I have a complicated brief, because this issue is bedevilled by the way people talk about ‘religion’ and ‘the secular’ (and latterly ‘secularism’) as if they knew some essential meaning attached to those terms. The same is true for ‘science’. It is also bedevilled by the way some educated people talk about what they believe ought to be the case rather than what actually is the case. That is because certain historic crises between science and religion have acquired an almost mythic status, in spite of recent revisionist history, for example the writings of John Brooke at Oxford. I am thinking of the Galileo controversy, of what Paul Hazard called the crisis in the European mind from the late seven-teenth to the early eighteenth century, and of the controversy over Darwin.

I might be expected to revisit the philosophical issues involved, for example about the status and freedom of humanity, whereas I shall be for most of the time simply a sociologist. Being simply a sociologist is not as simple as it seems because I have to begin by explaining what we do and do not take seriously about theories of secularisation. That involves a rapid dive into the inwardness of sociology with respect to concepts and procedures. It has to include a ragbag of different kinds of explanation from the advent of the car to the political alienation of the working class to the long-term processes of rationalisation. There is also the distinction between science as a mode of understanding and the indirect consequences of science and technology for our everyday lives. My brief is to discuss the former yet sociological theories of secularisation are more concerned with the latter.

The classic procedure of sociology in the absence of controlled experiment is cross-cultural comparison, so the core of what I argue has to turn on something so obvious, so crucial, and to my mind so compelling, that it is a sociological question why educated people do not take it on board. In terms of cross-cultural comparison, countries at roughly the same level with regard to scientific advance have religious profiles pretty well across the complete range. That means such general tendencies to secularisation as may exist (other things being equal) have to be set against the contingencies of history and the particularities of culture where things are hardly ever equal. I also have to speculate about mental space. Do we, or some of us, entertain mental spaces where there is an either–or, as between science and religion, or is mental space a manifold where we call upon different kinds of discourse and understanding according first, to our individual context and then to the degree of openness or closure of mental space made available to us by our culture.

I have to touch on the kind of religious mutations brought about by modernity, apart from straight secularisation, such as the recent re-emergence of magic, spirituality and what Paul Heelas and Linda Woodhead call resacralisation. Finally, I have to open up a theme implicit throughout, which is the power of master narratives, especially the Enlightenment master narrative, to cloud over the sociologically obvious in favour of an ‘obvious’ but false hypothesis about religion receding as science advances. The French may think master narratives have had their day but in my view that only tells us they have lost the plot. Enlightened intellectual histories of the relationship of science and religion, especially their dubious assumptions about the avantgarde, obscure the actual data about secularisation and ensure a principled ignorance about our global reality. That is why it keeps surprising us. As Sir Bob Geldof, himself an atheist, has reminded us, we do not understand what is happening globally unless we understand religion. Waiting for the world to play catch-up with France and Sweden is a mug's game.

So much for my awkward brief as regards the recession of religion with the advance of science. As I have hinted, my colleagues in the field of the sociology of religion, for example Grace Davie, Steve Bruce and James Beckford in the United Kingdom, might well regard rebuttal of so crass an idea as precious time and energy misspent. Putting it mildly, this is what the late Imre Lakatos, philosopher of science, would have called a ‘radically deteriorating research programme’. Once upon a time, somewhat before Darwin, many social scientists had their own evolutionary schemes where such ideas about science and religion were seriously entertained: no longer. One of its last representatives, to my knowledge, was the late Baroness Barbara Wooton in the 1950s. It was her confident affirmation of an either–or, with respect to social science and religion, let alone natural science, that forced me to think furiously about why what was obviously true to her was obviously false.

What then do sociologists of religion take seriously? They take seriously the privatisation of religion, although not everybody is convinced about it, for example, Jose Casanova in his impressive rebuttal, Public Religions in the Modern World (1994). What the privatisation of religion refers to is its declining social significance for public rhetoric, legitimation, debate and policy. The former Downing Street adviser, Alistair Campbell, warned the then Prime Minister, Tony Blair, about to address soldiers destined for Iraq: ‘We don't do God’. Sociologists also take individualisation seriously, as linked to privatisation. It is argued we put together our own personal bricolage of religious attitudes and ideas and build our own individual spiritual pathway. So much for the church. An example of this might be a comment by one of the Beatles: ‘Why should ten thousand people believe the same thing?’

At this point I have to tip the ragbag of theories on to the floor. There is the effect of geographical and social mobility in breaking up dense communal relations, permeated by religion, and in breaking up the unity of the generations. There is consumer hedonism hostile to long-term commitments of any kind and unwilling to invest time and energy in maintaining institutions, religious or otherwise. There are the rival attractions of a picnic in the car, or Sunday morning football for the under-13s, or weekend entertainment in front of the television. There is the impact of secular control of education and the media, so that even personnel employed by the church or in church schools make pre-emptive strikes in favour of secular criteria and values alien to religion. Then, in the longer historical perspective, there is the displacement of religious solidarity by national solidarity or the claims of party political ideology. It is not easy to turn these assorted ideas and influences into joined-up sociological narrative, but they clearly have little to do with the impact of science, apart from the indirect consequences of scientific invention. Science as a technical procedure available in this context or that is not the same as science as a mode of understanding. The philosophy of science, Feyerabend, Popper, Fodor, Hesse, etc., hardly enjoys mass sales. Sociologists take seriously their empirical indices and such basic organising frameworks as rationalisation, social differentiation and (already mentioned) privatisation. As for empirical indices, religious belief and practice are routinely correlated with age, generation, gender, status, occupation, education, urbanisation and so on to produce patterns of change. However, the crude data are not entirely selfinterpreting, so that you have to put questions to them, such as how far a decline in female practice is due to changing female lifestyles. How is it that, according to Andrew Greeley, any association between advanced education and irreligion has virtually disappeared in the youngest European generation. If we find, with Rodney Stark and Robert Wuthnow, that natural scientists and technologists are more religious than practitioners of the humanities and social sciences, that probably nudges us away from any idea that hard science is incompatible with religion. Stark's data show that the least scientific disciplines are the most irreligious, coming to a climax in my subject, and that irreligion does not increase with increasing exposure to hard science. (One oddity worth scouting out is why individual scientists do not believe the data about what they believe, any more than they accept the argument about the very minor role played by complex scientific findings in religious decline compared (say) to the invention of the Walkman. Attack a paradigm and the first reaction is to go into denial.)

Now for our two crucial organising frameworks: rationalisation and social differentiation. You will notice we have entered an intellectual world dominated by schemata and nouns of process ending in ‘-ation’. Nouns of process have a way of implying movement from state A (religious) to state B (secular). So do schemata like the shift from Community to Association. That way we find ourselves tilted on a historical slope where any particular religious present is pregnant with a secular future. Michael Wheeler, in his study of the religion of John Ruskin, complains of precisely this way of thinking and shows how unjust it is when trying to understand Ruskin. Another version of this approach is Herbert Butterfield's ‘Whig Interpretation’ of history which is one of the progressive master narratives governing our perception of the direction of change. I come to these master narratives later, only pausing to notice there are other master narratives of rotation, such as Sorokin's ideational and sensate periods, or narratives of decline such as Oswald Spengler's The Decline of the West.

Rationalisation is a concept we owe to Max Weber, and in his work it is more part of a narrative of decline towards the ‘iron cage’ than of progress. It does not refer to the advance of abstract reason but rather to the consequences of increasing technical efficiency in rendering our world impersonal and disenchanted. Here we encounter some of the indirect consequences of science, for example when we consider the increasingly rational and bureaucratic way in which we today organise health. Once there were hospitals named after St Bartholomew or St Thomas, but that religious frame has now been hollowed out, and the sense of a vocation to heal is onfined to the private motivations of individual carers and doctors. Clearly health is a sphere where the white coat of medical science is most pure and respected. On the other hand the grey suits of health bureaucrats and administrators are less pure and less respected.

The advance of rational bureaucracy was a theme developed by the late Bryan Wilson, and for him it belonged very clearly to a narrative of bleak decline. As a major theorist of secularisation he envisaged a regression from the personal, communal, religious and conscientious to the impersonal, fragmented, irreligious and demoralised. Again this has little directly to do with the advance of science, at least of natural science. But I do find it interesting that an agnostic should so bravely enter this contentious area to risk the coals of fire prepared for anyone linking the decline of religion to the decline of morality, personal responsibility and conscientiousness. The counter-arguments are obvious and have recently been made by Paul Gregory. Is the contemporary United States so obviously virtuous? Was the England portrayed in Hogarth's The Rake's Progress or in Jacobean drama so obviously responsible and clean living? Here enters the criminologist and sociologist of morality, Christie Davies, and his impressively documented The Strange Death of Moral Britain (2004). He plots the indices of the decline of religion against moral and criminal indices, fully alive to all the caveats and personally conscious of both gains and losses. His focus is the same as Bryan Wilson's, that is, the rise in nineteenth-century Britain, alongside modernity, of a personally responsible and conscientious kind of religion, propagated by the Sunday School, and by voluntary charitable organisations and voluntary youth organisations. This was a religious mutation rather than secularisation. Different commentators date the moment this rise went into decline differently, but there is considerable agreement about a parallel rise and fall occurring all across Europe. You will readily grasp that any correlation between fewer enrolments in Sunday Schools and greater knowledge of the Special Theory of Relativity is adventitious.

Christie Davies also relates these parallel changes to a shift from moral to ‘causalist’ understandings of human behaviour that owes something to social science and maybe genetics, as well as to a utilitarianism shifting from principles of moral justice to principles of minimum harm. Again these are gains and losses but I do not think anyone can be unaware of what has happened. Even criminals themselves have become expert manipulators of ‘We are sick, we are sick, we are sick, sick, sick; we are sociologically sick’. The social scientific translation of a Christian vocabulary relating to forgiveness and victimage is also pretty clear. Philip Rieff has written eloquently of The Triumph of the Therapeutic and I recently heard the family of Herod the Great cosily characterised from the pulpit as ‘dysfunctional’. So much for mired in sin.

Yet sociology should maintain its usual reservations about the seemingly obvious here, where sociology rather than natural science corrodes religious language. When I hear young people referring to the apparently demoralised worlds in which they so often live, I hear expressions and judgements about ‘bastard’ and ‘inappropriate’ behaviour saturated in morality. Even my social work relatives and friends resort to moral judgements outside the immediate environment of what for them is clearly a moral vocation. As for the close neighbours of ‘dysfunctional’ families, their comments are downright theological. These are families from ‘hell’, and going to ‘hell’. Therapeutic language may have triumphed but it persistently gravitates back to moral evaluation. The adjective ‘pathological’ is a judgement posing as a description.

Of course, morality can be separated from religion both with respect to individual behaviour and in the context of moral philosophy. That is not at issue, nor am I suggesting that any inherent limit on the erasure of moral language has to be linked to an inherent limit on the erasure of religious language. I am saying, however, that religious language is bound at the hip to moral language and that moral language of itself throws up analogues of religious understandings, such as the gratuitous, the redemptive, the salutary, and the justified, as well as locutions implying vertical as well as horizontal dimensions. Personally, I doubt whether either social or genetic determinism can erase the links between religion and morality implied by linguistic usages.

My discussion of rationalisation has, following Bryan Wilson, included the role of technically efficient and impersonal means, in relation to morality. I now turn to ‘social differentiation’, which is my own primary framework for analysing secularisation, though you will notice how it overlaps the framework of rationalisation. In my view social differentiation, while bringing about certain convergences on the secular, also allows room for my stress on historical particularity and cultural contingency.

The concept of social differentiation has roots in the increasing division of labour and refers to the increasing autonomy of semi-discrete sectors of social life. One by one these sectors, for example, law, social control, social legitimation, communication, health and education, acquire their own proper autonomy from ecclesiastical regulation and influence and from religious modes of thinking. Theology is no longer the automatic queen of sciences. Just as we are all aware of the impact of rationalisation so social differentiation is part of everyday experience regarding the autonomy, even conflict, of teacher with priest and of secular therapy with the priestly cure of souls when it comes to the dilemmas sometimes faced by health professionals in Catholic hospitals. If I wanted a small example with large implications of autonomy I might choose the partial edging out of the theologian on public bodies concerned with bio-ethics by philosophers or theologians talking philosophically. That change is part of a liberal claim to exercise universal objective jurisdiction as against all particular modes of thought and interest. In my view Jürgen Habermas rightly suspects this claim, specifically in relation to religion. But it is not clear what the alternative might be, especially given that liberalism is so often a Judaeo-Christian mutation, as both John Gray and Charles Taylor have argued.

Since I believe social differentiation allows for historical and cultural contingency and variability I need to offer an illustration, and it is this. The order in which different cultural sectors are differentiated varies historically, and the partial and then absolute separation of church and state in the United States had momentous positive consequences for religion from 1750 to the present day. By contrast, the late separation of church and state and/or church and elite in Western Europe had momentous and continuing negative consequences for religion. I also happen to believe rationalisation is equally inflected by history and culture, having lived in Dallas, Texas, which is simultaneously one of the most rationalised and religious metroplexes in the world. At the same time rationalisation is patient of a much more universalising treatment at the expense of historical particularity. There are those who, perhaps rather ethnocentrically, see post-Protestant Sweden and post-Catholic France as templates of the universal future. It might be nice but I suspect ‘ought’ and desire have once again run ahead of ‘is’ and historical plausibility. Iraq like Sweden? There is a way to go yet.

So far, we have looked at some of the variables sociologists take seriously, at empirical indices and interpretative frameworks like privatisation, rationalisation and social differentiation. We have insinuated worries about historically tilted slopes, Whig interpretations of history, and the extent to which we are working within master narratives, particularly Enlightenment ones and even, maybe, Kuhnian paradigms.

The way should be clear for my presentation of cross-cultural and cross-national data showing no consistent relation between the degree of scientific advance and a reduced profile of religious influence, belief and practice. However, in a way all too typical of my sociological tribe, I need first to step back for an essential point about sociological interpretation, at least as I position myself in the tradition of ‘verstehen’ or understanding, and ‘Geisteswissenschaften’ or cultural sciences as expounded by Max Weber and Wilhelm Dilthey. I emphasise understanding of signs as well as the explanation of indices. For our purposes I do that in the context of the depth and ambiguity of signs, in particular language, and the expansive manifold of mental space. There are emergent properties of action in the human universe not present in the physical or biological universes. I find them illustrated in the structure of motives in drama, including the constraints of options and alternatives for action in plotting, and the freedom of interpretation based on the way mutual anticipations are built into cues and scripts. This can be called the dramaturgical approach; and the point is to locate the right kind of scientific intentionality for any given problem, not to transfer what works in one sphere of knowledge to another where it is inappropriate.

Clearly the depth and ambiguity of symbols, signs, especially language, and the manifold character of mental space, bear quite directly on any notion of a zero-sum relation between science and religion. In mental space there is no such necessary either/or, because many different modes of ideation and discourse can be entertained simultaneously and selectively drawn upon according to context. Beyond that, the degree of openness and the degree of closure in mental space itself varies culturally, as between, for example, the constricted one-dimensional space of the former DDR where either/or prevails, and the rather conspicuously wide open spaces of the United States, where angels, UFOs and aliens happily cohabit with rocket science.

Allow me to press further this matter of the nature of mental space and to illustrate our recourse to different kinds of concept and discourse according to context. There is, for example, no difficulty about saying ‘Your life in their hands’, emphasising the autonomy of medical science and the doctor, and ‘He's got the whole world in his hands’, emphasising our dependence on profound spiritual resources. Consider also the difference between ‘Great is truth and shall prevail’ and ‘You shall know the truth and the truth shall set you free’ or ‘I am the way, the truth and the life’? Here we enter into the phenomenology of being and freedom. I am not starting a discussion of being and freedom or of the autonomous nature of language games here, but only stressing how sensitive we have to be about kinds of discourse to do justice to social complexity. There is more to sociology than causation, correlation, jargon, inverted commas and hobnail boots.

Now for my cross-cultural data: I am going to stay in the main with three revolutionary, scientifically advanced and rationalised societies, each with a widely different religious profile. I shall make one or two further observations, in particular about two other scientifically advanced and rationalised societies, Uruguay and Singapore, one where Christmas does not happen, except privately, and the other where everything happens in public and in private.

From a socio-historical perspective, the United States, founded on the basis of an Anglo-Protestant pluralism, experienced a relatively friendly relation between religion and Enlightenment, for example, the overlap of Episcopalianism and Masonry. Subsequent migration brought yet more Lutherans, Catholics, Jews and others, for whom religion was a residual anchor of distinct and sometimes regional identity in the American melting pot. Popular religion and elite Enlightenment mostly cooperated to support an inclusive civic nationalism. This was progressively based on common citizenship rather than an organic exclusive nationalism of ethnicity and language. That process came to an interim conclusion in the 1960s with the Civil Rights Movement. But the culture wars since the 1960s have not severed the ties of a common culture, for black and for white, for Hispanic and Anglo alike. By European standards American culture wars, however hot in 2007, are mere skirmishes.

In France and in Quebec matters have been quite otherwise. In Quebec an organic, ethno-religious and linguistic nationalism emerged, persisting up to the 1960s when religion and linguistic nationalism to some extent went their separate ways. Religious practice dropped as dramatically as it did in Holland over the same period. In France itself a civic and linguistic nationalism established by revolutionary violence warred against a religiously informed organic nationalism. The climax came after 1870 with the Third Republic. Religious practice began to fall, church and state were separated in 1905, and laicité has been dominant ever since. I need only mention one more difference, which is between the decentralised governance of the United States and the centralisation, particularly of education, in France. Decentralisation inhibits attempts at ideological monopoly by secular elites in the United States, where they are relatively unimportant. Centralisation assists monopoly by secularist elites in France, where they are important. Where else did the term ‘bien pensant’ come from?

As for East Germany (the former DDR) centralisation reached its apogee under the enlightened and secularist dictatorship from 1945 to 1989, following the pagan interlude of blood and soil, 1933–45. Modernisation was dramatically held back and active religion reduced to a small minority, in spite of the fact that a Stasi-infiltrated church was everywhere the springboard for bringing down the dictatorship in 1989.

What now of irreligious Uruguay and riotously religious Singapore? Uruguay is a fairly simple case, given that the reds more or less beat the whites, and the maximum immigration from Europe coincided with maximum church–state tension in the sending countries, Spain and Italy. A secularist university was created producing an elite for a ‘Latin’ country that does not even acknowledge Christmas. Uruguay is also 80 per cent urban.

Singapore is clearly one of the most urban, multicultural and rationalised countries in the world. Religion in Singapore has been state-controlled in a rather neo-Confucian manner to underwrite the values of national cohesion and development. The mutations in religion coming with modernisation include some separating out of the rich syncretic and animistic mulch from classical Buddhism, an expansion among the socially mobile of a personal, close-knit, almost cellular form of Christianity, and straight secularity.

I stress again how this separating out and this rise of a consciously personal religion is characteristic of many modernising societies, as it was characteristic of modernising Europe. One finds similar changes in Brazil, Turkey and Indonesia. In an 80 per cent Muslim country like Indonesia, highly educated women adopt the hitherto rare practice of wearing the headscarf as a sign of modernity and a protest against syncretism. Unfortunately the concentration in our western media on fundamentalism as a reaction against modernity neglects this alliance of Islam with modernity, as well as a similar alliance throughout the developing world between Christianity, hygiene, education, technology and transnational connections. Only in our traditional ex-Christendom or among nationalist intelligentsias elsewhere, is that connection not understood or else condemned as disturbing traditional lifestyles. The assumption is that one can opt out of change and the global reality, keeping traditional culture intact for the tourist and the anthropological gaze. As one aspiring and teetotal Pentecostal in Ecuador said to the anthropologist: ‘Did you expect me to wear feathers, Miss?’

I also stress the crucial variable implicit in all the considerations so far put forward. That is the role of the church and/or religion in relation to degrees of cultural distance and disparities of political power and establishment. This brings us back to the standard contrast between Western Europe and the United States. I want to illustrate how that works out with respect to a couple of related contemporary developments. They are the return of magic and the rise of spirituality that Heelas and Woodhead speak of as resacralisation.

Regis Debray in his extraordinary biography of God over the last three thousand years claims that ‘The Twilight of the Gods is Morning for the Magicians’. That is a phenomenon present both in Western Europe and the United States. However, in Western Europe there is a negative relation between Christianity and magic, indeed a further separating out of the two by modernity, at least north of the Mediterranean littoral, whereas in the wide open spaces of the United States the gods and the magicians enjoy high noon together. The same applies to spirituality. In Western Europe spirituality, with art and music as further sources of consolation and inspiration, is often encountered outside the churches. Spirituality tends to offer an alternative to faith, as it does, for example, in Germany. However, in the United States, according to Robert Wuthnow, spirituality and faith mostly move together. That I put down quite simply to my key variable. Religion in the United States adapts to every cultural level in the context of a pervasive reliance on voluntary association, federalism and a religious pluralism with no specific institutional connection between a church and the state or between elite power and culture. The contrast is clear between the United States since 1750 and more than a century of established Christendom in retreat in Western Europe.

Eastern Europe, for complex reasons connected with Ottoman domination and the secularist post-Second World War dictatorships imposed by the Soviet Union, has been patchily different. The dictatorships successfully suppressed religion in the DDR, the Czech Republic and Estonia. However, in Greece, Romania and Serbia, and now in Russia, the revival of Eastern Orthodoxy is very evident, in spite of quasi-monopoly and a de facto or de jure establishment. The ethno-religious sentiments of repressed nations are very powerful and a long history of repression or defeat can for quite a while cancel out the effects of religious establishment on the achievement of independence. Islam's reaction to defeat as a religion programmed for victory is a large-scale instance of the same principle. In Russia it is not a matter of repressed nationhood but of a failed ideology giving place to an age-old union of faith and nation. The mutation separating out magic and religion with the onset of modernity is not exactly evident in Eastern Orthodox Europe. In Greece one finds a startling degree of iconic protection and spiritual insurance available, together with a belief in the devil only rivalled in the United States. Given Greek history it is not difficult to see why, but the advance of science is hardly evident.

I suggest the pictures I have offered of modern (and rapidly modernising) societies, from the United States to Russia, do not say much for any hypothesis about the decline of religiosity with the advance of science, at least so far as science is received by the people at large. What matters is the reception of science and technology with respect to religion, not some intellectual and sometimes mythic history of the relationship generated in the academy. The key lies in reception theory and it is notable that hardly anyone, even Owen Chadwick in his splendid The Secularisation of the European Mind in the Nineteenth Century (1975), has brought together the social or popular history of religion and the intellectual history of ideas. Charles Taylor is a rare exception. In the academy, or what Peter Berger calls global faculty club culture, too many of us, including theologians, rely on the intellectual history of ideas to understand religion and secularisation and on the debates in the tiny world of intellectuals who have this particular interest. Only this can explain why my late and admired colleague, Ernest Gellner, sometime Professor of Anthropology at Cambridge, should upbraid me for describing the United States as ‘religious’, given its intellectual representatives were people like Richard Rorty and John Rawls. On every conceivable criterion the USA is more religious than Western Europe, no matter what people living in Cambridge, Massachusetts, may think.

Intellectual history remains affected by a master narrative treating religion as inchoate and backward superstition or as pathetic froth obscuring the surface of the real until blown away by revolution. Intellectual history is also too closely wedded to what John Weightman analysed in his The Concept of the Avant Garde. It is not that the concept of the avant-garde never applies but that it's dangerous when taken for granted, as it is almost daily in the press. It seems to me that the avant-garde in the romantic and antinomian 1890s gradually extended its influence down the social scale, first between the two world wars, and then in the 1960s. On the other hand, Louis Menand's account, in The Metaphysical Club, of the secularisation inaugurated in the United States by the move to philosophical pragmatism, is only an irrelevant fifth wheel as concerns the religious history of Americans as a people. Theirs is a religious practicality not a philosophical pragmatism.

There is one particular and virulent version of the Enlightened master narrative that has had some modest popular success, though I think it even more implausible than the idea that religion recedes as science advances. This narrative couples the idea of scientific truth and scientific innocent virtue in dispersing religious error tied to culpable vice. The appeal of this coupling comes less from any popular dismay about inadequacies in the argument from design than from a popular unease about a seeming association of religion with violence and intolerance. The damage caused by religion is what people worry about and hear about, not esoteric if important issues about design flaws in Creation. How many of those who may vaguely have heard of the Selfish Gene and its author think about the vivid popular language in which the idea is couched other than as an invitation from science itself to let rip your genetic constitution? The link between religious error and evil, whether promoted by scientists in the public eye, or over pints in pubs, seems plausible, and could take more time to disperse through sociological analysis than is available here. That is because it depends on a naïve and simplistic pointing at facts in Northern Ireland, the Middle East or wherever, backed up by a studied neglect of the all too complicated question of the way religion does and does not mesh with power and the dynamics of power. There is an equally studied neglect of an obvious process whereby a society partly converted (say) to Christianity, in turn partly converts Christianity to its own social structures, whether they are feudal, mercantile or capitalist. People who are scientific in relation to their own subjects, become carelessly opinionated about issues in the public world or the human sciences. It happens that there is no continuing Enlightenment institution in secular contexts comparable to the church in religious contexts to take the moral flack hurled at the corruptions of power. Torquemada can be held up as a real Christian exemplifying faith deploying political power, whereas Joseph Stalin cannot be held up as exemplifying secular Enlightenment in power because he was not really enlightened. Again Christianity can be blamed for what happened when adopted as the faith of the Roman Empire, whereas Darwinism can innocently wash its hands of what happened when converted by capitalist society into Social Darwinism or deployed by Nietzsche. Yet the metaphors of Darwinism are decidedly more susceptible to malign conversion than the metaphors of Christianity.

Claims to historical innocence are always suspect. If this sort of analysis does have any resonance, promoted as it is by biologists and geneticists skiing off-piste in the human sciences, it raises the possibility that religion might recede, however mildly, not with the advance of scientific truth but with the promotion of scientific error. It is worth adding that those scientific theories which build religion into our mental basement, as a neurological programme acquired in remote prehistory to assist our survival, appear to restrict secularisation rather severely – on scientific or pseudo-scientific grounds. In Pascal Boyer's rendering of continuing emissions from the mental basement, only the hyper-intelligent can escape religion: an idea with obvious attractions for members of the faculty club. Religion remains immortal, at least among the Untermenschen. Lord Winston, the eminent gynaecologist, is now publishing a book about God, sympathetically attempting to bridge the gap between science and religion by showing how religion has been built into the psyche for our better adaptation and survival. However, he also notes, quite correctly, that if that were so we ought all to be religious. It follows that it was only half built in as a kind of side effect. Whether that explains why the neuro-transmitters of religion have been turned off in East Germany, or are blazing away in the United States, or are strongly on for over half the population of South Korea and simply off for the rest, remains mysterious.

Sociology is a humble affair and definitely not constructed after the model of rocket science or quantum mechanics. That is because it is a subject with a human subject matter. It does, however, operate within certain evidential constraints on matters where mere opinion is supposed to be king, and I hope that within those constraints I have offered considerations undermining the idea that religion recedes as science advances. At least, if I were an atheist anxious to disturb the faith of intelligent young friends, I would recommend a course in biblical criticism, or in psychobabble and sociobabble, or, best of all, a vigorous drench in romantic literary Weltschmertz. But not, definitely not, a bracing course in astrophysics. They might too easily suppose they were tracing ‘the Mind of the Maker’.