Wider fieldsWhere does it meet other fields?
The social-scientific study of religion is the study of how religion works in societies, minds and cultures, without deciding whether its claims are true. Sociology studies religious groups and institutions; psychology, belief and experience in persons; anthropology, religion as lived within a culture; and the cognitive science of religion, why minds find religious ideas easy to hold and pass on.
Study v1.1, 9 November 2026
Religion studied by sociology, psychology and anthropology, which supply the data on belief and Bible engagement. Surveys show Western affiliation and practice falling, flattening in the United States since 2019, and high practice in Africa.
Surveys show Western affiliation and practice falling, flattening in the US since 2019, and high practice in Africa; US Bible use is 38 per cent (2026); Britain's reported 2025 revival was retracted.
Scholars divide on secularisation, whether social science explains religion away, the reality of religious experience, insider and outsider research, and whether "religion" and "spirituality" are usable categories.
Built in the 1960s by Peter Berger, Thomas Luckmann and Bryan Wilson around the functional differentiation of modern societies; Wilson's minimal claim is that religion "ceases to be significant in the working of the social system".
Steve Bruce defends the classic paradigm against all its revisions, holding that the world is secularizing as a consequence of modernity.
Testing the World Values Survey in almost 80 societies, Norris and Inglehart argue that religiosity persists where life is insecure and erodes where it is secure, so that rich publics grow more secular while the world as a whole has more people with traditional religious views than ever.
Berger, who in The Sacred Canopy (1967) had argued that pluralism relativizes religious meaning and encourages secularization, wrote by 1999 that "the assumption that we live in a secularized world is false" and called the world "as furiously religious as it ever was".
He argued that modernity provokes powerful counter-secular movements, with Europe and a Western-educated elite as the exceptions.
Casanova's Public Religions in the Modern World (1994) separates three claims: the differentiation of secular spheres, the decline of belief and practice, and the privatization of religion.
He holds differentiation to be the defensible core, drops inevitable decline, and names the "deprivatization" of religion.
Taylor's A Secular Age treats secularity as a change in the conditions of belief, in which faith becomes "one human possibility among others" rather than the default.
Models religious people as rational actors choosing among religious "firms"; Stark developed the perspective with Roger Finke, Laurence Iannaccone and R. Stephen Warner.
Stark and Iannaccone argued in 1994 that Europe's low practice is a supply-side problem of regulated religious markets, not a lack of demand, and proposed dropping the concept of secularization. Mark Chaves called the pluralism argument "a house of cards", and even supporters conceded that it fits Europe poorly.
Davie described Britain as a place where Christian belief outlasted church attendance.
David Voas and Alasdair Crockett, using the British Household Panel Survey and British Social Attitudes, found that belief has eroded at the same rate as belonging and that the decline is generational.
Berger coined the term in 1967 for an approach that treats religion as a human creation and deliberately leaves the truth of religious belief open: the sociologist brackets the question.
The same Berger urged sociologists to understand religion "from within", from the believer's perspective, in the manner of Weber's Verstehen.
Explains religious ideas largely as by-products of ordinary cognition: a hypersensitive agency detection device, the habit of reading minds and a bias toward purpose.
Its practitioners divide on what that implies for belief; the Stanford Encyclopedia of Philosophy sets out three answers to whether such theories "explain away" religion: that they debunk it, that they support it, or that they are irrelevant to its truth.
Studies what ordinary people do with faith in daily life rather than what institutions prescribe.
McGuire's Lived Religion (2008) urged sociologists to rethink the categories they use, attending to bodies, objects and practice; Ammerman's Sacred Stories, Spiritual Tribes (2014) found that religious communities still animate much everyday spirituality. Ammerman's interviews suggest that for most people religion and spirituality are not separate.
Answering Clifford Geertz's account of religion as a cultural system, Asad argued that such a universal definition carries an implicitly Christian form, and that defining religion is itself a historical, Western and Christian act.
Robbins argued in 2007 that an anthropology of Christianity had "largely failed to develop", because anthropological theory stresses continuity while many Christianities stress radical change.
He argued that anthropologists had treated Christianity as a thin veneer over older beliefs rather than as a culture with its own logic.
Max Weber's The Protestant Ethic and the Spirit of Capitalism is first published, arguing that Protestant religious values helped make possible the rise of capitalism in the West.21
Émile Durkheim's The Elementary Forms of the Religious Life treats religion as the way a society represents and renews itself.22
Victor Turner's The Ritual Process, from his work among the Ndembu of Zambia, gives the field "liminality" and "communitas".24
Rodney Stark and William Sims Bainbridge set out a deductive theory built from exchange and "compensators" in A Theory of Religion, with 344 propositions.8
Berger writes that "the assumption that we live in a secularized world is false", arguing that modernity provokes powerful counter-secular movements.5
Pascal Boyer's Religion Explained (2001) and Justin Barrett's Why Would Anyone Believe in God? (2004) set out the "standard model" of the cognitive science of religion.27
The STEP trial of intercessory prayer, which randomized 1,802 heart bypass patients in six hospitals, is published; strangers' prayer did not reduce complications.28
Joel Robbins argues that an anthropology of Christianity had "largely failed to develop", because anthropological theory stresses continuity while many Christianities stress radical change.18
Gallup finds US church membership below half for the first time, at 47 percent, against 73 percent in 1937 and 70 percent in 1999.29
The Bible Society withdraws its 2025 "Quiet Revival" report on 26 March after YouGov finds fraudulent respondents in the opt-in polls on which it rested. The EPIC survey of 7,453 US congregations, released on 24 April, finds median in-person attendance at 70, the first rise in 25 years of tracking.3233
Separate the claims being tested: the decline of belief and practice, the differentiation of institutions and declining religious authority are distinct, and a country can secularize institutionally while keeping high participation6.
Prefer probability samples, in which every person has a known chance of selection, to online opt-in panels, which recruit volunteers through advertising and email lists34.
Read every figure with its wording, mode and sample: British Social Attitudes asks attendance only of those who name a religion, and the General Social Survey's move to several modes in 2021 affects religion trends2635.
Treat self-reported attendance as generally over-reported36.
Use interviews, ethnography and congregational studies to reach meanings, practices and the stories people tell, taking the congregation as the unit1537.
Test causes by experiment, and give large, preregistered and replicated work more weight than a single study, a small sample or an unregistered analysis3839.
Keep association apart from cause: large representative samples can show an association that cross-sectional data cannot show to be a cause40.
Follow the ethics of human subjects research, with review and informed consent, and treat religious belief as special personal data under South Africa's POPIA and the European GDPR41424344.
The works below shaped the field and are cited in this study, from the earliest to the most recent. Hover over a book, or select it, to see its author and the sources behind it.
Listed because this study draws on their published work; no association with or endorsement of Numa Verse is implied. Suggest a correction.
The words that carry the most weight in this field, each defined from the sources of this study. Turn a card over to read its definition.
Scholars divide. Western Europe's decline and America's slower one are widely accepted, as is Africa's high practice; what divides scholars is whether decline is the path of modernity, the product of insecurity's end, or the result of regulated religious markets329. In the United States the Christian share has been relatively stable since 2019, at 62 percent in 2023 to 202430.
The American Bible Society, in ministry research fielded on a probability panel, counts 38 percent of US adults as Bible users in January 2026, reading Scripture outside worship at least three or four times a year, and 25 percent as reading weekly or more; Bible users were 50 percent in 20216364. In Gallup's 2022 survey, 20 percent of Americans called the Bible the literal word of God, 49 percent inspired but not all literal, and a record 29 percent fables and history recorded by man65.
A reported "Quiet Revival" rested on two YouGov opt-in polls; the Bible Society withdrew its 2025 report on 26 March 2026 after YouGov found fraudulent respondents32. The random-sample Labour Force Survey and British Social Attitudes showed no revival, and weekly attendance among Christians fell from 20 to 13 percent between 2018 and 20253466.
In the STEP trial, published in 2006, 1,802 heart bypass patients were randomized into three groups; strangers' prayer did not reduce complications, and deaths did not differ28. A Cochrane review of ten trials with 7,646 people found intercessory prayer "neither significantly beneficial nor harmful for those who are sick"67. Such trials do not measure prayer's meaning for those who pray.
The Stanford Encyclopedia of Philosophy sets out three answers to whether evolutionary and cognitive theories do so: that they debunk religion, that they support it, or that they are irrelevant to its truth13. Peter Berger's methodological atheism is a fourth: the sociologist brackets the question12.
Reviewed and approved by Duane Herholdt, BA (Theology), Chief Executive Officer, Numa Verse
Study version: v1.1, 9 November 2026 (theology map v1.27)
Last reviewed: 9 November 2026
