The Unnatural Nature of Science Page 16

  In order to focus on a particular example of a theory in the social sciences, I have chosen psychoanalytic theory. Freud has provided us with a set of seductive ideas that have had a major impact on how we try to explain human behaviour. I will consider whether or not these ideas constitute a science. I shall try to show that if psychoanalysis is a science at all, it is a very primitive science and, in an important sense, premature. An analogy can usefully be drawn between embryology in the eighteenth century and psychoanalysis today. The problem of whether embryos are preformed or whether form emerges during development can provide a useful model for a science at a primitive or even premature stage.

  The first clear-cut body of embryological knowledge is associated with Hippocrates, in the fifth century BC, who viewed development in terms of what he considered to be the two main constituents of all natural bodies: fire and water. This may now seem no less absurd than Thales’s passion for water (Chapter 3), but at least it was an attempt to explain the nature of things in terms of a general theory. But the so-called triumph of Greek embryological thought belongs to Aristotle, who believed that the embryo was made out of menstrual blood and that the male dynamic element gave it shape. Aristotle asked whether all the parts of the embryo come into existence together, or are they formed in succession, like the knitting of a net? He thus defined the preformation/epigenesis debate which was to continue for 2,000 years. Having opened chickens’ eggs at different times, he argued in favour of the knitting analogy and thus for epigenesis – that is, the gradual generation of embryonic structures. But his rejection of preformation – that everything was preformed in miniature from the beginning – was based on philosophical arguments, not on observation.

  Aristotle’s theory of development based on epigenesis as distinct from preformation was not based on any real evidence. Though we now know that embryos do develop by epigenesis, he was correct for the wrong reasons: it was little more than an inspired guess. However, he posed important questions about the nature of development, and his influence was enormous. So there was little progress in thinking about development until the late nineteenth century. Fabricius and William Harvey, in the seventeenth century, while providing detailed observations on developing embryos, were essentially Aristotelians, using their observations, with little justification, to support epigenesis.

  The notion of preformation in contrast to epigenesis, namely that all embryos had existed from the beginning of the world, was first formulated in detail in the 1670s by, among others, the biologists Malebranche in France and Swammerdam in the Netherlands. Preformation was based on the concept of the first embryo containing all future embryos. There were grave doubts as to whether epigenesis was possible, and it also seemed to undermine God’s powers. Could it really be possible for the marvellous development of embryos to be explained by mechanisms based on scientific principles such as physical forces? Preformationists believed, rather, that all the organs were present from the very beginning of development and merely grew larger. If preformation was correct, then the embryo must contain, in miniature, all the future animals (or plants) to which it would give rise. It appeared to Malebranche as not unreasonable that there are an infinite number of trees in a single seed: such a thought seemed extravagant only to those who measured God’s powers by their own imagination.

  The eighteenth-century preformationists had an answer to all criticisms. When the French preformationist Charles Bonnet was confronted with the argument against preformation that, if the first rabbit had enclosed within it all future rabbits, it would have had to contain 1010,000 preformed embryos, he merely answered that it was always possible, by adding zeros, to crush the imagination under the weight of numbers. He described the preformation theory as one of the most striking victories of the understanding over the imagination: ‘one of the greatest triumphs of rational over sensual conviction’.

  The rise of preformationist theories may have been a response to a series of philosophical problems. If matter could form organized beings, little role was left for the Divine Creator. Moreover, physical mechanisms for development, as suggested by Descartes, seemed impossible, for the mechanisms as exemplified by the laws of motion were blind, in that they were without direction. How could such forces generate the perfection of an organism? One answer, offered by the French biologist Buffon, was an interior mould which gave the embryo its form by means of ‘penetrating forces’. Another proposed that the embryo contained a ‘building master’. But such concepts provided no real answer at all, and they were severely criticized by preformationists such as the Swiss scientist Albrecht von Haller. The invariable production of always similar, always divinely constructed animals appears to be too great for the simple forces that produce something like a salt crystal.

  By contrast, Caspar Friedrich Wolff was an ardent epigeneticist who based his ideas on two principles: the ability of plant and animal fluids to solidify and a vis essentialis, or vital force, which together could explain development. Wolff was a rationalist dominated by the idea of sufficient reason. He was unconcerned about the nature of his vis essentialis: ‘It is enough for us to know that it is there, and to recognize it from its effects …’ He objected to the preformationists’ reliance on God rather than on a cause of generation. Much of his famous debate with Haller related to the development of blood vessels surrounding the embryo, which is a striking early event in chick development. Haller believed that the blood vessels pre-exist but only gradually become visible, whereas Wolff claimed that the blood vessels arose during development. Haller charged that Wolff was making an unwarranted assumption: namely, that if one cannot see a structure it does not exist.

  One of the most attractive features of physics is that it can provide a set of basic mechanisms which can explain an extraordinary variety of phenomena. The basic mechanisms themselves may, like Newton’s laws, not be easy to understand, but they are much simpler than the varied movements of objects which they can explain. A major difficulty with the preformation/epigenesis debate was that the explanations being offered were as complex as the phenomena themselves. Preformation itself was, of course, hardly an explanation, for it simply said that everything was there from the beginning, and the concepts ‘building master’ and ‘vital force’ in relation to epigenesis were no less complex than embryonic development itself.

  Was there evidence that could have settled the question in favour of either epigenesis or preformation? The answer seems to be ‘no’, for this was a problem that lingered on for at least another century and a half. Embryology remained essentially descriptive, and there was no causal analysis. It was only at the end of the nineteenth century, with the beginnings of experimental embryology, that the preformation/epigenesis issues began to be more clearly defined and some of the issues were settled. Understanding that all organisms are made of cells was essential and put the whole problem in quite a different light, for it then became possible to realize that new cells are generated by cell multiplication during development. Advances in cell biology, aided by the invention of better microscopes, made it clear that embryos could develop by epigenesis, and eventually, only recently, it was realized that DNA provides the programme for embryonic development. Preformation died slowly, and even at the end of the last century August Weismann’s theory, in which determinants in the egg nucleus were the controlling elements, was preformationist in concept.

  The preformationist/epigenesis controversy in the eighteenth and nineteenth centuries can be thought of as being premature, reflecting the premature state of embryology as a science. The problem was just too difficult for the time, and advances had to be made in other areas of biology, particularly cell biology, before progress in understanding development could be made. To use again Peter Medawar’s wonderful aphorism, science is the art of the soluble.

  This history of embryology can offer parallels with the current status of psychoanalytic theory. Is psychoanalysis a science? Is it helpful to consider whether or not it is common sense? Is
Popper’s falsification criterion helpful? The issue here concerns the explanations offered by psychoanalysis for aspects of human behaviour, not its therapeutic effectiveness.

  ‘The intention is to furnish a psychology that shall be a natural science; that is to represent psychical processes as quantitatively determinate states of specifiable material particles.’ This is Sigmund Freud’s opening sentence of his manuscript Project for a Scientific Psychology in 1895. At the end of his life, Freud insisted that his psychoanalytic enterprise had the status of a natural science, and he claimed that the explanatory gains from proposing an unconscious mind ‘enabled psychology to take its place as a natural science like any other’.

  Today, however, there are those who argue that Freud was guilty of ‘scientific self-misunderstanding’. It is argued that the criteria and methods of the physical sciences are inappropriate in thinking about psychoanalysis and other complex aspects of human behaviour. Instead a hermeneutic reading is proposed, by which it is meant that psychoanalysis should be viewed as an interpretive procedure, rather like interpreting a written text. Narrative explanations are always dependent on the context within which they are given and are thus complexly related to all the various cultural influences that affect the context. But, as Adolf Grünbaum has cogently argued, this approach obfuscates the whole issue. The statements of psychoanalytic theory are about tendencies or likelihoods of some behaviour occurring and are indeed causal statements: they are about cause and effect. Concepts relating to the unconscious, libido, Oedipus complex, and, above all, repression have entered our everyday thinking and are used to provide causal explanations. We need to know how reliable such explanations are, for there can be no doubt that psychoanalysis has transformed the way in which we think about human behaviour.

  The concept of repression of unwanted thoughts which are too painful or disgusting is central to psychoanalysis. It derives from the original ideas of Breuer and Freud in 1893, and can be considered to be the cornerstone on which the whole of the structure of psychoanalysis rests. In the course of the treatment by hypnosis of patients with hysterical symptoms, Freud and Breuer observed that there seemed to be a release from the symptoms if the patient had a cathartic experience which revealed the underlying cause. They concluded that for each distinct symptom affecting the neurotic patient, the patient had repressed the memory of a trauma that had closely preceded the onset of the symptom, and that the trauma had some analagous features with the symptoms. The famous example was Breuer’s patient Anna O., who had a phobia for drinking liquids. They claimed that she had repressed the sight of a dog drinking water from a friend’s glass, which had disgusted her. By recalling the incident, the repression was lifted and there was a dramatic disappearance of the symptom. This idea was developed by Freud into a model in which not just recent traumas were important but in which symptoms would, in general, reflect childhood repression with a sexual content. (Ironically, recent evidence shows that Anna O. was far from cured and had several relapses over a period of years.)

  The mental apparatus was divided by Freud into an ego, an id and a super-ego, each of which was involved in controlling the flow of psychic energy. The id is that part of the unconscious mind that is governed by irrational forces such as aggression, while the ego operates rationally and the super-ego acts as the moral conscience. Thus the ego withdraws energy from all associations which are unpleasant, and this results in repression of a memory or emotion which is still stored in the unconscious. It is, apparently, no easy matter for the ego to keep unconscious thoughts under control, and so the unconscious desire to injure someone, for example, requires defensive manoeuvres which result in the desire appearing under a different guise, and such repressed desires may also come out in a dream. While some of these ideas are novel and surprising, they do have an element of common sense about them, for they essentially locate in the mind three people with behaviours and feelings with which we are all more or less familiar. Behaviour results from the conflicting attitudes each has.

  When Popper came across these ideas, in around 1920, he described the partisans of psychoanalysis as seeing confirming instances everywhere: their world was full of verifications of the theory. No matter what happened, the theory was always capable of explaining it. In this continual confirmation he saw the weakness of what he regarded as their inductivist methodology. Popper thus abandoned verification as a strong basis for a scientific approach and proposed that only falsification is an important criterion for a science. For Popper, ‘these … clinical observations which analysts naïvely believe confirm their theory cannot do this any more than the daily confirmations which astrologers find in their practice.’ But does that mean that if psychoanalysis were falsifiable, or if at least parts were, it would be a science?

  Grünbaum has argued that Popper’s criticisms are unfair, for did not Freud modify his theories in the light of his clinical experience? For example, Freud even considered giving up the psychoanalytic method of investigation when his theory of hysteria based on seduction collapsed after he had come to the conclusion that his patients’ reported seductions were fantasies. (One of the peculiarities, and weaknesses, of the psychoanalytic method is that there is usually no way of telling whether a patient’s recollection of early events is fact or fantasy. This has bedevilled the whole question of claims concerning sexual abuse.) Another case relates to Freud’s theory of paranoia. His hypothesis was that repressed homosexual love is necessary for someone to be afflicted by paranoid delusions. One of his patients was a young lady who thought that the man she was involved with was arranging for their lovemaking to be photographed, in order to disgrace her. In the initial psychoanalytic session, Freud could find no sign in her of a homosexual attachment, so either his theory was wrong or the young lady’s report about her lover was correct. If indeed the young woman was paranoid, Freud was prepared to abandon his theory that the delusion of persecution invariably depends on homosexuality. But rather than praise Freud for accepting falsification, we should be critical of his conclusion. For how could Freud ever be sure that there was no homosexual attachment? Even after years of analysis, how could he claim the absence of such a homosexual feeling? There is nothing that requires that it had to emerge during the analysis. Thus, far from providing a good example of the refutation of psychoanalytic theories, it does just the opposite: it shows how difficult, if not impossible, it is to falsify such a theory.

  Freud regarded dreams as the royal road to a knowledge of the unconscious activities of the mind and claimed that repressed infantile wishes are the causes of all dreams. In brief, when asleep, the ego’s vigil on the id is relaxed and unwelcome thoughts try to enter consciousness and might disrupt sleep if they were allowed to enter. The unwelcome thoughts are paired with material retrieved from recent experience, and the true ‘meaning’ or ‘latent content’ of the dream is thus disguised – transformed – into bizarre forms and with symbolic representation. (What, one wonders, does the pairing, the transformation?) Freud himself revised his basic theory that all dreams were basically wish-fulfilment. He felt that he had satisfactorily disposed of objections to this idea based on the occurrence of anxiety dreams and punishment dreams – these, he explained, were the fulfilment not of instinctive impulses but of the censoring, critical super-ego – but, even so, there was one set of dreams he regarded as inexplicable: the dreams of victims of traumas such as war. Such victims regularly relieve their traumatic experiences in their dreams, and Freud could not see what wishful impulse could be satisfied by harking back to such exceedingly distressing experiences. But since there is no clear indication as to how a trauma is defined or on whether a wish has been fulfilled, the reliving of traumas is really no less, or more, explicable than any other aspect of Freud’s theory of dreams.

  The cases just quoted mimic both scientific method and science, but they barely qualify as science at all, because both the phenomena and the theory are so ill-defined. The problem with psychoanalysis is not philo
sophical but lies in the nature of the theory and the state of the subject: many (probably all) of the concepts in the theory are so loosely defined that the phenomena cannot be defined unequivocally and independently. Take trauma, for example. How does one decide what constitutes trauma? If seeing a dog drinking from a glass is traumatic, then there must be thousands of events in our lives which are traumatic. Could there be a trauma-free person? A further weakness is the role of repression in causing neuroses: it is apparently a necessary but not sufficient condition. What, then, provides the sufficient conditions under which repression of traumas causes clinical symptoms? Unless these are clearly specified, one is left with no theory at all: at best there is a weak correlation of poorly defined events. Even worse, the raw data of psychoanalysis are not verifiable by multiple independent observers.

  Then take the concepts of id, ego and super-ego, each of which has a character almost as complex as the phenomena they are trying to describe. Add to these a concept of psychic energy, different development stages such as oral and anal – each replete with a gamut of human emotions such as disgust, anger, desire and jealousy – and then allow both positive and negative interactions. It becomes, in principle, impossible to make any meaningful predictions or explanations: the concepts are so vague that it is almost impossible to imagine any behaviour that would either verify or falsify the theory. But these ideas do in a sense enlarge our common-sense understanding of how people behave and that is why the ideas are so seductive. And one should not deny the usefulness of collecting data which might, for example, link early experiences with later behaviour patterns. It could be very important to show, for example, that the effects of early separation of a child from its parents has a serious implication for later development.