The Blackwell Companion to Sociology (29 page)

research on tolerance is one area in which scholars have consistently analyzed

religious effects. The general relationship between religiosity and racial prejudice in the USA seems to be curvilinear. Those who are only peripherally involved in

religion are the most racially prejudiced, and both the non-religious and those

who are heavily involved in religion are among the least prejudiced. People's

motivations for attending religious services are also important. Those who

report attending for religious reasons are less prejudiced, those who attend for social reasons more prejudiced (Gorsuch, 1988).

Most past research on tolerance suggests that CPs are less tolerant than other

North Americans. However, a number of scholars suggest that most tolerance

scales (for example, on the General Social Survey) are biased against conserva-

tive religionists because they primarily test tolerance for left-wing and secular groups (atheists, feminists, communists, homosexuals) and exclude most right-wing and religious groups (fundamentalists, anti-abortion protesters, gender-

role traditionalists, ``creationists''). Some evidence supports this claim. For

example, an analysis of the 1987 Freedom and Tolerance Survey, which gave

respondents open-ended questions about their outgroups and intolerance, shows

108

Christian Smith and Robert D. Woodberry

that CPs are just as tolerant as other religious or non-religious groups (Busch, 1998). Certain groups within CPism are extremely intolerant, but they are not

large enough to influence the overall CP mean significantly. Furthermore, CP

intolerance seems to work against not only liberal and secular groups like

atheists, feminists, and homosexuals, but also some right-wing groups such as

militarists and racists. Generally, CPs are not less tolerant of Jews, blacks,

Asians, Catholics, Hispanics, or immigrants than other Americans (Woodberry

and Smith, 1998; Smith, 2000).

Religion and Social Justice

An important emerging dimension of recent work in the sociology of religion has

focused on religion's role in movements of social justice. Moderns have inherited from the skeptical and revolutionary Enlightenment (May, 1976) the view that

religion is naturally conservative, defensive, and allied with ruling elites ± as, for example, the French Catholic Church was in the eighteenth century. Certainly in

many cases religion has proved to be conservative, elitist, and allied with forces of oppression. But religion can cut both ways. It also readily inspires, mobilizes, and supports movements for social justice and democratization (Smith, 1996a).

Historically, in the USA, religious actors and organizations were crucial in the fight for religious freedom in the eighteenth century and against slavery in the nineteenth century. In the mid-twentieth century, the black church in America

was central in mobilizing and sustaining the black civil rights movement. People and organizations of faith in the USA in the decades since then have also played crucial roles in the anti-Vietnam War movement and in the Sanctuary movement

that protected illegal Central American refugees, and were active in the nuclear freeze movement, the Central America peace movement against Ronald Reagan's sponsored wars in El Salvador and Nicaragua, the environmental move-

ment, various progressive organizations such as Amnesty International, Habitat

for Humanity, the American Friends Service Committee, and other peace, anti-

poverty, and anti-discrimination movements and organizations.

Around the world, religion has played a significant role in the liberation

theology movement in Latin America, Africa, and parts of Asia, the overthrow

of the Somoza dictatorship in Nicaragua in 1979, the insurrection against El

Salvador's military regime in the 1970s, Solidarity's resistance to the Polish

communist state, the toppling of the Marcos dictatorship in the Philippines,

resistance to the Pinochet dictatorship in Chile, the anti-Apartheid movement

in South Africa, and the fall of the East German regime in 1989. In many

movements for peace and justice, religion has often been crucial in providing

activists with legitimation for protest rooted in the sacred, supplying insurgents with moral imperatives for love, justice, peace, freedom, and equality, and

offering activists powerfully motivating icons, rituals, songs, testimonies, and oratory. Religion can secure in movements self-discipline, sacrifice, and altruism, furnish movements with trained and experienced leadership and financial

resources, and supply congregated participants and solidarity incentives.

Sociology of Religion

109

Movements often reduce start-up costs by using pre-existing religious commun-

ication channels, authority structures, and social control mechanisms. Religion

can also provide potential common identification among gathered strangers, and

shared transnational identities beyond nation and language. Finally, religion can offer movements transnational organizational linkages, provide the protection

of `òpen spaces'' in civil society, and provide activists with political legitimacy in public opinion (Smith, 1996a). Much exciting work remains to explore the role

of religion not only in sustaining, but also often in challenging, social injustice around the world.

Improving Religion Measures

Many of the standard religion measures used in survey research have become

inadequate for the task. The significance of denominations, for example, in the

American religious field has changed, such that many survey denomination

questions have little analytical usefulness. Many standard theological indicators (for example, ``literal Bible'') yield invalid estimates of what researchers want to study. Improved measures of religious self-identity, practices, beliefs, and organizational location need to be developed and refined and used on surveys. Most

surveys ± especially longitudinal surveys ± have no questions on religion or only a few poorly constructed ones. Fortunately, this is beginning to change. High

quality data are becoming increasingly available and are starting to transform

scholars' perceptions of religion.

In North America, surveys typically have asked if respondents are Catholic,

Protestant, or Jewish. However, this distinction increasingly has little predictive power and researchers have developed more nuanced religion measures. The

greatest progress has been in developing better ways to differentiate groups of

Protestants. Little work focuses on differentiating types of Catholics ± although David Leege and Michael Welch (1988, 1991) of Notre Dame have made

progress in that direction. We still know almost nothing about how to identify

different types of Muslims, Buddhists, Hindus, and other minority religious

groups on surveys (see Leege and Kellstedt, 1993; Green et al., 1996; Woodberry

and Smith, 1998).

In North America, religious denomination is probably the most common way

scholars identify religious groups. With good denominational lists and categor-

ization schemes, scholars can effectively differentiate mainline-liberal and conservative Protestant denominations, as well as pentecostals and black

Protestants. Unfortunately, most denomination questions on surveys lack suffi-

cient detail. Protestant denominations, especially small ones, can be tricky to

code correctly. And many surveys do not distinguish subgroups within larger

religious families. For example, Lutherans are very heterogeneous ± the Evange-

lical Lutheran Church of America tends to be liberal, while Missouri Synod

Lutherans are quite conservative. Good religion analysis requires getting these

kinds of groups coded correctly. Fortunately, scholars (Green et al., 1996;

Steensland et al., 2000) are developing useful schemes for recoding detailed

110

Christian Smith and Robert D. Woodberry

denomination lists, which are being adapted by a number of major surveys,

including the American National Election Survey, the Southern Focus Poll, and

the Adolescent Health Survey.

Another way to identify different religious groups is with religious belief

measures. In North America, ``biblical literalism'' or ``biblical inerrancy'' are the most common belief measures scholars use ± in this case, to identify conservative Protestants. However, these Bible measures are crude. People vary

widely in what they mean by ``literal'' and ``without error.'' And since this

measure excludes many better educated evangelicals who are not literalists or

inerrantists, it makes ``conservative Protestants'' appear less educated and from a lower class than they actually are. Another problem is that scholars generally

only use a single belief measure to categorize religious respondents, which can

cause significant measurement error and biased coefficients. Or they use belief

measures additively, such as identifying conservative Protestants as people who

are biblical literalists, and `born again,'' and have shared their faith with others.

But this stringent procedure can confound the problem, excluding respondents

who actually belong. Ideally, researchers should use multiple beliefs as indicators of a latent belief system (Woodberry and Smith, 1998).

Even so, denominational measures and most belief measures do not distin-

guish different subtypes of conservative Protestants or Catholics, between which there are actually striking differences. Recently scholars have begun asking

respondents which religious tradition or movement they identify with ± funda-

mentalist, evangelical, pentecostal, charismatic, traditionalist Catholic, liberal Catholic, etc. These generally predict attitudes and behaviors better than

denomination or the religious belief measures. As to American religiosity, scho-

lars typically use measures of church attendance, prayer, or subjective import-

ance of religion. These religious measures work equally well for different groups of American Catholics and Protestants (Woodberry, 1998), though attendance

typically has the strongest impact on people's beliefs and behavior. But it is not clear how well these measures work for non-Christian groups. As new surveys

increasingly contact non-Christian respondents, scholars need to develop new

measures that address other forms of religiosity.

Finally, several new methodological techniques promise to enhance our study

of religion. For example, past research has tended to look at differences between religious traditions, but neglected the diversity within traditions. However, some scholars have developed ways to test whether the internal diversity is greater in some religious groups than others (DiMaggio et al., 1996; Gay et al., 1996).

Scholars have also begun to mix qualitative and quantitative research more

effectively. Past qualitative research on religion has suffered because it was

hard to determine whether case studies, `ìnsider documents,'' or interviews

were representative or not. On the other hand, survey research often did not

capture the important nuances, contradictions, and ambivalence in religious

language and culture, and religion researchers often projected alien meanings

onto their research subjects. Fortunately, several groups of scholars are begin-

ning to bridge the gap between qualitative and quantitative research. For ex-

ample, Christian Smith and colleagues (1998) conducted two-hour face-to-face

Sociology of Religion

111

interviews in 23 states with a random sample of churchgoing Protestants whom

they had contacted in a national telephone survey of the United States. This

directly linked national survey data with detailed information about how

respondents understand concepts and the reasons they give for what they believe

and do. Bradford Wilcox (1998) has combined broad reviews of religious family-

advice books with careful quantitative analysis. These kinds of mixed-method

strategies hold much promise for advances in religion research. Finally, religion research is just beginning to explore multilevel designs. For his National Congregations Study, for example, Mark Chaves and colleagues (1999) asked 1998

General Social Survey respondents where they attend church, then used this

information to identify and investigate a representative sample of all US

churches. This procedure could also be used to contact representative samples

of religious-based schools, voluntary organizations, or political groups. Religion research will only improve by moving beyond simply cross-sectional surveys to

creative mixed-method, longitudinal, and multilevel designs.

Globalizing Religion

Religion

Religious transformation is taking place at a global scale. In the past two

centuries, and especially since the Second World War, Christianity has spread

rapidly in Asia, Islam and Christianity in Africa, Protestantism in Latin America, and Hinduism, Buddhism, and Islam have spread to North America, Europe,

and former British colonies. Many new religious movements (NRMs) have also

sprung up by combining elements of different traditions. Population migrations

with religious implications are under way, and old religious cleavages are

reasserting themselves anew in places like Serbia, Palestine, India, and Indonesia.

Meanwhile, Pentecostalism ± a native of the early-twentieth century US West

coast ± is spreading rapidly in many parts of the world. Large segments of many

societies have changed their religions, and members of thesèìmported'' reli-

gions have been disproportionately influential in their home societies, in both the West and the non-West.