Religion and Society

Sociologists disagree about how to define religion, and this essay describes the three main approaches — substantive, functional and social constructionist — and assesses their strengths and weaknesses, arguing that each captures something important but none is fully satisfactory.

Substantive definitions. These define religion by its content: belief in the supernatural, a god or the sacred (Weber). The strength is clarity — this matches everyday ideas of religion and makes it easy to identify. The weakness is that it is exclusive and ethnocentric: it imposes a Western, theistic model and can exclude belief systems without a god, such as some forms of Buddhism.

Functional definitions. These define religion by what it does — providing meaning, solidarity and answers to ultimate questions (Durkheim, Yinger). The strength is inclusivity: it covers diverse belief systems across cultures. The weakness is that it is too broad — by focusing on function, it could include things most people would not call religion, such as nationalism or strong political ideologies.

Social constructionist definitions. These argue there is no single fixed definition: religion is whatever members of a society define as religious (Aldridge). The strength is that it avoids imposing one culture's view and recognises that definitions vary. The weakness is that it makes comparison and measurement very difficult.

Assessment. Each approach trades off against the others: substantive definitions are clear but narrow; functional definitions are inclusive but vague; constructionist definitions are flexible but unworkable for comparison. The choice matters because it shapes major debates — a substantive definition makes secularisation look strong, while a functional or inclusive one makes religion look alive and well.

Conclusion. There is no perfect definition of religion. For cross-cultural sociological study, a functional or carefully-bounded definition is most useful, but every choice involves a trade-off between clarity and inclusivity. The difficulty of defining religion is therefore real and unavoidable, and good sociology states which definition it is using and why.

Religiosity — the degree of religious belief and practice — appears to vary between social groups. This essay assesses the patterns by gender, age, ethnicity and class and the social explanations for them, arguing that there are real and well-explained differences, although measurement problems mean we should be cautious.

The patterns and their explanations. Women are generally more religious than men: explanations include differential gender socialisation (women raised to be more nurturing and passive), women's traditional roles leaving more time and need for religion, and risk-and-compensation theories. Older people are generally more religious than younger people, explained by the ageing effect (reflection near death), the generational effect (raised when religion was stronger) and disengagement. Many minority ethnic groups are more religious, explained by cultural defence (religion preserving identity against discrimination) and cultural transition (religion supporting migrants). Class shapes which religion people follow more than how much — working-class groups may favour sects or Pentecostalism, middle-class groups established churches or New Age spirituality.

The case for caution. These patterns are not certainties. Measuring religiosity is difficult: belief, belonging and practice differ, and Davie's 'believing without belonging' shows attendance figures underestimate belief. Patterns also vary between societies and faiths, and the choice of definition affects what counts. Feminists add that the gender pattern reflects religion's patriarchal control of women, not simply greater female piety.

Evaluation. The patterns are well-evidenced and consistently explained by social factors — socialisation, roles, identity and deprivation — which supports the sociological view that religiosity is shaped by society, not nature. But measurement problems and cross-cultural variation mean the patterns are tendencies, not laws.

Conclusion. Some social groups are clearly more religious than others, and these differences are best explained by social factors. However, because religiosity is hard to define and measure, the patterns should be treated as well-supported tendencies rather than fixed rules. The view is therefore largely correct, with important qualifications about measurement.

Measuring religious belief is one of the central problems in the sociology of religion. This essay assesses these difficulties, arguing that while measurement is genuinely problematic, some useful information can still be gathered if its limits are understood.

The core difficulty: belief, belonging and practice differ. Religiosity has at least three dimensions — private belief, formal belonging (membership) and practice (attendance) — and they often do not match. Davie's 'believing without belonging' captures this: many people still believe but no longer attend or belong, so attendance figures underestimate belief. Conversely, some attend for social reasons without strong belief, so attendance can overstate it.

Problems with statistics. Religious statistics raise validity problems — attendance measures behaviour, not belief. They raise reliability problems — different organisations count membership differently, and self-reported religiosity may be exaggerated or understated through social desirability. 'Religiosity' is also hard to operationalise: different measures (belief in God, attendance, prayer, self-description) give very different pictures. Official figures also miss new, private and non-institutional forms of religion, such as New Age spirituality.

The case that measurement is still possible. Despite these problems, sociologists can gather useful data by using multiple measures together (triangulation) and by being clear about what each measures. Large surveys can track trends over time, and qualitative research can access meanings that statistics miss. The difficulty is real but not total.

Evaluation. The strongest point is that the choice of measure shapes conclusions: if we only count attendance, religion looks like it is dying (supporting secularisation); if we count private belief and new spirituality, it looks much healthier. This is why the secularisation debate is so contested. Measurement is therefore not neutral — it is bound up with how religion is defined.

Conclusion. Measuring religious belief is genuinely difficult because belief, belonging and practice diverge and statistics are limited. However, careful, multi-method research can still produce valuable, if provisional, knowledge. The key is to recognise that every measure captures only part of religiosity, so conclusions — especially about secularisation — must be drawn cautiously.

It is widely observed that women tend to be more religious than men across many faiths and societies. This essay evaluates this view and the explanations for it, arguing that the pattern is real but that its causes — and its meaning — are contested.

The case that women are more religious. Across many societies women show higher levels of attendance, membership, prayer and belief than men. Several explanations are offered. Differential gender socialisation means women are raised to be more nurturing, caring and emotionally expressive — qualities aligned with religious participation. Women's traditional roles (caring, child-rearing, historically less paid work) may give more time for, and need of, the meaning and support religion provides. Risk-and-compensation theories argue women are more risk-averse and turn to religion for security, and that those who are deprived or marginalised (more often women) seek compensation in religion.

Qualifications and counter-arguments. The pattern is not universal or fixed: in some religions and societies men are more publicly religious, and women's higher religiosity may be declining as their roles change. Feminists offer a sharp reinterpretation: women's greater involvement reflects religion's patriarchal control of women rather than simply greater female piety — religion has historically confined women to particular roles and used them to transmit faith. Secularisation theorists note that women's religiosity is now falling as female employment rises, suggesting the pattern was rooted in social roles, not gender essence.

Evaluation. The empirical pattern is well-supported, and the social explanations (socialisation, roles, risk) are convincing because they explain it without appealing to anything natural. The feminist reinterpretation adds important depth: the same pattern can mean female devotion OR patriarchal control. The fact that the gap narrows as women's roles change strongly supports the social-factors explanation.

Conclusion. Women are generally more religious than men, and this is best explained by social factors — socialisation, roles and risk — rather than by nature. But the pattern is changing as women's roles change, and feminists rightly warn that higher female religiosity may reflect religion's control of women as much as women's faith. The view is correct but must be understood through its social causes and its contested meaning.

Defining religion is a notoriously difficult problem in sociology. The view that there is 'no satisfactory way' to do so reflects the failure of any single definition to win agreement. This essay evaluates that view by examining the substantive, functional and constructionist approaches, arguing that while no definition is perfect, this does not make the task pointless.

The case that there is no satisfactory definition. Each approach has a serious flaw. Substantive definitions, which define religion by belief in the supernatural (Weber), are clear but exclusive and ethnocentric — they impose a Western, theistic model and may exclude non-theistic traditions such as some forms of Buddhism. Functional definitions, which define religion by what it does — providing meaning and solidarity (Durkheim, Yinger) — are inclusive but far too broad, potentially counting nationalism, sport or political ideologies as 'religion'. Social constructionist definitions (Aldridge), which say religion is whatever members define as religious, avoid imposing one view but make comparison and measurement almost impossible. Because every approach fails in some way, the view appears justified.

The case against the view. A definition need not be perfect to be useful. Each approach captures something genuinely important about religion — the supernatural, social functions, or members' own meanings. Sociologists can adopt a definition appropriate to their particular research purpose and state it clearly, which allows meaningful study even without universal agreement. Indeed, the lack of a single definition reflects the genuine diversity of religion across cultures rather than a failure of sociology.

Evaluation. Judged by the criterion of producing one universally agreed, watertight definition, the view is correct — no such definition exists, and the trade-off between clarity and inclusivity seems unavoidable. But judged by the criterion of whether religion can be studied usefully, the view is too pessimistic: workable, purpose-specific definitions allow real research, and the choice of definition itself becomes an analytical tool (for example, it shapes whether secularisation looks strong or weak).

Conclusion. There is no single, universally satisfactory definition of religion, so in that strict sense the view is right. But this does not mean the task is hopeless: each approach illuminates part of religion, and sociologists can define it appropriately for their purpose while being transparent about the trade-offs. The absence of one perfect definition reflects religion's diversity, not the failure of sociology — so the view is exaggerated.

Why are some people and groups more religious than others? This essay assesses the view that religiosity is best explained by social factors — such as gender, age, ethnicity and class — against alternative explanations, arguing that social factors are the most powerful explanation but that measurement problems and other influences must be acknowledged.

The case for social factors. The patterns of religiosity correspond closely to social position. Women are more religious than men, explained by gender socialisation, roles and risk-and-compensation theories. Older people are more religious, explained by ageing, generational and disengagement effects. Many minority ethnic groups are more religious, explained by cultural defence (preserving identity against discrimination) and cultural transition (supporting migrants). Class shapes which religion people follow, with deprivation linked to sects and Pentecostalism (compensation and community). That religiosity maps so neatly onto social groups strongly suggests it is socially produced, not natural or purely individual.

Alternative explanations. Other factors also matter. Psychological explanations stress individual needs for meaning, comfort and certainty, which are not reducible to group membership. Theological or personal explanations point to genuine individual faith and experience. Secularisation theorists argue that broad social changes — rationalisation, science and modernity — explain declining religiosity better than any single group factor, while postmodernists argue religiosity now reflects individual 'spiritual shopping' rather than fixed social categories.

The measurement complication. Any claim about religiosity is limited by the difficulty of measuring it: belief, belonging and practice differ (Davie), so the patterns themselves depend on what is measured. This does not refute social explanations but cautions against treating them as exact.

Evaluation. Social factors are the strongest explanation because they account for the consistent, patterned differences between groups without appealing to anything natural, and because the patterns change as social conditions change (for example, women's religiosity falling as their roles change). However, social factors work alongside individual needs and broad processes like rationalisation, so they are best seen as the primary, not the only, explanation.

Conclusion. Levels of religiosity are best explained by social factors, which convincingly account for the differences between gender, age, ethnic and class groups. But the explanation is strengthened, not weakened, by recognising that individual needs and wider social change also play a part, and that measurement problems mean the patterns are tendencies. Social factors are the best explanation, but not a complete one.