ABSTRACT
The discovery that the young child is able to handle size relations fairly efficiently naturally leads one to speculate whether or not he can also combine these relations in an inference. No one disputes the discovery that young children generally fail and older children succeed with the traditional transitivity problems of the type administered by Jean Piaget and by J. Smedslund, What is at stake is what these failures and successes mean. The Kendlers, whose background is that of the American school of stimulus-response learning theory and very different from Piaget's, draw conclusions about inferential abilities which are at least as pessimistic as his. The child has not only to make the inference, but also to get hold of the information on which the inference must be based. It would certainly be very interesting to know how children cope with an active version of the inference problem.
