A problem every student of animal behaviour has to come to terms with is anthropomorphism, which means assuming that animals think or feel more or less as we do. Because of fierce criticism from behaviourists, explicit anthropomorphism is no longer scientifically respectable. John Kennedy's thesis is that anthropomorphism is not necessarily dead, but is lurking under different disguises. In fact, it still affects research, but is often unintended and therefore it goes unrecognized. He provides ample documentary evidence of the way researchers unconsciously slip into anthropomorphism. The book contains nineteen essays on behavioural concepts that have seldom been identified as anthropomorphic, but in fact bear that connotation and lead to mistakes. Some of these, such as search images in birds and the learning of grammatical language by apes, have been seen as errors after a time. A greater number, such as efference copy, goal-directedness, cognition, and suffering in animals, are still current though not yet regarded as erroneous. We can hardly hope to cure ourselves altogether of thinking anthropomorphically, and it can be very useful as a metaphor.

In this 1992 book, John Kennedy's point is that explicit anthropomorphism was well-nigh killed by fierce criticism from the radical Behaviourists, but that we have to recognize that today there is a new anthropomorphism which is much harder to avoid because it is unintended and largely unconscious. For that reason even those who if they were asked would firmly reject anthropomorphism nevertheless unwittingly slip into it from time to time. This book contains nineteen essays on behavioural concepts which have seldom been identified as anthropomorphic but in fact bear that connotation and lead to mistakes. Some of these, such as search images in birds and the learning of grammatical language by apes, have been seen through as errors after a time. A greater number, such as efference copy, goal-directedness, cognition and suffering in animals, are still current though not yet regarded as erroneous. The final chapter outlines things we can do to minimise the damage it does to the causal analysis of animal behaviour.