The Social Psychology of Religion (Psychology Revivals)

Michael Argyle,Benjamin Beit-Hallahmi

• Publisher: Routledge
• Year: 2013
• ISBN:
  9781135041489
• Edition: 1
• Topic: Psychologie
• Subtopic: Psychologie sociale

Originally published in 1975, this book is a completely rewritten, revised version of Michael Argyle's standard work, Religious Behaviour, first published in 1958. A great deal of new research had appeared since that date, which threw new light on the nature and origins of religious behaviour, beliefs and experience.

Trends in religious activity in Britain and the United States since 1900, and the state of religion in these two countries at the time, are examined. Evidence is presented on the origins of religious activity – including the effects of stress, drugs, meditation, evangelistic meetings, personality variables, and social class. Other studies examine the effects of religion, for example on mental and physical health, political attitudes, racial prejudice, sexual behaviour, morals, and the relation between religion and scientific and other achievements.

The findings are used to test the main theories about religion which have been put forward by psychologists and other social scientists, such as Freud's father-projection theory, cognitive need theories, and deprivation-compensation theories.

1 Introduction

 

 

 

 

 

Scope and aims

What we shall do in this book is to present the main empirical findings from social surveys, field studies and experiments about religious behaviour, beliefs and experience. We shall use these findings to test psychological and sociological theories about the origins, functions and effects of religious behaviour. We shall confine ourselves to religion in Britain and the USA between 1900 and 1973, which is where most of the research has been done.

We should define what we mean by ‘religion’. Several social scientists today offer definitions of religion which are very broad and include most systems of beliefs, philosophy or ethics. We decided to use a straightforward, everyday, limited definition of religion as a system of beliefs in a divine or superhuman power, and practices of worship or other rituals directed towards such a power.

Much of this research is about the antecedents of religious behaviour and beliefs; that is, the conditions under which they occur. These conditions in turn may be used to decide which of various possible psychological processes is responsible for producing religious behaviour. These conditions and processes can be regarded as the causes of religious activities, but they need not be regarded as the whole explanation.

A second kind of research is concerned with those correlates of religious behaviour in other spheres of behaviour, which can be regarded as the effects of religion. The basic question here, is ‘Does religion make a difference?’ (cf. Bouma, 1970). The answer is given in terms of both individual and social behaviour. The main difficulty is in isolating the influence of other variables such as social class and education. Lenski (1963) attempted to show that there was indeed an independent religious factor, which affected other social institutions and could be isolated from their influence. We shall examine the influence of the same religious factor, while holding constant other factors in social behaviour.

Some readers may be more concerned with individual religious beliefs and actions, and may worry that the individual believer may disappear behind the generalizations. This danger exists with any kind of systematic social-psychological inquiry. Our basic premiss is that individual behaviour cannot be isolated from its social environment. Our generalizations, when we formulate them, are intended to apply to individuals and their actions. Naturally they cannot predict every case of individual action, but if they have any value at all, they should help us to understand most cases.

Measurements and indices

In this section, the most important indices of religious activity will be described, and in each case two questions will be discussed: how satisfactory the index is as a criterion of religious activity; and how accurate the measurement is likely to be in terms of actual, as opposed to merely reported, behaviour.

Church membership is a widely used index. As an index it is rather unsatisfactory since it tells us very little about how active a person is or what he believes. Some members may have lapsed in enthusiasm while remaining on the books; others may be keen but never actually become church members. This index is often used for the study of changes in time, since the records are readily available for many years past. However, the criteria for membership tend to change with time: in the USA some churches concealed increases of membership before 1936 in order to evade taxes, while in the early years of the century some Protestant churches did not count children, and others did not even count women! Again, care must be taken in making any interdenominational comparisons, since the criteria vary from church to church. The Catholic church attaches a rather broad meaning to membership, and counts all who are baptized in the faith living in the area. The Church of England counts Easter communicants—rather more than the number of communicants on a normal Sunday. Other Protestant churches include only those over thirteen or fourteen who have made the effort to be accepted as members and placed on the roll of the church.

The usual way of studying membership statistics is via records, based on returns from individual clergymen and published in the yearbooks of individual churches. These are summarized in Whitaker's Almanack for Great Britain and in The Yearbook of American Churches for the USA. The figures go back to 1900 and beyond. Another way is to ask respondents in a sample social survey if they are church members. Rather larger estimates result from doing this. In 1954 in the USA, 79 per cent of people questioned in a Gallup Poll claimed to be church members (Rosten, 1955, p. 239), while the membership returns for that year only totalled 60·3 per cent (National Council of Churches of Christ in the USA, 1955). This may be because people tend to exaggerate their religious activities (as, it will be shown, in the case of church attendance) and because ‘membership’ is a rather vague term. It may be confused with occasional churchgoing or with ‘affiliation’.

Frequency of church attendance is another valuable index. It has the advantage over membership that its significance is the same for different dates and denominations. On the other hand, it may be argued that the Catholic church puts on greater pressure for sheer attendance than some other churches. There may be some church attenders who have no real religious beliefs or feelings. However, these people would not be expected to be active in more private kinds of worship, such as saying prayers or reading the Bible. The usual way of studying overt religious activity is by asking people how often they go to church. It is probably better to be specific and ask them if they went last Sunday; the concrete question put like this leaves less room for distortion. It was found in one English survey (Odham, 1947) that people who claimed to go to church ‘weekly’ actually missed an average of seven Sundays a year. In the case of church attendance it is also possible to count the number of people who go to church, and this has been done in a number of local studies by sociologists.

The saying of private prayers, and other forms of private religious act, can be ascertained from answers to questions, but not easily validated against direct observations of behaviour. As an index of genuinely religious activity, this particular criterion is a good one since non-religious motives are less likely to interfere; as a measurement it is unsatisfactory because of the impossibility of checking and the likelihood of exaggeration. All that is generally asked is the frequency with which prayers are said; the answers could be made much clearer by also inquiring into the duration of such prayers. As things are, it is difficult to interpret the very high frequencies reported: about 46 per cent of people in England report the saying of prayers daily, though about 14 per cent go to church weekly.

Attitudes towards religion or towards the church provide another index. By an ‘attitude’ is meant the extent to which a person is favourable or unfavourable to the organization or set of practices as judged by his verbal expressions. Some writers use ‘attitude’ to refer primarily to overt behaviour, in which case verbal measures must be validated, or checked against more direct measures. However, since attitudes are invariably assessed by verbal methods, and since attitudes measured in this way often differ from actual behaviour, it is probably best to treat them independently.

The objection to attitudes as an index of religious activity is that a person may be favourable to religion without either holding the beliefs or engaging in the practices. There is some evidence that this does not often happen. The importance of attitudes is that considerable progress has been made with refined means of measurement; this is one of the most effective kinds of measurement in psychology. One of the best known of all attitude scales is that originally devised by Thurstone and Chave (1929) for attitudes towards the church. Another widely used instrument is the Allport-Vernon-Lindzey Study of Values (Allport, Vernon and Lindzey, 1960; Hunt, 1968) which compares the relative strength of interests in six areas: religious, aesthetic, social, political, economic and scientific. Examples of attitude scales measuring religious attitudes are found in Shaw and Wright (1967), and Robinson and Shaver (1972).

Specific beliefs can be assessed only by asking single questions, and the precise wording of these is important. It is important that surveys should not be compared unless the wording of questions is identical. On the other hand overall measures of the extent to which a person accepts orthodox religious beliefs can be obtained by the construction of lists of items, on the attitude scale model. Kirkpatrick (1949) and others devised series of items to measure ‘religionism’, consisting of questions covering a wide range of conventional beliefs and having a high degree of internal validity. The use of open-ended questions can be illuminating: Gorer (1955), for example, asked his respondents to describe what they thought the after-life would be like. The answers to open-ended questions need coding or classifying before any statistical results can be obtained, and such questions also take longer to answer. It is more usual to give subjects a series of definite alternatives, though sophisticated people often find it difficult to agree with any of these: Leuba (1934) found it impossible to report the religious beliefs of philosophers because they could not understand the questions or agree with any of the answers. The kind of question which goes ‘Do you believe in God? Yes, No, Don't Know’ can be improved by specifying different forms or degrees of belief.

An important recent development in the measurement of religious beliefs is the application of the Semantic Differential, Q-sorts and similar measures. These enable an individual's perception of God to be compared with his perception of, for example, mother, father, doctor, or policeman.

Religious experiences have been asked about in a number of studies. Although relatively uncommon, they are of great interest, and are thought by some to be of great importance. Bourque (1969) used the question ‘Would you say that you have ever had a religious or mystical experience, that is a moment of sudden religious insight or awakening?’ About 32 per cent of the US adult population said ‘Yes’ to this question. Others have suggested that larger percentages will admit to religious experiences if references to mysticism or sudden conversion are omitted (Havinghurst and Keating, 1971).

Further indices that have been used in the investigation, and will be reported later, include professional employment as a clergyman, church worker or theological student—which can presumably be accepted as a good index, though the psychology of these people may well be different from that of the keen follower—and two sociological indices: contributions to church funds, and the publication of articles about religion. The first must, of course, be corrected for changes in real value and the level of wages. A suitable index is perhaps the average donation per head as a percentage of the average wage. Publications have been analysed by Hart (1933) and others in studies of changes with time. It may be objected that publications only reflect the opinions of a small minority, though of course publishers and editors have to stay in business, and to some extent give the public what they think it wants. In some analyses periodicals were weighted by the size of their circulations, so that some measure of the number of people actually reading the articles is included in the final index.

Multi-dimensional measures of religiosity have become widely accepted in recent years. Since it has become clear that being religious involves several discrete kinds of behaviour, which are possibly measurable as dimensions, the idea that a good measure of religiosity has to include several criteria has gained popularity.

Glock (1962) suggested the following five dimensions: ideological, ritualistic, experiential, intellectual, and consequential. The ideological dimension covers religious activities such as prayer and worship. The ritualistic refers to church attendance of various kinds. The experiential dimension covers intense religious experiences such as conversion, ‘speaking in tongues’, or mystical experiences. The intellectual dimension measures knowledge of religious dogmas and practices, and the consequential dimension deals with the effects of religiosity on conduct in other spheres. The measurement of these dimensions is done by using questionnaire items. Despite its theoretical refinement, this approach shares the problems of all questionnaire methods in relating to actual behaviour. Most studies using the multi-dimensional approach (e. g. Glock and Stark, 1966) measured these attitude dimensions and then correlated them with other attitudes. In a number of studies very low correlations have been found between the five Glock dimensions (e. g. Fukuyama, 1961).

Religious affiliation, or denominational membership, can be used as a dependent variable or as an independent variable, to predict or explain behavioural differences. Ideas about how members of various denominations differ stem from historical traditions, theological writings, or assumed psychological differences resulting from the first two. Examples of predictions based on denominational differences in traditions and practices are found in Lenski (1963), Johnson (1962), and others. One problem in conceptualizing the importance of denominational differences is the weight given to other factors, such as social class and education. Another problem is that of historical and social changes, which tend to modify or blunt denominational distinctions.

Limitations and caveats

We would like to spell out some of the possible limitations of our presentation, and to suggest some appropriate caution in using it. First, most of the studies we called upon were done in Great Britain and the USA. We relied on the copious published behavioural science literature in these two countries, where research techniques are most advanced. This limitation of our sources sets the boundaries for our generalizations. We can generalize from our findings to the situation in other English-speaking countries, and with less certainty to other Western countries. Since we are discussing societies which are advanced, industrialized, mostly secular, but with a predominantly Christian heritage, it is not easy to generalize about more traditional, non-Christian societies. Nevertheless, the findings and the theories presented here may be a source of ideas and hypotheses for further explorations in other cultures. Especially when we deal with general theories of religious behaviour, the ideas offered may be, and should be, tested, in a variety of social contexts.

We are also limited to a particular historical period, 1900–73. One may ask: ‘If things are changing so much, how can we generalize or predict anything?’ Some readers may be concerned over how much of what we report is already history. Since we are reporting changes and trends over two-thirds of a century, some of what we report is undoubtedly and deliberately historical. This is true of other findings in the social sciences. The changing structure of society affects behaviour at many levels, including the religious one. We are not necessarily trying to show universal and stable characteristics of behaviour, unless they are supported by empirical findings. We are certain that some readers will conclude that religious behaviour has enormously changed, while others will be impressed by how much continuity and stability are in evidence. Certainly some of the functional relationships which emerge appear to have a good deal of stability.

2 Religious activity in Great Britain and the USA, 1900–73

We shall not attempt to provide a complete history of religion in Great Britain and the USA during the twentieth century. We are concerned with absolute changes in religious activity, and with delinea...