Living things seek to understand their relationship with their environment

"In animal and human behaviour, trials are not chosen randomly."
  Keith Oatley

"Learning takes place not simply as the emission of a new response when an error occurs. Rather it involves understanding the particular structure of the mistake and deciding what to do next."
  Keith Oatley

A living thing apprehends its environment by detecting data that appear relevant to its behaviour pattern and interpreting them in the light of its mental model of its relationship with it, within as wide a context as is necessary for the purposes of determining the appropriate adaptive response. This means that it seeks to establish their meaning or to understand them.

This thesis is, of course, irreconcilable with current wisdom on the subject. Behavioural psychologists fall into a number of schools, but for a long time, the dominant one was that of Behaviourism. It is associated with the work of Edward Thorndyke and E. R. Guthrie, also with John Broadus Watson and particularly, and more recently, with B. F. Skinner. Though many behavioural psychologists, such as the members of the Gestalt school, and more recently of the 'cognitive' or 'mentalist' school that we associate with the name of R. W. Sperry, did not accept Behaviourism, it remains true that it still underlies much of the thinking in this field. One reason is that it is the view that best fits in with the paradigm of science that rationalises the methodology that experimental psychologists and ethologists must use, if they are to be taken seriously by the scientific community. This highly reductionist theory of behaviour is roughly as follows.

Living things apprehend their environment by acquiring sensations or 'sense data' - the atoms of perception - which provide them with measurable atoms of information or 'bits'. It is accepted that living things have a memory but the memory too is atomised - the atoms of memory being referred to as 'engrams' or memory traces. The atom of behaviour is the reflex, a mechanism whereby an environmental event, referred to as a 'stimulus', triggers off a blind and automatic response.

As far back as 1906, Sherrington wrote that the "simple reflex is probably a purely abstract conception ... if not a probable fiction". [1] Herrick describes it as a pure abstraction, a "mere manifestation of what is clearly a very co-ordinated pattern of behaviour". [2] However such criticisms have not deterred others from presenting the process even more mechanistically in terms of inputs and outputs, the stimulus being the input and the response the output, like a machine that switches on when the button is pressed.

Learning has to be explained of course and mainstream scientists have done so very ingeniously, making it appear random, atomistic and mechanistic as is required by the paradigm of science. The learning process is not seen by them as occurring in any necessary order; living things are seen as simply responding to their environment in a purely random way.

A rat trying to find its way out of a maze, for instance, will try out in random order a whole series of trial and error moves, corresponding to the Darwinist's random variations and the random genetic mutations of the neo-Darwinists. If one such move is crowned with success, it is said to be 'reinforced', the Darwinian and neo-Darwinian equivalent of being 'selected'. In laboratory experiments, rats are given rewards for making what the experimenter judges to be the right moves and are often also given penalties, such as being administered electric shocks, for the wrong ones - rewards and penalties being seen in the Benthamite tradition as the only motivations of living things.

Behaviourists attribute remarkable powers to reinforcement, just as neo-Darwinists do to selection. Thus Skinner informs his readers that "a man talks to himself ... because of the reinforcements he receives"; [3] that thinking is in fact "behaviour which automatically affects the behaviour and is reinforcing because it does so"; [4] that "just as the musician plays and composes what he is reinforced by hearing, or as the artist paints what reinforces him visually, so the speaker engaged in verbal fantasy says what he is reinforced by hearing, or writes what he is reinforced by reading", [5] while the creative artist is "controlled entirely by the contingencies of reinforcement". [6]

It is astonishing that serious people can really believe that the incredibly subtle and sophisticated behaviour of living things can be explained in terms of so crude and rudimentary a mechanism. If we really believed it, as Michael Polanyi points out, we would also have to accept that "if a dog were consistently offered food whenever it was shown the radiogram of diseased lungs and no food when shown the radiogram of healthy lungs, it should learn to diagnose pulmonary diseases". [7]

Pure trial and error learning, however, except perhaps in the most rudimentary possible forms of life, is mere fantasy. This has been realised even within the ranks of experimental psychologists. Thus Karl Lashley has insisted strongly that normal animals do not behave in a random fashion. Oatley agrees: "In animal or human behaviour," he maintains, "trials are not chosen randomly." [8] On the contrary, as Herrick notes, "the learning process is ordinarily directive and it is an organised activity, never a mere random fumbling." [9] I.

Krechevsky considers that "we must change our description of the learning process so as to recognise the existence of organised and systematic responses at all stages of the process." [10] Indeed, the rat,

"when placed in an unsolvable situation, does not respond in a helter skelter chance fashion, but makes all sorts of integrated, and informed attempts at solution. These systematic responses are, partly at least, initiated by the animal and are not altogether merely a resultant of the immediately presented external situation."

This means that such animals are not just reacting blindly to various stimuli but are seeking to interpret them correctly and thus understand their meaning; that, of course, involves establishing the role of the thing or event that attracts their attention within the context of the larger spatio-temporal system of which it is part.

When the Tahitians first saw a horse, introduced to the island by de Bougainville's French sailors, they immediately classified it in terms of the mammal they knew most resembled it ... the pig. It was obviously closer to the horse than the other two mammals of which they had any experience: the dog and the Polynesian rat. Very sensibly, they referred to the horse as a "man-carrying pig". This was not a robot-like response; it was a sophisticated attempt to understand this strange beast in the light of their experience of similar beasts. There is no reason to suppose that rats do not do likewise.

Some of Lashley's experiments cast considerable light on this question. He showed that rats with large cortical lesions could learn to solve a problem in about the same time as normal rats but their behaviour was simplified. They did so less elegantly. They stumbled on the solutions rather than find them in a more systematic or logical way. In other words, they did not really understand the nature of the problems they had to solve.

This reminds one of the behaviour of children brought up in isolation, whose model of the world remains stunted and rudimentary. Peter of Hanover, the celebrated 18th century isolate, brought to London as a curiosity, ended up working as a farm labourer. He would perform his tasks well but could never really understand their meaning. When asked to load a wheelbarrow with manure, he would do so very efficiently but, not understanding the point of what he was doing, would proceed to empty it and fill it up again repeatedly until made to stop.

To understand something is to determine its function within a larger spatio-temporal system. This means that widening the context in which we study it will increase our knowledge of its function. This is how a detective tries to understand a crime. Each clue is related to an increasingly wider set of events, gradually acquiring ever greater meaning. Other clues are treated in the same way, until eventually, the crime is reconstructed and the detective can be said to have understood exactly what has occurred. There is no other way of proceeding. The clue in isolation from its context cannot be interpreted, for it has no meaning and constitutes data, not information.

Living things, except for the very simplest ones, do not behave like robots. They are intelligent beings and whether mainstream science likes it or not, they seek desperately to understand their relationship with the world about them.

References