A reply to Professor Guthrie.

Abstract

I find that I agree to a surprising extent with Guthrie's strictures on my theory of learning. But I still do have a few difficulties. First, for a couple of general ones: (1) I think it is rather unfair for Guthrie to appropriate the term 'associative learning' and to identify it with his own particular brand of 'conditioning' and 'stimulus-responseism.' I, also, like the term 'associative learning.' In fact, I should like to use it to cover the first five of my seven varieties of learning. (2) I do not see how his conditioned-response-ism will ever explain 'latent learning.' For latent learning provides a set-up in which the learning takes place even when the correct responses are not made appreciably oftener by the animal during the learning period itself than the incorrect responses. But I suspect that perhaps Guthrie does not really believe our supposed experimental 'facts' concerning latent learning. But to come now, secondly, to the matter of the stringpulling situation itself. (3) My feeling from watching the animals (this, of course, is pretty awful anthropomorphism) is that the first time they pull the string in, it is because they have already learned— i.e., that they have then and there just 'got,' at least tentatively, the required 'expectation' and that it is because they have got it that they then pull the string in—not that they pull the string in and then get the expectation. 1 E . R. Guthrie, Tolman on associative learning, PSYCHOL. REV., 1937,44,525-528. J E. C. Tolman, The acquisition of string-pulling by rats—conditioned response or sign-Gestalt?, PSYCHOL. REV., 1937, 44, 2O3f. 1 Let me emphasize again and again that an 'expectation' does not require words nor consciousness—that it is just a 'set' for a certain environmental object-sequence.