The pro to‐industrial family economy: The structural function of household and family during the transition from peasant society to industrial capitalism

 The direction and extent of research into the history of  the family has been heavily influenced by the work of P. Laslett. This holds true not only for the favourable reception which Laslett's approach has experienced, for even his critics5 are in many respects indebted to him. So far, progress has followed two main paths of development. 1. In the theoretical area, there has emerged an historically oriented critique of the traditional social-scientific axioms concerning household and family. 2. In the methodological and technical area, there has been the deliberative accumulation of quantifiable data. The important new insights that have been gained by following these two paths should nevertheless not obscure the fact that the claim presented by Laslett that the new direction of research should be above all 'social structural history',6 oriented toward entire societies, examining the essential 'structural function of the family in the preindustrial world',7 remains fundamentally unfulfilled. In the course of Laslett's empirical work, theoretical criticism and methodological innovation did not achieve a positive reciprocal relationship that would have made his research praxis fruitful in other systematic endeavours, and so have helped build a substantial model and theory yielding new insights into the 'structural function of the family', and its transformation in the transition from traditional agrarian society to industrial capitalism. Unfortunately, with Laslett, theory formation and empirical investigation drifted apart as research progressed. After a promising, if not unproblematical start in Laslett's first book, The World We Have Lost, systematic and theoretical interests on the one hand and the practice of research on the other, began increasingly to impede and hinder one another. To be sure, the old hypothesis of the evolution from the large, extended, multigenerational household of the pre-industrial age to the nuclear family of the industrial age has been convincingly refuted by Laslett.8 However, Laslett's historical rejection of this traditional myth of social-scientific evolutionary theory did not lead him to the construction of a substantial theory which would have allowed a more precise location of household and family as functional elements and social-structural factors in the genesis of industrial capitalism. Laslett's 'Null Hypothesis' which has been tested cross-culturally and over time, but seems to apply above all to England, shows the nuclear family already to be the dominant household type before the industrial revolution. This hypothesis destroys the assumption that there is an historical covariance between industrialization and the formation of nuclear family structures.9 His proof that this constellation of a dominant nuclear family household type was relatively constant, not only before, but also during and after the industrial revolution, has called into question important assumptions of industrialization and modernization theory. These assumptions, like those of Parsonian structural-functionalist theory, originated in the premise of a symmetrical convergence or passive adaptation in which changes in family structure relate to the social and economic transformation of society.10 Laslett's findings reveal household and family as a relatively resilient and enduring structural element within the genesis of industrial capitalism.11 It is here that Laslett's systematic achievement, frequently overlooked by his critics, is to be seen, but he himself draws no systematic conclusion from these findings. For the categorical framcvork of his research offers no basis for a dialectical theory which would do justice to what E. Bloch referred to as 'gleichzeitige Ungleichzeitigkeit' (synchronous anachronism),12 which designates household and family as bearers of residual traditional structures in relation to those larger social and economic transformations characterizing the process of capitalist industrialization. It is only within this context that the structural function of household and family in the transition from traditional agrarian society to industrial capitalism can adequately be assessed. Any concrete undertaking along these lines would have to be a modest one. For the time being, it would be unreasonable to pursue a research strategy which would try to offer an alternative to the older universalistic approaches of the evolutionary theory or to the current Anglo-American concepts of modernization and industrialization.13 Rather, efforts to understand the dialectical nature of household and family during this fundamental transition to industrialism should aim at the formation of a model or theory based on a research terrain of manageable size, specified by status and region, for it is primarily in this way that an empirically oriented historian can offer fruitful theoretical hypotheses. If, through the choice of his methodology and research techniques, the historian peremptorily eliminates the opportunity to pursue this dialectical question, the result will be unfortunate not only for his own research, but also for the desired critical dialogue between historians and social scientists, especially in the area of the sociology of the family.14 For if the historian of the family leaves the formulation of conceptual questions to the social scientist, he conforms, whether he likes it or not, to the outmoded division of labour between history and the social sciences which traditionally limits the historian to the role of the 'supplier of data', and grants to the social scientist the role of 'producer of categories'.15 Moreover, the historian also unwittingly helps to perpetuate the persistence of old approaches and questions behind his own back, even when he acts, as does Laslett, as their critic. It is as a consequence of this lack of integration between conceptualization and empirical research that there may be seen in Laslett (but not only him) a trend of research that lays too much stress upon the isolated small group.16 This trend may be defined by its proponents as ' micro-social history ' but due to its inherent assumptions it cannot be carried out as 'functional micro-social history' within the framework of macro-historical questions and problems. In certain respects this orientation is indeed a necessary point of departure for historical research into the family. It is suited to a research situation in which the empirical 'unknowns' still outweigh the 'knowns' on the map of our knowledge. But where these tendencies harden themselves into a methodological and theoretical stance which proceeds on the assumption of a strict separation of empirical, antecedent data accumulation on the one hand, and the successive framing of materially substantiated hypotheses on the other,17 not only will one fail to 'minimize the unanticipated theoretical consequences of methodology',18 but there will also emerge serious disadvantages for the conduct of concrete historical research. These disadvantages result precisely from a situation in which the unreflected categorical assumptions establish themselves as a seemingly timeless framework of research and guide the work of the historian in the wrong direction. The problems with this approach manifest themselves in the case of Laslett's 'formal and restricted concept of structure"9 of household and family, a concept which restricts itself all too easily to kin relations and generational succession as decisive structural criteria, to which is then added co-residence as the only delimiting criterion identifying the household unit. It is indeed doubtful whether the employment of this structural concept of household and family makes it possible at all to render significant findings in the sense that 'social-structural history' demands. In the interest of too rigorous a claim for quantification, Laslett the methodologist sacrifices a priori and categorically the historical 'meaning' of those phenomena to be measured and compared in favour of a universal scale of comparison and measurement. If Laslett's structural criteria are used in isolation, without considering the necessary 'contextual identifications'20 which must be derived from the changing socio-historical conditions under which the family produces, reproduces and consumes, and if those structural criteria by way of definitions of 'nuclear family', 'extended family' and 'multiple family' become the sole basis for comparison and measurement, then the danger arises of computing the incomputable. It is true that the industrial proletarian grandmother may have lived in an 'extended family' as did the peasant grandmother, but this apparent uniformity by no means indicates an identity of household structures. The 'extended family' of the proletariat primarily functioned as a private institution to redistribute the poverty of the nuclear family by way of the kinship system. The extended family of the peasant, on the other hand, served as an instrument for the conservation of property and the caring for the older members of the family.21 However, it is not only that comparison and measurement are blunted by Laslett's 'restricted concept of structure'. The structurally relevant 'basic processes'22 of household and family, as they appear in production, reproduction and consumption, cannot be grasped using Laslett's structural criteria; neither can the function of this familial unit within the larger socio-economic context of the entire society. The categorical preconceptions and restricted methodological perspectives reduce the relevance of Laslett's approach towards the social history of the family in a decisive way. It is in view of this lack of empirical differentiation and the narrow theoretical scope of such a 'restricted concept' of household and family that it becomes necessary to develop different concepts, models and definitions based on a research area of middle range : small enough to enable deep analysis, yet large and typical enough to form the basis for generalization. These models and definitions should serve to analyse the  changing function of household and family in the social context of production, reproduction, as well as power relationships, and in addition to determine the repercussions of social and economic changes on family structure. In the following pages I hope to make such a preliminary attempt with a limited but significant historical example

I I. FUNCTIONA L INTERRELATIONSHIP S AN D TH E REGULATIN G SYSTE M O F TH E PROTO-INDUSTRIA L FAMIL Y ECONOM Y

The model whose main characteristics are sketched hereafter, and which has been more fully described in another context,23 attempts to outline the function of household and family during a critical phase in the transition of certain types of agrarian societies to the industrial system. This phase was characterized by the emergence, expansion and final decline of rural industries. The main period of this 'protean stage of industrial development',24 'industrialization before the factory system'25 or 'protoindustrialization',26 as it has come to be called, may be dated from the sixteenth to the early nineteenth century. The development of proto-industrialization indeed varied according to region and craft. But amidst all these differences it exhibited - apart from special developments in the mining and métallurgie trades - a common structural foundation. This consisted in the close association that existed between household production based on the family economy on the one hand, and the capitalist organization of trade, putting-out and marketing of the products on the other. Whether the home-industrial weaver, knitter, nailer or scythemaker entered the market as buyer and seller himself and so worked within the 'Kaufsystem', or whether he was organized in the 'putting-out system', he was always directly or indirectly dependent upon merchant capital. The functional interrelationship between family economy and merchant capital, and the peculiarly stable and at the same time flexible character of this configuration, constituted a comprehensive set of social relations of production, giving to the historical process of proto-industrialization the traits of a socio-economic system. The historical significance of this specific association of proto-industrial family economy and merchant capital emerges if one considers the social and economic conditions in which proto-industrialization originated. On a macro-historical level protoindustrialization appeared as the combined outcome of the destabilization and decomposition of traditional European peasant societies. Demographic growth and socio- economic polarization of the rural population led, above all in the high middle ages and in the sixteenth and eighteenth centuries, to the emergence of a numerous, underemployed class of small peasants or landless rural dwellers. This process formed an essential pre-condition for the penetration of industrial production into the countryside. Declining marginal returns in the small peasant or sub-peasant economies left only one alternative open to the rural dwellers: the part-time or full-time transition from landintensive agrarian production to labour-intensive craft production. From this viewpoint proto-industrialization could be called a special case of 'economic development with unlimited supplies of labour.'27 The transition of this under-employed rural labour force to the mass production of handicrafts could only materialize under the impact of a second essential factor: the emergence of a world market, extending beyond the confines of Europe, dominated by merchant capital. It was thus less the internal demand than the expanding demand from foreign markets which developed with the 'new colonialism' of the seventeenth century,28 which brought about industrial mass production in the countryside. In view of the low elasticity of supply of the town economy confined within the guild system, merchant capital was forced to fall back on to the potential of the rural labour force, and to transfer craft production to an increasing degree to the countryside. The emergence, rise and final agony of rural-handicraft industry cannot, however, be explained within this macro-historical framework alone. The macro-historical perspective needs to be supplemented by a micro-historical viewpoint. For proto-industry was domestic industry. It was closely tied to the inner dynamic which household and family of the rural artisans generated within a context increasingly determined by market and monetary relationships and by the capitalistic organization of trade, putting-out and marketing. As was the case with the peasant economy, the rural-industrial production process rested to a large extent on the household economy of small producers. Proto-industrial and peasant households did not, however, share only the common characteristic of a mere form of production. Beyond this common form both shared the functional and organizational unity of production, generative reproduction and consumption within the social formation of the 'ganze Haus'.29 To be sure, this unity lost its autarky as a self-subsisting economy within the 'ganze Haus' of the rural artisan. With the loss of their agrarian base, production and consumption increasingly became factors dependent on the market. But nevertheless, the family economy and 'ganze Haus' remained an effective socio-structural force even after their original agrarian base had largely disappeared. Indeed it was the inertia of the traditional family economy as a self-regulating unity of labour, consumption and reproduction which determined the central functional role of the household within the process and systemic pattern of proto-industrialization. The structural and functional link between proto-industrialization and household and family can most clearly be explained by the characteristic logic of the pre-capitalist family economy. This was first analysed empirically at the beginning of the twentieth century by A. V. Chayanov and initially presented in his 'Théorie der Familienwirtschaft im Landbau '.30 The heuristic fertility of Chayanov's analysis, originally concerned with the patterns of peasant society, can be seen clearly in its application to its proto-industrialized variant: for Chayanov the central feature of the characteristic economic logic of the family economy was that its productive activity was not governed by the objective of accumulating a monetary surplus or a net profit. 'The family economy could not maximize what it could not measure.'31 The object of its productive labour, rather, was to bring into equilibrium, into a ' labour-consumer balance ',32 the basic necessities of economic, social and cultural subsistence on the one side, and the expenditure of labour by the family on the other. Consequently, the family economy tried to maximize the gross product, not the net profit. It entered into exchange relationships as a producer of use values, even under circumstances in which its products necessarily became commodities whose exchange values were objectively determined by money relationships and merchant capital. If, for example, the returns of the family economy fell, it increased its expenditure of work, even beyond the amount which is customary in an economic system depending on developed capitalist wage-labour. In developed capitalist systems the interest of capital in exploiting the labour force is regulated and at the same time limited by the necessity to reproduce that labour force permanently. For the family economy, by contrast, the work effort, regardless of the amount of labour expended, virtually presents itself as an invariable overhead cost because of the lack of alternative possibilities for employing its labour force. It implies 'zero opportunity costs' as long as the family subsistence is not guaranteed and the possibility of marginal returns to labour exists. This holds true even when the economic returns would yield a deficit in the framework of a net profit calculation which would be based on comparable income scales for wage labour. The net return calculated on this basis would appear to fall below the family's cost of production.33 If, on the contrary, the returns of the family economy rise, for example because of more favourable economic conditions, the family economy has no need to increase its expenditure of work, rather it converts these additional returns into consumption and leisure; and a backward-sloping-supply-of-labour curve sets in. Thus although the behaviour of the family economy is influenced by the external conditions of production, it is not entirely determined by them. Rather, it depends in the main on the balance between production and consumption within the family. This labour-consumer balance does not originate, however, merely in the subjective preferences of the family members. Its logic is determined by socio-structural factors. It rests on the structural and functional nexus of the 'ganze Haus' as a self-regulating socio-economic formation, which organizes and combines production, consumption and reproduction through the common labour relations of the family members. Although this self-regulating system of the family economy was originally adapted to the needs of a peasant-artisan subsistence economy, it did not lose its effectiveness in the transition to proto-industrialization. On the contrary, the logic of family economic production became effective above all because of the inclination of the poor, landless producers to fall back on 'self-exploitation'34 in the production of craft goods, if this was necessary to ensure customary family subsistence and economic self-sufficiency. This dynamic had a macro-economic effect in reference to the emergence, the progress and also the internal contradictions of the proto-industrial system, above all in that the mechanism of self-exploitation by the family enabled the merchant or putting-out capitalist to realize a specific 'differential profit'. ' Competition enables the capitalists to subtract from the price of labour what the family produces in its own garden and small plots.35 The 'differential profit' realized in this way surpassed both the profits that could be gained from the social relations of production in the guild system, and the profits that could be derived from comparable wage-labour relations in manufactures. The following hypothesis therefore seems justified: the primary social relation of production in the transition from traditional peasant society to industrial capitalism was established not n manufactures, but in the characteristic nexus of small and sub-peasant family economy and merchant capital. Manufactures in any event only fulfilled the role of a supplementary system of production. In the proto-industrial phase, as Karl A. Wittfogel suggested, 'a famishing Lilliputian cottage industry choked off large industry ',36 preventing its emergence as a dominant form of production. If one considers the producing family of the rural-industrial lower classes from this viewpoint, it appears as the essential agent in the growth of emergent capitalism. The family functioned objectively as an internal engine of growth in the process of protoindustrial expansion precisely because subjectively it remained tied to the norms and rules of behaviour of the traditional familial subsistence economy. From this perspective the dominant impulse in the genesis of capitalism was not so much the 'Protestant Ethic' and the labour discipline subjectively inherent in this ethic, simultaneously enforced by capitalist wage-labour. Rather, the dominant impulse seems to have been the 'infinitely tenacious resistance... of pre-capitalist labour ',37 anchored in the family economy, which M. Weber completely pushed to the edge of his consciousness.38 The origin of modem capitalism is not in any case to be separated from the specific function, which the 'ganze Haus' of the small peasant household carried out in the final, critical phase of its development which was at the same time the period of its demise. This insight into the symbiotic relationship of family economy and merchant capital possibly points beyond its own specific case. It illustrates the essential function which the preservation of pre-capitalist enclaves has had and still has for the evolution and stabilization of capitalist societies.39 The example points at the same time, however, to possible contradictions and results of such functional relations of evolutionary and devolutionary processes. The proto-industrial system in any case contained within its structural foundation the essential seeds of its own destruction. In an advanced stage of proto-industrial expansion, this structural foundation became one of the negatively determining causes for the transition to industrial capitalism or led proto-industrialization into its cul de sac, i.e. into deindustrialization or 're-pastoralization' (F. Crouzet). This contradiction was anchored to the regulating system of the family mode of production as an effect opposed to productivity and surplus. It brought into operation the reverse side of the 'labourconsumer balance ' of the family mode of production because it affected the replacement of productive labour effort through consumption and leisure, through feasting, playing and drinking in exactly those situations of potential growth in which the capitalist putter-out could have obtained maximum profits. It was this contradiction which in the long run could not be squared with the dynamic of reproduction and expansion inherent in the proto-industrialized system. So it led either beyond itself to industrial capitalism or retreated backward from proto-industrialization into de-industrialization.

III. HOUSEHOLD FORMATION AND FAMILY STRUCTURE AS ELEMENTS OF THE PROCESS OF PRODUCTION AND REPRODUCTION The nuclear family without servants was the predominant type of household in rural cottage industry.40 This scarcely distinguished the proto-industrial household from other rural groups during the period of the disintegration of peasant society. It was, rather, typical of the overwhelming portion of the sub-peasant and land-poor class. Nevertheless, if one controls for social class it is clear that the average household size of the rural cottage workers was significantly higher than that of farm workers.41 Previous analyses have shown that the decisive factor determining larger household size was the larger number of co-residing children.42 This did not result from a higher level of legitimate fertility among cottage workers' families nor was a reduced level of infant mortality a factor.43 The higher average number of children is rather to be traced to the earlier age of marriage and possibly to an altered pattern of age-specific mobility in proto-industrial regions. R. Schofield, D. Levine and L. Berkner44 have demonstrated in any case that the traditional status of servanthood as an age-specific precursor to adulthood amongst peasant populations largely lost its significance among the rural cottage workers. Children of rural cottage workers remained longer in their parents' house yet also married earlier than the members of peasant or sub-peasant classes. 'A family has a better bottom than formerly: residence is more assured and families are more numerous as increase of industry keeps them more together.'45 Work within their family of origin by the children of weavers, spinners and knitters frequently took the place of work as servants in another household. But this alternative was not a result of free choice. Child labour, which both in its intensity and duration went far beyond that of the corresponding labour of farm peasant households, was in fact a vital necessity for the rural cottage workers' families.46 The extent to which the material existence of the proto-industrial family depended on child labour as the 'capital of the poor man' becomes clear in those cases in which children made no direct contribution to the 'working income' of the family. In those cases where they left home, children frequently were 'hired out' as already trained workers or they remained bound to the family by having to make regular payments to it.47 The longer residence of young people in their parents' house and the relatively low age at marriage resulted in a higher average household size among rural artisans. However, family structure did not follow the pattern of the larger peasant stem-families. That is, the more compelling integration of the child into the family work force, its longer period of socialization in the parents' home, and its early marriage did not bring about a closer connection between the generations in the sense of providing a stimulus to form large, three-generational households. On the contrary, the increasing dissolution of the agrarian basis and the transition to a proto-industrial mode of production under market conditions rendered ineffective the very causes which governed household and family formation among traditional peasants
Among peasant populations the necessary connection of household formation to resources which were scarce and which could be acquired only by inheritance formed the decisive structural determinant. It enforced restrictive marriage patterns4 * as well as the co-residence of the generations in the ' ganze Haus \49 The iron ' chain of reproduction and inheritance '50 at the same time functioned as a system of ' reproduction and patriarchal domination '. By controlling access to land as the only complete source of subsistence, the older generation controlled not only the pre-conditions of family formation on the part of the younger generation, but through inheritance it also controlled the structural connection of the family beyond the individual family cycle.51 Household formation and family structure among cottage workers, on the other hand, grew out of fundamentally different pre-conditions. Inherited property as the 'tangible* determinant of household formation and family structure receded in the face of the overwhelming importance of the family as a unit of labour. The foundation and continuing existence of the family as a unit of production and consumption was no longer necessarily tied to the transmission of property through inheritance. It was replaced by the possibility of founding a family primarily as a unit of labour. This not only reduced parents' control over the marital relations of the young, but it also loosened at the same time the structural connection of the generations, in so far as it had been guaranteed by property inheritance and patriarchal domination. It is true that parents were dependent on the labour of their children to an increasing extent, but they possessed no sanctions against adolescent children who wanted to leave the house and found a new nuclear family unit. Marriage and family formation slipped beyond the grasp of patriarchal domination ; they were no longer 'tangibly' determined by property relationships, but they did not lose their 'material' foundation in the process of production.52 The ' beggars' marriages ' between partners without any considerable dowry or inheri-tance, between 'people who can join together two spinning wheels but not beds',53 were frequently criticized by contemporaries, and constitute evidence for the new conditions shaping household and family. They were based on an increasing exploitation ( Verwertung) of the total family labour force. As Martine Segalen has demonstrated, the extraordinarily high rate of socio-professional endogamy among weavers54 in developed proto-industrial regions shows that household formation of rural artisans depended decisively on the highest possible work capacity of both marriage partners. The practical ability of the woman to work as an artisan before marriage determined her value as a marriage partner even more than her background as indicated, for example, by her father's occupation, property or social status.55 ' The better the weaving maids can weave, the better able they are to find a husband.'56 The new objective conditions of exploiting family labour in rural cottage industry required the choice of marriage partners who possessed technical skills; in this way, these objective factors allowed, subjectively, a more individualized selection of partners. Moreover, the same conditions demanded the formation of a new family economy as early as possible in the life cycle of young men and women.57 Maximum income opportunities were based on the maximum work capacity of both marriage partners and this reached its optimum at a comparatively early age. This not only eliminated the restrictive conditions which had limited the formation of new households within the family cycle and the succession of generations among full peasants. The rural industrial family mode of production created new pre-conditions of household formation which were determined by market conditions on the one hand and by the poverty of the rural producers on the other. These conditions inscribed not only the process of family formation among the rural industrial class, but they were also the chief factor determining family structure because they governed the entire life-cycle of the family.58 The constitution of the family economy primarily as a unit of work had specific demographic consequences as well. The pressure for the maximal utilization of the family work force not only required an early age at marriage and the teamwork of man and wife, it also favoured a form of reproductive behaviour which, by 'producing' a maximal number of child labourers, raised the productive capacity of the family and thereby its survival possibilities beyond that critical threshold of poverty on the margins of which the family often began its existence. Therefore it may be said that the demographiceconomic paradox of the proto-industrial system59 appears above all as a consequence of a mode of production based on the family economy. 'Women's earnings set a premium on early marriage, while the employment available for children encouraged large families and increased the supply of labour out of all proportion to the demand of the trade.'60 The imbalance between a fluctuating process of economic growth and a relatively constant process of demographic expansion, typical of proto-industrialization, had its basis in a prime factor of the social relations of production : the paradox, that those whose material conditions of inherited possessions rendered them least capable of rearing large numbers of children nevertheless produced them, may only be explained by looking at the specific conditions of exploitation under which the entire family labour force was placed in proto-industry.61 The drive toward marriage and intensive reproduction came about - within certain boundaries - more or less independently of the conjuncturally determined demand for labour; even under worsening economic conditions, a retreat to a restrictive, traditional marriage pattern characteristic of peasants and a corresponding mode of reproductive behaviour offered no viable alternative to the rural artisans. The adult proto-industrial worker was not able to exist as an individual; especially under worsening ' material conditions of production ', he had to depend to a growing extent upon the 'co-operation' of his entire family. 'No single-handed man can live; he must have a whole family at work, because a single-handed man is so badly paid he can scarce provide the necessaries of life... As soon as they [the children] are big enough to handle an awl, they are obliged to come downstairs and work.' (A domestic-industrial shoemaker from Northampton.)62 This pattern of reproduction of industrial workers, affecting household structure, family size, and relations of work, was not only an exogenous variable dependent on 'external conditions of reproduction'; it also acted as an endogenous variable shaping the family life-cycle from within. Functional and structural configurations of the working family were influenced by the reproductive process above all in so far as it regulated the 'dependency ratio' in its various phases through the family cycle. The ratio between workers and consumers that existed at the founding stage of the family was endangered by the reproduction process before it was again brought into balance. Before children could contribute to the household economy, they both hindered its productive capacity and increased its consumption. Successive births reduced the mother's ability to participate in family labour and thereby narrowed the margin of subsistence for both parents. It was precisely this temporal disjunction between production and reproduction within the proto-industrial family which trapped it between the Scylla of ' primary misery ' (arising from the conditions of the proto-industrial system) and the Charybdis of 'secondary poverty' (brought on by the family life-cycle). In bad times the longest working day does not suffice; the weavers who have between two and four dependent children fall heavily into debt and must regularly resort to poor relief. Only when two or three children sit at the loom can debts be repaid and savings made. If the brothers and sisters remain within the family and conduct an orderly economy, this offers a period when savings are possible. It is obvious how important it is to the parents to make their children work as early as possible, for they will not remain with them for long; the sons often marry at 22-23, the daughters at 18-19, both leave their parents and deliver them and their younger brothers and sisters into destitution. With the birth of children, the parents become poor ; with their maturation, they become rich; and with their marriage, they fall back into misery.*3 This dilemma of the family cycle was accentuated in its extreme above all under bad conjunctural conditions. Nevertheless, rural industrial producers were exposed to the ambivalent effects of the reproductive process not only under marginal conditions of income. The independent, intrafamilia! dynamic of 'demographic differentiation' (A. V. Chayanov) was exhibited precisely in those cases of small and medium-sized peasant households which were not yet in a permanent proto-industrial situation. In this case, the pressure of the reproductive process could turn peasants into temporary rural artisans. This temporary transition of small and medium-sized peasant households into rural industrial production occurred above all in those critical phases of their family life-cycle in which their subsistence could not be assured on the basis of agrarian production alone
The structural character of the dilemma to which the rural artisans were exposed under their marginal conditions of existence shows up above all in the constitution of extended households. The formation of extended families may here be seen as an effort to counterbalance both the 'primary misery' caused by the social relations of production and the 'secondary poverty' generated by the family cycle in the absence of developed forms of trade-union organization and the impossibility of an effective wage struggle. Complex household forms extending beyond the nuclear family occurred occasionally among propertied proto-industrial producers. According to the type of production, the stage of production and conditions of ownership, households with servants and apprentices were to be found more or less frequently. These households were sometimes those of traditional rural craftsmen65 or of small entrepreneurs who owned landed property and at the same time were engaged in the production and distribution of industrial goods.66 A third important group were the proto-industrial 'Kulaks'.67 The lines of demarcation between these extended household forms and another special type of rural work and settlement unit were blurred; in this special type, sub-peasant, satellite households of industrial producers were grouped as temporary lessees around a full peasant homestead, supplying its seasonal demand for labour and thus simultaneously serving as a kind of proto-industrial buttress to an agrarian organization of work
These household and settlement patterns should be considered as specific variants or mutations of the substantial farmer's or craftsman's family and they must be distinguished from the main type of extended family to be found among the landless or land-poor proto-industrial producers. This main type recruited its members above all from the closer circle of relatives or from a reservoir of non-related paying or working inmates.69 Thus it displayed a structure formally parallel to the extended household of the full peasant classes. Nevertheless, the two types differed fundamentally in their material, legal and institutional determinants. The extended family among the rural industrial workers was formed as a result of growing pauperization, increasing population pressure, of limited and congested living conditions and not least by the secondary poverty engendered by the family life-cycle. The classical stem-family, on the contrary, was formed essentially to conserve peasant family property. Viewed from a comparative perspective, the extended family of the rural artisans was much more the forerunner of the corresponding proletarian household configuration than a variation of the peasant stem-family.70 It did not function as an instrument of conservation of property, of well-being and of care for the aged, as was the case with the full peasant household, but as a private means to redistribute the poverty of the nuclear family chiefly by way of the family-and-kinship system. Such a situation of need could either arise temporarily during a critical stage of the family cycle or it could become a permanent condition of existence for the proto-industrial family as was the case in its final stage during the period of so-called de-industrialization. The sparse data which have been made available so far indicate that already during the expansionary period of rural industry during the eighteenth century, the classical, three-generational, stem-family pattern, consisting of grandparents, parents and children, did not occur to a significant extent.71 There does appear, however, a force within the family-and-kinship system integrating it in another direction. Nuclear family households which contained widows, unmarried sisters or brothers, nieces and nephews of the married couple did turn up fairly frequently. The conditions under which married couples co-resided in other households or left them again point towards the causes of forming such extended families: married couples who lived in another household - whether it was as relatives or as inmates - left it as often as the birth of children began the process of 'demographic differentiation '.72 Also within their familial subsystem the ratio of labour and consumption worsened. With more mouths to feed, the causes no longer existed which originally had made possible co-residence of this family in its 'host's' house. For the host, his 'guests' counted above all as labourers or paying inmates, which reduced the burdens and economic risks of his own family. Extended household formations among rural artisans therefore aimed primarily at balancing an unfavourable ratio of labourers and consumers. Thus the reversion to the kinship system or the recruiting of contributing inmates created a partial substitute for those functions which had been fulfilled within the traditional household by servants. Extended households seem to have been produced by primary misery and secondary poverty. The structural conditions and consequences of the specific connection of production and reproduction by which the proto-industrial family was formed, are only incompletely revealed, however, by the changes in the composition of the domestic group. Marginal conditions of existence substantially restricted the chances of engineering the survival of the family within this context. The proto-industrial family was by no means as free as the peasant household in its decisions as to recruiting additional members for its labour force. The adaptation of the household to early marriage and high fertility required by proto-industrialization entailed above all a change in the organization of work within the nuclear family unit itself. The range and penetration of this 'inner structural change',73 taking place within the organization of work, became manifest in the transformation of the division of labour between the sexes, in the configuration of roles within the family and in its social character. This 'inner structural change' cannot be sufficiently understood if it is conceived as a process of ' structural differentiation ' and ' role segmentation ' or as a mere prelude to the disintegration of the 'ganze Haus'.74 The history of the proto-industrial family economy formed a part of the long post-history of peasant society to the same extent that it formed a part of the pre-history of industrial capitalism. In this 'gleichzeitige Ungleichzeitigkeit' its main historical significance is to be seen. In any case its historical significance, even if conceived from the perspective of the history of a status-specific family type, cannot be reduced to that of an initial stage in the secular 'loss of function' of the family,75 which is so often myopically considered to be an immediate consequence of urbanization and industrialization and formally defined as the 'differentiation of occupational roles from the context of the kinship structure'.76 The proto-industrial family economy, it is true, was drawn into a progressive process of social division of labour. On the level of the family unit, this led to a loss of individual functions of production and thereby to the specialization of the productive unit as a whole; as a structural unit of work, however, the family economy during proto-industrialization was very cohesive indeed. In this respect it by no means underwent a process of disintegration. On the contrary, the necessity to work together under adverse conditions entailed a higher degree of functional integration and thereby also of structural cohesion than was necessary in the peasant family. 'In case of emergency, one man may be recruited out of two or three peasant families to protect the Fatherland without doing injury to agricultural production. This is hardly possible with families weaving woollen cloth. Their manufactures are like a machine consisting of many wheels which may not be touched.'77 Even when the internal organization of family labour underwent substantial change during the proto-industrialization period, and even when change in the existential basis of work affected the role configurations and the role relations of family members outside the immediate work process, still the framework of the 'ganze Haus' remained. The restrictive conditions under which the family economy had to ensure its survival in fact necessitated a 'maximum. . .of familial cooperation'.78 This had to be achieved by optimally redistributing and balancing the scarce labour resources of the individual family members. Under certain market conditions and within certain branches of production this imperative could go so far as to erase the traditional division of labour between the sexes and the age groups. The domestic production process of the rural industrial workers was thus characterized by a more flexible allocation of the role responsibilities of family members than was the case for the peasant (including even the small peasant and the sub-peasant classes). Especially lacking was that separation of work between men and women which was common, though not rigidly adhered to in peasant households, whereby as a rule the men worked out of doors in the fields, while the women were occupied with household tasks (including the practice of household crafts for the personal needs of the family, the cultivation of the gardens, milking the cows, care of the livestock and the marketing of the surplus produce of the household).79 Even when this sex-specific division of labour was largely erased, as it was among those small peasants or sub-peasants who were still overwhelmingly occupied with agricultural production, the man nevertheless remained, whether as a day-labourer, migrant-labourer or cottager, generally excluded from domestic cottage production. The sphere of woman's labour on the other hand extended its reach within this class and became increasingly important. Whether the wife became active as a spinner engaging in production of commodities for the market or whether she increased the marginal returns from petty agrarian production by intensive cultivation or by tending the livestock on the commons, often it was only her activity that assured the vital margin of the family economy's subsistence.80 'A woman cannot get her living honestly with spinning on the distaff, but it stoppeth a gap.'81 The proto-industrial household continued this sub-peasant pattern and at the same time changed it by making the man, so to speak, return to the household. At least in the textile trades he moved into a work situation that had been traditionally established by women though he did not give up his labour outside the house, at least as long as the partial agrarian basis remained intact.82 In this historical sense it seems justified to describe women as the 'vanguard of peasant household industries'.83 This holds true especially in those places where household industry was carried on in conjunction with a partial agrarian basis. Generally, however, the proto-industrial situation was characterized by a rather strong degree of assimilation between the production functions of men and women. Women in the roles of cutlers and nailers84 as well as organizers of the marketing of the industrial products85 were as common as men were in the roles of spinners86 and lacemakers.87 Occasionally this adaptation of familial work organization to the conditions of survival went beyond the disappearance of the traditional separation of labour between the sexes. It could lead to its reversal: where the necessities of production compelled women to neglect household 'duties',88 this 'Joss of function'could be compensated for by the men assuming traditional women's roles. Behaviour which to contemporary observers from the middle and upper strata of society all too quickly appeared as a reversal of the 'natural order', posed no particular role problem to weavers or to specialized households of spinners. It was here that 'men. . .cook, sweep and milk the cows, in order never to disturb the good, diligent wife in her work'.89 The distribution of family labour across the lines separating the labour between the sexes and age groups did not only determine the behaviour of family members in the sphere of production. Social behaviour and especially consumption and sexual activities were also influenced by the respective forms of co-operation between men and women and their outward constraints. As 'role functions' they were not separate from the process of production and reproduction, although in their symbolic, socio-cultural meaning they were more than mere extensions of that process. Although precise investigations are lacking, there are indications that role behaviour of the sexes in consumption among the rural artisans was by no means constantly tied to a division of labour in which men would function as privileged consumers ' symbolising the role of breadwinner' (N. J. Smelser), and were thus entrusted with status consumption in public, whereas women would be restricted to householding, to caring and preparing for the necessities of life.90 It is precisely in status consumption that an egalitarian role of both sexes came to be symbolized. This happened within the boundaries of the house as well as in the wider community. In the 'plebeian public' of the rural artisans both sexes frequently articulated their needs by drinking and smoking in common.91 Their communality manifested itself not only by passive consumption, it showed itself also in the active defence of traditional norms of subsistence.92 During food riots and actions against unbearable price rises, women by no means withdrew from the public eye. Very often it was the women who were 'more disposed to be mutinous; . . .in all public tumults they are foremost in violence and ferocity'.93 Even in immediate sexual encounters, the new conditions of production led to a changing social texture. Property constraints and patriarchal controls were loosened and, as we have noted, largely replaced by the need to select marriage partners on the basis of their productive capacity. Nevertheless, these new criteria permitted more latitude, more individualization and personalization in the selection of partners and thereby produced a gradual 'transformation of the world of erotic feelings'.94 Erotic expression, however, was not confined to a separate sphere, distinct from the work process, but was bound in a specific way to household production itself.95 'Where people of both sexes are always together, in the warmth of the same room, and where they... carry out work that occupies their head and heart so little', they spend their time in idle intercourse 'which is commonly concerned with lust and lasciviousness, with fraud and theft' and  those who have 'the dirtiest ideas imagine themselves to be and are held by the others as heroes'.96 It was not only at the symbolic level that the enlarged significance of sexuality in the everyday life of the rural industrial workers changed the position of the sexes and age groups. It led both to a lowering of the age of sexual activity97 and to increasing similarities in the sexual activity and behaviour patterns of men and women. The 'immorality' and 'shameless freedom of the sexes' which middle-class observers noted about the rural artisans in contrast to peasant behaviour98 was primarily a criticism also of sex-specific role behaviour. Observed through the behavioural biases of the upper class, this similitude of the behaviour patterns of men and women certainly appeared as an unbalanced relationship : Among these classes of men, the male sex is the reserved one and the women the ones who are on the make. . . The common maid understands the art of coquetry in its various forms just as well as the mature woman; she discloses her breasts without shame and certain other enticing parts of her body half way because she knows that is more alluring than all the way. If the young man still resists, she helps weaken him with liquor and if he doesn't respond to her invitation to her bed, she joins him in his. The usual plot of a romantic novel is thus reversed.99 This quotation unwittingly contains an important grain of historical truth pointing beyond its own specific content and once more illuminating the general problems with which historians of the family are confronted. Their critical praxis will indeed have to be informed by an understanding of those experiences which, according to the disapproving contemporary observer, formed part of the everyday life of the rural artisans. Quite apart from any reorientation of 'perspectives' this demands a much more fundamental change in the basic position of family history. The historian of the family must, in his conceptual approach, in the material questions which he asks and in the source material which he scrutinizes, try to 'reverse the usual plot of a romantic novel' as it too often has been retold by historians and sociologists of the family, even today.100 Ruling-class perspectives not only determined the way contemporary observers perceived rural artisans; they persisted in the explicit or implicit middle-class bias which still seems enshrined in the conceptual approach and even methodological perspectives of much that is written in the contemporary history and sociology of the family. In this way, social scientists, be they sociologists or historians of the family, become what Walter Benjamin referred to as 'adherents of historicism', against whom he recommended a form of 'cautious' historical materialism that could well serve as an agenda for future family history: "The adherents of historicism actually empathize.. .with the victor. And all rulers are the heirs of those who conquered before them. Hence, empathy with the victor invariably benefits the rulers. Historical materialists know what that means. Whoever has emerged victorious participates to this day in the triumphal procession in which the present rulers step over those who are lying prostrate. According to traditional practice the spoils are carried along in the procession. They are called cultural treasures, and a historical materialist views them with cautious detachment... A historical materialist . . . dissociates himself from. .. (them) as far as possible. He regards it as his task to brush history against the grain