Sources of Influence and Influences of Agency 155
not freely made – there is no social vacuum that would allow such freedom. Choices
are embedded in social relations that influence both the development and dynamics
of preferences, as well as the options available and choices taken (Akerlof 1997; Sen
1973,1993; Sherkat 1997).
Religious commitments are a function not only of socialized preferences but also
factors intrinsic to the individual and exogenous to the religious choice. Furthermore,
social influences may have nothing to do with the understandings that constitute re-
ligious preferences, and hence are not socialization influences even though they may
direct individuals’ behaviors. In the remainder of the chapter, I will discuss research on
agents of influence while keeping in mind the distinction between socialization and
social influences.
AGENTS OF INFLUENCE
Parents and Family
Across cultures and history, the family is the primary source of information about
supernatural explanations. Parents and relatives teach children understandings about
supernatural things, and this source of information has temporal and affective pri-
macy – both of which are important for influencing preferences. Surprisingly, many
studies in the sociology of religion contended that parents have limited influence on
children’s religious commitments (e.g., Hoge et al. 1994). These studies accepted com-
monly articulated assumptions about growing generational differences in values and
commitments – the generation gap thesis that led many scholars to assume that radical
shifts in religiosity were on the horizon (e.g., Wuthnow 1976; Bellah 1976). However,
most systematic research and more studies employing national samples and longitudi-
nal data from parents and children have demonstrated that parental influences domi-
nate religious beliefs and attachments throughout the life course (Acock and Bengtson
1978; Acock 1984; Willits and Crider 1989; Myers 1996; Sherkat 1998).
Parents and Children
The systematic study of parental influences on children’s religious preferences for reli-
gion began with Newcomb and Svehla’s (1937) study of 558 parents and children – in
which they found that mothers’ attitudes toward religion explained 34 percent of the
variation in sons’ religious understandings and 48 percent of the variation in daughters’
religious preferences. Since this early work, many studies have concluded that par-
ents have a substantial effect on children’s religious beliefs and behaviors (Hunsberger
1985; Acock and Bengtson 1978; Acock 1984; Willits and Crider 1989). Generally, these
studies assume that parental influences are limited to earlier periods of the life course
and that the crystallization of belief is achieved in the early life cycle. Later researchers
borrowed lifelong learning models from political socialization (cf. Sigel 1989) and in-
vestigated how parental effects continue over the life course. Parents help shape other
social ties, and this channels lifelong socialization. Indeed, life course events may make
parents more influential as young adults seek wisdom from parents on how to raise
children of their own and deal with stressful life events (Stolzenberg et al. 1995; Myers
1996; Sherkat 1991a). Examinations of parental socialization have tended to focus on
religious affiliation and participation – noting how parents’ participation early in the