Jump to ContentJump to Main Navigation
Religion in Secular SocietyFifty Years On$

Bryan R. Wilson and Steve Bruce

Print publication date: 2016

Print ISBN-13: 9780198788379

Published to Oxford Scholarship Online: October 2016

DOI: 10.1093/acprof:oso/9780198788379.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2017. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date: 26 February 2017

(p.231) Appendix One: Criticisms of the Secularization Thesis

(p.231) Appendix One: Criticisms of the Secularization Thesis

Religion in Secular Society

Steve Bruce

Oxford University Press


Although Bryan Wilson’s argument was very widely accepted by social scientists and by liberal (or doom-saying) Christians, there were dissenting voices in the 1960s, and these grew in number subsequently. This appendix will explain and to some extent evaluate the most significant criticisms.

Criticisms Historical and Empirical

David Martin is a contemporary of Wilson who shared a similar social background. His family was working class: he rather delighted in telling people that his father chauffeured Walter Runciman, liberal peer and grandfather of social theorist W. G. ‘Garry’ Runciman. Like Wilson, Martin found his way to university via a number of unorthodox and largely autodidactic channels.1 But he differed from Wilson in being a committed Christian (Methodist first and then Anglican) and in being sceptical of the neatness of Wilson’s secularization thesis.

Some of Martin’s earliest published works raised a number of serious criticisms long before the American supply-side theorists (of whom more shortly) discovered them.2 One objection was that, while all sociological explanation involves simplifying the complexity of social reality into an artificially neat narrative, Wilson’s thesis went beyond what is reasonable in such compression. Furthermore, he made his argument unduly convincing by his selection (or, a critic might say, creation) of the imagined starting points for the changes that he glossed as secularization. Wilson’s premodern world is a simplified version of the feudalism of the Middle Ages which exaggerated the influence of the Christian Church on society and on individual personality and behaviour. Some of that criticism is perhaps better directed against Peter Berger, whose popularizing of the phrase ‘sacred canopy’ implied that a (p.232) shared religion enveloped and encompassed the entire society and people.3 Historians (most notably Keith Thomas in his Religion and the Decline of Magic) have presented considerable evidence from church court records that many people’s Christianity seemed lacking in what we regard as marks of piety and might better be regarded as mere conformity to social expectations.4 For example, it was apparently common for people who attended church to mill around (churches were not then seated), engaging in secular conversations and business. In response, Wilson could have argued first that, precisely because church court records were an account of deviation, reliance on them exaggerates it. Second, we know from many sources that, though people were casual in their behaviour in church, they fell silent and kneeled when the Host was raised. Though the illiterate majority were not terribly well-informed Christians, they knew the Lord’s Prayer and the Hail Mary and made the Sign of the Cross. They knew the Ten Commandments, the Four Cardinal Virtues, the Seven Deadly Sins, and the Seven Works of Mercy. They paid their taxes to the Church. They were baptized, christened, married, and buried by the Church. Even the most humble left what they could to the Church in their wills, and the wealthy left large sums to pay for priests to say post-mortem masses for their souls. Almost everyone knew they had to make reparation to God for their sins, in this life or the next.5 And none of that is now true.

Third, the case that the popularity of such expressions of religious sentiment as church attendance, private and family prayers, church-sanctified rights of passage, and the like showed conformity rather than commitment has at its centre the rarely-asked question of just who was being obeyed or emulated. For patterns of behaviour to be so well established that those who do not share the beliefs that inform the behaviour nonetheless feel that their social status or character requires they conform, it must be the case that most people (or possibly most of the people who matter) are committed. Social practices such as trying to ensure that people tell the truth by having them swear oaths on the Bible might continue for some good time after the beliefs that inform them—there is a God who punishes sin and the Bible is a uniquely holy book—have become minority views, but they are not invented by an unbelieving culture. For conformity to explain anything, there must be a preponderance of genuine commitment, so noting that some apparent piety was coerced actually changes very little.

(p.233) A related empirical weakness is that in both subject matter and chronology Wilson skips from the Middle Ages as the acme of religion’s institutional power to the mid-nineteenth century, which was probably the highpoint of popular participation inspired by Protestant piety. Were we to start with the seventeenth or eighteenth centuries, we would have had a considerably lumpier trajectory.6 Wilson could, of course, defend his approach by noting that it is precisely the role of theory to pick out of uneven historical detail some explanatory thread. Other start dates or a more detailed and nuanced approach to whatever one takes as the premodern world that was lost to secularization would have muddied the explanation and made it harder to communicate but not fundamentally altered it.

More generally, historian critics have argued that the lack of attention to detail in what is taken to be the trajectory of decline means that better explanations are overlooked. For example, Callum Brown (and to a lesser degree Hugh McLeod) argue against the sociological notion of long-term decline caused by modernization by claiming that church involvement remained fairly stable until the 1960s, when major changes in popular culture and in the roles of women abruptly wrecked a religious culture that had been little dented by centuries of ‘modernization’.7 Arguably Brown’s ‘events’ alternative to foundational social change also involves a rather selective attitude to evidence: for example, it is true that church membership and baptism rates remained fairly stable through the first half of the twentieth century, but the more rigorous (and hence more sensitive) measures of church attendance, sabbath observance, and family prayers all showed considerable decline. In the second half of the nineteenth century attendance at most chapels was twice the membership; by the 1950s it was only half the membership. The most likely explanation of the changing relationship between the two indices of religious interest is that personal commitment to the key religious precepts declines first, while elements which have some strong secular appeal (such as the family ceremony of baptism) remain popular longer. Wilson could reasonably respond that such challenges are elaborations of his thesis rather than refutation; had he wished to write a very much longer book, they could have been accommodated.

A very different line of criticism concerns ideology. Martin argues that much use of the idea of secularization is partisan. Like Marxists who argue that capitalism will inevitably be replaced by the classless society while actively campaigning to make it so, those who wish to be rid of religion use the notion that modernity undermines religion in order to justify their political attempts to hasten that end. This criticism can largely be answered with greater (p.234) consistency of terms. We need to separate secularity (the condition of not being religious) from secularism (the philosophical justification for that outcome) and secularization. For sociologists, secularization is a social process driven largely by inadvertent and unintended consequences. For example, in Protestant societies, the gradual weakening of the state’s support for one single variety of Christianity was a response to the factionalism and schism that created large numbers of competing Protestant sects. When Thomas Jefferson famously promised a clear wall of separation between Church and state, he was responding not to atheists but to Baptists who feared that they would be dominated by Presbyterians and Episcopalians.8 There have been secularists—that is, promoters of secularism—but until the late twentieth century they were few in number, and their historical role was largely confined to philosophical justification for changes that were already underway. Anyway, even were it the case that the sociological notion of secularization was popularized by ideologically motivated secularists, that would tell us little or nothing about the validity of the secularization thesis. That must rest on the cogency of the argument and its fit with the available evidence.

A more complex criticism is that the secularization thesis is teleological in the sense that it rests on an ‘in order to’ rather than a ‘because’ foundation. Secularization is seen as an attempt to reach a specific end point and thus requires that social change be driven by some sort of plan: the hidden hand of an anti-God, if you like. From the 1970s on, that line of argument was reinforced by adopting the post-modern critique of anything that looked like an ‘Enlightenment project’. For postmodernists such value judgements as the belief that the present is better than the past and that the future will be better yet are entirely improper and so any sociological explanation of change that looks like a morally sanctioned developmental model is improper. Wilson would have no difficulty rebutting that criticism. The teleological criticism only looks at all plausible because it plays tricks with words: presenting the religious and the secular as if they were mirror images of each other when they are actually a comprehensive covering pair. ‘Secular’ in the secularization thesis means only that which is not religious. The religious is something specific, the secular is simply its absence: that is, everything else. The religious and the secular are like smoking and not smoking. And the ‘Enlightenment project’ argument, like much of this sort of criticism, is plausible only for philosophers who do not understand the idea of unintended consequences. No doubt many philosophers of the Enlightenment (especially in France) were secularists. August Comte famously expected the irrational religion of Christianity to be replaced by the rational alternative of—wait for it!— sociology. No modern sociologist holds such a view: the sociological secularization thesis developed by Wilson largely rests on unanticipated consequences.

(p.235) Finally, it is worth noting that, though he proposed ‘eliminating the concept of secularization’, Martin himself produced A General Theory of Secularization and two decades later a subsequent clarification.9 The main difference between Martin’s general theory and Wilson’s work is the degree of generalization. Both agree that the changes in economy, polity, and society glossed as modernization changed the nature and status of religion. Wilson argues that modernization reduces the social influence of religion and it does so in two forms. In the European form, people gradually abandon the churches. In the US form, church involvement remains popular but the churches themselves become somewhat secular, and their continued appeal rests on secondary secular social functions. Martin’s general theory supposes many of the same foundational causes as Wilson’s but argues that differences in religious culture and in political context create very different patterns of secularization. In the states of northern Europe, for example, the national Lutheran churches remain popular monopoly institutions for secular reasons (such as providing social services and legitimating national integrity). In the UK the more radically ‘reformed’ Protestantism produced so much competition between national churches, denominations, and sects that the state had to withdraw its support for a national church (or to be more precise, four such churches: one each for England, Scotland, Wales, and Ireland) and replace much of its activities with secular provision.

It is possible to reconcile the views of Wilson and Martin because in one important sense they are both on the same side.10 In the end much of the criticism of Wilson’s work stems from disciplinary differences. Because they tend to study changes in specific places over short time periods, British historians (there is a very different continental tradition) tend to dismiss sociological explanation as ‘sweeping generalization’.11 And many of those who dismiss sociological explanation as an Enlightenment project are philosophers and students of literature who study change through the works of prominent thinkers. Wilson and Martin are both sociologists who seek causes of change in unintended and inadvertent consequences of human actions. Once we take the view that Martin’s criticisms bear most heavily on liberal Christians and atheists who borrow some sociological language for their own very different purposes, the key differences between Wilson and Martin concern oversimplification and causal primacy. Can one compress the effects of modernization on religion into one bifurcated model or does one need more detail to avoid distortion?12 And is religion always epiphenomenal?

(p.236) At issue here is the extent to which cultural forces can autonomously influence economy, polity, and society. Although Max Weber’s famous ‘Protestant Ethic and the Spirit of Capitalism’ essay offered an early and popular example of religious change influencing society, the structural-functionalism which dominated sociology in the 1950s, and which Wilson implicitly endorses, tended to mirror structuralist Marxism in supposing that the real drivers of change lay in the deep structures of society and that culture was very much secondary. Martin is certainly right about the way in which British sociologists of the 1960s tended to dismiss religion as of no account except as a symptom of change, but Wilson’s work in general is hardly susceptible to this criticism: his many studies of sectarianism, for example, took very seriously the autonomous influence of religious ideas. And, just as the secularization thesis is not compromised by the fact that some secularists used it to bolster their agenda, it is also not ruined by having been adopted in a casual way by materialist sociologists (of Marxist or structural-functionalist hue) who mistakenly ignored the power of religious ideas and ideologies.

In brief, Martin’s criticisms have much to commend them, and they should have been taken more seriously by Wilson, but it seems possible to accept them without entirely ditching the secularization thesis as it is elaborated by Wilson. However, there is one school of criticism which both is sociological (in that it accepts that social change is often the result of abstract forces beyond the immediate knowledge of people in any particular time, place, or situation) and requires the wholesale abandonment of secularization, in both its Wilsonian and Martinesque forms: Rodney Stark’s supply-side approach to religious change.

Criticisms Theoretical

The rational choice, market, or supply-side approach to religious change associated with the American sociologist Rodney Stark and his students and colleagues offers a radical alternative to the secularization thesis.13 Wilson’s thesis supposes that what needs to be explained is the decline in religious interest (or, in the language of economists, ‘demand’) associated with modernization. Stark et al. suppose that the underlying demand for religion is stable and that such differences as we see over time and place are due to differences in the supply of religion.14 From that starting point, Stark has (p.237) produced a bewildering array of propositions, and economists such as Laurence Iannaccone have added more, but the key points are simple enough.

In the 1830s, Alexis de Toqueville, a minor French nobleman, toured America and made a series of penetrating comparisons of the old world of Europe and the new world of the United States. He was particularly struck by this irony. In France, where there was just one Church, large parts of the population ignored or despised it. In America, there were lots of competing churches and they were growing. Where religion is voluntary and there is a lot of choice, it will remain popular: where there is a state church and little or no competition, the incipient demand for religion will be suppressed, and religion will decline.

As the argument has been elaborated, there are a number of separate points. First there is the problem of unnecessary alienation or delegitimation. In pre-Revolutionary France, where the one Church was firmly on the side of the political status quo, those social classes which wanted to challenge that order became disillusioned with the Church, and, as there was no religion that was not the Catholic Church, with religion itself. As there was no state church in America promoting the interests of the ruling class, those people who wanted to express social and political dissent could do so without abandoning religion. They could become members of a lower-class dissenting sect.

Second, there is the issue of clergy motivation The state church system contains little incentive for the clergy to recruit a congregation. The nineteenth-century Church of England vicar or Church of Scotland minister derived his living from the rental value of lands given to the Church centuries earlier by monarchs and pious lords, or from a tax on local landowners. There was no direct material reason for such a clergyman to increase his congregation or to make himself popular. For personal reward and advancement, it was much more important that he curry favour with his social superiors who controlled Church appointments. The clergy of the American churches ate only if they attracted congregants who were willing to pay them. In a voluntary system there is a very strong incentive for preachers to make their message popular.

Third, there are a series of general virtues that are found in all competitive free markets (as contrasted with monopolies, especially if state supported). Where there is a state monopoly—and this is as true of religion as of cars—the product will not be as well adapted to the demands of consumers as when (p.238) there is a large number of organizations competing in a free market, all needing to win customers to survive. Instead of there being one state church, America had a vast range of competing religions and hence—the Heinz 57 varieties model—there was something that would suit everyone. That the voluntary system produces greater variety is not an accident but stems from that fact that, even if the state does not actively support a hegemonic church, the existence of one dominant organization raises the ‘entry’ costs for any new provider, and thus discourages innovation.

There is much of value in the supply-side model. Indeed, Wilson himself makes observations that fit it perfectly. For example, he notes that, ‘however much competition there may have been between denominations in America, [it] was on similar terms, without any one Church being able to claim superiority or advantages in the social and political sphere’.15 However, there are also problems which stop it adding up to a convincing explanation of US exceptionalism.

A problem with de Toqueville’s claim that the presence of a state church alienates people from religion if they become alienated from the political status quo is that, if it works at all, it works only for Catholic countries. Because Catholicism argues that the Vatican has a monopoly of religious truth, it encourages a radical choice. You either accept what the Church says or you abandon religion. The residue of that polarization can be seen in twentieth-century politics. Where other European democracies produced a range of centre, centre-left, and centre-right parties, Spain, Italy, Portugal, and France divided into two big competing blocks: the Catholic Church and its conservative supporters on one side and on the other the secularism of communist and socialist parties. But in Protestant cultures there is the intellectual possibility of breaking away. If you believe that all people can equally well discern the will of God and that there is no need for bishops and priests, then you can form your own religious organization. And many British people did just that. If they didn’t like the close ties between the landed gentry and the Church of England, they could join the Methodists. If the Wesleyan Methodists were too staid and middle class, they could join the Primitive Methodists. The contrast De Toqueville makes between state-supported religion and a competitive free market in religion does not work well for Britain, because, although there was a state Church of England and a state Church of Scotland, there were also lots of dissenting denominations and sects. According to the 1851 Census of Religious Worship, more than half of the people who attended church did so in some dissenting (that is, not state church) outlet.16

The point about differing clergy motivations in controlled and competitive religious markets, which Stark makes with a contrast of Lutheran clergy in (p.239) Germany and the USA, is a good one, but, again, it fails to explain the differences between Britain and America.17 The clergy paid from public finance may have lacked any incentive to do their jobs well, but there were plenty of ministers in dissenting denominations who had every good reason to seek popularity; by the 1850s, half of Britain’s clergy were outside the state churches.

Finally, there are many difficulties in conceptualizing and measuring religious pluralism or diversity that cast doubt on the evidence that Stark and his associates produce to support their contrast of different types of religious markets. De Toqueville built his impressions of America while travelling. America as a whole had, and still has, a great range of religious alternatives, but any one part of America—the places that people actually inhabit—showed a great deal of religious concentration. Utah was overwhelmingly Mormon. The southern states were conservative Baptist or Methodist (and they are very similar). Catholics were concentrated in Florida, Louisiana, East Texas, and Southern California. Large parts of the Mid-West were Lutheran. Pluralism would feature in people’s choices only if it was the case that they could pick between a variety of churches. In practice most could not. The religious historian Edwin Gaustad defines homogeneity as more than 50 per cent of the population belonging to the same denomination.18 By that measure in 1906, twenty-two of forty-eight states of the USA were homogenous. The figures for England were almost the same.

Furthermore, there were two obstacles to Americans benefiting from pluralism which were absent from Britain. When Stark and Finke count churches so as to calculate how diverse a place is, they count a black Pentecostal church and a white Pentecostal church as two. But that is misleading, because even now blacks and whites do not share churches. The same point applies to language or ethnicity. A German Lutheran church is not much of an option for a Swedish American. In contrast, with the exception of a few Gaelic churches in the Anglophone Scottish Lowlands and a few Welsh chapels in England, all British churches used English (or Welsh in Wales and Gaelic in the Scottish Highlands and islands), and so no one was excluded by language from any particular outlet. Since the 1980s London and the south-east of England have acquired a number of Pentecostal churches attended largely by West Africans and very few white folks, but this is new. For most of the nineteenth and twentieth centuries all the British churches have been equally attractive to people of all ethnic backgrounds. Arguably the absence of racial, language, and ethnic divisions meant that many UK towns and cities (p.240) offered their people a greater range of actual choices than did their US counterparts.

Finally we come to the main problem: comparison across time and space. One can find evidence for the virtues of competitive free markets in particular times and places: for example, it was often only the threat of losing members to dissenting sects that forced the British national churches to improve local provision. But the big picture fails to support the supply-side case. Every society in the modern world is more religiously diverse now than ever it was, and offers more freedom to choose between faiths than ever it has, but every society (including the USA) is now less religious than it was when it was less diverse. So the required association between diversity and religiosity in any one society over time is absent. It is also absent in the comparison of place. The most religious societies in the modern world are those that are most homogenous: Italy, Ireland, and Poland, for example, are considerably more religious than Britain, Holland, Australia, and New Zealand.

In summary, while the rational choice or supply-side alternative has specific application (something which Wilson anticipates), it fails to provide a persuasive comprehensive alternative to the secularization thesis.

One final word can be said in defence of Wilson’s work: as we will see in Appendix Two, religion in the Western world has continued to decline in power, popularity, and prestige since 1966. That is, even if Wilson’s explanation is wrong, he is, unlike many of his critics, trying to explain the right thing.


(1) D. Martin, The Education of David Martin: The Making of an Unlikely Sociologist (London: SPCK, 2013).

(2) D. Martin, The Religious and the Secular: Studies in Secularization (London: Routledge and Kegan Paul, 1969).

(3) P. L. Berger, The Sacred Canopy: Elements of a Sociological Theory of Religion (New York: Anchor, 1967). As I have argued, the term ‘pervasive world-view’ might give a better sense of how ideas subtly influence behaviour and thought than ‘sacred canopy’ (S. Bruce, ‘The Pervasive World-view: Religion in Pre-Modern Britain’, British Journal of Sociology, 48 (1997), 667–80).

(4) K. Thomas, Religion and the Decline of Magic (London: Weidenfeld & Nicolson, 1971).

(5) S. Bruce, God is Dead: Secularization in the West (Oxford: Blackwell, 2002), ch. 2.

(6) Martin, The Education of David Martin, 128–9.

(7) C. G. Brown, The Death of Christian Britain: Understanding Secularisation 1800–2000 (London: Routledge, 2001); H. McLeod, The Religious Crisis of the 1960s (Oxford: Oxford University Press, 2007).

(8) E. S. Gaustad, Proclaim Liberty throughout All the Land: A History of Church and State in America (New York: Oxford University Press, 2003), 35.

(9) D. Martin, A General Theory of Secularization (Oxford: Basil Blackwell, 1978), and On Secularization: Towards a Revised General Theory (Aldershot: Ashgate, 2005).

(10) Or at least I hope it is, because I regard myself as equally indebted to both.

(11) D. Nash, ‘Reconnecting Religion with Social and Cultural History: Secularization’s Failure as a Master Narrative’, Cultural and Social History, 1 (2004), 317.

(12) Martin, ‘Sociologist Fallen among Secular Theologians’, in The Religious and the Secular, ch. 6.

(13) R. Stark and R. Finke, Acts of Faith: Explaining the Human Side of Religion (Berkeley and Los Angeles: University of California Press, 2000); L. A. Young (ed.), Rational Choice Theory and Religion (London: Routledge, 1997).

(14) There are various possible justifications for this assumption. Some commentators believe that the human condition (in particular our ability to separate imaginatively the body and the person) naturally predisposes us to be religious. That proposition has difficulty dealing with the fact that in societies such as Denmark and Britain at least two or three generations have displayed very little religious sentiment without resultant social or psychological defects. Stark’s basis is very different. It argues that the common human condition of wishing for more than one has creates a demand for ‘compensators’ to make up the deficit, and religion trumps secular compensators such as those offered by political ideologies in that it can make promises about an afterlife. See R. Stark and W. S. Bainbridge, A Theory of Religion (New York: Peter Lang, 1987).

(15) RSS, p. 110.

(16) S. Bruce, Choice and Religion: A Critique of Rational Choice Theory (Oxford: Oxford University Press, 1999), ch. 3.

(17) R. Stark, ‘German and German–American Religiousness’, Journal for the Scientific Study of Religion, 36 (1997), 182–93.

(18) E. S. Gaustad, Historical Atlas of Religion in America (New York: Harper and Row, 1962).