an especially compelling example, Harper (1982) dispersed food at two different rates and showed that ducks sorted themselves into groups with numbers proportional to the amount of food at each location. Importantly, sorting often occurred before every duck had had the opportunity to sample the food. The social arena is another domain where numerical competence may be useful. Wilson et al. (2001) showed that the likelihood that male chimpanzees would respond aggressively to a simulated intruder's vocalization depended on the number of potential allies present. This suggests that males determine the number of conspecifics in their party to calculate their chances at successfully repelling an intruder.

Gallistel and Gelman (1992) made the important distinction between numerical categories and numerical concepts. An organism can be said to have a numerical category insofar as it can behaviorally group all exemplars of a given numerosity regardless of irrelevant dimensions such as surface area, density, and so on. In contrast, the richness of an organism's numerical concept is dependent upon the types of operations that it can perform on representations of number. In this section, we first review the data obtained from paradigms designed to test whether animals can form purely numerical categories; subsequently, we describe the evidence that animals can manipulate these representations of number in meaningful ways and thus have numerical concepts.

0 0

Post a comment