The debate amongst marxist feminists and socialist feminists about sex and class—that is, about the relationship between gender ideology and the material basis of women’s oppression and whether patriarchal structures exist independently from those of capitalism—is presently being pursued in two journals, New Left Review and Studies in Political Economy.footnote¹ All the participants agree that women’s subordination preceded capitalism and none explicitly espouses a dual systems approach, although it is the contention of Brenner and Ramas in New Left Review and of Pat and Hugh Armstrong in Studies in Political Economy that Michèle Barrett has in fact fallen into this trap. The major departure in the work of Brenner and Ramas and the Armstrongs to identify the material basis of women’s oppression lies in their attempts to theorize the biological component of sexual difference. Like Michèle Barrett (in her response to Brenner and Ramas) I am not wholly convinced by the attempt. If, as Barrett says, feminists have been ‘squeamish’ in the face of biologyfootnote² it is largely because of the difficulty they experience in skirting the mine-field of biological determinism. The arguments of these authors do little to reassure. Both sets of authors use women’s reproductive capacities to explain sexual divisions under capitalism. The Armstrongs argue that capitalism is ‘premised on free wage labour and on the separation of most aspects of workers’ reproduction from the productive process’,footnote³ and Brenner and Ramas that it was ‘likely if not inevitable’ (p. 48) that in the harsh circumstances of the nineteenth century women would take responsibility for children and domestic labour. Biology thus appears less and less the merely limiting factor the Armstrongs would have it be.

It appears to me that the approach of both Brenner and Ramas and the Armstrongs is not as productive as it might be. For as they all agree, irrespective of patriarchy’s origins, it is now part of the same system as capitalism. The Armstrongs suggest: ‘Women are simultaneously subject to capitalism, male dominance and their bodies. To pose the question in the form of alternatives is like asking whether ideas or material conditions structure women’s subordination. They are inseparable. They act together. Patriarchy and capitalism are not autonomous, nor even interconnected systems, but the same system. As integrated forms they must be examined together.’footnote⁴

And Barrett has written in her nlr response: ‘So we cannot counterpose ideology on the one hand and women’s economic situation on the other, because to do so is to ignore the degree to which the two are analytically related.’footnote⁵ This suggests to me, and I think the historical evidence supports it, that the main project at this stage must be to explore further the complex dynamic of the relationship between gender ideology and class. However, Pat and Hugh Armstrong have signalled their impatience with this approach when in their latest contribution they express dissatisfaction with Barrett’s idea of gender ideology as ‘a pre-capitalist vestige taken over and shaped as capitalism developed historically in various social formations’, and ask ‘when if at all, a more general explanation will emerge. Will it only be when in positivistic fashion enough bits of historical evidence have been accumulated?’footnote⁶ I accept that historical evidence is not necessarily binding because of its partial nature, but both Brenner and Ramas and the Armstrongs do take history seriously and in the case of the former have chosen to construct their argument firmly on the basis of empirical historical evidence.

I would maintain that it is difficult to see how much more progress can be made in building theory without a more rigorous attention to the historical evidence that is being used. In particular, enough work has been done, chiefly by socialist feminist historians, to indicate that the structure of Brenner and Ramas’ argument cannot adequately be supported, and I hope to demonstrate that the development of a materialist theory of gender ideology can proceed only after some modifications to their interpretation. In what follows I shall take in their original order Brenner and Ramas’ three areas of discussion: the role of trade unions and protective legislation in the formation of the family-household system; the importance of biological reproduction in explaining women’s position in family and workplace; and women and the welfare state.

Brenner and Ramas start their argument by accepting the importance of the family-household system for our understanding of the sexual division of labour, but they reject the idea that its emergence can be explained in terms of the introduction of protective legislation and trade union exclusive practices which, in their reading of Barrett, account for women’s precarious place in production and hence their dependence on men and their domestic role in the family. Almost certainly, protective legislation as an explanatory variable will not bear the weight of this argument, but it is difficult to see that Brenner and Ramas’ reading of Barrett is correct. In Women’s Oppression Today, protective legislation is treated as one of many strategies pursued by the labour movement in bargaining with capital. There is undoubtedly room for further research on its effects, particularly in the late 19th and early 20th centuries. The old account by Hutchins and Harrison, which Brenner and Ramas rely on, is undoubtedly the most exhaustive we have, but we need to be aware that its interpretation is Fabian and that in its day it was very much a part of the contemporary debate between liberal and socialist feminists over the issue.footnote⁷ The role of male trade unionists, however, is more difficult to assess. Brenner and Ramas contend that, ‘it is entirely unnecessary to resort to ideology to explain why trade unions were particularly adamant in their opposition to female entry into their trades’ (p. 45). Of course, male unionists were right to fear that women would undercut their wages. The Women’s Protective and Provident League, formed in 1874, was also aware of this problem, and thus when three women mule spinners were introduced at Lostock Mill on women’s rates of pay, and the mule spinners stopped taking women piecers as a result, the League took no action.footnote⁸ But unions were very slow to move away from a policy of exclusion towards one of organizing women workers. The statement of the traditional, liberal, craft trade unionist, Henry Broadhurst, to the tuc in 1877 that male unionists should use their ‘utmost efforts to bring about a conduct of things where their wives should be in their proper place at home’footnote⁹ is well known. But there is also Tom Mann’s statement, as an organizer of semi-skilled and unskilled workers, to the Royal Commission on Labour that he was ‘very loth to see mothers of families working in factories at all’, adding that he considered, ‘their employment has nearly always a prejudicial effect on the wages of the male worker.’footnote¹⁰ Here a clear ideology of women’s proper place and the fear of female competition are to be seen reinforcing one another, which suggests that an analysis of the production and dynamics of gender ideology (missing from Barrett’s work, as Brenner and Ramas point out) is crucial.

Where women were accepted as union members, they tended to be confined to special sections paying lower contributions and getting lower benefits. In the Bolton Amalgamated Bleachers’, Dyers’ and Finishers’ Association before World War I, men qualified as full members, boys as half members and women as one-third members.footnote¹¹ It was usually assumed that women would not play an active role in union affairs. The Blackburn Association of cotton weavers provided that no ‘person’ would be eligible for office in the Society if ‘his wife and children’ were not trade union members.footnote¹² Joanna Bornat has shown how women’s relationship to trade unionism and the workplace more generally was mediated by the institution of the family;footnote¹³ and it is clear that women’s domestic role, whether actual, or in the case of young women, anticipated, did place real limitations on their capacity to organize and on their expectations: most women accepted that because of their extensive domestic burdens their wage-earning would be secondary and their status at the workplace, therefore, inferior to that of men. But it is also necessary to explain, as Brenner and Ramas do not, how it is that the assumption that women will occupy a secondary place in the labour force persists in the minds of large numbers of women and even larger numbers of men after World War II, when women’s domestic burdens eased. The beliefs and behaviour of male unionists suggest first, that opposition was based on more than just fear of undercutting, and second, that their particular concern to prevent married women’s work was grounded in the belief that it was improper and unnatural. Brenner and Ramas are quite correct to assert that gender ideology is rooted in women’s and men’s actual experience and not merely imposed and internalized, but that experience has different meanings for working-class men and women, and their expressed consciousness of gender has, therefore, been different. Sally Alexander makes a similar point in a recent issue of History Workshop Journal. She argues that the experience of class is not always a shared and even one, and believes it necessary to use psychoanalytical tools to probe the subjective meaning of class for men and women through language.footnote¹⁴

Brenner and Ramas maintain that in the 19th century the harsh material realities of biological reproduction were such that sexual divisions were not negotiable and that there was little possibility of any outcome other than that women would bear primary responsibility for children and domestic labour. They also suggest that because of the extent of women’s domestic burdens, this sexual division of labour secured the greatest possible welfare for members of the working-class family (p. 51). This analysis has the great merit of recognizing that ‘the bourgeois family ideal’ was indeed a shared ideal; it is important to determine what was shared as well as what was not. Women labour leaders were as keen as men to see married women withdraw from the workforce, and Elizabeth Robert’s oral evidence from the North West has indicated that women who were forced through the sickness, unemployment or neglect of their husbands to do two full-time jobs, paid and unpaid, were pitied by thier neighbours.footnote¹⁵ As Ellen Ross has observed, the marriage contract between working-class men and women did not enjoin romantic love or verbal or sexual intimacy but rather required financial obligations, services and activities that were gender-specific.footnote¹⁶ Married women expected to manage the household and keep it together, caring for children and husbands and taking paid work as, and when, the family economy required it.