Excerpted from Wolfgang's Kohler's book The Mentality of Apes (1925).
Thorndike tested large numbers of dogs and cats in order to see what there is in the wonder-stories that are told about these domestic pets. The result was very unfavourable to the animals, and Thorndike came to the conclusion that, so far from "reasoning," they do not even associate images with perception, as humans do, but remain limited chiefly to the experiential linking of mere "impulses" with perceptions. This investigation did what was necessary in a negative way at the time, but, as is now being shown (also in America), it went a little too far. The tests were based upon those animal stories, and consequently were made so difficult that the result was bound to fall out badly; under the influence of the animals failures in these tests, Thorndike then drew generalizations about their capacities, which do not follow from those difficult experiments. However stupid a dog may seem compared to a chimpanzee, we suggest that in such simple cases as have just been described, a closer investigation would be desirable.
Regarding their principle, I must make a further objection to Thorndikes experiments. They were designed as intelligence tests of the same type as our own (insight or not?), and ought therefore, to have conformed to the same general conditions, and, above all, to have been arranged so as to be completely visible to the animals. For if essential portions of the experimental apparatus cannot be seen by the animals, how can they use their intelligence faculties in tackling the situation? It is somewhat astonishing to find that (in Thorndikes experiment) cats and dogs were frequently placed in cages containing the extreme end only of one or the other mechanism, or allowing a view of ropes or other parts of the mechanism, but from which a survey over the whole arrangement was not possible. The task for the animal was to let itself out of the cage by pulling or pressing the accessible part of the mechanism; thenthe cage door would open of itself. Thorndike also gives an account of experiments in which the animals were let out of their cages if they scratched or licked themselves. He contrasts these experiments with those involving the employment of any mechanical contrivance, as the former apparently imply no direct connection between cause and effect; but the causation is far from apparent even in the mechanistic experiments.