In "Shared Friendship Networks," Kalmijn analyzes and reformulates "dyadic withdrawal" in the progression of relationships in The Netherlands. Kalmijn's study involved analyzing the responses of a national survey of couples and singles that asked people to list their five best friends. He aims to assess the degree of network overlaps in couples' responses and use results spanning the life course stages to cull information about dating relationships.
Kalmijn presents two arguments to explain dyadic withdrawal: competition (our spouses and children compete with friends for our attention, time, and network resources) and balance (a sort of evolving natural selection to weed out inharmonious intransitivity between your spouse and your friends). Forty-three percent of friendship networks were symmetrically shared, which makes one think that perhaps spouses tolerate their partners' friends out of respect and politeness more than they sincerely form friendships with them. He mentions the very interesting existence of marriage capital, and argues that the smaller, more overlapping social networks of couples raises the exit-cost of a marriage and increases inter-dependency. While Kalmijn later dismisses any differences between divorced people and married or single, non-divorced people, I thought that this loss of friends after divorce and the high rates of divorce in contemporary America could be applied to McPherson et al's reading last week. He goes on to deduce that women are less socially dependent on their husbands, as the men cannot fulfil the function of their female, emotion-support based friendships, as opposed to the male, activity-based friendships. While I enjoyed reading this economic/rational transaction analysis of spouse relationship overlap, I do think that Kalmijn did not give enough credit, specifically in his balance principle, to the new identity as a couple's usurping of the individual and this effect on asymmetric, and specifically hostile "forbidden triads." I also wonder about cohort network durability and cohesiveness. Would Kalmijn have found that homophilous life structure development between two networks increased the chances of their friendship? You are probably more likely to stay friends and have frequent interactions with a friend (who you may have even been less "close" to before) who is having a baby at the same time you are, as opposed to the best friend who went off to medical school or to work as a travel guide. Another facet of this reading that stood out was found on page 236 about repondent behavior. Kalmijn reviews the total survey response patterns of respondents who did not report any friends and negates their responses as those of disinterested respondents, not those of people without any friends. Why was this sort of analysis not discussed in our McPherson et al reading? Again Klamijn's findings reiterate our past readings on educational levels having a positive effect on the number of friends.
In "Urban Families: Conjugal Roles and Social Networks," Elizabeth Bott blended ethnographic and case-study methods to analyze the degree of conjugal role segregation amongst twenty families in London and the correlation to their external social relationships. Bott finds that the more segregation between spousal roles, the more connected the network. She discounts a complete delineation to highly segregated roles to lower SESes, but she does find that they are more likely to be mechanics and working class than the joint professionals. A footnote likely dates this article as from 1954, and I wonder if the economic dependency of the wife on the husband also led to a social dependency, thus making their networks more connected (because the wife has little non-family contact outside of the home), but probably less democratic to the wife. Kalmijn noted in his work that professional contacts were less likely to be shared between spouses, and Bott builds upon this finding, that professionals have less connected networks, in which not all friends know each other. She does not address the size of the networks, perhaps the segregated couples had more connected networks because they were smaller and thus statistically less likely to have non-transitivity?
Bott also considered couples' attitudes towards the role-relationship of husband and wife. Do you think that egalitarian views of husbands and wives is a measure of social capital? Bott presents the example of the Ns in depth, a couple with an extremely connected network but also disgustingly extremely economically dependent wife. Would highly connected networks such as this impede geographic mobility? Bott mentions that connected networks are more likely to form when people continue to live in the same area that they were born, which would gel with the repeated findings in readings that educational levels tend towards more heterogenous networks. As women become more economically mobile, they also gain independent networks.
Claude Fischer, a proponent of subcultural theory, defends urbanism as a way of life by evaluating the validity of the common critique that cities by nature cause a loss of community by interviewing 1050 people microcosm-ically. He analyzes differences in kin reporting between rural and urban respondents and surmises urbanites tend to be more selective in interacting with degrees of kin, in naming their kin (include their sisters and maybe their favorite cousin, but not extended kin they have never met), and replacing nonkin for kin ties. Fischer differentiates between nonkin contacts and hints that urbanites can choose ties on bases of personality and affinity rather than more rural people's fettering to neighbors due to small populations. Urban residents tend to be involved in more organizations, but only because they are more educated, older, and affluent. These increased education and income levels also factor into the slightly higher number of "just friends" urbanites report. I enjoyed Fischer's reading for two reasons, he presented the idea of selectivity and the rise of "modern contexts" in defining your network, two concepts that I have been wondering repeatedly about in prior readings' presentation of modern social network analysis. Fischer concludes that town/city life do not cause these changes in community, but instead occur because of the different stylistic and social characteristics of the residents they attract. Community life involvement is dependent on various social determinants, and its decline, real or hyperbolized, should not be simply stereotyped as an urban anathema.
In "Different Strokes from Different Folks," we return again to our friend Barry Wellman, who explores the specialized supportive functions of different types of ties. His six explanations range from structural (tie strength and contact), structural (groups), cultural (kinship), positional (network members' traits), and relational (similarity/dissimilarity). Wellman conducts a study in East York, Canada, and finds that most active network members provide specialized support, being involved in no more than three of the total six support roles (emotional aid, small services, large services, financial aid, and companionship). I enjoyed this reading because it sort of hints at a rational exchange in our network evolution and also builds upon Wellman's previous assertion that contemporary network communities are more specialized (week 1 reading). He also makes a great point about the choice involved with telephone contact in relation to face-to-face contact and this effect on network tie strength. Wellman concludes by acknowledging the privileged First World only applications of his findings, and summarizes that people seek social support for network problems, emotional problems, and domestic assistance while leaving economic and political issues for institutions and markets.
Both Wellman's (29) and Bott's (20) studies used very small sample sizes, presumedly due to constraints on economic and time resources. Can you think of a better survey design to examine network communities more effectively (increase number of respondents or participants while not being exorbitantly expensive or taking 500 researchers 10 years to execute than the face-to-face surveys?