Is the secularization research programme progressing? Debate on Jörg Stolz’s article on Secularization theories in the 21st century: ideas,evidence,and problems

Abstract

The methodology of scientific research programmes, developed by Imre Lakatos, can help us to identify which theories are strong or weak. Applying this approach suggests that the secularization research programme is progressing, as Stolz argues. Some of the recent advances have been more successful than others, however. In particular, we have done better at understanding how secularization happens than why it happens.

Introduction

Jörg Stolz has climbed the Mount Sinai of secularization and returned with a proclamation in seven parts. Despite falling short of the full Decalogue, it is a wonder of discernment that should guide our work. I personally feel a sense of vocational renewal, and it is an honour to append my thoughts to his address.

Let me start with some personal history. As an undergraduate many years ago, I was fortunate to study philosophy, logic and scientific method at the London School of Economics. The department had been the home of Karl Popper, by then retired, and also Imre Lakatos, who died suddenly (aged 51) just before I arrived. It was an exceedingly small group; there were about half a dozen students in my year. We focused primarily on the history and philosophy of science, with the methodology of scientific research programmes being the main point of reference.

This analysis of scientific progress, devised by Lakatos, cleverly reconciled Thomas Kuhn’s sociology of paradigm shifts with Popper’s logic of scientific discovery. It explained why scientists continue to work on theories that struggle to account for all of the evidence, but also why it is often rational to do so. Importantly, it points to ways of identifying which research programmes (or paradigms, if you prefer Kuhn’s terminology) are progressing and which are failing.

Secularization can serve as a case study. During what Stolz labels the ‘contentious phase’ of work on religious change, many scholars claimed that the secularization thesis was dead. Some scholars – including a number who should know better – still do. (A recent presidential address at another learned society featured the assertion that ‘the presuppositions that informed secularization theory has been effectively refuted’; Edwards, 2019:10). Stolz argues not merely that secularization is viable as a theory, but that ‘we have collectively made important progress’ on it.

In what follows, I outline Lakatos’ approach and discuss how secularization might be treated in this framework. I consider the various contributions Stolz highlights to see how far they do in fact support the view that the overall programme is progressing.

The methodology of scientific research programmes

Lakatos (1970) provides the canonical presentation of the methodology of scientific research programmes (henceforth MSRP), though a good short introduction is available online (Musgrave, 2016, sections 2.2 and 3.4). As an aside on the vagaries of academic publishing, we now regard high-impact peer-reviewed journals as the quality benchmark, with chapters in edited collections having considerably lower status and conference proceedings almost none at all. Lakatos (1970) appears in the fourth (!) volume of the Proceedings of the International Colloquium in the Philosophy of Science, which these days would be rejected by even the most desperate vanity publishing house. In fact, it ‘became an international best-seller’ (Musgrave, 2016), at least in academic terms, and has stayed in print to this day with Cambridge University Press. Part of the explanation no doubt lies in the jaw-dropping quality of the colloquium (held in London in 1965), which brought together Tarski, Quine, Carnap, Kuhn, and Popper, along with Lakatos, not to mention Geach, Bar-Hillel, Dummett, Suppes, Bernays and Watkins, among others: a fantasy football team of mid-twentieth century giants in logic and the philosophy of science. The published volume also includes a chapter by Paul Feyerabend, though as far I can tell he was not at the original event. But I digress . . .

MSRP is underpinned by several ideas. The first is that science is driven by research programmes, not single theories. Such programmes consist of one or more fundamental principles, referred to as the ‘hard core’, that are then developed in multiple directions via testable theories and empirical research. The hard core must be retained for the programme to survive, but the hypotheses that surround it may be discarded if they are persistently unsuccessful at dealing with the evidence. In addition, a programme involves heuristic principles that guide the expansion of theory into new areas and also ways of coping with findings that threaten the core components.

So far, so descriptive. What Lakatos further offers are some normative indicators to identify research programmes that are doing well or badly. The simple Popperian view that a single counter-example is sufficient to falsify a theory is unrealistic: scientists wrestle with the evidence in various ways. That is not to say, though, that there are no standards, or that we can only judge in retrospect, once a scientific revolution has swept away the old paradigm and replaced it with a new one. Signs of progress include the theoretical conquest of new domains, ideally involving unexpected predictions that turn out to be correct; signs of degeneration include the multiplication of ad hoc hypotheses that limit the theoretical scope of the programme and serve only to protect its hard core.

Secularization as a scientific research programme

It has never felt natural referring to ‘secularization theory’, as if there was one accepted version that could easily be defined; we generally refer to it as a thesis or paradigm. It is a good example of a scientific research programme in this respect. The hard core is very simple: ‘The basic proposition is that modernization creates problems for religion’ (Bruce, 2002:2). The key questions that follow are ‘why, and how?’, and efforts to find answers constitute the research programme.

The basic thesis has been controversial, of course: opponents dispute the connection between modernization and religious decline. They point to the variability of the apparent impact of factors such as industrialization, urbanization and so on. They argue that both modernity and religion are as American as apple pie. And some suggest that the persistence of belief in the supernatural and the prevalence of alternative spirituality show that modernization transforms religion rather than undermining it. Scholars working on secularization have been obliged to respond to these objections.

An important strand of work on the research programme could be described as defensive, in the sense that it develops the theoretical superstructure, sets out auxiliary hypotheses, marshals evidence, addresses objections and criticizes alternative theories. Secularization: In defence of an unfashionable theory (Bruce, 2011) is the leading statement of this type, and Steve Bruce follows in the footsteps of Bryan Wilson, Karel Dobbelaere (to whom Stolz dedicates his paper), and others. A more empirically orientated strategy is to use large survey datasets to test rival accounts of religious change; Pollack and Rosta (2017) is arguably the most impressive example of this approach. (From a Kuhnian perspective this work might be derided as ‘normal science’, but scholarly seriousness is its primary characteristic, and ultimately it is the weight of evidence that matters.)

What Stolz provides is an overview of recent developments in the research programme; his synthesis is an important contribution. He argues that ‘progress has been made’, which seems fair. Certainly the contrast with competing research programmes is stark(!). The market model appears to have had its day; it led to some interesting work, but it always seemed designed for the United States, and few people still believe that the United States is immune to religious decline.

As for the claim that what we see is religious transformation, not decline, that research programme was born degenerate. While attracting commentators from across the Western world, it seems to boil down to the negative heuristic that any instance of religious decline needs to be explained away. People aren’t really defecting from religion, they are just ‘nothing in particular’. They may not go to church, but they still believe. And if they say that they don’t believe in a Christian God, they believe in something else. If they don’t believe, they hold something sacred. And so on, and on: when evidence knocks down one ad hoc hypothesis, another pops up.

Let us turn our critical attention to Stolz’s positive assessment of the secularization research programme. Applying the principles set out by Lakatos, does it still seem justified?

How does secularization happen?

If the central idea is that modernization creates problems for religion, and the key questions are how and why, my impression is that we have done better on the ‘how’ than the ‘why’. I feel uneasy making this assertion, because my own research has tended to concern ‘how’. There is no special virtue in this focus: the ‘how’ question merely seems more tractable.

In deploying the term ‘secular transition’, I argued that

The theory of secularization rests on a simple idea: social change tends to follow particular routes. Certain major transformations – such as the industrial revolution, the decline in mortality, or equalization in the status of women – occur exactly once in each society. . .We can use knowledge gained about one transition to illuminate the course and causes of another. (Voas, 2008: 25)

As mechanisms of change, I emphasized the role of convention (i.e. social norms), behavioural drift and the diffusion of innovations. I subsequently offered the model described by Stolz. And with others, I have argued that religious change mainly occurs between rather than within generations, implying that the pace of secularization is largely determined by the degree to which young people are socialized into religious identity, belief and practice.

Nearly all of this work is much more about how secularization happens than why it happens (except in a proximate sense). To that extent it can never be sufficient or fully satisfactory on its own. These efforts have, however, added empirical content to the research programme: the hypotheses lead to new predictions, some of which are unexpected, and at least some of which have been corroborated. Notwithstanding the continuing existence of anomalies (notably some unearthed by Stolz himself, e.g. the case of East Germany), the secularization research programme has been progressive in the way that Lakatos required, at least in this domain.

Why does secularization happen?

The other areas reviewed by Stolz – secular competition, existential security, education, pluralism, regulation – are mostly concerned with why modernization creates problems for religion. The idea of secular competition is the broadest and arguably most plausible of these conjectures. The principle is clear and compelling: modernization brings enormous growth and diversification in knowledge, technology, work, forms of authority, social identities, leisure opportunities and so on. Religion is crowded out as an unintended consequence: whatever it provided before, there is now an attractive secular alternative. Where religion weakens, the heuristic directs us to look for the secular competitor(s). Where religion persists, we are directed to see where it continues to hold a competitive edge.

A critic might object that the idea offers little beyond Bruce’s core proposition except post hoc stories tailored to circumstances. While that complaint seems unfair, it is true that the generality of the concept makes it nearly untestable. And if religion suffers in the face of secular competition, it becomes hard to explain why rapid modernization does not lead to immediate secularization (on the Arabian peninsula, for example).

Similarly, the idea that existential security is the crucial factor seems persuasive on first acquaintance (because of the plausibility of Norris and Inglehart’s account, and the strong association between the Human Development Index and secularity) but less so on further reflection. The associations at national level merely restate what the core proposition tells us: modernization is bad for religion. The relative weakness of the relationship at individual level undermines the causal story. Rescuing the theory is difficult without resorting to the kind of ad hoc hypotheses that signal degeneration in a research programme.

Education is another explanatory variable for which the effects are difficult to assess because of confounding with other aspects of modernization. There is also the problem that it may be associated with scepticism but – among believers – more frequent practice. Recent research in this area does not represent much of a theoretical advance. Some of the empirical work is impressive, though, and could be claimed as vindicating the predictive power of the programme.

The hypothesis that diversity (pluralism) creates problems for religion has also been debated for decades. Again, it is not obvious that there has been much theoretical progress, and as Stolz notes, there was a long empirical hiatus following the discovery of a serious methodological problem. After many years of work, however, Daniel Olson has solved that problem and produced results that should be decisive (Olson et al., in press). Although Lakatos would have baulked at the idea of a crucial experiment that supports one theory and falsifies another, the question of whether religious pluralism has a positive or negative effect on religious participation is critical to evaluating the supply-side market model on the one hand and the secularization thesis on the other. In answering that question for the United States, Olson has corroborated a much-contested prediction of secularization theory and struck what may be a fatal blow at his mentor’s ‘new paradigm’ (Warner, 1993).

And on the subject of the supply-side, there is finally the matter of regulation. People can be constrained by theocratic or atheistic authorities to behave in ways that are more or less religious than they would otherwise choose, and less draconian public policy may also influence the religious market. Stolz has contributed to theoretical progress by bridging the divide between rational choice as a social theory and work on secularization (Stolz, 2009). In practice, the impact of official regulation seems fairly marginal except in extreme cases (e.g. East Germany, Iran).

On balance, Stolz’s conclusion that ‘we have made significant progress in secularization research in the past 20 years’ seems right, though judged by the rigorous standards of MSRP, the position is less clear-cut. Ultimately we want to know why secularization occurs when and where it does, and we still struggle to specify which aspects of modernization are most important. The work goes on. Science offers the thrill of the chase, of discovering the unknown, and Stolz deserves our thanks for pointing the way.