1. INTRODUCTION
The notion that East-Central Europe is the locus of complex family organisation and familistic societal values has reached the status of a general dogma in Western social sciences and demography, and has wide currency in other intellectual circles as well.Footnote 1 A few scholars have criticised these mainstream perspectives on the topic from empirical, conceptual or epistemological points of view, and have suggested the need to move beyond the stereotypical and artificial divisions of Europe into ‘Western’ and ‘Eastern’.Footnote 2 However, as the persistent usage by demographers of the division proposed by John Hajnal to explain European demographic differentials since the Second World War suggests,Footnote 3 the position of the ‘revisionists’ is still largely absent from mainstream discourse. Thus, further attempts to persuade scholars to accept less stereotypical images of families from outside Western Europe are clearly needed. This essay examines once again the existing literature on historical family systems in East-Central Europe. However, it seeks to broaden intellectual horizons by placing the literature side by side with almost entirely unknown contributions of Eastern European scholars, thereby revealing the stark incompatibility of those two discourses. It is argued that the Western homogenising view of Eastern European family patterns stems equally from three specific attitudes: (1) the drawing of bold inferences from partial and inconclusive evidence; (2) the neglect of the substantial counterfactual testimony; and (3) the failure to consider the local, ‘native’ Eastern European literature on family and demography.
This paper is organised into four major sections. It opens with a re-examination of the evidence for the existence of a divergent ‘Eastern European family pattern’, followed by a review of the most vivid manifestations of this concept in contemporary scholarly literature. Next, well-established – albeit thoroughly ignored – evidence from Western mainstream demography and family history will be revived and compared with the dominant discourse. In the third and most extensive section, selected contributions of Eastern European scholars will be presented to reveal concepts of familial developments that came into being independently of the Western search for sharp contrasts in familial characteristics on the European continent. The paper closes with a general discussion and conclusion.Footnote 4
2. THE ‘GREAT EUROPEAN DIVIDE’: THE WESTERN PERSPECTIVE
One of the central tenets of sociological and historical studies of the family has long been the existence of a specific, peculiar ‘Eastern European type’ of family system. A century after the concept of ‘Eastern European’ geographic space emerged among enlightened intellectual elites in the West,Footnote 5 the idea that family developments in Eastern Europe diverged from those in the West was frequently articulated in nineteenth-century ethnographies. The German Romantic August von Haxthausen wrote extensively about the Slavic agrarian constitution and rural organisation, and argued that the Russian peasantry were invariably organised in large, extended and patriarchally structured families.Footnote 6
However, Eastern European ‘specificity’ was first captured and, consequently, juxtaposed with the patterns that are assumed to dominate the West in the works of Frédéric Le Play.Footnote 7 Le Play popularised the notion of a gradient of family and household types running from east to west, and claimed that patriarchal, patrilocal and multi-generational households could be found among ‘Eastern nomads, Russian peasants, and the Slavs of Central Europe’, as well as among the Hungarian population.Footnote 8 A review of Le Play's mid-nineteenth century household model mapFootnote 9 reveals a remarkably regular distribution of different family types across the countries of Europe, with clear divisions between the eastern regions of the continent on the one hand, and the northern and western regions on the other. The line he drew on his map between those three large macro-regions, which ran approximately from St Petersburg to Trieste and then down the Appenine Peninsula, marked the boundary between the patriarchal families of the East and the stem and unstable families of the West and North.Footnote 10
Le Play's conclusions about family structures in Eastern Europe were, however, derived from a study of only seven families who were largely concentrated in two highly dispersed regions, from the Urals to Hungary and Slovakia.Footnote 11 Le Play's approach, although innovative and valuable in many respects, could hardly fit the requirements of modern social science methodology, especially when it came to generalising from single case studies. Despite these obvious shortcomings, many contemporary scholars still hold Le Play's mapping exercise in high esteem.Footnote 12 By the second half of the nineteenth century, however, a basic image of the Eastern European family system had already been established.
This nineteenth-century assessment of Eastern European difference sank deep into the collective consciousness, and was later perpetuated in modern historical demography and family history, despite the discipline's otherwise strong revisionist drive regarding other aspects of older scholarship.Footnote 13 The myth of the existence of a demographically uniform Eastern Europe, where people married young and lived in patriarchal households, was presented most enduringly and pervasively in Hajnal's 1965 article on marriage patterns in Europe.Footnote 14 Hajnal summarised his theses, developed on the basis of an analysis of aggregate statistics from around 1900,Footnote 15 in a very concise statement: ‘The marriage pattern of most of Europe as it existed for at least two centuries up to 1940 was, so far as we can tell, unique or almost unique in the world. There is no known example of a population of non-European civilization which has had a similar pattern.’Footnote 16 The ‘European pattern’, the distinctive features of which Hajnal considered to be a high age at marriage and a high proportion of people who never married at all, pervaded, according to him, ‘the whole of Europe except for the eastern and south-eastern portion’.Footnote 17 Reiterating Le Play's original spatial exercises, Hajnal introduced an East–West gradient in European demographic behaviours with much greater force, and argued that ‘the European pattern extended over all of Europe to the west of a line running roughly from Leningrad (as it is now called) to Trieste’.Footnote 18 Hajnal significantly hardened Le Play's initial distinctions between Eastern Europe and the rest of the continent, and was keen to equate the marriage pattern of several countries located ‘east of the line’ with the marriage characteristics of ‘non-European civilizations’.Footnote 19 In his 1965 article, Hajnal also linked the European pattern of late marriage with the stem family as defined some 100 years before by Le Play, but he seemed to treat it as more akin to simple rather than joint (patriarchal) family systems. Hajnal's text can also be read as strongly suggesting the incompatibility of early marriage behaviour (ascribed to Eastern Europe) with the simple- or stem-family systems believed to prevail in other parts of the continent.
In the meantime, Peter Laslett elaborated on, reiterated and retold Hajnal's original hypotheses. Laslett started with envisaging the specificity of the English nuptiality pattern by indicating that marriage in the English context meant the creation of a new economic unit.Footnote 20 Soon after, however, he took the specificity of English marriage and household behaviour to be representative of Western Europe as a whole, and, at the same time, he described it as a characteristic which had probably distinguished ‘the west and the north of Europe from the east and the south’.Footnote 21 In Household and family in past time, what illuminated the Western familial pattern (‘West’ meaning mainly Western Europe and America) were the starkly conflicted marriage and household patterns observed in ‘Far Eastern Europe’ (rural and urban Serbia) and in Japan.Footnote 22 Laslett's perspective on pre-industrial Eastern Europe as representing the greatest intra-European departure from the ‘English standard’, and from Western Europe as a whole, can be easily discerned from his various, but geographically non-systematic, accounts in the volume.Footnote 23
Laslett's view of intra-European differences in family systems was crystallised in a 1977 paper.Footnote 24 Despite seeing the geography of European family systems as ‘being complex and puzzling’,Footnote 25 and acknowledging the limited availability of data for continental Europe, Laslett was not discouraged from making bold interpretative inferences from single case studies. He outlined several features of North-Western European households in the Early Modern era, subsuming them under the label of ‘the Western family pattern’. He then considered large parts of what is often referred to as East-Central Europe (Latvia and Estonia in the North, Poland, the Czech lands, north-eastern Austria, and also Transdanubian Hungary)Footnote 26 as belonging to a hypothesised ‘large intermediary area’ between Western and non-Western family systems.Footnote 27 It was the area with an explicitly ‘in-between’ position (close to the European ‘core’, but still not fully ‘Western European’), with a considerable degree of ambiguity and a tendency towards permutations and possible combinations of marriage and family patterns.Footnote 28 A ‘European Far East’ was represented in Laslett's account by Russia (the Baltic states were often included).Footnote 29 This concept provided researchers from the Cambridge Group for the History of Population and Social Structure with a long-anticipated place where the complex family was the ‘universal background to the ordinary lives of ordinary people’,Footnote 30 and therefore supplied Laslett with the missing element in his mapping exercise.
This now fully articulated notion of the extremes of familial organisation within the confines of the European continent was first substantiated by Peter Czap's study of a single Russian community of Mishino (some 170 kilometres south-east of Moscow; 128 households in 1814, with a population of 1,173 persons) during the eighteenth and nineteenth centuries, the first work in English to make extensive use of the ‘Cambridge’ approach.Footnote 31 The outcome of Czap's study posited that there was a specific ‘eastern European family type’ (as opposed to the Western one) which was characterised by a high proportion of three- or more generational multiple-family households, a mean household size significantly greater than five persons, and a more or less general propensity for early marriage within the population.Footnote 32
However, there is a certain level of ambiguity regarding the spatial implications of Czap's ‘model’ case study. Initially, Czap had no reservations about calling marriage behaviour on the Mishino estate ‘a robust non-European marriage pattern’ (without inverted commas),Footnote 33 and he also seemed to be genuinely satisfied with an assertion in a later piece of research that his findings were applicable to large areas of pre-emancipation Russia.Footnote 34 Nonetheless, he expressed serious doubts as to whether his ‘eastern European family type’ could have prevailed beyond the ethnic Russian territories of the continental part of the Empire.Footnote 35 Better still, he suggested that the southern agrarian provinces were separated in the eighteenth century from the central industrial region surrounding Moscow by a socio-demographic and familial frontier, although he was unable to define the position of this frontier more precisely.Footnote 36 However, in the early 1980s, the search for sharp contrasts in familial characteristics and the wish to brand major areas of Europe as having a particular type of household system were well underway. It was due to this methodological orientation, and also to the general ignorance of Western family historians about Russia and Eastern Europe generally, that Czap's tentative hypothesis suffered the mixed fortune of being regularly assumed to be representative of the whole country, and even of the whole continent to the east of Hajnal's line.Footnote 37
After the publication of this new evidence on the eastern part of the continent, further mapping endeavours were possible. In two papers published in the early 1980s, Hajnal used the concept of the ‘Western European family’ to distinguish between two kinds of household formation system in pre-industrial times: the simple household system in North-Western Europe, and the joint household system (stem-family systems in which one of the sons married and took over the farm after his parents retired were considered compatible with the North-Western Europe family system).Footnote 38 By explicitly calling the paper he published in 1982–1983 a ‘sequel’ to his famous 1965 essay, Hajnal seemed to suggest that the two supra-national, large-scale family systems he was studying could be spatially conceptualised as referring to territories west and east of his famous line. However, Hajnal's geographic references were flawed and imprecise. Since in 1983 there was no substantial body of evidence available to him for studying European east-of-the-line territories, he made no explicit commitment about the kind of household system that was characteristic of Eastern Europe in the past. Nevertheless, he was oblivious to Czap's reservations regarding the interpretation of the Russian data, and used his Mishino case study as if it were representative of all of the European East, or at least of all Eastern European serf populations in pre-modern times.Footnote 39 Similarly, he disregarded Laslett's remarks cautioning that there was a striking variegation of East-Central territories when he jumped thousands of kilometres south-west from Mishino to use circumstantial data from Hungary and Croatia to supplement his case for the prevalence of joint-family systems in other Eastern European regions.Footnote 40
By contrast, Laslett refined Hajnal's argument and departed from a simple dichotomous East–West model of European familial differences, arguing instead for two additional sets of tendencies, which he saw as bridging the most extreme contrasts in domestic group organisation on the continent. For this ambitious mapping exercise, Laslett used a household composition dataset that was very limited, even compared with the data used in his earlier speculative essay from 1977, and treated data from several single location points as illustrative of regions of Europe which seem to have had distinguishable forms of family and household (these ‘domestic group tendencies’ were West, West/Central or middle, Eastern, and Mediterranean).Footnote 41 Laslett took the ‘eastern tendencies’ illustrated with the Mishino data as ‘justifiably associated with the domestic group structure of European Russia as a whole and some of its surrounding areas’.Footnote 42 On the other hand, his notion of ‘West/Central or middle’ tendencies was accompanied by strong geographical and definitional uncertainties. He asserted that the area in which such tendencies had been found lacked a clearly marked pattern, and he therefore proposed treating those domestic group constituencies as an intermediate category, though closer to that of the West than to the Mediterranean or Eastern categories.Footnote 43 Laslett's ‘sets of tendencies’ in the middle zone complied geographically with the claims of Central Europeanists of the 1980s, and therefore included ‘small nations between Germany and Russia’.Footnote 44 Large parts of historical Poland – especially the Belarusian, Lithuanian and Ukrainian territories – would probably be allocated within the zone of the alien, ‘Eastern’ familial tendencies.Footnote 45 Strikingly, in this influential paper positing a ‘four region hypothesis’ in domestic group organisation, only two locations were mentioned from the vast area stretched along the West–East axis from Oldenburg and Vienna in German-speaking countries, to the Ryazan province south of Moscow in the East (two Hungarian settlements). This immense space ‘in-between’ was a ‘terra incognita’ in European family systems.
Laslett's and Hajnal's tentative generalisations have long been held in esteem. This reverence, as well as the prolonged scarcity of research material available for Eastern Europe, encouraged other scholars to indulge in intellectual equilibristic and bold generalisations. Tony Wrigley from the Cambridge Group for the History of Population and Social Structure argued that in Eastern Europe, ‘huge and complex households predominated with extensions both laterally and vertically’, presenting ‘a dramatic contrast’ with Western European standards. For ‘Middle Europe’, however, Wrigley envisioned less uniformity and a wider prevalence of family forms intermediate in size and complexity.Footnote 46 French historical demographers would not admit to having any doubts that ‘[f]rom Serbia in the south to Courland or Estonie [sic] in the north, passing through Poland and Russia, one encounters certain common features: a household size which is much larger than in the West, and a strong propensity for multiple households’.Footnote 47 Reflecting upon several decades of research, Andrejs Plakans recently remarked that the propositions about the historical characteristics of Eastern European family life – namely that ‘low ages at first marriage, low proportions not marrying, high proportions of complex households (…) – have retained a great deal of validity, and can therefore serve as a set of baseline characteristics for exploring the rest of the nineteenth century’.Footnote 48
Laslett's and Hajnal's hypothetical generalisations provided a ready and badly needed framework for scholars from other fields wishing to understand recent family, or even socio-political, developments. Within this framework, the tantalising claims and tentative inferences of family historians and demographers were eagerly transformed into ‘solid’ scientific evidence that helped to substantiate their own claims. Goran Therborn, for example, referred to Hajnal's nuptiality hypothesis in his global history and sociology of the family to demonstrate that ‘the classical European family divide, running from Trieste to St Petersburg […] is still visible in 2000’.Footnote 49 In a similar spirit, demographers took Hajnal's bipolar division of the continent from around 1900 at face value, and often used it too hastily as an additional tool for explaining European-wide differentials in demographic transformations since the Second World War.Footnote 50 Among some anthropologists, as well, Hajnal's notion of the North-West European simple family system continues to serve as an essential framework for explaining the relatively weak kin ties in Northern Europe, as contrasted with the ‘descent’-oriented and more familistic regimes prevalent in Eastern Europe and the Mediterranean.Footnote 51 Many recent demographic studies of contemporary European household structures still take Hajnal's and Laslett's mapping exercises for granted and use them as a reference point for the comparison of patterns observed in present-day Eastern Europe.Footnote 52
3. DISSIDENTS IN THE WEST: THE RECONCEPTUALISATION OF THE WESTERN AND THE EASTERN EUROPEAN FAMILY
All this was taking place – irrespective of the growing criticism in the West of the accuracy of Laslett's and Hajnal's accounts of regional variation in European household forms during the 1990s – largely because of Richard Wall's voluminous research. Wall's first piece, in which the topic of intra- and inter-regional variation in familial organisation was first tackled, was a relatively unknown study from the 1970s.Footnote 53 It opened by describing Laslett's vision of a typical English household – i.e. small, containing surprisingly few children and kin but large numbers of servants – as turning ‘into something approaching a stereotype, more particularly in regard to mean household size, to be applied to all manner of English communities regardless of location or time period’. Wall then made an attempt to propose a ‘correction [to that picture] by charting variations in household size and structure between the seventeenth and nineteenth centuries and between one part of the country and another’.Footnote 54 According to him, not only had demographic rates changed between the pre-1750 and the 1750–1821 periods in England, but, consequently, various aspects of household organisation and related phenomena (for example, the leaving-home process) changed too. The most dramatic modifications of this kind were related to the presence in households of kin and servants, and the number of households headed by females, even though the percentage of three-generational households remained rather low in England at less than 13 per cent. Wall also observed that there was a substantial range of variation between individual settlements in England: in some of them, it was possible to find almost no co-resident kin at all, whereas in other groups, they accounted for more than 10 per cent of the population. All in all, between 1650 and 1821 the share of households with kin was about 10–13 per cent in the selection of English communities, but by 1851 there was a dramatic shift, with the share rising to 20 per cent.Footnote 55 The changes were not tremendous, but they nevertheless seemed to be too great to be ignored: ‘Even while it remains relatively robust to the major demographic push of the eighteenth-century, the English household was not static. Nor, given the variations between area and area, would it be correct to see English households as variations on one basic type.’Footnote 56
Wall developed this theme further in a paper published in 1991, which took advantage of the increased availability of published research and data pertaining to the analysis of household systems in historic Europe.Footnote 57 Wall's point of departure was to claim ‘the illogicality of relying solely on the presence or absence of kin as a definition of the household system’. This was justified by the fact that the number of persons of all types found within a household varied considerably in different parts of North-Western Europe (including Spain).Footnote 58 Nevertheless, the figures he provided for the number of relatives (other than spouses and offspring) per 100 households by relationship to household headFootnote 59 already tell us a large part of the story, and are more important here than other methodological considerations. Regional samples from Belgium, Denmark, England, Iceland, the Netherlands, Norway, Spain and Switzerland revealed figures for relatives per 100 households which varied from five in one Dutch region to the mid-20s in countries such as Belgium and Denmark, to 35 or more in rural districts of Norway and Iceland. In the case of servants, the range in variation was even larger. For Wall, these intra-European differences were quite modest when placed alongside the structure of household in some non-European populations. Nevertheless, this ‘considerable variation in household structure even within the confines of northern and central Europe’ led him to tentatively discern four household patterns from the pool of available data.Footnote 60 He was convinced that the data he had assembled gave ‘a better perspective on what range of variation can be expected within an area which Hajnal's North-West European household system is supposed to dominate’. He added, quite stunningly, that ‘so great is the degree of variation that it must be doubtful whether Hajnal's generalization captures much of the reality of family and household patterns of north-west European societies in the past’.Footnote 61
The theme of variation, divergences and similarities in European family systems resurfaced rather forcefully in two papers published by Wall in the late 1990s and early 2000s,Footnote 62 both of which investigated the accuracy of Laslett's and Hajnal's accounts of regional variation in European household forms. Using the sample of English populations from the 1851 census, Wall showed that up to one-fifth of households in some of these populations were complex (by complex, he meant extended and multiple families altogether). He therefore raised serious doubts about whether it was justifiable to claim that the proportions of complex households were indeed ‘very low’ in England, as Laslett did on many occasions. While he did not reject entirely the notion of English specificity, Wall perceived it differently from his predecessors. ‘What is (…) most distinctive about the English experience,’ he noticed, ‘is its uniformity, relative to the variation in household forms occurring in other parts of Europe.’ France, Spain and Italy had regions dominated by simple family households, but, equally, they had populations with proportions of multiple family households far in excess of the shares seen in any of the English communities. According to Wall, the latter feature was also a prime characteristic of Eastern Europe. Yet even there the occurrence of multiple families might have occasionally resembled rates observed in other parts of Europe, in Italy in particular.Footnote 63 Not only could many distinctive patterns be identified within the confines of pre-industrial North-Western Europe, Wall asserted, but the pace and timing of familial change in the different parts of the continent probably varied considerably.Footnote 64
If the supposed foundations of the North-Western European familial specificity were substantially shattered by the research of people such as Wall, the emerging orthodoxy proclaiming the East–West familial dichotomy was on even shakier ground. This ‘demographic brotherhood of thought’ regarding the familial characteristics of the eastern part of the continent has actually turned out to be a smokescreen that hides substantial differences in research perspectives.
Among the earliest heterodox investigations into Eastern European household structures were Plakans' studies of the big Latvian parish of Nerft in eighteenth-century Kurland (which included 17 noble estates, 771 farmsteads and 11,040 individuals).Footnote 65 An intriguing outcome of this careful examination of surviving household lists was the observation that, despite being representative of the family pattern that contrasted sharply with what was known for the West, the complex family in Latvia was not a universal feature in the lives of ordinary people.Footnote 66 Plakans' remarks went largely unnoticed by Cambridge scholars.
In the meantime, June Sklar carefully collected census-type evidence for every political entity of the Eastern European region from around 1900, which she then minutely decomposed into smaller political units.Footnote 67 Following Hajnal in this matter, Sklar summarised her analysis in a very concise statement: ‘(…) the East European regions that were to become Latvia, Lithuania, Estonia, Czechoslovakia, and Poland after World War I did not exhibit the Eastern European pattern of marriage behaviours, but were actually closer to the West European pattern.’ By referring to values of the singulate mean age at marriage, she argued that ‘nuptiality in these regions at around 1900 followed the West European late marriage pattern’, with the mean age at first marriage fluctuating between 24 and 27 years for women, and between 25 and 30 years for men. Sklar observed commonalities across Estonia, Latvia, Lithuania, Poland and Czechoslovakia in the proportions single in different age groups, with only a slight departure from this general tendency found over the territories that were later to become Poland. She concluded that persons in all these areas ‘married rather late, and moderate proportions never married at all’.Footnote 68 In light of this abundant evidence, Sklar felt at ease in concluding that, in Eastern Europe around 1900, both ‘Western’ and ‘Eastern’ European marriage patterns prevailed, with the latter being followed by the Balkan countries. She substantiated her claims by using basic data to show that, around 1900, there were only negligible differences between household sizes in the Baltic, Czech and the Polish provinces and household size in Sweden (with the mean household size fluctuating between 4.7 and 5.2 persons), but that there were more significant differences in relation to Bulgaria and Serbia (which had mean household sizes of 5.8 and 7.2, respectively). More speculative, but still very interesting in the context of the discussion of Hajnal–Laslett models, were Sklar's comments about the relationship between marriage and residence patterns, and the way they both were buttressed through kinship rules and practices in various parts of Eastern Europe. On the basis of ‘historical and observational studies’, Sklar maintained that ‘in the Czech, Baltic and Polish territories, the independence of [the] nuclear family was reflected in the custom that the typical peasant farm should support one family only (…)’, and that the peasant practice was ‘to leave a farm undivided to one son who would marry and remain on the holding while “paying-off” his brothers and sisters (…)’. Both of these observations vividly recalled Hajnal's own description of Western European stem-family societies. According to Sklar, the emphasis upon the independence of the nuclear family in the Czech, Baltic and Polish provinces produced strong pressures that tended to favour late marriage, sometimes leading to celibacy among the non-inheriting offspring. In contrast, the integration of the nuclear unit into the parental household in the Balkans created pressures favouring early marriage.Footnote 69
Sklar's observations were close to Hajnal's own description of marriage as being contingent on the availability of self-sufficient positions or niches, and on the inheritance practices he saw as underlying the formation of typically North-Western European households. Unsurprisingly, Sklar seemed fond of relocating the dividing line suggested by Hajnal more towards the East, so as to include the countries such as Latvia, Lithuania, Estonia, Poland and Czechoslovakia into the zone of ‘Western’ marriage and household characteristics. Her repositioning of the demographic fault line in Eastern Europe suggests that parts of the Lithuanian, Belarusian and Ukrainian ethnic territories of the Polish state may have been included in the ‘Western’ zone.Footnote 70 By suggesting there were at least two distinct marriage patterns in the region, Sklar's paper refuted notions of historical Eastern Europe as a demographic monolith for the first time in modern population history. Her study found not only a transition zone along the North–South axis, which seemed to delineate East-Central European from Balkan marriage patterns, but it also revealed noteworthy differences within East-Central Europe itself. Finally, Sklar's analysis opened up options for recasting Eastern European marriage and family patterns at the turn of the nineteenth century, but only a few researchers took advantage of the opportunity presented.Footnote 71
Helena Chojnacka, a student of Ansley Coale at Princeton, proceeded along similar lines, and found a true spatial diversity of marriage behaviours in Tsarist Russia around 1900. Three belts of marriage regimes stretching from the West to East displayed a gradual decrease in nuptiality when moving from the south to the north of the country. Chojnacka confirmed Sklar's earlier observation, and suggested a correction to Hajnal's hypothesis: ‘[A]pplying Hajnal's terminology, the non-European pattern – defined as early and quasi-universal marriage – can be applied in the south and central regions of European Russia, but not in the north. The latter is much closer to the unique European marriage pattern.’Footnote 72 Although Chojnacka was not able to establish a clear relationship between different patterns of marriage and different types of families, she nevertheless tentatively suggested that ‘an extended patriarchal-type family’ was dominant ‘among the Great Russians, with a variety of modifications among the White Russians, and to a lesser extent among the Ukrainians’. Among the latter, she claimed, ‘the nuclear family was more common’.Footnote 73 As we can see, no claim for the universality of the prevailing family type on Russia's western fringes was made here.
Hajnal's hypotheses were also partly questioned by the authors of the Princeton monograph on Russia.Footnote 74 Their collection of figures on the singulate mean age at first marriage and the proportion ever married for Western European, Eastern European (including European Russia) and non-European (Asian and African) societies provides grounds for challenging Hajnal's attempt at equating the ‘Eastern European pattern’ with the marriage characteristics of ‘non-European civilisations’ as misleading. In both indexes, the contrast between ‘Western’ and ‘Eastern’ European populations (the latter being Bulgaria, Hungary, Romania and Serbia, as well as Ukraine, Belarus, Lithuania, Latvia and Estonia) was reminiscent of distinctions between the latter group and non-European populations from the Far East and North Africa.Footnote 75 Moreover, there was by no means an unequivocal spatial order to marriage and family patterns even to the east of the Hajnal line. Again, three distinct patterns of first marriage were detected within European Russia, with the Baltic republics sharing the late experience of first marriage long customary in Western Europe (Im of 0.56 or less in 1897), and Belarusian and Ukrainian territories displaying an ‘intermediary pattern’ (Im of 0.62 to 0.68) between the above and the pattern of early marriages characteristic of territories stretching almost horizontally from the Black Sea to the Urals. The examination of spatial distribution of singulate mean age at marriage values for Russia's westernmost provinces in 1897 indeed revealed quite substantial differences in marriage ages, which, however, did not unfold along a West–East axis, but rather from North to South.Footnote 76
A similar perspective was advocated in Josef Ehmer's study of historical marriage patterns in the crown lands of the Austro-Hungarian monarchy, 1880–1890.Footnote 77 Ehmer pinpointed a striking divergence of the Galician nuptiality regime from trends common among populations of the Kingdom's other provinces.Footnote 78 In Galicia at the end of the nineteenth century, he suggested, the age at marriage tended to be much lower: in some regions almost all men were married by the age of 30 years, and permanent celibacy was nearly unknown.Footnote 79 When entering East Galicia, Ehmer observed, ‘we are leaving behind the European Marriage Pattern and Household Formation System’.Footnote 80 However, even though he treated this area as a demographic monolith, Ehmer suggested there was a demographic fault line running across the province which divided it into two parts along ethnic lines. The western part, with the great majority of Poles (up to 90 per cent of the local population), was characterised by relatively large proportions of never-married males, whereas the situation differed greatly in districts dominated by the Ruthenian (Ukrainian) population. Importantly, in Ehmer's view the eastern Ukrainian family pattern represented an example of the East-Central European family type, which was supposed to prevail over the entire Carpathian area and extend into eastern Ukraine as well. It was distinguished by the pattern of earlier marriage that ‘might really be a transitional form towards the Eastern European Marriage Pattern’,Footnote 81 and also by patrilocal household formation and a strong tradition of complex, patriarchally structured family forms, but by a relatively small household size. Ehmer concluded that the marriage patterns of the Polish-speaking population in western Galicia departed only slightly from the more Western-like tendencies of the other crown lands of Austro-Hungary.Footnote 82 Ehmer's contribution supplied yet more proof of the need to variegate the view of family tendencies east of Hajnal's dividing line. Still, Ehmer's picture of variety in East-Central Europe was painted with a broad brush, and the concept of a transitional zone between Western and Eastern marriage and household patterns located somewhere in East-Central Europe, to which he subscribed, needed to be fleshed out with a more substantial body of evidence.
The concept of a transitional zone between Western and Eastern marriage and household patterns was later fuelled by another Austrian scholar. In two publications, Markus Cerman pointed out that Early Modern Central Europe may be thought of as representing the transitional area with respect to the European marriage pattern and different household formation systems. By focusing on Austrian and Bohemian data from the seventeenth and eighteenth centuries, Cerman blurred the existing geography of marriage patterns in that part of Europe by asserting that there were surprisingly high proportions of married males in Bohemia relative to Austrian areas as early as in the seventeenth century, and by suggesting that there was an additional North–South fault line in the region.Footnote 83 However, he still argued that in the Early Modern period Austria shared the same more ‘Western-like’ household and family formation system with Bohemia, but not with Slovakia, where higher proportions of complex households were found together with a lower mean age at marriage. Cerman was reluctant to see the Slovakian family pattern as representing an ‘Eastern Hajnal-type family system’; instead, he claimed it was far more appropriate to view Slovakia as being part of a ‘very broad transitional zone, whose dominant household patterns were strongly influenced by local and regional socio-economic and legal contexts’. In Central Europe, he continued, ‘there existed not only an extreme variant of the Western European pattern (…) in rural areas of Austria, but also significant variations from this Western European pattern in other regions such as Slovakia and Hungary (…). The famous Hajnal line which appears prominently in the literature in its role as structural border between Eastern and Western family systems appears therefore to be diffused by the presence of areas where family forms were more mixed.’Footnote 84
Since the early 1980s, attempts have also been made at canvassing the Balkan family and demographic realities.Footnote 85 Although Maria Todorova was very cautious in inferring generalised claims from the micro-censuses she examined for nineteenth-century Bulgaria, she nevertheless asserted that the Balkan region should not be incorporated as a whole into the non-European or Eastern European marriage and family patterns. Although her analysis of data from both rural and urban areas of Bulgaria generally confirmed the contrast made by Hajnal between Western and Eastern European patterns of female marriage, it also confirmed the assumption that there was relative uniformity in Western Europe and the Balkans regarding the average size of the household and the distribution of the households by composition. ‘The characteristics of the family and the household [in north-eastern Bulgaria during the 1860s],’ Todorova concluded, ‘do not make possible the establishment of some essential difference from the West European model.’Footnote 86 As in many places outside of Eastern and South-Eastern Europe, the Bulgarian family pattern was characterised by the predominance of simple-family households (some 60–70 per cent of all domestic units), alongside a substantial proportion of extended-family and multiple-family households (12–16 per cent of the latter), with, however, an allowance for individual cases in which the pattern was leaning significantly towards the Eastern type. Accordingly, Todorova reconceptualises the South-Eastern European area as having a great deal in common with Central and Southern Europe, particularly with regard to the occurrence of multiple families.Footnote 87
Recently, Siegfried Gruber used micro-census data from Serbia and Albania to extract information on historical household formation and marriage patterns in both countries, and found some interesting inter-country differences behind the general structural similarity of family patterns. The male age at marriage was higher in Albania, and Albanian men were married over a longer period of their lives than the Serbs. There were similar proportions of multiple-family households in rural areas in both countries, but they contained many more cousins in Albania than in Serbia. Almost all of the Serbian households were divided in each generation, while Albanian households were divided less frequently. Albanian cities clearly had higher percentages of multiple family households than Serbian urban locations. The proportions of unmarried people were quite similar, as was the average household size in both countries. Gruber's conclusion was that there are more indicators for different patterns than for only one family pattern in the two Balkan regions.Footnote 88 When confronted with the diversity of family arrangements stemming from an increasing number of micro-structural studies of Balkan communities, other scholars have rejected the very notion of ‘typical’ Balkan household arrangements.Footnote 89
4. SPEAKING FOR THEMSELVES: EASTERN EUROPEAN STUDIES ON FAMILY AND MARRIAGE
Although the first independent studies on Eastern European household structures appeared at almost exactly the same time as when the Cambridge Group framework for comparative analysis of families was completed and made known to a wider research community,Footnote 90 their sensible voices went largely unheard by Western scholars. Either they were mentioned only in passing, without affecting the general picture usually drawn, or they became known to a wider public too late to halt an ongoing stereotyping of Eastern European demographic realities.Footnote 91 These studies were, however, also preceded by even more voluminous literature from the period between the mid-nineteenth century and the early 1960s which anticipated many threads of contemporary English, Austrian or French studies on the history of family and kinship, even though they worked on different methodological premises and had different research goals. In this section, these older perspectives on familial issues will first be briefly reviewed, and then more contemporary literature will be considered.
4.1 Nineteenth- and early twentieth-century contributions
Among the objects of heated debates among Eastern European scholars since the late nineteenth century has been the issue of ‘intra-familial relationships’ (a term used to describe the totality of issues pertaining to familial land ownership, inheritance, kinship, co-residence, and, to a degree, residential propinquity of relatives). In practice, the points of disagreement have been related to the origins, size, legal character and spatial distribution throughout Eastern Europe of the so-called zadruga-type family forms; that is, a family community that in modern studies is frequently categorised as belonging to the residential community group.Footnote 92 Following Valtazar Bogišić, nearly all of the Southern-Slavic literature has concluded that the zadruga is a relic of ancient all-Slavic forms of ancestral organisation which can be traced back to the era of first settlement, and several East-Central European authors have also signed up to this theory.Footnote 93 This image, popularised in a simplified version in Western literature, would soon sink deep into the collective consciousness, and would, with time, condition the framework of debates on the geography of family forms in EuropeFootnote 94 by equating those archaic forms of communal social organisation with a supposed propensity to multi-generational co-residence over the whole eastern part of the continent, and among Slavs in particular. However, some of these early scholars also provided a striking acknowledgment of the diversity of family forms in East-Central Europe as early as the late nineteenth century, and offered particularly perceptive differentiations of various patterns of family forms developing in the western and eastern lands of the historic Polish–Lithuanian Commonwealth.
Karel Kadlec assumed that in medieval times the Central European variant of zadrugal forms known as niedział (literally, ‘something undivided’) represented a prototypical form of family life common to all Slavic peoples, which was meant to be a commune of people bound by ancestral kinship who jointly manage a shared estate under the guidance of one leader.Footnote 95 However, he also pointed out that zadruga-type communes survived for long periods only in southern Slavdom and in Russian countries, while in regions inhabited by western Slavs, they disappeared more quickly.Footnote 96 Among Poles, they disappeared earlier than among Czechs, most probably before the end of the sixteenth century. The disintegration of niedział progressed along different patterns among Czech and Slovak populations, as well. Among the rural Czech population, it was only occasionally found in the eighteenth century, while in Slovakian territories (especially around the Carpathians) its remnants were detected a century later. The more rapid process of the individualisation of family life and property laws in western Slavdom was caused mainly by the influence of Western concepts, especially the terminology of German law, and was manifested in the simpler structures and smaller sizes of local ‘undivided family communes’ relative to the structures found in Russian lands, and especially in the Balkans.Footnote 97 In Poland, as in the Czech territories, family collectives were rapidly reduced to forms of tighter communes embracing the joint familial property in a narrow sense of the term, most frequently between the father and his unmarried sons.Footnote 98
The crowning jewel of Polish discussions of zadruga-type family forms was the work of Oswald Balzer and Henryk Łowmiański. Balzer found big family communes in medieval Bohemia, in Poland proper, as well as in the Polish eastern borderlands, where they assumed forms identical with patterns known from Southern Slavdom or Ancient Rus. However, these zadrugal forms in Eastern Europe varied in durability. They disappeared fastest from the territories of the Polish Crown and Bohemia, and, if they lasted longer, then did so usually as relatively simple and small two-generational communes.Footnote 99 On the western fringes of Ukraine, family communes lasted well into the sixteenth century, both among the gentry and the peasant population. In some minor regions they did in fact survive up until the eighteenth century, but then only among peasants. Eastern European family communes differed also with regards to their life-cycle characteristics. In Poland and Bohemia, they took the form of temporary joint-property groups (sometimes, but not always, also co-resident entities), which usually split either immediately or shortly after the demise of the head. Further to the east of Poland, undivided family units were more durable: out there, communes among brothers lasted over their entire lives.Footnote 100
Łowmiański was the first to reinterpret communal property systems on Lithuanian–Ruthenian lands of sixteenth-century Poland in strictly demographic categories. Importantly, according to Łowmiański, the property communes jointly managing the land were composed of separate households, or dyms (hearths). The number of dyms making up a commune could vary substantially, and in Lithuanian regions was lower than in Volhynia and Polessie (northern Ukraine). Furthermore, dyms also differed considerably in size: in the southern belt of the Lithuanian–Ruthenian lands, they were bigger than in the more northern regions of the Grand Duchy.Footnote 101
Those discrepancies were but a signal of much more substantial differences in the material and social culture of the Slavic people since the earliest medieval times.Footnote 102 Among Slavs the disintegration of lineage groups into small families had already occurred during the period of intense settlement between the seventh and tenth centuries; however, this process did not result in the conjugal family gaining primacy everywhere.Footnote 103 At least in sixteenth- to eighteenth-century Poland, grand families of the scale of extended Balkan zadrugas did not occur, allowing only for the occasional occurrence of households of more than one married couple.Footnote 104 However, in some regions of Slavdom, the strong lineage system survived until very recent times.Footnote 105 Small and nuclear families from the eighteenth-century Polish Crown could be juxtaposed with residential communes from Belarus, where at that time there was a greater share of multiple households than of single households (even up to 60 per cent), as the population avoided a general parcelling of households and extensively used familial property communes.Footnote 106 According to Łowmiański, the grand Belarusian families from the late Early Modern period were the continuation of a previously vanished institution prevalent in the western lands of the Polish country.
4.2 Czech, Slovak and Hungarian literature
These early suggestions regarding the presence of an East–West gradient in family composition in Eastern Europe provided a unique agenda for more quantitatively elaborated studies on the structure of the family. Unfortunately, contemporary Eastern European family historians sought to develop these earlier insights only to a very limited extent.
They resurfaced most extensively in the Czech and Slovak literature. The investigation into family and household structures in the former Czechoslovakia dates back to the late 1980s, when ‘The 1651 Register of Subjects According to Their Religion’ (Soupis poddaných podle víry), covering almost all of the lands of historical Bohemia, was first examined using quantitative techniques.Footnote 107 Since then, one of the basic premises of Czech and Slovak scholars studying historical household structures has been that there may be an intermediary marriage and household formation pattern in Central Europe.Footnote 108 Pavla Horska was the first to introduce the concept of a ‘Central European model of the family’,Footnote 109 by which she meant a nuptiality pattern that was transitional between the North-Western and Eastern European models. She also asserted that, during the seventeenth, eighteenth and nineteenth centuries, the family household in the Czech countries never seems to have been of the patriarchal, ‘zadruga-like’ type: it was most frequently composed of the parents and children, as ‘elsewhere in Western Europe’.Footnote 110 Several studies confirmed the overwhelming dominance of nuclear households in sixteenth- and seventeenth-century Bohemia (up to 79 per cent), followed by extended households (up to 32 per cent), but only a very small share of domestic units shared by more than one family (up to 9 per cent). In addition, a significant fraction of the young unmarried population in Bohemia was found to have worked as unmarried servants in the households of non-kin.Footnote 111 All of these features made it possible to treat the Bohemian variant of the Central European pattern of the family as being more or less compatible with patterns observed in Western Europe.
At the same time, however, Horska and others warned that an important demographic fault line passed through Czech lands during the Early Modern period. Whereas in Bohemia more complex family types may have been widespread only before the seventeenth century, the ‘great family’ was much more usual in the Moravian Carpathians and Slovakia, where it frequently involved the co-residence of married brothers and sisters in a manner resembling the joint-property systems of a fraternal zadruga type.Footnote 112 A feature that differentiated such residential arrangements from the Eastern or South-Eastern European realities was, however, specific power relations within households, whereby a co-residing brother could occupy an inferior position and was entitled to have a share in the household consumption only if he performed various labour services for the brother-head. The non-negligible geographical pattern was also believed to have existed in the Czech lands with regards to nuptiality, as the age at first marriage declined as we proceed from the north-western to the south-eastern parts of the region.Footnote 113 Soňa Švecová drew on ethnographic literature to link these two different family and demographic regimes in the area of the former Czechoslovakia with two types of property devolution: the one-heir system known as ‘rodina jednonástupnická’, and joint-property systems known as ‘rodina nedielová’.Footnote 114 She also proposed that there was a decisive turn along the way from the Eastern type of household formation (rodina nedílová) to the Central European pattern (one-heir system), which could be found in Bohemian lands between the late sixteenth and early seventeenth centuries, but not in Slovakia.Footnote 115 In the latter, the development of nuclear- or stem-family arrangements was prevented by the family joint-property system, equal inheritance among sons, the real partition, and, finally, by the strictly agrarian environment.Footnote 116 Instead, three- or four-generation families with partilocal marriage and patriarchal power relations were quite prevalent, a pattern which often persisted well into the twentieth century.Footnote 117 Švecová was persuaded to consider the Slovakian family pattern as belonging to the ‘Eastern’ type of Hajnal's typology, with the Western Carpathians representing within the Central European setting the border between two different family models which she herself had suggested.Footnote 118
A similar variegation of family patterns was also found for late eighteenth – and early nineteenth-century Hungary. Andorka refuted the notion that poly-nuclear households would have represented something of a general pattern in Hungary. Although they were fairly widespread in the Transdanubian region, places where the share of nuclear households was much greater and extended and multiple families were much less prevalent could be easily found in other areas of the country.Footnote 119 A later study of seven localities from the period between 1747 and 1816 suggested that the household structure in Hungary ‘seems to have been intermediate between Western Europe on the one hand and Serbia and Russia on the other’, but with allowances that important differences may have existed within the country itself.Footnote 120
Tamás Faragó pinpointed those differences more precisely by allocating the marriage and household organisation patterns of several rural communities in eighteenth-century Hungary to three specific categories: the Western European, the Eastern European, and, finally, the East-Central European family model (Ostmitteleuropäische Familienmodell).Footnote 121 The latter category was supposed to encompass behavioural patterns representing a transition between the ‘North-Western European pattern’ identified by Hajnal and Laslett, and the Russian reality. Capturing the diversification of family forms in the territories of the historical Hungarian Kingdom also remains a leading research goal for contemporary Hungarian historical demography. Both Tamás Faragó and Péter Őri found a large degree of patchiness of patterns of marriage and household formation across pre-1800 Hungary, which evades classification according to the simple dichotomous model.Footnote 122
4.3 Polish scholarship
For the period 1960–2000, there have been a dozen or so works that deal more directly with the structure of the peasant household during the serfdom period in Poland. Most of these were isolated case studies describing family forms with various typologies devoid of any reference to the models of familial organisation developed in the West.Footnote 123 Nevertheless, all of them reported more or less unequivocally a decisive predominance of simple family households in the historical Kingdom of Poland between the sixteenth and eighteenth centuries, even though the territorial basis of these investigations was limited almost exclusively to the western and south-western parts of the country.Footnote 124 Nonetheless, Polish researchers generally hypothesised that there was a predominance of nuclear households over the whole of the historical region of Poland in Early Modern times, tentatively assuming the existence of different family systems operating on the country's eastern periphery. Acknowledging the homogeneity of manorial politics and the effect it had on the peasant family, Witold Kula proposed that the nuclear family spread over the entirety of the Polish corvée-obliged rural population.Footnote 125 Maria Koczerska, in turn, extended the simple family model over the population of the nobility, among whom it had replaced more kin-based residential arrangements as early as in the fourteenth and fifteenth centuries.Footnote 126 With recourse only to a very modest body of data (mostly from the seventeenth and eighteenth centuries), Irena Gieysztorowa proposed an operational hypothesis in which she observed that the age at marriage in historical Poland declines on an eastward trajectory, an assumption which was recently authoritatively repeated by Cezary Kuklo.Footnote 127 Quite in line with contemporary views offered by other Central European scholars, Gieysztorowa has accentuated the borderline character inherent in the patterns of Polish marital behaviours. This view was supported by later, more comprehensive comparative analyses of nuptiality across Polish territories, in which it was argued that the marriage pattern in eighteenth-century Poland may have been a cross between the Western and Eastern patterns, although much closer to the ‘unique’ North-Western European pattern than to patterns observed in the Hungarian, Russian and Ukrainian territories.Footnote 128
The hypothesis on the visible nuclearisation of family forms throughout the majority of Korona lands in the Early Modern period, which was introduced above, was supplemented by other studies of both rural and urban communities. Michal Kopczyński's study of a few dozen parishes in central and western Poland from between 1650 and 1800 showed that the majority of peasant households had a nuclear structure, and that the relatively high mean size of the domestic group resulted from the spread of hired servants and co-residing lodgers. The marked increase in the number of multi-generational families on the lands covered by his investigation was only brought about by peasant enfranchisement of the second half of the nineteenth century.Footnote 129 Referring to Kuklo's study of six urban communities of the eighteenth-century Polish Crown provided even greater certainty regarding the domination of the simple-family model in central Poland's lands. In the urban centres, the two-generational family was prevalent (making up 66–85 per cent of all domestic units), followed by unusually high proportions of solitary households.Footnote 130 According to Kuklo, the household structure in the Polish town of the pre-industrial era must be classified as ‘typically West-European’.Footnote 131 For the cluster of rural communities in eighteenth-century Silesia (today south-western Poland), I found a moderate age at marriage, the dominance of simple-family households, and a high incidence of life-cycle servants. I found strong indications of a stem-family pattern in those places, accompanied by cases in which the modes of household formation did not vary much from the neo-local principles prevalent in North-Western Europe, or which followed exactly this type of pattern.Footnote 132 As I have argued, if the European great divide in family systems suggested by Hajnal really existed, it was certainly not located in Upper Silesia. It would be necessary to search for it farther to the East.Footnote 133
Indeed, Anna Laszuk concluded that, in the mixed Polish–Belarusian rural areas in the north-eastern Polish Crown lands, the domination of the Western type of family was not unambiguous in the late seventeenth century. By and large, however, the simple family type still occupied a superior position and the share of joint-family-type domestic groups was small, and only among the nobility rose to more than four per cent of total households.Footnote 134 According to another author, Zdzisław Budzyński, the negligible importance of multi-generational families in the Polish eastern periphery resulted from the widespread practice among newly married couples of gaining economic and residential independence. The individualisation of property and residence, the argument goes, was the core organisational principle of the family household on the Polish eastern outskirts, both in the eighteenth and the nineteenth centuries. Differences between different ethnic groups (e.g. Poles and Ruthernians) in this regard were supposed to be small.Footnote 135
A more precise identification of the long-expected familial and demographic border in historical Polish territories was attempted in Szołtysek's studies of living arrangements in different regions of Poland–Lithuania (more than 14,000 peasant households from the eighteenth century were analysed).Footnote 136 The analysis initially revealed the juxtaposition of a more complex family system in the eastern communities with a homogeneous, but simple, family pattern in the western Polish lands. However, a closer look at the data showed that, at the end of the eighteenth century, not two but three household and family patterns with substantial numerical and qualitative differences existed in the historical Polish territories. The structural progression within larger regions, Szołtysek demonstrated, nearly always moved in the same direction: from less kin-centred, more nucleated, and neolocal households in the west, to much more notable levels of household complexity in Poland's more eastern territories. However, even on the eastern periphery (in Belarus, for example), the family pattern still differed markedly from paradigmatic examples of the Eastern European family type detected in Russia. These findings were taken as indicative of the existence of a wider Eastern European area with a similar family pattern at the end of the eighteenth century, with basic commonalities in household size and structure prevalent across Lithuania, Belarus, Red Ruthenia and western Ukraine; as well as Slovakia and the northern part of Hungary. On the basis of these results, more evidence was provided disputing the assumption that large parts of East-Central Europe were dominated by a homogeneous family system. These findings further demonstrated that Hajnal's dichotomous notion of Western and Eastern Europe from around 1900 cannot be transposed onto earlier periods.
4.4 Family and household studies in Lithuania, Belarus and Ukraine
Until recently, Belarusian, Lithuanian and Ukrainian scholars showed a general lack of interest in studying domestic groups in socio-historical perspective.Footnote 137 However, as early as the early 1960s, Angelé Višniauskaitė demonstrated that the ‘grand indissoluble family’ (an equivalent to the term ‘joint family’ commonly used in Western terminology) never constituted a dominant household form in ethnic Lithuania between the sixteenth and the end of the nineteenth centuries.Footnote 138 The nuclear family system in Lithuania was a direct consequence of lineage relationship decomposition, which affected the Baltic countries starting in the thirteenth and fourteenth centuries, followed by a marked decline in family communes. Both of these processes were additionally strengthened by the agrarian reforms of the mid-sixteenth century, which brought about the introduction of the three-field system and the manorial system.Footnote 139 An increase in peasant compulsory labour obligations imposed upon them by landlords in eighteenth-century Lithuania caused the accumulation of family labour on the holding, and, in consequence, led to a drastic rise in the share of multiple family households in Lithuania (33 per cent of all domestic units in the years 1700–1800). Paradoxically, however, the only moment when in some parts of Lithuania really complex multi-focal families could be found was during the 1930s and 1940s, that is, when capitalism was already a fact of life.Footnote 140
Some Belarusian scholars applied similar approaches in their handling of the problem of household structure in various Belarusian ethnic territories between the end of the sixteenth century and the mid-seventeenth century. Kapyski's analysis of 252 settlements revealed that, on average, one household was comprised of no more than 1.2 conjugal family units, and that more than 85 per cent of all households had only one conjugal family unit. Most of the remaining multiple-family domestic groups contained two small families co-residing. Also, in Belarus, the transition from the sixteenth to the seventeenth century was marked by an increasing simplification of peasant residential patterns, and one-family households made up the majority of domestic units over the entire Belarusian territory.Footnote 141 Referring to a more extensive dataset (over 30,000 peasant hearths), Valentin Golubev has proposed a regionalisation of family forms in various parts of the Belarusian part of the Grand Duchy of Lithuania at the end of the sixteenth century, with a major change occuring on a trajectory of movement towards eastern territories. Whereas the proportions of one-family hearths to the general number of peasant households amounted to 73 per cent in the western part of the region, it dropped to 66 per cent in the central region, and even to 46.5 per cent in its eastern part.Footnote 142
Referring to his study of several communities from central Belarus, Viachaslau Nosevich asserted that, based on his review of the sixteenth-century data, there was no reason to draw a sharp distinction between family structures in Eastern and Western Europe. He demonstrated that nuclear-family households were absolutely dominant in Belarus in the second half of the sixteenth century (between 70 and 89 per cent of total households), and that in some places this pattern developed even earlier.Footnote 143 At the same time, however, he noted the emergence of a more complex family pattern in central Belarus during the eighteenth and nineteenth centuries, which he linked to the gradual increase in feudal obligations imposed on the peasantry by the Eastern European landlords.Footnote 144 Nonetheless, over almost the entire eighteenth century, the rural population in Belarus was found to follow a pattern of rather moderate household complexity, which was in marked contrast to the features characteristic of nineteenth-century Russia. According to Nosevich, this ‘balanced’ household pattern may have been widespread and persistent in some other parts of Eastern Europe, including northern Lithuania, Ukraine, Estonia, Karelia and parts of Hungary.Footnote 145 Towards the end of the eighteenth century, the family pattern in Belarus gradually transformed into the more communal forms already typical of the vast regions of Russia, with the share of multiple-family households rising significantly above 50 per cent. It was this nineteenth-century phenomenon, but not its various antecedents, that made the distinction between family structures in Eastern and Western Europe so attractive to Western scholars.Footnote 146
The Ukrainian literature on family history brings yet another surprise. The overall description of the Ukrainian family system was drawn up with an emphasis on the powerful drive towards the independence of both individuals and family units in various historical periods, and on the uniquely ‘nuclear’ character of the Ukrainian peasant family.Footnote 147 The simple family, it was argued, decisively prevailed in Ukraine as far back as in the second half of the nineteenth century, making up an estimated 84 per cent of all peasant families.Footnote 148 The behavioural dimension of this characteristic was a norm dictating that kinsfolk only occasionally formed joint production and residential units. Even in those cases in which this happened due to poverty or other circumstances, there were no seniority principles, no joint property rights, or concept of the community of work among the co-resident families.Footnote 149 Other scholars acknowledged the coexistence of both small and ‘big, undivided joint-families’ in Early Modern Ukrainian lands, but noted that a typical extension strategy involved the addition of only one son who stayed at home in expectation of taking over the farm after the father's death. Moreover, the co-residence of married brothers sometimes encountered in sixteenth-century Ukraine was generally temporary in character.Footnote 150
While most of the folklore studies were focused on the late nineteenth century, some authors were suggesting that, in parts of Ukraine, the ‘grand patriarchal family’ did not actually exist throughout the entire Early Modern period.Footnote 151 Quite often, it was argued, those distinct family households were in essence patronimic communities of related persons who frequently co-operated economically on their shared plot of land.Footnote 152 Other scholars presented a more variegated picture, and proposed various caesuras to mark the beginning of the spread of the simple-family form across Ukrainian territories. According to some, a more pronounced disappearance of joint families in Ukraine was not seen until the beginning of the seventeenth century, though the process remained incomplete.Footnote 153 According to others, the popularisation of single-family households in both right- and left-bank Ukraine did not occur until the 1770s or even later, although the simple two-generational household definitely became a dominant family type by the mid-nineteenth century.Footnote 154 The risks involved in uncritically transposing the concept of the small nuclear family onto the realities of the period from before the half of the nineteenth century are clearly indicated by more contemporary historical–demographic research in Ukraine.Footnote 155
5. CONCLUDING REMARKS
A substantial number of the nineteenth- and twentieth-century works of family historians, historical demographers, and political economists and sociologists working on demographic issues were concerned with the spatial designations and divisions of Europe. Having been preoccupied with establishing borders, drawing borderlines, and distinguishing between different demographic and family systems in historical and contemporary Europe, scholars of those genres developed their own ‘symbolic geographies’ of the Eastern European demographic space.Footnote 156 Early generations of experts in family organisation and structure were in the habit of searching for striking contrasts in familial characteristics, and were often prone to speak in terms of dichotomous East–West contrapositions. This tendency to underplay variations in family living arrangements in the European East was criticised even among Western scholars. As early as the 1990s, David Kertzer argued that ‘eastern Europe, like western Europe, displayed a diversity of household systems in preindustrial times’ which were linked to ‘regional differences in political economic arrangements and ecological conditions’.Footnote 157
Indeed, the picture of the formation and structure of family forms in vast eastern territories that has been painted in Eastern European historical–demographic literature appears largely incongruent with the postulates of Western scientists. In substantial stretches of Eastern Europe (including territories to the east of the border area suggested by Hajnal) the nuclear-family model was found to have been prevalent, at least during some historical periods. Over substantial sections of this part of the continent – again with the inclusion of some regions from outside the ‘line’ – neo-localism seems to have remained a dominant practice of household formation here and there (e.g. in Ukrainian lands), undoubtedly constituting the very fabric of a prevalent familial ideology. Contrary to a widely held view, according to which Eastern European complex family patterns supposedly made economic sense for both the Eastern European peasants and the landlords, given the circumstances of re-feudalisation to which the two sides found themselves subjected,Footnote 158 in-depth studies of manorial practices have suggested that seigniorial authority often provided strong incentives for neo-local household formation among the subject farmers. While many scholars showed an inclination toward a belief in the geographical diversity of family forms in the European East, others maintained that some ‘borderlands’ strayed from patterns dominant throughout the majority of its territory. The works of Lithuanian and Belarusian scholars clearly indicate that, in some historical periods, the actual differences between the East and the West in terms of the composition of residential groups were much less pronounced than expected, or were even negligible. The variation, both spatial and temporal, in the external forms of familial life, so characteristic of Eastern Europe, is an important argument in favour of jettisoning the concept of the ‘dividing line’ entirely, or of substituting it with the notion of temporally fluent transitional zones which are always unstable and subject to transformations occurring in distinct contexts and for different reasons.
This does not mean, however, that all of the claims of Eastern Europeanists should be accepted uncritically. Eastern European literature on family forms was mainly exiguous. Well into the 1990s, it was screened off from the main current of European thought. The substantive weight of the observations made by these researchers was frequently diminished by the fact that the available source material was researched only cursorily, and by the application of a methodology largely detached from the mainstream solutions and concepts. The method of deduction from applying examples, instead of undertaking a comprehensive review of the problems, led to the coexistence of conflicting and often irreconcilable perspectives on the issue. The archived material presented to support certain arguments has often left much room for dispute with regard to the accurate categorisation of familial forms.Footnote 159 With a few notable exceptions (especially the Polish studies appearing since 1967 in the journal Polish Demographic Past), quantification was used only rarely, making the assessment of the representativeness of the presented findings difficult, if not impossible.Footnote 160 Most observations were restricted to the serfdom period, leaving aside a vast topic of familial behaviours in the post-enfranchisement era.Footnote 161 Last, but not least, in some national discourses the entire dispute pertaining to the historical roots of various family types has sometimes taken on a highly ideological character.
What is needed here is a research programme carefully documenting the contours and variability of household, nuptiality and life-course patterns for various Eastern European regions by using spatially organised historical statistical data. The essential part of this new agenda would be to identify and differentiate the composition and behaviour of multiple sub-populations in a given area or society of Eastern Europe. By revealing significant variations in household formation, marriage, residence patterns and welfare functions of the family group separating these sub-populations, a much more nuanced geography of family patterns, both in terms of its spatial and temporal aspects, could be achieved.Footnote 162 Instead of utilising traditional simplistic notions of dividing lines and ‘ideal family systems’, this new scientific programme would inevitably adhere to a more sensitive focus on the nature and permeability of frontiers and transition zones, and the ways in which familial and demographic borders have been crossed and diffused, both across space and over time.
It is definitely possible to undertake a research programme of this kind for the Eastern European space. We have just begun to take advantage of the ongoing micro-data revolution – a combination of digitisation, Internet access, and harmonisation of surviving census and census-like materials – to investigate critically the differences in European historical family systems over space and time (see www.censusmosaic.org). East-Central Europe – with its mixture of historical legacies, cultural propensities and ecological factors – will definitely remain at the centre of historical debate. A range of comparative projects have been started to provide new insights into the variation and evolution of European family systems and to improve our understanding of the causes underlying their continuity or discontinuity over space and time.Footnote 163
Although partly impressionistic, the findings of Eastern Europeanists discussed above revealed enough diversity of family forms and in the rhythms of their development in historical Eastern Europe to free us finally from a simplistic view of the continent's familial history, and particularly the perspective implied by the notion of a ‘dividing line’. The crux of the argument here is that breaking away from a homogenising view of Eastern Europe's family and demographic past can help scholars contextualise more thoughtfully recent demographic processes taking place in the eastern part of the continent. It may also serve policy analysts to understand better the role of historical heritage in the socio-political, economic and demographic currents of the new member states of the European Union, as well as of some potential candidates for membership. In this context, a consideration of particular meanings calls to mind the words of the most cited historian of East-Central Europe, Oscar Halecki, who wrote back in the 1950s that ‘one of the main defects of (…) the basic distinction between Western and Eastern Europe lies in the impression obviously created that all of what is geographically “Eastern” is alien, or even opposed, to “Western” – that is, truly European – civilisation’.Footnote 164 Historical studies of the family can draw lessons from these insights.
ACKNOWLEDGEMENTS
An earlier version of this article was presented at the 35th Social Science History Association Annual Conference, Chicago, November 2010, in the session ‘In homage of Richard Wall and the twenty-fifth anniversary of Continuity & Change: Household, family and marriage across time and space’. I wish to thank Andrejs Plakans and an anonymous reviewer for their valuable comments.
