88 SOCIAL SCIENTIST questions in understanding the social divisions of labour and the labour process. These questions cannot be confined within an analysis of the family or household; nor can they be understood within a restricted relations of pro duction model. Neither production nor reproduction models, however sophisticated, can address the questions which relate to the relations between these spheres of activity. Since both women and men "work and participate in households, the prime focus of enquiry at a micro and macro level should be the modes of integration between these. The Separation of Home and Work A further theoretical and empirical treatment of these questions would help to clarify the generally accepted notion of the separation of-'home and work. A reformulated concept of work would facilitate the separation of the ideology in such phrases as 'when man is at home, he is not working, when he. is working he is not at home" from the actual structuring of the working lives of men and women. Such a phrase has never been applicable to women nor, I would suggest, should it be treated as unproblematk in the case of men. To say that increasingly to earn one's liveihood meant going out of the household unit is one thing, but we cannot assume that this applied right across the board. Nor can we assume that the household ceased to be a workplace with the replacement of domestic industry by factory activity in some manufacturing industry. Analytically, the separation of the workplace from the site of the household is not necessarily co-terminous with capitalist relations of production. Historically, it can be argued that this has been the dominant trend in the development of capitalism. But coalmining and shipbuilding, for obvious reasons, were never household activities whatever relations of production existed. Agriculture has always been more problematic; the development of capitalist relations of production in agriculture existed from early days when day-labourers (particularly landless ones) hired (hcmselves out, but the position of those who hired themselves out for a year and became in some way part of the household of the farmer, a system existing until very recently, do not fall easily into the home and work separation. Where are the analytical boundaries to be drawn most usefully? The mid 20ih century conceptions of home, dominated by ideological constructions of domesticity and maternity (for women) and leisure and servicing (for men) cannot be assumed to be analytically useful for understanding work or the relations of production. Empirically the conception is questionable, and therefore it must be questionable theoretically. The official records such as census data, reflect a changing categorisa-lion of work/economic activity and can not be taken as a description of the actuality of household arrangements. We sliould use such data with extreme care in our attempts to understand the relations between households and the external economy and the economic relations within households. Marx argued that the domestic system of production would disappear in the logic of capitalist production. Bythell, writing in the late 1970s, saw its disappearance at the beginning of (Ills century as due to the inequities of the sys-