A Comparative-Ecological Approach to the Study of Learning

Abstract

The purpose of this chapter is to discuss the effects of a parameter which is present in every learning experiment, but about which we know relatively lit-tle-the species being tested. The effects of the species variable have been studied relatively infrequently by animal learning psychologists for at least three reasons. (1) Psychologists have tended to assume that a few general processes will account entirely for most, if not all, of the important phenomena of animal learning in a broad range of species. This assumption has been criticized fre-quently in recent years (e.g., Seligman, 1970; Seligman & Hager, 1972), but still dominates the field of animal learning. (2) There are serious conceptual problems in attempting to interpret species differences in performance in a learn-ing experiment (Bitterman, 1960, 1965). The basic problem is that if one species performs better than another on a learning task, this does not necessarily repre-sent a species difference in learning ability. It may represent an inadequacy in the design of the experiment for one (or both) species. (3) Psychologists have been without a systematic, biologically sensible method of selecting species for com-parative study. As Hodos and Campbell (1969) have pointed out, the psychologi-cal view of species has been dominated by the " scala naturae " concept. The scala naturae concept holds that there is a continuous and linear sequence of species, beginning with the most primitive and culminating in the most com-plex, man. As Mayr (1982) has discussed, this idea became increasingly untena-ble as biologists became increasingly familiar with the tremendous diversity of life. Mayr (1982) claims that the scala naturae concept was dealt its final death blow in 1812, when Cuvier "asserted emphatically that there are four distinct phyla of animals... and that there was absolutely no connection among them" (pp. 201-202). In any event, there can be no doubt that in clinging to the scala