The problems with the study of the family by Europeanists are two-fold. Firstly the terms they use like ‘family’ are often vague and unsatisfactory for analytic purposes—though they may serve as general signposts. Secondly, there is little comparative perspective. Yet this is needed not only to define terms but more importantly because every statement about the relation of the ‘nuclear family’ to capitalism or the invention of childhood is making a claim not only about Europe but about the rest of the world.

The history of the family in Europe has become tied in with notions of that continent’s unique achievements, especially in terms of ‘capitalism’ and ‘modernization’. The knot needs to be untied, not to weaken the relation between production and reproduction, which Wally Seccombe emphasizes in his two books A Millennium of Family Change and Weathering the Storm, but to rationalize at least the pre-industrial situation by reassessing its uniqueness at all levels in relation to other regions of Eurasia.footnote1 When we do that, the strong association between certain ‘family’ variables and the development of mercantile capitalism—and modern systems of knowledge—in Europe becomes much more attenuated.footnote2 Later on, the rapidity and innovatory character of large-scale industrial production did dramatically change the domestic lives of the masses, although those changes had and have their parallels in other parts of the world where there has been a sudden growth in urban living, with all the usual consequences of crowded accommodation, dependent production, the weakening role of extended kin groups—though not necessarily of extended domestic ties—relative anonymity, and looser community controls. However, here, as always, it is the nature of the link that is at stake. Was it the same for all classes, for both sexes? Were there not continuities as well as discontinuities that related to the different social domains of production and reproduction which allowed them a degree of autonomy? Did political and economic hegemony, both of which demand their own space, always dominate the family?

Wally Seccombe has written a comprehensive two-volume history of the family in north-western Europe from the early medieval period to the present, trying to link the systems of reproduction with those of production, and offering a critical perspective on much previous work in the field. It is the most valuable summary to date of the situation in this crucial part of the world. Some of Seccombe’s basic assumptions about its role are built into to his general Western models, but at another level his reassessment of the work of others is a signal success. Starting from a Marxist perspective, he concentrates on the transitions from feudalism to mercantile capitalism, to proto-industrialization and then to industrial capitalism, attempting, in the process, to do what Marxists have signally avoided doing in the past, that is, to reconcile demographic with social history. This division has tended to give us two kinds of family history: the harder variety centres upon numerical statements about the size of ‘families’—in fact, households—and the softer on changes in mentalités. The history of the family must be built around both—the numerical material buried in parish registers, exploited by the Cambridge Group, as well as the diaries and legal documents that form the basis of the work of Stone, Ariès, and others. Lacking was any serious attempt at drawing together both these threads and their insertion, in any rigorous way, into the wider context of the major social changes through which the West has passed. These tasks Seccombe confronts head-on. In addition, he claims to take account of women’s history, though here the force of his claim is not quite so strong.

Seccombe pursues this massive undertaking with very considerable scholarship and many insights. It is one of the great strengths of his survey of the history of the north-western European (mainly British) family that he consistently treats domestic groups not only as reproductive units—for the workforce as well as for kin-groups—but as productive ones, in pre-industrial societies where domestic groups usually work together but also in industrial ones where he sees reproduction as linked to productive processes mainly through wage incomes. His study will remain a source of reference and debate for many years to come. While other historians have attempted to relate factors such as the age of marriage to the price of corn, or ‘modernization’ of the family to the conditions of the market, he tries to associate the economy and the family in a much more comprehensive manner. To this end, he uses a version of the broadly Marxist concept of the mode of production, trying to link relations of kinship and marriage to productive relationships, though not in any absolutely determinative way.

There is no doubt that such an exercise needed to be undertaken. Seccombe has done so in a manner that summarizes a great deal of research and throws much light on changes in the family. If I do not see this work as entirely successful, that is partly because success is difficult to achieve in such a contentious field and partly because in dealing with this subject one needs to look outside Europe even to understand what happened within that continent. If Seccombe’s account does not always command immediate assent all around, that is partly a comment upon the nature of a much disputed territory which conceals a minefield of hidden agendas and ethnocentric prejudice. The family has been seen as contributing to, or the consequence of, a transformation of the economy and society that is considered to have taken place only in England, in north-western Europe, or possibly more widely in Europe. That is, the question of the structure of the family—demographically, developmentally, emotionally—is seen as part of the Uniqueness of the West and the European Miracle. Coming at the problem—like the sources he is dealing with—from a determinedly European standpoint, Seccombe is less critical of some of those assertions than he should be.