Sources of Influence and Influences of Agency 157
influenced their parents’ beliefs in the Bible, leading their elders to become less ortho-
dox in their interpretation of scriptures. Later in the life course, older parents pulled
the adult children back toward more conservative religious beliefs. A similar pattern
might be expected in revolutionary Iran, as young religious activists led their parents
and other relatives toward preferring particular Islamic beliefs. Later, as the revolution
lost its flare and the realities of living adult life under religious constraints sunk in,
older Iranians from more moderate generations probably became more influential in
defining their children’s religious commitments.
Spousal Influences
Marital ties are also important sources of influence, and religious intermarriage is one
of the strongest predictors of changes of religious affiliation (Lazerwitz et al. 1998;
Lazerwitz 1995a, 1995b; Sandomirsky and Wilson 1990; Sherkat 1991b). Importantly,
however, the direction of switching follows a particular pattern – what Stark and Finke
(2000) call “Greeley’s law” – that the more religious spouse has more influence over
the direction of change. Typically, this has meant that intermarriage with Catholics
generates switching into Catholicism, and that people who marry members of exclusive
sects tend to switch into the sect. Of course, intermarriage is also related to underlying
religious preferences, as people with strong valuations of particular religious goods will
be unlikely to marry someone who doesn’t share their desires. This selection bias tends
to minimize the influence of spouses on religious choices. When people have strong
religious preferences they will be unlikely to choose a mate who differs, and those with
weak religious preferences who are more likely to intermarry would exert little influence
on their partners (McCutcheon 1988; Johnson 1980).
People choose their friends and spouses in accordance with preferences; hence, val-
ued others are likely to reinforce existing desires rather than arouse new ones. Because
preferences also drive educational and occupational choices, this will tend to consoli-
date social ties across varied fields of social life (Darnell and Sherkat 1997; Sherkat and
Blocker 1997). Homophily strongly influences the composition of voluntary groups,
and social movements of all kinds are populated by people with similar backgrounds
and opinions (McPherson and Smith-Lovin 1987). Together, these theoretical expecta-
tions and the supporting empirical research suggests that macrostructural connections
are less important for the formation of preferences for collective goods, and instead that
individuals’ preferences drive their connections to social groups (whether families, oc-
cupations, neighborhoods, or social movements). Here, I argue against the macrostruc-
turalism that dominates explanatory frameworks in social exchange theory (e.g., Lawler
et al. 1993), and call for less minimalist conceptions of actors. Thickening the view of
actors’ motivations will help identify how people choose many of the structures of
which they are a part, thus lending agency to the framework and allowing for testable
hypotheses regarding the influence of networks on individuals, and of individuals on
networks.
Family Research and Socialization
The late twentieth century saw a flurry of sociological research on the religion-family
connection, yet data constraints hamper progress in the assessment of how family re-
lations influence religious beliefs and commitments and vice versa. Very few studies
track both parents and children over the life course, and fewer still have employed