5. Kinship & strong ties: concentric circles round the family
All these connections between the Le Play cleavages and other social, economical and political processes make one suspect that household cannot be taken as the only framework for current social changes. Let us try and enlarge the analysis to other dimensions of social relationships. Forty years ago Elizabeth Bott  underlined a chief difference between 'small-scale (primitive) societies' and 'urban industrialised societies':
"In England and other Western European industrialised societies work groups are seldom recruited on the basis of kinship, individuals may earn a living without depending on relatives for their means of livelihood, productive resources may be owned by individuals who are not related to one another (..). This reduced importance of kinship in economic affairs is associated with a narrower range of kin recognition, with absence of corporate groups of kin (..), with less frequent and intense contact among relatives".
No doubt the model featuring a large range of kin recognition, the importance of kinship and corporate groups of kin in economic affairs is dominant in many non-Western societies [Note 29]. But it is also very close to the Mediterranean model of the family-firm and stem-family. Elizabeth Bott's intensive study [Bott 1957] into a small number of London families and their social networks clears our mind of the "commonplace of sociology and anthropology, that kinship does not play a very important part in industrialised societies". Litwak and Szelenyi  stressed later that the growing thinning out, in Western societies, of face-to-face contacts does not mean a loss of importance of the primary groups.
The co-resident household must therefore be analysed as the core of an integrated system, surrounded by at least two circles which are analytically distinct. The first one includes that part of the kinship that is operationally or symbolically close to the household. The second circle, which develops round household and kinship and can extend beyond them, consists of the network of subjects connected with members of the household by strong ties, i.e. 'frequent ties, giving emotional or instrumental support' [Note 30]. If some processes of anthropological embedding, concerning family models, are connected with the current differentiation of demographic behaviour, something similar is also true of kinship and network patterns [Note 31]. Some sort of systemic connection links together the forms and sizes of the three relationship circles.
The first seminal rule of linkage between the family pattern and the connectedness of family networks [Note 32] was formulated by Bott , who distinguished two kinds of families (segregated conjugal [Note 33] versus joint conjugal role relationship [Note 34]) and two kinds of social networks around a family (a 'close-knit' network, with many relationships among the component units, versus a 'loose-knit' one, with few such relationships). Bott concluded: "the degree of segregation in the role-relationship of husband and wife varies directly with the connectedness of the family's social network". In other words, the asymmetrical family appears to have a more dense strong-ties network, i.e. a social network where there are more kinship ties than strong ties with non-relatives (to the extent that the proportion of kin in the network can be taken as a good proxy for its density):
"Kin are of special importance in any type of network. First, kin are especially likely to know one another, so that the kinship region of the network is likely to be more close-knit than other sectors. Second, relationships with and among close kin are relatively permanent" [Bott 1971].
An intriguing result emerges from European sociological research: as for both Le Play's household patterns and Bott's 'family and kinship' patterns, the size and form of the strong-ties networks are not homogeneous throughout Europe. Hollinger and Haller , emphasising from the 1986 survey of the International Social Survey Program [Note 35] significant differences between the north-western and central culture areas of Europe and the Southern ones, show that the closer the family structure and higher than elsewhere the frequency of face-to-face contacts with kin [Note 36], the fewer are the strong ties with non-relatives. To sum up," the importance of friends in people's social support networks is inversely proportional to the importance of extended kin".
Truly, keeping the current habit of identifying Bott's asymmetrical family with Le Play's stem-family, in the Hollinger & Haller contribution, there is evidence of a sort of paradox [Note 37]: contrary to the inverse relationship between social support networks and the importance of extended kin, which was found in the other six countries surveyed, Italy (the only Mediterranean country in the International Social Survey) is characterised both by a strong presence of kin and a large circle of non relatives strong ties. Nevertheless, if the Italian family is evolving toward a nuclear form but retains marked role asymmetry [Palomba and Sabbadini 1993], we should expect (in keeping with Bott's and Hollinger & Haller's more general rule) a smaller network of non kinship strong ties.
Recently some social surveys in the Netherlands [Gierveld, Tilburg and Lecchini, 1995] and in a few regions of Central-Northern Italy [Note 38] have made it possible to compare the size and form of the network of 'emotionally and/or instrumentally significant' ties among older people (over 65 years old) in the two countries. The networks came out as very different, both in size (14.5 members in the Dutch network, only 5 in the Italian one) and composition [Note 39]. But the small size of the strong tie circle is not peculiar only to older people.
In a recent Italian survey, called the 'Social Barometer' [Note 40], the age-specific curve of network size [Micheli and Billari 1998] has a parabolic shape, with the smallest sizes at the extreme ages, both among young people (where the network belongs mostly to the outer, non-kinship circle) and among older ones (where the strong ties begin concentrating into kinship) (fig. 1). However the peak of network size in full adulthood remains clearly inferior to the average Dutch levels. And such a restricted network is also an intrinsically kinship network [Note 41]. All this is fully coherent with Bott's typology, confuting the old stereotype of an Italian family surrounded by a circle of countless friends and relatives.
The general rule is then confirmed again, even though its rationale escapes us. Why is Europe split up into two patterns of social interaction, with different proportions and roles among co-resident relatives, non co-resident kinship and non-kinship ties? Socio-cultural interpretations stress the role played by the cleavage between private-oriented and public-oriented [Note 42] societies, and the statistical linkage between public orientation and basic education spread can support the "Second Demographic Transition" hypothesis, i.e. an underlying process of 'modernisation' from North to South of Europe. Actually, the clear preference expressed by young people for non kinship strong ties and therefore the overweighting of the outer circle in their network, as results from Social Barometer data, could also be an effect of the spread of high-school education.
Nevertheless, the difference between network size and density in Italy and the Netherlands is so great that it is difficult to support a simple hypothesis of uniform change in Europe. Moreover the diffusionist approach, unless it unrealistically assumes the absence of a Southern European family model, simply refers to some previous process of historical formation of the cleavage [Note 43]. And however it is interpreted, it cannot itself be based upon family models alone.
Kinship, Family and Social Network: The anthropological embedment of fertility change in Southern Europe
Giuseppe A. Micheli
© 2000 Max-Planck-Gesellschaft ISSN 1435-9871