Bob Kentridge 1995

Comparative Psychology: Lecture 1.

General Overview of the aims of comparative psychology.

One of the aims of comparative psychology is to use insights gained from the study of psychological processes in different species of animals to add to our understanding of human psychology. Any way of achieving this aim must depend to some extent on understanding the evolutionary relationship between animals and man. There are, however, two broad ways of approaching this issue. One way of regarding this relationship is to think of man as having evolved the most sophisticated psychological machinery possible. The study of animal psychology can then allow us to uncover the simpler foundations of man's psychological processes which are often obscured by his 'higher' abilities in the direct study of human psychology. This is probably the assumption that is implicitly made when considering the results of most psychological research on animals. A more sophisticated and more biologically correct approach is to consider how the different evolutionary lineages and pressures of natural selection of different species give rise to differences in their psychological abilities. The overwhelming majority of work in comparative psychology has focused on learning. In addition to this specialisation of area there has also been an emphasis on two particular approaches to characterising learning - classical conditioning and instrumental learning (or operant conditioning). Finally, two species, the rat and the pigeon, have been studied far more than all the others put together, perhaps leading to the simplistic view of 'comparative' psychology I alluded to earlier. We can see how this situation came about by briefly studying the historical development of comparative psychology. As this historical survey approaches the present we will study contemporary theories of animal learning and consider their applicability to human psychology. Finally I will briefly discuss some aspects of the psychology of human memory and will consider whether animal work sheds any light on them.

An aside about evolution.

Although the extent to which comparative psychology pays heed to evolutionary theory varies widely we should, nevertheless, have a firm basic grasp of evolutionary concepts. This grounding need not go much further than basic general knowledge, it is, however, worth clarifying a couple of points at this stage. There is, as I mentioned, a tendency to view work on animal learning as a way of revealing simple mechanisms upon which our 'higher' cognitive faculties are based. There is a danger that this may go hand in hand with a simplistic conception of evolution in which there is a "great chain of being" with 'lower' organisms at its base, out of which develop more sophisticated organisms, like rats for example, leading finally to man at its pinnacle. This is, of course, wrong.

Insert a "Great chain of being" image here.

All of the species alive today have been evolving for just as long as each other. There is therefore little justification for asserting that any living organisms are 'higher' than others in an evolutionary sense. It is, however, quite reasonable to assume that the capabilities of organisms today arose out of the selective adaptation of their ancestors. So the 'basic' psychological processes we might hope discover about in comparative psychology are those of our evolutionary ancestors, not those of our so-called lower contemporaries. It is hard enough to discover the true physical structures of extinct organisms from their fossils, let alone their behavioural and psychological capabilities, so, although we may be interested in this sort of 'fossil psychology' we stand no chance of making progress by studying the remains of our ancestors directly.

Insert a "Tree of evolution" image here.

We can, however, make quite reasonable inferences using a 'taxonomic' approach and making some simple assumptions about evolutionary histories. If, rather than making the mistake of assuming a single linear evolutionary history, we think of an evolutionary tree it is reasonable to assume that more primitive traits will be exhibited by wide ranges of species - by tracing back to the earliest common branch point we can estimate the point in our evolutionary history that such traits must have appeared.

Insert a "Latest common ancestor" image here.

As I mentioned earlier, this is one approach to the interpretation of comparative data. The other is approach does not aim so much to discover the evolutionary history of our psychological powers as to discover how particular environmental pressures in evolution select for particular psychological or behavioural adaptations. This approach is much more oriented towards the study of learning and behaviour in its ecological setting, whereas the study of fundamental processes in learning ca proceed quite successfully in quite artificial laboratory settings. Having briefly outlined the framework in which comparative data should ideally be viewed (but often isn't!) lets us now begin our survey of the history of comparative psychology.

Early studies of the mental abilities of animals.

It is reasonable to begin our survey in the time of Darwin since it is really only with Darwin's 'Origin of Species' (1859), or, more pertinently, 'Descent of Man' (1871) that people began to admit that the study of animals could shed light on the abilities of man. George Romanes met Darwin three years after the publication of 'The Descent of Man' having begun to study the behaviour of animals with a view to understanding their mental capacities. Romanes did not carry out experiments on animals, but rather hoped to gain insights into mental evolution by classifying observations of the behaviours of different species made by himself and others and thereby deriving general principles of animal intelligence. This seems a perfectly sensible approach to comparative psychology until we examine the kind of observations Romanes was attempting to classify. We find, for example, that one of Romanes' observers noted that an earwig, which she called 'Tom', "crawls up a certain curtain every day with the apparent expectation of getting its breakfast' since her two younger sisters were in the habit of feeding it with sugar each morning. There are many more objective observations in Romanes book 'Animal Intelligence', together with a few more ridiculous ones (which he might have included because of the social status of his observers). Romanes went on to attempt a classification on the basis of conventional zoological taxonomy of species in his later books - essentially hoping to use the same methods that Darwin had used when for the evolution of form applied to psychological abilities. His first and most widely read book 'Animal Intelligence' was dominated by observation without analysis. It amply demonstrates the need for a systematic and objective approach to gathering data on animal behaviour. Romanes was by no means alone in his study of animal intelligence, it was a burning issue in the late 19th century. Before we start tracing the development of modern theories of learning, however, a few further examples illustrate the problems faced by early investigators. First, there is the issue of instinct. Douglas Spalding also began to take an interest in the animal behaviour at about the time of 'The Origin of Species'. He was, however, particularly drawn to the topic by the work of philosopher's who argued that the mind was a blank slate (tabla rasa) whose structure was formed by experience. John Stuart Mill, who was a contemporary of Darwin's asserted, for example, that even the ability to perceive distance or spatial relations was not present at birth, but rather had top be learned by experience. The philosophical position that our minds are formed by experience is called empiricism, however, debates over the validity of assertions like those of Mill did not invoke any experimental, or as we might say, empirical!, evidence, rather, they stuck to philosophical arguments. Spalding, however, decided to perform some experiments. He determined, by comparing the accuracy with which chicks pecked at grain when they had been raised for the first few day of their lives with their heads in a hood or free to see, that visual judgement of positions in space was innate. The first point to note about this is that Spalding had carried out a well controlled behavioural experiment. He went on to carry out many more such experiment, discovering which abilities were innate and which must be acquired in many different species of animals. We may argue that instinctive behaviours are not as psychologically interesting as those which must be learned in response to a changing world - this is certainly the way in which comparative psychology developed in the first half of the twentieth century. The study of instinct, however, demonstrates the care which must be taken before we can attribute any behaviour to learning. Take, for example, the behaviour of the digger wasp. Before laying an egg a female digger wasp will kill or paralyse a caterpillar, then dig a hole, into which she puts the caterpillar before laying a single egg on top of it. The caterpillar serves as a food source for the developing wasp larva. This is quite sophisticated behaviour in itself, what is even more remarkable is that the mother will restock the hole with more caterpillars according to the amount that the larva has eaten. The mother even maintains two or three young in different stages of development, feeding each according to their needs. This certainly appears to be very intelligent behaviour, it is even flexible - the wasp doesn't always feed her young the same number of caterpillars, but it is purely instinctive. If we wish to understand the mental capacities of an animal we must do more than simply study its behaviour, we must discover the origins of that behaviour. Another famous example from the early years of animal psychology emphasises just how difficult this can be. One consequence of the great interest in animal learning was that attempts were made to educate animals as one might educate a child. One of these attempts appeared to be spectacularly successful. At the turn of the century the German press reported that a retired schoolmaster, Herr von Osten, had succeeded in training his horse, known as 'Clever Hans' to answer questions on general knowledge, judge musical interval and perform simple arithmetic. Questions were communicated to Clever Hans in German and he responded by tapping one of his hooves (one tap for true, two for false, or tapping out the number in answer to arithmetic problems). Hans' performance was so remarkable that a committee of German Psychologists, Zoologists and circus animal trainers were sent to assess it. They found that Hans could answer questions even when von Osten was absent! At this point the case was handed over to a student at the Psychological Institute in Berlin, Oscar Pfungst, for more detailed investigation. Pfungst carried out the crucial experiment, he established Hans' performance depended somehow on the knowledge of his questioner. When Hans answered correctly when asked questions to which the questioner knew the answer, but his responses were random when the questioner did not know the answer (although, of course, other observers must knew the answer and so could assess Hans' answers). Having established this Pfungst then systematically removed various sources of information by which the questioner might be unintentionally signalling the answer to the horse. By this process of elimination he discovered that questioners would make very small upward head movements when Hans' hoof taps had reached the correct answer. During his training has must have learned that he would be rewarded if he stopped tapping at this point. Pfungst could in fact induce Hans to answer questions incorrectly by intentionally making this movement. If we consider these examples we can see that a taxonomy of behavioural observation cannot really shed light on the evolution of psychological faculties unless it is possible to establish whether a behaviour is learned or innate, and if it is learned exactly what is that is being learned. The outcome of this at the end of the nineteenth century was a movement in psychology to eliminate both anthropomorphic and mentalist explanations of behaviour and to develop objective methods of studying learning. We can see from the example of Clever Hans that not only must we establish whether behaviours are innate or acquired, but we also need to know exactly what is being acquired and how. In order to do this a number of highly controllable tasks were developed for the training and testing of animals as we shall see in weeks to come.


Most of the historical background was taken from Boakes, R.A. (1984) From Darwin to behaviourism: Psychology and the minds of animals. Cambridge: Cambridge University Press. This is an extremely readable enjoyable book. The digger wasp's behaviour is described in Warren, J.M. (1974) Learning in vertabrates. In D.A. Dewsbury and D.A. Rethlingschafer. Comparative Psychology: A modern survey. Tokyo: McGraw-Hill Kogakusha, which is odd since its an example of a non-learned behaviour in a non- vertebrate. Dewsbury & Rathlingschafer is a good source of serious comparative psychology - if you want to know whether a sea-anemone or a horseshoe-crab can be classically conditioned look here.