Language, Signs, and Cognition in the Chimpanzee

LANGUAGE, SIGNS, AND COGNITION IN THE CHIMPANZEE Roger S. Fouts & Roger L. Mellgren 1 Introduction: Learning and Biological Considerations. Being trained as an American psychologist in the recent past has usually meant that there is a strong environmentalist bias in one's intellectual history. By this we mean that the radical behaviorist philosophy of John Watson still permeates the training of many scientists interested in understanding the behavior of organisms. The essence of this approach is the viewpoint that a newborn baby can be taken from its natural mother and by manipulating the baby's environment we can produce a beggar or thief or bank president or any type of person we desire. Perhaps we have tempered our environmental approach in the recent past, but the strong influence of radical behaviorism is difficult to undermine. How can I prove that you cannot take my child and turn him or her into a murderer or an astronaut? If we say that this is not the case, then we argue from negative evidence. That is to say that the appropriate experiment has not been done, and so it is wrong to conclude that environmental influences cannot produce vastly different adults given the same genetic makeup. (To be sure, studies of identical twins, raised separately, have attempted to deal with these kinds of problems, but so many methodological problems exist with this kind of study that we do not feel compelled to draw any conclusions from them.) Each time we are able to change a person's behavior by manipulating the environment we may add further evidence to the importance of the concepts of radical behaviorism. * We wish to thank students and former students who aided in the collection of data and observation reported here: E. Sue Savage, Janet Kate, Joe Couch, George Kimball, and Shiela Thompson. Support for the research reported and preparation of this paper was provided by the National Institute of Mental Health (MH23511-02) Sign Language Studies 13 Consider for a moment the opposite situation. Suppose we manipulate the environment and fail to observe a change in behavior. What do we conclude? Hayes & Hayes (1951a,b,) attempted to teach their chimpanzee Vicki to talk. The results of their attempt were at best very limited: Vicki learned to say "mama", "papa", "cup", and "up", albeit with a heavy chimpanzee accent. From the standpoint of the radical behaviorist several alternative conclusions are possible. First, we might suppose that the Hayeses were not familiar enough with current behavioristic procedures of reinforcement, shaping, fading, etc., and that they were unsuccessful due to this inadequacy. Second, we might postulate that for some reason the chimpanzee is not capable of learning language. Of course, this conclusion is not justified because it depends on negative evidence. To conclude that the chimpanzee is incapable of language because it failed to speak in one or a number of experiments would be comparable to concluding that life on other planets does not exist simply because we didn't find life on the moon or on some other planet. A third possibility, not likely to be espoused by the radical behaviorist, is that perhaps a chimpanzee is capable of language behavior but not in a language that is isomorphic with spoken human language. This more enlightened view might be categorized as a "modern behaviorist" position. Most reasonable linguists would not require that the subject be able to speak in order to determine if the subject possessed language or was capable of language. The strict or radical behaviorist makes the mistake of assuming "equivalence of associability" (Seligman 1970). This position assumes that general laws govern learning processes and the particular response we choose to measure may be arbitrarily chosen. The assumption is that by manipulating the subject's environment and observing changes in the arbitrarily chosen response, laws with great generality for all responses and organisms will be discovered. Of course, the notion that responses are arbitrarily chosen is clearly not true. A pigeon pecking at a lighted key or a rat running down an alley are not arbitrary organisms and responses in any sense of the word. Both organisms and their responses are used because they are reliable and produce data which...