What are we talking about when we talk about women?

Ancient religion has attracted some outrageous scholarship. And women’s religion in the ancient world – from cave people to the early Christians – has been blessed with far more than its fair share of lunacy. Part of this lunacy has, it is true, been confined to the wilder shores of popular imagination: vestal virgins having a dangerously good time with the highest-ranking senators of Rome; primitive mother goddesses ruling the roost in the never-never land of Stone Age matriarchy; beautiful Christian virgins speedily converting their thuggish Roman (would-be) lovers, then firmly leading them by the hand into the lion’s mouth. I am not only thinking of the licensed inaccuracy of film and fiction, however. Otherwise serious academics still offer arguments about women and religion that would be promptly – and rightly – laughed into the dustbin in almost every other field.

Take the currently fashionable questions about women and Christianity: why the attraction of Christianity for women? Why the apparent enthusiasm among some early Christian women for asceticism, sexual renunciation and self-starvation? One favoured scholarly answer stresses the liberating possibilities of early Christian communities. Christian asceticism enabled women to (re)gain control over their own bodies and find an escape route from the patriarchal authority which condemned them to the subservience of constant child-bearing. This argument works – up to a point. But up to what point? What happens when self-starvation turns, as it sometimes did, into lingering suicide? Is it still liberating? Or does it show the exercise of a different (and yet more lethal) kind of control over women’s bodies – no more a path to female autonomy than being torn limb from limb in the jaws of a lion? Would such crude arguments have been taken seriously in the sphere of supposedly rational politics rather than that of irrational religion, if they had been about slaves or oppressed provincials rather than about women?

In Her Share of the Blessings Ross Kraemer does a much better job of making sense of women’s religions in the ancient world than most of her predecessors. Not only does she provide a sophisticated discussion of the major cults in which we know women played a significant role – from the cults of Adonis and Dionysus in Classical Greece, through the official (and unofficial) rites of Roman women, to Judaism and the Jesus movement – but she also makes a brave attempt to set this patchy, diverse, often contradictory evidence in a general theoretical framework. General theory and the study of women’s religious experience in Antiquity have not gone hand in hand recently and Kraemer is somewhat breathless in her search for an appropriate model. ‘Androcentric’ approaches (‘women’s religious experiences were and are different from men’s because women are inherently more religious/emotional/superstitious than men’) are rapidly rejected. So, too, are various versions of the ‘deprivation’ or ‘compensation’ model (‘religion – particularly ecstatic religion – compensates women for the deprivation or inadequacies of their real lives’). As has long been recognised, many of the women most heavily involved in, for example, early Christianity came from the most privileged strata of society, at least in the economic sense. As Kraemer points out, if you react to this problem by making deprivation a merely subjective term (‘I feel myself to be deprived’), you end up with such a universal category that it is of hardly any help in the analysis of different religious movements. Kraemer opts, in an unnervingly wholehearted way, for a version of Mary Douglas’s old system of social and religious classification: ‘grid and group’.

Mary Douglas mapped all the forms of social organisation onto two axes. The first, the ‘grid’ axis, represented the degree of regulation that individuals experience in their daily lives – from the ‘strong grid’ life of the army, for instance, with its fixed hierarchies, rules and rigid symbolic codes, to the ‘weak grid’ life that is characterised precisely by the absence of all those things. The other axis (the ‘group’) measured the degree to which individuals feel themselves to be part of a social community and the degree to which that community imposes conformity (of thought, action and behaviour) on its members. So the ‘strong group’ life of a small commune is contrasted to the ‘weak group’ life of – at its most extreme – the solitary hermit. For Douglas, this system was a convenient tool in classification, as every society could be assigned its place on the axes: modern, free-market capitalism turning up as ‘weak grid, strong group’, the society of 19th-century plantation slaves as ‘strong grid, weak group’, and so on. It also enabled Douglas – or so she imagined in over-optimistic moments – to plot and predict the kinds of religious experience and belief that would match up with the secular, social organisation at each of the points on her axes. Kraemer tries to extend Douglas’s scheme to include the religious experience of the ancient world – and particularly the experience of women.

Kraemer reasonably supposed that the everyday social organisation of most women in Antiquity would fall into the category of ‘strong grid, weak group’ (like plantation slaves: highly regulated, but with little sense of group identity). Religion, in contrast, offered women various alternative patterns of social involvement. In the ecstatic cult of Dionysus, with its image of wild bands of women roaming the hillsides, there was a total and flagrant inversion of the norm: ‘weak grid, strong group’ replacing the ‘strong grid, weak group’. But other cults, like the worship of Isis, offered a more moderate alternative – ‘somewhat weaker grid, slightly stronger group’ – or inverted only one axis. Some of the women’s rituals in the state religion of Rome, for example, acted to strengthen the group axis, while not breaking down the hierarchical distinctions of the everyday ‘strong grid’.

Kraemer uses this classificatory scheme adeptly to point up the distinctions between the various cults that claimed women’s involvement, and to reassert and redefine the differences as well as the similarities that marked women’s religions. This is particularly important in her long discussion of the role of women in early Christianity, where she explores the differences in grid/group structure between the sects and sub-groups of the Jesus movement. She asks why a religion whose earliest stages ‘espoused an egalitarianism that had significant consequences for women’ turned so rapidly to the legitimisation of an extreme form of misogyny and wonders whether the answer was the ‘weak grid’ structure of several of the earliest Christian communities giving way to the ‘strong grid’ organisation of the established Church – whose insistent rules and hierarchies acted almost inevitably against the autonomy of women.

There are, however, more problems than advantages in this dependence on the grid/group scheme. It is not just the crudity of Douglas’s system or its extravagant (and widely disputed) claims to be a universal model for human society. Nor is it just the suspicion that Kraemer’s vision of the alternative structures of women’s religion is really the old ‘compensation’ model coming in again through the back door, dressed in a few new (or not so new) theoretical clothes. The central problem lies in what any model of this type almost inevitably conceals from view: the nature and definition of the ‘women’ that it takes as its subject.

Part of the difficulty lies in the complexity of ‘women’ as a historical and social category: rich women or poor? City-dwellers or peasants? Fourth-century BCE Athenians or fourth-century CE Romans? It is not just a problem of complexity, however, nor of the difficulty of deciding exactly which women we are talking about. It is more a question of how to understand the whole nature of the relationship between women and religion. It is easy enough to imagine, as Kraemer and most other writers do, two separate agents here: a pre-defined gender category (‘women’) acting upon, interacting with, and acted upon by the social forces that make up ‘religion’, religious institutions and beliefs. But what happens to their line of argument if we question the gender category itself, if we see the definition of women and their roles as unfixed, fluid, constantly under debate, never pre-defined? And, more important, what happens if we see religions and religious institutions as one of the most powerful mechanisms used by ancient cultures to negotiate that fluidity, to define ‘what it is to be a woman’.

If we see the problem in those terms (as to some degree we must), it then follows that there is an earlier stage in the argument that Kraemer and most other writers have missed out: not how women handled, viewed or reacted to religious institutions, but how religion had already operated to define their gender and to construct their view of themselves as women. We would all agree that it would be inadequate to analyse the involvement of blacks in contemporary British society and politics without at the same time considering how British society and politics acted to construct a model of race, ethnicity, racial separation or integration. In the same way, Kraemer does only half the job because she fails to stress that gender is constantly being defined within religion; religion and gender are mutually constitutive, not separate poles.

What would a different kind of argument about women’s religions in Antiquity look like? How could we start to recover those mutual relationships between religion and gender? Part of the answer might be simply to cast the net wider than Kraemer has chosen to do. She follows the standard procedure in studies of this type: cherchez la femme. Amid the confusing debris that makes up the ‘evidence’ for ancient religion, she tracks down and reconstructs (often with great scholarly expertise) the activities of that half of the ancient world who were biologically female, but passes over the wider problem of gender construction and its difference from biological sex. And she gives little thought to the whole range of strange (to us) and transgressive gender options that all ancient religions insistently presented: men who were not really ‘men’ (the castrated priests of the Great Mother Goddess, or the effeminate Dionysus); women who were not really ‘women’ (the vestal virgins whose symbols and privileges made them at least partly ‘men’, or the ‘male’ goddess Athena): men (like the Roman priest of Jupiter) whose rules of religious conduct made them inimitable paragons of the Roman husband, the paraded limit of what – in an impossibly perfect world – a Roman man might be.

The first and most pressing question in any study of women in ancient religions must be to ask how these wildly conflicting gender constructions operated within ancient societies; and secondly, how they intersected with the ‘real-life’ religious choices of the biologically female members of those societies. What difference did it make that right in the centre of Rome (just by the back door of the imperial palace, in fact) you could see the self-flagellating eunuch priests of the Great Mother – with long (‘female’) flowing locks and brightly coloured robes? What difference did it make that Roman satirical poets constantly harped on the licentiousness of the priests, their unrestrained sexuality, their passion for oral sex? How did the transgressive display of these eunuchs affect the Roman observers’ view of their own sexuality? What difference did this parade of castration at the centre of the cult of the Great Mother make to the ordinary male or female adherents of that cult? And what happened when they saw these eunuchs standing shoulder-to-shoulder with the quite differently transgressive vestal virgins – men who were not ‘men’ standing next to women who were not ‘women’?

Roman religion (like most ancient religions) constructed within itself a dazzling and puzzling range of gender images. The unchallengeable authority of the gods licensed the parade of quite extraordinary transgressions, against and around which Romans debated and defined what male and female roles were, and should be. For all its good sense, for all its brave attempts at a theoretical overview, Kraemer’s study misses out the crucial question of how Roman religion played its part in defining what a ‘woman’ was.