ABSTRACT
Our attempts to teach counting to the chimpanzees have enjoyed notably Httle success-so little that even the elementary stages of counting now 100m as a far greater challenge than the elementary stages of language. Of course, this view is not incompatible with the development of the child-it displays a considerable grasp of language before showing any numerical skills. However, the view is contradicted by at least one reported success of teaching chimpanzees to count (Ferster, 1964). Unfortunately, this study is marred by the absence of a test requiring the animals to respond to magnitudes different from those used in training, and the number of magnitudes used in training were "smali," easily within the chimpanzee's memorization capacity. Although it does not do simply to darken this reported success by placing it in the shadow of our failure, I cannot but remain skeptical until shown the results of a transfer test. 1
ONE·TO·ONE CORRESPONDENCE
Aletha Solter (1975) has worked out a training program to teach counting to very young children which begins by teaching one·to-one correspondence and progresses from there through several intermediate steps to the cardinality rule, which makes counting possible. In attempting to teach one-to-one correspondence to the chimps, we used Solter's approach. A small set of wooden cubes was the counters, and a set of blocks was one of several target sets that it was hoped the chirnps could ultimately be taught to count. By the usual combination of modeling and passive guidance we attempted to teach Sarall, Peony and Elizabeth to put one counter on each block, in effect to carry out this rule: Leave no countee without a counter and do not put more than one counter on any countee. This simple rule can be given various interpretations, such as that of Orthodox ludaism: every man shall wear a hat but no man shall wear more than one at a time.
