Bob Kentridge 1995

Comparative Psychology: Lecture 2.

Why reflect on history?

In this lecture I am, once again, going to give you a taste of the historical background behind a major part of comparative psychology. Before starting this lecture proper I would just like to reflect on why I'm doing this rather than ploughing straight into a description of classical conditioning, which is what I hope we will reach by the end of today. Later in the course we will look at both classical and operant conditioning. The theories and methods associated with these phenomena are quite different, yet on closer examination they share many features in common. If we do not understand the context in which these two approaches developed the decisions made by the scientists who worked on them may appear rather arbitrary. In the end it is understanding these learning phenomena rather than their histories which is most important, but by hearing about their histories I hope it will be easier to understand why they developed in the ways they did. I also hope that some of the story of their development will make it easier for you to remember their salient features.

Reflexes and the physiology of nervous activity.

In last weeks lecture I described some early work in comparative psychology. One of the main lessons both we and 19th century psychologists learned from this work was the need for objective non-anthropomorphic explanations of animal behaviour. In later lectures I will describe research that developed from this realisation. First, however, I will discuss work derived from a discipline which already took the objective study of animals very seriously - physiology.

Descartes and the concept of the reflex.

We can trace the origins of physiological explanations of behaviour back to the French philosopher Rene Descartes' "Discourse on Method" (1637). In this book Descartes asserted that the actions of animals (but not man) could be explained entirely in terms of physical principles. Having taken this position he then went on the investigate what functions the various bodily organs might serve in the operation of animals, largely through dissection studies. Descartes was particularly scathing about the quality of medical knowledge in the seventeenth century, most of which was not based on empirical evidence and called for public support of experimental physiological research. Although Descartes held very firmly to the view that man, but not animals, had a soul, and that therefore the behaviour of animals, but not man, could be explained mechanistically, he did suggest that involuntary behaviours in both man and animals may be explained on a common basis. Descartes was inspired by animated statues in the palace at St.Germain which were controlled hydraulically. For example, a statue of the goddess Diana bathing naked would retreat into some rosebushes when an approaching art-lover trod on a panel concealed in the pathway. Pressure on this panel opened a valve which released a flow of water which caused the statue to move. Descartes suggested that 'animal spirits' flowing through the nerves of animals or humans served a similar function in automatic behavioural responses in man and animals or reflexes. The term 'reflex' is derived from the notion that the flow of animal spirits produced by a stimulus is somehow reflected by the brain into an outgoing flow which eventually produces some behaviour. Unfortunately, although the theory contains some grains of truth (at least insofar as a reflexes are associated with quite direct actions of incoming sensory nerves onto nerves projecting to some sort of effectors), the assumptions upon which this theory were based had been shown to be false experimentally within Descartes' lifetime - nerves are not hollow, and the contraction and expansion of muscles is not achieved through inflating or emptying them of some fluid. Nevertheless, Descartes' concept of the reflex and of the explicability of human and animal behaviour mechanistically, although not widely accepted at the time, shaped much following physiological research.

The electrical nature of nervous activity.

Acceptance of the importance of reflexes was limited while the actions of nerves were still not understood. In the interim a wide range of behaviours were suggested to be reflexive in action, for example, digestion, coughing, sneezing, pupilliary reaction to light and so on. In addition, it was suggested by La Mettrie in his "Man a Machine" (1748), although it offended many deeply, that the behaviour man as well as that of animals could be explained entirely mechanistically. Nevertheless, it was hard to take these ideas seriously while there was no sensible mechanistic explanation of nervous action. The breakthrough came with flood of late eighteenth century work on electricity, for example Benjamin Franklin's accounts of his experiments with lightening published in 1751. Demonstrations of the effects of electricity, produced by generators of static electric charges, were also popular theatrical entertainments at this time. It had been suggested earlier in the century that electricity might form the basis of Descartes' elusive 'animal spirits', however, the absence of appropriate insulation in the nervous system seemed to rule the idea out. This all changed with Luigi Galvani's experiments with frogs and static electricity. He claimed to demonstrate that electrical energy was generated in the nervous systems of animals. As his experimental results did not justify his conclusions (something still not uncommon) a period of controversy during which many more experiments carried out in Italy by Allessandro Volta, Galvani's nephew Giovanni Aldini and others lead to much greater understanding of electricity and it's role in conducting signal in the nervous system (although precise understanding of this would not happen until the mid-twentieth century). Precise understanding was not, however, necessary for the theory of reflexes to be much more attractive now that a reasonable mechanism of nervous signal conduction had been suggested. The focus of work on neurophysiology and reflexes then moved to Germany, culminating in the measurement of the speed of nervous signal conduction in reflexes by Herman von Helmholtz (whose work on electricity also made an impact in physics).

Inhibition and the problem of spontaneous activity.

The solid conceptual status of the reflex as a neurally mediated automatic response to a stimulus now inspired some German scientists to suggest that all behaviour was in fact automatic - spontaneous activity was impossible - all behaviour was the result of reflexes, however, some reflexes were clearly simpler than others. The major problem facing this argument was evidence (initially from experiments on brainless and sometimes legless frogs) that should one reflex fail to be effective for some reason the stimulus inducing it would, after some time, begin to elicit a different behaviour. This is very difficult to explain if all nerves can do is excite activity in muscles, glands or other nerves. Luckily, the ability of the vagus nerve to decrease heart-rate - an inhibitory action, had recently been discovered. The validity of the suggestion that all behaviour was automatic therefore rested on discovering whether (and by what means) reflexes could be inhibited. A Russian, Ivan Sechenov, was the first person to appreciate the significance of inhibition and reflex action. During work in Germany and France Sechenov showed that by placing slat crystals in certain parts of a frogs brain he could reversibly inhibit its leg-withdrawal reflex. He returned to St. Petersburg and in 1863 published a monograph, 'Reflexes and the Brain', describing this work. Sechenov's description of reflex had begun to diverge from Descartes' notion of simple fixed 'reflections' of stimuli. First, Sechenov suggested that the strength of stimuli and o the responses they elicited need not be similar - very weak stimuli might trigger quite intense reactions. Second, Sechenov suggested that reflexes are ubiquitous and malleable, for example, he suggested that, as it was his habit to think of politics before going to bed each night it might happen that if were to lie down in the daytime the properties of his bedroom might elicit thoughts of politics in him. Sechenov felt that inhibition played a significant role in both these extensions of Descartes' concept of the reflex. Having demonstrated the existence of centrally mediated inhibition of reflexes, however, Sechenov did not go on to test these later inferences, perhaps because of the political and social conditions in Russia towards the end of the nineteenth century. Nevertheless, Russian experimental physiology thrived. Sergei Botkin, who had accompanied Sechenov on some of his studies in Germany, became Professor of Clinical Medicine at the Military- Medial Academy in Moscow where he maintained an animal laboratory for the experimental study of physiology. Botkin, however, was too busy with teaching and administration to run his small laboratory so, in 1878, he appointed a highly recommended young physiology graduate from St. Petersburg, Ivan Pavlov to be its director.

Ivan Pavlov, his students, and Classical Conditioning.

Pavlov's early work, for which he was eventually gained the directorship of the Oldenburgski Institute of Experimental Medicine in St. Petersburg - intended to be a Russian equivalent of the Pasteur Institute, along with other high posts, was not aimed at understanding the mental role of the reflex. Rather, it was in the physiology of the digestive system. He was interested in the way in which particular foodstuffs might elicit the release of gut chemicals particularly suited to digesting them and how these reflexes integrated with other muscular reflexes involved in digestion to produce a balanced and integrated process. In order to pursue this study Pavlov pioneered procedures in which careful surgery was carried out on animals, usually dogs, from which they recovered to lead near-normal lives (at least physiologically normal ones) in contrast to the more extreme acute studies on frogs which had typified experimental physiology up to that point. Two techniques Pavlov perfected were first the insertion of tubes into an animals throat so that, when the experimenter desired, food could be eaten by the animal without reaching its stomach, and second, the insertion of a tube into the stomach so that gastric secretions could be collected for subsequent analysis. This second tube had an important intellectual and financial role in Pavlov's laboratory. Financially it was important because drinking the gastric juices of animals had become a popular cure for stomach ailments around St. Petersburg and Pavlov's laboratory was able to supply large quantities of this hard to obtain commodity. Intellectually the gastric tube was important because it allowed people to observe that Pavlov's dogs would start secreting gastric juices not just when food was present in their mouths, but at the mere sight of food in the test room, or indeed at the sight of their regular feeder. This secretion, elicited by psychological expectations of food rather than food itself, became termed 'psychic secretion' and was assumed to be quite different from food-induced gastric secretions. Although the phenomenon was observed regularly in Pavlov's laboratory it was basically disregarded until nearly the end of the century - 1897. At this time Pavlov, now in his 50's, realised that physiological methods could be used to study psychological phenomena and that these phenomena must be described and explained in physiological terms if they are to be understood. He had, to some extent, been anticipated by some of his student who had already begun to work on psychic secretions. Stefan Wolfsohn had discovered that the amount and nature of saliva produced by different foods varies according to the type of food. The same was true of sights of different foods. Another student, Anton Snarsky, went on to show that apparently arbitrary signals could elicit the same effects. Dilute acids produce a copious amount of a particular type saliva. Snarsky found that if he coloured this acid solution black then soon the sight of any black liquid would induce large amounts of this sort of salivation. This is probably the first recorded attempt intentional classical-conditioning experiment; all the crucial features are present - a pre-existing reflex and an arbitrary stimulus which eventually becomes capable of eliciting the same response produced in that reflex. Work in Pavlov's laboratory was now firmly directed at understanding classical-conditioning and related phenomena. Pavlov strongly rejected mentalist explanations of conditioning couched in terms of an animals 'expectations' or 'desires' , in the process falling out with Snarsky, much of his work was concerned with pinning down the neural basis of classical-conditioning - a task made somewhat more difficult by the discovery by two British workers, Starling and Bayliss, who showed that food- related information may be transmitted hormonally rather than neurally. At this point it is probably sensible to leave history and turn to a description of the basic concepts of classical-conditioning before going on to describe related phenomena and discuss what exactly an animal might be learning during classical-conditioning.

Basic concepts in Classical Conditioning.

The archetypal classical-conditioning experiment is one in which dogs learn to salivate at the toll of a bell which has previously sounded when they were presented with food. In this procedure, before any attempt has been made to pair the bell with food presentation, that is, before any conditioning has taken place, the only reflex present is salivation in response to the presentation of food. The presentation of food is therefore the unconditioned stimulus (abbreviated US or UCS) and the reflexive salivation produced in response to this stimulus if the unconditioned response (UR or UCR). The US-UR reflex should be innate, rather than learned. The other stimulus used in the experiment, the sound of the bell, is initially quite neutral in its effects on the dog, at least insofar as the response under investigation is concerned - it does not provoke salivation on its own and it does not suppress it if the dog happens to be salivating when the bell is tolled. As the procedure aims to condition the dog to respond to the bell the sound of the bell is referred to as the conditioned stimulus (CS), even though, initially, no conditioning has yet taken place. In the course of the experiment the animal experiences the US and CS at the same time (approximately), eventually the animal will produce a response to the CS alone. This response is called the conditioned response (CR). Although it may be indistinguishable from the UR initially it is not uncommon for subtle differences to develop between the UR and CR, for example, the quantity and chemical composition of saliva produced in response to the CS alone after conditioning may differ from that elicited by the US (hence the need for distinct terms for UR and CR). These concepts can be summarised as follows:


Again, the history is from Bob Boakes' book The basic concepts of classical conditioning should be spelled out in any general psychology textbook or textbook on animal learning.