Freud wrote voluminously on sexuality and gender. My purpose here is not to give a summary or a history of his ideas and those of his followers, but to explore the contribution they make to a social theory of gender. Accordingly I will start not with the Oedipus complex and Freud’s discussions of femininity and

masculinity but with his general theorizing about the relationship between the psychic and the social.

On this theme Freud developed two main lines of thought. One made sweeping and mostly fanciful analogies between the psychic life of children and neurotics on the one hand, and ‘savages’, crowds and groups on the other. The other, much more profound, was concerned with the social repression of sexuality in European culture. This led him in 1908 to a remarkable psychological critique of marriage and the double standard as institutions that hinge on the suppression of women’s sexuality. The social pressure that denied women sexual happiness was the major force behind the growth of neurosis. In later writing, culminating in Civilization and its Discontents (1930), Freud generalized this argument about repression and laid more stress on its historical dynamic.

A conflict between civilization and nature is of course a traditional theme of Western thought. But Freud is not merely doing the rounds of antique philosophy. The clinical tools of psychoanalysis, especially the concept of repression, enable him to formulate the issue in a much more precise and complex way and move it in a new direction. In ‘Civilization and its Discontents’ he argues a similarity ‘between the process of civilization and the libidinal development of the individual’, which is not just an analogy but a relation: ‘civilization is built upon a renunciation of instinct’. The ‘renunciation’ is not projected into a mythical past or a philosophical heaven, as in Rousseau’s Social Contract (and indeed Freud’s Totem and Taboo). It is something that operates in every person’s life in the process of psychosexual development, and takes effect at the level of the individual unconscious through the mechanism of repression. Neurosis is a material, not a metaphorical, consequence of the social pressures on the person required by the advance of civilization. The structure of the adult personality is formed by this pressure, principally by the way it is experienced by the young child in the family. The formation of the super-ego as part of the person’s psychic equipment is itself a social mechanism. In Freud’s striking simile: ‘Civilization, there­fore, obtains mastery over the individual’s dangerous desire for aggression by weakening and disarming it and by setting up an agency within him to watch over it, like a garrison in a conquered city.’

But if ‘nature’ is thereby rendered partly social, so, conversely, the social is brought within the sphere of nature. Civilization is

not seen as something external to the events and processes of psychic life: it is seen as a product and extension of them. Its achievements represent a sublimation of the impulses that would, but for repression, find expression in a more raw and bloody fashion. In these later and more complex formulations Freud has shifted from a position that makes civilization the cause of neurosis, to one – admittedly not clearly stated but certainly implied – that sees civilization as continuous with neurosis, as part of the same structure. At this point he expounds the important concept of a historical dynamic in repression:

This conflict [of Eros and Thanatos] is set going as soon as men are faced with the task of living together. So long as the community assumes no other form than that of the family, the conflict is bound to express itself in the Oedipus complex, to establish the conscience and to create the first sense of guilt. When an attempt is made to widen the community, the same conflict is continued in forms which are dependent on the past; and it is strengthened and results in a further intensification of the sense of guilt… If civilization is a necessary course of development from the family to humanity as a whole, then… there is inextricably bound up with it an increase of the sense of guilt, which will perhaps reach heights that the individual finds hard to tolerate.

Which brings back the problem of unhappiness and closes the circle, for him – but is precisely the point of departure for others.

With the technical concepts of psychoanalysis Freud thus dismantled the traditional antinomy of nature and culture. He replaced it, most unexpectedly – it is certain that he did not himself see the theoretical significance of this – by the concept of a historical process. This process operates simultaneously at the macro-social and the individual levels. In it both human personalities and their troubles, and collective social achievements, are integrally produced.

As Marcuse saw, this conception makes it possible to sociologize psychoanalytic concepts. The Oedipus complex can be seen as a product of a definite historical type of the family. Repression itself is no abstract consequence of human relation in general, but takes definite form and intensity in specifiable historical contexts. This does not reduce psychoanalysis to sociology, any more than it can be reduced to biology. But the concept of a historical process allows a much more powerful connection between the social and the unconscious and those theories which, for instance, have simply tacked psychodynamic notions such as ‘identification’ onto a sociology of social control. It makes possible a socially critical use of psychoanalysis, which is no longer the theory of an eternal contest between instinct and reality.

In the early years of Freudian theory its potential for social criticism was developed mainly by Alfred Adler, a socialist whose original and fertile psychoanalytic writings on power, masculinity, war and motivation are now mostly forgotten. (Adler later became the father-figure of a small cult based on a bland ‘individual psychology’ which had no use for his early radicalism. The parallel with the fate of Wilhelm Reich’s socialist writings is close.) Adler emphasized the social shaping of motives and criticized Freud’s persistent attempts to derive them-Trorn-binlngy He had a clear view ol the importance of power relations in social structure, and attempted to develop a ‘psychology of power’, a psychoanalytic account of responses to powerlessness in childhood and social insecurity in adulthood. Most notably he focused part of this argument on the power relations of gender.

Perhaps influenced by the socialist feminism of the day, Adler noted that femininity was devalued in European culture, and argued that this devaluation shaped the psychological patterns of childhood. Children’s weakness vis-a-vis adults was interpreted by them as femininity, ‘childish value-judgements’ crystallizing around the cultural polarity of masculine and feminine. Submission and striving for independence coexisted in the child’s life, setting up a contradiction between masculine and feminine tendencies. Usually this led to some kind of compromise, the tendencies balancing, if under tension, in the adult personality. But the compromise was not always reached. Anxiety about weakness could lead to overcompensation in the direction of aggression and compulsive­ness, which Adler dubbed, in a famous phrase, ‘the masculine protest’ (applicable to both sexes). He saw this as the key structure in neurosis, but also generalized it as a forcible critique of hegemonic masculinity: ‘To this [children’s uncertainties about sexuality] is added the arch evil of our culture, the excessive pre­eminence of manliness. All children who have been in doubt about their sexual role exaggerate the traits which they consider masculine, above all defiance.’

Adler’s sketch of a critical psychology of gender, unfortunately, bore little fruit. After the break between Adler and Freud in 1911, and another radical wave in the 1920s which included a debate on Freud’s theories of femininity, psychoanalysis evolved in much more conservative directions. A continuing interest in the psychology of women by a few analysts such as Karen Horney moved away from the social grounds on which Adler had attempted to locate gender issues. By the 1940s and 1950s, in the writings of John Bowlby, Theodor Reik, Erik Erikson and their contempor­aries, mainstream psychoanalysis had become a powerful ideologi­cal support of the patriarchal family and conventional definitions of gender. The subsequent feminist criticisms of psychoanalysis were thus well warranted, though much of the fire was misdirected at Freud rather than his interpreters. A striking indication of the political shift is the attitude to homosexuality. Freud had scrupulously refused to define this as an illness, and had cautioned against attempts to ‘cure’ a homosexual orientation. By the 1950s psychoanalysts freely defined homosexuality as pathological in itself, and proposed and undertook psychoanalytic treatment. They had a singular lack of success.

It was theoreticians outside the clinical world of official psycho­analysis who pushed on with the social dimension of Freud’s thought. For various reasons, including the fact that most were men and few had practised analysis, the question of gender dropped gradually out of focus. The interest rather was in the psychological underpinnings of capitalism.

In the 1930s and 1940s the focus was on the emergence of fascism out of liberal capitalism. Karl Mannheim, trying to understand the popular success of fascist movements, brought psychoanalytic arguments into sociology in his Man and Society in an Age of Reconstruction. Wilhelm Reich did something similar from further left in The Mass Psychology of Fascism. A whole programme of research attempting to weld psychoanalysis to a reconstructed Marxism came from the Frankfurt school: the Horkheimer collec­tion Studien uber Autoritat und Familie, Erich Fromm’s Fear of Freedom, and most famous of all, Adorno et al.’s The Authoritarian Personality. This work moved in the direction of characterology, and inadver­tently began producing a typology of masculinity. ‘Mack’ and ‘Larry’, the celebrated case studies of high and low authoritarian­ism, can equally well be read as studies in the formation of contrasting masculinities.

In another line of development Herbert Marcuse took up Freud’s argument about the rising level of repression with the advance of

civilization. Eros and Civilization argued that some of this was the ‘surplus repression’ required to sustain an exploitative class society, and that this bore down mainly on non-genital sexuality, narrowing spontaneous eroticism. In One Dimensional Man Marcuse turned back on his own tracks and argued that advanced capitalism now allowed a controlled release of repressions. This was, however, in forms that stabilized the social order rather than disrupted it. The result was a socially repressive ‘desublimation’ of instinctual urges.

Both these lines of argument have the same goal, which I will call a theory of embedding. Their purpose is to explain how a social movement like fascism or a social system like advanced capitalism can establish links with unconscious mental processes and thus gain mass support regardless of its irrationality and destructiveness.

The two psychoanalytic concepts central to theories of embed­ding were the displacement of cathexis and the constitutive role ■ of repression. Most social applications of psychoanalysis postulate a displacement of emotions originally formed in childhood and attached to family members, onto objects Or relations in the larger social world in adulthood. The energy of the displaced emotions is largely due to the blocking of original impulses in the process of repression, the pattern of which is decisive in the formation of adult character. The same formative encounters thus shape personality and generate macro-social effects, by dividing conscious from unconscious and determining a particular content for each level.

When in the 1970s a new generation of feminist theorists roped in psychoanalysis to supply part of a critical theory of gender, it was largely as a theory of embedding. The social structure whose stability was at issue here was patriarchy, especially the sexual division of labour. What was felt surprising or in need of explanation was women’s acceptance, even active endorsement, of social arrangements that oppress them. The search for explanation was partly a matter of finding the psychic gain in the social loss, and partly a matter of tracing out through psychoanalysis the forms of adaptation to necessity in the shape of men’s social power.

Nancy Chodorow’s argument in The Reproduction of Mothering, though its details derive from the object-relations school of psychoanalysis, follows the basic logic of Fromm and Adorno. What comes out of the analysis is a character typology, in this case a fundamental distinction between femininity and masculinity.

The typology is used to provide a psychodynamic explanation of the acceptance of a social structure, in this case the sexual division of labour in childcare. Because of the different patterns of attachment to the mother, girls grow up with less sharply defined boundaries of the self and a greater need for emotional completion in relationships, boys with more clear-cut ego boundaries and a greater need for separateness. Women both want to mother, and are psychologically adapted for mothering, more than men.

Juliet Mitchell’s Psychoanalysis and Feminism is also concerned with character structure, though in her case the focus is a reconstruction and correction of Freud’s account of femininity. The scope of Mitchell’s social theory is considerably wider than Chodorow’s, as she construes Lacanian psychoanalysis in terms of an Althusserian theory of ideology, and patriarchy in terms of Levi-Strauss’s theory of kinship and culture. It is, however, less dynamic.

Like most of the structuralist theorists on whom she draws, the tendency of Mitchell’s argument is ahistorical. Embedding is construed as the insertion of people into places already laid out for them in a structure whose characteristics are predetermined. This is a common feature of feminist arguments based on Lacan’s reworking of psychoanalysis, for instance by Luce Irigaray. In Irigaray’s work, for all its psychological subtlety, there is a thoroughly categorical social analysis of gender. This is partly because of the structuralist style, partly because the focus is not on the division of labour but on the symbolic aspect of social process. ‘Patriarchy’, or the ‘law of the father’, in these writings is less the structure of social relations than the structure of how the world is imagined.

History is more fundamental in the most complex and original work of feminist psychoanalysis, Dorothy Dinnerstein’s The Mermaid and the Minotaur. This is to an extent in spite of her own theory, as she takes the sexual division of labour to be an evolutionary constant in human society before the present. Dinnerstein focuses on women’s work as mothers – ‘the female monopoly of early childcare’ – as the main determinant of the adult emotional patterns that sustain patriarchy. In explaining this she depends much of the time on a normative standard case.

She departs from the formula, however, in a series of brilliant sketches of emotional interactions. Here she catches, as no other social analysts and only a few novelists (Nadine Gordimer, Patrick

White) have done, the ambivalences of women and men towards each other, and the ways childhood resonances in their interactions lock the participants into patterns they would not rationally choose. Dinnerstein teases out the tendrils of the pre-Oedipal emotional relationship to the mother in explanations of adult women’s sexuality, men’s hatred for women, women’s acceptance of exclusion from ‘world-making’ in the public sphere, and, ultimately, the ecological crisis of contemporary civilization. Unlike structuralist appropriations of psychoanalysis, Dinnerstein’s does lead to a conception of a historical dynamic and a historical appreciation of the present moment. The book ends, in fact, with a remarkable psychodynamic account of the emergence of the New Left and the new feminism in the United States.

The crucial difference from structuralism is Dinnerstein’s focus on routine practice: the practice of childrearing on the one hand, the practices of interaction among adults on the other. This allows her to see the emotional patterns as structurally based but not structurally fixed, and to give a central place to emotional contradiction (in the Freudian tradition called ‘ambivalence’). The approach allows a recognition of the development of practice, of its historical cumulativeness. Habermas, in a moment of optimism, spoke of this capacity to form new structures as the ‘evolutionary learning process’ of societies. In Dinnerstein’s view the cumulation is rather in the direction of global disaster: ‘the prevailing symbiosis between women and men has something deadly wrong with it… It supports a growingly perilous societal posture; it helps lock us, as a species, into a suicidal stance toward the realities on which our collective survival hinges.’

Here we reach the limits of ‘embedding’ arguments, where the consequences of embedding bid fair to destroy the social order being embedded. We are also left with two theoretical difficulties. First, Dinnerstein’s argument is built around a normative standard case, though it carries conviction mainly when it complicates or departs from it. Can those complications and departures be made more systematic? Second, Dinnerstein’s argument, like Chodorow’s and Mitchell’s, assumes a fairly straightforward displacement of affect from individual family figures in childhood to whole categories of people in adulthood, most notably by undifferentiated use of ‘woman’ for both. But this displacement is not straightfor­ward, as psychoanalytic studies abundantly show. A classic example is the enormous difficulty and complexity of a young boy’s

attempts to construct a stable pattern of attachment documented in Freud’s ‘Wolf Man’ case history. As in Freud’s theoretical writings about femininity and masculinity, Dinnerstein leaves us unclear why the displacement must happen, and what agency or choice the growing child/adult has in the matter.

Theories of embedding state the effects of gender formation; psychoanalysis also offers maps of the route towards those effects. They centre on the Oedipus complex, ‘the fateful combination of love for one parent and simultaneous hatred for the other as a rival’ as Freud summarized it, that emerges with the maturing of the child’s erotic life around four to six years old. This appeared in Freud’s psychiatric casework as the nucleus of the neuroses, and mainly on this evidence became the key moment in psychoanalytic theories of human development. It was, Freud assumed, structured differently for girls and boys. The basis of this assumption was the different places of the mother and the father in family dynamics. Some feminist reworkings of the Oedipal theory have stressed the power dimension of this relation (Firestone), others the division of labour (Chodorow, Dinnerstein); they are not clearly distinguished in Freud.

The outcomes of the differently structured Oedipal crises are the bases of Freudian accounts of femininity and masculinity. With girls the process involved a messy and incomplete abandonment of the desire for a penis, eventually transformed into desire for a baby and the man who could give one. With boys the mother was retained as erotic object, a cathexis eventually displaced onto other women; but this desire was repressed by fear of the castrating father, itself provoking identification with him, the internalization of prohibitions, and formation of a strong super-ego. No such process stimulated super-ego formation in women. These were, so to speak, standard resolutions. They could and often did go wrong. Further, closer analysis would show underneath these patterns traces of contradictory ones. The ‘complete’ Oedipus complex, Freud insisted, involved ambivalent feelings of desire and hatred towards both parents.

The main modification of this argument within the psychoana­lytic tradition has been a greater attention to the pre-Oedipal period. Freud himself began it. In the ‘Wolf Man’ case history, his most extensive study of the problem of masculinity, Freud noted that in the crisis of the boy’s relation to his father, the feminine aim in relation to the father was repressed because of

the fear of castration: ‘In short, a clear protest on the part of his masculinity!’ This argument clearly presumes a pre-Oedipal masculinity, which was important enough to the little boy to force the repression of the strongest attachment in his sexual life. Freud never spelt out the nature of this early ‘narcissistic masculinity’, but it is clear that it cannot be identified with the ‘activity’ end of the activity/passivity polarity that plagues psychoanalytic discussion of gender, including Freud’s earlier writings. It probably does not have much to do with sex of object-choice either; ‘narcissistic’ is an apt adjective. It is difficult to escape the conclusion that it is the cultural value placed on maleness in a patriarchal society that is at issue here, and is a decisive presupposition of the classic Oedipus complex. In this context the emphasis in the French analytic tradition on the symbolic valorization of the phallus seems justified.

The pre-Oedipal years also emerge as crucial in Chodorow’s and Dinnerstein’s arguments. What Chodorow calls mothers’ ‘Oedipally toned’ responses to little boys push them along the path to the Oedipus complex, while the closer identification of mothers with their young girls leads to more permeable boundaries of the self. Chodorow and Dinnerstein both emphasize the sexual division of labour as the crucial condition of these pre-Oedipal effects. This suggests that the psychoanalytic sequence might be interpreted in terms of the structures defined in chapter 5. The structures interact in different ways as the child grows older. The dominant interaction affecting the emotional life of the child changes.

Period Major structural interaction

Pre-Oedipal Cathexis x division of labour

Oedipal Cathexis x power

If this is correct, we may conclude that the emotional dilemmas of the encounter with power in the Oedipal stage depend on the outcome of the earlier interaction of cathexis and the division of labour.

The psychoanalytic theory of the normative standard case can be revised sociologically along these lines. But this still leaves unclear what Freud left unclear, just how ‘standard’ this case is supposed to be. The long controversy about the supposed ‘univer­sality of the Oedipus complex’ across cultures has not been particularly helpful. Those combatants who were sympathetic to

Freud usually took for granted that Oedipus was standard equipment in European families and debated whether he could be found in ‘primitive’ cultures. Nor are we helped by Dinnerstein’s honest confession that none of the five people to whom she is closest matches at all well ‘the composite portraits of “normal” masculinity and femininity’ presented in her book.

The roots of this difficulty are connected to the unitary conceptions of gender discussed at the start of chapter 8. It is assumed that the systematic character of gender and sexual relations requires one central pattern of personality, plus deviations. The argument of the later part of chapter 8 may help sort out the place of the Oedipus complex as well. More than one major pattern of childhood emotional development is produced; and they are produced in relation to each other.

Freud succeeded in isolating and analysing one important pattern of psychosexual development in European families. For masculinity this was quite probably, in the social context of upper-middle-class Vienna, the hegemonic pattern. Some of his descriptions of femininity imply the ‘emphasized’ pattern sketched in chapter 8, though his confusions and hesitations about femininity show he found this harder to pin down. That too is consistent with the argument about the absence of a ‘hegemonic’ femininity in patriarchal society.

The crucial point is that no one pattern of development can be taken as universal even within the specifical social context Freud studied. Researches like Anne Parsons’s work on southern Italy have documented alternative ‘nuclear complexes’, and this point has to be applied within cultures as well as between them. The Oedipal drama is constructed in quite specific situations. Not only are there multiple pathways through childhood, the routes can and do change as gender relations change in history. Erikson’s suggestion that the focus of neurotic conflict has changed during the twentieth century – formerly about sexuality, now about identity – is a useful warning, though greatly overgeneralized. The now familiar research by Philippe Aries and others on the history of the family implies longer-term changes in the psychodynamics of child development.

In this light, the formulations of femininity and masculinity by dissident psychoanalysts like Adler and Jung acquire a new interest. Instead of seeing them as displacing Freud’s theory of the Oedipus complex (which Adler and Jung certainly intended), we can see them as accounts of other paths of psychosexual development that were available to be discovered through the research method that all schools of psychoanalysis share, the reconstructed clinical case history. The alternatives include Adler’s notion of the compulsive ‘masculine protest’, already described; Jung’s somewhat similar notion of a strong, authoritative masculine ‘persona’ based on the repression of weakness and dependency which then coexists as the unconscious ‘anima’; the Frankfurt school’s models of‘authoritarian character’. These can be brought together as elements of the psychoanalysis of a historically constructed range of masculinities. The Authoritarian Personality, sketching the ‘democratic personality’ and the social and political conditions for resistance to fascism, even made a beginning on the immensely difficult problem of the relations between these different dynamics.

This is as far as I can take this problem at the moment. The next step, clearly enough, is the empirical investigation of the range of developmental paths and psychodynamic patterns in a given milieu, and their interconnections.* The methods of such a study cannot be formulated without a closer examination of the idea of life histories, which I will attempt later in this chapter and in the next. Before doing so it is necessary to take a closer look at the concepts shared by all schools of classical psychoanalysis, repression and the unconscious.

In Freud’s original and most influential version, repression is applied to libido, to erotic impulses, and the whole theory of the formation of personality thus rests on an account of libidinal development. As Mitchell points out this is what revisers of Freud have found hardest to handle, and have most commonly dropped, in particular the theory of children’s sexuality. With a few exceptions, the libido theory was preserved only by the most orthodox school of psychoanalysts, whose increasing social and political conservatism blanked out its subversiveness.

The attempt to recover the erotic dimension, and to bring out the social criticism implicit in the theory of repression, has been central to radical reworkings of psychoanalysis which have seen Freud as a theorist of liberation more than a theorist of embedding. Norman Brown in Life Against Death, Marcuse in Eros and Civilization and An Essay on Liberation, and Reimut Reiche in Sexuality and Class Struggle present versions that largely ignore gender. But the same themes have been applied to women’s liberation by Dorothy

Dinnerstein, and to gay liberation by Dennis Altman in Homosexual: Oppression and Liberation and Mario Mieli in Homosexuality and Liberation.

The concept of repression, then, points to an erotic dimension of liberation, though it does not fully specify it. All the works just cited are forced to go outside the psychoanalytic framework to formulate the process of liberation. A purely psychic liberation is rather hard to sustain. Freud’s own therapeutic purposes were deliberately limited, given his acceptance of the established social order and family structure.

If patterns of femininity and masculinity are composed as patterns of repression, the reconstruction of gender does involve a reworking of erotic attachment and expression. Mieli sees this optimistically and expansively, as early gay liberation generally did. Dinnerstein is rather more sombre about the ‘project of sexual liberty’, though she sees it as central to whatever chance humans have of surviving. An important tendency in feminism sees the path forward as a withdrawal of women’s emotional attachments and energies from men to focus them on other women. Though there is a long way to go with any of these strategies, the political experience of the last twenty years is already consistent with psychoanalytic theory in at least two respects. Patterns of erotic attachment have proved deep-seated and difficult to change. The attempt to change them provokes strong and complex resistances at the personal level as much as the political.

The concept of repression is central to the main complication that psychoanalysis introduces to the concept of sexual character, since repression constitutes the unconscious. The conceptions of sexual character discussed in chapter 8, for all their differences, mostly share the assumption that personality is homogeneous – expressed in gender scaling by the requirement of statistical consistency, in humanistic psychology by the notion of a core ‘gender identity’, and so on.

Classical psychoanalysis violates this assumption comprehens­ively, and this is one of its most important contributions to the analysis of gender. The notion of ‘primary process’ thinking in the id and ‘secondary process’ in the ego; the idea of unconscious impulses which cannot be directly expressed in consciousness; the notion of the dream or symptom as a ‘condensation’; and the structural model of personality (id, ego, super-ego) – all were ways in which Freud tried to formulate the idea that repression marks out qualitatively different parts of the human psyche. They not only can contradict each other, they normally are in contradiction. Repression itself is a mechanism of contradiction.

We do not need to accept Freud’s particular formulations to appreciate the importance of the idea. It implies that femininity and masculinity are normally internally fissured and in tension. To use a rather static model, they are normally layered. (To correct the image one must bear in mind that the relationships between the layers are as important as their content, and the layers themselves writhe around, so to speak, as the person moves on a life trajectory.)

This can be accepted without the libido theory on which Freud based it. Indeed its most clear-cut application to gender is not by Freud but by Jung in his essay ‘Anima and Animus’. Here Jung suggests that sexual character is systematically layered in the sense that the public face of femininity and masculinity, compatible with the conventional social role, is always constructed by a repression of their opposites. A kind of unconscious personality (the ‘anima’ for men, ‘animus’ for women) develops as "a negative of the socially acceptable one, and their incompatible demands underlie many of the tortured emotional dynamics of marriages.

Freud’s case histories present more complex layerings again. The ‘Wolf Man’ is the classic example, Freud peeling off in turn an inhibited adult heterosexuality, a promiscuous but emotionally shallow adolescent heterosexuality, a passive and highly ambivalent homosexual attachment to the father, arriving at last at the pre – Oedipal narcissistic masculinity that has already been mentioned. Dinnerstein’s argument likewise revolves around the idea of layering, with the pre-Oedipal attachments as hidden determinants in the most complex adult emotional relationships.

Jung’s argument dramatizes the most problematic aspect of this: that femininity and masculinity can coexist in the same personality. Freud wrestled with this idea endlessly, talking variously of bisexuality, of activity/passivity, and so on; he was never very satisfied with his answers and we need not be either. Jung’s analysis is better since not reductionist – the starting-point is social role rather than biology – but is too schematic to carry us far.

The evidence for layering is convincing, and requires us to see the tension between different layers or tendencies in personality as a constitutive force in gender relations. Yet its theorizing seems fragile, vulnerable to any criticism of the concepts of repression and the unconscious.

Two critiques are of particular importance. The one formulated by Sartre depends on an account of consciousness and motivation which will be discussed in the next section. The other mounts an essentially political critique of the concept of the unconscious. Its clearest statement is in The Politics of Sexuality in Capitalism by a London radical group of the early 1970s. The ‘Red Collective’ argue that what Freud took to be effects of the unconscious are in fact effects of power, both class power and patriarchal power, their causes being invisible only so long as these structures are not brought into question by political practice. Psychoanalysis is an emanation of the establishment, and the concept of the unconscious is a mystification justifying dominance of the thera­peutic situation by the therapist.

The Red Collective is right to point to the social practices in which psychological concepts are formulated. The book is theoretically interesting, as well as moving, in reporting the group’s own attempts to work through the personal politics of sexual relationships. Its criticism of the notion of the unconscious is apt as far as the unconscious is thought of as a place where certain mysterious things happen, comprehensible only to the expert. Freud has to bear some responsibility for this, as much of the time he used spatial metaphors in talking of the unconscious.

But fundamentally Freud’s concept was processual, based on the dialectic of impulse, repression and symbolization. Even more, it was strategic, a concept arising in connection with practices of transformation. As Freud insisted at boring length, he did not discover the unconscious mind; that had been known to poets and philosophers for centuries. He was simply one of the first to try to do something about it, in the sense of working out therapies for people who had got it badly snarled. All Freudian theory developed in relation to a practice of attempting to transform lives. The concepts of repression and the unconscious mark both obstacles and potentials encountered in this practice. And that is, I think, the fundamental sense in which they are important to other practices of sexual politics.

Structuralist and poststructuralist ‘readings’ of Freud have been misreadings exactly so far as they separate the concepts from their context in practice. That separation is central to the pessimism these readings have led to. To bear a different fruit, analytic concepts have to be reconnected to practice. The Red Collective rightly argues that this will have to be based on different social foundations from the doctor/patient format. The medical model of psychoanalysis proved easy to absorb into conventional medicine, and to turn into a politics of ‘adjustment’ and non-violent social control.