Wally Seccombe’s ‘Marxism and Demography’ provides a welcome and insightful synthesis of Marxist perspectives and recent advances in demography and family history.footnote1 Seccombe rightly takes Marxist theory to task for relying on a sterile refutation of long-discarded Malthusian theory in place of addressing the material realities of changing population sizes and structures. Seccombe, further, provocatively states that differing modes of production are associated with different demographic patterns, and he outlines a model of the European transition period to demonstrate this contention. This integration of the demographic factor is an important advance over the Dobb-Sweezy-Brenner debates on the transition. Finally, Seccombe notes and criticizes the mutual lack of influence between feminist theory and demography. Clearly, we must attend to women as well as men as historical agents in order fully to understand changing demographic patterns. Equally clearly, feminist theory needs to consider women as individuals in variously structured populations in order to assess the changing meanings and options of their historical agency.

The implication of Seccombe’s analysis, then, is that if demography plays a crucial role in the transition, gender relations are as key to changes in the mode of production as are those of class. Seccombe, however, does not really follow out this implication. He instead posits demographic changes as solely the results of shifts in economic structures. We do not wish to deny a connection between the two, but rather to point out that Seccombe’s model does not adequately consider gender. After a critical introduction in which he calls for gender analysis, he returns in his actual model to the mainstream demographic assumption of the reproductive couple who engage in a consensual cost-benefit calculus in their reproductive decision-making.

Following out the implication of integrating feminist analysis into a Marxist demography, we would offer the following general correctives to Seccombe’s model. First, Seccombe, in positing that Europeans in the transition followed a neoclassically derived cost-benefit analysis in reproductive decisions, bans cultural variation from his model. In so doing, he rules the analysis of changing cultural constructions—of gender, childhood, parenthood, family—out of court. The recent scholarship on changing constructions of gender and ‘family sentiments’ is rich and should not be ignored. For example, the nineteenth-century rise of the ‘moral mother’ and the related cult of domesticity in the West, with their emphases on women’s religiously, tinged responsibilities for newly envisioned helpless and vulnerable infants and children, surely entered into women’s fertility decision-making—and perhaps gender struggle over fertility.footnote2 Seccombe recognizes ‘the problematic nature of macro-aggregation’ (p. 33). But he needs also to consider the problematic nature of micro-aggregation—fertility decisions are often the subject of gender-based dispute.

The consideration of cultural variation, then, brings in questions of differential male and female perspectives on fertility and their variation across time and culture. It also has a more obviously material implication: while Seccombe considers the ‘cultural conditions shaping the relation of marital sex to procreation’, he does not see that these conditions may determine that decisions are subject to more than the ‘prevailing relations between spouses and . . . the institutions of a given community’ (p. 31). The space between ‘spouses’ and ‘community’ may contain a number of kinspeople of cultural significance to fertility decision-making: e.g., potential grandparents, aunts and uncles. We must not reduce the historical and cultural variations in the decision-making unit by conceiving it through a nuclear family model that describes very little of even contemporary kinship reality.

This emphasis on the cultural constructions of gender and fertility decisions allows us to posit that there is no single kin and fertility arrangement necessarily associated with a given mode of economic organization. There is a certain ‘relative autonomy’ between kinship organization and the way in which use values are produced. Being open to this possibility allows us to escape the proto-industrial determinism of Seccombe’s transition analysis. The alternative is to begin any historical investigation with a thorough study of the already existing gender and class relations before tracing out their impact on each other.