Researching Religion: Why We Need Social Science 0198786581, 9780198786580

Table of contents :
1 Prelude
2 The Value of Social Science
3 Defining Religion
4 Measuring Religion
5 Bias in Social Research
6 Ethics in Social Research
7 Conversion
8 Social Theory and Religion
9 Action Rational and Irrational
10 Does Danger Make People Religious?

Citation preview

Title Pages

Researching Religion: Why We Need Social Science Steve Bruce

Print publication date: 2018 Print ISBN-13: 9780198786580 Published to Oxford Scholarship Online: August 2018 DOI: 10.1093/oso/9780198786580.001.0001

Title Pages Steve Bruce

(p.i) Researching Religion (p.ii) (p.iii) Researching Religion

(p.iv) Great Clarendon Street, Oxford, OX2 6DP, United Kingdom Oxford University Press is a department of the University of Oxford. It furthers the University's objective of excellence in research, scholarship, and education by publishing worldwide. Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries © Steve Bruce 2018 The moral rights of the authors have been asserted First Edition published in 2018 Impression: 1 All rights reserved. No part of this publication may be reproduced, stored in Page 1 of 2

Title Pages a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by licence or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this work in any other form and you must impose this same condition on any acquirer Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America British Library Cataloguing in Publication Data Data available Library of Congress Control Number: 2018932145 ISBN 978–0–19–878658–0 Printed in Great Britain by Clays Ltd, St Ives plc Links to third party websites are provided by Oxford in good faith and for information only. Oxford disclaims any responsibility for the materials contained in any third party website referenced in this work.

Page 2 of 2

Preface

Researching Religion: Why We Need Social Science Steve Bruce

Print publication date: 2018 Print ISBN-13: 9780198786580 Published to Oxford Scholarship Online: August 2018 DOI: 10.1093/oso/9780198786580.001.0001

(p.v) Preface Steve Bruce

A few words about the background to this book may help the reader understand its tone and its purpose. It is an end-of-career work: a summation of everything about theory and method I wish I had said to my students. It is argumentative. In part that is a function of a personality that in previous centuries would have been described as dyspeptic or bilious. Until relieved of my gall bladder by the National Health Service I was often literally those things, but I have always been them metaphorically. Most of what I want to say about the use of social science theory and methods is a product of disagreeing with others. I try to make my case positively, but it is impossible to avoid entirely a critical tone when the best way of showing why we should study religion in this way is to demonstrate that doing it that way leads to error. It must be said that many of the studies I criticize are, in many respects, excellent. The cheap shot has sometimes proved irresistible, but more often I have engaged critically with the work of scholars I respect because there is little merit in shooting fish in a barrel; it is much more productive to work out how good work could have been better. Like writing a book on prose style, authoring a critical commentary on research begs colleagues to find holes in one’s own positions, so I should stress that, when I refer to my own research in what follows, it is not because it is flawless but because I am familiar with its flaws. A paper that I co-wrote thirty years ago with the late Roy Wallis on the use of what people say about their actions in explanation was subtitled ‘defending the common-sense heresy’. I am reminded of that subtitle when I pass by the University of Aberdeen’s portrait of Thomas Reid. One of the lesser lights of the Page 1 of 2

Preface Scottish Enlightenment, Reid was appointed to a chair at Aberdeen in 1752, 240 years before I enjoyed the same privilege. While here he wrote his 1764 classic An Inquiry into the Human Mind on the Principles of Common Sense. I cannot claim to have been much influenced by Reid, whose (p.vi) main concern was to reconcile secular philosophy with Presbyterian Christianity, but I was struck by the good sense of the following: If there are certain principles, as I think there are, which the constitution of our nature leads us to believe, and which we are under a necessity to take for granted in the common concerns of life, without being able to give a reason for them—these are what we call the principles of common sense; and what is manifestly contrary to them, is what we call absurd.1 I take from that two general principles. First, we should resist theoretical and methodological postures that require us to deny abilities we routinely display in our everyday lives. Yes, it is hard to disentangle people’s motives from the accounts they later give of their actions, but every day we do exactly that when we listen to a student’s hard-luck story about the dog eating her essay or a child explaining why she had to take the car without first asking permission. Second, we should avoid taking seriously scholarly postures that do not inform the advocate’s day-to-day life. The day I meet postmodernists whose relativism does not disappear the minute they start talking about salaries and workloads is the day I will take relativism seriously. There is no false modesty in saying that there is nothing new in what follows. Every principle and practice that I advocate or defend was once a standard part of the social science armoury. That any of it needs restating is a result either of a lack of preparation (in the case of those arts-trained scholars of religion who stray into matters best understood with the tools of social science) or of an unfortunate fondness for novelty (in the case of social scientists who dismiss their predecessors as fools or knaves). Insofar as this contribution to the study of religion has a programme, it is to recover the tendrils of common sense from the absurdities that threaten to choke the social scientific study of religion. Finally, this is not a research manual but a reflection of principles that should underlie such manuals: more of a ‘why’ than a ‘how to’ book. Notes:

(1) T. Cuneo and R. van Woudenberg (eds), The Cambridge Companion to Thomas Reid. (Cambridge: Cambridge University Press, 2004), 85.

Page 2 of 2

Acknowledgements

Researching Religion: Why We Need Social Science Steve Bruce

Print publication date: 2018 Print ISBN-13: 9780198786580 Published to Oxford Scholarship Online: August 2018 DOI: 10.1093/oso/9780198786580.001.0001

(p.vii) Acknowledgements Steve Bruce

Much of what follows has been initially explored in previous publications and I am grateful to Oxford University Press, which has published many of my books, and to various journals for the opportunity to air my views. Many topics have also been discussed in conference and seminar presentations, and I am grateful to the organizers of, and participants in, all such events for the stimulus they provide. My working life has been spent in just two institutions: The Queen’s University of Belfast and the University of Aberdeen. I am hugely indebted to both for allowing me to teach in fields that intrigued me and for allowing me time to pursue my research interests. The University of Virginia, Charlottesville, and the University of Edinburgh hosted me during periods of research leave. At various times, my research has been supported by the Economic and Social Research Council (ESRC), the Nuffield Foundation, the Leverhulme Trust, and the Carnegie Trust for the Scottish Universities. I am particularly grateful to the ESRC both for funding my research on loyalist paramilitaries in Northern Ireland and for disguising the nature of that research. My primary intellectual debts are owed to David Martin and Bryan Wilson. They are frequently treated as competing and they certainly differed over much of what is now known as the secularization thesis, but I learnt a great deal from both. I also benefited enormously at Stirling University from the graduate supervision of Roy Wallis, who appointed me to a lectureship at Queen’s University in 1978. His work on new religious movements was an inspiration, but of greater enduring impact was his sage advice on theory and methods. In

Page 1 of 2

 

Acknowledgements recent years I have benefited greatly from the statistical skills of two congenial collaborators: Tony Glendinning and David Voas. An underrated source of encouragement and inspiration is the casual chat one enjoys with colleagues in a good department. To borrow the well-known quotation from Thomas Hobbes, the life of an academic may not be nasty, brutish, and short but it is certainly solitary, and corridor conversations with colleagues are important sources of affirmation and stimulus. I am fortunate to have shared a (p.viii) corridor with two very able young scholars: Marta Trzebiatowska and Andrew McKinnon. Finally, I must acknowledge two groups of people who have made my work possible. The staff of a large number of libraries, archives, and local records offices have helped me find relevant historical material for my research. Less easy to identify but every bit as important is the very large number of people in the UK and the USA who have invited me to their services, workshops, and therapy sessions, chatted to me about their beliefs, replied at length to my questions, fed me artery-clogging amounts of home baking, and, I hope, helped me to a better understanding of religion in the Western world. Naming them would be impossible, but without such generosity there would be no social science research and no opportunity to reflect on its methods and theories.

Page 2 of 2

 

List of Figures and Tables

Researching Religion: Why We Need Social Science Steve Bruce

Print publication date: 2018 Print ISBN-13: 9780198786580 Published to Oxford Scholarship Online: August 2018 DOI: 10.1093/oso/9780198786580.001.0001

(p.x) List of Figures and Tables Steve Bruce

Figure 5.1. Locations matrix 109

Tables 3.1. Belief in God, Great Britain, 1990 and 2000 (%) 53 3.2. Belief in God and self-description as religious, spiritual, or neither, Scotland, 2001 (%) 53 10.1. Church involvement in three Durham mining areas, 1851–1941 (% of adult population) 223

Page 1 of 1

 

Prelude

Researching Religion: Why We Need Social Science Steve Bruce

Print publication date: 2018 Print ISBN-13: 9780198786580 Published to Oxford Scholarship Online: August 2018 DOI: 10.1093/oso/9780198786580.001.0001

Prelude Basic Principles of Social Research Steve Bruce

DOI:10.1093/oso/9780198786580.003.0001

Abstract and Keywords So that its positive elements are not lost in the detailed critiques of errors in the study of contemporary religion, this prelude lists and briefly justifies seven basic principles that should guide the researcher. Before we can explain people’s beliefs and behaviour, we need to understand them. As talk is cheap, we should study what people do as well as what they say. Participation is always helpful because joining in activities allows us to check that we understand them. Although we should always be interested in how people see the world and hence should listen to them, we must always be sceptical of what we are told. We should always consider how representative are our subjects. We should hesitate to attribute to others motives we would not impute to ourselves. And, most importantly, we should always be polite and show gratitude to those we study. Keywords:   principles of social research, understanding, participant observation, study of religion, contemporary religion

Many general research texts, because they wish to appeal to the widest possible market, are inclusive; they treat all forms of social research as if they were of similar merit. This book is exclusive and offensive; it presents a particular vision of social science through a series of arguments. It could well have been entitled ‘Mistakes and how to Avoid them’. So that its positive principles are apparent, I summarize them here.

Page 1 of 4

 

Prelude 1. Understanding matters. Before we can explain beliefs and behaviour, we must understand them, and it is generally a good idea to begin by listening to the people we want to understand. Direct conversation is obviously the primary method for communicating with the living (and that is often better done by sidling up to people for casual chats than by formally interviewing them), but for large numbers we have the sample survey, and even the dead can be heard in the fragmentary remains of their talk: chapel minute books, letters, autobiographies, recordings, diaries, and the like. 2. Actions speak louder than words. An obvious problem of the survey questionnaire (and every other record of talk) is that it costs people nothing to assert that they believe this or that or have done this or that and sometimes they straightforwardly lie. Hence, if possible, we should always be interested in what people actually do. We watch and we try to devise measures of activities. And follow the money: one of the best way to understand people’s priorities is through studies of expenditure. 3. If you can, join in. Most religious groups are only too happy to welcome new participants, and a great deal of what they do is patently (p.2) public rather than private. Participation is a good way of learning people’s culture and thus complements my first two principles. Being able to produce a competent performance of taking part in a church service or group meditations is some reassurance that we understand at least minimally what is going on. The ethical issues of covert and disguised participation are discussed in Chapter 6. Here I will say only that many commentators on research are more precious about this than are the people they purport to protect. 4. Always be sceptical. A principled commitment to listening to those we wish to understand does not mean that we should be credulous; after all, there is ample evidence that people dissemble, even to themselves. Nor does it mean we should act as spokesperson for those we study. For reasons elaborated in Chapter 7, we should always think about what interests might be in play when respondents talk about their religious beliefs and actions. We should be even more sceptical about second-hand reports of what people believe or do. It does not concern religion, but the following is too good an example of credulity to miss. The Sunday Telegraph reported: ‘A fetish sex website … is getting thousands of visits each month from computers used by MPs, peers and their staff at Westminster. One such site was visited over 3,000 times in a month; another over 100,000 times.’1 The scale of those numbers should have raised doubts, but newspapers did not want to lose a good story by thinking too hard about that. The correct explanation, of course, was not a sudden outburst of prurient licentiousness among our elected representatives. It was that a bot, which someone had accidentally allowed to implant itself in the House of Commons computer

Page 2 of 4

 

Prelude system, was automatically linking to porn websites. The old adage remains relevant: if it is too good to be true, it is probably not true. As well as being sceptical of reports of people’s beliefs and actions, we should also be aware that we can rarely treat our research subjects as though they were experts on anything other than themselves, and sometimes they are not even that. 5. Consider how representative are your subjects or whom they can plausibly represent. This is particularly a problem for qualitative or ethnographic research. Quantitative researchers are familiar with the need to study samples of people that are representative of the wider population to which they wish to generalize, and they have well-practised techniques for deriving representative samples.2 In the (p.3) natural sciences the ideas generally come first and the research is designed to test them. Social researchers often start with only some vague notion of what might be interesting or informative about their subjects, hoover up as much information as they can, and then see what explanatory arguments might be illuminated by their data. There is nothing wrong with this, except (a) missing data and (b) inappropriate generalization. To the extent we have no idea beyond vague curiosity to guide our observations, it is always likely we will miss something that later turns out to be important. I spent hours in a far distant records office noting figures for Church of England Easter Day communicants from parish registers of services in order to chart the changing popularity of churchgoing in Hertfordshire village. Only later did I realize that, over my time period, taking communion at Christmas Eve grew to outstrip Easter Sunday in popularity. I had to go all the way back and record the Christmas figures. In order to understand the appeal of Pentecostalism, we may well study one Pentecostal congregation and treat the people we come to know well in it as if they were representative of the whole congregation and as if that congregation was representative of Pentecostal churches. Moreover, in claiming that the presence or absence of some social characteristic explains why Pentecostalism does or does not appeal, we may be generalizing about the entire population. We always need to think just how far we can generalize from our sample survey or from our study of twenty white witches. Some topics are so well worked over that we have very large amounts of evidence; for example, about the proportions of the population that go to church, that grew up Muslim, or that believe in life after death. We can be far less confident about why people are drawn to Wicca, even if we have spent a year with a coven. There is nothing wrong with saying ‘My research leads me to hypothesize that witches tend to be x, y, and z’. All too often, however, researchers write ‘Witches are x, y, and z’, with only a grudging concession that their study is limited to one very small group of people.

Page 3 of 4

 

Prelude 6. Be reluctant to impute to those you are trying to understand motives you would not impute to yourself. In many ways the social sciences fall short of the natural sciences, but we do have the advantage that we and the people we study share a common humanity. Our research subjects may indeed be quite unlike us, but more often they are more like us than we may realize; simply asking ourselves whether whatever (p.4) we are writing would sound plausible as an explanation of our own beliefs and actions should prevent egregious nonsense. It is also worth bearing in mind, as we write, how our report would read to one of the people we are observing and explaining. All too often I read research papers in which those described seem like credulous fools. If we do not think of ourselves as credulous fools, we should be reluctant to suppose it of others. 7. Be polite and show gratitude. Finally, we must remember that our research subjects are doing us a favour. Our intrusion into their lives is very unlikely to do them any harm, but equally well it almost never does them any good. If approached in the right spirit (think ‘fulsome expressions of gratitude’), people are surprisingly willing to give time and effort to explaining themselves or to directing us to useful research resources. We should never leave them feeling exploited. Notes:

(1.) Sunday Telegraph, 22 September 2013. (2.) Finding representative samples for surveys has become increasingly difficult. In the last two decades of the twentieth century it was a commonplace that the young and dispossessed were difficult to survey because they were hard to contact or reluctant to take part, but by 2017 pollsters were finding that ‘of every ten people in rich countries they contact, at least nine now refuse to talk’ (‘Democracy’s Whipping Boy’, The Economist, 17 July 2017).

Page 4 of 4

 

The Value of Social Science

Researching Religion: Why We Need Social Science Steve Bruce

Print publication date: 2018 Print ISBN-13: 9780198786580 Published to Oxford Scholarship Online: August 2018 DOI: 10.1093/oso/9780198786580.001.0001

The Value of Social Science Steve Bruce

DOI:10.1093/oso/9780198786580.003.0002

Abstract and Keywords Many elements of religious studies (expounding beliefs or interpreting texts, for example) have no need of social science, but as soon as we make assertions about changes in the popularity of religion, or of certain types of religion, or say why certain sorts of people are more likely than others to be religious, then we stray into territory that can be mapped only with the techniques of social science. This chapter presents a series of ‘things we cannot know unless we do numbers’ and, by answering the most common criticisms of what is now derided as ‘positivism’, makes the case for an old-fashioned scientific view of social science. Keywords:   positivism, relativism, bias, big data, statistics

Introduction Religion may fascinate even those who have little of it. Most liberal democracies are considerably less religious now than they were in 1900, but, if for no other reason than that we argue about its remaining privileges, religion remains important in those societies, and it is obviously still a powerful social force in the rest of the world. There are, of course, different ways of being interested in religion, and social science has little to contribute to many of them. Religious apologists, for example, who simply want to advertise their faith to a sceptical audience can do so without coming close to social science. But when they try to make their product more attractive by claiming that it is growing in popularity, or that it is increasingly attractive to young people, or that it is socially beneficial, then they require the tools of social science.

Page 1 of 29

 

The Value of Social Science Compared to the Germans or Americans, the British seem unusually resistant to social science. Tory government ministers believe that a good education requires Latin and Greek but not sociology, political science, or statistics. The editor of a leading current affairs magazine begins an article by asking: ‘Which is the better way of learning fundamental truths about life’ and offers only two options: ‘reading great philosophers or reading great novelists?’1 Novelists, great or otherwise, are often taken to be reliable reporters. Lara Feigel’s The Love Charms of Bombs: Restless Lives in the Second World War, for example, sounds like social history: it is actually a study of the lives and works of six novelists. A reviewer questioned Feigel’s wisdom in treating her writers’ novels as biographical source material.2 She might have raised the bigger question of why we should treat novels (or novelists) as historical source material. We could make a good case that descriptions of mundane matters—the (p.6) detail that fills in the background of novels—can be treated as reliable reportage because the contemporaneous author is likely both to know them and to have little reason to distort them. Arnold Palmer’s Moveable Feasts: Fluctuations in Mealtimes uses novelists’ accounts of the timing and contents of meals to produce an excellent history of changes in eating habits.3 But, for matters less mundane, fiction seems an unlikely source of fact, and yet the British reading classes prefer it to social science. The second important element of the background to British studies of religion is a particular consequence of our largely secular culture. Because there are fewer religious people in Britain than in the USA, and because religion itself seems less important, British social scientists are less likely than their US counterparts to study religion. This in turn means that relatively more of the contemporary research on religion is produced by religious professionals who have been drawn into asking questions for which their disciplinary background offers little training. Interdisciplinary research is often excellent, and, though this book is a defence of social science, I am not particularly precious about sociology. As Philip Abrams said: ‘it is the task that commands attention not the disciplines’.4 But academic disciplines are not shipping flags of convenience. They have discrete bodies of knowledge, repertoires of questions, and specialist skills, and the three are necessarily linked. In order to know certain things one has to ask certain things and be able to do certain things. People who have little experience of designing and managing attitude surveys, for example, are more likely inadvertently to produce unrepresentative results than people with such experience. A major twenty-first-century project to assess the religious beliefs of English university students found over half describing themselves as religious or spiritual and just over a quarter describing themselves as Christian. Of those, almost three-quarters had attended church regularly. That is, some 18 per cent of students were regular churchgoers. This was presented as evidence of the enduring, or even growing, popularity of religion among young people. And it Page 2 of 29

 

The Value of Social Science would indeed be a remarkable finding when many other sources tell us that around 7 per cent of English adults regularly attend church and that they are disproportionately elderly. Unfortunately, the survey had a response rate of only 9.4 per cent to an email questionnaire sent to a sample of universities, which included a disproportionate number of former Church of England training colleges in quiet rural towns. The students who were invited to complete the questionnaires were randomly (p.7) selected, but it is almost certain that students who were religious or spiritual were more likely to respond than students who were indifferent or hostile to religion. Even worse, there was nothing to prevent early responders (such as members of religious student societies) encouraging their friends to complete the questionnaire. The results have considerable value for understanding the people who completed the forms, but they can tell us nothing about English university students in general.5 If one purpose of this book is to establish the relevance of social science for the study of religion, a second is to promote a particular kind of social science. The distinction between qualitative and quantitative research is often exaggerated, especially when those research styles are attached to ancient and unresolved arguments about ontology and epistemology or, in English, the nature of the world and the proper form of its study.6 That philosophers cannot agree suggests there is very little point in us now joining those debates. Nonetheless, we have to understand some of those arguments, and I will, where necessary, explain them and suggest pragmatic responses. All that needs to be said at this point is that British studies of religion are far more likely to be based on qualitative than quantitative research and that this creates a number of shortcomings that should be addressed.7 This does not mean that I denigrate ethnography, have a principled preference for descriptions in numbers rather than words, or lack interest in what sense people make of their religious beliefs. It does mean that the sort of social science often derided as positivism is essential for some of our work. The book has a third purpose. Although its primary focus is social research, its examples are drawn from studies of religious belief and behavior, and so it also presents a very large number of what I hope are interesting and important observations about the nature of religion in the modern world. In this chapter I want to make the case for social science in the study of religion. In later chapters I will pursue these topics in greater detail.

In Defence of de Facto Positivism W. G. Runciman suggests we can usefully distinguish four tasks of the social scientist: reportage, description, explanation, and evaluation.8 We can report on events, situations and occurrences: what happened (p.8) in Paris in 1792? We can describe how the participants felt or understood their situations: what did the revolutionaries think they were doing? We can explain why things happened: Page 3 of 29

 

The Value of Social Science what were the causes of the French Revolution? And we can morally evaluate actions: were the consequences good or bad? For reasons discussed in detail in Chapter 5, I am less sure than Runciman about the fourth task. I take the third— explanation—to be the primary goal of the social scientist. But we have to be clear about what we are explaining; hence accurate reportage and a clear grasp of our subjects’ understanding of their situation, motives, intentions, beliefs, and values are essential first steps. But how should we approach these tasks? In the early days of social science there was, as the name suggests, a consensus that we should emulate the methods of the natural sciences. We could not often experiment, but the logic of developing and testing ideas that had served the study of chemicals well could do the same for the study of people. There was a lively debate in Germany around the end of the nineteenth century over the nature of the human or social sciences.9 Herbert Blumer in Chicago spent the 1950s arguing against emulating the natural sciences.10 But it was with the massive growth in the social sciences in the 1960s that such competing perspectives as symbolic interactionism, Marxism, and later feminism and postmodernism proliferated, and the roof fell in on positivism. The errors of positivism are now legend; its critics legion. And yet, I will argue, there is no alternative to a pragmatic positivist approach if we wish to produce testable propositions about the real world. The de facto positivism that I believe to be inescapable differs considerably from the grand vision of Auguste Comte and might best be described as the bastard child of Emile Durkheim and Max Weber.11 It is an ugly beast that will satisfy no philosopher of science, but it describes pretty well what most of us who study naturally occurring social phenomena actually do. Positivism is that approach to knowledge characterized by all or most of the following: 1. the aim to produce linguistic or numerical theoretical statements; 2. a concern with the logical structure and coherence of these statements; (p.9) 3. an insistence that at least some of these statements be testable; 4. ‘testable’ here meaning verifiable, confirmable, or falsifiable by the systematic observation of reality; 5. the belief that science is cumulative, 6. and trans-cultural, and 7. rests on results that can be separated from the personality or social position of the investigator; 8. the belief that science contains theories or research traditions that can talk to each other;

Page 4 of 29

 

The Value of Social Science 9. the belief that science sometimes incorporates new ideas that are discontinuous from old ones; and 10. the assumption that, underlying the various scientific disciplines, there is basically one science about one real world.12 Some of those are straightforward, but I will elaborate a little. The first two are aspirations for theory formation and are hardly controversial, unless, as Gerhard Lenski has done, one insists that our statements should take the form of mathematical axioms or, as Stanislav Andreski believes, statistics are misleading.13 We should try to frame our tentative explanations of social behaviour or social structure in ways that allow them to be tested. We cannot get far if our ideas are like the proposition advanced by one of my students that ‘secularization is all about modernity’. ‘All about’ might be specific enough for Pixie Lott’s toe-tapping pop hit It’s All About Tonight, but we need to be clear what causes what and each ‘what’ needs to be specified as narrowly as possible, as do the strength and direction of the causal relationships. Nor is it helpful, as many people in the arts do, to misuse the word ‘truth’—as in composer John Taverner’s ‘Stockhausen was a seeker after truth’—to mean conviction or enthusiasm.14 I could improve my student’s proposition by saying that individualism + religious diversity + an egalitarian ethos + political stability + state neutrality + the confidence in human faculties given by science and technology undermine traditional shared religious belief systems. This is clearer, but it would be better still if we could define each of those terms so as to identify the things in question unambiguously and specify how much of each has the desired effect. We know that states that are basically democratic (hence the egalitarian ethos in the equation) find it hard to impose conformity to a single national church if there is (p.10) considerable religious diversity, and they generally give up the struggle. But just how much religious diversity is needed to have this effect? And are there different types of religious diversity? For example, differences of religious affiliation within the upper classes might be more consequential than differences between classes. Similarly, diversity within a long-established religion may be more consequential than the arrival of an immigrant alternative: so long as the ‘natives’ share a common religion, that some migrant groups worship other gods might be of little consequence. But, even if we can be clearer about diversity, we need to do a lot more. I have listed six contributory factors—which is already an artificial simplification of my explanation of secularization—but not specified how they interact. Does any combination have the effect of undermining religion or, just as spray-painting a lump of iron ore is less useful than spraying steel after it has been shaped into car body parts, does the order in which the various factors come into play matter? Are all six causes necessary or can a large dose of one substitute for the absence of another? For example, the Lutheran states of Northern Europe had Page 5 of 29

 

The Value of Social Science very little religious diversity before their state churches started to haemorrhage adherents, but they did have a lot of ‘individualism’ (one meaning of which in this context is a commitment to individual liberty). So can extra weight on one factor compensate for the absence of another? Already it is clear how difficult it is to translate the concerns of the student of religious change into the theory style of the natural sciences, and, for reasons I will discuss later in this chapter, it seems certain that the social sciences will never come close to the precision of physics. Our wooliness is not a function of being a relatively new discipline; sociology will not, as Lenski hoped, grow up to become physics. Nonetheless, we must aim for clarity if we wish to explain anything. The third and fourth propositions are vital. There is little point in generating competing explanations of social behaviour unless these can be tested in a manner that eliminates the worst and improves the best. For most social researchers testing means holding against the testimony of the real world. That is, we suppose that, with skill and ingenuity, we can discover facts about the object of our enquiry that allow us to develop and hone explanations. There are alternatives. Faced with a contentious assertion about, for example, the effects of contraception on teenage sexual behavior, we could ‘take it to the Lord in prayer’ or seek guidance in Holy Scripture. We could (p.11) smoke hashish and test theories against subsequent drug-fuelled revelations (if we could remember them). We could argue that whether a theory is true or false is to be known, not by how well it fits existing evidence, but by how well it encourages particular responses. For example, some Marxists argue that what distinguishes knowledge from false consciousness is not its fit with current evidence but its revolutionary potential.15 Some feminists argue that their research methods are justified by the ‘transformative’ potential of the knowledge so generated.16 The problems with such views are many, and it is enough to mention just two. First, it is difficult to agree on moral judgements. Testing through praxis requires that we agree on what would count as an improving transformation, and often we do not. Ann Oakley might find widespread agreement that the point of studying women giving birth is to reduce patriarchal oppression, but we are unlikely to agree that the way to test competing theories about the structures of terrorist organizations is to help them improve their killing operations. Second, we cannot now know which is the best theory if the transformative justification requires that we wait around to see what happens when those people who read our papers and books react to them. Fortunately, most researchers would agree that testing propositions means seeking evidence that either supports or refutes them from the here-and-now.

Page 6 of 29

 

The Value of Social Science Our ability to make accurate observations of the world that can provide theorytesting evidence is doubted by philosophers who present reasonable arguments against the assumption that observation is, in principle at least, unproblematic. However, as such arguments have been pursued for centuries without resolution, I suggest we instead consider how well people manage observation and evidence-gathering in their day-to-day lives. My case is the pragmatic one that, between the perfectionist extremes of competing epistemological positions, there is sufficient common ground for us to go about our business. Consider the example of the 1851 Census of Religious Worship—the only attempt by a British government to measure in detail the religious affiliations of its people. This was a massive bureaucratic undertaking that used the population census structure of a hierarchy of officials responsible for distributing, collecting, checking, and returning the forms to administer a questionnaire to all known worship outlets. A church or chapel official was asked about the age and size of the church or chapel and about the (p.12) number of people attending services on Sunday, 30 March. Although the lowest level of officials put considerable effort into chasing up missing returns, a small number of generally very small outlets did not report. We can also have reasonable doubts about the accuracy of the returns in which all the figures are multiples of ten. More debilitating, the forms asked about attendances, not attenders, and, as many churchgoers then attended twice, we have to reduce the totals appropriately to get an estimate of the percentage of the population that attended church or stick to presenting a likely maximum figure (if no one attended more than once) and a likely minimum (if everyone who attended at all attended twice). But we should all agree that the census reports are still better evidence than the estimates of Mrs Effie McClumpher of Gargunnock. More than that, we can, with ingenuity and effort, improve the census estimates considerably. Alasdair Crockett applied the sorts of mathematical and logical skills common in codebreaking to improve the conclusions by calculating possible ranges of variation.17 For example, many returns give totals in neat multiples of ten, which suggests that the reporters guessed rather than counted. Instead of concluding we should ditch the census, we can first of all estimate, by examining a sample of the original returns, what proportion suffer this defect. Then we can estimate what difference such sloppy reporting makes to overall accuracy by assuming first that all the 20s might be 15 and second that they might all be 24, all the 30s might be 25 or 34, and so on. We cannot produce exact figures, but we can calculate plausible ranges and, for many of our concerns (such as ‘is churchgoing more or less popular in 1951 than in 1851?’), figures that can be represented as ‘plus or minus 6 per cent’ will be perfectly adequate. We regularly demonstrate that we do not need to solve all the problems of epistemology by giving practical demonstrations of such principles as the more information, the better and the more informed a source, the more

Page 7 of 29

 

The Value of Social Science reliable. That is, we have less difficulty knowing what would count as evidence in practice than in theory. Whether knowledge advances by the accumulation of supporting examples or by the absence of negative examples—that is, by what Karl Popper called verification or falsification—is another of those unresolved philosophy of science arguments that for most purposes we can sidestep.18 Whether the advance of science depends on verification or falsification and whether either is ‘really’ possible did not stop the early anatomists improving our understanding of the (p.13) circulation of blood, and it need not stop us knowing, for example, what would count as a good test of the proposition that the danger inherent in their work made miners and fisherman particularly prone to evangelical religion (the example is pursued in Chapter 10). We might haggle over every step in the process, but we can see the basic contours: find communities of miners and fisherman and compare their church- or chapel-going with that of communities of manual workers whose jobs are considerable safer. We could become more sophisticated. We could compare the religious activity of lead miners (whose work was arduous but no more dangerous than most manual labour) and colliers (whose work, because it involved poisonous and flammable gases, was unpredictably dangerous). We could compare the religious commitments of inshore and deep-sea fishermen. We could compare the churchgoing of communities based on both industries over time as they became safer. And so on. Provided we agree that there is a single real world out there, we should be able to test propositions about it. The fifth proposition is important because, if knowledge acquired through research cannot be cumulative, research is pointless. None of us can study everything. We each build on what others have done. I can try to explain why the people of the Hebridean island of Lewis remained religious longer than the people of the apparently similar Orkney and Shetland islands only because I can read the research of people who have studied in detail elements of the economic, political, social, and cultural lives of those islands over a time period longer than my lifetime. In turn I hope that the conclusions I come to about those three islands will be refined and tested by others studying settings that are similar in some respects and different in others. The sixth and seventh points are an extension of the fifth. Natural scientists suppose that propositions about the relationship between temperature and pressure hold true for the behaviour of gases in whatever country one performs the experiment; social scientists rather more humbly can hope that the propositions they develop are not too narrowly culturally confined. Here we need to avoid a common misunderstanding. I do not mean that social science ignores cultural differences: that would be absurd, when such cultural differences are often the very things we study and comparison is key to that Page 8 of 29

 

The Value of Social Science study. As Durkheim put it: ‘comparative sociology is not a branch of sociology; it is sociology itself insofar as it ceases to be purely descriptive and aspires to (p. 14) account for facts’.19 What I mean is that we aim to identify a common set of concerns and conceptual devices for the study of more than just the one piece of culture or society that interests us now. For example, the comparative study of kinship rests on definitions that find their appropriate expression in a large number of times and places; in order to say anything worthwhile about the differences in the roles adopted by ‘fathers’ in Lutheran Norway in 1900 and Catholic Colombia in 1980 we need to have some transcultural concept of father that underlies all the different versions we see. And, if we can develop some proposition that sufficiently survives testing in a variety of cultures, then we suppose we have found out something of general applicability. If my colliers and fishermen example works for modern Britain (and I know it works for the Pacific islanders studied by Bronislaw Malinowski because I borrowed the example from him), does it work for those occupations in a wide variety of religiously very different societies? We could add Chinese and Ukrainian coal miners to our study. The intercultural application of our theories is not the universality of Boyle’s gas laws and Newton’s laws of motion. It is the lesser but nonetheless necessary ambition of identifying regularities in human behaviour in terms that are not themselves confined to the culture of the observer. The principal objection to these points is relativism. It can be argued on a wide variety of grounds that objectivity is never possible, that every explanation is simply another narrative, that all narratives are social constructions that reflect the interests and preconceptions of those who devise them, and hence that knowledge is always only knowledge from a particular time, place, and social position. A small number of thinkers argue that even the natural sciences are social constructions in this sense.20 Paul Feyerabend’s Against Method is a classic statement of the extreme anti-science position: he called himself an ‘epistemological anarchist’.21 It is more common to exempt the natural sciences but condemn the social sciences to imprisonment within the ideological blinkers of its proponents. Some scholars condemn most research to the relativist prison but allow themselves a revelatory get-out clause. So some feminists dismiss most social science as a patriarchal social construction but claim that their emancipation allows them to discover the truth. Marxists allow an escape from false consciousness to true knowledge through their alignment with the interests of the working class (or, more usually, with what middle-class intellectuals tell the proletariat are their interests). (p.15) Again, whether we can produce knowledge rather than just another ‘account’ or ‘voice’ is one of those enduring philosophical arguments, and again my solution is the same. While there are serious practical difficulties in producing knowledge that is accepted as such beyond any one particular time and place, this need not prevent us trying, and we can find a number of justifications for so doing (which I will consider in

Page 9 of 29

 

The Value of Social Science various places in this book). Here I will offer just the two most obvious ripostes to relativism. First, the people who argue that knowledge is relative and that objectivity is not possible do not consistently pronounce the same death sentence on their own assertions about the world. Second, that inconsistency is repeated and multiplied in everyday life. Relativists treat their pay scales, workloads, electricity bills, mortgages in a positivistic fashion: measuring, comparing, and evaluating. It is true that electricity consumption has greater ‘facticity’ than ‘affection towards child’ or ‘religious commitment’, but only in degree, not kind. If everyday matters can be treated as though they have an explorable existence beyond the consciousness of the observer, I see no reason in principle for not applying the same assumptions to more complex issues. The importance of point 8 is that it shows the way out of endless division into ‘schools’. People naturally associate themselves in groups and traditions of thought. In the case of popular religions, those schools are largely incommensurate. That is, you cannot become a Muslim so long as you believe Lutheran things and vice versa. While Muslims and Lutherans have much in common, they are divided by competing claims that cannot simultaneously be true. Either God’s sending Jesus, his only son, to die as a sacrifice for our sins was a once-and-final act or Allah is God and Mohammed is his only prophet. One can, of course, combine elements of both in a new third religion, but, taken in their own terms, Lutheranism and Islam are incommensurate. Academic disciplines sometimes seem as sect-riven as religions. On the question of whether religious diversity undermines religion or strengthens it, the secularization school and the supply-side or rational-choice school (discussed in Chapter 9) seem thoroughly antagonistic, but they need not be. We can devise tests that allow us to decide which of these is right about what. The current adherents of those competing positions are unlikely to change their minds, but the rest of the scholarly community can arbitrate the dispute. (p.16) The ninth point—that knowledge changes—is important because it distinguishes science from other ways of thinking. Traditional Jews, Christians, and Muslims believe that their faiths were delivered complete and perfect. The good Jew, Christian, or Muslim is bound by the past; the good scientist is not. For long periods scientists may continue to work within a set of assumptions that contain a fundamental error, but, when that error is finally admitted, there will be what Thomas Kuhn called a ‘paradigm shift’.22 ‘This is what we have always done’ is a good argument in revealed religion or in Trobriand Islander fishing magic; it is not a good argument in natural science and nor should it be in the social sciences. It may often seem that grand theories in social science rise because they are novel and fall when the next generation finds them dull and dismisses them as the intellectual equivalent of Dad Dancing, but at the lower level of social science explanations of particular phenomena there are many Page 10 of 29

 

The Value of Social Science examples of ideas being dropped because they have failed to survive against the run of the evidence. The final point is so central to the scientific enterprise that it could easily have been the first. Physicists believe there is just one world. If Boyle’s gas law—at constant temperature, the volume of an ideal gas is inversely proportional to its absolute pressure—is right, it is right in Nigeria and Sweden. There is not a physics that works only for Swedes or only for Nigerians. Again I have to phrase this very carefully to avoid being misunderstood. For there to be any value in social research, the social scientific study of even culturally diverse phenomena must proceed on the basis that there is just one real world to be studied, otherwise we could not plausibly claim that cultures are actually diverse. I do not mean that, for example, all witchcraft is the same. I mean that, if equally well-educated and observant Nigerian, Swedish, and American social scientists study the witchcraft of Scotland in the eighteenth century, they will eventually come to the same conclusions.23 That they come from different cultures may cause them initially to concentrate on different facets and to ask different questions, but there is just the one thing that is being studied and they will eventually reconcile their differences, or future generations will arbitrate, by devising crucial tests to eliminate the mistaken versions. I should add that nothing I have said means that any of the above is easy and some of it may sometimes be impossible. My attempts to study the connection between inherently dangerous occupations and (p.17) receptivity to religion (the subject of Chapter 10) may fail because, although we can specify what would count as good evidence, we may not be able to find it. Nonetheless, I believe that the general logic of science offers the only way to prevent social science degenerating into literary criticism or fiction. Putting twenty people who know nothing about the matter in a sealed room and asking them to write an account of the passage of the 1832 Reform Bill may produce twenty delightful narratives. But, if we want something other than fiction, we need an approach based on something like the above description of positivism. Various criticisms of this de facto positivism will be addressed in a variety of places in what follows. Here I want to address the most pressing. Is Human Behaviour Different?

Some social scientists suppose that the raw material of social science is not in principle different from that of the natural sciences. Famously Durkheim argued that societies had characteristics (such as a suicide rate) that were independent of the people who composed those societies and that the task of sociology was to explore the relationship between such variables in the same way that Boyle explored the links between volume, pressure and temperature. Equally well known is Weber’s supposedly competing view that sociology should be concerned with understanding subjectivities and should aim to render human Page 11 of 29

 

The Value of Social Science behaviour understandable. As is often the case with such polarities, whatever each may have said in their programmatic statements about research, each in practice did what the other proposed. Durkheim’s Suicide is full of guesses about subjectivities and individual motives and Weber’s ‘Protestant ethic and the spirit of capitalism’ essay is a masterpiece of causal connections.24 Social science has to be concerned with understanding, because the cause of people’s behaviour lies in how they, not the disinterested observer, see the world. The action of the man running from the house shouting ‘Fire’ is explained by his belief that his house is on fire, and that remains the explanation even if he is mistaken. As W. I. Thomas put it in his dictum concerning the ‘definition of the situation’: ‘If men define situations as real, they are real in their consequences.’25 We would now say ‘people’ rather than ‘men’, but the point remains valid. However, I do not see that the role of (p.18) individual subjectivities prevents us from aspiring to a basically scientific approach to human behaviour. It just means that we have an extra layer to work through: we cannot confine ourselves to external appearances. So, for example, we cannot simply say that, because yoga in India is a Hindu spiritual discipline, the young mothers who take the yoga class in my village hall on a Friday evening are being Hindu or even being spiritual. We need to ask them how they understood what they were doing. And, when we do, we find that almost all of them are engaged in a purely secular exercise programme. While the need to understand before we can explain is sometimes presented as a weakness of the social, as compared with the natural, sciences, it can be seen as a strength. To quote Marshall Sahlins: ‘we have the possibility of knowing the cultures of others in ways that are … more powerful than the ways natural scientists know physical objects’.26 One way of testing any causal explanation is to unpack it to the level of the acting individual and ask whether it rings true. We should hesitate to apply to others any model of action that we would not readily apply to ourselves. Reflexive implausibility is not always grounds for rejection—we may be persuaded that those people really are unlike us—but, as I argue in Chapter 9, it should raise questions. Generalities and Laws

One major adaption of science for the study of people is the replacement, as our aspiration, of the universal law by the statement of probabilities. A lot of science also rests on probabilities rather than certainties, but it aspires to propositions that brook no exception. Exceptions just point scientists to the next stage of improving knowledge by better generalizations. Boyle’s law holds for all gases and all ways of measuring temperature and pressure. The social scientist has to admit that human beings are, severally and jointly, too complex for social action and social structure ever to be described in universal laws unless those laws are so bland as to be contentless. The best we can do is to develop statements of general likelihood. ‘The more sources of legitimate authority that are accepted Page 12 of 29

 

The Value of Social Science by the members of a religious movement, the greater the propensity to schism’ is the sort of proposition we can reasonably develop and test.27 When we say that poverty causes ill health, we do not mean that every poor person will always be in worse health than every rich person. Some people (p.19) have good genes; others have fortunate circumstances. What we mean is that, when we compare 200 people living on less than the minimum wage with 200 people earning more than £100,000 per annum, we will find more people with health problems in the first group than in the second. We also mean that the connection is not an accident or an indirect reflection of some third-party cause. Conservatives are fond of blaming the poor for their ill health and suggesting that, if they ate tofu instead of grease burgers, they would be as healthy as their wealthier counterparts. When we say poverty is the cause, we could mean either that poverty is the cause of unfortunate food preferences or that it is the cause of other considerations (such as access to good health care in infancy) that would not be changed by a change in the diet of the poor. That we trade in probabilities rather than universal laws is important when we turn our research on its head and work backwards from our general claims to any one individual. This may seem obvious, but it should be spelt out because the point is sometimes missed. I can say with confidence that, in modern Britain, very few people who were not socialized into a faith during childhood will acquire one in adulthood. That is, religious preferences are generally acquired in the family and are unlikely to change. One source allows us to put a figure on the proportion of people raised with no religion who later became religious: about 5 per cent.28 Like all probability statements that is fine for talking about Britain as a whole, but it does not allow us to know whether the greatgranddaughter of Mrs McClumpher of Gargunnock, raised with no religion, is going to become religious in later life. But also, and this is the point some people get wrong, it does not mean that we can refute the general proposition that religion is acquired in childhood by pointing to the junior Ms McClumpher’s adult conversion to Scientology. Because it aims to express an invariant connection, Boyle’s law would be proved wrong if one day the ideal gas behaves differently. As Karl Popper argued, universal laws are destroyed by single exceptions.29 Probabilities are refuted only by exceptions that amount to more than the leeway permitted in the statement of probability. This may seem obvious, but I routinely find that a member of an audience for a talk on the decline of Christianity in Britain will respond that national data on church decline cannot be right because his or her church is growing. To conclude this preliminary explanation and defence of positivism, I should add that many criticisms of positivism are entirely (p.20) reasonable. For example, it is true that an act of faith is involved in assuming that we can infer the future from the past and present. There is no way of proving that well-established trends will continue (other than to wait and see); but, equally well, scepticism about the validity of induction does not of itself justify expecting any particular Page 13 of 29

 

The Value of Social Science alternative. I am not committed to de facto positivism because it has answered all the queries that philosophers of science have raised. I accept it as the least bad approach that is available to us. Whatever faith is required to believe that, if we correctly understand some pattern of action and can see no reason to suppose a new variable will intervene, then we should assume that the next case will also fit the pattern, it involves less credulity than the metaphysics of Marxist historicism or Christian millennialism. Most social scientists are de facto positivists despite the problems of positivism because the alternatives are worse. If this view offers a reasonable preliminary explanation of why social research should, and more often than not actually does, rely on a pragmatic positivism, I can move on to consider different sorts of research method. As I have noted, the common distinction between qualitative research (often now called ethnography) and quantitative research is exaggerated. Many scholars do both, and, as I will argue by way of rebutting some of the more extreme criticism of quantitative research, ethnographers and fieldwork researchers use the same principles of observation and argument as social statisticians. But first I will try to convince the sceptic of the value of numbers by presenting a variety of problems in the study of contemporary religious behaviour that can be addressed only by statistical analysis of quantitative data.

Things you Cannot Know without Numbers British sociology of religion has a proud tradition of detailed studies of small religious groups. Bryan Wilson is probably now best known for his presentation of the secularization paradigm, but he had an enduring influence on the sociology of religion through his supervision and inspiration of a generation of students of small sects and new religious movements: James Beckford, Roy Wallis, Eileen Barker, Susan Budd, Michael Hill, Alan Rogerson, and James Whitworth are just some examples. Although he had no objection to statistical (p.21) research, Wilson encouraged his students to immerse themselves in the movements they studied. The case-study method was further encouraged by the institutional history of sociology in Britain. Sociology came late to Britain, once anthropology was well established, and in many universities it was based initially in joint departments. For example, the first head of the Sociology and Anthropology department at Stirling was Max Marwick, a renowned anthropologist of African witchcraft. There were notable exceptions—the London School of Economics and Nuffield College, Oxford—but most British sociologists of the 1970s and 1980s were probably trained in departments where the anthropological model of research was far more influential than what was sometimes cordoned off as ‘social statistics’. There is much in the British tradition of ethnographic studies of religion of which to be proud: all those named in the previous paragraph produced monographs that are still read forty years on. However, there are limits to what one can infer from ethnography. For example, Tanya Luhrmann’s excellent study Page 14 of 29

 

The Value of Social Science of English pagan groups gives us great insight into the motives, attitudes, beliefs, and practices of the small number of people she studied.30 Anyone who reads her book will learn a great deal about modern paganism. What they will not learn is the number of pagans in Britain and whether paganism is now more or less popular than it was at some earlier time. If there is a typical pagan, one cannot learn that from Luhrmann, because her sample is too small to be representative. If we cannot distinguish typical pagans from, say, typical Unitarians, we cannot hope to explain the appeal of paganism. This may seem obvious, but it frequently confounds ethnographers. Discovering that nineteen of the twenty witches in your coven have some common feature may seem like justification for taking that as a cause of being attracted to witchcraft, but, if most of the population that has not joined the Craft also has the same characteristic, it is not. That is, the search for the cause of some action can reasonably start with just those actors, but there needs to be a second stage in which we check that the supposed cause is not just as common with those who do not act like that. The limits of ethnography have long been well known. Blumer was one of the founders of symbolic interactionism—the theoretical approach that through the 1960s and 1970s underpinned much good qualitative sociology—and he was one of the severest critics of what he called ‘scientism’ in sociology. Over a long and influential career, Blumer repeatedly criticized the pretensions of variable or quantitative (p.22) research. Yet when he was invited to evaluate one of the masterpieces of the qualitative tradition—W. I. Thomas and Florian Znaniecki’s The Polish Peasant in Europe and America—Blumer concluded that its casestudy method did not meet the test of the four criteria he expected of any scientific instrument: that the data generated be representative, adequate, and reliable, and allow decisive testing of competing theoretical schemes.31 ‘Representative’ identifies the key weakness. How do we know that the people the ethnographer is studying are typical of that collectivity? I am sure she did not, but Luhrmann could have studied only those North London witches who fitted her tentative explanations. And how do we know that the studied collectivity is representative of the wider population to whom the ethnographer generalizes? The problem is particularly acute with case studies, because it is precisely the exotic and the unusual that catches the eye. In 1843 a third of the ministers and laity of the Church of Scotland split off to form the Free Church. There are many detailed studies of the early Free Church; far less attention has been paid to the two-thirds who did not leave. The same point could be made about almost every religious schism: change and novelty are more attractive than stability. Over the decades survey researchers have developed protocols for ensuring that samples

Page 15 of 29

 

The Value of Social Science studied are as representative as possible and for adjusting data to make them more so. Ethnographers barely raise the problem, let alone solve it.32 A Spiritual Revolution?

The limits to what can be inferred from even the best ethnography would be less troubling if ethnographers bore them in mind, but we routinely find such limits ignored. Darren Kemp, for example, makes specific assertions about the growing importance of New Age spirituality that cannot be tested by his method of describing a variety of New Age ideas and practices.33 Michael York concludes a descriptive account of the New Age in three settings by saying: Whilst the numbers involved with new forms of religiosity remain hard to identify precisely, we can at least recognise the ubiquity and growth of the diffuse religious consumer supermarket, which demonstrates an increasingly vital presence in both urban Holland and rural southern France.34 (p.23) The references to size—‘ubiquity’, ‘growth’, and ‘increasingly vital presence’—all imply that there is a lot going on. What appears to be an important qualification—the difficulty of precise identification—is neither here nor there. Just looking at my car and my cat, I find it hard to ‘identify precisely’ their weights, but I have no trouble knowing one is vastly larger than the other. Like the good magician (in the entertainment rather than the New Age sense) who deflects our attention from the real site of action with a lure, York pleads guilty to imprecision in the hope that we will not notice the absence of evidence for his primary claim. In a discursive essay on religion in modern Scotland, Steven Sutcliffe approvingly cites a theologian, writing: ‘most unchurched people in Scotland today are more likely to construct their worldview from aspects of the New Age outlook than from elements of mainstream Christianity’.35 We could justify the phrase ‘more likely’ in the sense that, because I am elderly, I am more likely to be mistaken for Madonna than Taylor Swift: true but trite because I am not going to be mistaken for either. I suspect he wishes us to take ‘more likely’ to mean ‘likely’ or ‘quite likely’. There is no evidence for this claim. Sutcliffe himself uses awkward prose to argue against the view that the changes in the religious culture should be seen as secularization: our attention should shift … from theorisations of secularisation as a ‘dereligioning’ factor in the modern world to a revisionist theory of secularisation as the laicisation and domestication of religious discourse and action in the culture at large … This relocates rather than erases religion: we look now for vibrant—if necessarily unstable—sites beyond the boundaries of those large social institutions (the churches) which have historically been treated as the hegemonic signifiers of ‘religion’.36 Page 16 of 29

 

The Value of Social Science The implication is that we will find what we are looking for. He goes on to list observations we can summarize as follows: Scotland now has citizens who belong to a wide range of non-Christian religions, it has a major New Age centre at Findhorn, Moray, and there is a lot of other New Agey stuff going on (which he illustrates with a detailed account of a fire-walking ceremony). Actually, the proportion of Scots who claimed a non-Christian religious identity in the 2001 Census was just 1.9 per cent, the Findhorn Foundation is staffed and visited mostly by foreigners, and, although the range of alternative spiritual activities one can identify is very large, the numbers of people in any way involved in any of them are tiny. (p.24) We do actually have a good large-scale test of the popularity of New Age spirituality. Heelas and Woodhead made a pioneering effort to measure what they called the ‘holistic spirituality milieu’ in Kendal, a town in the north-west of England that seems sufficiently typical to be taken as more or less representative of Britain.37 Their conclusion (based on a praiseworthy effort to identify every form of group activity and one-to-one service provision) is that somewhere between 1 and 2 per cent of the population was involved in such activities as yoga, tai chi, aromatherapy, spiritual healing, and the like in a typical week. An obvious difficulty is the classification of activities as spiritual or secular. Heelas and Woodhead allow us to add the insider’s perspective to our own judgements because they asked their respondents a number of questions about the extent to which they saw their activities as spiritual and about the importance of spiritual growth. The results fit pretty closely with what an outsider would expect. In the one-to-one activities, the practitioners were more likely than their customers to see what they were doing as spiritual. And the most popular activities were those least likely to be seen as spiritual. If we are to regard holistic milieu spirituality as compensation for the decline of the Christian churches, we need to get the scale right. The churchgoing population of Kendal is 7.9 per cent. If we increase that to 15 per cent to give some idea of those who are at all interested in Christianity, we are left with 85 per cent of the population or almost 32,000 people who are not in the mainstream of religion and who are thus, in theory, available for recruitment to some alternative. Set against that potential market, the 270 people involved in the holistic milieu for spiritual reasons seems very small. The spiritual innovation, which some scholars wish to describe as the zeitgeist, is attracting only a tiny fraction of the people made available by the decline of mainstream religion. The textual and ethnographic research of scholars such as Kemp, York, and Sutcliffe makes claims for significance that their type of research cannot sustain, and the sort of research and statistical analysis that could sustain such conclusions fails to do so.

Page 17 of 29

 

The Value of Social Science Believing but not Belonging?

For two decades it was common to claim that the decline of the British churches represents, not a loss of faith as such, but only a loss of faith in religious institutions. It was asserted that Britons were, (p.25) as Grace Davie subtitled her book Religion in Britain since 1945, ‘believing without belonging’.38 This assertion is often supported only by evidence that, at any one particular time point, some weak measure of religious interest scores higher than more rigorous measures: for example, the numbers of people who say they believe in God is generally considerably higher than the numbers who go to church. But no attention is paid either to the meaning of the first claim or to trend figures. Having approvingly cited Davie, academic and Church of England cleric Martyn Percy wrote: ‘statistical surveys continually support the thesis that England is a place where the vast majority of the population continue to affirm their belief in God, but then proceed to do little about it’.39 This is misleading. What the surveys actually show is that the English divide fairly evenly between four positions: belief in a personal creator God, belief in some higher power of life force, belief that ‘there is something there’, and not believing in God. That is, three-quarters of the English do not believe in the sort of God that the Revd Percy and the Church of England worship. More telling, the 2016 British Social Attitudes survey showed a majority saying they had ‘no religion’.40 Furthermore, the believing-but-not-belonging case is made by comparing the long-run trend of declining church involvement with snap shots that show that, at any one time, religious beliefs are more popular than churchgoing. This unbalanced contrast should prompt the social scientist to ask what is the trend for religious belief. Alasdair Crockett and David Voas combined all the responses to the British Social Attitudes surveys from 1983 to 2002 to produce a body of data for about 60,000 respondents and graphed indices of religious identification, church attendance, and support for specific religious beliefs.41 They found that three separate measures of religious interest—religious identification, religious belief, and religious activity—changed over the twentieth century along similar trajectories. In every decade, claiming a religious identity was more popular than assenting to specifically Christian beliefs, which in turn was more popular than attending church. But all three trend lines were downwards. And, if one wants to make anything of small differences in the trajectories, the belief line actually fell faster than the other two: that is, the British are better described as ‘belonging without believing’! The belief line certainly did not stay stable while the other two declined. Crockett and Voas might be wrong. The important point is that the ‘believing without belonging’ depiction of contemporary religious culture can be tested only with big data. No (p.26) small-scale community study, no matter how well done, can test that description.

Page 18 of 29

 

The Value of Social Science Does Ageing Make us More Religious?

Without the right sort of information and the right approach to analysing it, it is easy to be distracted by red herrings. For example, it is well known that church congregations are markedly older than the population at large. This is sometimes explained by the idea that, as people get older, the deaths of friends and relatives and the approach of their own demise makes them more mindful of religious questions. For that explanation to be valid it would have to be the case that congregations were full of people who began to attend only when they got old. Large-scale surveys clearly demonstrate that a very different explanation is more plausible. Congregations are disproportionately elderly because they have failed to recruit young people. We now know that the ‘death makes people more religious’ proposition is a false start. We might have guessed that from ethnographic studies of religious communities if the researchers had pressed respondents on their detailed careers of churchgoing, but large-scale surveys make it much easier to compare the religiosity of age cohorts. Why Are Conservative Churches Growing?

So that it does not seem that I am claiming omniscience, I will add a mistake of my own. After Dean Kelley’s Why the Conservative Churches Are Growing provided a robust presentation of the idea that conservative Protestantism in the USA was faring better than its mainstream and liberal alternatives because of intrinsic features, very many scholars (myself included) spent much time and energy arguing over precisely which features of conservative religion explained its appeal.42 It turned out we were all wasting our time. Using large-scale data sets, demographers demonstrated that the differing fates of different religious organizations were largely explained by differences in fertility.43 People switching from mainstream to conservative denominations accounted for only 4 per cent of the total change; 76 per cent of religious change is explained by family size. Even if Unitarian and Southern Baptist families were equally effective in retaining their children in their respective faiths, the former would have declined and the latter would have grown because Baptists typically had more (p.27) children. Hence, if there is anything in the relative success of conservative Protestantism in the USA that should exercise us, it concerns matters some way from the original Kelley-inspired argument; instead of trying to explain sectarian religion’s superior appeal, we need to explain why Southern Baptists like large families and why they marry young and thus enjoy aboveaverage fertility. The large-scale survey work of demographers told us something vital that those of us engaged in detailed ethnographic studies of conservative and liberal churches could not see. Does Being Religious Make Us Conservative?

Another of my mistakes can illustrate the importance of statistical analysis. I have frequently assumed, on the basis of my conversations with church people, that regular churchgoers are more likely than the rest of the population to be conservative on such contentious socio-moral issues as divorce, abortion, and Page 19 of 29

 

The Value of Social Science homosexuality because they are religious. Using data from the 2001 Scottish Social Attitudes survey, Tony Glendinning and I initially found quite strong connections between religious affiliation and frequency of church attendance, on the one hand, and socio-moral attitudes, on the other. Had we stopped there, we could reasonably have concluded that ‘being religious’ was a major determinant of conservative attitudes. We would have been wrong. One of the advantages of turning all of the components of my supposed linkage into numbers and having a lot of people to look at is that we can apply the statistical techniques of multivariate analysis or MVA. In brief, this involves simultaneously comparing the relative contribution of a large number of possible causes to the outcome we wish to explain. MVA demonstrated that age was a better predictor of attitudes than our measures of religiosity: the more pious and the more morally conservative were also the older people in the sample. Being religious and being conservative were both a function of age. This should be obvious, but quantification is sufficiently unpopular that I will spell it out. My conversations over the years have allowed me to know the churchgoing habits and the socio-moral attitudes of, at most, a few hundred people. The Scottish Social Attitudes survey allowed me to know the same things about 2,000 people. And, because that sample of Scots was designed to include men and women, old and young, rural and urban, and so on in proportion to the population at large, those 2,000 respondents were far more representative of Scots as (p.28) a whole than the people I had come to know through my social networks and occasional conversations with strangers. Because we could combine such various measures of religious interest as churchgoing, selfdescription as religious, and assent to various questions about religious beliefs into a single scale of religious commitment, we could, with considerable sophistication, explore the connections between these ‘variables’ and such other characteristics as age, gender, education, and occupation. MVA allowed us to perform an intellectual feat well beyond the capacity of any ethnographer. Second-Order Relationships

One advantage of big data is that they allow us to explore patterns in human behaviour that are too subtle to be seen by the naked eye of the ethnographer. An important version of that assertion concerns second-order relationships. If we suspect that being religious makes people more law-abiding, we could compare the conviction rates of churchgoers and non-churchgoers with otherwise similar social characteristics. But it might well be, as Rodney Stark has argued, that the effects of faith differ depending on whether the individuals in question live in predominantly religious or secular settings. That is, the effects of particular characteristics might vary with other variables (in this case, the extent to which those characteristics are shared with others) in ways that are not at all apparent until we can search for subtle patterns in large-scale data.44

Page 20 of 29

 

The Value of Social Science More examples will be given in the following chapters but these are enough to make the general point: to the extent that we can believe the individuals studied are representative of the studied collectivity, ethnographic work is important for educating us about religious beliefs and behaviour and for understanding (in the Weberian sense) what the world looks like to others, but, if we want to develop and test explanations, we need to generate and test appropriate evidence and that generally requires quantitative social science.

Social Science versus Lay Narratives One major difference between the natural and the social sciences is that the former are the preserve of experts while in some form or (p.29) another the latter are part of the day-to-day lives of ordinary people.45 My mother did not do amateur physics: she used electricity but did not feel obliged to have her own explanation of how it worked. But she did popular sociology all the time. Her response to almost any news item was to devise a general theory about the people or activity being reported. The natural sciences are so esoteric that few people who were not trained subatomic physicists, for example, would feel equipped to engage in debates about photons and bosons. The social sciences deal with the stuff of everyday life and generally use ordinary language. Even our professional jargon tends to use commonplace words. The familiarity of the social sciences cuts both ways. It allows lay people to insult us by claiming that they knew it all anyway: I do not know the source, but some wag is supposed to have dismissed the celebrated 1930s University of Chicago Sociology department’s research on the layout of cities as ‘taking three years to find the red light district’. Our use of familiar terms also allows people with little or no training in social science to enter the field. A very ‘high’ theory of science supposes that, by virtue of their training, scientists have expertise that is denied the laity. I do not believe that graduation with a social science degree is like the ordination of a Catholic priest in conferring magical power, but I do want to suggest an amateur-toprofessional range of competence that is defined partly by acquired attitudes, partly by learned skills and information, and partly by the social organization of disciplines. The professional end of that range should be distinguished by the following. First, we strive for objectivity. It may be impossible entirely to transcend our own interests and prejudices, but that does not mean that we should not try to be disinterested. The sentiment expressed by Robert Solow may now be a cliché, but it remains valid: that it is impossible to achieve a completely aseptic environment does not mean that I would have been as well having my gall bladder removed in a sewer as in a hospital operating theatre.46 Second, we strive for consistency. The Queen in Alice through the Looking Glass may have believed six impossible things before breakfast. My mother believed Page 21 of 29

 

The Value of Social Science six different things most days and was not in the least troubled by her inconsistency. When Wallis and I realized that we wished to claim both that modernization eroded religious belief and that evangelistic activity was higher in nineteenth-century towns and cities than in the countryside, we felt obliged to produce a (p.30) reasonable way of reconciling what Callum Brown took to be a major weakness in the secularization paradigm.47 Third, we are less easily satisfied. My mother was able to do instant sociology because her interest in the enterprise stopped at the point where she could articulate a plausible explanation that fitted her prejudices. As these were many and varied, she usually latched on to explanations pretty easily. Professionals spend weeks, months, and years developing ideas and testing them against the best evidence we can find: for good or ill, I have spent a large part of forty years trying to understand secularization. And a lack of satisfaction can lead to us changing our minds. Of course, professionals can be obstinate in clinging to their treasured explanations but—as my ‘strong religion’ example shows— evidence, plus the need for consistency, can change minds. Thirdly, professionals are trained and practise. The social science disciplines have their perspectives and their techniques. They can be taught and they can be learnt. Subdisciplines also have accumulated knowledge. My mother knew on the basis of a few anecdotes that the Moonies were brainwashed. A serious student of the sociology of conversion would read scores of detailed studies of conversion before, I hope, coming to a very different conclusion. But, and this is where the social structure of academic disciplines becomes important, the validity of social science does not rest on the attitudes or skills or any one practitioner. My critics might be right that I am so wedded to the secularization paradigm that I fail to see signs of religious growth. The enterprise will survive my character defects because there are other scholars reexamining my data and theories and presenting alternatives. And, even if none of my generation change their minds, the next generation will decide to what extent I or others were correct.

Can We Be Objective? Chapter 5 will address the possibility of objectivity at length, but a few words should be said here in this general defence of social science. One element of the critique of positivistic social science deserves special attention because it is attractive and, if correct, fatal to the enterprise. To return to the point already introduced, a wide variety (p.31) of social science perspectives such as postmodernism, feminism, Marxism, and French (as distinct from American) structuralism either argue or assume that the interests of researchers contaminate their work. The problem is obvious: we are all creatures of our environment, socialized into a particular set of beliefs and values that influence

Page 22 of 29

 

The Value of Social Science how we see the world. We are members of social groups with particular interests. Hence we cannot be objective. In the trivial sense this is patently true. However, it need not follow that all research is equally polluted or that the pollution is always a problem. The point about preconceptions derives its power from the unstated assumptions that there are only a small number of such preconceptions and that they are carried by large and well-defined groups. If there were as many preconceptions as there are people, the point would be irrelevant: that such diverse people could agree on research findings would justify us thinking they were more than just another ‘voice’. For it to be a problem for social research, it would have to be the case that people shared their interests in a small number of clusters defined by gender, age cohort, ethnicity, or nationality, for example. Yet it is hardly likely that all or even most of the people of some putative interest group (especially when, in the gender case, it is half the society) really are identical. One critic of the secularization paradigm has asserted that ‘the gender-blind approach which has long characterised the sociology of religion not only ignores gendered difference but assumes the masculine point of view is normative’. She adds: ‘Because they take the male experience to be normative, the narratives of secularisation under discussion assume that modern men and women are affected in exactly the same way.’48 I am not sure what would count as ‘the masculine point of view’ or ‘the male experience’ because I can think of little of interest to sociologists that all men experience in the same way. Even when one were to narrow it down to ‘all manual working men 1850–1900’ and concentrate only on the world of work, it is hard to imagine that the experiences of a Cornish fisherman (whose produce, once landed, was marketed by women, who also managed all onshore parts of the operation) and a Lanarkshire miner (whose produce was handled by the colliery company) would be the same. If we are seriously going to impute interests to people on the basis of their membership of some collectivity, it has to be one defined by actual (as distinct from assumed) similarities. Imputing common interests sufficiently strong to distort social research to classes and (p.32) class fractions seems a little more plausible: we might now divide all people into five or nine interest groups rather than just two. But, even still, it stretches plausibility to suppose that agreed class interests within each of those nine groups will sufficiently transcend other differences to produce a common and mistaken perception of anything other than class. Second, even where we suspect systematic bias, it is important to distinguish sins of omission and sins of commission. Social scientists have indeed neglected some research topics, but it does not follow that what they have had to say about what they did study is therefore mistaken. One could argue that the 1970s interest in new religious movements reflected an unfortunate fondness for studying the exotic over the mundane and led to the neglect of arguably more serious religious currents, but it does not follow that the work on new religious Page 23 of 29

 

The Value of Social Science movements was in itself faulty. Max Weber’s famous thesis explaining the ‘spirit’ of modern rational capitalism as being an inadvertent and accidental consequence of ‘the Protestant ethic’ pays no attention to gender differences, but it is hard to see that the omission in any way weakens his case. If he is right, for example, that the Reformation’s rejection of confession and penance as ways of periodically wiping the slate clean increased ‘salvation anxiety’, it is difficult to imagine that men and women reacted differently, and, even if they did, the puritan culture and behavioural rules that such anxiety produced pressed equally on men and women. It would be interesting to consider whether there were gender differences in what Calvinist men and women took as signs of being predestined to heaven rather than hell—did men favour business success while their wives looked more to the health and happiness of their children?—but there is nothing about the questions that Weber did not ask that invalidates the answers he gave to the ones he did ask. Third, even if bias causes us to make a mistake or even consistently to make a certain type of mistake, it does not follow that all our work is pointless. Given that most scholars study a wide variety of topics in their careers, inherited and reinforced preconceptions may well have little or no impact on some of their work. I may, of course, understand myself less well than some stranger who has never met me, but I really do not think I have any personal interest in the matter of secularization: as a private citizen I do not care whether religion is declining or growing in popularity. The subsidiary case could be made that, though I initially had no dog in that particular race, (p.33) I have now written so much about secularization that my professional standing has become my greyhound. That is possible, but recanting their earlier support for the secularization thesis did not harm the careers of Peter Berger or Rodney Stark. But, even if my lack of faith contaminates my research on secularization, why suppose that it distorts my views on such other topics as the tendency of demotic religions to factionalism and schism, or the deleterious effects on the morale of congregations of meeting in buildings vastly too big for their current numbers, or the reasons why the charismatic movement recruits better in the south of England than in Wales and Scotland? That is, those who would argue that social science is made impossible by our natures have to make the case for each matter in hand and cannot throw male chauvinism, orientalism, imperialism, or any other supposedly single interest on the table as though it was trumps in a scholastic card game. Fourth, sceptics have to answer the empirical case that scholars do often take intellectual positions that apparently clash with their personal desires. As I note in Chapter 5, the religious preferences of social scientists and social statisticians do not map neatly on to their position on the secularization debate: varieties of Christian, atheists, and agnostics can be found on both sides.

Page 24 of 29

 

The Value of Social Science Fifth, we can return to the structure of the profession. That Roger Straus became persuaded of the value of Scientology before he encountered any other Scientologists may explain why he so strongly argued for an explanation of religious conversion that sees it as a consequence of deliberate seeking, but the rest of us who evaluate competing explanations will bring to the issue a variety of other views and, in many cases, a genuinely open mind. Straus’s interests might explain why he advocates a particular theory: they do not determine the critical professional reception of that theory. In brief, while social scientists differ from most natural scientists in studying their own kind, and thus may have reactions to their research matter that are absent from the laboratory, it need not follow that social science is fatally contaminated. I would argue for a more pragmatic attitude. In most cases objectivity is possible because our interests are not strong enough, are not specific enough, and are not sufficiently widely shared to prevent them from being eliminated as a serious obstacle by the normal principles of clear theory formulation and testing in a community of contending and disparately interested scholars. Ernest Gellner delivered the best riposte to Feyerabend’s (p.34) abandonment of method: ‘it is one thing to say that, in a complex and difficult world, almost anything might contain some truth, and therefore it behoves us to be humble; and quite another to say that anything goes and therefore we should be arrogant’.49

Conclusion It is important to repeat the point I made at the start of this chapter. Even if we confine ourselves to academic disciplines, there are very many ways of viewing religion. Were we recruiting staff for a university religious studies department, we would want to hire theologians, philosophers, experts in religious texts and rituals, and even architects, to work alongside historians, psychologists, and sociologists. My claim for social science is the limited one that certain kinds of questions about religion can be answered only by the methods and approaches of social science: if one is interested in the social causes and consequences of religious belief and behaviour, then one has to do social science. I have also made a preliminary case for the value of quantitative social research. While detailed ethnographies have enormous value in helping us get ‘inside’ religious belief and behaviour, they are severely limited by problems of scale and representativeness in their value for generating and testing explanations. In subsequent chapters, the methodological and theoretical issues introduced here will be pursued in greater detail. Notes:

(1.) J. Connell, ‘The Week’, The Week, 15 September 2012. (2.) R. Hill, ‘Flings’, London Review of Books, 21 February 2013.

Page 25 of 29

 

The Value of Social Science (3.) A. Palmer, Moveable Feasts: Fluctuations in Mealtimes (Oxford: Oxford University Press, 1984). (4.) I also agree with Abram’s reluctance to isolate sociology from history: ‘in terms of their fundamental preoccupations, history and sociology are … the same thing. Both seek to understand the puzzle of human agency and both seek to do so in terms of the puzzle of human structuring. Both are impelled to conceive of that process chronologically … Sociology must be concerned with eventuation, because that is how structuring happens. History must be theoretical, because that is how structuring is apprehended. History has no privileged access to the empirical evidence relevant to the common explanatory project. And sociology has no privileged theoretical access’ (P. Abrams, Historical Sociology (Ithaca, NY: Cornell University Press, 1982), pp. x–xi). (5.) The project in question is ‘Christianity and the University Experience in Contemporary England’. Results are from the project website— (accessed February 2011). Credit is due to Peter Brierley, who, although an evangelical Christian who should have been pleased by the result, spotted its implausibility; see P. Brierley, ‘Student Religion’, Future First (April 2011), 3. (6.) For one of the most effective rejections of this bifurcation, see the work of Anselm Strauss: A. Strauss, Qualitative Analysis for Social Scientist (Cambridge: Cambridge University Press, 1987). (7.) That British academia is generally short of quantitative skills has been recognized by the British Academy (the learned society that is to the humanities and the social sciences what the Royal Society is to natural science): anon, ‘Count Us In’, British Academy Review, 26 (2015), 17–19. (8.) W. G. Runciman, A Treatise on Social Theory, i (Cambridge: Cambridge University Press, 1983), 1–56. (9.) Key protagonists were Wilhelm Windelband, Wilhelm Dilthey, and Heinrich Rickert, and one major outcome was Max Weber’s distinctive approach to methods: W. J. Cahnman, ‘Max Weber and the Methodological Controversy in the Social Sciences’, in W. J. Cahnman and A. Boskoff (eds), Sociology and History: Theory and Research (Glencoe, IL: Free Press, 1964), 103–29. (10.) H. Blumer, Studies in Symbolic Interactionism (Englewood Cliffs, NJ: Prentice-Hall, 1969). (11.) For an excellent discussion of twelve possible meanings of positivism, see P. Halfpenny, Positivism and Sociology: Explaining Social Life (London: Routledge, 1982).

Page 26 of 29

 

The Value of Social Science (12.) This list is paraphrased from that given in I. Hacking, Scientific Revolutions (New York: Oxford University Press, 1999). (13.) G. Lenski, ‘Positivism’s Future—and Sociology’s’, Canadian Journal of Sociology, 16 (1991), 187–95; S. Andreski, Social Science as Sorcery (Harmondsworth: Penguin, 1975). (14.) I. Hewett, ‘Songs in the Key of Life’, Times Review, 9 November 2013. (15.) For a discussion of the Marxist theory of praxis and its epistemological consequences, see A. Feenberg, The Philosophy of Praxis: Marx, Lukács and the Frankfurt School (London: Verso, 2014). (16.) See, e.g., L. Stanley and S. Wise, Breaking Out: Feminist Consciousness and Feminist Research (London: Routledge, 1983). (17.) A. Crockett, ‘A Secularising Geography? Patterns and Processes of Religious Change in England and Wales, 1676–1851’, Leicester: University of Leicester, Ph.D. thesis, 1998. (18.) K. Popper, Conjectures and Refutations (New York: Basic Books, 1962). (19.) E. Durkheim, The Rules of Sociological Method (New York: Free Press, 1964). (20.) This is the ‘strong programme’ in the social studies of science, best exemplified by the work of Harry Collins: H. M. Collins, Changing Order: Replication and Induction in Scientific Practice (London and Beverly Hills: Sage Publications, 1985). (21.) P. Feyerabend, Against Method (London: New Left Books, 1975). (22.) T. Kuhn, The Structure of Scientific Revolutions: Fiftieth Anniversary Edition (Chicago: University of Chicago Press, 2012). (23.) Beyond assuming a minimal degree of cross-cultural comprehension about common concerns, I am deliberately avoiding trying to settle the intractable problem of translatability, especially with regard to ritual action. For a comprehensive review of options, see the contributions to B. R. Wilson (ed.), Rationality (Oxford: Blackwell, 1970), and S. P. Turner, ‘Translating Ritual Beliefs’, Philosophy of the Social Sciences, 9 (1979), 401–24. (24.) E. Durkheim, Suicide: A Study in Sociology (London: Routledge Kegan Paul, 1970); M. Weber, The Protestant Ethic and the Spirit of Capitalism (London: George Allen and Unwin, 1976). (25.) W. I. Thomas and D. Thomas, The Child in America: Behavior Problems and Programs (New York: Knopf, 1928), 572. Page 27 of 29

 

The Value of Social Science (26.) M. Sahlins, ‘Human Science’, London Review of Books, 9 May 2013, 29. (27.) This is Roy Wallis’s theory of factionalism: see R. Wallis, ‘A Theory of Propensity to Schism’, in R. Wallis, Salvation and Protest: Studies of Social and Religious Movements (London: Frances Pinter, 1979), 174–92. (28.) The figure was arrived at by my colleague Tony Glendinning analysing the responses to questions on current religion and religion of upbringing in the 2001 Scottish Social Attitudes survey. (29.) K. Popper, Conjectures and Refutations (London: Routledge and Kegan Paul, 1963). (30.) T. Luhrmann, Persuasions of the Witch’s Craft: Ritual Magic in Contemporary England (Cambridge, MA: Harvard University Press, 1991). (31.) H. Blumer, Critiques of Research in the Social Sciences: An Appraisal of Thomas and Znaniecki’s ‘The Polish Peasant in Europe and America’ (New Brunswick, NJ: Transaction Books, 1979). (32.) For a detailed and incisive discussion of what Goldthorpe calls the problem of variation, see J. Goldthorpe, On Sociology (Oxford: Oxford University Press, 2000), ch. 4. (33.) D. Kemp, The New Age: A Guide (Edinburgh: Edinburgh University Press, 2003). (34.) M. York, ‘Alternative Spirituality in Europe: Amsterdam, Aup and Bath’, in S. Sutcliffe and M. Bowman (eds), Beyond New Age: Exploring Alternative Spirituality (Edinburgh: Edinburgh University Press, 2003), 131. (35.) S. Sutcliffe, ‘Unfinished Business—Devolving Scotland/Devolving Religion’, in S. Coleman and P. Collins (eds), Religion, Identity and Change: Perspectives on Global Trans-formations (Aldershot: Ashgate, 2004), 91. (36.) Sutcliffe, ‘Unfinished Business’, 88. (37.) P. Heelas and L. Woodhead, The Spiritual Revolution (Oxford: Blackwell, 2004). (38.) G. Davie, Religion in Britain since 1945 (Oxford: Blackwell, 1994). (39.) M. Percy, ‘Losing our Space, Finding our Place? The Changing Identity of the English Parish Church’, in S. Coleman and P. Collins (eds), Religion, Identity and Change: Perspectives on Global Transformations (Aldershot: Ashgate, 2004), 32.

Page 28 of 29

 

The Value of Social Science (40.) Nat Cen, ‘British Social Attitudes: Record Number of Brits with no Religion’; http://www.natcen.ac.uk/news-media/press-releases/2017/september/ british-social-attitudes-record-number-of-brits-with-no-religion/ (accessed September 2017). (41.) A. Crockett and D. Voas, ‘Generations of Decline: Religious Change in Twentieth-Century Britain’, Journal for the Scientific Study of Religion, 45 (2006), 567–84. (42.) D. Kelley, Why the Conservative Churches are Growing (New York: Harper & Row, 1972). (43.) M. A. Hout, A. Greeley, and M. Wilde, ‘The Demographic Imperative in Religious Change in the US’, American Journal of Sociology, 107 (2001), 468– 500. (44.) R. Stark, ‘Religion as Context: Hellfire and Delinquency One More Time’, Sociology of Religion, 57 (1996), 163–73. (45.) One interesting feature of the ‘democratization’ implicit in the internet is that people do seem increasingly willing to assert strongly held views about scientific matters of which they actually know nothing. For a poignant example, see the public responses to the July 2017 court battle between the parents of terminally ill baby Charlie Gard and the doctors treating him. (46.) R. Solow, The Interpretation of Cultures (New York: Basic Books, 1973), 30. (47.) R. Wallis and S. Bruce, ‘Secularization: The Orthodox Model’, in S. Bruce (ed.), Religion and Modernization: Sociologists and Historians Debate the Secularization Thesis (Oxford: Oxford University Press, 1992), 8–30; C. Brown, ‘Did Urbanism Secularize Britain?’, Urban History Yearbook (1988), 1–14. (48.) L. Woodhead, ‘Gendering Secularization Theory’, Kvinder, Kãn og Forskning, 1 (2005), 1–25, (accessed 17 November 2009). (49.) E. Gellner, Spectacles and Predicaments: Essays in Social Theory (Cambridge: Cambridge University Press, 1979), 188.

Page 29 of 29

 

Defining Religion

Researching Religion: Why We Need Social Science Steve Bruce

Print publication date: 2018 Print ISBN-13: 9780198786580 Published to Oxford Scholarship Online: August 2018 DOI: 10.1093/oso/9780198786580.001.0001

Defining Religion Steve Bruce

DOI:10.1093/oso/9780198786580.003.0003

Abstract and Keywords We cannot explain if we cannot identify what needs explaining. This chapter challenges the idea that religion is particularly difficult to define. It demonstrates that what are often called functional definitions are actually assertions about possible consequences of religious belief and argues for a substantive definition that accords with what most people think of as religion: systems of beliefs, behaviour, and organizations built on the assumption of a divine being or beings with the power of moral judgement. It also considers what we mean by spirituality and by the secular. The links between the secular, secularism, secularity, and secularization are clarified. Keywords:   defining religion, functional definitions, spirituality, the secular, substantive definitions, secularization, secularity, secularism, secularist

Introduction As the serious study of any phenomenon requires that we know what we are to study, we must have some interest in definition, but students of religion make rather heavy weather of identifying their business. I am not thinking here of Edward Tylor’s many ways of defining religion in Primitive Culture.1 He at least concludes his lengthy review of the alternatives by choosing one. What I have in mind are the various postmodern approaches that argue that there is actually no such thing as religion because religion is a modern social construct and one usually constructed for bad purposes. Tim Fitzgerald, for example, argues that religion as an idea is a consequence of the modern separation of church and state (with some influence from the earlier separation of science from religion).2 Talal Asad finds an older and wider set of ancestors for his Genealogies of Page 1 of 28

 

Defining Religion Religion but also concludes that religion is a construction of European modernity: ‘there can be no universal definition of religion, not only because its constitutional elements and relationships are historically specific, but because that definition is itself the historical product of discursive processes’.3 Worse yet, the definition ‘authorizes … particular forms of history-making’ that are morally bad.4 As with most eye-catching novelties, a sensible point is made absurd by exaggeration. First, the origins or uses of a concept have no necessary bearing on the reality it purports to comprehend because discovery is not the same as invention.5 Isaac Newtown’s discovery of gravity was a ‘historical product of discursive processes’, as Asad would put it, but prior to its discovery people had no trouble sticking to the earth’s surface: Newton conceptualized gravity but he did not invent it. (p.40) Second, the uses to which some idea is put does not exhaust the idea and even demonstrating that an idea has been put to bad ends does not of itself prove that the idea is badly conceived and should be discarded. Third, the postmodern critique of the idea of religion applies equally well to every other concept and definition we use, including those used by Fitzgerald and Asad in their postmodern critique. In a fine demonstration of a failure to understand the nature of language, the first page of a British government study of religious beliefs starkly announces ‘the actual term “religion” is an invented or constructed category’, as if there was some contrasting category of nouns for social processes and institutions that, like rain and snow, are naturally occurring.6 Fitzgerald wants to abandon the study of ‘religion’ in favour of the study of the supposedly material political and social relationships that are putatively disguised by the idea of religion. But to identify what supposedly lies behind religion itself requires definitions and concepts that will be equally susceptible to the same dismissal by family tree and discourse.7 As it offers no rational scientific grounds for arbitrating between competing discourses or voices, if applied rigorously, the postmodern critique (like all other forms of relativism) permits only cacophony or silence. A related general problem is apparently identified by Malory Nye when he argues that to define religion by nouns implies too static a view of what is always changing and shifting. I am not sure we need concede his premise. It is true that religious organizations and traditions always change, but such changes tend to be slow enough so that, at any particular time, analysts have no more trouble distinguishing an Anglican from a Baptist than do Anglicans and Baptists, and most of them manage to attend the right building on a Sunday. But, even if we were prone to an overly static view of our business, I fail to see how Nye’s

Page 2 of 28

 

Defining Religion preferred solution—replacing nouns by verbs—helps anything. Replacing ‘religion’ by ‘religioning’ adds only ugliness. The postmodern relativist will, of course, reject the realism that underpins my three criticisms, but this is one of those incommensurable divides about which one can do nothing but takes sides. There is a very different approach to defining religion that the social scientist should also reject: separating true from false religion. It is common for religious people to make a partisan distinction, and they can do it in opposing directions: the term ‘religion’ may be claimed as an honorific or rejected as an insult. Some believers insist (p.41) that what they believe about the supernatural is religion because it is true and that all competing faiths are superstition. Some neatly reverse that approach: what they believe is just the truth and what everyone else has is mere religion. Emile Durkheim curiously committed himself to a version of this when he wrote: ‘If then we wish to discover the true nature of religion, we must observe it at the zenith of its evolution; it is in the most refined forms of Christianity and not in the puerile magic of the Australian aborigine or the Iroquois that we must expect to find the elements of the definition we are seeking.’8 I see no sociological value in such partisanship because, while objective reality may be relevant for understanding why some plan of action fails to achieve its goals, as I noted in Chapter 3, what usually matters in explaining human action is what actors believe to be the case. Since the mid-twentieth century, most sociologists of religion have attempted to avoid judging the truth claims of competing religions by adopting a posture of methodological agnosticism. What matters for understanding Mormons is not whether we believe that the Angel Moroni delivered the Book of Mormon to Joseph Smith but whether Joseph Smith and his followers believed this to be the case. In his determination to be unimpressed by protestations of religious neutrality, Fitzgerald has argued that, by adopting methodological agnosticism, the academic study of religion has actually endorsed a particular theological position: a sort of wet liberal ecumenism in which all religions are vaguely true. Fitzgerald may be right that the Unitarian–Universalist brand of Christianity has an attitude to other religions that is close to the sociological principle of methodological agnosticism, but that is not the fault of sociology. He may also be right that the scholars of religious studies he directly criticizes (such as Ninian Smart) were both methodological agnosticists and Unitarian–Universalists, but again that is not the fault of sociology. The religious studies version of methodological agnosticism may, as Fitzgerald argues, suppose that all proper religions are in some sense true, but what sociologists such as Peter Berger mean by the practice is importantly different.9 It does not implicitly or otherwise support any particular theology because it treats all beliefs as human projections. Some may also, of course, be true, but claims about the divine are beyond empirical testing by the discipline’s methods and hence beyond our Page 3 of 28

 

Defining Religion remit. For example, it is always possible (as a devout Catholic might argue) that the fragmentation of Protestantism is God’s punishment for heresy, but, as social scientists (p.42) have no way of testing that assertion, we will continue to explore the contrasting organizational consequences of, on the one hand, having a single final point of access to God’s will and, on the other, permitting that all right-minded believers can comprehend the will of God. Our raw material consists of the beliefs and actions of identifiable people. Which (if any) of them accurately grasps what (if anything) may lie beyond the material world is none of our business. It is worth stressing (because many students of religion see their subject as unique) that methodological agnosticism is not a sociological response to religion in particular but is applied to beliefs in general. One good reason for it was discussed in Chapter 2: if we wish to explain why people act as they do, the explanation will take the form of identifying which of their (not our) beliefs were brought into play in interpreting their (not our) perceptions of the circumstances in which they acted or reacted. The other good reason is that our appreciation of the role of social construction has taken us beyond the initial problematic of the sociology of knowledge. Karl Mannheim’s famous Ideology and Utopia and his preceding work divided knowledge into truth and ideology and offered sociological explanations of only the latter.10 As in the early sociology of science, it was assumed that correct beliefs required no explanation: only error needed to be explained. Since the 1960s, sociologists have generally treated all knowledge as sociologically explicable. Of course, that a particular set of ideas is effective is one good reason why people may be persuaded to accept them, but in many fields (responses to quack medical therapies, for example) efficacy is often largely a matter of socially constructed shared perceptions. That this seems to lock us into a circle is sometimes a problem, but it is sufficient for my purposes here to establish that most modern sociologists (and not just sociologists of religion) routinely bracket the truth value of the claims made by people whose behaviour they are trying to understand. Much of the difficulty in defining religion comes from arguments about which of a largely agreed set of characteristics should be definitive. The particular version of this that we know as functionalism will be discussed shortly, but I will first identify a variety of other examples. Because it is widely assumed that people (severally or jointly) have certain enduring needs that are met by religion, it is common in descriptions of largely secular societies for a wide variety of secular beliefs and activities to be re-badged as religions. For (p.43) example, football may be described as a religion because fans may be extremely committed, may find that attending matches gives them an effervescent or liminal experience similar in some respects to the collective ecstasy found in shared religious activities, and may treat football grounds as sites of pilgrimage.11 Various forms of music and dance culture have similarly been Page 4 of 28

 

Defining Religion treated as religious on the grounds that they share some features with religion more conventionally defined.12 My difficulty with such an approach is that it obscures more than it illuminates. Defining football as a religion discourages a detailed consideration of the differences between sport and religion and achieves by fiat what should be established by empirical demonstration. Broadening the notion of religion to include anything that shares any of its features runs counter to one of the key purposes of definition: to isolate the distinctive features of the phenomena that interest us. We can see the problem in a November 2009 British court judgment. Tim Nicholson was sacked by Grainger PLC after refusing to obey a legitimate instruction. His employer, who had travelled from London to Dublin without his Blackberry mobile phone, instructed the employee to fly to Dublin to deliver the phone. The employee refused on the grounds that the gain to his boss of being rapidly reunited with his means of communication was patently less than the damage to the environment of undertaking an unnecessary plane journey. The employee wished to use the legal protections given to religious beliefs as grounds for seeking compensation for having been sacked. He argued that his environmentalist principles were sufficiently important and sufficiently strongly held to be the equivalent of a faith. Mr Justice Burton accepted his argument and allowed him to bring the case. The obvious difficulty with that judgment is that, rather than clarifying what distinguishes religion from any other ideas or sources of inspiration that a plaintiff may claim are precious, anything taken very seriously is now a religion.13 The above identifies one of the key problems in defining social phenomena: because they are often complex, we have to select which of a range of characteristics we shall regard as definitive, and we may disagree about this. There is a second related but analytically separable problem, which I will now address. In addition to a range of potentially definitive characteristics (viewed in a static sense) phenomena are often attended by a range of typical causes and (p.44) consequences, and we may wish to incorporate some of these in the definition of the thing itself.

Functional Definition Two of the founders of sociology are conventionally described as having defined religion in terms of its supposed consequences.14 In his Critique of Hegel’s Philosophy of Right Karl Marx said: ‘religion is the sigh of the oppressed creature, the heart of a heartless world, just as it is the spirit of a spiritless situation. It is the opium of the people’.15 The upper classes use religion to oppress the lower orders, and the lower orders use it to console themselves. Durkheim’s functional definition of religion makes its capacity to unite all people in a common consciousness the definitive feature. Some confusion arises because it is not at all clear that the supposed consequence or function is

Page 5 of 28

 

Defining Religion uniquely associated with religion or religion with it. That is, they may not be coterminus. The Marxist view of religion is co-terminus with its function read in one direction. Once the real source of oppression has been removed by socialism, there will be no pain that requires the opiate of religion, and religion will end. But the same does not work in the other direction. Presumably Marxists allow that there are forms of ideological obfuscation other than religion (bourgeois nationalism, for example) and other forms of rebellion-dampening ‘opiate’ (for example, non-metaphorical opiates). Thus, while we can say that Marxists think religion is a thing that has the consequence of quietening the masses, it is not strictly the case that such a consequence uniquely defines religion. This is even more so the case for Durkheim’s definition of religion, because it is not a necessary part of the definition that religion will disappear if social cohesion vanishes. Social diversity will produce cultural diversity, which will, in turn, weaken religion, but for a few generations competing religions will remain extant. Nor is it necessary that only religion can serve the purpose of providing social cohesion. That is, there are functional equivalents of religion. We do not need to go too far down this path to appreciate that what are commonly described as functional definitions of religion are actually assertions about the origins or common consequences (p.45) of religion. It is, of course, possible to be entirely rigorous about functional definition and assert that by ‘religion’ we will mean whatever provides either ideological obfuscation or social cohesion. But, if followed strictly, that practice would add nothing at all to our understanding of either obfuscation or cohesion: it would simply be a renaming. Wherever we saw social cohesion we would assert that the thing that produced it was religion (even if, for example, it was something quite different, such as the external threat of violence). That such renaming would be little of explanatory value is inadvertently recognized by those who purport to use a functional definition of religion in that they do not present purely tautological statements. Instead they identify function separately from religion as such. Robert Bellah, in Tokugawa Religion, says: ‘It is one of the social functions of religion to provide a meaningful set of ultimate values on which the morality of a society can be based.’16 If religion is not just the functions it serves, it must be identifiable separately from those functions, and it is usually identified by conventional substantive definitions. Jack Goody says of Talcott Parsons and Bellah: ‘it is perhaps significant that in their pragmatic treatment of religious phenomena the above authors adhere much more closely to the “traditional” sphere of discourse’.17

Page 6 of 28

 

Defining Religion For a truly functional definition of religion we can look at what Edward Bailey calls ‘implicit religion’. In his manifesto for a research centre to study it, he offered three definitive features of implicit religion: 1. Commitment. … it directs our initial attention towards the empirical human experience (of being committed), which has the merit of placing the ‘content’ or ‘object’ of the experience, within the wider context, of experience-ing. It is concerned, you might say, with ‘religion’ as such, before turning to its ‘theological’ component. 2. The second definition is integrating foci. Just as Commitment covers the whole ‘ladder’ of levels of consciousness, so integrating foci covers the various ‘sizes’ of society, the individual and the social, the face-to-face group and the societal … 3. Intensive concerns with extensive effects.18 Bailey appears to be saying that commitment, social belonging, and feeling strongly about something are actually religious (even though in most instances of those things the people involved think they are secular). (p.46) The obvious problem with this conceptualization is that it is very hard to think of anything that would be excluded. Depicting gardening as religious seems rather pointless. It does not help us understand religion or gardening, and, paradoxically, it hinders functional analysis. It is perfectly proper to be interested in the purposes or effects of some activity, attitude, or institution, and the notion of ‘functional equivalent’ is an important one. We learn a great deal about social evolution and social mutation if we compare, for example, the different ways that societies allocate resources. But, in order to demonstrate that some social institution has the functions imputed to it, we have to be able to identify the institution independent of those supposed functions. If not, we are merely renaming. To say that any social activity (for Bailey, ‘integrating foci’) is implicitly religious tells us no more about the world than does ‘this three-sided object is a triangle’. Actually it tells us less. Because ‘three-sided object’ is specific, we at least still know what a triangle is; with Bailey’s renaming we no longer know what religion is. If, instead of renaming parts of the world (and without warrant imputing social functions to them), we try to develop testable empirical claims about functional equivalents, we must begin with substantive differences. To say that football may share some features of a religion is to raise interesting research questions about the resources fans devote to football, the psychological states they attach to success and failure, the attitudes of some fans to the arenas in which their teams play, their capacity for unfounded optimism (and conversely their ability to construe historic victim myths), and so on. Put in those terms, the identification

Page 7 of 28

 

Defining Religion of parallels also raises the corresponding question of differences. Football fans typically: • do not have shared rituals that guarantee divine approval; • do not marry only those who support the same team; • do not shun those who do not share their allegiance; • do not suffer debilitating trauma if they lose faith in their team; • do not construct elaborate moral and social codes around their footballing interests; • do not found political parties to advance a divinely ordained agenda; • do not construct creation myths; and • do not destroy the God-offending stadia of competing teams. (p.47) Constructing such a list is educational in that it immediately suggests crucial differences between football and religion. For example, football requires competing teams. As Celtic fans discovered in 2012 when Rangers, the other half of the Glasgow pairing in what some Victorian wag correctly dubbed ‘the Old Firm’, was relegated to the third division for financial improprieties, the absence of credible competition removes much of the pleasure of winning trophies. Games are stylized conflicts that require that the losing teams survive to play a return fixture. Such would not have been the preference of the medieval Crusaders or of modern Islamists; they sought or seek to eradicate the competition. Football is also confined in time and space. Football fan behaviour is genuinely ‘carnivalesque’ in the way that extreme commitment is toggled on and off: fans may behave like prats shortly before, at, and shortly after the match, but they are expected to return to the normal world. And there is a wellunderstood audience for fan behaviour. It is fine to insult the fans of opposing teams; it is not fine to insult the civilians you pass on the street or with whom you share a train. Arguably religion in secular societies is similarly confined to the leisure world of the weekend and to those who willingly accept it, but most religious people regret that marginalization and in principle reject it. To borrow a 1970s advertising slogan intended to discourage casual pet purchases, Christianity, like a dog, ‘is not just for Christmas’. Most significantly, football allegiance has never been the basis for distinctive claims about the origins or maintenance of the world. Even for diehard Barcelona fans, ‘Lionel Messi is God’ is a metaphor; for true believers, ‘God is God’ is not. In brief, examining the parallels between football and religion can be interesting and illuminating, but that task is not helped at all by defining football as a religion. To do so is to establish by definition what should be demonstrated factually. The notion of implicit religion seems particularly pointless. If there are no limits placed on the word ‘implicit’ by a more traditional definition of religion, then everything that is not overtly religious is implicitly religious, and fat people are implicitly thin and short people are implicitly tall. Rejecting the concept of implicit religion should not be confused with ignoring the very important point Page 8 of 28

 

Defining Religion that phenomena can be religious (in the conventional sense) to varying degrees. At one extreme we have the self-consciously religious: praying, attending worship services, reading sacred texts, contemplating the life of the Buddha, and so on. At the other extreme we can imagine actions and (p.48) states of mind that are influenced slightly by religious ideas and motifs, and we can even imagine that the people in question may be largely unaware of those religious influences. For example, a father may take pride in the effort he puts into his parenting role and implicitly suppose that there is some obscure karmic order to the universe that means that good actions will be rewarded even where there is no obvious this-worldly mechanism for such a result and where the reality often seems otherwise. Someone who supposes that a cheat will ‘get his comeuppance’ may be relying on an unconsidered or implicit hope of a supernatural order. If the speaker really supposes that, all this-worldly evidence to the contrary, bad people will always be punished, we may reasonably say that such a view is implicitly religious. However, we should also want to distinguish such views from thoroughly secular alternatives. It is quite possible to have a secular model of the long-term rewards of good parenting or the long-term penalties for cheating. In summary: there are great difficulties with functional definitions. First, as we see by the way that proponents either fall back on studying beliefs and institutions that are religious in the more obvious substantive sense or treat everything as religious, they are difficult to apply consistently or usefully. Second, they assume the very thing that needs to be explored: just what functions does this or that religion perform in this or that setting? Third, they pretty well rule out the possibility of change in the popularity of religion and, despite all the evidence of our eyes, they rule out the stable but largely secular state of Denmark.19 If functional definitions are better understood as assertions about consequences that are predicated upon conventional substantive identifications of religion, we have a good case for arguing that the substantive approach is of more value, even in the matter of identifying the uses and consequences of religion. The clear distinction between the identification of the phenomenon and claims about its consequences offers protection against tautology. Because we are no longer asserting them by definition, we appreciate that we need to examine the effects of religion. We also bring back into focus important issues that are disguised by a functional definition, such as whether and why religions differ in their capacity to serve the function in question. Most importantly, we permit the possibility that a society can have more or less religion and thus allow the possibility of secularization.

Page 9 of 28

 

Defining Religion (p.49) Substantive Definition For reasons I will explain shortly, a definition that fits with broad contemporary common-sense reflection on the matter is usually not a bad place to start. Moreover, the utility of a definition must in the end depend upon the success of the explanations in which it is employed. That is, the purpose of a definition is to bring together analytically similar phenomena, aspects of which we believe we can explain in the same terms. I define religion substantively because this allows me to formulate a number of theories that I believe have considerable explanatory scope. Religion, then, consists of beliefs, actions, and institutions that assume the existence of supernatural entities with powers of action, or impersonal powers or processes possessed of moral purpose. Gods are an example of the former; the Hindu principle of karma is an example of the latter. Such a formulation seems to encompass what most people mean when they talk of religion. There are, of course, objections to such a definition. As a pre-emptive rebuttal of those such as Richard King who reject such a definition on grounds derived from Edward Said’s critique of orientalism, I see nothing in this definition that confines it to the West and to theistic salvation religions or that insults Hinduism and Buddhism.20 Leaving aside for a moment the fact that most Hindus and Buddhists do actually worship deities, this definition seems to fit perfectly well even the most philosophical brands of Hinduism and Buddhism. Some critics of substantive definitions regard the fact that people argue about definitions (and, in particular, contest their own place in any classificatory system) as invalidating any particular definition. This does not seem a pressing problem. I say more about when analysts should argue with their subjects shortly, but that particular groups and individuals contest and manipulate definitions does not of itself vitiate attempts at classification. For example, dictators and their dissident critics argue over whether a country deserves to be described as a democracy, but that does not prevent political scientists defining democracy (and its alternatives) or using the notion in social scientific description and explanation, and nor should it. In an interesting reversal of Fitzgerald’s claim that popular definitions of religion inadvertently endorse a particular kind of religion, André Droogers argues that substantive definitions of religion are problematic for the opposite reason: because they assume that religion (p.50) is false.21 His logic is that religion (or other crucial elements of a substantive definition such as the sacred) is always one half of a contrast pair: in this case religion-and-science. And, as social scientists work within the paradigm of science, those who study religion effectively brand their subject matter as false. One can see the problem but not accept the conclusion. Although it will not satisfy adherents of any particular religion (who will insist that their faith be exempted from phenomenological bracketing), the attitude of methodological agnosticism seems to resolve this Page 10 of 28

 

Defining Religion difficulty. I do not see how social scientific analysis is hampered by treating all beliefs as being susceptible of social scientific explanation. There is nothing in the disciplinary armoury of the sociologist that allows me to know whether televangelist Pat Robertson was right to regard the Haitian earthquake of 2009 as divine punishment for voodoo, but, for my purposes of understanding US fundamentalists, it is enough that Robertson believed this to be the case. The most common objection to substantive definitions of religion is a watereddown version of Asad’s assertion about the historical specificity: there is just too much variety. It is certainly true that a universally applicable ahistorical definition of religion (and of anything else in the non-material world) is impossible. However, for those who are not complete relativists, there are three sensible responses. The first is empirical. Just how much common content is there in what peoples in different times and places have viewed as religious? Provided one is content with a minimalist definition, there seems a great deal of common ground. The ancient Greek warrior making offerings to his Gods before battle, the medieval knight promising to build a monastery if God gives him victory in his crusade, and the modern Protestant businessman praying for God to bless his new factory are engaged in recognizably similar activities and, if time travel ever allowed them to meet, we could expect them to be mutually intelligible. Similarly, we can imagine the superstitious British fishermen discussed in Chapter 10 having little difficulty explaining their rituals of reassurance to Bronislaw Malinowski’s Trobriand Islanders, and vice versa.22 The second response is to argue that there is nothing particularly wrong with a very broad conceptualization of religion, provided it is accompanied by a wellinformed sense of the variety of forms it can take. Not all conceptual work needs to be done in a single master (p.51) definition. On the basis of detailed ethnographic studies of a number of contemporary Christian congregations, Martin Stringer argues that ordinary believers may have superstitious, disordered, pragmatic, and immanent beliefs that accord little with the coherent, transcendent, and transformative view of Christianity found in official teachings and creedal statements.23 He takes this to mean we should change our definition of religion. I take it to mean only that we should be aware that members of religious organizations can differ in their appreciation, acceptance, and conformity to the teachings of their organizations. The third response is pragmatic. A detailed universally applicable definition of religion is clearly necessary if one aims to produce a theory of the origins or purposes of religion that will apply to all times and places. But, despite the criticisms of ‘sweeping generalization’ sometimes levelled at sociology, most of us have the humbler ambition of producing low-level generalizations about this or that sort of religion in this or that time and place, and for such purposes a Page 11 of 28

 

Defining Religion definition that covers a decent part of the human record—for example, Europe in the last ten centuries—seems perfectly viable.

Defining Spirituality Religion is relatively easy to identify because, although a private religion is not ruled out by definition, the socio-logic dictates that only those systems supported by significant numbers of people (or, in pre-democratic societies, numbers of significant people) survive long enough to trouble the analyst. Defining spirituality is much more difficult because, as I will argue, what is usually meant rests heavily on the subjective and on the amorphous. My reading of both actors’ responses and commentators’ glosses is that the term ‘spirituality’ may be applied to at least four analytically separable things. First, it may simply represent the subjective component of religion and be a modern synonym for what the Victorians called piety. This is what we have when our survey respondents claim to attend church frequently, assert entirely conventional Christian beliefs, and describe themselves as both religious and spiritual. In this sense, ‘spiritual’ signifies that they are really serious about their religion, and it may also carry an implied criticism of some others whose religious behaviour is mere conformity. (p.52) Second, it may represent a claim to sensitivity. Describing urban working-class culture in the north of England in the 1940s, Richard Hoggart perceptively made the point that ‘Christian’ was often used as a way of asserting general virtue; it was a synonym for ethical. ‘I’m as good a Christian as you, though I don’t go to church,’ they say. With that often goes the implied reversal: ‘You’re as bad as I am, even though you do’. The regular churchgoer, it is inferred, may be less virtuous than some of those who never attend. If he is very regular and something of a figure at church, he may well be a bit of a hypocrite—whereas the man who makes no pretensions, but does his best, is probably much nearer being a Christian. After all, doing your best to be an ‘ordinary decent’ person—that is what Christianity means, really.24 My guess is that, with the decline of Christianity, much of that usage has passed to the term ‘spiritual’ but with a new twist. Hoggart’s non-churchgoers wished to avoid being labelled immoral and to claim respectability. Current use of ‘spiritual’ seems intended more to fend off accusations of insensitivity and shallowness. Increased personal liberty seems to have made people less concerned about conforming to externally imposed moral codes but created a new concern for possibly critical judgements of our personalities. Being spiritual may be claimed against an implied alternative of being thought dull, insensitive, beastly, and unromantic.

Page 12 of 28

 

Defining Religion Third, ‘spirituality’ may be used by analysts to designate a middle ground between the conventionally religious and the avowedly secular. This is certainly done by critics of the secularization thesis who first postulate complete secularity as the state required to fulfil the secularization thesis and then claim those who are neither conventionally religious nor paid-up members of a rationalist association as spiritual. Consider the evolution of the common survey question on belief in God. From the 1950s to the 1990s Gallup and other polling organizations usually asked people to choose between variants of rows 1, 2, 4 and 5 Table 3.1: There is a personal God; There is some sort of spirit or life force; I don’t really know what to think; and I don’t really think there is any sort of God, spirit or life force. The first of those is classic theism: the expected Christian response. Row 4 is intended to capture agnosticism and row 5 atheism. Quite what the second category is capturing in a non-Hindu or Buddhist population is not clear, but something very interesting happened when Gordon Heald of Opinion Research Business (ORB), who had previously worked for Gallup, added the row 3 option: There is something there. Table 3.1. Belief in God, Great Britain, 1990 and 2000 (%) Survey questions

1990

2000

There is a personal God

32

26

There is some sort of spirit or life force

41

21

Not asked

23

I don’t really know what to think

15

12

I don’t really think there is any sort of God, spirit or life force

10

15

There is something there

None of these Total

1 99

3 100

Sources: 1990 Gallup; 2000 data provided by Gordon Heald of ORB.

Page 13 of 28

 

Defining Religion

Table 3.2. Belief in God and self-description as religious, spiritual, or neither, Scotland, 2001 (%) Survey questions

Religious

Spiritual

Neither

Can’t Choose

Total

There is a personal God

60

15

21

3

99

There is some sort of spirit or life force

33

31

32

4

100

There is something there

31

8

57

4

100

I don’t really know what to think

8

2

85

5

100

I don’t really think there is any sort of God, spirit or life force

5

3

90

2

100

None of these

6

4

87

2

99

Source: 2001 Scottish Social Attitudes survey (n = 1,598).

Page 14 of 28

Defining Religion (p.53) As the respondents in the two surveys are not the same and the polls are ten years apart, we cannot be sure that the differences in responses are caused by a change in the measurement instrument (rather than representing, for example, real change in beliefs or enduring differences between two populations). But the fact that so many people in 2000 chose ‘There is something there’ suggests that many of those who chose ‘higher power’ in 1990 were not actually committed adherents to some notion of cosmic consciousness but were non-religious people who were reluctant to commit themselves to agnosticism or atheism. Support for that view can be found in data from a Scottish survey. Table 3.2 shows in its rows the responses to the standard Gallup/ORB questions and in its columns the division of respondents according to how they answered the question ‘Whether you attend religious (p.54) services or not, would you say you were religious, spiritual or neither?’25 As we can see, 60 per cent of those who claimed to believe in a personal God also described themselves as religious. Those who said they believed in some sort of spirit or life force spread equally across religious, spiritual, and neither. But consider those who said ‘There is something there’. Only 8 per cent called themselves ‘spiritual’ and more than half said ‘neither’. This suggests that the ‘There is something there’ option represents something closer to the procedural point of agnosticism than to nontheistic spirituality. It reflects an unwillingness to commit or a lack of interest in the issue. And, to work back a stage, it certainly calls into question the high levels of spirituality inferred from surveys that offer only four options and take the ‘life force or higher power’ respondents to be spiritual. Thus far I have suggested three usages of spirituality that are largely distractions from our purposes in constructing a useful definition: as a synonym for piety, as a synonym for sensitivity, and as a usually inappropriate way of enrolling the undecided or uninterested in the ‘really religious’ camp. At the risk of sounding like a stuck record, I will now repeat the point I made about defining religion. If, for a moment, we stop making programmatic assertions about the difficulty of conceptualizing our subject matter and look at what we do, we find among analysts and ordinary people a considerable amount of agreement. I suggest that we can readily identify spirituality (of the sort that interests those of us who study the changing fate of religion in the modern world) by three features. Its ideological core is a belief in some sort of supernatural force or entity that differs from that of conventional religion in having no location outside the self, except in some vague notion of an all-pervasive cosmic consciousness. A phrase common in the New Age centre at Findhorn—‘coming into your power’— nicely captures the sense that, rather than allowing access to something external, enlightenment makes you aware of what you already have. Its second component is the idea that believing in the spirit changes how one sees and feels about the world: by becoming more environmentally sensitive, for example. Its third component is the idea that becoming aware of our spiritual nature should Page 15 of 28

 

Defining Religion make us better people. These three features fit perfectly well with what Paul Heelas and Linda Woodhead mean by spirituality: ‘subjective-life forms of the sacred, which emphasize inner sources of significance and authority and the cultivation or sacralization of unique subjective-lives’ and which they contrast with more conventionally (p.55) religious conceptions ‘which emphasize a transcendent source of significance and authority to which individuals must conform at the expense of the cultivation of their unique subjective-lives’.26

Should Sociology Have itsown Language? We can now turn from common problems in defining religion to the specific question of a sociological definition of religion. Chemists do not need to consider the relationship between their concepts and the concepts used by their subject matter. The elements are unaware of the characteristics that are used to construct the Periodic Table; the Transition Metals do not complain that they are every bit as good as the Noble Gases. The subjects of the social sciences have consciousness, and they use the same language as the analysts. Our work depends upon understanding our subjects, which requires that we understand their language. Hence actors’ and analysts’ terms are inevitably tangled. It is hard to think of any concept used in mainstream sociology that is not in common usage. A sociologist who studies ‘the family’ may also invite his or her family for Christmas. The sociologist of crime may also be a criminal. As analysts we sometimes use concepts (social network, for example, or social capital) that are rare in the bus queue, but even then we suppose that we could, if required, translate these into terms that ordinary people would recognize. We all understand deprivation. The notion of relative deprivation—feeling deprived when one compares oneself to someone else, to one’s past, or even to one’s expectations—may take the lay man or woman a moment’s thought, but it is perfectly intelligible. On the further fringes of the discipline there are technical terms that would be alien to most people: adjacency pairs, pre-sequencing, and repair sequences are examples from conversation analysis. But even these terms are readily translatable into lay language. Indeed, the ultimate check that one had, for example, correctly identified ‘Do you want to see a film?’ as an ‘invitation’ requiring acceptance or decline rather than as a ‘question’ requiring a yes or no answer would be to ask the speaker his or her intentions. There is no mystery about how sociological concepts develop. We shift between collecting examples of our subject’s talk about what (p.56) interests us, refining that varied usage into a simplified and systematic formulation, and applying this again to lay usage. And we work backwards and forwards between observation and explanation. Sociological concept formation differs from its popular cousin in three ways. First, the professional formulation is generally abstracted from a wider range of usage than its lay counterparts. Because of our interest in generalization and Page 16 of 28

 

Defining Religion comparison, the sociological definition of the family, for example, is likely to be informed by knowledge of a far greater range of instances than is that used by lay people when they consider their own, and other known, families. Second, sociological concepts are generally broader than lay concepts because they are designed for purposes more abstract and enduring than those that motivate the lay person and they are refined in arguments between large numbers of scholars from diverse backgrounds. An ethnomethodologist colleague of mine found a fine example of a clash between professional and lay distinctions while working in a mental institution in the 1970s. The psychiatrists classified the patients according to elaborate diagnostic categories. The nursing orderlies classified their charges as either ‘wetters’ or ‘wanderers’, depending on whether a patient’s incontinence or disorientation caused most work. Third, sociological concepts will generally have an orderliness not required in everyday life. For example, whatever one thinks of the utility of the classic church/sect/denomination/cult typology of religious organizations, because it was developed by a large number of scholars trying to identify the causal relations between features of the religious life with the greatest explanatory power, it has far greater internal consistency than the dichotomy of ‘churches’ and ‘dangerous cults’ common in the mass media.

Arguing with our Subjects Sociology’s reliance on lay language does not commit us to accepting actors’ accounts in any or every particular instance. Having refined our usage, we stick to it, even if it means arguing with our respondents. Consider the notion of social class. This is now conventionally described in a limited number of ways, based largely on the extent of power, authority, freedom, responsibility, training, and expertise involved in occupations. When we ask Scots and English people their (p.57) social class, we find a systematic variation in the difference between how respondents describe themselves and how analysts classify them. Scots who are, in the observer’s sense, middle class are more likely than their like-situated English counterparts to describe themselves as working class. When they do so, we do not accept that valuation and report that the Scots are more likely than the English to be working class. Instead we argue with our subjects about their self-description. We disagree with them and we try to explain why they get it wrong and we think we have an explanation: identifying strongly with the working class creates a distinctive political culture that is objectively well suited to Scottish conditions and that allows Scots to think of themselves as different from, and superior to, the English. Strictly speaking, the above is an argument not about definition but about identification. The Scots who mis-describe themselves share the same definition of class as the researcher; indeed, any rhetorical benefits of the misidentification rest on the definition of working class as involving manual labour. Page 17 of 28

 

Defining Religion The mis-identifying Scots want to claim that they are manual workers in spirit, if not in body. Apart from the example given earlier of some believers wishing to restrict the term ‘religion’ to themselves or to apply it only to others, I can think of few cases where researchers are likely to argue with their respondents about definitions at such an abstract level. A more common problem is respondents’ dislike for the terms we use at the next level down: for example, church, sect, denomination, and cult. And even here the point at issue is not usually fit with the specifics of technical terms but dislike for their wider connotations. If we simply numbered the types 1, 2, 3, and 4, reluctance to accept the designation ‘sect’ and ‘cult’ may well diminish, because what upsets sectarians and cultists is not the detailed description of their characteristics but the implications in everyday usage of fanatic, extremist, and false believer. The first reason why researchers cannot cede to their research subjects the right to define key terms is that our research has to be directed to what best tests and advances our tentative explanations of what interests us. I have argued that the church, sect, denomination, and cult typology has considerable explanatory value as a chronological commentary on the changing status of religion under modernization. Various aspects of modernization (in particular individual liberty and egalitarianism) render impossible the compulsory church type of religion, where a single body represents an entire people to God and vice versa, and instead encourages the formation of (p.58) competing voluntary sects. The diversity thus created encourages the state to become increasingly neutral and forces both church and sect to become increasingly denominational. This attempt to identify and explain crucial changes in the history of Christianity in the West by applying what was originally devised as a rather static model of organizational types may be illuminating or it may be nonsense. But it cannot be tested if the various groups that it attempts to understand themselves determine which label best describes them. They will all be ‘churches’. We have the professional sociological proposition: modernization makes the church type of religion impossible. And we have the lay rebuttal: ‘No it doesn’t. We are all churches.’ Actually the reality is worse: what almost all our subjects will say is: ‘Our group is a church. Our competitors are sects or cults.’ Put more generally, definition and conceptualization have to be a disinterested professional activity because those we study may be ill informed, partisan, or both. Even when they are well informed about themselves, they may well have highly skewed and partial views of their competitors. Put even more generally, this preference for professional over lay designations is a necessary response to a lack of consensus among those we study. If we are researching anything bigger than a very small group, we are not going to find a

Page 18 of 28

 

Defining Religion degree of consensus that allows us to cede definition and conceptualization to our subjects. Thus far I have been concerned mainly with rather grand abstractions such as religion and class. There is one sort of definition where the Fitzgerald and Asad elision of discovery and invention is proper. Although we may reasonably think that ‘religion’ is a timeless concern of human life—or at least on pragmatic grounds we would be wise to avoid seeking its origins—we cannot object to the claim that individual religions are products of particular social circumstances. We know when Mormonism began and who founded it and we can thus try to explain its origins. We know when Methodism was founded and who founded it. However, such statements may be contested by believers using precisely the distinction between discovery and invention that I used earlier. Analysts say that Mormonism was invented by Joseph Smith. Mormons say that Smith only ‘discovered’ Mormonism; he did not invent it. Smith was the messenger of a divine revelation rather than the deviser of a human creation. It tells us much about how religion is justified that very few of the creators of new religions actually claim authorship; either it already (p.59) existed in the mind of God and its earthly progenitor has simply taken delivery, or it previously existed on earth but had become buried under layers of human additions, and, like the art restorer removing layers of Victorian varnish from an old oil painting, the discoverer has simply revealed what had been lost. When the Tibetan Buddhist Chogyam Trungpa first published Cutting through Spiritual Materialism, he conventionally described his teachings as the channelled wisdom of a previous incarnation; only decades later did he claim authorship.27 The one clear example of claimed and copyrighted invention that comes to mind is Scientology. L. Ron Hubbard deliberately changed his secular psychotherapy of Dianetics into the religion of Scientology because he was offended by Dianetics practitioners editing and augmenting his invention and he wished to reclaim it. Given that most followers of any particular tradition or church or sect will want both to acknowledge its founder’s work and to claim the greater legitimation that comes from divine authorship or an ancient history, analysts may often argue with believers about the origins of their faith. We will assert that we are not disputing believers’ assertions because, following Berger’s lead, we simply report rather than endorse what believers say, but this may not placate the believers, who suspect that, by not endorsing their claims, we are implicitly rejecting them. They are right. There is no way of avoiding the fact that to explain the origins of Mormonism in terms of the social circumstances that saw a proliferation of new religions in the nineteenth century north-east of the United States is pretty much to argue with Mormons about where their faith came from.

Page 19 of 28

 

Defining Religion Arguing with our subjects about definitions of particular religions is a commonplace part of social analysis. It is obvious that this must be the case for social scientific concepts such as social class. Although we base the definition on the realities of people’s lives, once we have refined it, we apply it even when it involves classifying as middle class a white-collar professional worker who describes himself as working class. What may be less obvious is that we also have to do this with concepts that are more clearly the property of our subjects. As the beliefs of Joseph Smith and Brigham Young and other Mormon leaders define Mormonism, the Mormon Church has the first punt at deciding who is a Mormon, but we can imagine others (for example, the tiny Fundamentalist Church of Jesus Christ of Latter-Day Saints led by jailed polygamist Warren Jeffs) claiming to be Mormons. This may not matter for most research purposes, but suppose, like Rodney (p.60) Stark trying to explain why some types of religion recruit better than others, we want to make testable claims about the growth of Mormonism. We need to count the number of Mormons. That requires that we, the analysts, decide how to apply a definition that begins with the subjects’ definition. Even where some concept is initially defined by our research subjects, we may have to arbitrate inconsistent and competing claims for identification. This may seem like nit-picking, but consider how we define a Christian. While recognizing variation, we could broadly say that Christians are people who either believe the statements of the Apostles’ Creed or the Nicene Creed or who respect such creeds as being important historical statements of what their churches used to believe, which still, in some vague sense, represent the essence of Christianity.28 For some matters we can leave it to the Christian churches to decide who meets their requirements, but for many research purposes we need to count the number of Christians in a society (or rely on the work of others who have made such a count). That may be for social policy reasons: we may wish to decide what sort of privileges (such as the right to publicly funded schools or chaplain posts in the armed forces) should be accorded the Christian churches. It may be for social scientific explanatory reasons: we can make no progress at all in explaining secularization if we cannot determine roughly its onset and trajectory and in European societies until very recently that means counting Christians. But, as we saw with the 2001 and 2011 censuses in the UK, people who patently fail to meet the requirements of the Christian churches may nonetheless claim the designation. In practice we work around the problem by devising new labels (‘nominal Christian’ or ‘heritage Christian’, for example) and we study why and how people choose labels for occasions and purposes. But we have to recognize that even the style of research that privileges understanding operates at a remove from its subjects. Unlike those of the natural sciences, few of our concepts are ours alone. Notions such as ‘social class’, ‘family’, and ‘religion’ have to have some consistent relationship to the world of the people we study or we will be wasting our time. This is more so the case for such notions as ‘Methodist’ and ‘Scientologist’, which involve Page 20 of 28

 

Defining Religion some element of belonging. In the first instance these will be defined by those we study, but, having agreed on a particular conceptualization, we try to use it consistently, even if it means arguing with some of our subjects. This is not the analyst deciding who is a real Christian. It is the analyst (p.61) saying that, having taken the statements of those whom most Christians authorize to decide who is a Christian as the basis for our definition, we implement it consistently and that will sometimes require that we argue with some people about their right to the designation. For completeness, it is worth mentioning that analysts may well need to correct some accidentally misleading designations. For example, the Office of National Statistics helpfully collates data on the frequency of religiously sanctified and secular weddings, but, because they are not celebrated in a government Registry Office, humanist ceremonies are included in the lists of the religious even though they are usually chosen precisely because they are not religious. In analysing the political correlates of religious identities in Northern Ireland, Ian McAllister and Bernadette Hayes label the non-religious as ‘religious independents’ and thus may inadvertently create a false impression of Ulster’s religiosity.29 A third example of such designations perhaps involves, if not a deliberate intent to mislead, then at least a subconscious display of religious bias. Sylvia Collins-Mayo et al.’s study of young people’s religiosity actually found very little but blurred that fact by calling a variety of largely secular concerns ‘immanent faith’ and referring to the ‘holy trinity’ of self, family, and friends.30

Defining the Secular While we are defining religion, we should consider the secular. Some conservatively religious people regard the ‘secular’ as the mirror image of religion. In the USA the Christian Right tried to overturn increasingly secularist interpretations of the constitutionally required separation of church and state by arguing that secularism was itself a religion and hence that it was unfair to constrain government approval of religion while allowing the secular to triumph.31 In the UK, conservative Christians have argued that the strong commitment to secularism of scholars such as Richard Dawkins is itself ‘religious’ because it involves undue psychological investment and anything that people feel very strongly about is a religion. We should reject both these partisan word games. The religious and the secular are not mirror images; they are comprehensive alternatives, like smoking and not smoking. One describes a particular activity or attitude and the other describes nothing more than the (p.62) absence of that activity or attitude. In the UK at least, avowed secularists have always been thin on the ground. Although secularization has increased the social space available to them, such people are its symptom and not its cause. The primary driver of secularization is

Page 21 of 28

 

Defining Religion the increase in the proportion of the population that is ignorant of, and indifferent to, religion. In brief, the secular is simply the absence of the religious. Secularization is the social process by which features of modernization have inadvertently weakened religion. Secularity is the consequence of secularization. Secularism is something very different: it is the deliberate promotion of secular social arrangements. And the secularist is someone who promotes secularism.

Definitions and Quantitative Methods As one of the purposes of this book is to advertise the value of quantitative research, it is worth pointing out that statistical analysis of large-scale data sets allows a useful technique for testing the validity of definitions: factor analysis. We can search for the degree of connection between different indices of the general phenomenon that interests us. Of course, the qualitative researcher does this too, and Anselm Strauss’s excellent Qualitative Research Methods offers a variety of suggestions as to how this can be done, but the mental juggling involved in trying to assess the strength of connection between four or five variables is beyond most of us. More will be said about this in Chapter 4; here I want to make the point that whether our definition of some phenomenon has any reasonable fit with the thinking and behaviour of our real-world subjects can actually be tested statistically in rather fine detail. We can ask survey respondents if they think of themselves as religious and we can ask them a variety of ancillary questions that allow them to elaborate their understanding of religion and that test to what extent they conform to the definition implied in their responses to such questions. We then test the strength of connection between the various responses. The bright qualitative researcher might be able to do this for fifty people: Richard Hoggart’s account of what religion meant to his working-class neighbours in Hunslett from which I have quoted is an excellent (p.63) example. But quantification allows the responses of thousands of people to be analysed for coherence and cohesion. The 1991 British Household Panel Survey data show considerable consistency in answers to a variety of religion questions. The sample splits with 40 per cent saying ‘No’ and 60 per cent saying ‘Yes’ in response to the question of whether they had a religious identity. Those responses mapped closely onto responses to the proposition that religion was, or was not, of personal significance. Nobody who said that religion was of no significance claimed to be religious and correspondingly no one who said they were religious was on the negative side of the personal significance array. There was a similar correspondence with church attendance: those who never or almost never attended church also said that religion was of no personal significance, while the regular churchgoers, not surprisingly, thought their religion was personally significant. The same correspondence was found with a church membership question.32 The details are less important than this simple observation: a variety of measures of religious interest match up in the ways we would expect, which, unless it is a Page 22 of 28

 

Defining Religion remarkable coincidence, means that our survey questions, and the concepts embedded in them, generally reflect consistent understandings of, and orientations to, religion. That is: our terms are not meaningless. Fitzgerald may think there is really no such thing as religion. Ordinary people disagree with him.

A Pragmatic Justification for Relaxed Usage This brings me to my final theme: a practical defence of our current practice. Herbert Blumer argued that, contra those who think the social sciences can imitate the conceptual accuracy of the natural sciences, the best we could do is to create ‘sensitizing concepts’ that give ‘the user a general sense of reference and guidance in approaching empirical instances’.33 Blumer sees this as a weakness. I do not. If one takes the view that what matters about definitions is utility, a defence of my relaxed attitude to defining religion is that it does not prevent us doing good social science. To illustrate the point, I will briefly mention some debates in the sociology of religion with which I am familiar. (p.64) Since the 1980s, Rodney Stark and his colleagues have elaborated the supply-side or rational-choice model of religious behavior, which I discuss in more detail in Chapter 9. The relative merits of this model and the classic secularization thesis have been argued at tedious length, and numerous tests have been devised and presented. The disagreements are so throughgoing that commentators have referred to a ‘paradigm shift’ (to borrow a term used by Thomas Kuhn in his account of scientific revolutions).34 From the point of view of this chapter’s theme, what is interesting is that none of the arguments advanced in this extensive conflict depended on definitions of religion. The same is true for competing explanations for the growth of Pentecostalism in Latin America and Africa. A clear divide can be discerned between those who explain its popularity in terms of Pentecostalism’s usefulness for those attracted to it and those who stress the supposed political interests of those who promote it. That story can be told in functional terms—whose interests are being served? —or in terms of plausibility—to what extent does Pentecostalism derive legitimation from its association with US power and prestige?—but those arguments do not depend on us arguing about whether Pentecostalism is really a religion. One controversy that, at first sight, does seem to hinge on definitions is the case made by Grace Davie that many Europeans who are not active participants in organized religion are nonetheless ‘vicariously religious’. She believes that, with varying degrees of consciousness, the apparently religiously indifferent appreciate other people doing religion on their behalf. I find this unconvincing, but the problem does not lie with definition.35 I have no difficulty at all with the idea of vicarious religion. We see it in the common medieval Christian practice of Page 23 of 28

 

Defining Religion paying others to say masses for one’s soul postmortem. We see it in the Thai Buddhist division of labour that sees lay people feeding the monks in return for the monks earning transferable religious merit on their behalf. What is at issue for understanding religion in the modern West is whether, for example, we should regard avowedly non-religious Britons who watched the funeral service for Princess Diana as being vicariously religious. In two small-scale community studies I agreed with Davie’s empirical observation that the clergy continue to perform important roles for non-believers when, for example, they act as spokesmen for a population that has suddenly become newsworthy and organize public demonstrations of sentiment: my examples concern the foot-and-mouth cattle disease that seriously damaged a Devon farming community in (p.65) 2001 and the murders committed by Derek Byrd in Cumbria in 2010.36 My conclusion is that in these and related cases the clergy are invited to perform important community functions for purely secular reasons: they form a body of professionals that is trained in public speaking and organizing large events, that has an obligation to the public at large, and that is not party political. Most residents of the villages in which Byrd randomly murdered the strangers he encountered turned out for his victims’ church funerals, but that does not mean they were displaying vicarious religion. It means that the funerals, though they were quasi-religious events (the obviously religious was toned down by the officiating priest), were treated by non-believers as an opportunity to show communal solidarity. But whether Davie or I am right, my point here is that the argument does not hinge on the definition or conceptualization of vicarious religion; what is at issue is the accuracy of the identification of her examples. My general point is the simple one that in these and in many other arguments within the sociology of religion, definition is rarely at issue. We are, quite properly, arguing about evidence. There is always a danger in justifying some course of action by its apparent results: readers may not be as impressed as I am by the value of these debates. Like the Marxists of the 1960s who derided the arguments between Parsonian structural-functionalists and symbolic interactionists as trivial because both groups of scholars were suffering from false consciousness, scholars who fret about the definition of religion may regard the illustrations I have offered as proof that a relaxed attitude to identifying our subject matter produces pointless work. Again I can only offer a pragmatic rebuttal. Do we learn more about religion in the modern world from Fitzgerald’s arguments about definitions or from empirical responses to Stark’s analysis of religious markets? Or, to separate the two parts of Stringer’s Contemporary Western Ethnography and the Definition of Religion, is the value of his detailed descriptions of what his respondents believe enhanced at all by his attempts to change how we define religion? I would argue not.

Page 24 of 28

 

Defining Religion Conclusion In reviewing common arguments about definitions, I have made the following points. Philosophically minded scholars spend too much time agonizing over the correct way to define religion (if they permit (p.66) that such a thing exists). While it is quite proper to be interested in the functions of religion, defining religion in terms of its supposed uses or consequences is unhelpful. It is also unworkable, as is demonstrated by the fact that supposed functionalists actually fall back on substantive definitions. As there is little hope of (or need to) develop sociological concepts that do not rest on common usage, a substantive definition of religion that fits pretty well with the way ordinary people use these terms seems entirely adequate for social scientists who do not aspire to grand ‘theories of’ religion. It is actually not difficult to define religion, spirituality, or the secular provided we do not expect such master abstractions to do all our conceptual work. I have also argued that very few arguments in recent social science studies of religion have hinged on competing definitions of our subject matter. There is a further pragmatic justification for defining religion by its substantive core. We study ordinary people. We ask them about their beliefs, attitudes, activities, and commitments. We cannot usefully ask them if they adhere to ‘some source of answers to questions of ultimate significance’. We can ask them: ‘Are you religious?’ Our definitions and concepts have to fit with the way our research subjects think about these things. And when we test large bodies of data, we find considerable overlap between diverse expressions of religious interest and sentiment. Where we differ from lay people is in consistency. Having derived our concepts from common usage, we apply them in an evenhanded fashion, even when that means arguing with those we study. As we will see in Chapter 4, measuring and scaling personal religiosity, for example, is sometimes tricky, but it is no more difficult than measuring levels of education or social class or levels of health and well-being. Provided we hold consistently to the principle of methodological agnosticism and are not drawn into debates about the truth or falsity of religion, religion is no more or less difficult to define than any other social phenomenon. Notes:

(1.) E. B. Tylor, Primitive Religion, 2 vols (London: John Murray, 1871). (2.) T. Fitzgerald, The Ideology of Religious Studies (Oxford: Oxford University Press, 2000). (3.) T. Asad, Genealogies of Religion: Discipline and Reasons of Power in Christianity and Islam (Baltimore: Johns Hopkins University Press, 1993), 9. (4.) Asad, Genealogies, back cover.

Page 25 of 28

 

Defining Religion (5.) I am grateful to my colleague Robert Segal for pointing this out. (6.) Home Office, Religion in England and Wales: Findings from the 2001 Home Office Citizenship Survey (London: Home Office, 2004), 1. (7.) M. Nye, ‘Religion, Post-Religionism, and Religioning: Religious Studies and Contemporary Cultural Debates’, Method and Theory in the Study of Religion, 12 (2004), 447–76. (8.) W. Pickering (ed.), Durkheim on Religion: A Selection of Readings with Bibliographies (London: Routledge and Kegan Paul, 1975), 75–6. (9.) P. L. Berger, The Sacred Canopy (Garden City, NY: Doubleday, 1967), 180–1. (10.) K. Mannheim, Ideology and Utopia (London: Routledge, 1936). (11.) D. Hervieu-Léger, Religion as a Chain of Memory (New Brunswick, NJ: Rutgers University Press, 2000), 104–7. (12.) G. Lynch and E. Badger, ‘The Mainstream Post-Rave Club Scene as a Secondary Institution: A British Perspective’, Culture and Religion, 7 (2004), 27– 40; R. Sylvan, Traces of the Spirit: The Religious Dimensions of Popular Music (New York: New York University Press, 2002). (13.) S. Adams and L. Gray, ‘Climate Change Belief Given Same Legal Status as Religion’, Daily Telegraph, 3 November 2009. (14.) For an insightful contrast of substantive and functional definitions, see P. L. Berger, ‘Some Second Thoughts on Substantive versus Functional Definitions’, Journal for the Scientific Study of Religion, 13 (1974), 125–33. (15.) K. Marx, Critique of Hegel’s Philosophy of Right (Cambridge: Cambridge University Press, 1970), 17. (16.) R. Bellah, Tokugawa Religion (Glencoe, IL: Free Press, 1957), 6. (17.) J. Goody, ‘Religion and Ritual: The Definitional Problem’, British Journal of Sociology, 12 (1961), 154. (18.) E. Bailey, ‘Implicit Religion: What Might It Be’. Inaugural chair lecture, Middlesex University, November 1997. (19.) Anthropologist Timothy Jenkins, for example, uses his definition of religion as a form of collective self-imagining and self-worship ‘which presents visions of social flourishing; of right social order and of what it is to be human’ to make religious anything more thoughtful than tile-grouting: T. Jenkins, ‘Congregational Cultures and the Boundaries of Identity’, in M. Guest, K. Tusting, and L.

Page 26 of 28

 

Defining Religion Woodhead (eds), Congregational Studies in the UK: Christianity in a PostChristian Context (Aldershot: Ashgate, 2004), 36. (20.) R. King, Orientalism and Religion: Post-Colonial Theory, India and the Mystic East (London: Routledge, 1999); E. Said, Orientalism: Western Conceptions of the Orient (Harmondsworth: Penguin, 1995). (21.) A. Droogers, ‘Defining Religion: A Social Science Approach’, in P. Clarke (ed.), The Oxford Handbook of the Sociology of Religion (Oxford: Oxford University Press, 2009), 263–79. (22.) B. Malinowski, Argonauts of the Western Pacific (New York: Dutton, 1922). (23.) M. Stringer, Contemporary Western Ethnography and the Definition of Religion (London: Continuum, 2008). (24.) R. Hoggart, The Uses of Literacy (London: Chatto and Windus, 1957), 98–9. (25.) As always, I am grateful to my former colleague Tony Glendinning for the statistical analysis. (26.) P. Heelas and L. Woodhead, The Spiritual Revolution (Oxford: Blackwell, 2004), 6–7. (27.) D. J. Mukpo, Dragon Thunder: My Life with Chogyam Trungpa (Boston and London: Shambala, 2008), 222–4. (28.) Precisely this shift in attitude to creeds is found in the ‘declaratory’ act passed by the Free Church of Scotland in 1892. This ‘re-purposed’ the Westminster Confession of Faith (the classic statement of Calvinist Protestantism) from being a currently applicable condensation of what the true Christian should believe to being an inspirational historical document to be interpreted in the light of present knowledge and circumstances. (29.) I. McAllister and B. Hayes, ‘Religious Independents in Northern Ireland: Origins, Attitudes and Significance’, Review of Religious Research, 37 (1995), 65–83. (30.) S. Collins-Mayo, B. Mayo, S. Nash, and C. Cocksworth, The Faith of Generation Y (London: Church House Publishing, 2006), 32–3. (31.) For example, October 1986’s Smith v. Alabama Board of Education. Helped on his way to a functionalist definition of religion by sociologist James Davidson Hunter’s expert testimony, Judge Brevard Hand found that secularism was a religion. Hand’s decision was overturned on appeal: R. D. Rubin, ‘The New Christian Right and the Death of Secularism as Neutrality in the United States’, Journal for the Study of Religion and Ideologies, 13 (2006), (accessed May 2016). (32.) Tony Glendinning deserves the credit for this analysis of the 1991 BHPS data on religion. (33.) H. Blumer. Symbolic Interaction (Englewood Cliffs, NJ: Prentice-Hall, 1969), 148. (34.) R. S. Warner, ‘Work in Progress toward a New Paradigm for the Sociological Study of Religion in the United States’, American Journal of Sociology, 98 (1993), 1044–93; T. Kuhn, The Structure of Scientific Revolutions: 50th Anniversary Edition (Chicago: University of Chicago Press, 2012). (35.) G. R. C. Davie, The Sociology of Religion (London: Sage, 2007). Here as in so many other places I am grateful to David Voas for his comments. See S. Bruce and D. Voas, ‘Vicarious Religion: A Critique’, Journal of Contemporary Religion, 25 (2010), 243–59. (36.) S. Bruce, ‘A Sociology Classic Revisited: Religion in Gosforth’, Rural Theology, 10 (2012), 39–49, and ‘Religion in Ashworthy 1958–2011: A Sociology Classic Revisited’, Rural Theology, 11 (2013), 92–102.

Page 28 of 28

 

Measuring Religion

Researching Religion: Why We Need Social Science Steve Bruce

Print publication date: 2018 Print ISBN-13: 9780198786580 Published to Oxford Scholarship Online: August 2018 DOI: 10.1093/oso/9780198786580.001.0001

Measuring Religion Steve Bruce

DOI:10.1093/oso/9780198786580.003.0004

Abstract and Keywords Because the empirical study of contemporary religion is dominated by ethnographic case studies, this chapter argues for the importance of the statistical analysis of big data. It begins by arguing that description in words involves counting every bit as much as description in numbers. It explains different sorts of scales. It makes the case that counting the Godly is no more difficult than measuring any other sort of social characteristic and considers the relative merits of various ways of doing it. It details both the virtues and vices of quantitative analysis. Particular attention is given to the vital problem of representativeness: how far can we generalize from the particular people or groups we study. Keywords:   ubiquity of counting, scales, representativeness, quantitative analysis, qualitative analysis, quantitative research, qualitative research

Introduction Although E. P. Thompson was one of the more sociologically aware of English historians, he derided social statistics as ‘the mumbo-jumbo of those latter-day astrologers … who for 200 years have been trying to persuade us that nothing is real that cannot be counted’.1 Whether quantitative social research should be dismissed as mumbo-jumbo depends on who or what Thompson has in mind, but he is wrong to scorn the proposition that ‘nothing is real that cannot be counted’. Or at least a good case can be made for saying that serious study of aggregate human behaviour requires measurement or something very like it. In this chapter I will argue that counting, either to oneself or out loud, is an

Page 1 of 32

 

Measuring Religion essential and inescapable part of comparative observation. I will also explain and defend the sort of counting that goes on in quantitative social research. This issue has particular salience in the study of religion, because there is a strong tendency for those who come to it from an arts or humanities background to suppose that there is something uniquely ineffable about religious commitment. At a seminar in a then-distinguished Religious Studies department, I presented statistical data to show that, however generously one measured it, it was impossible to create an estimate for the popularity of New Age or holistic milieu spirituality that brought it anywhere close to the losses to the Christian churches. Alternative spirituality is interesting as a source of insights into the culture of those who inhabit that world where health and well-being intersect with fringe religion: typically middle-aged female arts graduates working in the caring professions. But it cannot be presented as evidence for either the grand claim that all people are essential religious or for the lesser claim that the (p. 71) New Age is taking up the slack left by the decline of mainstream Christianity and thus leaving our religious economy in the steady state required by the rational choice or supply-side critics of secularization who are discussed in Chapter 9. To general approval, a postgraduate student whose work was predicated on the assumption that alternative spirituality is both popular and growing in popularity responded to my data by saying: ‘Well you can’t really measure whether people are spiritual or not.’ The obvious riposte, which had clearly not occurred to the student, is that such scepticism ruled out her optimistic evaluation of the New Age every bit as effectively as it challenged my pessimistic assessment. She wanted to be sceptical of my measurements but not of her own implicit and vague assessments. So that we are clear about the importance of measuring religious belief and behaviour, it should be stated that without such measures we can say almost nothing of any value about the popularity of any particular form of religion or of religion in general or make any assertions about the appeal of certain kinds of religion to certain sorts of people. As a preliminary it is worth noting that some observations require little or no actual measurement because the differences that concern us are obvious. I do not need to measure the Great Pyramid of Giza and my house to know that the pyramid is bigger than my house. Likewise, we do not need to measure the power of the Christian churches now and in the Middle Ages to be confident that the latter is much greater than the former. However, many differences are not so obvious; if, for example, we want to claim that alternative spirituality is more popular now than it was in 1960, then we do need measurement. To avoid confusion, I will separate what we might make of the data produced by our measurements from the largely practical matter of how and what we can measure. What we might infer from, for example, church attendance will be discussed at the end of the chapter. I will begin with a defence of measurement.

Page 2 of 32

 

Measuring Religion The Ubiquity of Counting Quantification is often discussed as though it were an option: you can observe things and describe them in words or you can count things (p.72) and describe them in numbers. Actually counting is not optional. Detailed ethnographies, textual analysis, or the analysis of lengthy interviews may involve less (and less obvious) quantification than the analysis of social surveys, but they involve it nonetheless. Matthew Wood’s excellent study of New Agers in Nottingham makes the entirely plausible observation that one of the appeals of the shared perspectives that united the group is the absence of an external constraining authority.2 He came to his observation by listening to his people talk about their attitudes to being bossed about and concluding that more of them are in favour of individual autonomy than are in favour of external imposition. That is, he has implicitly counted expressions of sentiment in one direction and in another and concluded that one lot outweighs the other sufficiently to present a certain view as a common characteristic. He does not present an explicit count, but the implicit weighing of the evidence is obvious. Max Weber does not present any statistical evidence (arguably, he offers very little evidence at all) that the Puritans suffered greater salvation anxiety than their Catholic predecessors, but we and he assume that, if push comes to shove, he could so do. Furthermore, observation without comparison is of little value. Indeed, one could argue that observation is almost always comparative even if the comparison is not overt. When I observe the blue tit eating from the ball of nuts and lard hanging outside my window, I can perhaps describe its chest colour without comparing it to anything else (although I am likely to be thinking that the blue tit is yellower than the coal tit), but I cannot usefully describe its size except by comparing it to the other birds. But we can go further and argue that social observation, if it is to be the basis for explanation, must be comparative because, unless we can compare one thing to another, we cannot identify what is unusual about it that requires or points to an explanation. And when we compare, we are almost always, implicitly or explicitly, counting the occurrence of this or that. To say, for example, that one of the main distinctions between Protestant and Catholic Christianity is the relative reliance of the former on the Bible and the latter on authoritative pronouncement from the church hierarchy is to rely on implicit measures of importance. How can we know that Protestants place more emphasis than Catholics on the Bible? We could count the number of copies of the Bible printed in Protestant and Catholic countries, count the frequency of family Bible reading, count the number of times that (p.73) Protestant and Catholic clergy defend their beliefs by reference to biblical texts (as against, for example, reference to papal encyclicals), and so on. Of course, frequency is not the only evidence of importance. One may know that I love my children because I talk about them a lot (the cosmopolitan style) or because the very few words I say about them express profound feeling (the style of the Aberdeenshire peasant). But even in purely linguistic description there is an Page 3 of 32

 

Measuring Religion implied counting. We do not need to do this because we are linguistically competent, but we could lay out words for affection on a scale from hate, dislike, do not care, do not mind, like, and love and note that there is a single property here that runs from strongly negative to strongly positive and that we can read as ‘more or less’. More or less is measurement. A good example of how neglecting this obvious fact can lead to misrepresentation is found in a paper by Boaz Huss entitled ‘Spirituality: The Emergence of a New Cultural Category and its Challenge to the Religious and the Secular’.3 His case is that the ‘emergence’ of spirituality (which can only mean there is now more of it than there once was) makes redundant the old religious/secular dichotomy (which must mean that there is now a non-trivial amount of spirituality). So we have two assertions about the size of a social phenomenon. But the paper offers no evidence for these two empirical claims. Instead his presentation is literary description with a few illustrations and no attempt to demonstrate their representativeness. To appreciate that some sort of ordering and counting underlies most forms of description is not a justification for using numbers inappropriately, and I see no value in pretending to an impossible exactitude. C. Wright Mills famously mocked the pretensions of what he called ‘abstracted empiricism’.4 He was right to be sceptical of so-called mathematical sociology that translates words into letters and sentences into equations, but his own work was replete with empirical claims based on implicit ordering and counting.

Scaling The value of statistical analysis and the sort of statistical analysis that is appropriate depend on the sort of data collected. Some limits can be seen clearly if we consider four different sorts of scale: nominal, (p.74) ordinal, interval, and ratio. A nominal scale is not really a scale at all: it is simply a sorting of cases into piles identified by a name or nomos: hence ‘nominal’. An example is the division of Christianity into Orthodox, Catholic, Lutheran, Reformed, and Anglican. An ordinal scale is a statement of order or ranking—more or less—that implies nothing about the relative gaps between points on the scale. For example, we would implicitly be using such a scale if we described the degree of spiritual power imputed to the clergy (or sacerdotalism) by different Christian churches in a ranking. We put the Brethren at the bottom because they reject the category of clergy altogether. We could put the Presbyterians, Methodists, and the low end of the Church of England next because, although they regard the professional minister as little better than any elder, lay preacher, or lay reader, they find them convenient, and they reserve to the clergy the right to administer communion. Then we would place the high end of the Church of England, which supposes that the properly ordained clergyperson has some supernatural power Page 4 of 32

 

Measuring Religion that invests his or her activities with sacramental effect. And at the top we would have the Catholic Church, which is high Church of England max. With a bit of refinement that could be the basis for measuring sacerdotalism and using it in some correlation if, for example, we wanted to test the proposition that the more demotic churches in Britain have declined less rapidly than those with a high theory of the clergy. What we can do with information presented on an ordinal scale is limited by what distinguishes it from the interval scale: we cannot set a value on the spaces between the items on our scale. We cannot plausibly say that the amount of extra sacerdotalism that is added as we move up from Brethren to Presbyterian is the same as that involved in going from low Anglican to high Anglican. Hence, although we might for convenience assign these four positions the values 1, 2, 3, and 4, this would be misleading, because it implies the intervals between the positions are the same. But we can plausibly describe the typical level of sacerdotalism in various societies by comparing them to the ‘mode’, which is simply the most popular category. Ordering has the additional virtue of forcing us to think clearly and transparently. Joseph Bryant is critical of Guy Swanson’s comparative modelling of elements of his explanation of the origins of religion, but his critique is possible precisely because of the clarity of Swanson’s ordinal scale comparisons.5 (p.75) An interval scale assumes the gap between the positions is regular. The Celsius scale, for example, describes temperature in terms of regular intervals created by taking the difference between the freezing and boiling points of water under a pressure of 1 atmosphere and dividing it into 100 identical units. Interval scales allow us to do what we cannot properly do with an ordinal scale: to calculate the arithmetic mean and median to describe the typical or central tendency. But such averages are not much use unless we know the spread of the results: the same average life expectancy could be produced by everyone dying at roughly the same age and by some people dying very young and a similar number living long enough to receive the Queen’s centurion birthday congratulations. Clearly those would be very different sorts of societies. The advantage of the interval scale is that we can calculate and describe such differences in the spread of findings with such devices as the interquartile range and the standard deviation. For completeness I add the ratio scale, which is an interval scale with a fixed zero point. An example is the Kelvin scale for temperature, which starts at ‘absolute zero’ (or –273.15 Celsius). The great advantage of the ratio scale (which can be used for most of the properties that interest physicists such as

Page 5 of 32

 

Measuring Religion plane angle, mass, length, duration, and energy) is that it allows a full range of mathematical operations. There are two points to this brief toe-dip into statistical theory. One is to recognize a limit on analysis in the social sciences. As relatively few social characteristics can be meaningfully described with a ratio scale, the sophistication of statistical analysis commonplace in the natural sciences is impossible or, if displayed, misleading. Nonetheless, scaling is vital, even for qualitative work. Ethnographers, for example, use implicit notions of more and less for every characteristic they describe. For example, Tanya Luhrmann’s excellent study of witches in London in the 1990s often distinguishes degrees of involvement and status within the groups.6 How do we know someone is more involved than someone else? We construct an implicit scale based, for example, on number of meetings attended. We recognize that someone is a leader and someone else a subordinate by implicitly counting and scaling the numbers of times their guidance is sought by others, the number of times they lead the rituals, and so on. The second point is to show the limits of qualitative observation. Generally, it is confined to the first two types of scale, while quantitative (p.76) analysis is usually able to construct interval scales and thus perform a wide variety of revealing and—for many explanations—necessary statistical analyses.

Is Counting the Godly Especially Difficult? As I noted in Chapter 3, scholars of religion seem unusually likely—note the implicit counting and comparison here—to make heavy weather of defining religion. They are also unusually likely to claim that assessing the extent of some characteristic of religion is especially difficult, if not impossible. I would like to suggest otherwise by demonstrating that many of our concerns can be quite properly measured and those measures can be presented on scales that allow basic statistical comparisons such as arithmetic means. For example, ‘Is the typical Scot more or less religious than the typical English person?’ is a quite proper interest. That may seem like the insensitive and clumsy question that gets quantitative analysis a bad name among arts graduates, but it is essential for answering the sort of question they do like to ask. For example, we might suspect from casual observation that the charismatic movement is far less popular in Scotland (and in Wales and the north of England) than in the south of England and guess that there is something about the religious culture of places with a long and strong tradition of demotic Protestantism that makes the charismatic movement unattractive, but before we can set off down that route we need to eliminate the simpler possibility that Scots are simply less interested in all forms of religion.

Page 6 of 32

 

Measuring Religion Beliefs

An anthropologist of religion has suggested ‘it is possible to hold a reasoned distrust both of statistics and of attitude surveys’.7 True. But a reasoned distrust of qualitative research is far more appropriate. Measurement in the study of religion is no harder than in any other field of the social sciences. We can assess the popularity of religious beliefs by asking people what they believe; though they may be less direct about their asking, this is what ethnographers do. Once we are (p.77) familiar with the range of options among the people who interest us, we can compress them into survey questions that force a choice between options (as in ‘Which of the following come closest to your view of God?) or that offer a range of responses to some assertion (as when we ask people to ‘strongly agree, agree, neither agree nor disagree, disagree, or strongly disagree’ with ‘the Bible is the revealed word of God’). That the beliefs in question may concern the supernatural does not of itself make measuring their incidence any more difficult than is measuring belief in such other diffuse or abstract notions as ‘democracy’, ‘justice’, ‘fairness’, ‘good health’, and so on. As always in such exercises, there is a trade-off between depth and breadth of enquiry. In total I must have spent well over 200 hours talking to the Revd Ian Paisley in the back of his car as he was chauffeured around Northern Ireland, a similar amount of time talking to his clergy and political colleagues, and I read all his published works and listened to many recorded and live sermons. I learnt a lot about what Paisley believes. But if I want to know what the typical middle-class Scot believes I would need to do the same with at least a few hundred people and that is impossible. To survey the beliefs of 2,000 people we need to compress our questions into simple forms with a narrow range of permitted answers. Each of these exercises is worthwhile, and a good case can be made for saying that the detailed immersion involved in the first style of research is a necessary condition for doing the second type well, but in principle I see nothing wrong with the second type. Survey questions can be good or bad, as can questions asked in lengthy interviews or slid into casual conversation, but there are a few general points worth making. I will leave until Chapter 7 the very large question of whether the social influences at play in the act of accounting for oneself prevent accounts being used as data about people’s motives, intentions, beliefs, and the like, and concentrate on more mundane matters of suitability of attitude surveys. It is unreasonable to dismiss out of hand a commonly used research tool, but it is reasonable to think about the circumstances in which it is more or less useful. First, there is little point in asking direct questions about matters that are sufficiently sensitive to make dissembling likely: as Seth Stephens-Davidowitz has demonstrated, the number of times that people in a survey claimed to have had sex with a condom was considerably higher than the total number of condoms sold in their states.8 He does not know if they were exaggerating their

Page 7 of 32

 

Measuring Religion prowess or (p.78) their caution, but he is probably right that direct questions about sexual activity are probably not worthwhile. Second, the more concrete or factual the matter the better. We need to limit possibilities for misunderstanding. How many rooms are there in your house? How many sinks do you have? These are useful indicators of wealth, and they are perfectly clear. We need not be concerned that the question is misunderstood or understood differently by different respondents. ‘Do you believe in God?’ is quite a different matter and is probably sensible only as a preliminary to a series of more detailed questions about what sort of God one might believe in and the certainty of that belief. Third, survey questions work best for issues to which the respondent has probably given some thought. Many survey questions seem unlikely to be tapping any enduring beliefs because they ask about matters that may hardly ever have crossed the respondent’s mind. ‘Is the Bible the revealed word of God?’ may work for committed members of Protestant churches because it is something they think about a lot. Indeed, self-consciously liberal and conservative Protestants may define themselves largely by answers to precisely that question. It may work less well for Catholics, and it may be entirely meaningless for the religious indifferent and ignorant. The excellent British Religion in Numbers (aka BRIN) website contains summaries of very large numbers of recent surveys of religious beliefs, and one of the most obvious conclusions from even a cursory examination of the graphs to be found there is that contemporary responses are all over the place. I can think of four possible explanations. First, the population may divide into distinct constituencies of people who have clear, consistent, and stable views on such matters and the variation in the findings may be a consequence of different constituencies being differentially sampled on different occasions. That is, any sample being surveyed may be more or less representative of some wider population. Second, it may be that the decline of the churches has removed shared understandings and left people to devise their own idiosyncratic and inconsistent positions, to which they are nonetheless strongly and consistently committed. That is, the variation may reflect the diversification effect of secularization. Third, it may be that a large part of the population now gives almost no thought to those questions. Because I like most food and am indecisive in restaurants, when asked to order I pick at random or (p.79) choose the first or most memorable item on the menu. Many survey respondents seem to do similar things, which is why survey question wording and placement are important. I

Page 8 of 32

 

Measuring Religion suspect that if we asked the same people very similar questions a month later we would get different responses. Fourth, survey questions seem most useful when, rather than attempting to ascertain a complex attitude by direct questions, they take the respondent through scenarios that offer fairly concrete expressions of what we take to be a general attitude. For example, rather than ask directly ‘Do you think religion should have more influence on public life?’, one can ask a series of questions about specific examples. The 2001 Scottish Social Attitudes survey explained that television companies were required as a condition of their licences to make and show religious programmes, and asked whether that requirement should be continued. It also asked whether all state schools should be required to hold public prayers. Having first checked that there is a high degree of consistency in responses to these and related questions, we could combine the answers to a wide variety of questions into a single scale for the attitude of being ‘pro-public religion’ and look for correlations between scores on that scale and other characteristics. I should add that discovering what people think is as much a problem for qualitative styles of research as it is for the survey analyst. With surveys we at least know what was asked. In research based on free-form interviews or casual conversations, we often have no idea at all what preceded a respondent saying something that is quoted as evidence of a particular belief or attitude. Oral histories are sometimes treated as if transcriptions of respondents’ monologues were uncontaminated evidence. Where what the interviewers said is included in such transcripts we can often see questions that might reasonably be regarded as ‘leading’. There is nothing wrong with leading; it is often necessary to encourage the taciturn or steer the garrulous. My point is that there is no justification for treating the highly stylized, confined, and brief questions of the survey and the free-form length discussions of the interview as respectively polluted and pure.9 One final thought about the value of surveys for assessing beliefs: the big survey has the advantage over the small case study in that it allows us to hope that mistakes cancel out. Though there may also be specific sources of error in any sort of research, there will also be random mistakes in description, observation, or assertion. Consider a national survey of the victims of crime. Respondents may mistake an (p.80) accident for a crime and vice versa. Coders may code robbery as burglary or vice versa. The more data we have, the better the chance that such random errors will cancel out and leave the overall data reasonably reliable. A single researcher talking to twenty people offers little insurance against misunderstanding. A team of trained survey administrators quizzing 2,000 people with a set of questions that have first been tested in pilot studies can reasonably hope that idiosyncrasies and misunderstandings will cancel

Page 9 of 32

 

Measuring Religion out.10 And when many large surveys produce consistent results, our anthropologist’s ‘reasoned distrust’ begins to look like special pleading. Behaviour

Most religions are marked not just by specific beliefs but also by associated actions: the Godly are supposed to behave differently from the unGodly. The Christian churches all require regular attendance at acts of public worship. Paul’s Letter to the Hebrews (10:25) instructs Christians not to forsake ‘assembling ourselves together’. The Catholic Church’s Code of Canon Law refers to Sunday attendance as an obligation on all the faithful and its Catechism states that deliberate absence from mass without a very good reason is a grave sin.11 The Church of England’s canon law says: ‘It is the duty of all who have been confirmed to receive the Holy Communion regularly, and especially at the festivals of Christmas, Easter and Whitsun or Pentecost.’12 The Shorter Westminster Catechism—the 1647 summary of Presbyterian teaching—asks in question 88: ‘What are the outward means whereby Christ communicateth to us the benefits of redemption?’ and answers: ‘The outward and ordinary means whereby Christ communicateth to us the benefits of redemption, are his ordinances, especially the word, sacraments, and prayer; all which are made effectual to the elect for salvation.’ The word (that is, the Bible) could be read privately by the well informed, but hearing its meaning expounded in sermon would be more useful to most people. Likewise, one can pray alone, but public and communal prayer is encouraged. The sacraments require public ministerial administration, and this typically takes place in church. In brief, while regular church attendance does not define the Christian, it is expected and it is normal, which is why the churches themselves use attendance counts as a measure of popularity. (p.81) In principle, measuring church attendance is simple: you count the people who go to church. However, most explanations that use church attendance as evidence of religious interest are usually concerned with more than the one or two congregations we could attend ourselves. We are normally interested in the churchgoing of large populations. Which brings us to aggregation. Some, but only some, churches, denominations, and sects collect and collate such attendance data and some of them do not make it available to the public. We can get some idea of national patterns for England from figures produced by the Church of England and the Methodists because they account for a large proportion of the institutional religious activity, but nonetheless converting those into a single figure for the whole country (by, for example, calculating what proportion of national church ‘membership’ is accounted for by Anglicans and Methodists and scaling up their combined attendance figures pro rata) involves a great deal of estimating.

Page 10 of 32

 

Measuring Religion Furthermore, it is clear that the Church of England’s reported ‘average Sunday attendance’ data (collected by averaging four Sundays) now underestimates churchgoing because attending midweek is now common. For example, average Sunday attendance in 2001 was 1,041,000, while adding average weekly attendance took it to 1,205,000. On average, for the ten years 2000–9, adding weekly to Sunday attendance raises it by 18 per cent. The problem with adding weekly attendance to Sunday attendance is that we have little idea of how many midweekers also attend on a Sunday, and so we have to present Anglican church attendance as being somewhere between a maxima of the weekly figure (if no midweekers also attend on Sunday) and a minima of the Sunday figure (if all midweek attendances were by people who also attended on Sunday). And there is no useful information about the extent of religious activity by nonChristians. If mosques, temples, and gurdwaras do collect attendance data, they do not collate them or report them. For the last two decades of the twentieth century and the start of the twentyfirst, we have the church censuses conducted by Peter Brierley and the various organizations for which he worked. With remarkable diligence, he and his associates compiled a database of all extant Christian congregations and mailed all with a request for a count of attenders at services on an ordinary Sunday in October and various other information about the attenders and the congregation. Particular effort was put into identifying congregations of small (p.82) denominations and sects. The return rates were good. For example, 38,607 forms were sent out for the 1989 English church census and 70 per cent were returned completed. Brierley conducted censuses in England in 1979, 1989, 1998, and 2005, in Scotland in 1984, 1994, 2002, and 2016, and in Wales in 1982. Extrapolating from the 1998 survey, Brierley estimated church attendance in England as 7.7 per cent of the adult population in 2000. His 2005 census showed it then to be 7.0 per cent. The Welsh figures are similar at 7.9 and 6.6 per cent. The Scottish estimates are higher: 13.1 per cent in 2000 and 12.0 per cent in 2005.13 There is one good reason for thinking that Brierley’s censuses are reasonably accurate. In many of his published reports, Brierley extrapolated future figures from previous ones. In the cases where there was a subsequent census, his extrapolations turned out to be close to the figures derived from the subsequent study. A third source of church attendance estimates is the data generated by social surveys. Many bodies commission one-off surveys that include questions about church attendance. Unfortunately, these vary enormously in quality at every stage of the process, from the initial sampling to the reporting of data, and it cannot be an accident that one-off surveys generally produce results that please their funders: those funded by secularist organizations show religion going down Page 11 of 32

 

Measuring Religion the pan, while surveys produced by religious organizations find evidence that religion is more popular than we suppose. More useful, because they are more consistent in their questions and make their raw data available for reanalysis, are recurrent surveys such as the British Election studies, the British Household Panel Survey, and the British Social Attitudes surveys (and their regional counterparts). The problems of sample surveys are well known. First, the people questioned may not be representative of the population at large. To an extent this can be corrected by weighting. If our final slate of respondents is 60 per cent female, we can correct the gender imbalance by scaling down the responses of women to 51 per cent. Of course, such weighting works only for the limited range of demographic characteristics known from other reliable sources.14 Second, surveys report, not actual activity but claims about activities. Those claims may well be false, especially if the matter in question is something about which a social consensus of value may exist or be suspected. People may shade their answers to comply with what they think the questioner wants or expects. Dramatic proof of (p.83) unreliable responses was produced by Kirk Hadaway, Penny Long Marler, and Mark Chaves, three American sociologists who found it hard to reconcile the routinely reported survey figures for US church attendance with what Kirk Hadaway knew from the data he collated in his work as in-house researcher for the United Church of Christ—one of the major American denominations. They took Ashtabula County, Ohio, and contacted every church they could identify and asked the minister to report attendance on a particular Sunday. Where they had no response, they stationed observers to estimate attendance. They also commissioned a standard telephone sample survey that included a question about church attendance. When they compared the two sets of results, they found that the survey count was 83 per cent higher than the figure arrived at by augmenting reports from clergy by direct observation at non-cooperating churches.15 They repeated the experiment on a smaller scale with a large Baptist congregation in Alabama. It had an adult Bible study before the main morning worship service and most people attended both. The Bible study took a register. A telephone survey of the congregation, with a question about attendance the previous Sunday buried in a welter of spurious questions, was arranged for the following week. Compared with the evidence of the register, the survey results showed overclaiming of over 60 per cent. Interestingly, Hadaway, Marler, and Chaves do not follow Stephens-Davidowitz in plumping for the ‘everybody lies’ explanation. They make the more subtle point that regular churchgoers tend to hear the question not as concerning specific behaviour the previous week but more as ‘are you a churchgoer?’, and, because they think of themselves as ‘people who go to church’, they reply in the affirmative even though they did not attend that particular week. Hence, while

Page 12 of 32

 

Measuring Religion religion is still popular, surveys will fail to reveal the full extent of declining frequency of attendance. As both these tentative explanations—compliance effects and time lag—explain the overclaiming as a roundabout function of the popularity of churchgoing, it seems reasonable to suppose that British people are less likely than Americans to overclaim, but survey figures for British church attendance are still considerably higher than the numbers arrived at by Brierley’s censuses or produced by the churches. For example, the British Social Attitudes survey for 1991 shows that 11.3 per cent of respondents claimed to attend once a week or more often, 2.4 attended once a fortnight, 6.4 per cent attended once a month, and 12.5 attended twice a year. If we convert all to the (p.84) appropriate weekly figure and total them, we have a weekly attendance figure of 14.7 per cent.16 This is around twice the figure Brierley derives from his church censuses. There is one form of survey that avoids compliance effects: the time-use diary study (or TUS). Here a hopefully representative sample of people are asked to write down what they are doing every ten minutes. In the 2001 UK TUS they were also asked to note if they were doing it at home or outside the home and if they were doing it by themselves or with others. Because such surveys have no ostensible purpose (or at least not one obvious to the respondents), it seems unlikely that people will much exaggerate any one sort of activity.17 To pre-empt irrelevant criticisms of survey data, I should note that the welldocumented failure of pollsters to anticipate correctly the results of the 2016 UK referendum on the European Union membership or the 2017 Westminster general election is almost certainly explained by two considerations that are not normally pertinent in the study of religious belief and behaviour: the failure to tap successfully one crucial stage in political preferences translating into voting and the difficulty of tracking late changes in preferences. Without reliable information on the likelihood that respondents will trouble to vote, their stated party or policy preferences may mislead. And campaigns themselves make a difference. Events in the days shortly before a vote (and such events would include the publication of opinion polls) may change either people’s minds or their determination to turn out. There are many other religion-redolent actions other than churchgoing that we can count. For certain sorts of religion, membership or formal affiliation is significant, and, although one needs to be well informed about the ways in which such association is recorded, it is not conceptually or practically difficult to measure. Primitive Methodist chapels, for example, reported to their district headquarters every three months the numbers of new members gained by baptism or by transfer from another chapel, the numbers lost through transfer, death, or ‘ceased to meet’, and the resulting total. Chapel stewards sometimes got lazy; we can guess that identical figures being reported each quarter or Page 13 of 32

 

Measuring Religion multiples of ten appearing unfeasibly often are signs of casual compliance. But cross-checking membership statistics with records of subscriptions paid shows that every three or four years, when a circuit of chapels got a new superintendent minister, the figures were accurately updated.18 (p.85) Identity

One frequently available mark of religious interest that can be counted is identification as given in government censuses and social surveys. Although there has long been a religious identity question in the Northern Ireland census, Britain came late to asking about religion: so late in fact that it was done only in 2001 and 2011. The number of people who claim some religious identification is always considerably higher than the number who assent to religious beliefs or who take part in activities predicated on such beliefs—witness the 72 per cent of the English and Welsh who in 2001 ticked the ‘Christian’ box on their census forms—but there is one facet of such identifications that we can suppose comes close to reflecting people’s actual religious beliefs: the ‘No Religion’ category. We can imagine good secular reasons why English and Welsh people who would not claim to be religiously observant should nonetheless on occasions claim a religious identification. My own casual conversations with many such people after the 2001 census suggest that one consideration was a wish to make a statement about the cultural heritage of their country. Placing the religious identification question directly after a question on ethnic identity made that reading more likely, as did the fact that the tick-box options ran Christian, Muslim, and so on. One acquaintance made the point forcefully when she said something like: ‘I’m not religious myself but I’m more of a Christian than a Muslim and so is the country and so it should be.’ But because until the 2010s it was a deviant category and because it offers the least opportunity for claiming membership of some community, we can have some confidence that those people who ticked the ‘No Religion’ box really do mean that they do not identify with any particular religion. As an aside it is worth noting the value of a multiplicity of sources. We suspect that the 2001 72 per cent Christian identification for England and Wales tells us more about ethnicity and politics than about religion, because the corresponding figure for Scotland, at 65 per cent, was lower despite all the sources showing church membership and attendance tobe higher in Scotland than in England and Wales. The difference may well be explained by Scotland having a much smaller Muslim population, by the Scottish religion question not following directly on the ethnicity question, and by the Scottish choices (Church of Scotland, Catholic, (p.86) Episcopalian, and so on) directing attention away from ethnicity and towards actual involvement with religious organizations.

Page 14 of 32

 

Measuring Religion Other Measures

Most of us would not object to the general principle that people’s optional spending is a good measure of their interests, and I see no reason why we should not apply that to assessing the popularity of religion. Church incomes from voluntary donations are often available and are a decent measure of shifting popularity, as is the amount of money, relative to other bequests, that is left to religious organizations. One of the ways in which people have traditionally demonstrated their religious commitment is inviting religious organizations to sanctify important points in the life cycle. To put it the other way round, one of the ways in which religious institutions have impressed themselves on the lives of ordinary people is by providing offices for the celebration of key life events. Baptism, confirmation, wedding, and funeral have for centuries marked the key stages in life (and, in the case of funerals, its end). Measuring uptake of these options is relatively easy, and for some we have long runs of reliable statistics. For example, we have fairly reliable data on the number of infants baptized by the Church of England since 1885, and, as it baptized the vast majority of English infants, such data can serve as a good index of the popularity of baptism as a whole. For Scotland we have figures for religious and civil marriages since 1855. As with any other measure, we need to be sensitive to contexts, especially when we compare across time and space. For example, there may well have been relatively more religious marriages in Scotland than in England because the Church of Scotland has long married divorcees. The decline in religious celebrations of rites of passage that Bryan Wilson documented in 1966 has either continued apace or accelerated. In 1962 the Church of England baptized 55 per cent of England’s infants. In 2008, the figure was 13 per cent, and only a small part of that gap is filled by non-Anglican baptisms.19 There is little mystery as to why not-especially religious people will choose a church wedding. Particularly if it is conducted in a medieval church, the ritual confers solemnity. As well as providing far more impressive backgrounds for photographs than almost any civil Registry Office, the church wedding (p.87) shows that the parties are serious about their vows, even when they are invoking as a witness a God in whom they do not believe. In the 1960s, about 70 per cent of weddings in England were religious. In 2003 less than a third were.20 Data on the religious (or otherwise) nature of funerals are not collected by the government, but Co-operative Funeralcare is by far-and-away the UK’s largest provider, with over 100,000 funerals a year. In 2011 it reported that only two-thirds of its funerals followed the rites of a particular religion. Humanist celebrations counted for 12 per cent, and 21 per cent of its funerals were ‘contemporary’—a personalized celebration of the life of the deceased. Only a third of funerals now have only religious music; the remainder use contemporary or classical music or a mixture of both. The top three funeral songs in 2009, according to Co-operative Funeralcare, were ‘My Way’ (by Frank Page 15 of 32

 

Measuring Religion Sinatra or Shirley Bassey), ‘Wind Beneath My Wings’ (Bette Midler or Celine Dion), and ‘Time to Say Goodbye’ (Sarah Brightman or Andrea Bocelli).21 The above is by no means an exhaustive list of ways that we can count the Godly or assess the strength of religious sentiments.22 Good sociologists will find data in unlikely places. Over my lifetime the typical funeral service (whether in a church or a crematorium) has shifted from commending the departed to God to celebrating the life of the deceased. We can quantify that change by analysing changes in the design of graveyard headstones, as Curt Dahlgren has done. He photographed 1,485 headstones and found that over the twentieth century religious symbols had been increasingly replaced by ones that referenced the deceased’s secular interests.23 Societal Measures

All these above measures represent primarily individual choices (though, of course, such choices are shaped by social norms). We can also infer a great deal about the power and popularity of religion from features of societies. For example, the number of full-time religious professionals that a society supports is a good indicator of either the power or the popularity of religion. If the clergy are supported by some form of taxation, as was the case with the state churches of England, Scotland, and Wales until the early twentieth century, and as is still the case in some European countries, their number is an index of the popularity of religion with those responsible (p.88) for setting taxes. If religion is entirely voluntary and the clergy are paid by their adherents (as is now the case in the UK and has been the case in the USA for over two hundred years), their number is an index of the importance that people give to organized religion relative to other spending. We can also infer something about public estimations of religion from the social standing of the clergy. Of course, context is important, but it is significant that the twentieth century saw a marked decline in the social status and income of the clergy, relative to other professional groups. This can be measured with considerable accuracy.24 There has been a marked decline in the number of people who enter the ministry as a first career or, to put it the other way round, an increase in the numbers of people who come to the ministry as a second career.25 In the late 1960s, the average age of men in training for the Church of England was 25; in the late 1990s it was 40. One measure of social value or power is earnings and we can note that clergy incomes have declined relative to similar occupations. In 2001 the average Anglican clergy salary was about £17,000, which, even with free accommodation, was well below the national average for graduates.26 A 2012 Office of National Statistics listing of 400 occupations by salary has the clergy at Page 16 of 32

 

Measuring Religion number 293: its figure of £21,485 is far below that of ‘health practitioners’ at £54,684.27 Another method is to ask people how much they trust various occupational groups; this is something Ipsos MORI has done since 1983. Then 83 per cent of the public trusted clergy to tell the truth. In 2013 the figure was 66 per cent, well behind doctors (89), teachers (86), and judges (82 per cent).28 Popular or Folk Religion

A common criticism of quantitative research is that focus on measurable indices of religious interest gives undue weight to institutional or formal religion, which is contrasted with popular, folk, or informal religion. There is something in this. It is tempting to concentrate on marks of religious interest that are easily identified and measured, and it is easier to count church attendance than to assess the popularity of quasi-religious superstitions. The ‘churching of women’ was certainly easier to identify in the Middle Ages when it was an officially (p. 89) approved Church of England service than in the second half of the twentieth century when it had diminished to a superstition. A new mother was a source of bad luck until she had ‘been churched’—a requirement that could be met by her simply stepping into the local chapel any time the building was open.29 However, that undue focus on institutional religion may be misleading does not, of itself, prove that folk religion is of enduring importance: that needs to be independently established. And it is hard to see how popularity can be assessed without counting. To say that a particular group of people hold a particular superstition must mean that the researcher has asked around enough people and often enough got certain kinds of replies to questions and certain responses to promptings to be confident that many people really do believe x rather than y. Moreover, if popular or folk religion is to be cited in arguments about the relative popularity of religious sentiments, it is not good enough to show that there is some of it about. Difficult though it may be, we need plausible estimates of relative popularity at different points in time. Without that we cannot rule out the possibility that popular religion is declining in tandem with institutional religion for two reasons: similarity and parasitism. It seems likely that popular religion is undermined by precisely the same social changes that weaken institutional religion. The growth of social diversity brings with it cultural variation that weakens the hold of any particular cultural practice, as does anything that weakens community cohesion. Arguably, popular religion is more vulnerable to secularization than its institutional counterpart because, unlike the faith preserved and taught by the churches, it relies on person-to-person transmission within an enclosed community. When communities are broken up (for example, by large-scale housing redevelopment or population shifts), their idiosyncratic traditions are rarely re-created in the new settings. The development of new technologies has undermined the importance of parent-to-child transmission of knowledge, as has increased reliance on the mass media, on schools, and on peer-to-peer transmission. When housework changed only slowly, daughters could learn both skills and attendant Page 17 of 32

 

Measuring Religion superstitions from their mothers. The introduction of new machinery killed off most such superstitions, and I see no sign that new ones are being devised: for example, I know of none connected with microwaves or coffee-makers. Second, in many ways popular religion is parasitic on the more formal variety. The new mothers of the fishing village of Staithes (p.90) cannot follow their mothers and grandmothers in ‘churching’ themselves by stepping into a chapel, because two of the three chapels have closed and the third is open only for services. More abstractly, as knowledge of formal religion declines, so does the opportunity to rework its content into popular themes. Like many children of the late 1950s, my sister and her friends skipped to a chant that contained references to Easter and the important days leading up to it; my daughters and their friends skipped to a chant that referenced McDonald’s burgers and a BMW. The folk appropriation of institutional religion is fascinating, but there are good grounds for suspecting that the two rise and fall together. The counter case must be supported with evidence that popular religion remains popular, and that requires that it be measured in some way.

Groups and Boundaries There is one crucial element of the measurement process that I have thus far not mentioned directly, though I have implicitly relied on it a number of times. Gross counts are rarely useful for comparing over time or place. Hadaway and colleagues could have reported Ashtabula County church attendance in absolute numbers, but to compare the results of their two different sources—the church census or body count and the telephone survey—they need to express both in the common currency of percentages and that draws our attention to the problem of the bottom line of the fraction: the total number of people who could have attended or could have answered the telephone survey in the affirmative. It is a simple point and one that most people intuitively grasp: whether a characteristic of some group is important or significant or revealing depends on the characteristics of the wider population from which that sample has been drawn. Telling us that a particular number of the Kendal holistic milieu participants were women is of little or no value unless we know the relative number of men and women in the Kendal population: the bottom lines of the fractions. David Goodhew claims to have found church growth in York. His study lists fifteen new churches that were established between 2001 and 2011. In total their claimed attendance in 2011 was around 715. (p.91) Most of the reported attendance figures are multiples of ten, which does not give confidence in the accuracy of the count, but we will take the figures as given. Another study gives a total attendance figure for York in 2011 of 6,801. Goodhew did not, in 2011, survey the York churches that were extant in 2001, so we cannot know if they lost (or gained) attenders over the decade, but for the sake of argument I will assume that their attendance remained the same and that all the 715 people Page 18 of 32

 

Measuring Religion attending the new churches were genuinely novice church attenders rather than defectors from the other churches. That would be an increase in churchgoing in York of 10.5 per cent in a decade. Which would indeed be remarkable, except that in the same period the adult population of York grew by 11.3 per cent. Thus, even if we permit the contentious assumptions that all the people attending newly formed churches were novice churchgoers and that none of the other churches had lost attenders, the growth in church attendance failed to keep pace with the growth in the adult population. This is not growth: it is decline. Another study describes an ‘Anglican resurgence’ in London: ‘Partly through some favourable external circumstances but mainly through a new cultural strategy and spiritual renewal, churches started growing again in 1991 and have continued to grow for the last 20 years.’30 But then the adult population of London had also been growing. Between 1991 and 2011 London’s adult population increased by around 22 per cent.31 The 15 per cent growth claimed for the Church of England between 1990 and 2009 might thus better be described as failing to keep up with population growth. To summarize, measuring the phenomenon that interests us is often not enough. For our counts to have much value as the basis for identifying what is significant and for making comparisons across time and space we usually need to identify (and measure) something else: the pool or baseline we will use to convert absolute counts into percentages. The general principle is intuitively understood, but, as the York and London church growth claims show, it is sometimes ignored.

The Virtues of Quantitative Research Having spent most of my career on various types of qualitative research, I am happy to acknowledge that there are research problems (p.92) that are best approached through the analysis of texts, participant observation, unstructured interviews, and the like. I also believe that good quantitative analysis requires a considerable degree of familiarity with the topic one is studying, and such familiarity may be acquired by qualitative research. However, the balance, certainly in British studies of religion, is so firmly in favour of ethnography that redress justifies further statement of what, in other contexts, would be the glaringly obvious. There are a number of respects in which, for matters that could equally well be studied by both quantitative and qualitative methods, the former can reasonably be regarded as superior. Spotting Patterns

As already noted, one obvious advantage of the statistical analysis of big data is that it reveals subtle patterns that are invisible to the naked eye. There is now software that helps us analyse qualitative data, but there is never a sufficiently large number of ‘data points’ in observational or interview research to show the sort of second-order relationships at work in Louis Zurcher and George Page 19 of 32

 

Measuring Religion Kirkpatrick’s study of the reasons why Americans support anti-pornography crusades.32 I think their argument—that people who are ‘status inconsistent’ in being overpaid relative to their levels of education are subconsciously anxious about their position in American society and express this by supporting conservative movements—is wrong. I see no reason to suppose that Americans do worry about being over-rewarded and hence I am more drawn to a simpler explanation that takes each of those variables separately. People who are well paid are likely to think well of the status quo. And overt support for pornography (as distinct from covert consumption) is more likely among the university educated. But, right or wrong, the point is that any relationship between status, education, and attitudes can be sensibly explored only with large-scale quantifiable data. Re-examination and Critique

Quantitative data are often available for re-examination. Much ethnography could have been made up. The data from the 2001 Scottish Social Attitudes survey are available to anyone to download and reanalyse. Nobody has to take my word for it that, according to the (p.93) 2001 Scottish census, it is not Catholics (thought by some to be victims of discrimination) who have the lowest profile in the labour market: it is people who were raised with no religion. Sceptics can analyse the data themselves. Even without reanalysis of the original data, quantitative research is usually reported in ways that make it relatively easy to see why one disagrees. In The Red Hand I presented a number of arguments about the unusual features of ‘prostate terrorism’ and illustrated them with snippets from a vast body of notes taken over many years. Readers may suspect I am wrong, but they cannot see the bulk of the evidence that led me to my conclusions. Even lodging interview transcripts and other such qualitative material in a public access data archive does not remove the need for trust in the discernment of the qualitative researcher, because such material often omits vital elements. A transcript of one of my prison interviews will not show the reader the subtle gestures and facial tics that helped me judge the meaning or veracity of what was being said. The conventions for reporting survey results generally require a clear separation of data presentation and analysis and often allow us to see why we are not persuaded by the conclusions. For example, Tadeusz Doktór assesses the growth of New Age spirituality in central and eastern Europe using data from the International Social Science Program (ISSP) surveys of 1991 and 1998 and reports a conclusion very different from the one Glendinning and I arrived at with the Scottish Social Attitudes data and Heelas and Woodhead drew from their Kendal project.33 While we found a strong positive connection between class (taking that as an amalgam of occupation, income, and education) and interest in the New Age, Doktór concludes that New Age beliefs are more popular in the least-developed countries of Europe. The discrepancy is almost certainly a technical one. Scholars in Britain and America use some version of Page 20 of 32

 

Measuring Religion seeing the self as divine or discovering divine power within oneself as the constitutive feature of New Age spirituality and, not surprisingly, find most of it among prosperous well-educated people. Doktór has used interest in the occult and forms of divination, which is a rather different matter and which might sensibly be expected among those who feel themselves somewhat powerless.34 The explanation of the disagreement is less important than this procedural point: with Doktór’s work the reader can see how he came to his conclusions and can, if minded, go back to the original ISSP and SSA sampling frames, questions, and responses. (p.94) Agreed Procedures for Analysis

Quantitative research has widely agreed principles for data analysis. The experts may argue about which test of significance is most appropriate for what sorts of data, but there is a consensus that limits the sorts of claims one can make from any particular body of data. Qualitative research relies entirely on the skill of the analyst to describe accurately and to tease out causal connections; the amount of evidence that can be presented in any publication is always just a tiny fraction of the observations that led the ethnographer to his or her conclusion. There is sometimes enough descriptive material for the reader to come to alternative conclusions, but mostly we have to accept the analyst’s possibly idiosyncratic judgement of what data represent. Consider the work of two anthropologists, one of whom spent fifteen months involved in the New Age milieu of Glastonbury.35 The main conclusion of the study is that the New Age milieu sustains a mode of social organisation and a body of beliefs and ideas whose features in many crucial respects display striking similarities with the social and cultural forms of ‘original human society’—that is to say, human society in its evolutionary basic form.36 Despite having read the book closely and sympathetically, I am not convinced that people whose commitment to individual autonomy prevents them from creating and sustaining social institutions should be described as a ‘tribe’. I can point to a number of anecdotes in the book that raise my doubts, but it is in the nature of such ethnographic work that I cannot see Ruth Prince’s working out. Tests for Consistency

In theory, the ultimate test of any conclusion is its fit with the real world. In practice, there is often no test of this, though some sorts of study are fairly closely linked to real events. Opinion polls on voting intentions are tested by the result of the subsequent election, though apparent error can sometimes be explained by changes that occurred between poll and election or by factors that, though perhaps they should have been, were not considered in the poll. In the same way we can ask how the key themes of Bryan Wilson’s Religion in Secular Society (completed in 1966) accord with what happened (p.95) over the Page 21 of 32

 

Measuring Religion subsequent half-century. Such checks are occasionally available for qualitative work. Although it was not intended to be predictive, the reader of my book Paisley, which was finished before Ian Paisley had to choose between accepting or rejecting a devolution settlement for Northern Ireland that would see him share power with Sinn Fein, can ask if my qualitative analysis of his beliefs and values pointed in the same direction as Paisley’s subsequent actions.37 But mostly such tests are rare for either quantitative or qualitative research. However, quantitative work has an advantage over qualitative work in a second test of validity: repetition. The same or similar questions have been asked of the same or similar populations in hundreds of surveys in recent decades. As have already noted in commenting on differences in responses to the same religious belief questions, variation does not necessarily mean the method is wrong, any more than similar results prove it valid, but replication does take us some way to confidence. Replication is rare in qualitative work, largely because ethnographers, like anthropologists, value novelty: they want their own tribe. If there are no links to external reality that can test the validity of some research method and findings, there is a useful internal test: consistency. Here quantitative work has the advantage over qualitative research. It is sometimes possible to see inconsistences in ethnography, but the presentation of ethnographic research rarely permits such internal examination. In contrast, there are well-rehearsed and agreed procedures for checking how and to what extent various responses to survey questions match up. The sceptic could respond that premodern alchemy had agreed protocols and was still nonsense. Of course, consensus among practitioners is not a guarantee of validity but it is something and it is something that is generally lacking from ethnographic research. Limits on Inference

Arguably the most common fault of poor social research is that it overreaches itself. The criticism I have most often made of articles I have been asked to referee for journal publication is that the analyst has advanced conclusions that cannot be supported by the limited evidence. This is generally less of an issue with quantitative analysis, (p.96) because there are clear and consensually agreed rules about many stages of data analysis. Of course, good researchers working in any style will sense when they are drifting from interpretation to speculation, but statistical analysis has the advantage of tests that alert us to common problems. For example, such tests can tell us when, for the number of variables we are trying to correlate, we have too few cases. Running out of people is now a major problem in the statistical analysis of British survey data on religious beliefs and behaviour. Let us suppose we are trying to identify what sorts of people attend church as a preliminary to trying to explain the appeal of churchgoing. We start with 2,000 respondents. At most only 10 per cent will Page 22 of 32

 

Measuring Religion regularly attend church: that is 200. We know that there are strong associations between gender and churchgoing. So we have 100 men and 100 women. To examine the effects of age, we divide the sample into three age groups. Assuming they divide evenly we have only 33 people in our ‘churchgoing elderly women’ box. Now divide by social class, simplified into just three bands. If spread evenly there will only be eleven ‘upper-class elderly churchgoing women’. Once the numbers get that small, it is obviously dangerous to make anything of apparent patterns, and most statistical packages automatically test for that problem. Social statisticians also have agreed tests for establishing the likelihood that any observed pattern in the data is a product of chance. I should mention one type of overreach common to much research of every sort: treating respondents as ‘experts’, not just about their own experiences and subjective appreciations of the world, but also about objective realities. It is a good principle that, if you want to understand people, you first ask them what they are doing and why, but it does not follow that you should accept their judgements about anyone or anything else.38 For example, a study of antiCatholic discrimination in the Scottish labour market built its claims to find considerable sectarian discrimination on assertions made by respondents, not about themselves, but about others. Some claimed that discrimination was common, though they had never suffered it, and justified that claim with vague assertions about the religious composition of their workplaces, when it was clear that they did not know the precise Catholic/non-Catholic mix of their workplace, nor, given the population of their area, what it should have been if it matched the religious composition of the catchment area.39

(p.97) What we Make of Data One of the grounds for the sort of scepticism expressed in the E. P. Thompson quotation at the start of the chapter is the accusation that the collectors of statistics are naive about what they represent. It is routine for members of the audiences to which I make presentations of statistics on church attendance, for example, to point out (as though it had never occurred to me) that the significance of some measure changes over time or that apparently similar measures can have very different meanings in different contexts. Of course it does and they can. In Britain at the start of the twentieth century church attendance was much higher than church membership. Most Methodist chapels, for example, had twice as many attenders as they had members. For every person who was willing to accept the obligations of membership (which included regular attendance at midweek society meetings), there was another who wished only to attend on the Sabbath. By the end of the twentieth century, the relationship had been reversed, with many churches and chapels having twice as many people on their books (for Anglicans the electoral roll; for Methodists society membership) as regularly attended. For a simultaneous regional difference in the meaning of these indices we can note the apparent paradox of the nineteenth-century Scottish highland and island region, accurately noted for Page 23 of 32

 

Measuring Religion the piety of its residents, having lower levels of church membership than the lowlands. The explanation is simple. The pious Calvinist Presbyterians of the highlands and islands were reluctant to presume that they were assuredly part of the elect that was entitled to take communion. Indeed, one of the services of the three-day communion event was called ‘fencing the tables’ and it involved a minister explaining the characteristics of the true Christian in such a way as to make it clear he thought these rare. Hence the common phenomenon of people who attended church every Sunday of their lives and conducted daily family prayers and kept the Sabbath but never asked for membership. If we want the equivalent of a lowland church membership rate for the highlands and islands, we have to add the total of ‘adherents’ (which was latterly recorded) to the number of members. Clearly counting things is of value only if we understand what they mean, and that requires two sorts of theoretical work. We need to know what something means to the people we are trying to (p.98) understand. Why did the people of the Middle Ages leave money in their wills to pay for others to say masses for their souls postmortem? Why did people in the fifteenth century and in the twentieth century have their babies baptized? Why did Scottish highlanders attend all three days of the annual communion season but not take communion? Most forms of natural science do not require that stage. But the second sort of theoretical work is common to both natural and social science: collecting observations has little value unless we have some idea that shapes our interpretation of, or is tested by, the data. In that sense, theory and data are intimately intertwined. This is not to concede the relativist’s claim that the researcher’s preconceptions and biases prevent accurate observation. It is only to make the much less contentious point that the significance of some measurements is derived from their relationship to some ideas that form part of a coherent set of propositions. This works in two directions. We can either begin with a theory and derive from it some operational measures that in turn generate the data or we can start with the data and work backwards to connect them to our theoretical interests. In the first case we have to acknowledge the key difference between the natural and social sciences: psychologists apart, we cannot normally create experiments that limit the variables operative in any setting to just the things that would test our propositions and hence we are stuck having to make the best we can of what is naturally occurring. Take the example of the secularization paradigm. Although declining interest in religion is not the only change encompassed in the notion of secularization, it is a key change; hence our interest in measures of that abstraction we call ‘religiosity’. We can argue about the best operational measures of that abstraction. While I think it worthwhile trying to assess people’s beliefs directly by various forms of asking them, I am sufficiently mindful of compliance effects that wherever possible I would augment an interest in stated beliefs with measures of actions predicated on those beliefs. So Page 24 of 32

 

Measuring Religion I would be cautious of making too much (which the churches regularly do) of surveys that show that many people who do not regularly attend church say they would like to attend more often. As I cannot think of many obstacles that would prevent the serious religious seeker from finding some suitable outlet for his or her desire to worship, I guess that much of that talk is an artefact of the measurement process: the act of being asked causes the respondent to inflate the salience of some vague (probably nostalgic) notion (p.99) of churchgoing. Actions at least demonstrate a degree of commitment. But what actions form the best measures of religiosity? That will, of course, be context dependent: religions differ in their requirements, and even a single religion, confession, or denomination may emphasize different requirements at different times and for different classes of people. As we cannot experimentally create our own evidence in which any action has only one meaning, we need to appreciate the wide variety of meanings that may inform any act. Data sceptics are fond of such blanket dismissal as rejecting church attendance as a mark of religious interest in the past because ‘people just went to church to conform’. In times and place that may be true, but it does not prevent even such church attendance being an index of religious interest because the pressure to conform itself shows the social power of religion: my great-grandfather may have felt that respectability required him to attend the Methlick parish church in 1870; in 1930 my father felt little such pressure and attended only rarely; if I now attended, I would be part of a deviant minority. The second obvious riposte to the ‘just conformity’ dismissal is this: if few people actually believed, the practice of attending church would never have acquired sufficient popularity to become a social requirement. An interesting aside: it seems clear that among the oldest attenders I meet in churches there is some pressure on men to accompany their wives (and it is usually that way round). The 2001 time-use study showed that, for church attenders who had a coresident partner, almost half of those partners did not attend. That seems like a good indicator of change: unlike a dinner party invitation, church attendance no longer comes with an implied requirement for spouses to act like conjoined twins. All this criticism amounts to is the sensible proposition that, before we can claim a certain measure as embodying our desired variable of ‘religiosity’, we need to know the full range of motives that might inform a particular action. Given that we cannot experimentally control our respondents so that the only thing that varies is that which interests us, our best solution is to go in the other direction and collect as many measures as possible. We cannot be sure just what part of having one’s children baptized, or being married in church (as distinct from a civil ceremony), or regularly attending church, or being a church member, or claiming a religious identity in a census, or assenting to religious beliefs best represents religiosity, but we can view each of these as a (p.100) Page 25 of 32

 

Measuring Religion circle in a Venn diagram and hope that where the circles overlap we have a reasonable measure of what interests us. So, to make the case that Britain in 2000 was less religious than Britain in 1900, we measure as many indices of religious interest as we can find and compare all of them.

The Dangers of Quantification The British study of religion is weighted so much to the qualitative and ethnographic that my primary task here is to restore the balance. However, there are some weaknesses of quantitative work that should be mentioned. One concerns selective amnesia. Once one has generated statistical data, it is easy to forget the origins of those variables, which are often surrogates for what we would really like to know. An example of this problem is an article on Islamophobia I refereed for a journal. It tried to explain differences between societies in levels of Islamophobia by their relative proportions of Protestants and Catholics. It went from identifying a correlation to explaining that difference in terms of the effects of Protestant and Catholic theology. In so doing, the author forgot that the Protestant and Catholic identifications were purely nominal. As the vast majority of the Catholics and Protestants had not darkened the door of a church since childhood, it seems unlikely that their adult behaviour would be much shaped by their churches’ teachings. Had the author consistently called his sample ‘nominal Protestants’ and ‘nominal Catholics’, he or she would not have made that mistake. It is even easier to make that sort of mistake when using a statistical package such as SPSS for analysis. The raw data are fed in one end and results pop out of the other and just how those data were generated, and hence their possible meanings, get lost. An example can be found in the work of Rodney Stark and his colleagues on the relationship between religious diversity and popularity of churchgoing.40 As a measure of diversity, they use the Herfindahl Index. This is 1 – ((a/z)2 + (b/z)2`+ (c/z)2 …) where z is the total number of churchgoers and a, b, c, etc., are the numbers who attend particular denominations. The scores range between 0 and 1, and the closer the number gets to 1, the greater the amount of diversity. If one works out the scores by hand, one discovers an odd characteristic. Imagine a (p.101) town with five equally popular denominations: that will give a score of 0.80. Now imagine a town in which one denomination has half the churchgoers, three each have 10 per cent, and four others have 5 per cent each. That town scores 0.61. So the second town is apparently less diverse, even though it offers eight choices instead of the five available in the first town. The Herfindahl Index actually describes how people have spread themselves across options rather than the total number of options that are on offer. That might seem like a small difference, but it is crucial when, as Stark does, one then argues that the more choices that people have available to them the greater the chance of any one person finding something that suits him or her and the higher the rate of churchgoing. An important difference, Page 26 of 32

 

Measuring Religion which is obvious if one does the sums by hand, becomes obscure when the computer analyses the data. A third general problem is that the statistics used to analyse data may themselves generate spurious patterns. David Voas, Dan Olson, and Alasdair Crockett demonstrated—by creating impossible combinations of denominations —that many of the correlations that Stark and associates discovered were actually artefacts of the statistical procedures used rather than accurate reflections of real associations in the real world.41 An apparently similar problem is what is sometimes called ‘dimensionality’ or ‘overfitting’. Stephens-Davidowitz makes the point with regard to possible genetic explanations of human traits.42 There are so many genes (or ‘independent variables’) that with enough cases we are likely to generate entirely spurious connections; search a big enough data set and some correlation will appear. However, this does not seem like a good reason for abandoning quantitative analysis. It is true that a single analysis of a massive data set may produce a silly answer, but in most fields of social science there is a well-established body of knowledge that allows us to separate the meaningful from the spurious correlation. And we can repeat the analysis on different data sets. Various spurious correlations will appear, but those correlations that persist and are repeated will be promising candidates for explanation.43 Finally, there is the common problem of forgetting the limits of much statistical analysis: correlation (that is, two variables changing regularly, in step) does not of itself prove causation. Demonstrating that two variables are commonly related (having lower than average levels of education and being conservative, for example) does not (p.102) prove that A (lack of education) causes B (being conservative). It could be that B causes A (conservative people are more likely to eschew higher education for a life in commerce) or that both A and B are caused by some untested-for variable C. In this example, it might be that age (that is, being elderly) explains both being conservative and having missed the opportunities for higher education that young people have recently enjoyed. More sophisticated statistics (multivariate analysis, for example) much reduce the chance of such errors, but, especially in the simplified form reported by polling agencies in press releases and than mangled by journalists, it is still remarkably common to see causation assumed from some correlation. In defence of quantitative research, it should be noted that social statisticians know about these problems and try to address them. Most ethnographers simply offer us their unverifiable observations and unsupported explanations and expect us to trust their skill and judgement.

Page 27 of 32

 

Measuring Religion Conclusion To the believer, a religious faith may be unique. To the social researcher, it is a cultural product that is no more difficult to study than any other. Two topics have been intertwined in this chapter: the need for measurement in any sort of social research and the possibility of measurement in the study of empirical aspects of religion. As an undergraduate I was much impressed by Stanislav Andreski’s satirical attack on quantification in Social Science as Sorcery.44 I would then have unreservedly agreed with the Thompson quotation at the start of the chapter. A long research career has gradually persuaded me that quantification is not incompatible with Max Weber’s insistence on understanding. But my most embarrassing admission is that I failed to appreciate that quantitative social scientists were far better informed about the weaknesses in their methods than were the critics of those methods and that they had developed a large number of ways of correcting the weaknesses that, in my youthful and confident naiveté, I thought were fatal.45 As I argued in Chapter 2, there are very many facets of religion as it is lived that cannot be comprehended without measurement. (p.103) Are the Scots more or less likely than the English to be attracted to the charismatic movement? Are women more or less likely than men to be religious? Why are old people apparently more religious than young people? And have they always been or is it a product of ageing? If we cannot answer the simple question of ‘more or less’, the possibility of understanding, let alone explaining, pretty well anything is slight. And if we can accept comparative description expressed in such words as ‘more’ and ‘less’, there can be no principled objection to statistical description. The only principles that matter are that the statistics be appropriate and the descriptions accurate. Notes:

(1.) E. P. Thompson, ‘Anthropology and the Discipline of Historical Context’, Midlands History, 1 (1972), 48. (2.) M. Wood, Possession, Power and the New Age (Aldershot: Ashgate, 2007). (3.) B. Huss, ‘Spirituality: The Emergence of a New Cultural Category and its Challenge to the Religious and the Secular’, Journal of Contemporary Religion, 29 (2014), 47–60. (4.) C. Wright Mills, The Sociological Imagination (Oxford: Oxford University Press, 1999). (5.) J. M. Bryant, ‘Positivism Redivivus? A Critique of Recent Uncritical Proposals for Reforming Sociological Theory (and Related Foibles)’, Canadian Journal of Sociology, 17 (1992), 29–53; G. Swanson, The Birth of the Gods (Ann Arbor: University of Michigan Press, 1960).

Page 28 of 32

 

Measuring Religion (6.) T. M. Luhrmann, Persuasions of the Witch’s Craft (Oxford: Blackwell, 1989). (7.) T. Jenkins, ‘Congregational Cultures and the Boundaries of Identity’, in M. Guest, K. Tusting, and L. Woodhead (eds), Congregational Studies in the UK (Aldershot: Ashgate, 2004), 120. (8.) S. Stephens-Davidowitz, Everybody Lies: Big Data, New Data, and What the Internet Can Tell Us about Who We Really Are (London: Bloomsbury, 2017). (9.) A good example of entirely reasonable leading can be found in the transcripts in L. Jamieson and C. Toynbee, Country Bairns: Growing Up 1900– 1930 (Edinburgh: Edinburgh University Press, 1992). (10.) There is in survey work an equivalent of the unscrupulous ethnographer who makes up stories and quotations. I very briefly worked as a tree planter. As we were paid by the number of young trees taken from the store, a few planters cheated by burying handfuls in a large hole. Although it is rare, similar behaviour is sometimes found among survey interviewers. But, unlike fabrication in ethnography, interviewer deviation is itself studied, and a number of ingenious ways of preventing and detecting it have been devised. See P. Winker, N. Menold, and R. Porst (eds), Interviewers’ Deviations in Surveys: Impact, Reasons, Detection and Prevention (Frankfurt: Peter Lang, 2014). (11.) T. Horwood, The Future of the Catholic Church in Britain (London: Laicos, 2006), 13. (12.) Church of England, Canons of the Church of England, 7th edn (London: Church House, 2010), B 15, (accessed May 2014). (13.) P. Brierley, UK Religious Trends (London: Christian Research Association, 1999), table 2.14. (14.) There is an additional problem that David Voas raised in a discussion of why YouGov survey results tend to differ systematically from those of other polling agencies. The volunteers can be selected to be representative on the obvious demographic measures, but YouGov’s method may well attract volunteers who are subtly unusual: ‘Who knows whether people who think that it’s fun or profitable to complete online surveys are representative of anyone else in the characteristics that we’re interest in? Having the right mix of age, sex etc. doesn’t adjust for personality, beliefs, values, civic engagement, religion, etc.’ (D. Voas, personal communication, 18 June 2015). (15.) C. K. Hadaway, P. L. Marler, and M. Chaves, ‘What the Polls Don’t Show: A Closer Look at US Church Attendance’, American Sociological Review, 58 (1993), 741–52, and ‘Overreporting Church Attendance in America: Evidence Page 29 of 32

 

Measuring Religion that Demands the Same Verdict’, American Sociological Review, 63 (1998), 122– 30. (16.) Data kindly supplied by Alison Parks of NatCen. (17.) In brief, the 2001 TUS study shows that less than 9% of the population do anything that they or the data coders considered to be religion and that the vast majority of this (7.2%) happens on a Sunday morning, outside the home, and in the company of others; that is, it is churchgoing. I find it reassuring that this figure is close to those arrived at by Peter Brierley’s church censuses. (18.) For an example of what a good social scientist can do with such data, see C. D. Field, ‘Joining and Leaving British Methodism since the 1960s’, in L. J. Francis and Y. Katz (eds), Joining and Leaving Religion: Research Perspectives (Leominster: Gracewing, 2000), 57–85. (19.) I am grateful to David Voas for these data. (20.) ONS, Office for National Statistics, Marriage and Divorce Statistics Historical Series, vol. 31 (London: Office for National Statistics, 2005), table 3.29. (21.) C. D. Field, ‘The Ways we Say Goodbye’, British Religion in Numbers, 24 January 2011, (accessed March 2011). (22.) Increased digitization and use of the internet has created the possibility of analysing very large amounts of anonymized data on internet searches and spending. It may well be that future generations will find ways of analysing the causes and consequences of religious beliefs and behaviour with such data, but it is probably significant that Stephens-Davidowitz studies sports performance, political opinions, body-image concerns, health worries, and investment decisions, but, apart from a discussion of expressed wishes to kill Muslims, his only mention of religion comes in an examination of the odds on loans being repaid. Apparently mentioning ‘God’ in the application is strongly correlated with defaulting; see Stephens-Davidowitz, Everybody Lies, 259. His explanation is not that the Godly are less reliable than the rest; it is that those who appeal to God are paying insufficient attention to financial realities. (23.) C. Dahlgren, ‘Dödens semiotic. Gravstenen som odödlighetens medium’ (‘The Tombstone as a Medium of Immortality’), in C. Dahlgren, E. M. Hamberg and T. Pettersson (eds), ‘Religion och sociologi. Ett fruktbart möte. Festskrift till Göran Gustafsson’, Religio, 55 (2002), 97–111. (24.) See B. R. Wilson Religion in Secular Society: Fifty Years On (Oxford: Oxford University Press, 2016), appendix 2.

Page 30 of 32

 

Measuring Religion (25.) S. Gilliat-Ray, ‘The Fate of the Anglican Clergy and the Class of’97: Some Implications of the Changing Sociological Profile of Ordinands’, Journal of Contemporary Religion, 16 (2001), 216–17. (26.) N. MacErlean, ‘The Clergy’, Guardian, 25 November 2001, (accessed January 2016). (27.) ThisisMoney, ‘Compare your Pay to the National Average in your Job: League Table of Official UK Salaries across 400 Trades and Professions’, This Money, 28 January 2013, (accessed January 2016). (28.) Ipsos MORI, ‘Trust in Professions’, (accessed June 2016). (29.) D. Clarke, Between Pulpit and Pew: Folk Religion in a North Yorkshire Fishing Village (Cambridge: Cambridge University Press, 1982), 122–4. (30.) J. Wolfe and B. Jackson, ‘Anglican Resurgence: The Church of England in London’, in D. Goodhew (ed.), Church Growth in Britain: 1980 to the Present (Aldershot: Ashgate, 2012), 32. (31.) The Office of National Statistics NOMIS website gives the following figures for population of London aged 15 and above: 1991: 5,445,357; 2011: 6,643,200. (32.) L. A. Zurcher and R. G. Kirkpatrick, Citizens for Decency: Antipornography Crusades as Status Defense (Austin: Texas University Press, 1976). (33.) T. Doktór, ‘Churches, Sects and Invisible Religion in Central and Eastern Europe after the Transformation,’ in D. M. Jerolimov, S. Zrinščak, and I. Borowik (eds), Religion and Patterns of Social Transformation (Zagreb: Institute for Social Research, 2004), 299–313. (34.) The 2001 Scottish Social Attitudes survey showed a clear class difference: the only one of our four fields of ‘New Age’ spirituality in which interest from working-class women came close to matching that of middle-class women was divination. (35.) R. Prince and D. Riches, The New Age in Glastonbury (Oxford: Berghahn, 2001). (36.) Prince and Riches, New Age, p. xiii.

Page 31 of 32

 

Measuring Religion (37.) S. Bruce, Paisley: Religion and Politics in Northern Ireland (Oxford: Oxford University Press, 2007). This was finished in 2006, before the St Andrews negotiations that led to the creation of a devolved government with Paisley as First Minister and Martin McGuinness of Sinn Fein as Deputy First Minister were completed. I correctly anticipated Paisley would accept the deal; I was entirely wrong in thinking that in office he would shun McGuinness. Having been led by the Lord to accept devolution, he threw himself into his new role with such enthusiasm that he and McGuinness were nicknamed the ‘Chuckle Brothers’ after a local comedy act. (38.) I leave aside the trickier problem (on which the entire edifice of psychoanalysis is built) that people may not even be experts about themselves. (39.) P. Walls and R. Williams, ‘Sectarianism at Work: Accounts of Employment Discrimination against Irish Catholics in Scotland’, Ethnic and Racial Studies, 26 (2003), 632–62; S. Bruce, T. Glendinning, I. Paterson, and M. Rosie, ‘Religious Discrimination in Scotland: Fact or Myth?’, Ethnic and Racial Studies, 28 (2005), 151–68. (40.) This is discussed at length in S. Bruce, ‘Pluralism and Religious Vitality’, in S. Bruce (ed.), Religion and Modernization: Sociologists and Historians Debate the Secularization Thesis (Oxford: Oxford University Press, 1992), 170–94. (41.) D. Voas, D. Olson, and A. Crockett. ‘Religious Pluralism and Participation: Why Previous Research Is Wrong’, American Sociological Review, 67 (2002), 212–30. (42.) Stephens-Davidowitz, Everybody Lies, 246–71. (43.) As ever in matters statistical, I am grateful to David Voas for clarifying my vague suspicions. (44.) S. Andreski, Social Sciences as Sorcery (London: Penguin, 1972). (45.) For a good example of an insightful discussion of the issues involved in interpreting a body of data, see D. Voas, ‘Religious Involvement over the LifeCourse: Problems of Measurement and Classification’, Longitudinal and Lifecourse Studies, 6 (2015), 212–27.

Page 32 of 32

 

Bias in Social Research

Researching Religion: Why We Need Social Science Steve Bruce

Print publication date: 2018 Print ISBN-13: 9780198786580 Published to Oxford Scholarship Online: August 2018 DOI: 10.1093/oso/9780198786580.001.0001

Bias in Social Research Steve Bruce

DOI:10.1093/oso/9780198786580.003.0005

Abstract and Keywords It is now commonly argued that objective or value-neutral social science is impossible; that all research is corrupted by the biases and interests of the researcher. This chapter critiques various forms of that position and demonstrates both that the expected sources of bias are too narrow and that there is ample evidence they can be transcended. A social scientific study of religion is possible because researcher interests are often too complex to produce single preferences, because many topics are impervious to interests, because in many cases evidence trumps preferences, and because the competition between researchers allows subsequent generations to decide who has the better of any argument. Keywords:   value-neutrality, objectivity, bias, prejudice, community of scholars

Introduction This chapter is concerned with three questions. First, can researchers study some social phenomenon without their preferences, personal or ideological, distorting what they see and how they report it? Second, do researchers help readers make due allowance for potential bias by declaring such preferences? And, third, should the researcher take sides? I will argue affirmative, negative, and negative. The danger of contamination from bias is greatly exaggerated. Having researchers declare either specific allegiances or general orientations makes little difference to our ability to evaluate research. And taking sides is almost always impractical or unethical. The problem of bias in social research

Page 1 of 19

 

Bias in Social Research will be introduced with a critical discussion of an unusually clear argument for its prevalence.

Standpoints and Standards Mary Jo Neitz presents some interesting comments on the relationship between the personal religious commitments of researchers, the status of the religion they study, and the results of their research.1 In brief, she argues: ‘Researchers’ own locations—as an insider, an outsider, an apostate or an advocate—have consequences for the questions they ask, and to whom they address them, as well as how they interpret the data they collect.’2 Perfectly reasonably as she is an American writing for a largely American audience, Neitz’s argument is in the first instance concerned (p.109) with study of religion in a country where a large part of the population is religious, where there still is a dominant religious tradition, and where many academics are religious. Matters look rather different in Europe. That few Europeans are religious has a number of consequences that will be explored en route to a defence of the possibility of objective and value-neutral research. I will argue that the work of British sociologists of religion suggests that motives other than those produced by ‘location’ in Neitz’s sense may produce reasonably impartial research. I could write a lengthy defence of the ability of my American colleagues to transcend their personal preferences, but here I will make the general points with the UK examples I know well. Locations and Interests

In classic sociological mode, Neitz offers a four-box typology based on these two considerations: is the religion being studied dominant or marginal and is the researcher inside or outside the religion being studied (see Figure 5.1). Cell A: Neitz argues that ‘most work conducted by sociologists of religion in the United States belongs in this cell’, and she Figure 5.1. Locations matrix may well be right.3 That most Source: Neitz, ‘Insiders’, 131. US sociologists study mainstream religion and themselves have some religious commitment apparently has two deleterious consequences: ‘minority traditions may not show up in sufficient numbers to be included in the analysis’ and researchers will produce an inaccurately positive view of the benefits of the sort of religion they share.4 Cell C: Neitz believes this box to be small because people who were not raised in, or do not still belong to, some version of the dominant religious tradition ‘may be busy studying things they think are more important’. Nonetheless, those in this box are still likely to produce improperly positive evaluations of the Page 2 of 19

 

Bias in Social Research dominant religion because (p.110) ‘disciplinary practices, including the methodological and rhetorical ones discussed below, provide a set of taken-forgranted assumptions about the utility of the dominant religion’.5 Cell B: This cell consists of adherents studying their own marginal religions, and the pitfalls she mentions include the frosty reception professionals apparently give studies produced by insiders. It is worth noting the lack of symmetry in the problems she associates with cell A and cell B. For cell A, she worries that mainstream religion adherents will be biased in favour of their own faith, but for cell B she worries that the rest of the profession will ignore work by supporters of minority religions. Why is she not concerned that sociologists who are members of the Brethren, the Moonies, or the Scientologists will systematically distort their research so as to puff their preferred religion? My guess (and I will return to this later) is that Neitz shares the common sociologists’ sympathy for the underdog. Cell D: Finally we have marginal religions being studied by outsiders. Her focus here is new religious movements (NRMs). She believes that sociologists have often defended NRMs (especially in arguments about brainwashing and deprogramming). Although probably true, this does not sit easily with the claims made about cell A: having asserted that sociologists are more likely to approve of their own mainstream religious traditions, she now says that they defend marginal minority religions. My knowledge is limited, but the US scholars of NRMs with whom I am acquainted do not seem particularly different from those Neitz puts in her cell A, except that circumstance provided them with an exotic group to study. Neitz recognizes that, like all such models, her location matrix simplifies: Researchers have complex identities and may find themselves ‘inside’ on some dimensions and ‘outside’ on others. Similarly the line between ‘mainstream’ religions and marginal ones … is also one that can shift over time: there are many examples of sects that become mainstream denominations.6 Such difficulty of application is to be expected of any model that simplifies in order to make a point. However, too much stretching and one has to wonder if the model has identified the main variables. In the British context both axes of the location matrix are difficult to operationalize for reasons additional to those given by Neitz. (p.111) I will take the religious commitments of sociologists first and then consider the division of religion into mainstream and marginal.

The Standpoints of British Students of Religion An important British distinction, which figures only peripherally in Neitz’s paper, is the separation of the professional from the amateur sociologist. Quite why the sociology of religion should be relatively unpopular among British sociologists is Page 3 of 19

 

Bias in Social Research not obvious. It may be, as Neitz says in explaining a lack of US outsiders studying religion, that ‘people who have no emotional connection (positive, negative or even ambivalent) to religion may also be less likely to have an intellectual connection’.7 In the UK context, it is more likely to be a function of secularization. Emotional connections aside, people who think religion is of little importance as an independent variable for whatever they wish to explain are unlikely to study it, and in the UK religion has become so unpopular that such neglect is not unreasonable. In the 1950s there were a large number of community studies in the UK that gave significant space to religion. But when Marilyn Strathern (arguably one of the UK’s leading anthropologists) led a detailed study of a north Essex village in the late 1960s, she chose not to include religion in its purview.8 Between 1949 and 1952 Margaret Stacey conducted a detailed study of the town of Banbury, which, in its published form, devoted a long chapter to detailing the religious attachments of its residents and considering in particular the local political associations of various forms of religious identification: Conservatives tended to be Church of England while Labour attracted Methodists and other Nonconformists. In 1966 Stacey led a restudy, which was not published until ten years later. The second book gave only ten pages to religion, and two of those were taken up with a diagram. The second volume brought forward some of the religion data from the first so that comparisons could be made, but there were significant mistakes and unexplained changes in the transcription of the original religion data, and it seems clear—and the one surviving member of the team has confirmed this— that the 1966 restudy team was simply not that interested in religion.9 (p.112) Whatever the explanation for sociology’s neglect of religion, the result is that the bulk of apparently and putatively sociological research on religion in the UK in the twenty-first century is done by people who were not trained as social scientists and who do not work in social science departments. For example, of the UK-based participants in the 2014 annual British Sociological Association Sociology of Religion conference, probably only a quarter were sociologists (defined here as having at least one sociology degree). They were far outnumbered by theologians and graduates in religious studies: around 40 per cent. The only other discipline with more than a handful of representatives was anthropology (at 13 per cent). Among professional sociologists of religion, insiders are rare. David Martin is an ordained Anglican and Grace Davie describes herself as ‘a moderately active member of the Church of England’.10 Bryan Wilson, Roy Wallis, James Beckford, Eileen Barker, and I are outsiders. What is significant about our lack of personal religious commitments is that it shows a stimulus missing from Neitz’s model (which has consequences for her scepticism about value-neutral research): the importance of sociological problems as the motive for studying religion. These outsiders were drawn first to sociology as a discipline and then to the sociology of religion as a field of research. The reasons for both attractions are many. Roy Page 4 of 19

 

Bias in Social Research Wallis was attracted to sociology because it offered a way of understanding his own social background (a working-class lad who was the first of his family to go to university), because the people who taught it were people like him, and because its concerns were intellectually stimulating. He pursued the sociology of religion because he was interested in the role of charisma in leadership and persuasion and because he was impressed by Bryan Wilson’s scholarship. Barker’s route in was similarly both intellectual and personal: ‘When we had to choose our options … the sociology of religion, which was taught by David Martin, seemed more interesting than anything else on offer.’11 My interest in the sociology of religion came from my girlfriend giving me Peter Berger and Thomas Luckmann’s The Social Construction of Reality. That caused me to wonder how any minority, religious or secular, maintains a distinctive world view, and the most convenient deviant minority was the Christian Union: a group of evangelical students in a largely secular campus university. That such a productive cohort of professional sociologists were religious ‘outsiders’ has one important feature to (p.113) which I will return: they were professionally disinterested. Other considerations may have distorted their research—for instance, they may simply have been bad at their jobs—but it was not, as the conventional view of interests requires, distorted by their personal relationship to what they studied. As we will see shortly, the British group that possibly fits Neitz’s cell A are not sociologists interested in religion but scholars in Theology, Divinity, and Religious Studies departments who become interested in the sociology of religion (but not normally in sociology per se) as a source of ways of understanding the changing nature and position of religion. Arguably this is itself a function of secularization: Religious Studies, Theology, and Divinity departments have complemented their previous roles as exponents and immanent critics of religion with an interest in the social causes and consequences of religious change. That is, they have shifted from teaching religion to teaching about religion. The second major difference between the USA and UK concerns Neitz’s second axis: the distinction between dominant and marginal hardly works for the UK, where the most popular religion is now ‘None’ and where the proportion of the population that is involved in organized religious activities in a typical week is about 7 per cent. All religions in the UK are marginal. The two state churches— the Church of England and the Church of Scotland—retain some flummery (blessing royal weddings, for example), but they lost their state funding at the start of the twentieth century and they no longer have the power to stigmatize alternatives. The then highly deviant nineteenth-century sects are now entirely assimilated. BBC Radio Aberdeen’s road traffic reporter produced a significant (because unconsidered) sign of respectability-having-been-achieved when she referred to a particular junction on Aberdeen’s notoriously crowded ring road as ‘the Latter Day Saints junction’. The Mormons are now just the owners of a wellPage 5 of 19

 

Bias in Social Research placed spire. The Pentecostal sects of the 1920s and 1930s are now as mainstream as any denomination. The Conservative–Liberal Democrat coalition that came to power in 2011 offered full state funding to a wide variety of new schools under its ‘Free Schools’ programme. Its success probably owed a lot to the accidental need for the government to show it supported religious diversity, but a school in Skelmersdale run by Transcendental Meditation was admitted to the programme. More generally, yoga and meditation are part of mainstream culture, and mindfulness is now (p.114) presented within the National Health Service as a cure for a variety of mental illnesses. Paganism is accepted by the armed forces as a legitimate religion, and the once resolutely Presbyterian Aberdeen Press and Journal happily publicizes the annual visits of Kevin Carlyon (self-styled High Priest of British White Witches) to cast protective spells for the Loch Ness Monster. With David Voas, I have argued that the respectable/deviant distinction that informs the classic divide between church and sect and between denomination and cult is due for revision, and, with the exception of the recently imported religions of Islam, Hinduism, and Buddhism, the same case can be made against most of what is implied in any dominant and marginal distinction.12 The third major difference between the USA and UK concerns what it is that sociologists argue about. Neitz’s main example of researcher location distorting research concerns whether religion is good or not. This seems like a theological rather than a sociological question, but, even if we cast it in some Durkheimian form as the importance of a shared religion for social cohesion, it is hard to see it as a major concern of European students of religion: that ship has sailed. A better candidate would be secularization, which has exercised historians, sociologists, theologians, and religious studies scholars who have borrowed from the sociology of religion. The one religion that is obviously ‘marginal’ (and that has attracted relatively more research interest in the UK than in the USA) is Islam, but it has not been much studied by sociologists of religion. There are many reasons for this (including obstacles of language and restrictive gender roles), but one is that the subject has been treated under the rubric of ethnic relations and migration studies. This is speculation, but I suspect this may have a lot to do with a wish (supported by British Muslims) to separate the religious core of Islam from its currently unpopular cultural and political appurtenances. Critical Sociology and Value-Neutrality in Research

These observations about the sociology of religion in a largely secular society are intended as background to my major disagreement with the epistemological and methodological implications of Neitz’s location matrix. Neitz, like the participants in a recent roundtable published by Critical Research on Religion, accepts ‘the demise of the (p.115) once-hoped-for goals of objectivity and

Page 6 of 19

 

Bias in Social Research neutrality’.13 Despite this, she makes a series of testable claims, so I feel justified in testing them. But first it is necessary to consider what is meant by ‘critical sociology’. As someone who still hopes for objectivity, I reject the shared view of the roundtable participants that a critical approach means shifting attention (I assume this is a baseball metaphor) from the ball to the pitcher, and I will return to that in a moment. For the realist, critical sociology used to mean a toneddown version of conventional Marxist epistemology. The robust application of ‘scientific’ Marxist thought would deny objectivity to any research that was not itself the robust application of Marxist thought. The softer agenda promoted by Antonio Gramsci and the Frankfurt School does not deny the possibility of truth outside Marxist assumptions but it does suppose that power distorts: in any field there are hegemonic interests that cloud the eyes of most scholars and should be exposed.14 This is what Neitz means when she says that we should attend to the standpoints of researchers and their relationship to those they research. Hence critical sociology of religion requires some dominant religion to which sociologists belong and which they study. To fill out her cell A location: a sociologist who is an elder of a Baptist Church and whose research on mainstream Christianity is funded by a religious foundation will improperly conclude that mainstream Christianity is socially beneficial. The ‘improperly’ in that construction is essential: we could have no objections if we believed (because, for example, every other sort of researcher concurred) that the conclusion of our Baptist elder’s research was reasonable. To make the standpoint case we would have to find research that led to unreasonable conclusions that suited the insider’s personal religious interests. Research on religion by professional British sociologists offers very few examples. Grace Davie—an insider—has gone to greater lengths than most to promote alternatives to the secularization view. First she argued that the decline in the churches was primarily a loss of faith in joining organizations rather than a loss of faith: hence the British were ‘believing but not belonging’. Then she argued that, while conventional religion might have declined, ‘vicarious religion’ was alive and well. By this she meant that an apparent willingness of the British to have the churches and church officials do religion on their behalf showed that religious sentiment persisted.15 Both arguments have been taken up enthusiastically by Christian church officials, and Davie is a popular speaker in church circles. (p.116) But neither argument is so obviously wrong that it must be explained by personal religious sentiment clouding professional judgement. David Martin, perhaps the most prolific and influential insider, made a series of trenchant criticisms of the secularization thesis in the early 1960s, long before Rodney Stark discovered it was wrong. But, in an important exception to the expected line-up of interests and arguments, he also produced one of the most

Page 7 of 19

 

Bias in Social Research sophisticated accounts of secularization: a book entitled A General Theory of Secularization hardly places its author on the list of secularization deniers.16 Is there a coincidence of personal interests among the non-religious sociologists who work within the secularization paradigm? Few have committed themselves to print on this point, but Wilson offers a good example of a man who could describe and explain changes that he personally found unwelcome. Although he is routinely cited as the fons et origo of the modern secularization paradigm, he authored one of the most dismal projections of society without religion. He expected that, just as the external exoskeleton of the beetle serves the same purpose as the vertebrate’s internal skeleton, the maintenance of social order in the absence of shared divinely endorsed moral values would require external supervision: the all-seeing eye of God-internalized-as-conscience was being replaced by the all-seeing eye of the surveillance camera.17 Where one may find better support for Neitz’s proposition is among the theologians and religious studies scholars who borrow from the sociology of religion. The man who mistakenly claimed church growth in York is an Anglican clergyman.18 The survey of student religiosity mentioned at the start of Chapter 2 was led by a theologian, and only one of the four team members had significant social science training.19 With no thought to the certainty that those who so completed the questionnaire might be unrepresentative precisely in their interest in religion, the results were initially presented as a description of British students in general, who, remarkably, appeared to be vastly more religious than the population at large. Later reports corrected the mistake, but not before a number of Christian journals and websites had latched on to what looked like the first shoots of a religious revival. As it further demonstrates that ideological preference need not match research conclusions, it is worth noting that Peter Brierley, an evangelical Christian and social statistician, spotted the implausibility of the data and pointed it out in his regular summary of religious research.20 (p.117) Other examples could be added to suggest that theologians straying into sociology may be insufficiently critical of findings that apparently fit their personal wishes. However, before we suppose that personal preference creates self-serving credulity, we should consider an alternative: the problem may simply be professional inexperience. Social scientists used to presenting descriptive statistics as a proportion of the total population available would not have seen the decline of church attendance in York as growth, and people trained in survey design and administration would not have generalized from a patently unrepresentative sample. That is, expertise may be as or more important than standpoint.

Page 8 of 19

 

Bias in Social Research To summarize my argument thus far, I can find what might be British examples of Neitz’s cell A—insiders producing misleadingly positive research findings that they accept because they like them—but these concern non-sociologists straying outside their primary area of expertise, and the explanation may well lie in lack of familiarity with technical matters rather than in standpoint. Which brings me to another important point in Neitz’s presentation. The use of the term ‘standpoint’, pioneered in feminist sociology, cheats the argument, because, as a locational metaphor, it implies that everyone must have one and only one. We must have a location and, as we are not Schrödinger’s cat, we cannot be in two places at once. We all have a standpoint whether we admit it or not, so we had better admit it. Arguably that might work for gender, because (with few exceptions) we all have one and only one. But even with gender there is a conceptual difficulty: the assumption that the characteristic of gender brings with it a shared set of interests. Feminists and others have argued about this without resolution beyond the recognition that the interests of both genders will vary internally with age, class, ethnicity, wealth, and so on. That is, the issues on which there is a clear and singular female standpoint are likely to be few. As a group, the religious may be more coherent than the non-religious, but neither obviously shares a common and limited set of interests, even in the study of religion. At this point the sceptic may shift to Neitz’s subsidiary case, which is that the dominance of certain methods (surveys and statistics are singled out) leads to ideological distortion: ‘quantitative studies with large samples are more likely [than ethnographic work] to reflect dominant values’.21 Such work is almost entirely absent from British sociology of religion. But, popular or unpopular, it is hard to see (p.118) a causal connection between stats and an especial likelihood to reflect dominant values, unless it is that large-scale surveys require large-scale funding, and the funders may be reluctant to see certain lines of enquiry pursued. I cannot speak for the USA, but the UK offers little evidence for that fear. It is true that ‘cheap and cheerful’ surveys commissioned by Christian organizations primarily to get some press attention suspiciously often manage to find something to be cheerful about, but the larger commercial polling organizations have been quite happy to ask questions that produce such findings as that most Britons think religion is harmful.22 And recurrent social science polls have asked questions that allow us to show that most Britons do not want religious leaders to have more influence and that most Britons prefer ‘non-religious’ to ‘deeply religious’ people.23 The sceptic can then shift ground to argue (as Thomas Kuhn famously did in his notion of paradigms in scientific research) that, while professional sociologists of religion may have no personal interest in religion, they will have a professional interest in defending previous positions. But, given that such famous sociologists as Peter Berger and Rodney Stark did their careers no harm at all by recanting Page 9 of 19

 

Bias in Social Research their early support for the secularization thesis, it seems improper to assert that professional inertia is necessarily a source of bias. Finally I come to a neglected ground for a defence of the still-hoped-for goals of objectivity and neutrality: the absence from many research topics of any feature likely to elicit systematic and collective distortion. For interests to contaminate research, it is not enough that scholars have interests: it must also be that the research involves matters that involve those interests. For example, we can imagine that a subject such as the social role of the Catholic Church in Africa (especially its refusal to promote contraception) offers an obvious powerchallenging role for critical sociology, but much of what concerns the sociology of religion in a largely secular country seems unlikely to be seen consistently differently from different standpoints. The following are topics that have been covered in my sociology of religion course: the role of the Reformation in modernization, the extent and causes of secularization, the use made of religion in cultural defence and cultural transition, the relationship between types of religious beliefs and types of religious organization, social mobility and religious change, the nature of charisma, the appeal of new religious movements, the nature and appeal of New Age spirituality, difficulties of identifying and measuring religious commitment, (p.119) explanations of conversion, and the gender gaps in religion and spirituality. A particular standpoint might produce a consistent view on some, but others are either so technical or so broad that one cannot reason a standard line. That one believes, for example, that the individualism, egalitarianism, and literacy encouraged by the Reformation played some part in modernization gives no clue as to how one would view competing explanations of religious conversion. And whether the Bromley and Shupe role theory of conversion offers a more convincing explanation than the Lofland and Stark value-added seven stages model (discussed in Chapter 7) seems impossible to shoehorn into any conventional model of standpoints or interests. Objectivity in research need not rest on the saintly detachment of the scholar; it can be an accidental consequence of the research subject matter being too multifaceted to produce simple alternative conclusions that would fit with shared individual religious preferences.

Professional Ambition and Personality There are three important extraneous influences on research that Neitz does not mention: professional rivalry, personal ambition, and personality. Professional rivalry is often cited by those postmodern relativists who see all academic work as ‘constructing’ (rather than understanding and explaining) its research subjects. Although he rejects Bruno Latour’s belief that analysts have a free hand in such construction, Mike Savage’s rich history of British social research explains many of its changing features as resulting, not just from improving understanding, but also from changes in the relative power of interested professional groups. Change should be seen ‘not as solely a matter of Page 10 of 19

 

Bias in Social Research scientific progress or the advance of knowledge, but also as related to the remaking of cultural hierarchies, and the re-drawing of class, gender, and national boundaries’.24 Provided we remember that such a claim can be made only by someone who separates knowledge from interests and thus itself refutes relativism, it is difficult to take exception to this view. My response is that it simply encourages us to apply to social observers the same questions we apply to what they observe. Perhaps because success is both more clearly defined and better rewarded in the natural than in the social sciences, ambition seems (p.120) less of a problem in studies of religion than in, say, trials of new drugs, but its potentially distorting effect is still worth mentioning because it shows how narrow is the view of bias that Neitz entertains. A possible example is the career of Carlos Castaneda. He wrote The Teachings of Don Juan, A Separate Reality, and Journey to Ixtlan while he was an anthropology student at the University of California, purportedly drawing on research diaries of his experience of a Yaqui Indian shaman Don Juan Matus. The books earned Castaneda his undergraduate and postgraduate degrees; they also earned him a great deal of money. Their veracity was soon questioned. Richard de Mille’s Castaneda’s Journey argued that ‘logical or chronological errors in the narrative constitute the best evidence that Castaneda’s books are works of fiction’.25 Other anthropologists pointed to a large number of inconsistencies and implausibilities.26 It seems pretty obvious (to put it minimally) that Castaneda exaggerated and invented. One can only guess his motives, but two degrees, fame, and wealth seem like good incentives. A second unconsidered issue is the effect of personality. Because the problem is rarely discussed, I have to draw on my limited personal experience for examples. For fifty years since its first publication in 1966, Bryan Wilson’s Religion in Secular Society has been cited as the prime source of the modern secularization thesis, but his work on sects has also been influential. That research was characterized by an unusual degree of familiarity with, and an unusual sympathy for, small Protestant sects. He was active in defending the Exclusive Brethren: he wrote a number of popular newspaper and magazine articles criticizing the bishops of the Church of England who sat in the British parliament’s upper house for their mean-spirited (and successful) attempt to show public disapproval for the Exclusives. The matter in hand was an accident. The Brethren took an unusually extreme view of the Pauline injunction: ‘be ye not yoked with unbelievers’. Most Christians take this to mean one should not marry a non-Christian, but the Brethren take it to forbid any sort of affiliation with any body that contains non-Brethren people. They will not attend university, for example, or belong to trade unions. Legislation introduced in the 1963–4 parliamentary session created a new system for testing and approving the competence of pharmacists. The Brethren argued for an amendment to the Act that would allow them to sit all the professional examinations but not actually join the new Royal Pharmaceutical Society. The bishops in the House of Lords Page 11 of 19

 

Bias in Social Research opposed (p.121) the amendment, and it fell. As a result, Brethren pharmacists had to give up their businesses. I do not know whether Wilson volunteered to defend the Brethren or was invited by them, but he tried to help. That, combined with his sympathetic presentation of the belief systems of the Elim Pentecostal, Christian Science, and Christadelphian sects, had led me to suppose that Wilson had been raised in one of these groups. But this is not the case. So that the adults could enjoy some privacy in a tiny house, he was sent to a Methodist Sunday school, but his religious socialization seems to have been brief and shallow. The more I have learnt about Wilson from those knew him best—he was such a private person few knew him very well—the more I have come to suspect that his sympathy for Protestant sectarians came more from a similarity of personality than from any shared ideology. I think, consciously or subconsciously, he saw something of his own driven, buttoned-down, and sublimated personality in the sectarian. I suspect a similar harmony of researcher and research subject personality in my own work. In at least three research themes—the sociological differences between conservative and liberal Protestantism, the career of the Revd Ian Paisley in Northern Ireland, and the fate of New Age spirituality—I have found myself differing from colleagues in ways that might be summarized as being unusually sympathetic to puritanical and doctrinaire religion. That obviously owes nothing to what Neitz would call my religious standpoint because I do not have one, but it may reflect a subtle consequence of a military upbringing. A third example concerns Roy Wallis. In a career cut short by a shift into university management and then early death, Wallis made important contributions to the sociology of new religious movements and that realm where the human potential movement shades into contemporary spirituality. As a student I recall being particularly impressed with his explanation of the appeal of movements, such as Esalen, est, and Insight, which promised to help middleclass strivers who had deferred gratification and subordinated personal relationships to the ambitious pursuit of career success to regain contact with their spiritual sides and create authentic trusting human relationships.27 When I got to know Wallis better, I realized that his description of the potential customer for world-affirming new religions was in many respects autobiographical. Wallis was far more in tune with human potential movements such as est and Insight than I ever was, and that sympathy came not from ideological standpoint but from personality. (p.122) In the abstract, it is impossible to say whether the personality traits of social researchers are helpful (because they allow empathetic understanding) or unhelpful (because they bias research). But we can conclude that professional ambition and personality are as likely as religious partisanship to distort research.

Page 12 of 19

 

Bias in Social Research Confession or Bragging All of this suggests that there is little to be gained from researchers declaring their most obvious identities, interests, or ideological sympathies. A biography sufficiently detailed and insightful to explain ideological preferences, professional ambitions, and personality traits might help us understand what appear to be errors, but, as there are many research topics that are unlikely to be seen in any particular way because of a particular standpoint, and as the empirical record shows that many social scientists are perfectly capable of coming to conclusions that clash with their own preferences, there seems little point in either confessing to, or bragging about, those preferences. Furthermore, the introductory confession may harm research more than it enlightens the reader of that research. Researchers may feel that, having handed the reader the key to decoding their work, they need not try to be objective; that is, it may serve as a licence for the researcher to be straightforwardly partisan. There is something related in much modern anthropology where the abandonment of the search for objectivity has led to research reports telling us little about the matter in hand and a great deal about how the researcher feels about or reacts to the matter in hand.28 Such personal reflections can be interesting—if we wish to learn a lot about the personal growth of young anthropologists—but they will disappoint readers of ethnographies who expect to learn more about those studied than about the author.

Siding with the Underdog Strictly speaking, partisanship and bias are not the same thing. Researchers could conduct research in an entirely value-neutral manner and come to conclusions quite unaffected by their personal preferences (p.123) but then— as a separate matter—act in a partisan manner towards their subjects. To what extent taking sides leads to bias will be discussed shortly, but first I will question the proposition that social researchers should take sides. In 1966 the renowned American sociologist of deviance Howard S. Becker delivered a conference address published as ‘Whose side are we on?’.29 He argued that taking sides is necessary because value-neutral research is impossible and so the only issue is which side we back. And he thought we should back the underdog. In contemporary language, researchers should articulate the interests of their subjects. A popular methods text begins by saying that ‘the researcher is merely here in the first instance to give “voice” to other people’.30 This attitude has become popular with feminist sociologists and with anthropologists.31 According to Dorothy Smith, feminist sociology replaces the commitment to objectivity with a commitment to making women’s voices heard, to deconstructing gendered power relations, and to empowering women.32 Page 13 of 19

 

Bias in Social Research There are practical and ethical difficulties with this. The practical problem is the lack of cohesion and consistency among those to whom we are supposed to ‘give voice’. How can the student of conservative Scottish Presbyterianism take sides when the conservative opponents of the national Church of Scotland divide into Reformed Presbyterian Church, Free Church, Free Presbyterian Church, Free Church (Continuing), and Associate Presbytery and to those we would have to add Brethren, Baptists, Pentecostalists, Congregationalists, and autonomous evangelical churches? As I note in Chapter 6, even something as dull as organizational typing can offend someone: it is common among conservative Protestant sects for each to eject the ‘sect’ label for themselves while happily applying it to others who, to the outsider, seem remarkably similar. The project of ‘giving voice’ to some population is rendered impossible when such people do not agree on what their voice should say. The injunction to take sides is implicitly justified by the assumed immorality of benefiting from those we study without giving something back. Reciprocity is an attractive notion, but it leads into an ethical minefield. Much of my time in Northern Ireland was spent studying terrorist organizations. Is anyone seriously suggesting that I should have repaid my research subjects for their cooperation by helping them become better terrorists? Anne Oakley may have had no trouble sympathizing with her pregnant women and young mothers.33 Russell and Rebecca Dobash may have had no trouble combining their (p.124) research on domestic violence against women with social policy promotion of the interests of battered wives.34 Although my career benefited from their cooperation, I could not assist loyalist paramilitaries, and I hope no one would suggest I should have so done. If taken seriously, the injunction to take sides means either we become accessories to illegal and immoral behaviour or we study only nice folks. We study battered wives but not terrorists, or marijuana smokers but not heroin dealers. Michael Burawoy, in his proposal for ‘public sociology’, thinks sociologists should support NGOs and social movements.35 He does not provide an exhaustive list of the latter, but I am pretty sure it would not include xenophobic, Islamaphobic, isolationist, and extreme right-wing movements, even when they are more popular than the left-wing movements he does like. Even if we could agree who are the good guys (and, as I have just suggested, that is rendered impossible by the internal variations within any population for which we might want to ‘give voice’), the principle of taking sides would impose an ethically unprincipled constraint on research. Although partisanship and bias in research are analytically separable, in practice they are very often closely allied. Some partisans justify their partiality by arguing that objectivity is impossible. Other argue that, even were it possible, it would be unethical: the Dobash position. What I conclude from wanting to be as objective as possible, from appreciating that our research subjects rarely share a Page 14 of 19

 

Bias in Social Research single interest that we could consistently represent, and from a desire to prevent social science being confined to the study of Snow White and the Seven Dwarfs is that the posture that best suits the interests of the greatest number is research that is as objective as we can make it. Anyone we study who does not appreciate being honestly and accurately represented may complain, but, as some will complain unless we simply endorse their views, a consistent striving for neutrality will at least allow us to tell our subjects, ‘like it or not, that is really how you look’. Some research subjects have such an exalted and misguided image of themselves that any description by anyone outside their inner circle will be rejected as unfair: Scientology, for example, is notoriously hostile to outsiders.36 But over the years my position has found support in unlikely quarters. After God Save Ulster: The Religion and Politics of Paisleyism was published, Ian Paisley phoned me and said: ‘Considering you’re a communist or a crypto-communist and a Fenian or a crypto-Fenian’—at that point I realized he was joking—‘it was pretty fair’.37

(p.125) Why we Should Study People we don’t Like In an earlier discussion of surveys I observed that specially commissioned oneoff opinion polls have a suspicious habit of producing at least some results that the funders can claim as supporting their values or agenda. Something similar could be said about almost all ethnographic studies of religious and spiritual groups and movements. That such studies almost always produce sympathetic portraits should be a cause for reflection. My guess is that it results from a combination of two factors: topic choice and researcher seduction. Perhaps naturally, many scholars choose to study people for whom they already have considerable sympathy. And that is compounded by a natural tendency to ‘go native’. This is not Neitz’s standpoint case; it is much more complex than ideological agreement. We spend four or five years learning some group’s language and culture. Extensive social contacts produce friendships and loyalties. Whether we are aware of it or not, we imagine how what we write may be received by those we write about. And, while such reflection may not produce the wholesale distortion of a natural scientist who fakes lab results, it may cause us to hold back particular impressions or observations because they might hurt the feelings of someone we have come to like. Looking back on my own work, I can see many places where I have been a little more generous to people I liked than perhaps I should have been. With only a little flippancy, I suggest the solution is to study people whose beliefs and values we do not share. Then, instead of reinforcing each other to provide inappropriately positive views of our research topics, our personal feelings and our professional commitment to understanding will balance out.

Page 15 of 19

 

Bias in Social Research Conclusion None of these discussions should be taken to mean that objectivity is easy, but there seems little point in social science research that makes no such effort or that sees its role as representing those it studies. It is perfectly reasonable for scholars to engage in hopefully value-neutral research and then, as a separate enterprise, to represent the interests of some of those they study. But the separation of research and representation is vital. Academics in liberal democracies enjoy the (p.126) privilege of being able to research with very little external interference. For them to prefer taking sides over value-neutrality is to waste a rare opportunity. There is quite enough partisanship in the world without those who have the luxury of objectivity giving that up to add yet another ‘voice’ to the already loud clamour of contending agendas and perspectives. That is especially the case for the study of religion, because most religions are quite vocal enough without social scientists acting as cheerleaders. Notes:

(1.) M. J. Neitz, ‘Insiders, Outsiders, Advocates and Apostates and the Religions they Study: Location and the Sociology of Religion’, Critical Research on Religion, 1 (2013), 129–40. (2.) Neitz, ‘Insiders’, 131. (3.) Neitz, ‘Insiders’, 133. (4.) Neitz, ‘Insiders’, 134. (5.) Neitz, ‘Insiders’, 134. (6.) Neitz, ‘Insiders’, 131. (7.) Neitz, ‘Insiders’, 131. (8.) M. Strathern. Kinship at the Core: An Anthropology of Elmdon, a Village in North-West Essex in the Nineteen-Sixties (Cambridge: Cambridge University Press, 1981). (9.) For a detailed discussion of Stacey’s two studies, see S. Bruce, ‘A Sociology Classic Revisited: Religion in Banbury’, Sociological Review, 59 (2001), 201–22. (10.) G. Davie, ‘Thinking Sociologically about Religion: Discerning and Explaining Pattern’, in T. Hjelm and P. Zuckerman (eds), Studying Religion and Society: Sociological Self-Portraits (New York: Routledge, 2013), 122. (11.) E. Barker, ‘Doing Sociology: Confessions of a Professional Stranger’, in Hjelm and Zuckerman (eds), Studying Religion and Society, 41.

Page 16 of 19

 

Bias in Social Research (12.) S. Bruce and D. Voas, ‘Secularization and Typologies of Religious Organizations’, Journal of Contemporary Religion, 22 (2007), 1–17. (13.) C. Martin, R. McCutcheon, M. R. Miller, S. Ramey, K. M. Simmons, L. D. Smith, and V. Touna, ‘Keeping “Critical” Critical: A Conversation from Culture on the Edge’, Critical Research on Religion, 2 (2014), 302. (14.) My view of critical sociology has benefited from conversations with my colleague Andrew McKinnon. (15.) G. Davie, Religion in Britain since 1945 (Oxford: Blackwell, 1994); Religion in Modern Europe: A Memory Mutates (Oxford: Oxford University Press, 2000). (16.) D. Martin, ‘Towards Eliminating the Concept of Secularization’, in J. Gould (ed.), Penguin Survey of the Social Sciences (London: Penguin, 1965), 169–82; A General Theory of Secularization (Oxford: Basil Blackwell, 1978). (17.) B. R. Wilson, ‘Morality in the Evolution of the Modern Social System’, British Journal of Sociology, 36 (1985), 315–52. (18.) D. Goodhew (ed.), Church Growth in Britain 1980 to the Present (Aldershot: Ashgate, 2012). (19.) M. Guest, K. Aune, S. Sharma, and R. Warner, Christianity and the University Experience: Understanding Student Faith (London: Bloomsbury Academic, 2013). (20.) P. Brierley, ‘Student Religion’, Future First (April 2011), 3. (21.) Neitz, ‘Insiders’, 217. (22.) YouGov, ‘John Humphrys Religion Survey, February 2007’, (accessed June 2010). (23.) S. Bruce and T. Glendinning, ‘Privatization and De-Privatization’, Journal for the Scientific Study of Religion, 50 (2011), 503–16; B. Clements, ‘The Sources of Public Feelings towards Religious Groups in Britain’, Journal of Contemporary Religion, 27 (2012), 419–33. (24.) M. Savage, Identities and Social Change in Britain since 1940: The Politics of Method (Oxford: Oxford University Press, 2010), 237. (25.) R. De Mille, Castaneda’s Journey: The Power and the Allegory (Santa Barbara, CA: Capra Press, 1976), 166.

Page 17 of 19

 

Bias in Social Research (26.) J. H. Kelley, Yaqui Women: Contemporary Life Histories (Lincoln, NE: University of Nebraska Press, 1978); M. T. Painter, With Good Heart: Yaqui Beliefs and Ceremonies in Pascua Village (Tucson, AZ: University of Arizona Press, 1986). (27.) R. Wallis, ‘Inside Insight’, New Humanist, 95 (1979), 94. (28.) J. Clifford and G. E. Marcus (eds), Writing Culture: The Poetics and Politics of Ethnography (Berkeley and Los Angeles: University of California Press, 1986); B. Scholte, ‘Toward a Reflexive and Critical Anthropology’, in D. Hymes (ed.), Reinventing Anthropology (Ann Arbor: University of Michigan, 1999), 430–75. (29.) H. S. Becker, ‘Whose Side Are We On?’, Social Problems, 14 (1967), 239–47. (30.) K. Plummer, Documents of Life (London: George Allen and Unwin, 1983), 1. (31.) For a feminist example, see L. Stanley and S. Wise, Breaking Out: Feminist Consciousness and Feminist Research (London: Routledge, 1983). For a collection of anthropological writing, see H. Armbruster and A Laerke (eds), Take Sides: Ethics, Politics and Fieldwork in Anthropology (Oxford: Berghahn, 2008). (32.) D. E. Smith, ‘Women’s Perspective as a Radical Critique of Sociology’, in S. Harding (ed.), Feminism and Methodology (Bloomington, IN: Indiana University Press, 1987), 84–96. (33.) A. Oakley, ‘Gender, Methodology, and People’s Ways of Knowing: Some Problems with Feminism and the Paradigm Debate in Social Science’, Sociology, 32 (1998), 707–32. For a critique of Oakley’s attitude to quantitative research, see G. Letherby, ‘Quoting and Counting: An Autobiographical Response to Oakley’, Sociology, 38 (2004), 175–89. Oakley reflects on her career in Patriarchy, Gender and Social Science (Cambridge: Polity Press, 2014). (34.) R. E. Dobash and R. P. Dobash, Violence against Wives: A Case against the Patriarchy (London: Open Books, 1980). (35.) M. Burawoy, ‘For Public Sociology’, American Sociological Review, 70 (2005), 4–28. (36.) For a detailed account of how Roy Wallis was harassed by Scientologists, see R. Wallis, ‘The Moral Career of a Research Project’, in C. Bell and H. Newby (eds), Doing Sociological Research (London: George Allen and Unwin, 1977), 149–67. (37.) The original Fenians were a nineteenth-century forerunner of the IRA, and the term is still used by Ulster Protestants as a derogatory way of referring to Irish nationalists.

Page 18 of 19

 

Bias in Social Research

Page 19 of 19

 

Ethics in Social Research

Researching Religion: Why We Need Social Science Steve Bruce

Print publication date: 2018 Print ISBN-13: 9780198786580 Published to Oxford Scholarship Online: August 2018 DOI: 10.1093/oso/9780198786580.001.0001

Ethics in Social Research Steve Bruce

DOI:10.1093/oso/9780198786580.003.0006

Abstract and Keywords It is right that social researchers consider the ethical implications of their work, but discussion of research ethics has been distorted by the primacy of the ‘informed consent’ model for policing medical interventions. It is remarkably rare for the data collection phase of social research to be in any sense harmful, and in most cases seeking consent from, say, members of a church congregation would disrupt the naturally occurring phenomena we wish to study. More relevant is the way we report our research. It is in the disparity between how people would like to see themselves described and explained and how the social researcher describes and explains them that we find the greatest potential for ill-feeling, and even here it is slight. Keywords:   social research, research ethics, ethics of data collection, ethics of reporting, informed consent

Introduction For perfectly good reasons, universities and research funders have become ever more concerned about the ethics of social research. Unfortunately, the principles that express such concern are often borrowed from the model of supervising medical research, with little or no thought to how well such a model fits social science.1 This chapter will consider the ethical problems of various sorts of social research. I will argue that the possibility of such research causing serious harm is so remote it need not much concern us. What is more likely is that the way we write about people may offend (or, to be precise, give opportunities for offence to be taken). However, as people may take offence at social science characterizations of, or explanations of the actions of, any group to which they Page 1 of 21

 

Ethics in Social Research can feel they belong, the problem arises with all social science writing and is not confined to reports of specific projects. Moreover, it is a perpetual feature of all commentary on social life. As a non-religious social scientist I could readily take offence at what religious people say about both the non-religious and social scientists. That such irritation opportunities are commonplace suggests that no special protection against them is justified. To confine the topic so that some aspects can be discussed at length, I will say little about entirely covert research. In certain circumstances, a good case can be made for spying on people. Even in the usually pacific sociology of religion, there are mad, bad, and dangerous people who should be exposed. Any researcher who had infiltrated and exposed the Jonestown community (over nine hundred dead in 1978) or Heaven’s Gate (thirty-nine dead in 1997) before lives were lost would have been a hero. Less dramatically, (p.130) a good case can be made for using deceit where it is necessary to gain access, to prevent behaviour being changed by the presence of an outsider, or to allow the researcher to share the experiences of the researched, and—this is the vital condition—where the topic is important enough. The ethics of obviously covert research are in many senses simpler than the ethical concerns of the quasi-covert research I will describe in this chapter. Most of us would agree that covert research is in principle unethical because it involves behaving in ways that we would not condone (or at least not boast of) in our non-work lives.2 We defend its use where research could not be done without deceit and where the benefits outweigh the deleterious consequences for general morality. The Branch Davidians offer a good example. This small sect had its origins in a split from the Seventh Day Adventist movement in the 1930s. Its distinctive features included communal living in a Waco, Texas, compound, belief that Christ’s return to earth was imminent, subordination to the erratic instructions of dictatorial leader David Koresh, and possession of what, even for Texas, was a lot of guns and explosives. In 1993, concerned that people were being held against their will, various law-enforcement agencies laid siege to the compound. This reinforced the Davidians’ belief that they were living in the end times. Precisely what brought the siege to an end is disputed, as is the extent to which those who died were volunteers rather than victims of Koresh’s megalomania, but, after two months, the officials stormed the compound and in the ensuing violence eighty-three Branch Davidians and four law officers died.3 Although I would not encourage any inexperienced student to join such a strange, secretive, and dangerous sect, covert research would clearly be justified in such a case. The justification for covert research is its fruits. Laud Humphreys almost certainly could not have studied homosexual liaisons in public toilets had he not adopted the role of a ‘watch queen’ (someone who likes to observe but not take part) and then used a personal contact to identify participants from their car Page 2 of 21

 

Ethics in Social Research number plates. Tearoom Trade tells us a great deal about married men who engaged periodically in homosexual acts and thus makes an important contribution to our understanding of sexuality. And it remains a classic of the genre.4 Unlike homosexuality in the 1960s, selling double-glazing is not illegal, but it involves sharp practices that those involved are very unlikely to discuss with outsiders. Hence John Bone decided to combine his (p.131) studies and need for a part-time job by working part-time first as a window and then as a fitted kitchen salesman. His The Hard Sell offers considerable insights into the use of interactional social pressure in sales that would not have been available to anyone who was not involved in the business.5 But, as entirely covert research is rare in the study of religion, I will leave it to one side and concentrate on the ethical issues of a more common research style.

Unobtrusive Observation and Disguised Participation Most historical and quantitative research raises no particular ethical problems because these are taken care of at one remove by our sources. In the secondary analysis of survey data, relevant ethical issues will have been resolved by the original researchers and by the keepers of the UK Data Archive.6 Likewise with archival work in local records offices: those depositing and managing collections set the conditions for their use. In one case, a collection of a solicitor’s office papers had been closed for fifty years. In another, access to Methodist chapel records required permission from the circuit superintendent, which he readily gave. An ethics committee (which shall remain anonymous) was troubled by my fieldwork, which, like much research on religion, involved the unobtrusive observation of, and participation in, ongoing social life. It had three components: (a) attending church services and other public religious events; (b) engaging members of religious bodies and of the general public in chat about their religious commitments when the opportunity naturally arose; and (c) taking note of significant things seen and heard as I went about my life, all without declaring my role as a social researcher or announcing the possibility that I might write about these encounters. I will add a little more detail. Whenever I can, I attend religious services. I unobtrusively make a few notes on cards in my Bible, or Book of Common Prayer, or on the provided Order of Service (as, for rather different reasons, do other members of many congregations), and I sometimes use such notes to illustrate some general point in writing intended for publication. More often than not the ‘data’ thus (p.132) collected simply improve the understanding that I bring to bear on such more formally acquired data as survey results. The second type of work, the directed casual conversation, though it is rarely discussed in methods texts, can be remarkably fruitful. Most people have views about religion in general or about specific types of religion that they are only too Page 3 of 21

 

Ethics in Social Research happy to share when given the slightest encouragement or, in the case of London taxi drivers, entirely without encouragement. Although such material is flattered by the term ‘data’, it is highly informative. It is thirty years of chatting to strangers that convinces me that most people in the UK, even regular churchgoers, have little knowledge of, or interest in, such once-highly-charged debates as how churches, sects, and denominations should be structured, the value of professional clergy, the proper age for baptism, the proper frequency of communion, the universality or otherwise of saving grace, and exactly what is going in what is variously called The Lord’s Supper, Holy Communion, Communion, or Mass. That is, the eighteenth- and nineteenth-century arguments that created our sects and denominations are now largely passé. In their place we have disagreements, which cut across traditional divisions, about the ordination of women and gays, pews versus chairs, old liturgies versus freestyle worship, speaking in tongues or not, and venerable hymn tunes such as Repton and Kingsfold versus Christian pop. There are many sorts of data that will lead the researcher to this conclusion, and where possible it is always nice to have reliable quantitative data, but, provided we remain alert to limits of generalization, such everyday conversation is immensely informative. Clearly a constitutive characteristic of both these types of research is that they are unobtrusive; we wish to avoid the compliance effects (or, more rarely, discompliance effects) that may come into play when people know they are being observed and that their words and actions may be the subject of future report, and we wish to avoid disrupting events. People are self-conscious in any interaction but to add ‘being researched’ to the list of other considerations in play is not helpful. The value of this is obvious, but I will give one small example. Before the start of a Methodist service in a small chapel in Northumberland, I overheard the church steward explaining to the visiting preacher that ‘we’ll be a bit thinner on the ground today because’ and then she listed four people who were laid low by various infirmities of old age. She ended this explanation with a (p.133) heartfelt sigh and the remark that ‘Aye, we’re falling by the wayside!’. Her disappointed tone seemed to apologize simultaneously for the indignities of old age, for the poor local turnout, and for the decline of Methodism. She might have said exactly the same in an interview, but knowing she was a spokesperson might well have produced a misleadingly upbeat assessment. Opportunities for casual conversations about religion with members of the public arise surprisingly often with only the merest prompting. Bed and breakfast hosts invariably ask what I am doing in their area. Replying that I am visiting churches and chapels, or that I am writing a history of the area’s religious traditions, often produces fulsome responses. One Devonshire guest house owner, when I said I was interested in the history of local Bible Christian chapels, told me in great detail about his deviant status as a Catholic among Methodists and about the ways he teased his neighbours who, ‘if it weren’t for me, wouldn’t see a Catholic from one year’s end to the next’. A Northumberland Page 4 of 21

 

Ethics in Social Research host, hesitantly at first but then with increasing vigour as she got into her subject, explained why she ‘had no time at all’ for organized religion. She had worked in international aid agencies and had come to dislike Islam (for its attitude to women) and the Catholic Church (for its attitude to contraception). Over the years the guest house breakfast table has been a rich source of information about what ordinary people (as distinct from church officials and representatives) think. And even the non-committal ‘humph’ followed by a quick change of subject is sometimes suggestive. The third type of work is so much a part of the everyday life of the social researcher that it seems a little strange even to call it ‘research’ but we do it, and the ethics committee did not like it, so it is worth spelling out. Someone with an ongoing interest in a particular field of human life will always be observing and noting relevant signs, symbols, words, and deeds. The acute social researcher hoovers up impressions of the world and learns from them. For example, the fact that, in order to appeal to the unchurched, independent evangelical congregations often play down overtly ‘churchy’ elements in their public presentation first occurred to me when, walking down a Glasgow street, I passed a building with XCEL on its facade. What I first assumed to be a gym or a college turned out to be a church. Red herring alert: surprisingly often discussions of the ethics of different research styles get sidetracked into arguments about relative quality of data such strategies produce.7 As I am not claiming that (p.134) unobtrusive observation is always superior to other forms of research, there is no need to consider the data beyond accepting that there is value in it. Having explained why we might engage in the unobtrusive observation of ongoing social life, I will now consider the possibility that such research harms those so studied. The comments will be structured in three parts: data collection, data analysis, and reporting. The second of these can be dealt with quickly. Data analysis generally raises few ethical problems because we do it in private. It is the first (because it involves contact with the public) and the third (because it offers the greatest opportunities to offend) that we should consider.

Harm in Data Collection The potentially most harmful part of medical research is the data collection phase because this is where the unusual intervention occurs. It is the drug you are given (or not given) that may kill you. We can all agree that people should not be the subjects of intrusive medical research without having the benefits or hazards fully explained to them and their ‘informed consent’ (hereafter IC) recorded. It is worth adding here (because it sets the proper context) that actual harm (as distinct from hurt feelings) is so rare in the conduct of social research that I cannot think of any UK examples, whereas I have no trouble finding examples of death and serious injury caused by medical research.8 At a stretch Page 5 of 21

 

Ethics in Social Research we might imagine participant observation being harmful if the researcher is less good at the job he or she has taken for research purposes than is the person who would otherwise have filled the role. But, countering that, participant researchers may well be excellent workers because they are not yet careworn, cynical, or casual. There is no reason to think that, when Erving Goffman was working undercover in a Shetland hotel on what became The Presentation of Self in Everyday Life, he was a bad waiter, or that Donald Roy, while working on a series of now-classic papers on output restriction in manual work, was a bad machinist.9 Indeed, in the two cases where graduate students I was supervising were rumbled, their colleagues became suspicious precisely because they tried too hard and were too enthusiastic. The lack of obvious harm explains why I begin by supposing that an ‘ultra’ IC view is inappropriate for social research. It is important (p.135) to have a reasonable estimate of how much protection is required because that will determine whether the conceptual and practical difficulties of seeking and gaining IC justify us going ahead regardless or require us to abandon unobtrusive research. The ethics committee’s first objection to my work was that it involved deceit: I was not always entirely frank about my status or my purpose. The committee seemed to share the view of Kai Erikson that people should never be studied without their express permission.10 Disguising one’s purpose (even as little as when tourist-who-might-write-book presents himself just as tourist) prevents people exercising that right. Though the law on privacy may occasionally be relevant, there is no ‘right’ in English or Scots law not to be observed in public, and nothing in the literature suggests it exists in other jurisdictions. Much, of course, hinges on what is involved, but where the study is purely observational and the behaviour in question is public, it is difficult to see how a consistent case can be made for Erikson’s principle. With the exception of the Exclusive Brethren, who believe that communion should be taken only by those who meet its narrow criteria for having been saved, Christian bodies are usually only too anxious to attract outsiders.11 Most churches have signs outside that say ‘Visitors Welcome’ or something similar. Partly because I respect the public/private division and partly because I do not want to stand out more than necessary, I sometimes feel the need to check that visitors are indeed welcome. Some services as so sparsely attended that I have feared I was joining the choir practice or a committee meeting. Once I accidentally entered a small rural Methodist chapel by the wrong door and found myself in a kitchen where women were laying out an array of cakes and sandwiches. Thinking I might have stumbled on a private event, I asked if they always put on such a fine spread for a Sunday afternoon service: ‘Nay. Most Sundays you’d be lucky to get a cup of tea out of us but this is a camp meeting.’ I asked if a passing tourist could join in, and one woman assured me that ‘you’re Page 6 of 21

 

Ethics in Social Research more than welcome provided you say nice things about my Victoria sponge’. Is my presence harmful? In a large church service, I am invisible. In a small rural chapel, I am an asset. I fill a seat, sing enthusiastically, make a generous donation, and chat politely over the coffee and cakes afterwards. Most researchers have personal limits to deception. For example, I never take communion. Although most churches are now liberal in their views of entitlement and introduce that part of the service with an inclusive invitation to ‘all those who feel able to join us in celebrating (p.136) the Lord’s Supper’ or ‘all members of the body of Christ’, I do not go forward (or in a Presbyterian church I pass the trays along the pew without taking a glass of wine or a wafer of bread). Ironically this attempt to maintain a line between passing as a sympathetic observer and pretending to be a true believer is probably the most disruptive thing I do. In small Anglican congregations where everyone goes forward to the altar to take communion, people notice that I stay in my seat and clearly wonder about it. To my embarrassment I have even had a steward suppose I am handicapped and offer to bring the wine and bread to me. And in charismatic congregations where, at various points in the service, congregants stand and raise their hands in the air, I remain seated, which, as in the previous example, is probably more disruptive than pretending to be filled with the Holy Spirit. But I believe it is important to maintain a distinction between padding out a congregation and pretending to be a believer, because, unlike the Branch Davidians, British Christians are not so obviously dangerous that fully covert research is justified. In choosing church services to protect from disguised participation, the ethics committee picked quite the wrong sort of event for its sympathies. Not only are church services public, but arguably most religions are ‘public’ in that they explicitly challenge the beliefs and values of others. Religion is often argumentative, provocative, and confrontational. Since 1850 the Catholic Church, for example, has openly campaigned against democracy, freedom of speech, gay rights, divorce, and much else we now value. Its officials use the Church’s pulpits to promote policy positions that have consequences for people who are not members of the Church: the Scottish clergy, for example, have frequently used pre-election sermons to tell the faithful how to vote. Conservative Protestant churches repeatedly demand that those outside the circle of the saved change their behaviour or suffer dire consequences. And they receive public benefits (such as tax relief and tax-payer funding for their schools) that subsidize their work. It seems reasonable to argue that, by accepting the benefits of charitable status, by advertising their services, and by challenging the values of others, churches are more like political parties than private clubs. I am tempted to go further and argue that, far from having a right to privacy, churches ought to be grateful that more non-members do not attend in order to barrack them. When Protestants sing ‘The Church’s one foundation | Is Jesus Christ her Lord | She is His new creation | By water and the Word’, they Page 7 of 21

 

Ethics in Social Research are criticizing the (p.137) Catholic Church’s claims to a divinely ordained and authoritative structure.12 Set against the disruption of services, which could be justified by their ideological pronouncements—I have lost count of the number of times I have been told my life is worthless by evangelical Protestants—the mere presence of a social researcher seems unobjectionable. Even if one is minded to sympathize with Erikson’s view that people have a right not to be studied, if there is to be any unobtrusive social research, such a right has to be set against the equally important right not to have one’s activities disrupted. Unless we stop social research, giving people the right not to be observed without their express permission will, in many circumstances, cause them much more discomfort than being observed. The ethics committee objected to my attending church services because I did not give congregants the right to ‘withdraw’ from my research. It is hard to imagine how one could present them with that opportunity without destroying the atmosphere of the service. Presumably an announcement of the presence of an observer is not enough. As many congregants welcome strangers, giving them the right to withdraw would require not just an announcement but also an explanation of what is afoot. So what does one say to the congregation? Qualitative researchers cannot usually state the end point of their research: it may be a brief mention in a journal article, a few pages in a book, or it may just add to the researchers’ general understanding without any particular external consequence. This problem is not solved by saying that we should declare our occupation. Most people, especially the elderly who make up the bulk of UK churchgoers, have little idea of sociology. Many years ago I attended a tiny Baptist congregation with the pastor, who had read some of my work. At the end of the service he introduced me to the seven people in the chapel as a visiting sociologist and invited them to ask me questions. A few were interesting, but most were either pointless or awkward because, apart from the pastor, no one had any idea of what a sociologist did. Two women took me to be a social worker. One harangued me for letting wrongdoers out of prison; the other criticized me for taking children into care. Since then, if invited to explain my presence (and it is common to be asked by the vicar shaking hands at the end of a service something like ‘You’re a new face. Have you just moved in to the area?’), I have usually gone with some vague formula such as ‘I am thinking of writing a history of the chapels in this area and I thought I would visit a few to get the feel of the place’. At other times I have just said (p.138) ‘I’m in the area on holiday’. Adding ‘As I’m from Scotland, Anglican services are a bit strange to me’ then justifies asking questions about the service. Obfuscating, yes. Harmful? Only if, like Erikson, one believes that being observed while in public is itself offensive. It is difficult to define and measure harm in social research, but there is one useful touchstone: departure from the ordinary. Not all unusual interaction is dangerous, but the obverse works: interaction that is commonplace is hardly Page 8 of 21

 

Ethics in Social Research likely to be damaging. One of the clues that drug trials are potentially harmful is that, leaving aside sexual predators armed with Rohypnol, people do not normally slip each other powerful drugs. Nor do they replace their neighbours’ therapeutic drugs with inert powders. In contrast, people observe each other in public places all the time. The lady sitting next to me in a tiny chapel in the Strath of Kildonan may, like me, notice that the organist inserts a few extra beats at the end of each line to allow the elderly congregants time to draw breath before attacking the next line. She may note the casual dress of the couple in the third row and may remark on it to her friend later. She may spot that the harassed cleric has confused the name of this congregation with that of another where he will be preaching later in the day. The social researcher who engages in disguised participation is ‘studying’ the event and the other members of the congregation in much the same way as all the other participants are studying it and each other. The unobtrusive observer differs from the lay participants only in reasons for being there. If the social researcher’s observations are no more intrusive than anyone else’s, we can be fairly confident nothing bad is being done in the act of observing. The ultra IC position is even harder to apply to research that takes the form of engaging members of the public in chat about their religious commitments when the opportunity naturally arises. And there is as little justification for it. Edward Shils takes the view that all the social space around an individual ‘belongs’ to that person and ‘no one should enter this without permission’.13 Actually we enter each other’s space all the time. As the people I meet know that I am not their husband, brother, or close friend, they are already as much on guard as they would be with any other stranger. That they are nonetheless willing to talk to me seems sufficient permission to remember that talk for possible future use. There seems nothing to gain by making myself further alien by requesting that people give their IC to my learning from something they say. My intervention in their lives is so slight that the American tourist in Teesdale to whom I gave a lift and (p.139) who told me about the Methodist ancestors whose graves she was seeking, or the guest house landlady who explained her dislike for organized religion, will almost certainly have forgotten those encounters. Even if they remember them, it is very unlikely they are harmed by such memories. And this is the crucial point: it is difficult to see why people should be protected against encounters with a polite researcher any more than they are against the rest of the social interaction in which they were engaged in a typical day. If the loquacious London cabbie does not want his opinions to be remembered by a passenger, he can keep them to himself. Once, while driving through some rundown town in the north-east of England, I spotted a disused chapel and stopped to photograph it. A strongly built young man in casual clothes stopped his expensive BMW, approached me, shook me very firmly by the hand, and offered to buy the chapel before my firm put it on the market! I disabused him of the notion that I was an estate agent preparing a Page 9 of 21

 

Ethics in Social Research sale portfolio and explained that I was fascinated by old chapels and liked to photograph them; which is minimally true but fails the ethics committee’s requirement for full disclosure. It turned out that he was a self-employed builder (hence the physique and the BMW) who specialized in converting old chapels into residential accommodation. He was chatty and informative about nineteenth-century buildings: ‘Not all of them are as solid as they look. Especially in the pit villages. A lot of them were built by bloody amateurs. One layer of brick work only. Corners not properly tied in. Miracle they’ve stood this long.’ With no sense of irony, he also said that converting Anglican churches was often difficult because ‘graveyards are a bloody nightmare’. Should I have asked him to sign an IC form? After I had informed him about what? That I might someday repeat his observations (without attribution) in an essay on chapel conversion? As Roy Wallis put it to my undergraduate research ethics class forty years ago: The sociologist cannot easily stop being a sociologist. He or she cannot readily switch it on and off. I may enter a conversation or a setting quite ordinarily in the course of a day and only subsequently or even long after the event realize that I have generated research data.14 When one does not know in advance what encounter or observation may later be recalled as ‘data’, it is impossible to ask for IC. One sign that the ultra IC position is untenable is that it places greater constraints on the academic researcher than on other participants in any (p. 140) public event. Any member of a congregation may report their impressions in any manner of ways, from comments in parish magazines to letters to newspapers. There is an excellent website called Ship of Fools, which has a mystery worshipper section where, under a number of standard headings, people anonymously report on church services they have attended. The ultra IC position would allow me to attend a church service as a private individual, write a mystery worshipper report for Ship of Fools, publish it online, and then quote that in my publications. Yet it prevents me from attending the same service and later using my observations firsthand in my writing. The ultra IC position would also prevent most of the reporting of such eminent journalists as J. B. Priestley, George Orwell, and James Cameron.15 It is worth adding that the ultra IC position is based on a false contrast between authentic social interaction and interaction in which one party has an ulterior motive. As Erving Goffman demonstrated in the 1960s, there is very little naive interaction. People are almost always in the business of putting up a front and managing the impressions they make on others, especially during encounters in public.16 The ethical issue in qualitative research is not whether the researcher is being deceitful. It is whether, if we describe attending a church as a confirmed agnostic with an academic interest in religious rituals rather than as a potential Page 10 of 21

 

Ethics in Social Research convert as ‘deceit’, the deceit is likely to be harmful, and with the best will in the world I simply cannot see that case. At the risk of being tedious, I will add a further clarification. In the debates about the ethics of covert research, often the only example of possible harm that is adduced is the shock of discovering that the nice man who joined last year was not really a convert but a pretender. I can accept that discovering your fiancée is a bigamist conman, or that your male lover is a woman, will be a shock.17 Given that many genuine converts backslide, discovering that one of the flock is pretending to be a sheep in order to write dull sociology journal articles seems hardly disturbing. What is involved in the sort of disguised participation I am describing and justifying here is so little a departure from normal interaction that to call such participation harmful is a gross distortion of language. The final problem with the ultra IC position is that it supposes research to consist of distinct time-limited projects with clear single purposes—a view of research quite at odds with the qualitative social researcher’s continuous learning from the social world. The alert researcher will always be sucking in impressions from the world that (p.141) interest him or her. Alongside our specific planned programmes of work, we are constantly taking impressions from our social environments and learning from what turns up. The general point is simple. As a form of data collection (if that is not too pompous a way to describe being alert to what is going on around us all the time), the unobtrusive observation of naturally occurring social life is quite unlike much medical research. And this is recognized by the American Sociological Association, the ethical statement of which deliberately excludes such research from those types requiring IC.18 As observation rarely has a fixed duration, a limited set of concerns, or a narrowly specifiable purpose, it is difficult to see how one could acquire IC without disrupting and distorting what interests us. But, as important, it is difficult to see why we should try. Social scientists who simply observe ongoing social life are by definition having no effect. Social scientists who behave like ordinary people and later make use of what they have learnt from everyday interaction are changing the lives of their subjects no more than is any other stranger with whom such people casually interact. My bed-and-breakfast host engages all her customers in polite conversation. Unless one adopts a precious and unrealistic view of what degree of disclosure of purposes is expected in everyday interaction, that I have a disguised primary purpose in recalling her conversation is not ‘harming’ her in any normal sense of the word. A final thought on the costs to our subjects of data collection: the ethical issues are rarely discussed comparatively. Different research styles impose different burdens on those we study. A formal interview may take two hours out of a busy

Page 11 of 21

 

Ethics in Social Research person’s life and a focus group may take two hours from each of ten people, but unobtrusive research costs the subjects nothing.

Harm in Reporting Although I am treating data collection and reporting as two distinct phases of research, some forms of reporting readily occur while collecting data and others can be arranged as ‘second phase’ ways of collecting data. An example of the former is the presentation of sociological commentary in conversation. Most often I play ‘the daft laddie’—look blank and invite education—but I sometimes find myself being treated as an expert (on the effects of chapel amalgamation, for example) (p.142) and debating with those to whom I talk. The latter is in effect a form of feedback that usefully invites correction and clarification. A more formal way in which reporting can feed back into data collection is through presenting respondents with rough drafts of intended publications for correction and comment. Ethics aside, I have often found this a useful way of eliciting clarification and further information. I gave a rough draft of The Red Hand to the senior Ulster Volunteer Force man who had been my patron. He identified a few small mistakes but then startled me with his reply to my asking ‘What did you think of it overall?’: he paused and then said he hated it. He let me stew for a moment and than elaborated: because he was consumed with the dayto-day problems of running a terrorist organization, he rarely stood back and considered the overall impact of the UVF’s actions. He found my draft very depressing, because it clearly showed that his gunmen had done more damage to the Protestant people they were supposed to protect than to their Irish republican enemies. If we subsume all forms of social research commentary, including those that occur while collecting data, under the heading of reporting, we can ask what harm such commentary can do to our research subjects. Ironically, the ultra IC model’s medical origins cause it to have very little to say about this stage of research. The form I had to complete for the ethics committee had ten questions about the conduct of data collection and five about consent but none about reporting. While there are many legitimate concerns about the ethics of publication (or more often the non-publication) of the results of drug trials, such reporting is not expected to have any consequences for the individual trial participants.19 Yet, for much social research, it is the reporting that holds the greatest potential for harm, slight though that is.20 Discreditable Disclosure

Social researchers may offend their subjects by reporting discreditable things. We should not exaggerate the problem. What the hoodlum says to a fellow gangmember about his actions may well be sufficiently different from what he says to the police or the court that reporting it will cause embarrassment or worse, but most religious people do not regard their beliefs as discreditable. To the extent Page 12 of 21

 

Ethics in Social Research that they refrain from discussing them with strangers, this is only because there is in the UK a strong informal prohibition on face-to-face proselytizing. My Northumberland chapel steward might have been (p.143) a tad embarrassed had her precise words on the frailty of her congregation been broadcast, but that churchgoers are typically elderly, that Methodism is declining, and that the country is becoming more secular are hardly secrets. Offence can often be avoided by anonymizing our research subjects. So the Revd Obadiah Slope becomes ‘a senior Anglican priest’. As social science is concerned more with the general and the typical than with the particular and the idiosyncratic, disguising identities of people and places does not normally devalue our work. We often do this without any effort by merging our data: thus my observations of Methodist chapels in Devon, Northumberland, and Durham become observations about ‘rural dissenting chapels in England’. If the identity of a people or place can be completely disguised, those we write about will not even know it. But, even if those I have interviewed over the years recognize themselves in something I write, that other readers will not recognize them may allay some of their anxieties (if they have any) about how they will be perceived by the world at large. Conflict between researcher and researched (and this melds into my final point) may arise, not because we are disclosing anything discreditable, but because the researched simply do not like our conclusions.21 For reasons I will give shortly, I do not see this as an ethical issue. While anonymity can usually be guaranteed for individuals, it is harder to disguise a religious community or an organization. Everyone with an interest in the subject knew that John Lofland’s ‘Divine Precepts’ was actually the Unification Church (aka Moonies), and his attempts at preserving the anonymity of his research subjects were quickly subverted by all the later commentators on the Moonies who cited his original publications.22 One reason why social researchers rarely engage in discreditable disclosure is that we are usually second on the scene. When I studied loyalist terror organizations in Northern Ireland in the 1980s, I initially took great trouble to hide my notes from potential security force interest. I later came to realize that very many of the senior officials of those organizations were police informers! And most of the groups and activities we study have already been worked over by journalists. Were I now to write a sociological explanation of the use of religion for purposes of sexual exploitation, almost all my data (on Catholic priestly abuse, for example) would be drawn from the public domain. In order to illustrate his case that simply reporting religious beliefs and activities may itself be harmful, Martin Bulmer has to posit an imagined (p.144) religion that requires secrecy: ‘in some cases the most cherished values and Page 13 of 21

 

Ethics in Social Research beliefs of the group may be threatened by publication, a fact recognized by anthropologists in the case of some ritual practices’.23 Far from feeling that disclosure threatens the sanctity of their rituals, British Christian churches publish their orders of service. And, given their status as tax-avoiding charities, one might add that, if the British churches are engaged in rituals they would prefer to keep secret, the tax-paying public has a right to know about it. However, there are more subtle and intellectually interesting problems in providing sociological commentaries on people’s behaviour that I would now like to consider. Disagreement with Subjects

It is always possible that those we research will be annoyed by our reporting (either at an early feedback stage or in publication) simply because they think we are wrong about something. This is always likely with sociology, because we tend to be as concerned with the inadvertent and unintended consequences of people’s actions as with their motives and intentions. If a time machine allowed Max Weber’s Calvinists to read his essay on the connections between the Protestant Ethic and the Spirit of Capitalism, they might well be irked by the suggestion that their salvation anxiety had contributed to the growth of rational capitalism, but there seems no obvious resolution to this that would still permit research. Why not can be seen if we imagine the obvious solution: we simply act as mouthpieces for those we research. Clearly we have to write it as we see it. And if people are annoyed, we have two responses. First, it is a commonplace of social life that people disagree with reports of their behaviour. Provided such reports are not libellous or gratuitously insulting, research subjects should not expect to be shielded from them. Second, research subjects are very rarely a homogenous group with one single response to anything. As I noted in Chapter 5, even if we thought our job was to give voice to our subjects, we would have to ask ‘which ones?’. The Wrong Tone

Believers may be offended by sociological writings because they do not have the right tone. I try to enliven swathes of dull sociological prose by occasionally being flippant or jokey. Some believers find this (p.145) offensive, because they expect their beliefs to be treated consistently with reverence. The issue here is not blasphemy; it is merely that failing to write in the manner of one kneeling at prayer may irk some people. The Offence of Technical Terms

A related issue arises from the fact that social science jargon is not sufficiently jargonish to prevent outsiders having an attitude towards it. Even if it possessed consciousness, it is hard to imagine trinitroglycerine objecting to what chemists call it, but social science uses the same language as lay people, and, even if we use words in a careful, consistent, and morally neutral manner, people can still Page 14 of 21

 

Ethics in Social Research be offended by pejorative associations. For example, I have repeatedly had negative reactions to the sociological use of the term ‘sect’. Ian Paisley, founder of the Free Presbyterian Church of Ulster (FPCU) and later First Minister of Northern Ireland, objected to my describing the FPCU as a sect. I carefully took him through what sociologists mean by the terms church, sect, denomination, and cult, and he agreed that, if religious organizations had to classified in such a manner, then ‘sect’ would indeed be his preference. But he still did not like the term, because to him, and to most people, it meant ‘bad religion’. Social Science as the Denial of Authenticity

It is also worth recognizing that many people, if they are confronted with it, will be uneasy with the social science explanation of their behaviour. Sometimes we like to think of ourselves as products of our environment or as being swept along by events, but we also like to think of ourselves as free agents. Our fortunes are an achievement; our misfortunes are the fault of such circumstances as poor child rearing, class oppression, and the like. Social scientists are not immune to such dichotomous attribution, and I will say more about that in a moment, but, even if we distinguish between free agents and people whose behaviour can be explained, our distinction is not likely to map onto that used by those whose behaviour we are trying to explain. Hence our subjects may object to their imagined-to-be-autonomous behaviour being explained. Philosophers may argue that causal explanations of people’s behaviour need not clash with the explanations they give in terms of good reasons, but it is (p.146) understandable that people may see sociological explanations of their actions as a denial of their integrity. A related problem concerns uniqueness and typicality. Most people can sometimes see themselves as a type (as in ‘I’m your typical Guardian reader’), but they also see themselves as more than just a carrier of shared social characteristics. To the extent that social science typifies and explains, it may well present people with an image of themselves that, like a badly taken photograph, just does not look right. This has particular application for sociologists of religion. Most of us have little difficulty in not obviously taking sides. Few of us would disagree with Peter Berger’s argument that we should practise methodological atheism (although these days we would better describe it as methodological agnosticism) by ‘bracketing’ the truth claims that religions make for themselves, and we routinely do this by working at one remove.24 Ian Paisley believes that God does not hear the prayers of papists. We report that ‘Ian Paisley believes that God does not hear the prayers of papists’. We neither endorse nor deny his claim. For the purposes of social science, all we need do is ensure that we are indeed correct in our reports of Paisley’s beliefs. We then try to explain why he holds this belief by looking for similarities with others who also believe it.

Page 15 of 21

 

Ethics in Social Research That we work at one remove from the actor’s view of the world allows us to argue that, for example, there is no clash between the sociological proposition that having full-time officials leads to oligarchy and oligarchy causes sects to lose their radical edge and become denominations and the Exclusive Brethren view that full-time clergy are a devil-inspired departure from God’s will. We are neither challenging nor supporting the Brethren, because these explanations operate at different levels. We do not argue with our respondents’ views; we translate them into our terms. Social science explanations and the perspectives of Christians can be readily reconciled if the Christians are minded to believe that God works through human regularities. The liberal Christian can accept evolution by saying that evolution is the way in which God created the world and that the Bible only does not say that because God has to reveal himself to his various peoples in ways they can understand and the people of Noah’s time understood artisanal creation but did not know of evolution. Similarly, the liberal Christian can argue that God invented sociologists and gave people the characteristics that reveal (p.147) themselves through regularities in behaviour that can be identified and explained by sociology. But many conservative Christians will wish to attribute even apparently mundane things to God’s direct actions. Ian Paisley’s Free Presbyterians are fond of interpreting all sorts of events and occurrences as a ‘message from the Lord’. A social science explanation of the same phenomena may not clash from our perspective, but it may from theirs. However, to stop at the divide between those believers who are minded to accept secular regularities in human behaviour and those who wish to attribute everything to God’s direct and apparently idiosyncratic interventions is to miss something important. More often than either easy acceptance or outright rejection of social science explanation, we find a partisan application of social science: respondents are happy to have the actions or beliefs of their competitors ‘explained away’ but bridle at being treated in the same way themselves. In this they are behaving in much the same way as pre-1960s sociologists. Karl Mannheim’s sociology of knowledge was in effect a sociology of error.25 Why people believed the truth required no sociological explanation; only ideologies (because they were wrong) needed explaining. The intellectual strategy we adopt in order to avoid taking sides—essentially The Social Construction of Reality position that all beliefs equally stand in need of sociological explanation—is, to many of our subjects, a highly partisan position.26 It means we are rejecting their accounts.

Taking Offence and Being Harmed Having identified various ways social science reports create opportunities for people to take offence at the way they are presented, we can now consider whether they cause such harm that we should give up writing social science.

Page 16 of 21

 

Ethics in Social Research Naturally the sociologist should be no keener than any decent lay person to offend, but it is hard to construe presenting our subjects in ways other than they would prefer as requiring informed consent. The current legal position on religious hatred is clear: simply reporting religious beliefs and even making critical remarks about them is not an offence. The Racial and Religious Hatred Act 2006 specifically said: ‘Nothing in this Part should be read or given effect in a way (p.148) which prohibits or restricts discussion, criticism, or expressions of antipathy, dislike, ridicule, insult or abuse of particular religions or the beliefs or practices of their adherents.’27 That we study people does not commit us to being more reverential about their beliefs and rituals than they are themselves. Very many believers can be flippant and humorous about their own people. Ian Paisley once discouraged interruptions from an over-enthusiastic congregant who repeatedly and loudly added ‘Amen’ to his sermon by ad-libbing ‘I see someone’s had his porridge today!’. I see no reason why we should be perpetually po-faced. I might add that almost all sociological writing about religion is vastly more respectful than is most commentary on sociology by religious people. It is equally the case that using social science jargon that the lay person does not like falls a long way short of harm. It may be unfortunate and we can try to minimize it. For example, I generally refer to all Christian organizations as ‘churches’ and use sociological terms such as ‘denomination’ and ‘sect’ only when I am explicitly drawing attention to the principles that inform that classification. But any kind of intellectual activity that is not purely descriptive involves imposing on our subjects terms that they either do not use themselves or do not use consistently. And, even were it possible to speak and write about one particular religious strand in only the terms its adherents use, we would inadvertently offend every other strand. It might just be possible for an anthropologist reporting on a people with a unitary culture to represent all of them in their own terms, but to represent any one Christian organization in its preferred terms will offer offence-taking opportunities to many other Christian groups. To the charge of offending people by offering causal explanations of behaviour for which the actors wish to claim either idiosyncratic free choice or divine prompting, we have no choice but to plead guilty. Born-again Christians have a right to claim that God has chosen them or they have chosen God, but sociologists have an obligation to point out that only 5 per cent of Scots who currently claim a religion were not raised in that faith. Some of those born-again Christians may be offended by the report of the statistical regularities of the family reproduction of faith; many are not and praise the pious family. But they cannot assert that this does them such harm that sociologists should either not study such regularities or should keep silent about their conclusions.

Page 17 of 21

 

Ethics in Social Research (p.149) We can present a good case for saying that social science explanation need not deny the integrity of the actor or divine intervention, but there is no getting away from the fact that some of our subjects will not accept some reconciliation in terms of different spheres. But I am tempted to say ‘so what?’. That almost everyone is willing to offer causal explanations of why their competitors have got it wrong is a warrant for us doing the same thing, but, it is to be hoped, with greater consistency. To recap thus far, no matter how well it protects the identity of particular subjects, reporting research may offend the thin-skinned because it fails to accord precisely with how they see themselves. Is this unethical? I do not see that it is any more harmful than the disagreements about motives and reasons that are a constant feature of everyday life. Something that lay people do to each other all the time cannot justify constraining social research. Finally, because it shows the poverty of an ethical objection to our research subjects simply not liking the way they are portrayed, we should note that the possibility of such a clash of perspectives is not confined to the subjects of specific studies (qualitative or quantitative, overt or disguised) of those who may feel offended by social research. Knowing that your organization or activity has been the subject of sociological enquiry may cause you to seek out relevant reports and thus increase the likelihood of finding something to dislike, but people who have not directly been the object of the social science gaze may still bridle at the way people like them are portrayed by social science. A 1960s worker in the car industry who was not part of the Affluent Worker studies could still be irritated by John Goldthorpe et al.’s sociological explanations of their attitudes and lifestyles.28 The same point can be made about discreditable disclosure. Social scientists often identify patterns in aggregate data that might offend any particular member of a collectivity if that fact was known about that individual. For example, American economist Benjamin Edelman has demonstrated that credit card spending on internet pornography is highest in the most socio-morally conservative US states.29 There are possible explanations other than hypocrisy, but nonetheless an Alabama Baptist might well feel insulted by this report. Should Goldthorpe have sought the informed consent of all members of the affluent working class, or all car workers, or all residents of Luton? Should Edelman have refrained from publishing his research findings? Of course not.

(p.150) Conclusion We should always be mindful of the ethical implications of our research. Being a researcher does not free us to act with little thought to the interests and sensitivities of those we study. But there is no good reason why we should be unusually precious about our work. What seems obvious to social researchers but occult to some ethics committees is that, because the vast majority of social research is quite unlike a therapeutic drugs trial, the informed consent approach Page 18 of 21

 

Ethics in Social Research is unnecessary and impractical. It is impossible to apply without disrupting the social activities we reasonably wish to study and without subverting the purpose of our research, which is to observe the naturally occurring rather than the staged event. One option is to give up unobtrusive social research. The other is to ask if there is any good reason to think that such research causes sufficient harm to justify abandoning an important research method. Unless one supposes with Erikson and Shils that simply being observed is a loss, we have to look for harmful consequences. What we do by way of data collection changes the lives of our subjects no more than they are always being changed by encounters with strangers in public. Where we need to think about ethics is in the reporting of our observations. Because we are concerned with the typical rather than the particular, our research reports rarely identify specific people, places, or organizations. They even more rarely report discreditable observations that are not already in the public domain. The most common possible sources of offence—social science language and sociological explanation—are quite rightly not seen as major ethical problems. I can see no grounds for constructing a right to be protected from being explained, and, given that ordinary people are quite happy to offer causal explanations of the behaviour of others (and even sometimes to explain away their own actions), I see no reason why we should try to construct such a right. Notes:

(1.) For articulate versions of this case, see P. Atkinson, ‘Ethics and Ethnography’, 21st Century Society, 4 (2009), 17–30, and R. Dingwall, ‘The Ethical Case against Ethical Regulation in Humanities and Social Sciences’, 21st Century Society, 3 (2008), 1–12. (2.) P. Spicker, ‘Research without Consent’, Social Research Update, 51 (2007), 1–4, and ‘Ethical Covert Research’, Sociology, 45 (2011), 118–23. (3.) S. A. Wright (ed.), Armageddon in Waco: Critical Perspectives on the Branch Davidian Conflict (Chicago: University of Chicago Press, 1995). (4.) L. Humphreys, Tearoom Trade: Impersonal Sex in Public Places (New Brunswick: Transaction, 1970). For critical commentary on his ethics, see the additional material in the 2009 edition. (5.) J. Bone, The Hard Sell (Aldershot: Ashgate, 2006). That his research method was justified is clear from the fact that the book was awarded the Best First Book prize by the British Sociological Association in 2007. (6.) The UK Data Archive is a fantastic but underused resource. The original data sets for such major recurrent surveys as the British Social Attitudes, British Household Panel, and British Election studies are available to bona fide Page 19 of 21

 

Ethics in Social Research researchers from the archive at the University of Essex. It also makes available a great deal of data (such as interview transcripts) from qualitative studies. (7.) R. Barbour, ‘The Ethics of Covert Research’, Network, 15 (1979), 9. (8.) My inability is shared by at least one other very well-informed commentator: P. D. Reynolds, ‘On the Protection of Human Subjects in Social Science’, International Social Science Journal, 24 (1972), 693–722. My point should not be confused with the very different observation that agents such as governments, just as they can use natural science for unpleasant ends, can try to do harm with the results of social research. But, as I note with the issue of discreditable disclosure, social scientists rarely uniquely possess important knowledge: the rich and powerful prefer hiring their own spies, and journalists usually get to a story long before academics. (9.) E. Goffman, The Presentation of Self in Everyday Life (Harmondsworth: Penguin, 1972); D. Roy, ‘Banana Time: Job Satisfaction and Informal Interaction’, Human Organization, 18 (1959), 158–68. (10.) K. Erikson, ‘A Note on Disguised Observation’, Social Problems, 14 (1967), 366–73. (11.) This restriction applies only to taking communion. Although visitors are not particularly encouraged by the Exclusives, their services are open to the public and are advertised (with a very small sign) on the face of their chapels. (12.) The lyric is by Samuel Stone. The hymn has been popular in most Protestant churches since the end of the nineteenth century. (13.) E. Shils, ‘Privacy and Power’, in E. Shils (ed.), Center and Periphery: Essays in Macrosociology (Chicago: University of Chicago Press, 1975), 344. (14.) R. Wallis, Lecture on Research Ethics, University of Stirling, 1974. (15.) J. Cameron, The African Revolution (New York: Random House, 1961); J. B. Priestley, English Journey (London: Heinemann/Gollancz, 1934); G. Orwell, Down and Out in Paris and London (London: Gollancz, 1933). (16.) E. Goffman, The Presentation of Self in Everyday Life (New York: Anchor Books, 1959); Encounters: Two Studies in the Sociology of Interaction—Fun in Games & Role Distance (Indianapolis: Bobbs-Merrill, 1961); Behavior in Public Places: Notes on the Social Organization of Gatherings (Glencoe, IL: Free Press, 1963). (17.) In 2017 a woman was convicted and imprisoned for pretending to be a man in order to have sex with a heterosexual female friend. The victim’s response is discussed at length in S. Hattenstone, ‘“I was pretending to be a boy for a Page 20 of 21

 

Ethics in Social Research variety of reasons”: The Strange Case of Gayle Newland’, Guardian, 15 July 2017. (18.) ASA Code of Ethics, (accessed January 2015). (19.) The entirely wonderful Ben Goldacre has long campaigned to prevent drug companies cherry-picking their data by having the results of all drug trials made public. See B. Goldacre, Bad Pharma (London: Fourth Estates, 2013). (20.) For various perspectives on this, see the contributions to C. B. Betchell (ed.), When They Read What We Write: The Politics of Ethnography (Westpark, CT: Bergin and Garvey, 1993). (21.) For a discussion of an example, see J. Lawton, ‘Gaining and Maintaining Consent: Ethical Considerations Raised in a Study of Dying Patients’, Qualitative Health Research, 11 (2001), 693–705. (22.) J. Lofland, Doomsday Cult (New York: John Wiley and Sons, 1981); D. G. Bromley and A. D. Shupe, ‘Moonies’ in America: Cult, Church, and Crusade (Beverly Hills CA: Sage, 1979). (23.) M. Bulmer, ‘Comment on “The Ethics of Covert Methods”’, British Journal of Sociology, 31 (1980), 61. (24.) P. L. Berger, The Social Reality of Religion (London: Faber and Faber, 1969), 12. (25.) K. Mannheim, Essays in the Sociology of Knowledge (London: Routledge and Kegan Paul, 1952); M. Barrett, The Politics of Truth (Cambridge: Polity Press, 1992). (26.) P. L. Berger and T. Luckmann. The Social Construction of Reality (London: Penguin, 1966). (27.) N. Addison, Religious Discrimination and Hatred Law (London: RoutledgeCavendish, 2007), 114. (28.) J. H. Goldthorpe, D. Lockwood, F. Bechhofer, and J. Platt, The Affluent Worker in Class Structure (Cambridge: Cambridge University Press, 1969). (29.) B. Edelman, ‘Red Light States: Who Buys Online Adult Entertainment?’, Journal of Economic Perspectives, 23 (2009), 209–20.

Page 21 of 21

 

Conversion

Researching Religion: Why We Need Social Science Steve Bruce

Print publication date: 2018 Print ISBN-13: 9780198786580 Published to Oxford Scholarship Online: August 2018 DOI: 10.1093/oso/9780198786580.001.0001

Conversion Motives, Structures, and Discourse Steve Bruce

DOI:10.1093/oso/9780198786580.003.0007

Abstract and Keywords Although religious conversion is rare (most people are born and socialized into their faiths), it provides an important challenge for social science. This chapter considers the relative merits of explanations (such as brainwashing) that see conversion as something done to the convert and more activist paradigms that see conversion as an accomplishment of the religious seeker. It also considers the extent to which we can take what people say when accounting for their actions as the raw material for our explanations. It argues against the view that talk is only ever another narrative. Although an account of one’s past (given in court, for example) may well be shaped by a desire to influence hearers, it can still be used as material for inferring motives. In everyday life we untangle layers of motives and distinguish between different degrees of selfunderstanding and honesty. We can do the same in social science. Keywords:   conversion, narratives, testimonies, brainwashing, religious seekers, ethnomethodology

Introduction This chapter has a theoretical and a methodological purpose. The topic of religious conversion will be used to reflect on a perennial problem in the social sciences: how do we explain behaviour without portraying people as dupes? I will also consider the value of an approach to research that fundamentally changes the use we make of the explanations people give of their actions. Conventionally, accounts of action (be they monologues, responses to questions, or written texts) are treated as evidence of the motives, reasons, psychological Page 1 of 19

 

Conversion states, or social conditions that inform such acts. What is sometimes called ‘discourse analysis’ concentrates either on innate features of such accounts (in which case it becomes a form of literary criticism) or on the social setting in which such accounts are given. Either way, there is a challenge to the previous practice of using what people say about their behaviour as the raw material for explanation, and we need some reasoned response to what is otherwise a considerable obstacle to social science. I will also explain religious conversion. Conversion has been of sporadic interest to social science. The emotionalism of eighteenth- and nineteenth-century religious revivals intrigued some sociologists, and in the late 1950s Alan Eister studied conversion to Frank Buchman’s Moral Re-Armament movement, but what fuelled the boom in conversion studies was the late 1960s growth of new religious movements (NRMs).1 The numbers of young Americans and Europeans who became devoted followers of the Revd Sun Myung Moon or Bhagwan Shree Rajneesh were always vastly smaller (p.154) than the numbers who shifted unobserved or unremarked between Protestantism, Catholicism, and indifference, but the new religions were sufficiently alien to make up in exoticism what they lacked in numbers. Why were affluent young students and graduates (that is, people who should know better) turning their backs on the culture of their parents and giving themselves into the thrall of strange messiahs?

Brainwashing and Biological Manipulation The most popular answer given by lay people was that this phenomenon was not conversion at all; it was brainwashing.2 The case was made with varying degrees of sophistication. As William Sargant did in his popular Battle for the Mind, academic underpinning could be provided by citing the work of Nobel Prizewinning Russian physiologist Ivan Pavlov on the conditioning of dogs.3 Pavlov is best known for what is now called ‘operant conditioning’. He repeatedly gave his dogs a definite signal (such as the ticking of a metronome or a weak electric current) before feeding. The food made the dogs salivate. An extended period of associating the signal and the feeding created in his dogs such a strong connection that when he gave the signal without presenting food, they still salivated. This pioneering work on the conditioned reflex was disrupted when the Leningrad laboratory was flooded in 1924. Once the dogs were rescued and the research resumed, Pavlov found in some dogs ‘the recently implanted conditioned reflexes had also now all disappeared’.4 Sargant offers this, along with the experience of shell shock in the trenches of the First World War, as proof that trauma can addle the mind and then argues that the hellfire and damnation preaching of Wesleyan Methodist preachers worked in the same way: hearers were terrified into losing their critical faculties and became sufficiently credulous to accept beliefs that normally would have had little or no appeal. Their brains, having been ‘washed’ clean of rationality, accepted irrational beliefs.

Page 2 of 19

 

Conversion Another source for the notion of brainwashing was the work of Robert Lifton and Edgar Schein on Chinese indoctrination programmes.5 During the Korean war, US prisoners were subjected to an extensive campaign of torture, brutality, and ‘re-education’ to (p.155) persuade them to embrace communism and to denounce American imperialism. The brainwashing paradigm asserts that brain function can be deliberately distorted by physical violence or the threats of such violence, malnourishment, sleep deprivation, drugs, loud rhythmic music (but only pop music, never Beethoven or Wagner), and hypnosis. Once befuddled, victims credulously accept beliefs they would otherwise not entertain. The idea of brainwashing has never been popular with serious social scientists—Benjamin Zablocki is the only one that comes to mind—and the reasons for rejecting it seem overwhelming.6 First, it did not work in Korea. Those who cite Lifton and Schein on coercive persuasion seem to have read only the titles of their books and not the contents. Despite having almost complete control over their prisoners (and the final sanction of death), the Chinese failed miserably. Almost no American prisoners were converted; mostly what was produced was superficial compliance under duress. The very few who became sympathetic to Maoism were people who were already well disposed to totalizing systems of thought. Lifton devotes a lot of space to analysing the predispositions of ‘Miss Darrow’ and ‘Father Simon’ (the two who came closest to being converted), and the main thrust of his explanation is not manipulation or coercion but the dynamic interaction of an authoritarian ideology and a pre-existing desire to surrender to an allencompassing totalitarian world view.7 In brief, although the work of Lifton and Schein is frequently and loosely cited as proof of brainwashing, the explanation of such change as was found lay, not in what the Chinese did to their prisoners, but in the appeal of a certain sort of ideology to a certain sort of personality. Sargant’s case is equally unconvincing. His use of Pavlov’s frightened dogs and of shell-shock victims to explain evangelical conversion skips over the enormous difference between the first two and the third. Both the dogs and the victims of shell shock were terrified out of their wits by a very real threat to their lives. Even the most unethical evangelist does not actually threaten to kill his audience: he uses words to threaten them with a ghastly fate. Unless one is persuaded by those words, there is no danger; only those persuaded by the words feel threatened. Yet the fact of being persuaded is explained by the threats conveyed by the words. The logic is circular. A second weakness is that Sargant illustrates only the first or preparatory part of the brainwashing process.8 At best his examples show only that trauma can cause psychological problems; none shows (p.156) the successful implanting of incredible ideas. Sargant’s case is further weakened by his dated critique of rock-and-roll. Like many intellectuals of his time, Sargant regarded loud Page 3 of 19

 

Conversion rhythmic music, especially when enjoyed by crowds of excitable teenagers, as a source of psychological damage: and that was just the Beatles playing through a cabaret-band-sized sound system! Lord knows what he would have made of Heavy Metal, Electronica, House, Hip-hop, Trance, and the rest. And yet those teenagers went on to take their parents’ places as accountants, civil servants, clergy, and bankers. In brief, the brainwashing argument lacked firm foundations before it was applied to the NRMs of the 1960s, but its application to the Moonies, the Divine Light Mission, the Hare Krishnas, and the like suffered from a further weakness: there was no evidence of coercion. The people who accepted an invitation to tea with the Moonies were never prevented from leaving. Those who agreed to attend a residential weekend were not tortured or drugged. Christopher Edwards was a Moonie for seven months. His Crazy for God asserts that he was brainwashed, but the worst he can claim is that he worked very hard for the cause; he ate junk food and was often tired; and he was constantly in the company of committed members who appealed to his youthful idealism, reiterated proclamations of love, made chaste displays of affection, and discouraged him from thinking about his old life or contacting his old friends. Displays of affection apart, this seems hardly different from the life of any ambitious young professional. By far the most powerful evidence against the brainwashing paradigm is the recruitment record of the late 1960s NRMs. The Moonies talked to hundreds of thousands of young people. Only a tiny fraction of those accosted in public places by canvassers accepted the offers of hospitality and only a tiny fraction of those ever became converts. Eileen Barker presents robust and clear data on Moonie recruitment and retention.9 Of those who were interested enough to attend the two-day workshop, 15 per cent had quit before the two days were up; 70 per cent had gone before the seven-day course; 82 per cent had gone before the three-week course; and 93 per cent had broken off contact within a year. Only 7 per cent of those who willingly subjected themselves to two days of Moonie ‘brainwashing’ were still involved one year later. When I asked a former senior Moonie (who later become a successful New Christian Right activist) about these data, he ruefully said: ‘Yeah, 93 per cent drop-out rate. Not exactly the secret of fire, is it?’

(p.157) Social Forces Although most rejected the psychological and neurological models of brainwashing, sociologists produced their own somewhat manipulationist explanations of conversion. Based on Lofland’s doctoral thesis study of the early endeavours of the Moonies in the USA, John Lofland and Rodney Stark produced a highly influential model of conversion that combined background social forces and structured social interaction.10 The recruitment of members to what Lofland rather charmingly continued to anonymize as the Divine Precepts long after Page 4 of 19

 

Conversion everyone knew he meant the Unification Church was explained by a model of seven elements. Three were initial predispositions: recruits suffered from some sort of long-term social strain, were comfortable with religious language and with a religious attitude to the world, and saw themselves as religious seekers. The other four elements were features of the situation. The recruits had reached some sort of turning point in their lives: their old attitudes and perspectives were no longer working for them. They developed close ‘affective ties’ with group members. Their ties to family and friends were cut. And they engaged in intensive interaction with group members. Arguably it was more of a summary of a few common elements in the biographies of the early converts than an explanatory model, but it did contain both variants of the social determinist view that conversion could be explained by what happened to the recruit: in this case the social structure had created in some people needs that conversion met and the recruit was led into conversion by the manipulation of his or her relationships by cynical group members. The weakness with both versions of social causation is that they overspecify. This is a common fault in much social science. Even refined attempts to provide a checklist of social circumstances that cause people to commit crimes end up identifying a very large number of people, only some of whom commit crimes. Likewise with attempts to specify the social sources of recruitment to new religions. For example, Roy Wallis’s analysis of new religions according to whether they are world-rejecting, world-affirming, or world-accommodating is accompanied by detailed descriptions of the social circumstances that create someone who should find each type particularly appealing.11 These are insightful and thought-provoking. It is easy to see, for example, how deferring gratification in pursuit of (p.158) career success can create a pool of middleclass professionals who sense that along the way they have lost the capacity for guilt-free pleasure and authentic social relations. Hence the appeal of quasispiritual movements that endorse material well-being but offer to compensate for years of repression and sublimation. However, although the Wallis explanations are intuitively plausible, like structural explanations of crime, they encompass far too many people: what he offers as causes of attraction to certain NRMs are commonplace, but the numbers of people actually attracted have never been anything except trivial. Lofland himself recognizes this problem when he says: It is unfortunate that a convention has grown up in sociology of treating various kinds of demographic characteristics, structural, and/or personal frustrations and so forth as reasonably complete accounts of factors that push people into groups that are dedicated to protest against the prevailing social order. Not that these factors are unimportant, or that such models are inaccurate, but that they are woefully incomplete.12

Page 5 of 19

 

Conversion Lofland’s model partly resolves that problem by adding that the potential member must have the opportunity as well as the motive (that is, he must actually come across the offer of est training or whatever) and by making the developing relationships between potential recruit and group members a key element of the journey. The difficulty is that this leads us into circularity. We know that most NRM recruits were young; they had left their families of origin but not yet formed their own families. We could describe this in terms of bonds or ties: recruits were unusually free of the bonds that tie the rest of us to our families, jobs, and homes.13 But such bonds are not objective constraints. A married woman should be prevented from entering the dating market by her tie to her spouse. But she may meet someone she prefers and leave her husband. Existing social relationships are obstacles to innovation, but people innovate nonetheless and revalue their previous social relationships. That being the case, it is difficult to use the lack of existing constraints as an explanation for conversion. The same applies to such predispositions as social strain. Very many different circumstances can be interpreted by the pre-convert as sources of unhappiness. Equally, the same objective circumstances can be viewed by different people in different ways: some people find contentment in an orderly mundane life that others will find stultifying. It is often the case that we know a particular set of circumstances was seen by the actor as social strain only after he or (p.159) she has joined a new religion; what looked like part of the explanation may actually be retrospective revaluation. In brief, the Lofland–Stark model looks at first sight like an explanation of conversion, and it certainly fits what we know about most people who joined world-rejecting NRMs—young, single, childless, away from home—but what appear to be variables in a causal explanation are either too vague (and thus do not distinguish between those who do and those who do not join) or are open to retrospective reassessment in a manner that makes the whole schema somewhat tautological. We know that the predispositions were sufficiently powerful and the social relationships that developed with group members sufficiently strong only after the person has converted.

The Autonomous Seeker As is common in the social sciences, the thesis produces the anti-thesis. Led by a number of important papers by Roger Straus, one of Lofland’s doctoral students who doubted the value of Lofland’s approach because it did not fit the manner in which he had become attracted to Scientology, sociologists began to argue against both social structural and social interactional forms of determinism.14 Conversion is not something that is done to the hapless individual; it is something the knowing individual accomplishes. Instead of being a passive (and possibly unaware) carrier of social forces—what Harold Garfinkel called a

Page 6 of 19

 

Conversion ‘cultural dope’—the convert should be seen as an active seeker after enlightenment.15 One particularly significant argument came from another study of Moonie recruitment. David Bromley and Anson Shupe turned the prevailing orthodoxy on conversion and recruitment on its head.16 They rightly note that most people think of conversion as occurring in the following order. First, background social structural conditions produce needs or predispositions. Second, the individual is exposed to new beliefs that address such predisposing needs. And, third, the potential convert becomes a committed member of the group and acts appropriately. Bromley and Shupe believe the sequence should often be reversed. After initial contact with the group, novices start to behave as if they believed; they try on the new role like someone trying on a new suit of clothes. Only if they feel comfortable with the (p.160) role do they start to absorb the beliefs that make sense of, or validate, the role. Finally, the new recruit learns, through the new beliefs that are the solution, to see his previous life as a problem, as lacking in something. That is, by acting as if they were true believers, new members acquire the perspective that creates, through retrospective reinterpretation, the needs that social scientists traditionally cite as the explanation for the conversion. This is an attractive notion: the image of a person tentatively exploring ideas by acting out the behaviour that, for committed believers, comes with the new ideas fits well with ethnographic descriptions of NRM recruitment. But it does leave one rather big question unanswered. If we ignore needs or predispositions, how do we explain why there are social patterns to Moonie recruitment? Why are black people or farmers or poor people not attracted to experiment with the role? As those who experiment with the believer role differ systematically from those who do not (and from those who find the idea so ghastly they reject the very first advances of the Moonie recruiters), some part of the explanation must come before the role play. I do not see how we can dispense entirely with some notion of predispositions. This is not the same as saying that the causes of conversion are social structural or social psychological. In the case of the Moonies, the predispositions seem primarily ideological and cultural: recruits tended to be idealistic, raised in Christian households, and keenly interested in matters spiritual, but not strongly committed to one Christian church. We could work back a stage to some social foundations: we could show why those characteristics were associated with a particular class, race, and age cohort. This would not be the hard social structural causation that supposes we do need a social cause to new religion recruitment, but it would anchor the Bromley and Shupe model.

Page 7 of 19

 

Conversion Discourse Analysis Once the manipulationist and activist alternatives were clarified, anglophone sociology lost interest in conversion. One recent British sociology of religion text has no index entry for conversion,17 and a quick count through recent issues of the Journal for the Scientific Study of Religion confirms my impression of neglect. Of the 250 (p.161) articles in 25 issues after 2000, only 2 were concerned with conversion and 1 was by a psychologist.18 This neglect may result from satisfaction with the state of our understanding. More likely it results from weariness with the topic. After all, there are only so many times one can have the passive versus active, or structure versus agency, argument. But there are two further considerations. In a moment I will discuss the interest in large-scale religious change that is an alternative to focusing on individual conversion. First, I want to consider a radical reorientation of conversion studies that is part of a general drift from causal explanation in sociology. Lofland himself took tentative steps down this road when he suggested that, instead of seeking the causes of conversion, we should instead analyse common ‘motifs’ in stories told about conversion.19 Such a redirection is a local instance of a general approach to the accounts that has been around on the fringes of sociology since C. Wright Mills in 1940 proposed that, ‘when an agent vocalizes or imputes motives, he is not trying to describe his experienced social action. He is not merely stating “reasons”. He is influencing others—and himself. Often he is finding new “reasons” which will mediate action.’20 That is, explaining yourself, giving an account of your actions, is not just objective reporting of the past: it is also a social act in which you seek a certain outcome. Second, in many settings actors have available to them templates that shape the way they account for themselves. The late Victorians were fond of drunken fathers who repented on their deathbeds; in the twentieth century a more common theme was the pious mother who tried but failed to prevent the testifier going off the rails.21 That is, accounts may be stylized to an extent that it is hard to believe they are an accurate representation of the speaker’s past. Courtroom testimony is an obvious example of both methodological themes. The actor has an interest in making a favourable impression—to be found not guilty —and there are templates that govern courtroom speech.22 Lawyers can coach their clients. The implications for conversion accounts are obvious. Converts who describe their conversions are often trying to impress others. They also often present their testimonies within a well-established frame that they themselves have learnt from others. Anyone who has spent any time in evangelical Protestant circles will appreciate that, while testimonies may not be as formalized as the casual Catholic confession, they are predictable enough for an outsider to construct a plausible testimony.

Page 8 of 19

 

Conversion (p.162) The same two observations can be made about the way in which some ex-members of new religions explain themselves. Like people facing criminal charges, they want to be let off; in this case they want to be freed of responsibility for their previous actions. They want to say: ‘It was not me who for a while believed that a chubby Korean was the Messiah; they did it to me!’ And the brainwashing paradigm offers a pre-existing template for giving a conversion account that allows ex-members to deny their own agency. That accounts are not (or not only) reconstructions of the past is taken by some social scientists to mean that we cannot use what people say about themselves as data for social explanation. For ethnomethodologists in the 1970s, the problem was scientific validity. Accounts were poor data for the past but good data for the present, so the only way to be truly scientific was to study the performance of account-giving. As a development of a generalised sociology of motives perspective in short, ethnomethodology promises analytic pay-off in its commitment to address the relationship between accountability and activity in highly specific sociocultural situations, e.g., the experience of accounting for conversion.23 As Brian Taylor put it: ‘converts’ accounts may be analysed as a means of bringing to the surface underlying patterns of a domain-specific accounting practice which generates talk as ‘convert talk’ and produces the speaker as a typical convert’.24 Bruno Latour makes the same point about religious speech in general.25 An obvious appeal of the idea that conversion stories should be taken, not as flawed material for causal explanation, but as ideal material for studying how converts talk about conversion, is that it eases the researcher’s life. As with the fondness in cultural studies for analysing social issues such as gender relations by treating magazine articles as revelatory ‘texts’, the shift from seeking the motives that caused action, to inferring from accounts the motives involved in account-giving, narrows the realm that interests us and thus makes research a great deal easier. It becomes easier still if we argue, as many postmodernists have done, that any sort of explanation (even one narrowed to the setting for account-giving, such as the courtroom and the religious crusade) is impossible: all we can do is literary criticism. We do not need to leave the office to study religious conversion: we just need a pile of conversion testimony texts. (p.163) At this stage I find myself sounding like a stuck record: as with so many other contributions to social science, what began life as a sensible qualification to previous practice was inflated into a misleading analytical dead end. Mills was obviously right to point out that some occasions of explaining ourselves to others are themselves actions that shape the explanation. We are often trying not only to explain what we did but also to appear rational or Page 9 of 19

 

Conversion sensible or sensitive or pleasant. But it does not follow that all accounts are of that nature and hence that we should abandon the search for causal explanations.26 As I noted in Chapter 2, even in its very narrow conversation analysis form, the ethnomethodology extension of the Wright Mills principle does not avoid imputing motives for actions. For example, ‘You doing anything this evening?’ can be heard as either a question or an invitation, and the difference between them lies not in the words that are spoken but in the guessed intentions of the speaker. Clearly, we are more likely to be accurate when imputing simple and short-range motives than when trying to dig out the reasons why someone joined a sect, but this is a matter of degree, not of absolute distinction. A second reason for rejecting the extreme position on accounts is that it is selfrefuting. When Wright Mills says ‘when an agent vocalizes or imputes motives, he is not trying to describe his experienced social action’, he is making an empirical assertion that is testable only against other sources of evidence about the reasons for his action.27 If, as I have done, we explain why Christopher Edwards claims to have been brainwashed by his desire to absolve himself of responsibility for joining the Moonies or by a wish to become a talk-show favourite and sell copies of his book, we are implicitly saying that he was not brainwashed and that, if we could find video tapes that showed his life before he met the Moonies and the entirety of his interaction with them, we would see a picture different from the one he is painting. Typically, ‘functions of accountgiving’ explanations seem so plausible that we tend to accept them without appreciating that we are in effect engaged in motive substitution. The actor says one thing; we say something else. We do not notice that this substitution assumes the very thing that the general approach denies: the possibility of finding an accurate explanation of the past. One might say in response that there is no need to investigate the actual past of, for example, religious revival converts, because the accounts given in such a context are so similar that they cannot (p.164) possibly be accurate. That is a good point. Law courts have been persuaded that the presence of identical phrases in the notebooks of a number of police officers shows that such notes were not made separately and contemporaneously but were the product of post hoc collusion. However, conversion testimonies are rarely alike enough to suggest plagiarism. Furthermore, even if we accept that similarity of conversion motifs shows the influence of the situation of testifying, this does not entirely rule out the use of such testimony as the raw material for explaining conversion, because it is possible to reconcile conventional use of accounts with the Millsian perspective through the process of selection. Unless we suppose that what would count as a socially acceptable account of some previous action is extremely constrained, it must be the case that there are a wide variety of motives that could be claimed. If there is a range of acceptable accounts, it may be that people pick the one that best represents their experience. That is, the Page 10 of 19

 

Conversion new focus on accounts and ‘situated motives’ does not, as the relativist postmodernist thinks, preclude the conventional social science practice of using what people say about themselves as evidence from which to construct an explanation of their behaviour. It draws our attention to the possibility that some elements are played up (or down) for interactional advantage. It makes us attend to the possibility that some themes have become conventional. But it does not prevent the explanation of action. That is, to justify motive substitution, it is not enough to show that accounts are similar; the similarities in accounts may reflect actual similarities in the previous lives of the speakers. But, in order to test that, we have to assume that it is possible to do conventional social science reconstruction and explanation. Third, to say that accounts cannot be taken naively as evidence about the reality behind them is not itself a sufficient justification for supposing that we can never test accounts or that in every case underlying realities are unknowable. After all, in our everyday lives we repeatedly deal with the problem. We are regularly confronted with dissembling children, spouses, colleagues, and students, and yet we do not become practical relativists.28 As we did in the Edwards’s example by showing that the details he offers do not justify the brainwashing gloss, we become skilled at seeking signs of veracity from the internal evidence of the account itself. And we often have other sources of evidence. This is important. Insofar as the ethnomethodological critique of using accounts as data for social scientific (p.165) explanation is plausible, it works best when speech acts are taken in isolation: when the account is all we have. When we have a wide variety of different sorts of ‘accounts’ (what the convert said in different settings, direct observation, contemporaneous diaries, letters, and the like), the ethnomethodological reservations become less pressing. It may not be easy to know the past, but I am not persuaded that we must give up trying. In brief, an interest in the discourse of conversion seems entirely legitimate, but there is no reason for it to be exaggerated into relativism or for it to displace the more conventional interest in explanation. I should stress that the ‘vocabularies of motive’ reinterpretation of what people say is perfectly reasonable when what they say does not fit well with their actions. Our identification of a compliance effect—when survey respondents offer responses that they feel are more acceptable or honourable—is an example of reinterpretation that is justified by the fact that we know from other sources (actual counts of church attendance, for example) that the account being given is incorrect. Another interesting example is the implausibility of responses to questions about why people have given up attending church. People are often reluctant to state outright that they gave up going to church because they realized they no longer believe and instead cite either accidental obstacles (such as moving house and ‘not yet’ finding a suitable church) or features of the Page 11 of 19

 

Conversion church they rejected (‘too much infighting’ is a common criticism). When one considers the enormous range of worship outlets within easy reach of most Westerners, such explanations start to look like polite ways of avoiding the real problem. It is unfortunate that many such studies ask their questions only once and do not press respondents on how hard they tried to find a church in their new district or how hard they tried to find a church that was not engaged in sectarian arguments with competitors.29

Explaining Large-Scale Change Another possible reason for waning social science interest in explaining individual conversion is that it can seem trivial, especially when the conversion involved joining movements that themselves now seem trivial. The older social science questions about large-scale religious change seem much more important in a world where fundamentalism (p.166) is on every news broadcast, where the collapse of communism has permitted a resurgence in religiously informed nationalisms, and where one of the biggest changes in the global religious landscape is the spread of Pentecostalism in Latin America and Africa.30 But, and this is an important reservation, while an interest in the large scale is quite proper, there is a danger in abandoning altogether attempts to explain individual conversion. To introduce a point taken up again in Chapter 9, like it or not, our accounts of large-scale change assume models of individual action, and, unless we are careful, we end up implying implausible models of individual motivation. The point can be seen in competing explanations of the success of Pentecostalism in Latin America. In the left-wing explanation, the important fact about Pentecostalism is that it encourages an individualist and quietist response to social, economic, and political troubles; without it there would be socialist revolution. Hence Pentecostalism serves the interests of the USA. Hence it is explained by the interests of American capitalism. In the right-wing version, religious conversion is a rational reaction to changing social circumstances. Swapping the traditional, fatalistic, organic, and community-based Catholic faith for Pentecostalism allows the development of a personality better suited to the demands of urban capitalist economies, offers a strongly supportive association of similarly situated individuals, encourages economically effective puritanical behaviour, gives ordinary people leadership roles, and develops such useful generic skills as literacy and group management. Both these explanatory traditions have their merits, but both can easily become caricatures. In the Marxist version, the imagined Pentecostal convert is a fool: led by others into a false analysis of his or her problems. In the positive adaptation model, the convert is a social-climbing cynic: switching religions because the new one seems more useful for getting on. In common with most examples of functionalist analysis—where some new behaviour is explained by its latent benefits—both versions neglect the point that, for any institution or pattern of social action to have latent functions, it must also have manifest ones Page 12 of 19

 

Conversion and those must be convincing to the actors. For example, the puritanism that comes with conversion to Pentecostalism has clear advantages for people whose precarious grip on prosperity would be much strengthened by sexual continence, the preservation of the family, continued employment, and avoidance of the conspicuous consumption of carnival. But the motivational (p.167) underpinning of these virtues is the belief that God requires them: that is what gives people the resilience to sustain an ascetic lifestyle and to defer gratification against the considerable temptation to have fun now. If latent functions can explain human action only if we can connect the secondary benefits of conversion to the primary appeal of the new beliefs without implying that the converts are stupid or cynical, we still need to address the issue of belief: why is it that some people find one world view more persuasive than another? The solution is often quite simple. The problem for the pre-convert is that the truth of the key propositions of any religion is not knowable in this life. We cannot ourselves test whether getting right with Jesus or accepting the Divine Precepts will assure us of salvation. We have to take such claims on trust. However, we can see the latent or secondary benefits of the new religion in the personalities, social circumstances, and lifestyles of those who have already converted. One of the bases of social trust is social similarity. In the early nineteenth century, the Scottish highlands (parts of which had remained Catholic long after the lowlands had embraced the Reformation) converted en masse to a particularly enthusiastic brand of evangelical Calvinism. In so doing, the highlanders rejected not only the ancient Catholic faith that had survived where clan chiefs had been strong enough to protect it from the Reformation but also the moderate Protestantism of the parish ministers of the state Church of Scotland. That widespread religious change is often explained in terms very similar to the explanation of the spread of Pentecostalism in Latin America. Drastic change in people’s economic circumstances called their old world view into question. In this case the change —generally referred to as ‘the Clearances’—was the enclosure of once-common land and the displacement of subsistence farmers to make way for sheep. The forced pacification of the highlands that followed the failure of the 1715 and 1745 Jacobite rebellions radically changed land use. Clan chiefs and their lieutenants who had once valued a large peasantry because it could be mobilized as a military force now saw their lands simply as sources of rent and forced the peasants to move to villages on the coast (where they were encouraged to take up fishing or, for a short time, kelping) or to emigrate to the colonies. In many parts of the highlands and western isles, the Clearances caused great hardship and created social unrest that was settled only in 1886 when the Crofting Acts gave tenants security of tenure and ensured access to common grazing lands. (p.168) Clearly the destruction of the economy and forced migration undermined the previous world view and put people in the market for a new understanding of their place in this life and the next. But that does not explain Page 13 of 19

 

Conversion why people chose evangelical Calvinism rather than the moderate Presbyterianism of the parish ministers. To explain that we need to appreciate one matter of substance and two issues of social relationships. Evangelical Calvinism was well suited to the people’s distress because, as with all chiliastic religions, it offered a radical critique of current circumstances and a promise of compensating reward in the next life. Their present problems were a result of sin: their own sins but, far more, the sins of their social superiors. And God would reward them in the next life and punish their masters as the social order was reversed: the first shall be last and the last shall be first. The new religion thus offered in the supernatural realm what could not be achieved in this world. Revolution was impossible but divine mercy and justice would eventually settle accounts. But that some new idea is apparently superior from the perspective of the bystander does not guarantee that those involved will see it that way. Whether innovations (practical as well as ideological) are seen as superior will depend to an extent on relationships between the audience and the carriers of the new world view. Hence our need to consider two sets of social relationships. The parish ministers were discredited by their social remoteness from their flocks and by their willingness to support the rapacious landlords, who paid their stipends and who funded their churches and manses. We can be confident that the nature of relations between congregations and their pastor was significant because in the rare cases where ministers did side with the people against the improving lairds—Alexander Sage of Kildonan in Sutherland and his son Donald Sage of Resolis in Ross-shire were notable examples—the people remained loyal to the state church.31 The second set of relevant relationships is that with the carriers of the new faith collectively known as ‘The Men’, not because they were not women but because they were not ministers. Some of the Men were entirely amateur preachers of the gospel; others were employed by the Church of Scotland or evangelistic organizations as lay catechists and schoolteachers. But the latter were so poorly paid that their occupations gave them spiritual status without separating them from the people they sought to influence. Like those they addressed, they were crofters who struggled to make a living from a vicious sea and a stony land. When the people heard the (p.169) Men preach, they heard people who were like themselves but better. Not better off, like the ministers, but more content and better assured of salvation. And the Men regularly proved their alignment with the dispossessed by preaching against their social superiors. In this example we can see how conversion can be explained without supposing that the converts were either the dupes of social forces beyond their ken or cynical pretenders to a new faith motivated primarily by a wish for its secondary benefits of consolation for their suffering and safe form of social criticism. The key is plausibility. By their personalities, their social positions, and their actions, the Men made their message more plausible than that offered by the parish

Page 14 of 19

 

Conversion ministers. This is the crucial middle term in the equation that, for convenience, we sometimes shorten to ‘social problems caused religious change’.

Conclusion Although the conversion of the Scottish highlands and islands was so effective that, when the Church of Scotland split in 1843, almost all those congregations joined the opposition Free Church, some people remained loyal to the old church. We can say the same of the spread of Pentecostalism in Latin America. We can construct a convincing explanation of rapid social change disrupting long-standing social structures and their associated cultures and thus putting people in the market for a new perspective. We can show how the basic themes of Pentecostalism were better suited to the circumstances of a rapidly urbanizing society than was the previously dominant organic Catholicism. We can show how the thousands of amateur Pentecostal pastors both offered a gospel well attuned to the situation of their audiences and embodied both that situation and the improvement to it that would follow conversion. And yet many did not convert. This brings me to the limits of social science. Some attempts to explain religious conversion fail because, as Lofland put it: ‘various kinds of demographic characteristics, structural, and/or personal frustrations and so forth … are woefully incomplete’.32 Paul Heelas and Linda Woodhead’s explanation of the appeal of holistic spirituality suffers from the defect of presenting as causes of such interest near-universal features of the modern world.33 When they explain why women are more likely than men to be attracted to (p.170) yoga, meditation, and the like, they present features of the situation of almost all women as explanations of the choices of remarkably few women. Wallis’s explanations of the appeal of various NRMs has the same defect of specifying a population vastly greater than the number of those who actually show any interest.34 Those explanations are poor and can be improved upon. For all Lofland’s later disavowal of his seven-stage model, I would argue that it or something like it, by attending to the relationships that develop between preconverts and group members, probably represents the closest we can come to explaining religious conversion, especially when we add the Bromley and Shupe observation that people often experiment by acting as if they believed and move into complete conviction only when they find their new roles more satisfying than previous alternatives. I do not think that further elaboration will improve our explanations, because we have reached the point at which personal idiosyncrasies become relevant. We can explain broad patterns of behaviour in terms of probabilities. Within those broad patterns we cannot explain why one particular individual converts while another does not. To an extent we can explain why the Men were less successful in some parishes than others by the attitudes of the parish ministers, but we cannot hope to offer a sociological explanation of the dissent of the few people of the parish of Lochs who did not defect to the Free Church in 1843 because at that level choices are informed by personal features that are not susceptible to social science generalization. It Page 15 of 19

 

Conversion may be that one family had a running dispute with their neighbours about the quality of their fences; the neighbours joined the Free Church, so they did not. Their son jilted our daughter, so we will remain with the parish church. Social science can hope to explain common patterns of behaviour; it cannot hope to explain the actions of every individual person taken separately. In addition to trying to explain religious conversion, this chapter has addressed a number of important methodological problems common to that endeavour. I have argued that it is possible to accept the view that conversion testimonies, like other forms of accounts of action, owe some of their features to the situation in which such accounts are given and to a desire to render one’s previous actions explicable and acceptable without abandoning attempts to explain conversion and limiting ourselves to literary criticism of account texts. I have also tried to show how, by attending to the role that social relationships play in making new ideas or practices more or less plausible, we can avoid the (p.171) common mistakes of implying that people are either motiveless or cynical in their motivation. Finally, I have conceded the limits of social science generalization. We can hope to explain why rural English Methodism recruited better from villages where ownership was divided between a number of landlords than from villages owned by a single estate. We can hope to explain why, in every modern division in Scottish Presbyterianism, the highlands and western isles have come out for the conservative evangelical wing. We can hope to explain why the world-rejecting new religions of the late 1960s particularly appealed to white, middle-class, young, single, and childless college students and graduates. We cannot hope to explain why any particular individual converted and another apparently similarly situated person did not. That is beyond the scope of social science. Notes:

(1.) F. M. Davenport, Primitive Traits in Religious Revivals: A Study in Mental and Social Evolution (New York: Macmillan, 1905); A. W. Eister, Drawing Room Conversion: A Sociological Account of the Oxford Group Movement (Durham, NC: Duke University Press 1950). (2.) For an excellent review of the field, see A. D. Anthony and T. Robbins, ‘Conversion and “Brainwashing” in New Religious Movements’, in J. R. Lewis (ed.), The Oxford Handbook of New Religions (Oxford: Oxford University Press, 2008), 243–97. (3.) W. Sargant, Battle for the Mind: A Physiology of Conversion and Brainwashing (London: Heinemann, 1957). (4.) Sargant, Battle for the Mind, 15.

Page 16 of 19

 

Conversion (5.) R. Lifton, Chinese Thought Reform and the Psychology of Totalism (New York: W. W. Norton, 1961), and E. Schein, Coercive Persuasion (New York: W. W. Norton, 1961). (6.) B. Zablocki, ‘Exit Costs Analysis: A New Approach to the Scientific Study of Brainwashing’, Nova Religio, 1 (1998), 216–49. (7.) Lifton, Chinese Thought Reform, 218–20. (8.) This is even more clearly the case in the studies of ‘possession’ in Voodoo and related spirit cults: W. Sargant, The Mind Possessed: A Physiology of Possession, Mysticism and Faith Healing (London: Heinemann, 1973). (9.) E. Barker, The Making of a Moonie (Oxford: Blackwell, 1984). (10.) J. Lofland and R. Stark, ‘Becoming a World-Saver’, American Sociological Review, 30 (1965), 863–74; J. Lofland, Doomsday Cult (Englewood Cliffs, NJ: Prentice Hall, 1966). (11.) R. Wallis, The Elementary Forms of the New Religious Life (London: Routledge and Kegan Paul, 1984). (12.) Lofland, Doomsday Cult, 33. (13.) Our ability to treat social bonds (or their absence) as a variable in causal explanations of conversion is debated in D. Snow and C. L. Phillips, ‘The Lofland– Stark Conversion Model’, Social Problems, 27 (1980), 430–7, and R. Wallis and S. Bruce, ‘Network and Clockwork’, Sociology, 16 (1982), 102–7. (14.) R. Straus, ‘Changing Oneself’, in J. Lofland (ed.), Doing Social Life (New York: Wiley, 1976), 252–73, and ‘Religious Conversion as a Personal and Collective Accomplishment’, Sociological Analysis, 40 (1979), 158–65. (15.) H. Garfinkel, Studies in Ethnomethodology (Englewood Cliffs, NJ: PrenticeHall, 1967), 68. For a critical evaluation of Garfinkel’s claims, see M. Lynch, ‘Revisiting the Cultural Dope’, Human Studies, 35 (2012), 223–33. (16.) D. G. Bromley and A. Shupe, ‘“Just a Few Years Seem like a Lifetime”: A Role Theory Approach to Participation in Social Movements’, in L. Kriesberg (ed.), Research in Social Movements, Conflicts, and Change (Greenwich, CT: JAI Press, 1979), 159–86. (17.) M. Hamilton, Sociology of Religion: Theoretical and Comparative Perspectives (London: Routledge, 2001); B. J. Zinnbaeur and K. I. Pargament, ‘Spiritual Conversion: A Study of Religious Change among College Students’, Journal for the Scientific Study of Religion, 37 (1998), 161–80; J. A. Belzen, ‘Religion as Embodiment: Cultural–Psychological Concepts and Methods in the Page 17 of 19

 

Conversion Study of Conversion among “Bevindejiken”’, Journal for the Scientific Study of Religion, 38 (1999), 236–53. (18.) Belzen, ‘Religion as Embodiment’; Zinnbaeur and Pargament, ‘Spiritual Conversion’. (19.) J. Lofland and N. Skonovd, ‘Conversion Motifs’, Journal for the Scientific Study of Religion, 20 (1981), 373–85. (20.) C. Wright Mills, ‘Situated Actions and Vocabularies of Motive’, American Sociological Review, 5 (1940), 905. (21.) J. Rose, Intellectual Life of the British Working Classes (New Haven: Yale University Press, 2001). (22.) J. M. Atkinson and P. Drew, Order in Court: The Organization of Verbal Interaction in Judicial Settings (London: Macmillan, 1979). (23.) B Taylor, ‘Conversion and Cognition: An Area for Empirical Study in the Microsociology of Religious Knowledge’, Social Compass, 23 (1976), 21. See also A. Blum and P. McHugh, ‘The Social Ascription of Motives’, American Sociological Review, 36 (1971), 98–109, and M. B. Scott and S. M. Lyman, ‘Accounts’, American Sociological Review, 33 (1968), 48–62. (24.) Taylor, ‘Conversion’, 21. (25.) B. Latour, Rejoicing or the Torments of Religious Speech (Cambridge: Polity Press, 2013). (26.) For a more detailed critique of ethnomethodology’s treatment of motives, see R. Wallis and S. Bruce, ‘Rescuing Motives’, British Journal of Sociology 34 (1983), 61–71, and S. Bruce and R. Wallis, ‘Accounting for Action: Defending the Common-Sense Heresy’, Sociology, 17 (1983), 102–11. A more sympathetic evaluation can be found in J. A. Beckford, ‘Accounting for Conversion’, British Journal of Sociology, 29 (1978), 249–62. (27.) C. Wright Mills, ‘Situated Actions and Vocabularies of Motive’, American Sociological Review, 5 (1940), 907. (28.) It is an obvious point, but it is always worth repeating; social scientists who argue for a relativist epistemology will be worth taking seriously only when they are consistently relativistic. Philosophers disparage this tu quoque (in English ‘You’re as bad!’) method of critique as a logical fallacy, and it is true that inconsistency and hypocrisy do not necessarily invalidate the speaker’s argument, but it is reasonable to take epistemological positions only as seriously as the proponent takes them.

Page 18 of 19

 

Conversion (29.) For examples of this sort of study, see P. Richter and L. J. Francis, Gone but Not Forgotten (London: DLT, 1998), and S. Aisthorpe, Investigating the Invisible Church: A Survey of Christians Who Do Not Attend Church (Edinburgh: Church of Scotland Mission and Discipleship Council, 2014). (30.) P. Gifford, ‘Some Recent Developments in African Christianity’, African Affairs, 93 (1994), 513–34, and African Christianity: Its Public Role (London: Hurst, 1998); D. Martin, Tongues of Fire: The Explosion of Pentecostalism in Latin America (Oxford: Blackwell, 1990), and Pentecostalism: The World their Parish (Oxford: Blackwell, 2001). (31.) D. Sage, Memorabilia Domestic: Parish Life in the North of Scotland (Edinburgh: Albyn Press, 1975). See also S. Bruce, ‘Social Change and Collective Behaviour: The Revival in Eighteenth Century Rossshire’, British Journal of Sociology, 34 (1983), 554–72. (32.) J. Lofland, Protest: Studies of Collective Behavior and Social Movements (New Brunswick, NJ: Transaction Books, 1985), 128. (33.) P. Heelas and L. Woodhead, The Spiritual Revolution (Oxford: Blackwell, 2005). (34.) R. Wallis, The Elementary Forms of the New Religious Life (London: Routledge and Kegan Paul, 1984).

Page 19 of 19

 

Social Theory and Religion

Researching Religion: Why We Need Social Science Steve Bruce

Print publication date: 2018 Print ISBN-13: 9780198786580 Published to Oxford Scholarship Online: August 2018 DOI: 10.1093/oso/9780198786580.001.0001

Social Theory and Religion Steve Bruce

DOI:10.1093/oso/9780198786580.003.0008

Abstract and Keywords Students of religious groups or activities are often pressed to provide an appropriate theoretical background for their work. The practical difficulty is that much social theory is actually philosophy with little empirical basis. This chapter considers the merits of four different sorts of social theory: normative theory that tells us what is good and bad; zeitgeist metaphors that capture the nature of modernity with some eye-catching word or phrase (for example, Baumann’s ‘liquid modernity’); agenda-setters (such as feminism or postcolonial theory) that want new questions asked in new ways; and sociological explanation grounded in reasonable extrapolations from empirical research. The drawbacks to the first three are identified as a way of advertising the virtues of the last type of theory. Keywords:   social theory, normative theory, zeitgeist metaphors, agenda-setting theory, sociological explanation

Introduction This chapter considers the use of social theory in the empirical study of religion. It is not easy to define social theory. ‘Theories’ plural or ‘theory’ preceded by a substantive qualifier (such as rational choice theory discussed in Chapter 9) are usually internally consistent bodies of explanations of particular institutions or patterns of social action, with clear specifications both of the general principles that have informed the explanations and of the implications of specific explanations for those fundamental principles. But ‘theory’ singular or ‘social theory’ tends to be highly abstract and sweeping generalization about the nature of social life—Judith Butler’s ‘performativity’ is an example—or broad

Page 1 of 17

 

Social Theory and Religion characterizations of social evolution, which, like Zygmunt Baumann’s ‘liquid modernity’, claim to capture the essence of modern world. There is a good reason why the sort of social theory that is discussed in courses with that title is often so abstract that it might be better described as social philosophy: to the extent that we wish to equip students who will specialize in a wide variety of areas of social life with ideas that will shape their research, it is desirable that those ideas be as general and as generally applicable as possible. Unfortunately, that generality often comes with the downside of a lack of specific application. To shape this presentation, social theory is divided into four types: normative theory, zeitgeist metaphors, agenda-setters, and sociological explanation. Though they have features that overlap, these are rather different animals. I will consider how these are currently used by students of religion and how that usage can be improved. To state my conclusion at the outset, I will make the case that the most useful sort of theory for students of religion as it is lived is that which is furthest from (p.175) the sort of philosophy that could be done without going outside and closest to the coalface of actual social research.

The Nature of Social Theory Normative theory tells us what the world should be like. It is not just description and explanation (and sometimes it is little of either); it is also a blueprint for a better future. Overtly normative work is more common in political theory than in sociology, but critical sociology, for example, is generally normative: that is, it takes sides. Zeitgeist metaphors are works that try to capture the essence of the modern age in a single jaunty image: disciplinary society, risk society, McDonaldization, network society, tribal society, and liquid modernity are all examples of such metaphors. Then we have the agenda-setters. Feminist theory, queer theory, and postcolonial theory are not in the first instance logically connected explanations of related social phenomena, though they can sometimes be that: they are ways of looking. For many adherents, their value lies in personal liberation. By identifying the social origins of problematic attitudes and circumstances, feminist theory, for example, may help women feel better about themselves.1 For the researcher, their primary value lies in alerting us to topics that have been neglected and to questions about those topics that have not previously been asked. I should add that much of the social theory that is heavy on these three tendencies is the work not of sociologists but of philosophers, psychoanalysts, and literary critics.2 We should not get bogged down in discipline boundary demarcation. Like W. G. Runciman, I tend to the view that ‘a distinction between sociology, anthropology and history will have meaning only in terms of incidental Page 2 of 17

 

Social Theory and Religion differences of technique’,3 but there is a clear difference between those subjects, which require direct engagement with an objective external reality, and philosophy or English literature, which can be practised in a sealed room. Not all closed room research is pointless. Peter Berger has pointed out that much mathematics has developed without reference to empirical research, yet the theorems that mathematicians have developed from thinking about the logical consequences of what are in effect just tautologies—a triangle has three sides only because (p.176) ‘triangle’ is what we call a flat three-sided object— nonetheless have significant application to the empirical world. For example, maths allows calculations of weight, speed, fuel consumption, orbits, and trajectories that allow us to fly men to the moon and bring them back safely (unless the Moon landings were really filmed on a lot in California). One could argue that philosophizing about the nature of social life is useful because it can give us testable assertions about the world. One example is the claim that all people are essentially religious; another is the assertion (tested in Chapter 10) that danger, by forcefully reminding people of their mortality and vulnerability, disposes them to be religious. My fourth tendency in social theory might be thought of as generalized social scientific explanation. Students of some specific social phenomenon develop their results, usually through systematic comparison, into ever more general propositions. Examples would include Emile Durkheim on suicide, John Goldthorpe on social mobility, Howard Becker on the socialization of dope smokers, and Harry Collins on working scientists.4 Here theory is an attempt to explain research findings in terms general enough to allow further refinement, extrapolation, and testing.

Theory in British Studies of Religion Before considering the merits (or otherwise) of each of these four tendencies for the social scientific study of religion, I will report a brief exercise in cheap-andcheerful research. It is useful to see just which social theorists figured most prominently in British studies of religion. Doubtless there are more rigorous ways of doing this but to get some idea of current usage I went through the bibliography of each article published in the Journal of Contemporary Religion (JCR) between 1995 and 2010 and noted how often theorists—defined in a cavalier fashion as those people who get anthologized and discussed in social theory books—were cited. The results were as follows. In those fourteen years, the top ten theorists in the JCR were, in reverse order: Theodor Adorno (with five mentions) and Clifford Geertz (with six). Roland Robertson shares the seventh equal spot with Michel Foucault. Zygmunt Baumann takes sixth place and just ahead of him is Anthony Giddens. Placed fourth is Emile Durkheim. (p.177) The top three were, in third place, Pierre Bourdieu; in second place, Max Weber; and, at number one, the most popular

Page 3 of 17

 

Social Theory and Religion social theorist with contributors to the Journal of Contemporary Religion is Peter Berger. In keeping with the principle that good research produces findings one does not expect, the most significant observation from my trawl is not the placings in this beauty contest, though the presence of Weber and Durkheim but not Karl Marx are surprises. It is the relative absence of any social theory. Almost half the 267 articles surveyed (48 per cent to be precise) cite no social theorists. In total, thirty-eight theorists are cited but seventeen of them (again almost half) are cited only once. Furthermore, at least half the citations seem ritualistic. For example, almost all Robertson’s citations accompany fleeting mention of globalization. Weber’s are justifications for focusing on understanding. Giddens is cited as an authority for the hardly contentious idea that social structure and personal agency both shape social action. And what put Berger in the top spot is not anything from his extensive work in the sociology of religion: it was his popularizing of the phrase ‘the social construction of reality’. In almost all cases, nothing in particular is said about how this or that bit of reality was socially constructed; the point being bolstered was almost always the vague one that culture matters. Of course, it does not follow that articles that cite no social theory are informed by none. The work of older scholars, such as Bryan Wilson and David Martin, is thoroughly pervaded by theory, though direct attributions are rare. And in some cases there is no need to reference particular theories because the influence is obvious. But, actually, many of the Contemporary Religion articles were theoryfree; they described in detail religious beliefs, attitudes, and practices. There is nothing wrong with that except lost opportunity. It is routine for articles in more narrow disciplinary journals such as the Journal for the Scientific Study of Religion (JSSR) or Sociology of Religion to justify their publication both by the novelty of the research reported and by their contribution to current analytical debates. The high theory of social science supposes that empirical research should advance development by testing some general (that is, theoretical) ideas. We know that in practice the theory–substance link often works the other way round. We research some religious movement or activity because it intrigues us, and, having collected our observations, we search for some theoretical argument to which our material can be turned. (p.178) But even this adventitious link between research and theory is often missing or tenuous in studies of religion. In a long career of refereeing journal submissions I have very frequently been confronted with theoretical justifications for research that are so obviously spurious that they border on the cynical: if nothing more germane occurs to the author who feels pressed to ‘include some theory’, almost any report of religious (or even ‘implicitly religious’) activity can be justified by claiming that it refutes a caricatured Page 4 of 17

 

Social Theory and Religion summary of the secularization thesis. But nonetheless very many articles in such journals as the JSSR do manage a fruitful interaction of theory and substance. The absence of that interaction in JCR articles does not necessarily render research publication pointless, but it is reasonable to describe it as missed opportunities to tell us how this particular study advances more general knowledge and how it relates to other similar studies. Readers can, of course, make their own connections, but, as the researcher knows the material better than anyone, he or she is best placed at least to begin the work of trying to explore the research’s wider implications. As I am about to criticize much contemporary social theory, one might think that to regret its absence is self-contradictory, like the judgement of the diner who complains: ‘The food was terrible. And such small portions.’ There is something in that, but rationality can be restored by noting in advance a point I will elaborate later: there is also a dearth of sociology in much superficially social scientific writing about religion in Britain.

What Sort of Theory should we Use? The case that more social theory would be a good thing having been made, we can go back to my four tendencies and consider what we might gain from each of them. Such summaries obviously involve a great deal of compression, but I hope I do not misrepresent the scholars discussed. Normative Theory

Bryan Turner has correctly noted that ‘professional sociology has difficulty addressing overtly normative issues’.5 There are two reasons (p.179) for this. The first problem is that, while taken in isolation, any normative theorist may be inspirational, it is hard to construct widely accepted grounds for choosing between competing visions of the good life. For social scientists working within the de facto positivist approach described in Chapter 2, the value of any theory is determined not just by internal consistency but also by how well it is supported by the available evidence and how well it survives repeated attempts to refute it. Not all normative social scientists ignore that sort of work; many are pro-cake and pro-the-eating of cake. Classically, Marxists have claimed that their work was both thoroughly scientific and, because it encourages revolution, better than science. But the cavalier treatment of evidence—a few research findings plucked out of context to show that the good life is both good and inevitable—generally shows disdain for the scientific method and makes it clear that understanding some past or present reality takes second place to promoting some extra-social scientific goals or values. I return to this point shortly, but the sporadic use of social facts to tart up what is primarily an apologia for a political position is rather similar to the structure of the sermons of the great American evangelist Billy Graham: the faith comes first, and a few nuggets of social observation are worked into the presentation to give it extra plausibility.

Page 5 of 17

 

Social Theory and Religion Anti-positivists reject conventional notions of evidence but, having done so, leave us with no grounds for accepting or rejecting any claims, other than our agreement (or otherwise) with the normative theory being punted. I recall one self-styled ‘critical criminologist’ dismissing the work of the great Edwin Sutherland6 with the rhetorical flourish ‘what has Sutherland ever done to promote popular struggles?’. (I would like to say that I am withholding his identity to spare his blushes. Actually, the author in question had so little impact on the field that I have forgotten his name.) But like Marx, that author clearly believed that the purpose of theory was to change the world rather than just (or even) explain it. Not being politically radical was enough to render Sutherland’s long career of serious research pointless. But, as any large number of us are unlikely to agree on what would count as good change, political rectitude is no help here. Edmund Burke said: ‘I do not vilify theory … No, whenever I speak against theory, I mean always a weak, erroneous, fallacious, unfounded or imperfect theory; and one of the ways of discovering that it is a false theory is by comparing it with practice.’7 Rather flippantly, I sometimes (p.180) think we should judge models of the good life in the way that potential religious converts judge new faiths: by how pleasant, congenial, or admirable are their proponents. Karl Marx’s promotion of the dignity of labour was quite at odds with his own reliance on the profits that his colleague Frederich Engels made from his inherited wealth. Walter Benjamin and Nicos Poulantzas committed suicide. Michel Foucault was a self-harmer and drug-abuser who tried to kill himself a number of times. Felix Guattari was a philanderer who bullied his wife into having affairs so that he did not feel guilty about his own infidelities. Louis Althusser did more than bully his wife: he murdered her. As these are not people you would invite round to your house, why look to them for instruction on the nature of the good life? Philosophers commonly reject this tu quoque or ad hominem argument on the quite correct grounds that the personality of an author is poor ground for judging his or her ideas, and normally I would agree, but there is one good reason—the absence of alternative tests—for making an exception in the case of normative theory. The problem with normative or critical social theory is that it offers no principles agreed outside that theory’s magic circle for distinguishing between alternatives. Old-fashioned ‘social science as science’ has shared principles to guide choosing between alternatives. Critical theorists reject those rules but cannot establish alternatives beyond how congenial we find their politics. Because Western social scientists typically tend to be left wing, that gaping hole in their method is often not noticed, but the problem becomes clear the moment we imagine scholars who are committed to a Christian or Muslim sociology applying for jobs in a British university. They would be rejected because we would argue that sociology has to go where the research evidence points: it cannot be constrained by a faith that claims the trump card of divine Page 6 of 17

 

Social Theory and Religion authority. This is not to say that pious Christians and Muslims cannot do sociology. The fourteenth-century North African Ibn Khaldun is often cited as the first sociologist, and many of the nineteenth-century American sociologists were initially inspired to study social problems by their Christian values.8 But conventionally Christianity and Islam draw their authority from divine revelation, claim an unchanging core of beliefs and values, and suppose that current social arrangements either are God’s will or should be changed to accord with that will. Any of those principles is incompatible with social research that does not have its conclusions written in advance. And the same is true for normative theory. And, if we do know the (p.181) endpoint in advance, the accompanying social science research is just window dressing. My second difficulty is that I cannot see how the majority of the research that is done by students of religion would be much improved by normative social theory. Normative theorists are useful for displaying one’s partisan loyalties or presenting a vision of the good society, but they offer little help in analysing detailed social phenomenon. That one is ‘on the side of’ the poor, the labouring classes, ethnic minorities, or women might make us feel righteous—what is now called ‘virtue-signalling’—but it does not help social analysis. Zeitgeist Metaphors

Zeitgeist metaphors are popular because they seem to capture, in a single memorable image, the essence of the modern world and thus provide a useful background for understanding the changing nature and status of religion. Thinking in the social sciences often involves the simple contrast: rural versus urban, men versus women, bourgeoisie versus proletariat, and the like. As ways of drawing our attention to significant variables, such bifurcations are useful. When applied chronologically (as periodization), such divisions offend historians but they can capture important changes. Durkheim’s shift in social solidarity from mechanical to organic is an example, as is Ferdinand Toennies’s community versus voluntary association contrast.9 The weakness of zeitgeist metaphors (especially when the contrast is missing or vague) is that most are far too general to help us much analyse small chunks of social life or, taking the process in reverse, to be tested by empirical research. Let us take the theory-research order first. Very many scholarly journal articles look like the following. The zeitgeist metaphor gets a big puff in the introduction. There then follows detailed analysis of some very small slice of human life, which makes no reference to the metaphor. In the conclusion, the metaphor reappears to add inappropriate heft to the conclusions of the study. I often suspect that the preferred theory could be replaced by almost any other and make little or no difference to the novel research being reported.

Page 7 of 17

 

Social Theory and Religion Now consider the research-theory order. I will give an example from a paper reviewed for a major journal. To preserve anonymity I will call the religious body in question the Snibbo Fellowship. (p.182) The paper began with an outline of Zygmunt Baumann’s claim that the modern world was characterized by ‘liquid modernity’ (hereafter LM), from which the author extrapolates the implication that the LM is secularizing because it undermines traditions and bodies of doctrine. The paper’s core was a detailed and informative description of the beliefs of the Snibbos, who collectively are characterized by a high degree of ideological fluidity and by an agreed unwillingness to be doctrinaire. The conclusion tried to relate the Snibbos to LM, but it was not clear whether the fact that the Snibbos had survived over a number of generations was being offered as refutation of LM’s implied commitment to secularization or as an illustration of what religion in an LM world looks like. Either reading was possible. The difficulty is scale. At most there are only 15,000 followers of Snibbo in the UK: a tiny fraction of the number of people who have some sort of religious commitment. Hence, if the survival of Snibboism is presented as refutation of LM, Zygmunt Baumann could reasonably dismiss it as irrelevant. No version of the secularization paradigm expects that religion will vanish entirely, and Baumann’s LM depiction of modernity is clearly about general tendencies and preponderant tones. Tiny exceptions cannot test the depiction. But then it is hard to see what would. If the theory is neither essential to the research findings, nor crucially tested by them, its inclusion is pointless. Although the work of Anthony Giddens is more detailed than most zeitgeist metaphors, even his ‘reflexive society’ is difficult to apply usefully. As Lorne Dawson says of his relevance to religion: ‘He seems to think about the role of religious orientations in a modern context in two not altogether consistent ways.’10 That I am not alone in suspecting a high degree of interchangeability between zeitgeist metaphors seems clear from the short life most enjoy. Were it the case that flashily attractive depictions of the modern condition were often confirmed by the studies that approvingly cite them or were of lasting benefit to empirical research, they should surely endure. But many do not. Twenty-five years ago, in the late 1980s, Roy Wallis and I wrote a general review of the British contribution to the sociology of religion.11 In it we took issue with two esteemed colleagues. Turner’s Religion and Social Theory (which can still be read with great profit) criticized British sociology of religion for not being involved in ‘any major theoretical debate in modern sociology’.12 In particular it had failed to engage ‘in neo-Marxist debates about modes of production and ideology, French structuralist discussions of subjectivity and power, and critical theory’s discussion (p.183) of knowledge, the state and legitimacy’.13 James Beckford made the same point in his Religion and Advanced Industrial Society when he said that ‘the sociology of religion has been intellectually isolated against, and socially isolated from, many of the theoretical debates which have invigorated Page 8 of 17

 

Social Theory and Religion other fields of modern sociology’.14 The important theorists we should have been attending to were, according to Turner, Althusser, Poulantzas, Jurgen Habermas, Barry Hindess, and Paul Hirst. Althusser and Habermas appear on Beckford’s list, which also contains Foucault, Antonio Gramsci, Alberto Melucci, Claus Offe, and Alain Touraine. By the late 1990s, probably only Habermas and Foucault were still read and them not often. Any late-twentieth-century list of prominent theorists would probably include Giddens, Ulrich Beck, Baumann, George Ritzer, and Manuel Castells—all fading as the generation of students they inspired reaches retirement.

An Aside on Genealogy and Governmentality Not all theory is as flexible as Baumann’s liquid modernity. Some popular big theory has quite definite purposes and specific content yet nonetheless is of little value to the social science student of religion. It has been said of Foucault’s ‘genealogical’ claims for modernity being characterized by the state’s interest in making us more readily controllable, that studies of governmentality typically show a clear preference for synchronic accounts. They portray forms of power/knowledge as monolithic, with state practices fitting seamlessly with practices of self-creation. This synchronic focus often leads to somewhat reified and homogenous accounts of modern power, with little sensitivity to diversity, heterogeneity, and resistance within and over time.15 Arguably all this would be improved by translation into English, but two concepts in particular require elaboration. By genealogy is meant a way of understanding and writing history that, like the search for one’s ancestors, picks out and traces back a few common threads. By governmentality is meant the way that governments or other powerful actors use knowledge to make policies that effectively control how we see ourselves and how we act. Both these notions combine in Foucault’s history of the evolution of punishment and modern medicine. (p.184) There are a number of weaknesses with genealogy. The first is that, as conventional historians would assert, it is poor history because it selects what fits its preconceived explanations. Like the snob’s claim to be descended from one illustrious ancestor while ignoring the vastly more numerous insignificant contributors to his or her gene pool, genealogy is too selective. For example, Foucault’s account of the increased rationality of discipline makes much of Jeremy Bentham’s design for a panoptical prison—one in which inmates would be constantly supervised (or at least could not assume that they were not being supervised). What those who are taken with this model fail to appreciate is that the Panopticon was a design, not a real prison. Far from being places of rational discipline run by supervising guards, most contemporary prisons are run by a small number of prisoners who bully, rape, sell drugs to, extort from, and Page 9 of 17

 

Social Theory and Religion generally terrorize the rest of the inmates. The humane rationality of Bentham’s Panopticon may well inform the rhetoric of prison management, but it has little or nothing to do with the reality. Second, Foucault’s genealogy fails to recognize the importance of the unintended consequences of social action. As Weber famously demonstrated in his argument that the religious reforms of the Protestant Reformation eventually and unintentionally created the spirit of capitalism, the long-run effects of changes are often a very long way from anything intended by those who set them in motion. Third, it ignores or contradicts actors’ stated motives. Whatever justifications doctors, psychiatrists, or penal reformers offer, they were really contributing to state control. As I argued in Chapter 3, we must sometimes argue with our subjects: they may not understand themselves or they may understand themselves but present a different picture to us. We can also study the unintended consequences of social action without reference to the motives of the actors: Weber’s Protestant ethic thesis is a good example. But what Foucault and his disciples advocate is something importantly different: now the unintended consequences of social action all miraculously and implausibly converge on increased governmentality. Agenda-Setting Theory

There seems more point to agenda-setting theory. Even if we reject the case, raised in Chapter 5’s discussion of bias, that scholars cannot (p.185) transcend either the ideological limits of their upbringing or the current interests of their class, gender, national, ethnic, or professional identity, we can readily accept that collective enterprises can be blinkered. And here the outsider has an advantage. There is always value in asking new questions, and a new set of interests (in the topic rather than the bias sense) will often cast new light on old problems. Having said that, I would caution against two common dangers in the attraction of agenda-setting theory: exaggerating what is distinctive about the new and what is defective about the old. It is a common fact of academic life that those who believe they have discovered some important but neglected facet of social analysis advertise their discovery with grand programmatic statements. As an example of such exaggeration, I offer the initial programmatic statements made on behalf of ethnomethodology, especially in its subfield of conversation analysis or CA. Harold Garfinkel made a great fuss of criticizing previous sociology for relying on unprovable assumptions about motives. Ethnomethodology was superior because it did not make guesses about ‘what goes on in people’s heads’.16 And yet it seems clear to non-partisans interested in the social organization of talk that the most basic tools of conversation analysis do rely on guesses about actor motives and intentions. For example, a question and a request are distinguished not by Page 10 of 17

 

Social Theory and Religion grammatical form but by the intention of the first speaker in the adjacency pair. So the person who replies to ‘Do you know the time?’ with only a curt ‘Yes’ will either have made a mistake or be teasing, as will become clear when one or other speaker tries to repair the fracture. Either the first speaker adds ‘No, I mean “what time is it?”’ or the second speaker adds ‘Ha. Only kidding. It’s 3.30’. Similarly, the classic ‘Are ye dancin’?’ grunted by an apparently indifferent Glasgow male to a wallflower female in a dance hall is not a question but an invitation. One could argue that the sorts of motives and intentions assumed by CA are considerably narrower and more mundane than those assumed in conventional sociology and are thus less likely to be mistaken. But it remains the case that the research done within the ethnomethodology canon, worthwhile as much of it is, now seems far less novel than the paradigm shift advertised in the original programme. We can find another example of exaggeration turning virtue into vice in Turner’s work on the sociology of the body. It was inspirational in reminding social scientists that people are not just minds or personalities, but he grossly overstates his case in defining religion by (p.186) ‘the function of controlling the sexuality of the body in order to secure the regular transmission of property via the family’.17 Religion does indeed often do that, but it does many other things besides. The second danger of agenda-setting theory is that, in asserting the primacy of its agenda, it ditches too much of the old research. To its eternal disgrace, much social science until the 1970s paid little overt attention to gender, and, when challenged by feminists, many sociologists offered at best weak justifications for that neglect. But agreeing that new work needs to be done does not tell us what to do about the old stuff. For example, an American feminist glosses the secularization theory of religious change as follows: The story, generated by elite white European male scholars, tells their understanding of the displacement of powerful state churches by secular authorities in the public realm, with a subsequent loss of status for the men who continued to lead those churches at all levels, and at the same time a migration of men like themselves out of the churches … the secularization story is told from the vantage point of the unmarked mainstream, whose religion is mapped onto the dominant culture … in telling the story of secularization in terms of decline, we tell a partial story of elite white men.18 This is a complex claim. By suggesting that ‘telling the story of secularization in terms of decline’ is just one option among many, it rejects the fact of secularization. It also offers an explanation of why male sociologists have got it wrong. It is certainly true, as I have argued at length elsewhere, that secularization is much more than just decline.19 But the fact of decline is Page 11 of 17

 

Social Theory and Religion unarguable. However we measure it, every index of interest in, and the power of, religion in the industrial liberal democracies of the West shows decline. Even in the USA, where decline started much later than in Europe, regular churchgoing fell from 40 per cent of the population in the 1950s to around 20 per cent in 2010. But, quite independent of that, the explanation of why commentators got it wrong (if indeed they did) is unconvincing. That Marx, Weber, and Durkheim were white Europeans is beyond contest, although, as almost all Europeans of their era were white, their colour is unremarkable. That they were part of the elite is simply wrong. That their views are in some sense explained by their loss of status and that the secularization paradigm is solely or even primarily the work of members of dominant churches are both false. Many church leaders noted, and (p.187) were concerned about, the decline of their churches, but that tells us nothing about sociological commentators on secularization. If one was really interested in the somewhat tangential task of developing a biographical treatment of the sociology of secularization, one would better begin by noting the prevalence of outsiders, often émigrés and displaced persons, who had an awkward and marginal relationship to power.20 Such leading lights of the Scottish Enlightenment as Adam Smith, David Hume, and Adam Ferguson were at best lower middle class. Lords Monboddo and Kames were self-made lawyers who acquired their apparently aristocratic titles on being raised to a particular level in the Scottish court system. Weber was wealthy, but his liberal politics and his periodic bouts of mental illness made him a marginal figure in German public life. Marx was the son of a Jewish convert who spent most of his life in exile and who was marginalized both by his politics and by his poverty. Durkheim was also by origins an outsider: his family were Jews. The early part of his career was precarious, and he ended his days in a right-wing France marginalized by his liberal opinions. The elite and establishment depiction becomes even less sustainable if one moves to the post-1945 contributors to the secularization paradigm. Ernest Gellner was born in Paris to a middle-class Jewish family from Bohemia: people whose state, the Austro-Hungarian Empire, had been dissolved in 1918. He was educated in Prague, in an English-language school, before his family moved to England. He studied at Oxford, but fought with a Czech émigré unit when he joined the army, and after the war returned to Prague to study before being again made stateless by the Communist takeover. David Martin’s father was a taxi driver, and the origins of Bryan Wilson, Roy Wallis, and James Beckford were equally humble. Even the male part of the explanation is true only for that period of European academic life when most scholars of every discipline were male. And the depiction of the consequences of social background for the key ideas is as inaccurate as the account of social background. Some versions of the secularization story may be ‘told from the vantage point of the unmarked mainstream’, but an important part of the explanation of secularization concerns

Page 12 of 17

 

Social Theory and Religion the success of subordinate and peripheral religious movements chipping away the privileges of established churches. To the extent that we are influenced by any agenda-setting theory, we can dismiss everything that has gone before, or we can treat the (p.188) effects of omission as a research problem in its own right and ask to what extent a different set of questions would have changed the research and its conclusions. To continue with the secularization example, I would argue that much of the explanatory part of the secularization paradigm stands up pretty well to a gender-minded re-examination. For example, the social–psychological claim that being exposed to a range of different religious perspectives creates problems of certainty and conviction for the believer is a properly asexual claim. It rests on universal assumptions about the difficulty of believing one is uniquely correct when one enjoys pleasant and rewarding interaction with people who hold different views. It may be wrong, but, unless we are to suppose that women are naturally and significantly much more (or less) anti-social, obstinate or dogmatic than men, it is not wrong because it fails to distinguish between the sexes. The same can be said for the proposition that effective technologies displace occasions for resort to religious solutions and thus reduce the presence of religious ideas. The illustrations that we could draw from the life experiences of women and men are different (as are the examples we would draw from different social classes), but the key principle —that religion loses authority as the range of events for which it appears to offers the best solution decreases—seems genuinely universal. Furthermore, gender differences in outcomes may be perfectly well explained by universal propositions. The idea that rewarding social interaction with people of different faiths weakens dogmatism fits well with the observation that generally male church adherence declines before that of women: in many settings, gender differences in domestic roles and in employment patterns meant that men were more likely than women to interact with strangers and to perform jobs that required the sublimation of private preferences to religiously neutral public roles. To summarize, there is obvious value in agenda-setting theory that alerts us to omissions and identifies new research problems, but we should beware of exaggerating both the virtues of the new and the vices of the old. Sociological Explanation

Finally I come to sociological explanation. Not surprisingly, as a sociologist who just happens to study religion, it is in general sociological explanation that I find a body of knowledge from which (p.189) students of religious phenomena can borrow with advantage. For example, in the chapter on religious conversion I drew implicitly but heavily on research in mass communication and in what was called ‘the diffusion of innovation’ to establish the importance of affective social Page 13 of 17

 

Social Theory and Religion bonds in persuasion.21 After the Second World War and inspired by theorists (such as Theodor Adorno) who had been made refugees by that war, many social scientists promoted the view that modern ‘mass society’ was an aggregate of isolated individuals that lacked the low-level interpersonal connections of what we now call ‘civil society’. As an atomized audience for mass media, such people were unusually susceptible to recruitment by extremist movements. Characteristics within the audience might be shared in the sense of ‘common to’ but not in the sense of being the subject of communication. Thinking began to change when the Paul Lazarsfeld et al. study of the 1940 presidential election found that media coverage did little to change minds. When people did switch, the influence was generally a friend.22 Elihu Katz, Lazarsfeld, Robert Merton, and others demonstrated that, rather than directly recruiting atomized individuals, mass-media communication actually worked through local opinion leaders in a two-step fashion. As Luther Gerlach and Virginia Hine demonstrated with the example of the role of radio evangelism in the spread of Pentecostalism, mass media worked best when they were mediated by ties of personal influence.23 The conclusion was reinforced by studies of fashion and of the diffusion of medical innovation.24 Social scientists who study mass communication should know about that body of social theory, as should social scientists who have had a good broad education in a social science discipline. But it may well be missed by someone coming to the problem of conversion from religious studies or practical theology who reads current sociology of religion and high social theory but not empirical sociology.25 One virtue of sociological explanation is that, if well presented, we can readily see on what the general propositions were based and what would count as critical tests. There is not the flaccidity one finds in zeitgeist metaphors. The student of the Snibbos thinks that Baumann’s liquid modernity idea somehow captures some general essence that both enhances our understanding of the Snibbos and can be tested by the Snibbos, but the notion of liquid modernity is so vague that it is hard to see it as especially illuminating in this example or in any sense tested by the example. For reasons elaborated in Chapter 9, I have (p. 190) considerable difficulty accepting much of the work in what is variously called the rational choice, supply-side, or religious market explanation of variations in levels of religious interest, but in presentation and logic much of it is a model of social science. The papers written by Rodney Stark, Roger Finke, and Laurence Iannacone are exemplary in their clarity.26 Their reasoning is so very clearly explained that it is easy either to build on or to criticize their research. Provided it is clearly elaborated, even misguided ‘close to the coalface’ reasoning advances science because it allows us to identify dead ends.

Conclusion It is understandable that the most popular social theory is generally also the most abstract and least well grounded. The grander the assertion, the greater its apparent application and hence the greater the number of scholars who can find Page 14 of 17

 

Social Theory and Religion some reason to cite it. Normative, zeitgeist, and agenda-setting theory also have the appeal of letting scholars declare their political posture and their virtue. Foucault’s notion of ‘governmentality’, for example, not only gives us a sweeping characterization of the modern world but also allows us to show that we are against oppressive government. The drawback is that there is a corresponding lack of specific value in improving our understanding of the social phenomena we research. For all that we can find inspiration in all sorts of social theory, the sort of theorizing likely to be of the greatest value in the study of religion as an empirical phenomenon is what Robert Merton called theories of the middle range, what Barney Glaser and Anselm Strauss called ‘grounded theory’, and what I simply call ‘sociological explanation’.27 Notes Notes:

(1.) For an example of such personal reactions to ideas, see N. Hannan, ‘A Book that Changed Me’, Guardian, 28 August 2013. (2.) For example, Edward Said, whose claims about orientalism have been extremely popular, was a student of English literature and art; see E. Said, Orientalism (Harmondsworth: Penguin, 1977). Judith Butler, whose notions of performativity have been similarly popular, is a philosopher and literary critic; see J. Butler, Excitable Speech: A Politics of the Performative (London and New York: Routledge, 1997). (3.) W. G. Runciman, Sociology in its Place (Cambridge: Cambridge University Press, 1970), 12. (4.) E. Durkheim, Suicide (London: George Allen and Unwin, 1970); J. H. Goldthorpe, Social Mobility and Class Structure in Modern Britain (Oxford: Clarendon Press, 1987); H. S. Becker, ‘Becoming a Marijuana User’, American Journal of Sociology, 59 (1953), 235–42; H. M. Collins, Changing Order: Replication and Induction in Scientific Practice (London and Beverly Hills: Sage Publications, 1985). (5.) B. S. Turner, ‘The Short History of Human Rights’, Contemporary Sociology, 40 (2011), 678. (6.) For decades Sutherland was one of the most prominent American criminologists; see E. H. Sutherland, Principles of Criminology (Chicago: University of Chicago Press, 1924); White Collar Crime (New York: Holt, Rinehart & Winston, 1949); ‘The Diffusion of Sexual Psychopath Laws’, American Journal of Sociology, 56 (1950), 142–63. (7.) F. Mount, ‘No Theatricks’, London Review of Books, 21 August 2014, 16.

Page 15 of 17

 

Social Theory and Religion (8.) On Ibn Khaldun, see Aziz Al-Azmeh, Ibn Khaldun: A Reinterpretation (London: Routledge, 1991). On the Christian commitments of early US sociologists, see C. D. Bryant and D. L. Peck, 21st Century Sociology: A Reference Handbook, I (Los Angeles: Sage Publications, 2007), 362. (9.) Durkheim, Suicide; F. Toennies, Community and Civil Society (Cambridge: Cambridge University Press, 2001). (10.) L. Dawson, ‘Privatization, Globalization and Religious Innovation: Giddens’ Theory of Modernity and the Refutation of Secularization’, in J. A. Beckford and J. Walliss (eds), Theorising Religion: Classical and Contemporary Debates (London: Ashgate, 2006), 114. (11.) R. Wallis and S. Bruce, ‘Religion: The British Contribution’, British Journal of Sociology, 40 (1989), 493–520. (12.) B. S. Turner, Religion and Social Theory (London: Heinemann, 1983). (13.) Turner, Religion and Social Theory, 3. While I am a great admirer of Turner’s work, especially on the sociology of the body, I think it entirely to the credit of sociologists of religion that they did not get involved in what, as this quotation makes clear, was essentially a quixotic quest to save Marxism from the failure of its predictions. (14.) J. Beckford, Religion and Advanced Industrial Society (London: Unwin Hyman, 1989), 13. (15.) M. Bevir, ‘Rethinking Governmentality: Towards Genealogies of Governance’, European Journal of Social Theory, 13 (2010), 425. (16.) H. Garfinkel, Studies in Ethnomethodology (Englewood Cliffs, NJ: Prentice Hall, 1967); J. Heritage, Garfinkel and Ethnomethodology (Cambridge: Polity Press, 1984). (17.) Turner, Religion and Social Theory, 8. (18.) M. J. Neitz, ‘Afterword’, in K. Aune, S. Sharma and G. Vincent (eds), Women and Religion in the West: Challenging Secularization (Aldershot: Ashgate, 2008), 222. See also L. Woodhead, ‘Gendering Secularization Theory’, Kvinder, Kãn og Forskning, 1 (2005), 1–25; (19.) S. Bruce, Secularization (Oxford: Oxford University Press, 2011). (20.) I describe this as tangential because whether some theory advances knowledge depends on the internal coherence of that theory, its consistency with previous work, and its fit with the evidence. If we accept the de facto positivist approach outlined in Chapter 2, why someone advances a particular theory is Page 16 of 17

 

Social Theory and Religion neither here nor there. Those who prefer to watch the pitcher rather than the ball (to continue with the metaphor introduced in Chapter 5) will prefer Neitz’s method of analysing the biographies of analysts to identify the interests that ensure they are wrong. (21.) E. Katz and P. Lazarsfeld, Personal Influence (New York: Free Press, 1955); E. M. Rogers, Diffusion of Innovations (New York: Free Press of Glencoe, 1962). For an account of the background to Katz and Lazarsfeld’s research, see J. Pooley, ‘Fifteen Pages that Shook the Field: Personal Influence, Edward Shils and the Remembered History of Mass Communication Research’, Annals of the American Academy of Political and Social Science, 608 (2006), 1–27. (22.) P. F. Lazarsfeld, B. Berelson, and H. Gaudet, The People’s Choice (New York: Columbia University Press, 1948). (23.) L. P. Gerlach and V. Hine, People, Power, Change: Movements of Social Transformation (Indianapolis: Bobbs-Merrill, 1970). (24.) Katz and Lazarsfeld, Personal Influence; J. Coleman, E. Katz, and H. Manzel, ‘The Diffusion of Innovation among Physicians’, Sociometry, 29 (1957), 253–70; H. Manzel and E. Katz, ‘Social Relations and Innovation in the Medical Profession’, Public Opinion Quarterly, 19 (1955–6), 337–53. (25.) It is not central to this discussion, but one may wonder if the familiarity of twenty-first-century young people with digital communication and the online world is changing how plausible and influential is direct mass-media communication. The claim is now routinely made that Islamic terrorists were radicalized by the internet. It often turns out that such people actually had personal connections that might have been more influential, but nonetheless it is interesting to consider whether increased familiarity has made mass media more persuasive. (26.) For extensive bibliographical details, see the notes to Chapter 9 and S. Bruce, Choice and Religion (Oxford: Oxford University Press, 1999). (27.) R. K. Merton, Social Theory and Social Structure Glencoe, IL: Free Press, 1957); B. G. Glaser and A. L. Strauss, The Discovery of Grounded Theory (London: Routledge and Kegan Paul, 1967).

Page 17 of 17

 

Action Rational and Irrational

Researching Religion: Why We Need Social Science Steve Bruce

Print publication date: 2018 Print ISBN-13: 9780198786580 Published to Oxford Scholarship Online: August 2018 DOI: 10.1093/oso/9780198786580.001.0001

Action Rational and Irrational Steve Bruce

DOI:10.1093/oso/9780198786580.003.0009

Abstract and Keywords The model of human action as resting on rational choices between alternative opportunities for maximizing our utility has been borrowed from liberal economics by social scientists keen to refute the conventional explanation of secularization. This chapter considers whether we can treat religion as a commodity that people buy more or less of according to individual rational choice. It argues that religion differs from soap powder both extrinsically (because we cannot readily compare its costs or benefits) and in terms of its social roots: in most societies switching religion threatens social bonds far more than does changing car brands. Finally, it advances an important general principle: that we should be very reluctant to impute to other people motives that we would not impute to ourselves. Keywords:   religion, utility-maximizing, economic rationality, consumerism, imputing motives

Introduction Chapter 8 challenged the idea that religious conversion is usually a product of an uncommon credulity caused by an unconscious or subconscious reaction to biological, psychological, or social stresses. My point there was that, however strange it seems to us, most conversion is generally rational from the point of view of the actor. In this chapter I want to consider an approach to religious beliefs and behaviour that arguably treats such things too rationally. To be more precise, I will consider the limits, for understanding religion, of the narrow ‘maximizing-utility’ view of rationality that underpins the rational choice perspective borrowed by some sociologists of religion from economics.

Page 1 of 18

 

Action Rational and Irrational So that the focus of my criticisms is clear, I will acknowledge that there are aspects of religion that do properly fall within the remit of economics. For example, social scientists have long been interested in the economic consequences of religious cultures. When Max Weber explored the modernizing effects of the Protestant Reformation in his classic Protestant Ethic and the Spirit of Capitalism, he was arguing with Werner Sombart’s claims for the effects of a Jewish ethic.1 More recently, Peter Berger followed a number of seminal works in the sociology of religion by promoting research on the links between religion and economic development.2 We can also observe that religious organizations have mundane economic interests and problems that shape their activities. For example, the merger of the three main strands of British Methodism in 1932 meant that many chapels were redundant. In the Cleveland area of the north-east of England, the Loftus Wesleyan Methodist circuit and the Staithes Primitive Methodist circuit overlapped so that the villages of Brotton, Skinningrove, Brotton, Loftus, Hinderwell, and (p.194) Staithes each had Wesleyan and Primitive chapels— and some also had Bible Christian chapels. In Skinningrove, two chapels were back-to-back—so close that one congregation could have sung its hymns to the organ being played in the other chapel. A fit man could have thrown a Bible from the Staithes Primitive chapel to its Wesleyan rival. By the time of the merger, Methodism had so declined from its 1870s highpoint that none of the chapels was more than a third full, and the officials charged with managing the newly formed Loftus and Staithes Methodist circuit could see obvious financial advantages in rationalizing provision: a classic economic concern. As an aside that bears on the value of imputing economic rationality to people in general, it is worth noting that chapel members often resisted rationalization. Because they were sentimentally attached to the ‘chapel from which I want to be buried’ or ‘the chapel my grandparents built’, they maintained their chapels until the membership literally died out, the building fell apart, or the dwindling membership could no longer provide enough lay preachers to conduct the required large number of very small services. Furthermore, mundane economic concerns sometimes influence the extent and the manner in which people express their religious commitments. People too often make considerable sacrifices for their faiths for us to accept any general proposition that churchgoers seek to minimize their ‘faith costs’ as they do their grocery bills, but we can sometimes trace connections between the vitality of a local economy and its religious life. Methodist chapels in the villages to the west of Durham city were built by and for coalminers in the last quarter of the nineteenth century. In good times, when the mines were working at full capacity and coal expensive, new chapels were built, existing ones were expanded and refurbished, and attendance and membership grew. As the rather sad notes attached to the quarterly reports of membership by chapel stewards showed,

Page 2 of 18

 

Action Rational and Irrational when the mines went on short-working or closed, chapel membership declined, as did attendance and financial support.3

Rational Choice in Religion However, recent sociological borrowings from economics have been more farreaching and contentious than these mundane observations. Since the early 1990s, a number of US scholars have tried radically to (p.195) reshape our understanding of religion by applying economistic principles to such core problems as the explanation of belief, conversion, membership, and commitment.4 Before I begin to explain my reluctance to be impressed by this body of work, I should repeat a point made in Chapter 8. It is possible to criticize the rational choice approach to religion, not only because it is wrong-headed, but also because, unlike much social theory, it makes its assumptions and propositions extremely clear. The plethora of hypotheses and theorems may at first sight seem daunting (especially when key words are converted into symbols), but, unlike the social theory that elides, blurs, obfuscates, and takes for granted, they permit critical examination. In that sense rational choice is good theory even though it is wrong theory. Inspired by Nobel-prize-winning economist Gary Becker’s view that the principles of economics can be fruitfully applied to all forms of social behaviour, Rodney Stark, Roger Finke, and Laurence Iannaccone and their students have produced a large body of work that seeks to explain both large-scale social patterns (such as the differences between societies in levels of secularization) and micro-decision-making (such as shifting between denominations) by considerations of market structure and cost–benefit calculation. Taking as his paradigm the supposed virtues of the free market over the regulated economy in meeting and stimulating needs for such consumer goods as cars, Stark argues that differences in religious vitality (usually measured by church membership or church attendance) can be explained by structural features of the religious economy. A free and competitive market creates a vibrant religious culture. Religious monopolies or hegemonies (especially if they result from state support for a particular religion) suppress innate demand and depress the religious culture.5 Competition in a free market is virtuous because it allows easy entry for new ‘products’, and, the greater the diversity of religions on offer, the greater the chances that everyone will find something that suits him or her. Competition also forces every provider to improve or lose adherents. Lack of state funding means that the clergy eat only if they attract and keep a congregation, which means they will work hard to be popular. And the absence of any connection with the state means that political dissent does not undermine the legitimacy of religion, as it did, for example, in France during the Revolution, where the Catholic Church’s support for the ancien regime caused many people to reject it and, because it was the only church on offer, to reject religion itself.

Page 3 of 18

 

Action Rational and Irrational (p.196) This rational choice or supply-side explanation of religious vitality (socalled because it assumes that the underlying demand for religion is stable and that defects in supply explain apparent secularization) has been subject to considerable testing. Finke, Stark, and Iannaccone find much evidence to support it, mostly in contemporaneous correlations of levels of religiosity in parts of the USA with apparently differing degrees of religious diversity. As I have shown at length elsewhere, attempts by others to replicate such research have usually failed.6 More fundamentally, David Voas, Dan Olson, and Alasdair Crockett have demonstrated that much of the apparent positive connection between religious diversity and vitality is actually an artefact of the statistical techniques used to demonstrate it.7 There is no doubt that, in particular times and places, competition between churches can lead to an improvement of provision and hence to an increase in opportunities for church involvement. In the early nineteenth century the Church of Scotland’s provision in the Orkney and Shetland islands was often woeful. Parish ministers were required to serve three or four churches separated by miles of boggy land unadorned by any form of road; some had to serve more than one island. In some cases the creation of new parishes was the result of popular local agitation, but what often prompted an improvement was competition from such dissenters as the Baptists, Methodists, and Seceder Presbyterians. We can also find examples of competition leading to increased commitment. The 1843 Disruption, in which a third of the Church of Scotland ministers and members left to form the Free Church, followed a decade of increasingly bitter rivalry between moderates and evangelicals that forced Scots to question their religious affiliations. That rivalry probably had no immediate effect in boosting the number of Christians, but people who had previously attended their parish churches regularly but never applied for membership now enlisted. That they had to choose sides may well have increased commitment and caused families to work harder to ensure that their children became wellinformed Free Church or Church members. But though we can often see short-term and local gains from competition, the overall and long-term effect seems deleterious. Increasing religious diversity forced the state gradually to withdraw from the active support of any particular church, and education, social control, and social welfare gradually become secular. At the level of social psychology, once the initial stimulus of intense rivalry had worn off, diversity had the effect of weakening religious dogmatism and (p.197) commitment, as positive social interaction with members of different churches caused believers to appreciate that their church did not have a monopoly of social virtue or theological rectitude. Taken at the level of nation states, the general supply-side model fails two obvious tests. Comparisons of different societies in the same period usually run the wrong way: for example, the generally more homogenous Catholic European Page 4 of 18

 

Action Rational and Irrational societies (Poland, Ireland, and Italy, for example) are more religious than the diverse Protestant ones. Comparison of the same society at different times also runs the wrong way: Britain has a much freer and much more diverse religious market in 2017 than it had in 1800, but the proportion of the population at all involved has fallen from at least 60 per cent to about 7 per cent. All modern societies have become more diverse over the twentieth century and all have become patently less religious in the process. The record of applying economic models (such as Iannaccone’s human capital approach) to the religious behaviour of individuals is little more impressive and one hint as to why that is the case may be found in the patent irrelevance for religion of some classic economics problems. One such is the free rider: how can organizations prevent people who do not make a proportionate contribution to the costs from enjoying the benefits of involvement? Football grounds, for example, have high fences to prevent those who have not paid the entrance fee from enjoying the game. But, far from discouraging free riders, most religious organizations welcome them. They would like churchgoers to fund the church, but they would rather have churchgoers who make no contribution than be less popular. In the nineteenth century, Shetland clergy gave clothes to poor parishioners who were inhibited from attending church or chapel by their lack of decent apparel. Consider the question of whether religious officials should provide religious offices for rites of passage to people who are not regular churchgoers. Most clergy do not mind such free riding. They will baptize the children of non-attenders either because they believe baptism to be effective despite parental indifference or because they see it as an opportunity to evangelize the unGodly (by, for example, insisting that parents attend a few classes on the significance of baptism before they can enjoy the rite). Clergy attitudes to marriage and burial are equally generous. I recently asked a Brethren lay preacher why he had officiated at the wedding of two people who were patently not Christian: ‘Well, it’s a chance to bring a wee word of God to folk who never hear it.’ When committed believers do rebel (p.198) against free riding, it is generally not because of any loss that readily fits the economist’s model. When ministers refuse to marry couples who have no church connection and are not willing to pretend to any, it is not because they wish to safeguard their resources for the use of proper members. It is because they see such misuse of religious offices as disrespectful; it offends the honour of their religion. Of course, with a little imagination, non-economic concerns such as pride and persuasiveness can be shoehorned into economic metaphors, but such translation aids understanding no more than does replacing English with Gaelic.

Religion and Economic First Principles Doubtless better-educated scholars came to this conclusion more quickly. It took me a few years of arguing with particular propositions advanced by Stark, Finke, and Iannaccone to appreciate that my specific disagreements shared a common ground: religion cannot be treated as a consumer durable such as a car or a Page 5 of 18

 

Action Rational and Irrational fridge because religious people do not so treat it. The problem can be readily illustrated by briefly considering how well what Becker describes as the foundations of economics can apply to religious behaviour. As an aside, I should say that, in order to keep this discussion manageable, I will not consider the very real possibility that the utility theory type of economics developed from the ideas of John von Neumann in the 1940s fails to explain even those matters that have traditionally been within its remit.8 Becker believes that ‘the economic approach is a comprehensive one that is applicable to all human behavior, be it behavior involving money prices or imputed shadow prices, repeated or infrequent decisions, large or minor decisions, emotional or mechanical ends’.9 Reasons for doubting the value of such imperialism can best be explained by briefly working through his succinct summary of the assumptions underlying economists’ approach to human behaviour. Economists assume first that people engage in maximizing or economizing behaviour: if we can buy an identical product in two shops at different prices, we will buy the cheaper. Second, there are ‘markets that with varying degrees of efficiency co-ordinate the actions of different participants … so that their behavior becomes mutually (p.199) consistent’. Third, ‘prices and other market instruments allocate the scarce resources within a society’.10 Most Western nation states now have relatively free religious markets, though, as we will see, most people do not use those markets in the way they use the market for used cars or woollen clothes. The role of such ‘market instruments’ as price in allocating scarce resources is highly contentious. Again more will be said about this but we can start by noting that, unlike those for most mass products, the potential consumers of religion do not buy more if the price goes down. The disagreements over the reasons why conservative forms of Protestantism generally survived the twentieth century better than the liberal varieties was mentioned in Chapter 2. To avoid elaborating on those debates, I will just minimally note that the least demanding forms of Christianity are not the most popular—which, until one starts adding qualifications about niche markets, snobbery, and exclusivity (and such qualifications often undermine, rather than qualify, them), basic resource allocation principles seem irrelevant. To begin with Becker’s first principle: in order to maximize utility we must be able to compare alternatives. In choosing between brands of soap powder, we can compare the weight in the box or the strength of the powder and decide which offers the best return. We can delegate such testing to fellow citizens (as when we read reviews of products on TripAdvisor or Amazon) or to a professional organization (as when we consult the Consumer Association’s product tests in Which magazine). But how can we compare the value of being a Jehovah’s Witness, a Jew, or a Jesuit? As the veracity of the core claims made by Page 6 of 18

 

Action Rational and Irrational religions can be known only after death or at the Day of Judgement—whichever comes sooner—we have no way of knowing now which, if any, is correct. That is, competing religions are literally incomparable. Some peripheral aspects of a religious faith can perhaps be tested against other religious or secular alternatives. We could compare going to church with going to concerts as a way of making friends or meeting potential spouses. We could compare the social status benefits of one denomination against another. We might even be able to test the assertion that becoming ‘clear’ with Scientology will have greater benefits for our mental health than would buying sessions of secular psychotherapy or taking antidepressants. But even here comparison is difficult. Although the Bromley and Shupe model of conversion discussed in Chapter 7 suggests that potential converts (p.200) ‘try on’ the role of believer, we must suppose that such this-worldly benefits of religious conviction as good health and contentment are enjoyed only by those who do actually believe. We cannot pretend to be saved in order to test the therapeutic value of being sure of one’s salvation. So, even with peripheral aspects of religious behavior, one of the conditions for maximizing utility is generally absent. And it is certainly absent for the core of religion. Until death we cannot be sure whether any religion is correct. Or, to put it another way, we cannot hedge our bets by buying small amounts of different religions. We can manage soap powder uncertainty by buying some of each of rival brands; we cannot diversify our religious portfolios as a hedge against choosing the wrong God. This, by the way, is why the French philosopher Blaise Pascal was wrong to argue that, in the absence of certainty about the existence of God, we should bet that he exists. Pascal assumes there is only one God and that, while there may be much to be gained from believing in him, there is little or nothing to be lost. However, if there are a number of Gods, believing in the wrong one will offend the right one. Spread betting does not work with religion. The key difference between soap powder and religion can be seen clearly if we consider the power of ideology to shift perceptions fundamentally. When we change soap powders from Cleanwell to Wellclean, we are not required to believe that Cleanwell is, in reality, not a cleaning agent but a source of pollution. Although many religious adherents can be polite about other religions, and it is sometimes the case that one religion will allow that another has an incomplete vision of the truth, most religions conventionally require that the claims of competitors be rejected. There is no God but Allah; my God is a jealous God. To an orthodox Catholic, Calvinism is not a comparable alternative: it is the work of Satan. If comparing the benefits of rival products is one necessary condition for maximizing utility, a second is the ability to compare costs. The price of soap powder is expressed and exacted in a common currency. The price in sterling of Page 7 of 18

 

Action Rational and Irrational Cleanwell and Wellclean is displayed, and, to prevent sharp practice in labelling, my supermarket also tells me the price per weight of washing powder or typical number of washes. But to talk of the ‘price’ of religious involvement is stretching a metaphor too far. Although selling enlightenment is common among purveyors of New Age or alternative spirituality and is found in some of the 1970s new religious movements (Scientology, (p.201) for example), it is anathema to the major religious traditions, which offer themselves freely to all who would believe and sometimes impose themselves even on those who would not. The pricing metaphor is not saved from shipwreck by substituting the shadow price of time for money. How ‘costly’ time spent on some activity is depends on the extent to which we find that passage of time rewarding. To some potential members of a Pentecostal church, a two-hour prayer meeting is an enjoyable experience: to me it is a cause of back and knee ache. In brief, the economistic or rational choice model of human behaviour requires that we be able to assess costs and returns from some neutral or consensually agreed standpoint before we make a commitment to one religion rather than another. But the nature of religion does not allow such comparisons. Rational choice models of behaviour depend on us knowing what is the rational choice. When faced with the possibility of buying the same breakfast cereal from two outlets, it would be irrational not to compare the prices and buy the cheaper. But, as Jon Elster puts it: ‘To the extent that we cannot tell, or cannot tell uniquely, what the rational choice would be, the theory fails … In a word rational-choice theory can fail because it does not tell us what rationality requires.’11 There is a second group of obstacles to maximizing utility that concern our freedom to choose. The economistic version of rational choice assumes that, as when we stand on a used car lot and swither between a Ford Mondeo and a Vauxhall Insignia, there are no constraints on choice that are not part of the maximizing calculation. In contrast, maximizing in religion is almost always heavily constrained by culture (and its embodiment in social relationships). Religious affiliation is so often closely associated with other forms of social identity that in many societies people cannot switch without considerable loss. In Kosovo, Bosnia, or Northern Ireland, religious affiliation is not a personal preference to be altered at will. It is a matter of communal identity, and those who change religion may well find themselves ostracized or worse. People are not shunned, expelled, or murdered for changing car brands; in many countries switching religion carries a very real danger of all three fates. Even in the USA—conventionally regarded as both the most religious diverse and the most religious of modern societies—religious choice is constrained by powerful social identities. One technical fault of Stark, Finke and Iannaccone’s claimed demonstrations of a positive correlation between religious diversity and religious commitment is (p.202) that they exaggerate diversity by counting as Page 8 of 18

 

Action Rational and Irrational alternatives black and white versions of sects and denominations. In a world that is still racially divided, this is a mistake. Certainly in the nineteenth and twentieth centuries few white Americans would have attended a black church and vice versa. They make a related mistake in treating ethnic and linguistic variants of a single tradition—Swedish and German Lutheran—as alternatives. Even in religiously tolerant societies, radical alternation of religion is likely to threaten social relationships. Still today in Scotland, where little of public consequence hangs on religious affiliation, leaving the Free Church for the Free Presbyterian Church may cause family rifts. For a Protestant to become a Muslim will require that most of his or her social world be restructured. In the UK there has been a considerable growth in Pentecostal churches that recruit almost exclusively from black (primarily African) immigrant populations. Although the UK is not as racially polarized as the USA, the failure of those churches to recruit ‘white British’ (a census category as well as a description) Christians suggests that cultural inhibitions remain strong. The point seems so obvious one feels almost embarrassed to spell it out, but, as it is central to the religious market model, we should do so: in most societies religious affiliation is not just an individual preference. The same point applies even more strongly to the producers of religious commodities: churches, sects, and denominations are constrained in a manner and to an extent quite unlike anything that bears on the manufacturers of goods or the providers of secular services. A distributor of Massey-Ferguson tractors who finds sales drying up can switch to selling John Deere tractors. If the agricultural sector collapses, he can get out of tractors altogether, sell his land for housing, and invest in yen or zlotys. Because religions conventionally claim to be divinely inspired and to be changeless, their providers are very slow to change. The Catholic Church in Latin America may envy the growth of Pentecostalism, but it cannot readily alter practices or beliefs in pursuit of the market. One realm in which we see very clearly the difference between religious ideologies and consumer products is in the way producers respond to failure. Small sects—and this is also common in extremist political parties—may reconcile themselves to their unpopularity, and even revel in it, with some version of the assertion that being small just proves that they are right all along about the power of Satan (or, for left-wing parties, the strength of false consciousness). When a (p.203) clothing chain’s sales fall, the directors do not say ‘Well, that just proves our clothes are too good for ordinary people’ and carry on regardless. They hire new design teams, change their pricing policies, and do their best to win back market share from their competitors.

Page 9 of 18

 

Action Rational and Irrational Implausible Models of Behaviour Having explained why religion is ill-suited to the economist’s model of people seeking to maximize utility, I want now to look closely at a few of the specific propositions advanced in the rational choice treatment of religious behaviour in order to illustrate a useful technique in social explanation. Social scientists have traditionally been divided over the extent to which, and the manner in which, we can treat abstractions such as community and society as existing independently of the individuals who at any one time constitute them. Emile Durkheim famously argued that sociology should be concerned, not with explaining individual actions, but with the causal explanation of features of societies, and he demonstrated his case with a challenging example. Precisely because there can be few actions that are more apparently ‘individual’ than suicide, he chose to demonstrate that the suicide rates of societies could be compared and explained by other features of those societies (primarily variations on the theme of social cohesion) apparently without resorting to individual psychology.12 In one trivial sense, Durkheim’s work does depend on individuals: the suicide rate is nothing other than a summation of a number of individual suicides. But it also depends on individuals in the more important sense that, although his explanations can be presented in the language of societal characteristics— anomie is positively correlated with suicide rates—they are always unpacked to the level of individual actions. So, when he tries to explain why the suicide rate of widows is lower than that of widowers, he spells out a theory about formerly married women being more content than formerly married men with hobbies and the company of pets. Whether this is insight or sexist nonsense, it is now a story about an imagined widow and an imagined widower. Rather than see this as Durkheim being inconsistent, we are better to see it as a skilled social researcher doing the obvious: trying to make sense of a pattern in statistics by imaging (p.204) the circumstances, beliefs, attitudes, motives, and feelings of the individual people whose actions form those patterns. We can state this is a useful principle for assessing any abstract explanation. A regular correlation between indices is a useful place to start, but, if an explanation is to be sustained, it must rest on a logic that remains plausible when we reduce it to the actions of imagined people. The point can be made with some examples. One of the ways in which increasing competition makes a society more religious, according to Stark, is by incentivizing the clergy. The clergy of a religious monopoly apparently have little reason to work hard to attract and sustain a congregation. This is especially the case if they are funded by some form of public taxation. Such clergy not only have no particular reason to attract a congregation but will prefer to be unpopular because they will still get their Page 10 of 18

 

Action Rational and Irrational stipends but do less work for them. Stark illustrates this argument by comparing Lutheran clergy in the competitive market of the USA (where they had to attract and keep a flock that would pay them) and in Germany (where they were paid from a tax levied on all who did not register their preference for another church): ‘the German clergy are better off with empty churches, which place little demand on their time, than with full ones’.13 That is exactly what we would extrapolate from a utility-maximizing view of human motivation: getting your salary without those annoyingly demanding congregants is better than being popular. But, if we apply this to Stark himself, we see the problem. Stark did not deliberately teach badly so that his courses would be unpopular and he would have less work for the same salary. He researched a great deal and published an extraordinary number of journal articles: work for which he was not paid. We can perhaps provide an economic explanation for such diligence in the early part of his career: although unpaid, journal publication helps build the reputation that produces promotion and pay increases. But why did he continue to do this after he had reached the peak of the profession? Stark lectured well because he took pride in his performance and wished to be admired. He published a great deal because he believed his work was important and wanted to bring it to the attention of his colleagues. He was driven by the conviction that he was right and that he had an obligation to persuade others. Although my output cannot rival Stark’s, I recognize in myself the qualities I impute to him: a desire to be admired and a strong commitment to our discipline. So why does he deny such qualities to German Lutheran ministers? After all, they add a third (p.205) motive: the belief that they are God’s agents. One possible justification is to argue that, as pride in a job well done is a constant for all clergy, it can be ignored, while we examine only the variable of the additional financial incentives that bear on clergy in a free market. But this will save Stark’s proposition only if the pride-in-the-job constant is presumed to explain only a small part of the commitment that clergy bring to their work, and I see no reason for supposing that. Studies of actual clergymen tell us the obvious: irrespective of the security of their income, almost all of them want to be popular because, as well as reflecting on their competence, popularity reflects on their performance of a divine commission. One very detailed study of rural Church of England clergy showed that, although they put a brave face on it, many were troubled by very small congregations: ‘If it gets very small—say five—I am disappointed,’ and ‘It’s annoying two or three Sundays a year if no-one comes’. Apart from the psychological problems of being confronted by indifference to what one holds dear—as voiced by the man who said ‘It hits you when nobody turns up’—small congregations inhibit ritual performance: ‘16 people without a choir trying to do Anglican chanting is a pretty dismal experience.’14

Page 11 of 18

 

Action Rational and Irrational Iannaccone has tried to cast new light on various aspects of religious behaviour by viewing religion as a form of human capital. For example, that most people who do switch churches move to something quite similar is explained by their wish to get the benefits of time and effort invested in learning the beliefs and behaviour of their first church, sect, or denomination. The same ‘return on investment’ principle explains why those who return to religion after a period away are more likely to join the sort of church in which they were raised than to join something very different. Apparently it also explains why religious conversion occurs early in life: old people do not convert because they know they do not have enough life left to get a decent return on their investment in the new faith! An obvious flaw with that explanation is that most 50-year-old men do not know if they have four or forty years left. The deeper problem with these explanations is that the same patterns can be explained (more plausibly, I would argue) by the fact that existing beliefs, even when we are considering changing them, affect the plausibility of alternatives. A few of us are attracted by the exotic, but most Presbyterian Christians will find Methodism more attractive than Scientology, not because they are reluctant to invest time and effort in learning something (p.206) so very different, but because they find something so very different implausible. One finding that Iannaccone claims to explain by the desire to conserve human capital is that families in which husband and wife belong to the same denomination go to church more often than mixed-denomination families. A household can produce religious commodities more efficiently when both husband and wife share the same religion. Single-faith households benefit from ‘economies of scale’: the same car drives everyone to church; there is no question as to how time and money contributions will be allocated to different religions; it is not necessary to debate the religion in which one’s children will be reared.15 An initial difficulty with his explanation is that it assumes the causal connection runs one way: that ‘spousal homology’, as it is known in the literature, causes greater religiosity. It is as or more likely that the connection runs in the other direction. Young people who are seriously committed to their faith will seek marriage partners from within their own congregation and denomination (and if they do not, their parents are likely to remind them!): that is, greater religiosity causes spousal homology. Or, to put the same logic the other way round, people who are not very religious are much less likely than religious people to take religion into account when responding to a possible marriage partner. But, once our ill-sorted couple has married, is it really likely that their subsequent church attendance will be determined by such increases in costs as needing two cars to attend church? I can imagine that disagreement about which church the children should attend reduces churchgoing. I can imagine that living with a decent, caring, and virtuous individual of a different faith erodes confidence in Page 12 of 18

 

Action Rational and Irrational the rectitude of one’s own faith. But, in a country where gasoline is cheap, I cannot imagine that extra fuel costs—after all, how far away are those churches? —is a consideration. To summarize, with Durkheim’s easily amused widows, Stark’s cynical Lutheran pastors, and Iannaccone’s penny-pinching hoarders of human capital, we have statistical patterns being explained by propositions that seem superficially sensible until we try to tell them as stories about people like us. I am not suggesting that explanations must always be rejected if we cannot imagine ourselves acting in that way. Even if I am right in thinking of myself as a typical Scot of my class, gender, age, and cultural background and can thus impute (p. 207) my reasoning to a wide range of others, there will be people out there who are unlike me. Iannaccone might be able to produce convincing evidence that, for the class of people of middling religious commitment, increases in marginal costs do indeed discourage churchgoing. But, until such data are presented, that the explanation just does not sound convincing is sufficient reason to be sceptical of it. To put it most generally, and to return to the point made by Marshall Sahlins in Chapter 2, the social sciences can compensate somewhat for the limits placed on them by their inability to create the experiments common to the natural sciences by making use of the researcher’s similarity to his or her subjects. In abstract arguments between those who see our primary purpose as understanding and those who think we should pursue causal explanation, those two tasks are set against each other as competing alternatives. I would argue that they are both necessary to the social sciences. Understanding without causal explanation is just description; explanation cannot even begin without some considerable understanding. And understanding offers a useful test of causal explanations. For convenience we often write about such abstractions as religious diversity, the secular state, rates of churchgoing, and the like, but, if any of the connections we draw between such abstractions is to be sustained, it has to remain plausible when we unpack it to the level of the imagined individual actor. Any explanation that imputes to others motives that we would not readily impute to ourselves should be regarded with considerable suspicion.

Secularization and Maximizing The supply-side explanation of religious vitality was promoted as an alternative to the secularization approach with its stress on the decline of demand for religion. If we accept the argument that religion is such a profoundly social cultural product that it is not well explained by the principles of economics, we might conclude that the theories of Stark, Finke, and Iannaccone will become more persuasive as societies become more secular. The model of the utilitymaximizing individual works best for something like the motor industry: where there is widespread and general demand for a product; where costs and benefits can be assessed from a neutral or consensual standpoint (p.208) that does not Page 13 of 18

 

Action Rational and Irrational change fundamentally at the point of choice; and where the matter in hand is a personal choice that is little constrained by enduring and powerful social identities. It is worth remembering that the secularization thesis, as it has been developed by scholars such as David Martin, Bryan Wilson, Karol Dobbelaere, Thomas Luckmann, and Peter Berger from the seminal work of Karl Marx, Max Weber, and Emile Durkheim, is not solely or even initially concerned with the decline of religion. As both cause and symptom of decline, we see the following changes: the supernatural diminishes as core beliefs are rationalized and psychologized; this-worldly rewards become more important than the afterlife; enjoying the therapeutic benefits of religion displaces the glorification of God as the primary purpose; external authority is either replaced by the choosing individual (as in the New Age) or becomes severely attenuated (as in liberal Christianity); major religions cease to claim sole access to the will of God and become increasingly ecumenical; states give up enforcing religious conformity; religion becomes free from other major social identities; religious requirements bear on ever smaller areas of life (as spheres such as the economy and the polity are released from the control of religion) and on ever fewer people (so that others are not required to follow the religious laws that bind the believer); people distinguish between the morality that God requires and the law that the state can be expected to enforce; an ever-greater variety of religions are tolerated; religion becomes a matter for the private life of the individual; and, as religion shrinks in importance, people become increasingly indifferent to the particularities of religion while generally remaining vaguely approving of religion in the abstract. One way of encapsulating those changes is to say that in a largely secular society—and only in such a society—religion can be treated as a consumer commodity.

Maximizing and the Cultic Milieu If we are looking for a sphere of religious activity that does resemble the free market of diverse providers competing for the attention of the autonomous choosing individual, we can find it in the world of New Age spirituality. The cultic milieu of alternative spirituality is a marketplace of competing providers. Its wares are sampled by (p.209) informed consumers who try Bachian flower remedies, aromatherapy, and feng sui, take a yoga course, attend workshops on Tibetan overtone chanting, fire-walking, or Bhagwan Rajneesh’s dynamic meditation, and experiment with Hopi ear candles and Sufi dancing. From a now-truly-global cafeteria of spiritual practices, beliefs, and rituals, Westerners can, with almost no fear of being stigmatized as dangerous deviants, construct their own idiosyncratic packages. And because they buy such services and products, they remain in control of their involvement. But what makes the cultic milieu possible, and underpins it, is the novel epistemological principle that the believer is the final arbiter of what is true ‘for her’ (and it usually is a woman); or, as they say in trade: the consumer is always right. This is, of course, a very Page 14 of 18

 

Action Rational and Irrational long way from the epistemological foundations of the major world religions, which assume that there is a God or Gods whose will is knowable through clear but limited channels controlled by properly sensitized, trained, or appointed experts.16 One way of summarizing the secularization thesis is to note that the complex pattern of social changes we gloss as ‘modernization’ has been accompanied by an increase in individual liberty. For some, that liberty has led to religion being abandoned altogether; for others, it has led to a selective attitude to the precepts of once-hegemonic faiths. That has been visible in the liberal wings of the major Christian churches since the late nineteenth century, and since the 1960s even conservative churches in the West have had to compromise much of their behavioural distinctiveness—Puritanism is now rare, even among fundamentalists—and become ever less dogmatic and authoritarian. While the consumerism of the New Age milieu may be extreme, it is part of a general trend in the religious culture of the Western world.

Conclusion An unfortunate consequence of recent developments in the language of the social sciences is that the two words in the phrase ‘rational choice’ have become so strongly paired that the economistic model of rationality—with the wellinformed, socially detached, and unemotional consumer maximizing utility by choosing between alternatives on the basis of calculable costs and rewards—has caused us to lose sight of other types of rationality. That which is not maximizing (p.210) utility is not necessarily irrational.17 In his earliest work Talcott Parsons rejected the narrow economistic view of the importance of instrumental rationality.18 He did not think it could explain social solidarity based on shared values. To appreciate what he meant, we can go back to Max Weber’s delineation of four types of rationality—affective, traditional, value rational, and means–end—which, he believed, could be found in differing amounts in different societies. Affective rationality is determined by the actor’s feelings or emotions. Traditional rationality is determined by conventions and habits: it is entirely rational to do what everybody else does, especially where there are no sure grounds for testing options, or to do what one has always done, until circumstances change so as to make previous practice obviously counterproductive. The third type, value-rational (Wertrational), refers to acting in accordance with beliefs and values (aesthetic, ethical, aesthetic, or religious, for example). What economists, and the sociologists influenced by economics, have done is to elevate Weber’s fourth type—the purposive, instrumental, or means– end rational (Zweckrational)—into the key principle behind human behaviour and neglected the other three types.19

Page 15 of 18

 

Action Rational and Irrational In this chapter I have intertwined a methodological principle and a point of substance. The methodological principle, which we can apply universally, is to be suspicious of any explanation that, when unpacked to the level of the imagined actor, seems implausible. Such explanations may, of course, be correct, but they will require a great deal of evidential support to be persuasive. To return to a theme of Chapter 7, if, after making due allowance for differences in background and environment, we cannot imagine ourselves being brainwashed by hell-fire preaching or by the solicitous attentions of young Moonies, why suppose that anyone else was? The point of substance concerns the unsuitability of religion for the rational choice treatment. Every one of what Becker, the Nobel-prize-winning economist who championed the imperial expansion of his subject, regards as basic requirements for an economistic analysis is missing from religion: for believers, religions are incomparable and, even if we did wish to maximize our utility by choosing or changing religion, we would be hamstrung by our inability to compare costs and rewards. To repeat Elster’s caution: when we cannot know what is the instrumentally rational choice, then rational choice, in the very narrow sense in which Stark and colleagues use the term, is not possible, and hence it cannot be used to explain religious behaviour (p.211) and belief. Furthermore, in most societies religion is too strongly associated with other social identities to be a matter of personal choice. Grasping that allows us to see a considerable irony in the rational choice theory of religion. It is only in largely secular societies that the decoupling of religion from other important realms of the lifeworld allows it to be treated as a field of utility-maximizing opportunity. When religion no longer much matters, then, and only then, does instrumental rationality become more important than value rationality and tradition. Notes Notes:

(1.) M. Weber, The Protestant Ethic and the Spirit of Capitalism (London: George Allen and Unwin, 1976); W. Sombart, Economic Life in the Modern Age (New Brunswick, NJ: Transaction Publishers, 2001); R. H. Tawney, Religion and The Rise of Capitalism (New Brunswick, NJ: Transaction Publishers, 1998). (2.) P. L. Berger, The Social Reality of Religion (London: Faber, 1969), and The Capitalist Spirit: Toward a Religious Ethic of Wealth Creation (San Francisco: Institute of Contemporary Studies, 1990); P. L. Berger and R. W. Hefner, ‘Spiritual Capital in Comparative Perspective’, (accessed January 2006). (3.) S. Bruce, ‘Methodism and Miners in County Durham 1881–1991’, Northern History, 48 (2011), 337–55; R. Moore, Pitmen, Preachers and Politics: The Page 16 of 18

 

Action Rational and Irrational Effects of Methodism in a Durham Mining Community (Cambridge: Cambridge University Press, 1974). (4.) For summaries, see T. G. Jelen (ed.), Sacred Markets, Sacred Canopies: Essays on Religious Markets and Religious Pluralism (Lanham, MD: Rowan and Littlefield, 2007), and L. A. Young, Rational Choice Theory and Religion (London: Routledge, 1997). (5.) R. Finke and L. A. Iannaccone, ‘Supply-Side Explanations for Religious Change’, Annals of the American Academy of Political and Social Science, 527 (1993), 27–39; R. Finke, ‘An Unsecular America’, in S. Bruce (ed.), Religion and Modernization: Sociologists and Historians Debate the Secularization Thesis (Oxford: Oxford University Press, 1992), 145–69; R. Finke and R. Stark, The Churching of America, 1776–1990: Winners and Losers in our Religious Economy (New Brunswick, NJ: Rutgers University Press, 1992); L. A. Iannaccone, ‘The Consequences of Religious Market Structure’, Rationality and Society, 3 (1991), 156–77. (6.) S. Bruce, Choice and Religion: Critique of Rational Choice (Oxford: Oxford University Press, 1999). (7.) D. Voas, D. V. A. Olson, and A. Crockett, ‘Religious Pluralism and Participation: Why Previous Research is Wrong’, American Sociological Review, 67 (2002), 212–30. (8.) For alternatives, see D. Kahneman, Thinking Fast and Slow (London: Allen Lane, 2011), and R. H. Thaler, Misbehaving: The Making of Behavioural Economics (London: Allen Lane, 2015). (9.) G. Becker, ‘The Economic Approach to Human Behavior’, in J. Elster (ed.), Rational Choice (Oxford: Basil Blackwell, 1986), 112. (10.) Becker, ‘Economic Approach’, 112. (11.) J. Elster, ‘Introduction’, in Elster, Rational Choice, 17. (12.) E. Durkheim, Suicide (London: George Allen and Unwin, 1915). (13.) R. Stark, ‘German and German–American religion: Approximating a Crucial Experiment’, Journal for the Scientific Study of Religion, 36 (1997), 182–93. (14.) D. Davies, C. Pack, S. Seymour, C. Short, C. Watkins, and M. Winter, Parish Life and Rural Religion: Rural Church Project Vol. III (Cirencester and Nottingham: Royal Agricultural College and University of Nottingham, `1990), 38–45.

Page 17 of 18

 

Action Rational and Irrational (15.) L. Iannaccone, ‘Religious Practice: A Human Capital Approach’, Journal for the Scientific Study of Religion, 29 (1990), 297–314. (16.) In theory Protestantism does not require such expert clergy, but in practice all but the smallest Protestant churches, denominations, and sects have professional officials. (17.) For reasons given here, though it is an improvement on ‘irrational’, I reject Bryan Turner’s description of religion as ‘non-rational’; see B. S. Turner, Religion and Modern Society: Citizenship, Secularisation and the State (Cambridge: Cambridge University Press, 2011), 72. (18.) T. Parsons, The Structure of Social Action (New York: McGraw Hill, 1937). (19.) M. Weber, The Theory of Social and Economic Organization (Glencoe, IL: Free Press, 1947), 115–17. For elaborations and criticisms of Weber’s views of rationality, see the contributions to S. Whimster and S. Lash (eds), Max Weber, Rationality and Modernity (London: Allen and Unwin, 1987).

Page 18 of 18

 

Does Danger Make People Religious?

Researching Religion: Why We Need Social Science Steve Bruce

Print publication date: 2018 Print ISBN-13: 9780198786580 Published to Oxford Scholarship Online: August 2018 DOI: 10.1093/oso/9780198786580.001.0001

Does Danger Make People Religious? Steve Bruce

DOI:10.1093/oso/9780198786580.003.0010

Abstract and Keywords This chapter takes an interesting proposition—that danger makes people particularly receptive to religion—and uses the examples of the supposed piety of miners and fishermen to explore three very different sorts of social explanation. It could indeed be the case that unpredictably dangerous work disposes people to consider their mortality or to find supernatural ways of dampening anxiety. Or it could be that the unusual social structure of fishing villages and mining communities (generally isolated and introverted) insulates religious traditions from secularizing forces. Or it could be that the piety of these communities is a social myth based on the romantic assumption that those who work close to the elements should be more open to the supernatural than is the cosseted urban office worker. As well as addressing the substantive proposition, it considers practical problems of measuring piety. Keywords:   danger, piety, fishermen’s religion, Methodism, mining communities, romanticism, superstition

Introduction This chapter uses the examples of mining and fishing to explore the connection between danger and being religious or superstitious. In a longer treatment, religion and superstition would be taken separately; there are many important differences between them and they may compete, but both involve the supernatural, both are vulnerable to secularizing social forces, and both are often explained as responses to uncertain danger.1

Page 1 of 27

 

Does Danger Make People Religious? The topic is interesting for the light it throws on two different but often intertwined explanations of religiosity: social function and social organization. Religion may be popular because it addresses some pressing personal or social need—the two are not necessarily the same—or because communities are structured in ways that facilitate the family transmission of religious beliefs and practices, the monitoring of performance, and the exclusion of outside influences.2 A third possibility—that the apparent association of danger and piety is largely in the eye of the modern beholder—will also be considered. Although, for reasons given in the first four chapters, I am firmly opposed to the postmodern notion that researchers can do no more than present competing narratives, it is important to appreciate how layers of social construction may intervene between the observer and subject being researched. Finally, this chapter illustrates the need for professional humility: we may be able to pose a research question and yet be unable to untangle sufficiently the possible causal threads, or be unable to find appropriate evidence, to come to a definite conclusion. That does not make social research pointless: every clear thesis, even when a particular attempt to test it is inconclusive, is another layer of foundation on which others may build.

(p.214) Mining and Fishing Mining and fishing once engaged a large part of the British workforce. In 1800 there were 42,000 miners; in 1900 1,127,000.3 Fishermen are harder to count because many fished part-time, but there were over half a million in 1900.4 The danger inherent in offshore fishing can be illustrated from the churchyard of St Oswald’s, Filey. Walking from west to east, one finds gravestones for the following:5 Ted and George Jenkinson lost at sea 25 November 1925; William R. Jenkinson drowned at Primrose Valley 29 June 1948; Ross Jenkinson lost at sea 29 October 29 1880; William Jenkinson; Richard Cammish Jenkinson and Thomas Castle Jenkinson all lost at sea in an explosion 15 April 1919; George Jenkinson drowned near Filey Brigg 16 January 1874; Thomas William Jenkinson killed at sea 25 July 1914; also son George drowned 5 December 1907; Edmund Ross Jenkinson lost at sea 28 April 1892; William Jenkinson drowned at sea 2 November 1861; George Jenkinson lost at sea in Filey Bay 14 December 1896; Matthew Jenkinson drowned in Boston deeps 3 July 1849; Matthew Jenkinson drowned in Filey Bay 1 December 1863; John R. Jenkinson, Robert Jenkinson, and George F. B. Jenkinson all drowned in the Research disaster 25 November 1925. And this represents only part of what one family lost to the sea: the Jenkinson daughters married into other fishing families whose men are marked with similar gravestones. Mining was also extremely dangerous. All miners risk rock falls, but, because bituminous mines were pervaded by flammable gases, liable to be ignited by sparks from metal implements or by the explosives used to loosen the coal, colliers constantly faced the possibility of mass fatalities. For miners, Page 2 of 27

 

Does Danger Make People Religious? ‘intimations of their mortality were an ever-present reality as they worked in the mine, sensed in every creak of timber, fall of stones, or flicker of the lamp’.6 The following are just some British disasters. On 12 December 1866, 361 workers died at Oaks Pit, Barnsley. In 1877, 209 colliers were killed in Blantyre. An explosion at the Prince of Wales pit in Abercarn killed 270 men and boys—more than three-quarters of the shift’s workforce. The Albion colliery in Cilfynydd saw 295 men killed in a gas explosion in 1894. In 1913 an explosion wrecked the pit at Senghenydd in Glamorgan: 439 men and boys died. From 1850 to 1914, an average of more than 1,000 miners were killed annually. No national data for crippling injury are available, but extrapolating (p.215) from the records of miners’ relief funds in Lancashire and Yorkshire between 1872 and 1896 suggests that there were then 100 long-term injuries for every fatality.7 All accidents are tragedies for the families involved, but, because mining and fishing losses often involved large numbers (either absolute or relative to community size), they became part of the history of the occupation and of the community. Newhouses was a small clutch of terraces in Whitehaven, built by Lord Lowther for his miners, housing around 200 people in 1900. Newhouses families were used to fatalities; the Lowther pits were unusually gassy. An explosion at the Kells pit in 1819 killed 18 residents; two years later 7 died at the William pit; in 1823 the same pit claimed 18; two years later it claimed a further 5. In 1910 an explosion at the Wellington pit killed 136 people, of whom at least 36 lived in Newhouses; one street alone saw 13 deaths.8 Such disasters became part of the community’s memory: memorialized in monument, in poetry, and in song. Because they often involved the suspicion that lives had been unnecessarily endangered by unscrupulous and greedy owners, mining disasters were particular likely to attract national attention and widespread sympathy for the bereft communities.

Danger and Religiosity Linked Mining and fishing communities have long been linked with religion and superstition in social science, in the reports of evangelists, and in the popular culture of those communities. The supposed causal connection is an example of the Marxist’s ‘material base causes cultural superstructure’ method. The link is not money or power but, as in Max Weber’s famous explanation of Calvinism’s part in creating the ‘spirit of capitalism’, the need to resolve psychological stress. The classic text is Bronislaw Malinowski’s study of the Trobriand Islanders: The Argonauts of the Western Pacific.9 The men of the islands fished in two very different contexts: interior lagoons and the open sea. Malinowski observed that, while fishing in the open sea was surrounded by a variety of magic rituals, such ritual was largely absent from fishing in the lagoon. The difference, of course, was the degree of unpredictable danger involved in the two activities.

Page 3 of 27

 

Does Danger Make People Religious? (p.216) John Poggie continued the Malinowskian theme in his 1970s studies of ‘ritual adaptation to risk’ among New England fishermen.10 The success of Methodism among Cornish miners is explained by a wish to display independence of mind but also by evangelical religion’s ability to manage fear.11 A study of the fishing villages of the north-east of Scotland was entitled ‘Faith, Fear and Folk Narrative’.12 A history of the Brethren in the same area wrote: ‘The perils the men were exposed to at sea certainly had an influence on their piety.’13 An anthropology of one village noted: ‘fishermen were said to be uniquely aware of their own mortality and thus their dependence on God in times of peril’.14 We have already met the Jenkinsons commemorated in a Filey churchyard. Although nineteenth-century Yorkshire fishermen were generally baptized, married, and buried by the Church of England, very many were Nonconformists, and fishing villages supported a variety of large chapels. Though tiny, Staithes, for example, had large Primitive Methodist, Wesleyan, and Congregational chapels. It became nationally known for the men’s custom of meeting on the pier before the Sunday evening service and singing popular revival hymns as they walked up the hill, with members peeling off as they came to their respective chapels. Polperro in Cornwall was similar, both in its plethora of chapels and in its nationally known fishermen’s choir. In the 1940s, St Monans on the Fife coast, a fishing village known to its neighbours as the ‘Holy City’, had ten churches to cater for ‘the spiritual needs and zealous propensities of a thousand souls’.15 As well as Catholic, Presbyterian, and Congregational churches, there were ‘Open Brethren, Closed Brethren, Fergusson’s Brethren and Duff’s Brethren’ and there were ‘Pilgrims who, like the Brethren, did not believe in salaried ministers … but who, unlike everyone else, did not believe in churches either, except in the figurative sense, and so brought God under the fabric of one another’s roofs’. There were Baptists and Evangelists and the Salvation Army: They hell-fired at you in the streets one-and-all … This was if you did not come to church. And if you did come to church they made it even worse for you. They leaned over their pulpits and pleaded with you to come forward and be saved till their faces turned purple and the veins broke out on their brows, under the storm-tossed wrath of their raging white hair, so that they looked like something from the world of Moby Dick.16 (p.217) The 1859 revival is often mentioned, but of particular interest is the revival of 1921, which was unusually narrow in its effect. It began in Lowestoft, where the local herring fleet of some 350 boats was joined for the herring season by some 400 Scottish drifters. A history of Providence chapel tells the story:

Page 4 of 27

 

Does Danger Make People Religious? the Baptist minister at London Road Baptist Church in Lowestoft on the East Anglia coast had invited Douglas Brown to preach at a mission there in March … he spoke Monday night and at meetings on Tuesday morning, afternoon and night. The power of the Holy Spirit moved among the people from the beginning. On Wednesday night ‘inquirers’ packed the adjacent schoolroom for counselling and prayer. Sixty to seventy young people were converted that night, along with older people. Each night more packed the ‘inquiry room’ after the service. So the mission was extended indefinitely … By the end of March, the meetings were moved from the 700 seat Baptist Church and other nearby churches to the 1100 seat St John’s Anglican Church … Revival meetings multiplied in the fishing centre of Yarmouth, as well in Ipswich, Norwich, Cambridge and elsewhere. Scottish fishermen working out of Yarmouth in the winter were strongly impacted, and took revival fire to Scottish fishing towns and villages in the summer. Jock Troup, a Scottish evangelist, has visited East Anglia during the revival and ministered powerfully in Scotland.17 While the coastal villages of the north-east of Scotland were strongly affected, this interest in enthusiastic religion had no resonance one mile inland. Beyond being confirmed in their suspicion that fisher people were strange, the farmers and farm workers were untouched.18 Peter Anson, who wrote extensively on fishing folk from the 1930s to the 1960s, explicitly linked danger and superstition: When I first began to be interested in fisher folk-lore, the capture of fish still depended to a great extent on the moods of the winds and the waves … Life both ashore and afloat was still surrounded by fears and forebodings … There was still mystery on every side.19 Welsh Nonconformity—the enthusiastic religion of the Baptist, Presbyterian, and Methodist chapels—and mining are often linked. John Harvey claims that Nonconformity became the religion of the Welsh coalfields in the nineteenth century and that ‘the culture of the collier was thus largely that of the chapel’.20 Colliers praying and singing hymns underground provided one of the motifs of the (p.218) 1904–5 revival.21 And the extraordinary preponderance of lay preachers among the founders and first officials of miners’ unions (especially in Durham and Northumberland) is sometimes expanded to imply widespread piety in the coalfields.

Social Functions The most commonly advanced or assumed cause of the apparent piety of mining and fishing communities is the unpredictably dangerous nature of the work: people are more religious (and an attendant observation, more superstitious) when they are aware of their vulnerability and lack of control over their environment. Highly explosive firedamp was the miner’s constant companion; Page 5 of 27

 

Does Danger Make People Religious? likewise the unexpected storm for the fisherman. Masons and carpenters died in accidents, but they could hope that skill, knowledge, and attention to safety allowed them to manage the risks. Although colliery engineers worked hard to minimize risk—adequate ventilation was a major advance, as was the safety lamp—mining remained hazardous. Religion and superstition reduced fear and anxiety by giving some sense of control, and religion compensated by assuring the vulnerable and their relicts of the sure and certain hope of the bodily resurrection. The Trobriand islanders’ rituals were instrumental, as was the superstition of the miner who refused to work his shift if he saw a nurse on the way to work. But religion in this context had a second purpose: if prayer could not change outcomes, it could make people feel better about their own vulnerability and about the deaths of colleagues and kin. By ensuring the next life, being saved was insurance in the sense of compensation rather than prevention. An aside: mining and fishing are good examples, because they allow us to sidestep one conventional objection to causal explanation. Psychologists and some sociologists suppose that needs are relatively fixed: a product of unchanging biology. Most sociologists tend to the view that needs are flexible. It is because we have been socialized to expect the world to be a particular way that we ‘need’ this or that or react in a particular way to any circumstance. In this context the danger inherent in fishing and mining is useful because the need in question is pretty well immune to the social constructionist argument. How could we interpret the death of our entire male kin in a (p.219) major mining disaster as a good thing? There have been martial cultures—the Spartans, for example—which seem able to socialize members into a genuinely equitable attitude to imminent and early death, but, if we regard ‘fact’ and ‘social construction’ as polar ends of a spectrum, we can probably agree that the dangers inherent in coalmining and offshore fishing have sufficient ‘facticity’ for us to leave aside the constructionist critique of needs. An additional function of religion and superstition, this time social rather than personal, is the integration of a community and its separation from the wider society. Serious religion could be expressed within any strand of Christianity, but fishermen seem to have been particularly attracted to exclusive sects such as the Brethren and the Baptists. Similarly, Peter Anson hits on an important social function of folklore when he says of the many fisherfolk taboos that ‘a stranger would have made endless mistakes without realizing it’, which was precisely one of its purposes.22 Like some forms of religious adherence, it separated the community from outsiders.

Social Integration That danger alone is an unsatisfactory explanation of religiosity is suggested by the fact that there are unpredictably dangerous occupations that are not Page 6 of 27

 

Does Danger Make People Religious? particularly associated with enthusiastic religion. Soldiering is a good example. Despite their periodic vulnerability, British soldiers have not been known for their piety. We might argue that in some contexts soldiers, like the test pilots described by Tom Wolfe in The Right Stuff, persuade themselves that their skill gives them mastery over fate, but the twentieth century has seen many wars where the scale and manner of death are likely to have provoked fatalism among even the most skilled.23 During the Second World War, the crews of Bomber Command took to the air knowing that the odds on surviving were at best 50:50. During the First World War, very large thousands of men died in the trenches of Somme and Ypres, mostly in circumstances entirely beyond their control. Yet there was no discernible increase in religious interest among either combatants or the general public.24 And only the most objectivist view of danger would expect anything different. Malinowski’s idea is not that any sudden threat (p.220) produces a magical response among those previously unfamiliar with such magic; it is that long-term living with threat creates a corresponding culture, which is passed from generation to generation. Social responses to any new stimulus will be shaped by the general extant culture. That the shared culture sustains and transmits the superstitions and the receptivity to enthusiastic religion means that (especially when the general culture is becoming ever more secular) we have a second possible explanation of the religiosity apparently associated with unpredictably dangerous work. It could be that the link works at one remove: that what sets mining and fishing apart is not directly the vulnerability of the workers but the internal cohesion of the communities created and sustained by risk. In explaining the appeal of the Brethren to the fisherfolk of the north-east of Scotland, Dickson notes that ‘the communal nature of fishing way of life was a factor here’.25 Webster makes the same point for Gardenstown.26 The social structure of fishing and mining communities provided a bulwark against secularization and thus allowed attitudes that were once far more widespread to survive in those isolated pockets. The case rests on proximity and homogeneity. Fishermen and miners are unusual in living cheek by jowl and separated from their neighbours. Pit villages were hastily thrown up around mine workings, and miners lived in colliery property, so that almost everyone in a crowded confined area worked in the same pit. There was a considerable degree of nepotism in employment, if one can regard a job as a collier as a benefit to the ‘nepots’. Many Durham mines lasted less than a century, but in that time it was common for father, son, and grandson to work in the mine. Because pit villages were densely populated and isolated, collier families intermarried. Colliers worked in teams that shared the profits of their labours. Thus not just their safety but their earnings were the basis for a strong

Page 7 of 27

 

Does Danger Make People Religious? sense of reliance on each other. The result was an unusually high degree of social integration and the possibility of close monitoring. The cohesion of mining communities was reinforced by the hostility of the wider society: ‘coal-mining communities were generally regarded with some distrust and suspicion by other workers’.27 Ditto for fishing villages. By definition they were on the coast, and so opportunities for social contact were only half those of other places. Their economies had little connection with those of the farming areas immediately inland. Fisherman worked unsocial (p.221) hours and they worked in teams, usually based on the extended family. Their families were directly involved in the business, women and children often being responsible for finishing the dressing of the catch and its marketing. ‘The old fisher cottages, tightly packed among the cliffs of Gardenstown, backed on to one another … this gave rise to a strong sense of the community living in each other’s pockets.’28 In such a context, life away from the prying eyes of neighbours was almost impossible. Those who failed to attend chapel, or to observe rituals such as the churching of women, would immediately be known to their neighbours as deviants and could in various degrees be disciplined by shunning and by being excluded from economically important networks of mutual support. We could construe the effect of the communal way of life as an extension of the social function observation: the community embodies and preserves the need for a response to the unpredictably dangerous nature of the work. Or we could drop the notion of need altogether and simply take the higher levels of religiosity existing in all of Britain at any time since the mid-nineteenth century as given, see the secularizing effects of modernity as the key cause of change, and explain the piety of fishing and mining communities as a consequence of their structure isolating them from that change. Even if mining and fishing had been no more dangerous than farming and building, the same pattern might have been evident.

Change and Secularization It is not easy to see how we could arbitrate between these competing explanations, but one possible test would be to see which of changes in the danger of the work or changes in community structure had the greatest effect on levels of religiosity. So how have mining and fishing changed? Both have declined greatly. In 1921 some 65,500 people were employed in fishing in Britain. By 2012 this had fallen to 12,445.29 Where Scotland’s fishing employed 25,500 people in 1921, in 2002 the figure was just 5,707.30 In 1900 there were more than a million miners in Britain. When the colliery companies were taken over by the state in 1951, the National Coal Board employed over 700,000 people.31 In 2013 there were only 4,000 people employed in mining—almost all of it open cast and (p.222) above ground. But before it virtually disappeared, mining had become considerably Page 8 of 27

 

Does Danger Make People Religious? safer. For the Durham coalfields, the death rate in 1873 was 2.03 per 1,000 employees. Naturally annual figures fluctuate, but the trend is steadily downwards. It falls below 1 per cent for the first time in 1919, and in 1955, the final year for which data are available, it was 0.55 per 1,000 employees.32 For the industry nationally an annual death total that had often passed 1,200 was down to 200 in 1962.33 By the 1990s the fatal accident rate had fallen to fluctuate between 0.01 and 0.03 per 1,000 employees.34 The safety record of fishing is less clear, but it seems to have got worse, improved markedly, and then remained stable. Between 1919 and 1934 the fatal accident rate was 267 per 100,000 fisherman years. This rose between 1935 and 1938 to 457 per 100,000 fisherman years. Since the end of the Second World War, it has been 130 ‘with no discernible reduction … from 1948 to 2005’.35 So what of religion and superstition? Undoubtedly religion has declined. The 1851 Census of Religious Worship recorded attendances rather than attenders. The combined attendance at all three services in the two Staithes chapels that primarily recruited fishermen was 880. If everyone attended only once, that is the total number of attenders. If, as is possible, everyone attended twice, there were 440 attenders. We have no more recent attendance data, but there was a combined membership of 105 in 1945 and just 14 in 2005. We know from many sources that in 1900 most Methodist chapels had attendances around twice the size of the membership and that at the end of the century not all members attended. That suggests not so much a decline of Christianity as a collapse. I leave the details aside, but a similar story can be told for Cornwall’s Polperro and Aberdeenshire’s Gardenstown.36 The Staithes rate of secularization was probably faster than that of Britain as a whole, but, before we take this as proof that the improvement in fishing’s safety record caused fishermen to abandon their now less-necessary religion, we should note that the second half of the twentieth century saw a major change in our other possible explanation of heightened religiosity: although Staithes remained much the same size, the number of fishing families declined, as their children became better educated and moved into white-collar work in the expanding welfare state. The spaces were filled by people who worked in nearby towns but wanted to live in a picture-postcard quaint fishing village. So the community that had previously insulated the (p.223) fisherfolk of Staithes from the general secularization of the country changed to become less involved in fishing and better integrated with the outside world. Because the vast majority of Christians were either Methodist or Church of England, and church and chapel records survive, I was able to compile reasonably accurate measures of membership and regular attendance in three County Durham mining areas.37 The first mines to employ a significant part of the population were in Teesdale. Towards the end of the nineteenth century the Page 9 of 27

 

Does Danger Make People Religious? Deerness Valley saw many shallow pits come into production. In the twentieth century, they petered out and production shifted to the deep mines of the East Coast. So the data series begin at different times. They also end at different times. After 1931 there were few miners in Teesdale, and, though coal was still being drawn from the Deerness Valley until the 1970s, colliers were too small a proportion of the population for area data to represent the religious sentiments of miners. In Table 10.1, the first set of three rows uses chapel membership and parish church electoral roll data to give a sense of the committed core. The second set of three rows uses regular chapel attendance and Easter Day communicant figures to indicate a looser involvement. The patterns are fairly clear. Both measures peak in Teesdale in 1881 and then decline fairly steadily (with a slight rise in the early 1930s). Had there been colliers in the Deerness Valley in the 1880s, they would likely also have been more religious than their successors. The Deerness Valley was certainly never again as religious as it was in 1891. The same is true for the East Coast.

Page 10 of 27

 

Does Danger Make People Religious?

Table 10.1. Church involvement in three Durham mining areas, 1851–1941 (% of adult population) Area

1851

1861

1871

1881

1891

1901

1911

1921

1931

13.72

19.67

25.77

27.53

25.40

24.47

21.82

20.73

21.61

11.61

11.47

8.92

8.99

8.35

3.79

3.54

3.06

1941

Core Teesdale Deerness East Coast

3.02

Core and Penumbra Teesdale Deerness East Coast

27.44

39.34

51.54

55.06

50.80

48.94

43.64

41.46

43.22

23.22

22.94

17.84

17.98

16.70

5.69

5.31

4.59

Source: S. Bruce, ‘Methodism and Miners in County Durham 1881–1991’, Northern History, 48 (2011), 337–55.

Page 11 of 27

4.53

Does Danger Make People Religious? (p.224) We can also accurately describe the religious interests of Durham miners who participated in an NCB-sponsored scheme to encourage miners to move south. Gareth Evans surveyed a sample who had moved to Cannock Chase in Staffordshire in the 1960s.38 He says: ‘For men the centre of their social activities tended to be the pub or the club … The churches and particularly the chapels were remembered as playing an important part in the life of the community, although their importance began to diminish from the early 1930s.’39 Membership of a social club was common: 86 per cent of miners and 61 per cent of their wives belonged. Not one respondent belonged to a church or chapel, and his detailed questioning led Evans to conclude that none had attended a church or chapel except for a funeral, wedding, or christening. The popularity of superstition is much harder to gauge than that of organized religion, but all the comment assumes it declined over the twentieth century. Paul Thompson’s Shetland fishermen, studied in the 1970s, contrasted their scepticism with the credulity of previous generations: ‘Some of the previous generation were “ridiculous. You couldna mention a mouse or a cat … They would put a knife in a mast if they wanted wind … Terrible—but I didna believe in that at any time at all.”’40 Writing in the 1960s, Anson is less sure. As a general stance, he argues that ‘folk-lore is still a living force among Scots fishermen’. However that phrase is introduced by ‘there is no doubt old traditions die hard’, and it is followed by ‘even if the younger men are often ashamed to admit it’, both of which imply the waning of superstition.41 Writing about Whalsay in the early 1980s, Tony Cohen makes no mention of superstition. In his autobiography, Scarborough skipper Fred Normandale, who fished from the early 1980s, describes in great detail the role of heavy drinking and practical joking in the working lives of Yorkshire fishermen, but makes only one mention of superstition: a fellow skipper reported on a dreadful maiden voyage on his new boat, which he blamed on the champagne bottle not having broken properly at the launch ceremony.42 It is thus quite possible to make the case that the secularization of the oncereligious miner and fisherman is the result of the need for religion having diminished. Unfortunately it is difficult to distinguish this from the decline in the size and cohesion of the base communities. This was partly a result of the industries shrinking and partly a result of those who remained in them becoming affluent enough to choose where to live. Although they were still more tied to an industry-related base than (p.225) most people, work for colliers and fishermen was becoming increasingly separate from other parts of their lives and those who wished to escape close community monitoring could do so. Demonstrating cause in the social sciences is always difficult, and here we have the classic problem of a pattern inviting at least two explanations. The secularization of the culture of miners and fishermen is incontestable. And it coincides with both the work becoming less unpredictably dangerous and the Page 12 of 27

 

Does Danger Make People Religious? supportive communities losing cohesion. Ideally we would like to separate the two, but it is difficult to see how we could do that with the naturally occurring data available, especially when logic suggests that the two possible causes are themselves mutually constructing and reinforcing: the tightly integrated nature of Newhouses or Gardenstown was in good part a consequence of the dangerous nature of the work in which their residents were involved. A closer look at timing suggests that neither danger nor community cohesion alone explains religiosity. The churches and chapels in mining areas were declining before the major improvements in safety seen in the twentieth century. And, as can be seen in the extraordinary communal solidarity shown during the 1984–5 miners’ strike, community cohesion persisted while religiosity declined. There is a similar lack of obvious connection between the improvement of safety in the fishing industry and the decline of religion in fishing communities: the chapels peaked long before the industry’s safety record improved. A better connection can be made with community cohesion, but the connections are complex and are mediated by social and geographical mobility. The proximate cause of the collapse of evangelical religion in fishing villages was the generation that was raised during and immediately after the Second World War getting educated and moving into the professions and white-collar occupations that proliferated with the growth of the welfare state. As they moved out of Staithes, Polperro, Peterhead, and St Monans, they also moved away from their religious culture.

The Social Myth There is a third alternative. The inherent danger of fishing and mining may well be a social fact that is relatively immune to the (p.226) distortions of social construction, but the other side of the causal connection—the piety of fishing and mining communities—may always have been exaggerated: a perpetual danger of ethnographic and historical studies that look closely at one community in isolation. Paul Thompson’s Living the Fishing, widely regarded as a sociological classic, claimed there were some 5,000 Closed Brethren in the north-east of Scotland—a figure that is more like the entire Scottish membership.43 Though we see the pit village of Bower’s Row (aka Allerton Bywater) through the eyes of Jim Bullock, who grew up in the chapel culture, most of the colliers were not members of the chapel.44 The Bullock household was a beacon of evangelical self-improvement, but neither adjacent family shared its values. Though Bullock does not put proportions on them, it is likely that the majority of families in the village were closer to the feckless drinking spendthrifts against which the Godly defined themselves than they were to the ranks of the Godly. The Whitehaven local history already mentioned documents the mining fatalities of the residents of Newhouses and describes in some detail housework (it was hard), sports (they liked them hard), and the drinking and fighting culture of the Page 13 of 27

 

Does Danger Make People Religious? colliers (they were hard). It does not mention religion.45 An amateur history of Byers Green, a small Northumbrian pit village, regards religion as less significant than drink: ‘they found a means of escape from harsh reality in the beer parlours, of which Byers Green once had six’.46 A Bewicke Main, Northumberland, miner and trade unionist mentions the chapel just once when he says: ‘The chapel was well attended by the Methodists of the village.’47 He could have said: ‘The chapel was well attended by the people of the village’, but he did not. A minister in 1924 said of the Lothian miners: ‘the same conditions prevail here as in Fife: the bulk of the miners have no church connexion’.48 We can return to Table 10.1. What was not mentioned earlier was the import of rows 1 and 4. The lead miners of Teesdale, whose work was arduous but not unpredictably dangerous, were patently more religious than the bituminous colliers of either the Deerness Valley or the East Coast pits. Even in 1931 they recorded a Methodist penetration rate of 21.61 per cent, and, despite the dominance of Methodism, the Church of England had a higher penetration rate in Teesdale in 1931 and 1941 than it had in either population of colliers.49 One interesting source is the miners’ library. In the late nineteenth century, but more so after the miners’ welfare fund had been (p.227) established by Parliament in 1920, many mining villages had lending libraries, the acquisitions for which were determined by miners. Jonathan Rose’s analysis of a number of extant registers concludes: ‘Welsh miners certainly had an enormous appetite for thrillers, Westerns and tepid sex, though they did not entirely ignore Charlotte Bronte.’ They did, however, entirely ignore religious subjects.50 Chapel members presumably filled their boots with religious literature from their chapels, but the rest displayed no unusual interest in things spiritual. The apparently serious popular religious culture of Staithes and Polperro has been mentioned, but it is possible that such public displays as the singing fishermen have created a misleading impression. One student of the Yorkshire fishing communities notes that, ‘despite occasional revivals, such as that of Filey in 1823, there was no general sustained swing away from Anglican to Nonconformist allegiance’.51 Robin Hood’s Bay, Filey, and Staithes had strong Methodist ties, but Scarborough, Grimsby, Whitby, and Hull continued to support the Church of England. According to the 1851 Census of Religious Worship, between 57 and 64 per cent of the people of Staithes went to church on Sunday, 23 March—above the national average but not remarkably so.52 In 1896 the pious fishermen of Newlyn in Cornwall were so angered by boats from East Anglia fishing on the Sabbath that they attacked them and dumped their catches; three days of rioting followed, with the police, army, and even the navy involved. This reinforces our awareness of the religiosity of Cornish fishing

Page 14 of 27

 

Does Danger Make People Religious? villages, but it also tells us that another major fishing region did not share those principles.53 Of all the Shetland islands, Whalsay, studied by Cohen in the 1980s, is most dependent on fishing. Yet Cohen says that very few Whalsay people were religious and does not report this as a recent change.54 He makes no mention at all of superstition. This may reflect a modern anthropologist’s wish to distance his discipline from its folklorist roots, but it is hard to imagine he would have entirely failed to mention something he thought was important. There are two further reasons for questioning the supposed causal connection between our unpredictably dangerous work and piety: the absence of any mention of it in general historical sources and the already-mentioned fact that fishing and mining were unpredictably dangerous long before fishing and mining communities became a byword for religiosity. The latter observation rules out any invariant (p.228) lawlike link between danger and religiosity, but we could rephrase the supposed connection in terms of variable receptivity: working in an industry with a high death rate does not of itself create a need for enthusiastic religion, but, if a religious revival is making evangelical Protestantism more popular, mining and fishing communities will be disproportionately receptive. Yet John Gillies’s tombstone-sized collection of accounts of eighteenth-century religious revivals makes no mention of colliers or fishermen.55 Thomas Pennant’s 1771 A Tour of Scotland has four pages on superstitions, but they all concern farmworkers. If fishing and mining communities were to the fore in the evangelical movement of the nineteenth century, they should have figured in the standard history of the creation of the 1843 Free Church of Scotland, but Brown does not mention them.56 Aberdeen encompasses the fishing villages of Torry and Footdee, but Allan MacLaren’s history of the aftermath of the 1843 Disruption does not mention them.57 Most significant is omission from either the first or the second Statistical Account of Scotland. The first Statistical Account was the project of agricultural improver Sir John Sinclair, who in 1790 secured the Church of Scotland’s approval to send a lengthy questionnaire to 900 parish ministers. It asked for details of geography, topography and population, of agriculture and manufacturing, and of the cultural and religious habits of the people of the parish. The responses were published in county volumes. Some entries were brief, but many were extensive essays by ministers who were themselves formidable scholars. The exercise was repeated in 1835 with the New Statistical Account. The volume of material is so great I cannot be sure I have read all the relevant entries, but those for predominantly mining or fishing parishes contain no reference to the piety of colliers and fishermen. Given that the general culture of the Church of Scotland was moderate, it seems reasonable to use dissent (unless it was Episcopalian or Catholic) as an index of unusual piety. The 1790s Peterhead entry notes a large Episcopalian population Page 15 of 27

 

Does Danger Make People Religious? (about a quarter), but Berean and Methodist meetings had failed, and the one small Seceder congregation drew its few attenders from far and wide. Fraserburgh likewise had Episcopalians but no dissenters. By the 1840s, Peterhead’s dissent had grown to almost 40 per cent of the population, but it was still overwhelmingly Episcopalian. By the 1840s, Fraserburgh had three competitors, but the Church of Scotland accounted (p.229) for 87 per cent of the population, and most dissenters were still Episcopalian. This entry does, however, contain a reference to fishermen: ‘One of the great evils with which religions has had to contend here … is the excessive use of spirituous liquors. Various regulations have been laid down for restraining that use, particularly during the herring fishery.’58 Twentieth-century Gardenstown has been much studied as an example of a Godfearing fishing community, but there is no sign that it was unusual in the late eighteenth century, when fishing was more important to its economy. The parish entry in the first Statistical Account of Scotland makes no mention of the religious behaviour of the fisherfolk of Gardenstown and notes only seven dissenting families in a parish of almost a thousand families. The entry in the 1842 account is no different.59 Jacob Primmer, an evangelical Presbyterian who served as a curate in the village in 1872, remarks on the good attendances—400 people from a total population of 1,000—and describes a revival in which he cooperated with the Free Church and Seceder ministers: ‘Many of the farmers and country people were constantly with us in the Tabernacle and fisher-folk were not behind hand.’60 Fisherfolk were not ‘behind hand’, but nor were they afore hand. The same point can be made about colliers. They do not stand out in general area studies. The claim that fishermen were unusually superstitious may also have been an exaggeration. People who study miners and fishermen remark on their superstitions. But if we look down the telescope from the other end, a connection is not so obvious. It is, of course, extremely difficult to quantify and thus compare effectively the extent to which any community or occupational group was, or is, superstitious, but we can consider a source that documents superstition per se. In Iona Opie and Moira Tatem’s magisterial compendium, there is no entry for colliers or miners, and, while there are a large number of entries under ‘sea/seamen’, most have inland as well as maritime illustrations. For example, under ‘Counting dangerous/unlucky’, we are told: ‘On no account must the boats be counted when at sea’, but also that Suffolk shepherds in the early twentieth century refused to count lambs until the lambing season was over and that the people of Argyll in the eighteenth century held it unlucky ‘to number the people or cattle belonging to any family … especially on a Friday’.61 That is, fishermen and colliers may have been superstitious, but it is not obvious that they were more so than anyone else; that (p.230) impression may have

Page 16 of 27

 

Does Danger Make People Religious? resulted from researchers studying communities in isolation rather than comparatively. Of course, my study of religion and superstition among people who worked in two unpredictably dangerous occupations is at best introductory. I would want to spend years rather than months on it to be confident of my findings, but I have done enough to raise the possibility that the apparent piety of colliers and fisherfolk is a social myth. If so, how do we explain it? That a theory is internally inconsistent or ill-supported by evidence is enough to damn it, but additional confidence in rejecting some idea can sometimes be drawn from an explanation of why the mistake has occurred. One possible source of distortion is the confusion of supply and demand. It is certainly true that evangelistic preachers were fond of using fishing and mining for illustrative material. Fishing is no surprise; the metaphor is there in the New Testament. Christ calls fishermen to be his disciples and his disciples to be ‘fishers of men’. How better to illustrate the principle that one should prepare now to meet one’s maker than to reference the mine explosions or the boats lost at sea that took young men before their time. And the darkness of the coal mine readily served as a metaphor for the pit into which sinners would be cast. On the other side, the miner’s lamp frequently served in the 1904 Welsh revival as a symbol of light and hope, as did the harbour lights and the safe haven. If one starts from religious discourse, one may well see mining and fishing as especially associated with piety. A second possible source of distortion is the study-in-isolation. Many historical and anthropological studies of mining and fishing areas are based on a single community and lack systematic comparison either with national data or with places that are comparable in all but the nature of their main form of employment. It is thus possible that what seemed distinctive would have lost that status had the researcher looked elsewhere. Anson, for example, collects and presents examples of superstitions that create the clear impression that his subjects were unusually superstitious, but I suspect that, like Thomas Pennant, he could have produced a similar body of observations about farmworkers.62 When there are comparisons, they may be the wrong ones. For example, that Welsh miners might have been markedly more religious than Sheffield steel makers in 1900 or 1930 is not the question. As there are many reasons why one society may be more or less (p.231) religious than another, narrower comparisons are needed: Welsh miners with Welsh farmers, for example. Systematic comparison is difficult because, although the marvellous Vision of Britain website has made reasonably reliable census data available, the areas for which we have good church attendance or chapel membership data rarely correspond with census units. And operationalizing a measure of superstition is impossible. So, in the absence of comparison, any signs of religious or Page 17 of 27

 

Does Danger Make People Religious? superstitious interest among miners or fishermen can be held to be symptomatic. A third consideration is the focus of enquiry. Lay people often assume that apparent differences in religiosity between occupational groups, communities, or classes are related to the core business of religion rather than, as I suggest in my second possible explanation already discussed, being related to apparently tangential concerns such as the structure of the community and its isolation from wider cultural trends. If the only cause of the survival of religion we consider is the appeal of its core concerns, then it is easy to suppose that the elemental nature of mining and fishing explains the supposed religious interest. Although it is not a necessary part of any argument, it is useful to consider why our generation may have got it wrong. A plausible explanation would be that, for a largely secular and comfortable culture that is remote from both strongly shared serious religion and unpredictably dangerous work, the two things seem to fit. There is certainly something about underground depths that causes us to impute heroic status to miners—their constant closeness to death supposedly making them more sensitive to what really matters in life. If any group of workers should be ‘nearer, my God, to thee’, it is miners. The following is the response to underground hymn-singing from an academic and broadcaster who in 2009 produced a re-enactment of a 1929 revival service underground: ‘It was the most incredible sound I have ever heard. The sound was so rich it was overwhelming. I couldn’t actually join in the first hymn because I was so choked up.’63 He would probably not have had the same reaction to Bread of Heaven sung in a car factory by car assemblers. Ditto fishermen. Battling with the elements seems intrinsically noble—a more manly take on Wordsworth’s romanticism. Fishing and mining are elemental. Religion deals with life and death. Therefore fishermen and miners should be more religious than office- and shopworkers. (p.232) Community may explain the myth in the same manner but at one remove. Because it is the case that conventional religion is best preserved by close-knit communities, it is easy mistakenly to work backwards and suppose that all close-knit communities are especially religious. And when the most arresting image of those communities is of people waiting at the pit head or the pier for the return of their dead, imputing an unusual degree of interest in the faith is natural. Finally, unpredictably dangerous work and religion both comfortably fit into the same heritage. They are part of a common past, especially so in Staithes, where the Primitive Methodist chapel has been converted into a museum that displays relics of religion and fishing side-by-side. Both are immortalized in the same sepia photographs. Most of us do not know any fishermen or miners or any

Page 18 of 27

 

Does Danger Make People Religious? seriously religious people. We know shopworkers and clerks, and they are not religious. Here ignorance allows myth-making. In brief I am suggesting that the causal connection between piety and danger may be something of a social myth, an exaggeration by a largely secular people who have been liberated from the metaphorical pit of evangelical religion and from the literal pit of the coal mine.

Conclusion This brief exploration of the links between danger and religiosity—theoretically interesting for the way in which three very different types of sociological explanation can be invoked—is unfortunately inconclusive. The initial observation—the supposed religiosity of miners and fisherfolk—has two equally plausible (and intertwined) explanations. The key weakness in the idea that danger stimulates piety is that the unpredictable danger in mining and fishing is almost a universal while piety is a variable. Hence the importance of other considerations, such as the structure of the supporting community. This muddies the waters, but we could save either explanation by regarding danger or community cohesion, not as causes of heightened religiosity but as retardants of secularization. The introverted nature of the communities (which can be explained as a combination of the nature of the work and the social structure of the community) meant that cultural change came later, which would fit the chronology. For the UK as a whole, churchgoing declined from at least (p. 233) 1851, but, for mining and fishing areas, the peak may have been closer to the end of the nineteenth century. We lack good church and chapel records from much earlier than the 1890s, and, indeed, the Teesdale lead miners apart, there were few Durham miners before the 1890s. So, until better data are produced, perhaps the best we can say is that there is little evidence that miners and fishermen were especially religious when everyone else was religious, and that their apparent concern for the afterlife comes late and may well owe more to their relative insulation from cultural trends. I have suggested a third possibility: that the heightened receptivity to enthusiastic religion (and to superstition) of mining and fishing communities is something of a modern social construction that tells us more about ourselves than about its apparent subjects. That third possibility is interesting. It suggests that social constructions of the past become more important the more remote the past. It also suggests that, when serious religion is itself rare, we are more likely to expect it among people who routinely face death because making sense of suffering is what religion does. Although our expectation that the instrumental bond between the possibility of imminent death and interest in one’s soul is reinforced by the tight cohesion of mining and fishing villages prevents this being an entirely individualistic view of religion—the utilitymaximizing opportunity of the previous chapter—there is still something of a Page 19 of 27

 

Does Danger Make People Religious? modern ‘people choose religion’ assumption about this. Shopkeepers were not as religious as miners and fishermen because they did not need to choose so to be. The way around such modern myth-making is good historical evidence, and here I have to admit defeat. In the time I was able to spend on the topic, I was not able to gather the evidence to take the argument beyond where I must leave it. Such evidence, of course, may simply not be available, and it is to such general methodological conclusions that I now turn. The hypothesis pursued in this chapter is fairly typical of the sort of puzzle that scholars interested in the social causes and consequences of religious belief and behaviour try to solve. We start with a general proposition that we might have inferred from our own observations or read in the work of others (in this case: unpredictably perilous work makes people unusually religious or superstitious) and we try to see how well it fits with such new evidence as we can collect. (p.234) Our first task is to ‘operationalize’ our concepts—an ugly word, but the mechanical metaphor is useful because it emphasizes the need to turn ideas into actions. Vague notions have to be expressed in ways that we can identify and measure. So we start with ‘Working in perilous circumstances encourages people to seek or accept supernatural responses to situations that cannot be pragmatically managed’. What sort of work fits the ‘perilous’ bill? Fishing and mining seem like good candidates, though we should immediately be thinking that inshore and deep-sea fishing probably differ in degrees of attendant peril, as do types of mining. And each may have become less dangerous as improvements in technology, weather forecasting, and geology gave increased rational control. Operationalizing ‘being unusually attracted to supernaturalist beliefs and behaviour’ is in principle straightforward. We need to measure how religious or superstitious are fishermen and miners, and, though some scholars make heavy weather of definition, I doubt whether most of us would have much difficulty in agreeing on what we (and importantly our research subjects) mean by ‘religious’ or ‘superstitious’. However, there is a difference between ‘knowing it when you see it’ (as US Supreme Court Justice Potter Stewart said of pornography) and measuring it. Of all social interests, the popularity of religion is relatively simple to measure.64 All Christian bodies either require or very strongly recommend regularly gathering together to worship God; hence church attendance is an excellent index of religiosity. And, because it takes the form of actions performed more or less often by more or fewer people, we can—as I did in Table 10.1—construct a scale and array our various populations at clearly different points. So we can make detailed comparisons across time and space. Measuring superstition is more difficult. We can easily recognize very superstitious cultures and cultures

Page 20 of 27

 

Does Danger Make People Religious? that are not at all superstitious. That is, the extremes of an imagined scale are fairly easy to document, but it is hard to capture positions on that axis. Once we have operationalized our concepts and determined some measures of them, we see how our independent variable (peril) and our dependent variable (either being religious or being superstitious) co-vary. That is, we seek correlation. But, before we can make anything of what we find, we need to be confident that those we study are representative of those who arguably possess whatever characteristic interests us. Was Gardenstown in the 1850s typical of Scottish north-east fishing communities? As the great footballer Kenny Dalglish put (p.235) it: ‘Maybes aye, maybes no.’ Or, if we have chosen to study Gardenstown because we want to live there for a year—that is, the observations come before the theoretical interest to which they are then directed—we need to consider how far we can generalize from our site to fishing communities in general and then further to people whose work is perilous. Representativeness is one obvious problem with case studies; its less obvious companion is distinctiveness. It is not enough to show correlation in the population that we study: that typical fishermen both face peril and are religious does not of itself establish that peril causes religiosity. We need to go the next stage and check that fishermen are more religious than farmers or clothiers. To return to an example already given, St Monans had a great many worship outlets, but did it have more per head of population than inland Angus settlements of the same period? That is obvious to social statisticians who analyse large data sets, because they very rarely try to correlate just one independent and one dependent variable. Their analytic method is based on simultaneous multiple comparisons. They are not asking: ‘Is perilous work positively related to an index of religious interest?’ They are asking how strongly related to religious interest are any or some of age, gender, ethnicity, occupation, education, or whatever else they know about that might be relevant. All the above is pretty straightforward, and it would not need spelling out were it not the case that representativeness and distinctiveness are barely discussed in any of the many examples I have given of scholars explaining the apparent piety of fishermen and colliers by the unpredictably dangerous nature of their work. They simply assume that the connection between peril and susceptibility to religion and superstition has been well established or is self-evident. The penultimate point I want to make from the hypothesis considered in this chapter is that all the conceptual clarity in the world may not get us over two final hurdles. First, we might not be able to disentangle characteristics. To test which of community cohesion or peril better explained religiosity, it would have been good to find fishing and mining communities that remained structurally Page 21 of 27

 

Does Danger Make People Religious? unchanged over a long period in which technology reduced drastically the number of work-related deaths. But most of my Durham colliery villages lasted less than eighty years—less than three generations—and, though the death rate halved twice in that period, the days of the disasters were still (p.236) sufficiently close for family and community memories to keep alive values and attitudes developed in the nineteenth century. The second hurdle is the absence of crucial evidence. In general, the student of religion is remarkably well served. Far more often than political views, cultural preferences, or leisure pursuits, religious beliefs and behaviour have been tapped by social surveys because, even when they are not the focus of any survey, they are often correlated with the primary topic of the survey: surveyors interested in attitudes to contraception or abortion, for example, may guess that religious orientation may be an influence. Religious organizations have often pioneered literacy and have kept exemplary records. For centuries parish clergy have maintained indices of their work and reported to their superiors. From the start of the nineteenth century until the merger of the various strands of Methodism in 1932, the chapels kept exemplary standardized records of the numbers joining and leaving the fellowship in various ways. The Bolton Quaker chapel has similar records from its foundation in 1678 to 1860. Because the church and state shared a common interest in such social ‘statuses’ as being born, getting married, and dying, we have, for example, centuries of government records of the portion of the population being married with a religious or a civil ceremony. But it is still sadly common that we can specify what would count as good evidence and then fail to find it: some of the records of chapels in the north-east of England that might have helped me compare fishing and farming areas were destroyed by German bombs in the Second World War. Sometimes records exist, but the geographical units they cover do not fit our research needs. For example, because many Anglican parish churches pre-existed the Durham collieries, the distribution of colliers did not map neatly onto parish boundaries, and some imaginative estimation was required to guess how many colliers attended the Church of England. As a general principle, any sort of area research gets more difficult over the twentieth century because improved roads and cheaper cars meant that people could work in one place but pursue their interests elsewhere. However, before we become pessimistic about the prospects of explaining religious beliefs and behaviour, it is worth remembering the most important feature of social science research: it is a collective and cumulative enterprise. My work on fishing and mining communities is the product of probably less than a year in total. There may well be mistakes that will be obvious to others. Some industrious doctoral student may well find sources of data that have (p.237) eluded me and think of new sets of structured comparisons that will allow my tentative thoughts to be taken forward. Any researcher can do only so much, but that is fine because, provided we are asking interesting or important questions, Page 22 of 27

 

Does Danger Make People Religious? and we are sufficiently transparent about our data collection and the inferences we draw from that data, others can continue or question our work. The virtues of research should survive the critical challenge that eradicates its vices, so that, sometimes slowly and painfully, we move further away from error. Notes:

(1.) For a discussion of the secularization of superstition, see S. Bruce, Secularization (Oxford: Oxford University Press, 2011), ch. 6. (2.) Friend thinks it is both: S. Friend, ‘A Sense of Belonging: Religion and Identity in Yorkshire and Humber Fishing Communities c.1815–1914’, Ph.D. thesis, University of Hull, 1992, p. 213. (3.) J. Benson, British Coalminers in the Nineteenth Century (New York: Holmes and Meier, 1980), 127. (4.) M. S. Haines, ‘Britain’s Distant Water Fishing Industry 1830–1914’, Ph.D. thesis, University of Hull, 1998, p. 233. (5.) I visited the Filey graveyard while researching Staithes and noted the very large number of Jenkinsons who had perished at sea, but I did not think to record their details. This listing is the work of I. E. Allen and A. A. Todd, Filey—a Yorkshire Fishing Town (accessed May 2012). (6.) J. Harvey, Image of the Invisible: The Visualization of Religion in the Welsh Nonconformist Tradition (Cardiff: University of Wales Press, 1999), 35. (7.) Benson, British Coalminers, 40. (8.) C. McCourt, Newhouses Revisited (self-published, n.d.). (9.) B. Malinowski, The Argonauts of the Western Pacific (New York: E. P. Dutton, 1961). (10.) J. Poggie, ‘Ritual Adaptation to Risk and Technological Development in Ocean Fisheries: Extrapolations from New England’, Anthropological Quarterly, 53 (1980), 122–9. (11.) T. R. Harris, ‘Methodism and the Cornish Miner’, Cornish Historical Association Occasional Papers, 1 (1960), 21. (12.) F.-J. Brown, ‘Faith, Fear and Folk Narrative: Belief and identity in Scottish Fishing Communities’, Ph.D. thesis, University of Aberdeen, 2010. (13.) N. Dickson, ‘Open and Closed Brethren and their Origins in the North East’, in J. Porter (ed.), After Columba after Calvin: Religious Community in North-East Scotland) Aberdeen: The Elphinstone Institute, 1999, 161. Page 23 of 27

 

Does Danger Make People Religious? (14.) J. Webster, The Anthropology of Protestantism: Faith and Crisis among Scottish Fishermen (London: Palgrave Macmillan, 2013), 5. (15.) C. Rush, Hellfire and Herring: A Childhood Remembered (London: Profile, 2008), 127. (16.) Rush, Hellfire, 208. (17.) Providence Chapel, ‘Forgotten Revival of 1921’; (accessed May 2013). (18.) I heard this sort of comment from my father, who was a child of local farmworkers, and from other very elderly relatives. (19.) P. F. Anson, Fisher Folk-Lore (London: Faith Press, 1964), 11–12. (20.) Harvey, Image of the Invisible, 33. Until the 1970s Welsh Presbyterians called themselves ‘Calvinistic Methodists’, but I will anachronistically employ the current usage. (21.) C. R. Williams, ‘The Welsh Religious Revival, 1904–5’, British Journal of Sociology, 3 (1952), 242–59. (22.) Anson, Fisher Folk-Lore, 25. (23.) T. Wolfe, The Right Stuff (New York: Farrar, Straus and Giroux, 1979). (24.) For a demonstration that there is no evidence of an upsurge in religious interest in Britain associated with the First World War, the Great Depression, the Second World War, or the 2008 financial crisis, see S. Bruce and D. Voas, ‘Do Social Crises Cause Religious Revivals’, Journal of Religion in Europe, 9 (2016), 26–43. (25.) Dickson, ‘Open and Closed Brethren’, 161. (26.) Webster, Anthropology of Protestantism, 5–6. (27.) P. F. Mason, The Pit Sinkers of Northumberland and Durham (Stroud: History Press, 2012), 98. (28.) Webster, Anthropology of Protestantism, 5. (29.) Haines, ‘Britain’s Distant Water Fishing Industry’. (30.) Marine Management Organization, UK Vessel Lists, http://www.gov.uk/ government/collections/uk-vessel-lists (accessed January 2014). (31.) Benson, British Coalminers,, 217. Page 24 of 27

 

Does Danger Make People Religious? (32.) Durham Mining Museum, Statistics, (accessed February 2014>. (33.) Durham Mining Museum, Statistics, (accessed February 2014). (34.) J. K. W. Davies, ‘A Bayesian Analysis of Some Accident Data’, Journal of the Royal Statistical Society, ser. D 39 (1990), 12. (35.) S. E. Roberts, ‘Britain’s Most Hazardous Occupation: Commercial Fishing’, Accident Analysis and Prevention, 42 (2010), 46. (36.) C. Burchill, ‘Fishing for Souls: Faith and Community in a Moray Fishing Village’, Ph.D. thesis, University of Aberdeen, 2008. Burchill misplaces Gardenstown: it is in Aberdeenshire, not Moray. (37.) We do not have data on occupations, but the villages used to construct these figures were so overwhelmingly mining company villages that we can use place as a surrogate for occupation. For methodological details, see S. Bruce, ‘Methodism and Miners in County Durham 1881–1991’, Northern History, 48 (2011), 337–55. (38.) G. Evans, Miners on the Move: The Settlement of Incomers to the Cannock Chase Coalfield from Northumberland and Durham (Stafford: Stowefields Publications, 2002). (39.) Evans, Miners, 16. (40.) Paul Thompson with Tony Walley and Trevor Lummis, Living the Fishing (London: Routledge and Kegan Paul, 1983), 333. (41.) Anson, Fisher Folk-Lore, 12. (42.) F. Normandale, Slack Water (Scarborough: Bottom End Publishing, 2004), 128. (43.) Dickson, ‘Open and Closed Brethren’, 160. (44.) At its peak, there were 75 men in the adult Bible class. If we add another 30 for those who attended only the Sunday service (a lack of commitment that would not have long withstood the pressure to be more involved), we are a long way short of the likely male population of the 120 houses in the village. Some, of course, would have been Anglicans or Catholics, but Bullock’s relatively few references to them suggests they were not numerous. (45.) And this despite Whitehaven having Primitive Methodist, Countess Huntington’s Connexion, Wesleyan, Wesleyan Association, Baptist,

Page 25 of 27

 

Does Danger Make People Religious? Congregational, Plymouth Brethren, and Christian Brethren chapels, and at least five independent evangelical mission halls. (46.) J. Stephenson, A Pocketful of Coal Dust: Reflections on North-Est Folk Life (self-published, 1979). (47.) N. Cowen, ‘On Mining Life and All its Ways’, Durham: Durham Records Office, typescript written c.1940s. (48.) Poor record-keeping means I have lost the details of the author and the article title, but this quotation comes from the United Free Church Record (1924), 213. (49.) A small part of that difference may be due to the presence of Catholics, often Irish migrant workers, in the Deerness Valley and East Coast pit villages. Unfortunately, I have been unable to find any reliable Catholic data. (50.) J. Rose, The Intellectual Life of the British Working Classes (New Haven: Yale University Press, 2001), 255. (51.) Friend, ‘Sense of Belonging’, 171. (52.) D. Clark, Between Pulpit and Pew: Folk Religion in a North Yorkshire Fishing Village (Cambridge: Cambridge University Press, 1982), 54–5. (53.) Wikipedia, ‘Newlyn Riots’ (accessed February 2014). (54.) A. Cohen, Whalsay: Symbol, Segment and Boundary in a Shetland Island Community (Manchester: Manchester University Press, 1989). (55.) J. Gillies, Historical Collections Relating to Remarkable Periods of the Success of the Gospel. Reprinted with a Preface and Continuation to the Present Time by the Rev. Horatius Bonar (Kelso: John Rutherford, 1754). (56.) T. Brown, Annals of the Disruption (Edinburgh: Macniven and Wallace, 1893). The Western Isles came out for the Free Church, but Orkney and Shetland, where proportionately more people fished, did not, and the preference of the Gaelic-speaking parts of Scotland for the Free Church is easily explained by political disputes over land use. (57.) A. A. MacLaren, The Disruption Years in Aberdeen (London: Routledge, 1971). (58.) J. Cumming, ‘Fraserburgh’, The New Statistical Account of Scotland: County of Aberdeenshire. (Edinburgh: William Black and Sons, 1842), (accessed January 2018). (59.) T. Wilson, ‘Parish of Gamrie’, The Statistical Account of Scotland: County of Banff (Edinburgh: William Creech, 1791), 469–78; T. Wilson, ‘Parish of Gamrie’, The New Statistical Account of Scotland (Edinburgh: William Black and Sons, 1842), 271–96, (accessed January 2018). (60.) J. Boyd Primmer, Life of Jacob Primmer (Edinburgh: William Bishop, 1916). (61.) I. Opie and M. Tatem. A Dictionary of Superstitions (Oxford: Oxford University Press, 1989). (62.) Anson, Fisher Folk-Lore. (63.) B. Pedley, ‘Praising the Lord by the Light of Safety Lamps’, The Times, 10 October 2009, p. 112. (64.) Potter Stewart in his concurring opinion on Jacobellis v. Ohio (US Supreme Court vol. 378, no. 184, 1964) said: ‘I shall not today attempt further to define the kinds of material I understand to be embraced within that shorthand description [“hard-core pornography”], and perhaps I could never succeed in intelligibly doing so. But I know it when I see it, and the motion picture involved in this case is not that.’

Page 27 of 27

 

Index

Researching Religion: Why We Need Social Science Steve Bruce

Print publication date: 2018 Print ISBN-13: 9780198786580 Published to Oxford Scholarship Online: August 2018 DOI: 10.1093/oso/9780198786580.001.0001

(p.241) Index Abrams, P. 6 accounts, use in explanation 15, 56, 77, 78, 148, 153–65 ageing 26, 103 agnosticism 53, 54 Althusser, L. 180, 183 Andreski, S. 9, 103 anthropology 21, 95, 112, 120, 122, 175 Asad, T. 39, 51, 58 Bailey, E. 45–6 Banbury 112 baptism 86, 98, 99, 132, 197 Baptists, Southern 26 Barker, E. 20, 112, 156 Becker, G. 195, 198, 199, 200, 210 Becker, H. 123, 126 Beckford, J. 20, 112, 183, 187 believing without belonging 24–6 Bellah, R. 45 Berger, P. L. 112, 118, 146, 175, 177, 193, 208 bias, research 30–3, 108–27 Bible 72, 77, 78 Bible Christians 133, 194 big data 25, 28, 92 Blumer, H. 8, 21, 22, 63 brainwashing 30, 110, 154–6, 162, 163, 164 Branch Davidians 130, 136 Brierley, P. 81–3, 116 Bromley, D. 119, 159, 160, 170, 199 Buddhism 49, 51, 59, 114 Calvinism 32, 97, 144, 167, 168, 200, 215 Catholic Church 33, 72–4, 80, 96, 136, 195 Page 1 of 8

 

Index in Latin America 14, 166 in Scotland 167 Catholics 133, 197 causation 101, 102, 157, 160, 203, 206, 207, 215, 225–6 Census of Religious Worship (1851) 11–12, 222, 227 census, population 23, 60, 85, 93, 202, 231 censuses, church 81–2, 83 charisma 119 charismatic movement 33, 76, 103, 136 Christian Science 121 Christianity and social science 180 church and state relations 39, 61, 113, 120, 236 social functions of 196 type of religion 57–8 church attendance 63, 71, 80, 81, 98, 101, 206 in America 83, 90, 186 in London 91 in UK 3, 6, 11–12, 24, 25, 27, 28, 80–3, 205, 232 in York 91 church membership 63, 85, 97, 195 Church of England 6, 7, 25, 58, 74, 81, 86, 89, 91, 205 Easter Day Communion in 3, 223 Church of Scotland 22, 85, 86, 114, 123 churching of women 88–9 clergy 73, 74, 83, 132, 136, 197 motivation 195, 204–5 role of 64–5 social status of 88 stipends of 88 coercion 155, 156 colliers, see miners, coal Collins, H. 36 communication theory 189 Communion 3, 74, 80, 97, 98, 132, 135, 136 community, enclosed 90–1, 215, 219–21, 225 comparison in explanation 13–14, 56, 72, 76, 197, 199, 230–1, 234, 235, 237 compliance effects 83, 84, 98, 132, 165 Comte, A. 8 confession of sins 161 confirmation, religious 86 (p.242) consent, informed 134, 142, 147, 149, 150 consistency 29, 30, 56, 63, 66, 79, 94–5, 179 conversation analysis 163, 185 conversion 153–73 correlation 74, 79, 100, 101, 102, 201, 204, 234, 235 counting, unavoidability of 70–3, 75 crime and religion 28 critical sociology 114–15, 118, 175 Page 2 of 8

 

Index Crockett, A. 12, 25, 101, 196 Davie, G. 24–5, 64, 65, 112 death 135, 155, 218, 219, 222, 228, 233, 235 defining religion 39–69 by function 44–8 by substance 49–51 defining spirituality 51–5 definition of the situation 17 demography 26, 27, 82, 158, 170 denomination, type of religion 57, 114, 132, 145, 148 denominationalism 58, 111, 147 deprivation, social 18–19, 55, 187 dimensionality 101 disciplines, importance of 6, 30, 101, 175, 189 discourse analysis 160–5 discovery (distinguished from creation) 39, 59 discrimination, religious 27, 57, 96 discursive processes 39 diversity, religious 9, 10, 44, 58, 89, 113, 195–6, 202, 210 domestic violence 124 Droogers, A. 49 Durkheim, E. 8, 13, 17, 41, 114, 176, 177, 181, 186, 187, 203, 207 Easter Day Communion 3, 80, 223 economic rationality 194–211 education 66, 79, 92, 93, 101, 107, 197 Elster, J. 201, 210 Episcopalians 228–9 epistemology 7, 12, 115, 173 ethics, research 129–50 ethnography 7, 20, 21, 92, 95, 104 disadvantages of 21, 22–3, 75–6, 226 evidence 11, 12, 16, 17, 30, 65, 71, 72, 79, 93, 94, 99, 164, 179, 230, 233, 236 evolutionary model of religion 41 Exclusive Brethren 120, 146 false consciousness 14, 65 falsification 12 family size 26–7 family transmission of faith 19, 26–7, 148, 213 feminism 10, 11, 14, 31, 117, 123, 175, 186 Feyerabend, P. 14, 33 fiction as historical source 5–6 Finke, R. 190, 195, 196, 198, 201 fishermen and religion 13, 14, 214–37 and superstition 50, 214–37 Fitzgerald, T. 39, 40, 41, 58, 63, 66 Foucault, M. 176, 180, 183, 184, 190 fragmentation 41 France 21, 187, 195 Page 3 of 8

 

Index Free Church of Scotland 22, 68, 123, 169, 170, 202, 228, 229 free rider problem 196 functions of religion 45, 46, 48, 64–6, 83, 166–7, 213, 218–19 funerals 64, 65, 86, 87, 197, 224 Garfinkel, H. 159, 185 Gellner, E. 33, 187 gender 31, 32, 82, 96, 114, 118, 120, 123, 185, 186 and church attendance 28, 119 and secularization 31–2, 186–8 and spirituality 119 genealogy (Foucault) 39 generalization 3 German sociology 8 Giddens, A. 176, 177, 182, 183 God, belief in 25, 52–4, 77, 78, 87 Hadaway, C. K. 83, 90 Heelas, P. 24, 54, 93, 169 Herfindahl index of diversity 100 Hinduism 18, 49, 52, 114 historicism 20 history (discipline) 5, 34–5, 175, 183, 184, 226 Hoggart, R. 52, 62 human capital 197, 205–6 Iannaccone, L. R. 190, 195–8, 201, 205–7 identification, religious 23, 25, 54, 63, 85, 99, 111 immigration 10, 202 (p.243) implicit religion 45–7 individualism 9, 10, 94, 119, 189 induction 20 Islam 15, 47, 114, 133, 180 Islamophobia 100, 124 intermarriage, religious 206 Jehovah’s Witnesses 200 Jews 16, 187 karma 49 Kelley, D. 26–7 Kemp, D. 22, 24 Kendal 24, 90, 93 King, R. 49 Kuhn, T. 16, 64, 118 language, sociological 29, 55–6, 66, 145, 150, 203 Lenski, G. 9, 10 life after death 3, 208, 233 Lofland, J. 119, 143, 157–9, 161, 169, 170 Luhrmann, T. 21, 22, 75 Lutherans 10, 14, 15, 74, 202, 204, 207 magic 16, 29, 41, 180, 215, 216, 220 Malinowski, B. 14, 16, 50, 215–16, 220 Mannheim, K. 42, 147 Page 4 of 8

 

Index marriage 158, 203, 220 civil 86, 87 religious 61, 86, 87, 99, 197, 216 Marx, K. 44, 177, 180, 186, 187, 208 Marxism 8, 11, 14, 20, 31, 44, 66, 115, 166, 179, 182, 215 mass-communication 189, 192 mass society 189 maximizing utility 193, 198–211 Merton, R. K. 189, 190 Methodists 74, 81, 84, 111, 121, 131, 132, 133, 135, 139, 143, 155, 193, 194, 197, 216, 223, 226, 227, 228, 232 methodological agnosticism 41, 42, 50 millennialism 20 miners coal 13–14, 215–37 lead 13, 226, 233 Moonies, see Unification Church morality 45, 123, 130, 208 Mormons 41, 58, 59, 113 multivariate analysis 27, 102 nominalism 60, 100 normative theory 174, 175, 178, 181 numbers, importance of 20–8 Nye, M. 40 objectivity 14, 15, 29, 30–3, 114–15, 118, 119, 123, 124, 125 paradigms in research 16, 118 participant observation 1–2, 92, 134 partisanship 41, 58, 61, 122, 123, 124, 126, 147, 166, 167, 169, 181, 202 Pentecostalism 3, 114, 121, 123, 189, 201, 202 in Latin America 64, 166, 167, 169, 202 political parties 111, 112 Popper, K. 13, 20 popular (folk) religion 88–90 positivism 7–17, 19–20, 30, 179 postmodernism 31, 40, 119, 162, 164 Poulantzas, N. 180, 183 praxis, Marxist theory of 11 prayer 10, 79, 80, 97, 146, 210, 217, 218 Presbyterians 97, 114, 123, 136, 145, 168, 171, 196, 202, 205, 216, 217, 229 Primitive Methodists 74, 80, 85, 194, 232 probabilities 18–19, 170 professionalism 29, 56, 58, 110, 111, 112, 115, 117, 118, 119–20, 122, 178, 213 puritanism 32, 72, 121, 166, 209 Quakers 236 questionnaires 1, 6, 7, 11, 116, 238 rational choice 193–212 rationality 154, 184, 193–212 reflexivity 18, 122, 182 Reformation 32, 118, 119, 167, 184, 193 Page 5 of 8

 

Index relativism 14, 15, 40, 119, 165 religious belief 25, 28, 40, 43, 66, 76–80, 85, 98, 99, 136, 142, 148, 177, 194, 205, 236 renaming secular as religious 42, 46–7 replication 95, 196 reportage 6, 7 reporting research 141–50 representativeness of research subjects 2, 21, 27, 82, 84, 104, 116, 117, 234, 235 (p.244) research ethics 129–52 revivals, religious 116, 153, 163, 216, 217–18, 227–8, 229, 231 rites of passage 61, 64, 65, 86, 87, 99, 132, 197, 216 Robertson, R. 176–7 role-play theory of conversion 159–60, 170 Runciman, W. J. 7, 8, 175 sabbatarianism 97, 227 Sahlins, M. 18, 207 salvation 31, 49, 72, 80, 144, 167, 169, 200 Salvation Army 216 Sargant, W. 154–6 scales in description and measurement 66, 73–6, 81, 82 scepticism 2, 5, 20, 33, 71, 73, 93, 95, 97, 99, 112, 117, 118, 224 schism 22 sociological theory of 18, 33 schools 60, 113, 136, 187 science natural 8, 10, 16, 17, 18, 29, 33, 60, 63, 75, 98, 119, 207 and religion 39, 50 social 17, 28, 29, 30, 33, 34, 55, 75, 119, 124, 125, 129, 143, 145–7, 149, 153, 164, 165, 169, 171, 177, 180, 186, 190, 207, 225 scientific method 8, 13, 16, 22, 33, 98, 99, 179, 207 Scientology 19, 33, 59, 124, 159, 199, 200, 205 second-order relationships 28, 92 sect, type of religion 27, 57, 121, 123 secularization 9–10, 15, 207 sociological explanations of 9, 20, 30, 31, 33, 52, 62, 64, 71, 98, 208, 209, 220–9 sensitivity 52, 54, 183 Seventh Day Adventists 130 Shupe, A. D. 119, 159, 160, 170, 199 Smith, Joseph 41, 58, 59 social class 19, 32, 56, 57, 59, 60, 66, 93, 96 social constructionism 14, 112, 177, 196, 213, 218, 219, 226, 233 social interaction 139, 140, 157, 159, 197 social mobility 118, 176, 225 social pressure 131 social research, basic principles of 1–4 social science defence of 5–38 disdain for 5 social status 199 Society of Friends, see Quakers Page 6 of 8

 

Index sociological explanation 42, 143, 146, 147, 149, 150, 170, 174, 188–90, 232 sociology of knowledge 42, 147 lay versus professional 8–30, 111 of religion 20, 31, 63, 65, 111, 112, 113, 115–18, 149, 160, 177, 182, 183, 193 of science 42 socio-moral values 27, 45, 116, 149 spirituality 22–4, 51–4, 66, 70, 71, 73, 93, 118, 119, 121 sport and religion 43, 46–7 Staithes 89, 193–4, 216, 222–3, 225, 227, 237 Stark, R. 28, 33, 60, 64, 65, 100–1, 116, 118, 119, 157, 190, 195–210 statistics, value of 5, 9, 70, 76, 86, 97, 101, 102, 103, 117 status inconsistency 92 Straus, R. 33, 159 Stringer, M. 51, 65 strong religion 26–7 structuralism 31 substantive definitions of religion, see defining religion Sunday observance, see sabbatarianism Sunday school 121, 212 superstition 88–9, 213–37 supply-side explanations of religiosity, see rational choice survey(s) 53, 54, 72, 76, 77, 78, 79, 84, 93, 98 questions 29, 77–80 representativeness of 2, 22, 27, 82, 96 response rate 6, 25 sampling 2, 78, 79–80, 82, 83, 84, 90, 117 Sutcliffe, S. 23, 24 symbolic interactionism 21, 65, 68 theologians and social science 25, 112, 114, 116, 117 Thomas, W. I. 17, 22 time use diaries 84 Turner, B. S. 178, 182–3, 185 Tylor, E. 39 (p.245) understanding 1, 8, 17–18, 28, 60, 102, 103, 119, 122, 125, 177 Unification Church 30, 110, 143, 153–6 Unitarian-Universalism 41 Unitarians 21, 26 United States 109, 113, 114, 157, 166, 186, 196, 201, 202, 204 vicarious religion 64–5, 115 Voas, D. 25, 101, 104, 114, 196 Wales 33, 76, 82, 85, 87 Wallis, R. 20, 112, 122, 140, 158, 170, 182, 187 Weber, M. 8, 17, 32, 72, 102, 144, 177, 184, 186, 187, 193, 208, 210, 215 Wesleyan Methodists 154, 194, 216 Wicca 3, 21, 114 Wilson, B. R. 20, 86, 94, 112, 116, 120–1, 177, 187, 208 witchcraft 16, 21 Woodhead, L. 24, 54, 93, 169 Page 7 of 8

 

Index worship 10, 25, 47, 49, 80, 98, 132, 140, 165, 234 yoga 18 York, churchgoing in 90–1, 116, 117 York, M. 22, 24

Page 8 of 8