A history of the secularization issue

The Structure of the Scholarly Community
Another device used in relation to these interviews was a short questionnaire in which I asked respondents to indicate, among a list of 25 books and articles including all the important works²⁶ we have discussed so far and a few others (see below), which they had "leafed through", which they had "read entirely", and which they considered as a "reference book". I also asked them to indicate, very approximatively, how many times they had personally met the authors of these articles or books. I weighted the results by counting one point for each mention of a paper "leafed through", 2 points for each that had been "read entirely", and 3 points for each "reference book". I thus obtained the following classification:
Author title points
Berger, The Sacred Canopy (1967) 51

Bellah, Civil Religion in America (1967) 48

Luckmann, The Invisible Religion (1967) 47

Wilson, Religion in Secular Society (1966) 44

Dobbelaere, Secularization: A Multi-Dimentional Concept (1981) 44
Cox, The Secular City (1965) 39

Herberg, Protestant-Catholic-Jew (1955) 34

Shiner, The Concept of Secularization in Empirical Research (1967) 30

Hervieu-Léger, Vers un nouveau christianisme? (1986) 29

Stark & Bainbridge, The Future of Religion (1985) 29

Parsons, Christianity and Modern Industrial Society (1963) 29

Bourdieu, Genèse et structure du champ religieux (1971) 27

Fenn, Toward a Theory of Secularization (1978) 26

Acquaviva, L'ecclisi del sacro... (1961) 26

Wilson, Aspects of Secularization in the West (1976) 25

Bell, The Return of the Sacred... (1977) 23

Isambert, La sécularisation interne... (1976) 23

Le Bras, Déchristianisation, mot fallacieux (1963) 23

Luhmann, Funktion der Religion (1977) 17

Matthes, Religion und Gesellschaft (1967) 16

Pfautz, The Sociology of Secularization... (1955) 15

Gauchet, Le désenchantement du monde (1985) 13

Lübbe, Säkularisierung... (1965) 11

Ferrarotti, Il mito della secolarizzazione (1984) 4

Luckmann, Neuere Schriften zur Religionssoziologie (1960) 3

As we can see, these works can be roughly classified in two groups ("very popular" and "not very popular"), with a small intermediary group. Of course, there is one problem with this classification. I conducted most of my interviews at the CISR Conference in Helsinki. Thus, this classification is only representative of the position of these texts as estimated by members of the CISR. However, as will be shown later on, if properly accounted for, this bias is particularly enlightening.

One way to counter this bias is to use a different source. Another evaluation of the impact of these authors can be made through use of the Social Science Citation Index. The approximative number of citations per year during the 10 years²⁷ following the first publication of a given piece of writing for different works can be represented in the following table (keeping the same order and the same clusters):

Author title citations
Berger, The Sacred Canopy (1967) 13²⁸

Bellah, Civil Religion in America (1966) 9 (3)²⁹

Luckmann, The Invisible Religion (1967) 10

Wilson, Religion in Secular Society (1966) 5

Dobbelaere, Secularization: A Multi-Dimentional Concept (1981) 2
Cox, The Secular City (1965) 11

Herberg, Protestant-Catholic-Jew (1955) 19

Shiner, The Concept of Secularization in Empirical Research (1967) 2

Parsons, Christianity and Modern Industrial Society (1963) 1

Fenn, Toward a Theory of Secularization (1978) 1

Acquaviva, L'ecclisi del sacro... (1961) 1

Wilson, Aspects of Secularization in the West (1976) 0

Bell, The Return of the Sacred... (1977) 2

Isambert, La sécularisation interne... (1976) 0

Le Bras, Déchristianisation, mot fallacieux (1963) 0

Luhmann, Funktion der Religion (1977) 3

Lübbe, Säkularisierung... (1965) 1

There are some interesting discrepancies between what members of the CISR consider important and what social scientists in general consider important. The most interesting of these discrepancies concern Dobbelaere's Trend Report and, to a lesser extent, Wilson's book³⁰. We must postpone the interpretation of this discrepancy until we have some insight into the social structure of the scholarly community. Three other discrepancies - Bell, Herberg, and Luhmann - can briefly be commented upon: they are indicative of the fact that some very well-known sociologists who have written on secularization have found more echo outside of the community than inside of it. Note also that the drop in popularity, which is very pronounced from Shiner down according to the SSCI, is much less pronounced within the community. This is probably due to the specialized culture shared by the members of this community, in which even works that are relatively obscure for social scientists in general are almost required readings. The "intermediate" group probably owes its particular position within the CISR to its popularity among social scientists in general.

This information about the popularity of certain writings offers us a first clue as to the shape of the scholarly community. But they are not sufficient: as we will see, an author can be very widely read without really belonging to the community. We therefore need to push the investigation a little further, by considering the interpersonal relationships, especially as revealed by the answers to the questionnaire, which allowed me to construct a very rough sociogram of the relationships between members of the CISR and the sociologists we have identified as carriers of the paradigm. In the following list, the same authors are classified according to the number of persons who had had very numerous (i.e., generally more than ten personal encounters³¹) contacts with them:

Author number

of

persons
Berger 3

Bellah 2

Luckmann 2

Wilson 11

Dobbelaere 14
Cox 2

Herberg 0

Shiner 1

Hervieu-Léger 6

Stark 4

Parsons 3

Bourdieu 2

Fenn 6

Acquaviva 2

Bell 1

Isambert 8

Le Bras 2

Luhmann 0

Matthes 0

Pfautz 0

Gauchet 0

Lübbe 1

Ferrarotti 2

The comparison of the first and of the third table is enlightening. The most striking discrepancies concern Berger, Bellah and Luckmann³². This is not surprising: Berger and Bellah have never attended a meeting of the CISR. Furthermore, after writing The Sacred Canopy, Berger changed his mind considerably on the subject of secularization, and to a large extent dropped his interest in this topic³³. Thus, not only is he an outsider because he is an American, and never came to the CISR, but also because he no longer really belongs to the community of sociologists of religion. Luckmann, although he came several times to the CISR, is also an outsider, largely for the same reasons as Berger: for two decades, he even more completely neglected the sociology of religion.

The only two cases where the position as a theorist (according to the first table) and the social position (third table) coincide are those of Dobbelaere and Wilson. And their social position can, in turn, explain the discrepancy between their positions in the first and in the second table. These discrepancies are explained by the differences in the judgements of the in-group with regard to the secularization paradigm (essentially the CISR) and of the out-group. Although they are not as widely read as Berger or Bellah by social scientists in general, their particular position in the CISR allows them to boost their readership in this community, and thus to equal the more widely read scholars. (These considerations will cause us to consider Dobbelaere's trend report, which we should have had to discard if we had checked its popularity only in the SSCI, as a very important piece of writing for the secularization paradigm.)

The number of ties Dobbelaere and Wilson maintain with the other respondents is very impressive. Indeed, they can be considered to be standing at the very center of the relational network that makes up the scholarly community which we were looking for. This fact is particularly important in view of Wilson's immense prestige and influence in British sociology of religion³⁴ and of the important informal role he retains within the CISR³⁵. Furthermore, the relations between Wilson and Dobbelaere themselves are particularly intense. First, Wilson's candidacy to the presidency of the CISR was promoted and strongly backed by Dobbelaere³⁶. Second, Wilson and Dobbelaere several times invited each other to teach in their universities³⁷. Third, Dobbelaere started working on his Trend Report during a 1977 Visiting Fellowship at Wilson's All Souls College in Oxford (Dobbelaere 1981, p. 1). Finally, Wilson and Dobbelaere have written at least two papers together³⁸. The fact that the core of the community is constituted of two sociologists whose combined influence is so important provides it with a particularly solid foundation.

The best way to illustrate this situation is through a brief history of the evolution of the debate. In 1967, Berger was very instrumental in proposing the first widely known sociological secularization theory. This, as Robertson remarked (1971, p. 298), occurred in conjunction with the growing interest in the sociology of knowledge, and in this respect, the names of Berger and Luckmann are always associated. At this time, Dobbelaere and Wilson - whose Religion in Secular Society (Wilson 1966), although the idea and the word "secularization" were contained in it, was not as immediately successful as Berger's book - were not very important names. Subsequently, both Berger and Luckmann turned to other interests, and never came to be fully integrated in the community of sociologist of religion. At the end of the 70s, a second wave of secularization theories, much more firmly anchored in a community of sociologist of religion - rather than in sociology of knowledge - erupted. At this time, Wilson wrote several papers and books dealing much more explicitly with secularization (1975, 1976a, 1976b). Fenn (1978) and Martin (1978) proposed fully-fledged theories, and, shortly thereafter, Dobbelaere summed the whole lot up in his Trend Report (1981). This second wave was more than a coincidental flare-up: in fact, it was closely related to the decompartmentalization of the CISR (see chap. 10), and these four sociologists are all closely related to this organization. Martin, Wilson and Dobbelaere all acted as presidents of the organization. Fenn always was a member without official duties - but he maintains a considerable number of personal ties, as can be seen in the classification. In short, the scholarly community in which the secularization paradigm is rooted is centered around the CISR - although, as Dobbelaere explains (1989b, p. 385), it extends to sister organizations like the SSSR,³⁹ the ASR⁴⁰, and the ISA. The persons who played - and still play - an important role in the vindication of the paradigm (which Berger certainly does not do) are the very persons who stood at the command of the CISR in the last decade.

To summarize, the secularization paradigm has undoubtedly been very explicitly vindicated by a concrete scholarly community⁴¹, gathered mainly around Wilson and Dobbelaere, in the framework provided by the CISR. Other sociologists who are important from a theoretical point of view and who stand very close to the center of this community are Fenn, Luckmann, and Martin⁴². Others - Parsons, Berger, and Bellah - do not belong to this community. But most of their writings and theoretical formulations have been accepted by the community: they can therefore be considered as passive carriers of the paradigm. The two other sociologists who appeared important from consideration of the literature (Herberg and Yinger) have not been accepted by this community, and will not be considered to be carriers of the paradigm.

The secularization paradigm has thus been recognized to have come into existence in the 60s, and to be grounded in an identifiable scholarly community: the CISR. Until now, I have been speaking somewhat vaguely of the "secularization paradigm". I can now introduce the terminology I will be using throughout the rest of this book. The new consensus that emerged in the 60s around the topic of secularization will be referred to as the CISR secularization paradigm or, more briefly, the CISR paradigm. The next step in the analysis will be to discuss the writings of the members of this community - and of the other important authors who have been accepted by it - in order to identify the elements of the paradigm.

CHAPTER 4

THE CISR PARADIGM: A FIRST OVERVIEW
I will now delineate the main characteristics of the CISR paradigm, without much critical elaboration. The evidence supporting this construction will be presented in chapter 12. In presenting the elements of the paradigm, I will often underscore their features by contrasting them with some of the assumptions held before the 1960s, that is, before the paradigm had been established. I will first present the most abstract, then the more concrete, levels of the paradigm.
The Meta-Theoretical Level and the Disciplinary Matrix
As noted in chapter 1, a paradigm in the Kuhnian sense consists of at least two levels of abstraction. The first is meta-theoretical, the second infra-theoretical. At the meta-theoretical level, the CISR paradigm rests on a series of beliefs and attitudes. Let us first examine the beliefs. There are essentially two such beliefs: a philosophical assumption, and an epistemological stance, itself largely derived from this assumption.

Most of the theoretical writings contained in the CISR paradigm share a common philosophical assumption, which holds that man is simultaneously and inseparably a rational and a religious being. This assumption contrasts very sharply with previous thinking in this field, which was characterized by a dichotomy. Some thinkers thought that man was a rational animal, and that religion was therefore doomed to disappear, while others thought that he was essentially a religious animal, and that modernity must therefore be combated.

Sociologists working within the CISR paradigm also share an epistemological stance, which holds that sociology must be independent from religion and theology, while at the same time remaining neutral by "bracketing" elements which fall outside of the empirical realm. In this case also, the CISR paradigm stands in very sharp contrast to the situation prevailing previously. Sociologists either considered that science could disprove religion and should fight it or, contrariwise, that sociology was to be a mere tool in the hands of the Church. The evolution leading to this new epistemological stance will be shown to be closely related to the appearance and recognition of sociology of religion as a speciality belonging to general sociology, yet distinctively autonomous inside of it.

So much for the meta-theoretical level. But before reaching the level of the exemplars proper, we must rapidly discuss a number of other elements that belong to the CISR paradigm - although it is not quite clear exactly how they fit into the Kuhnian model. These elements are not beliefs, nor exemplars, but belong to an intermediate level of abstraction, perhaps equivalent to what Kuhn calls the "disciplinary matrix" (Kuhn 1969, p. 182; Eckberg and Hill 1979, p. 926). This level is the result of technical choices made by the sociologist.

Thus for example, the CISR paradigm contains an implicit definition of the area which it covers. First, the paradigm is centered on a problematic that can be defined as the relationship between modernization⁴³ and religion. Thus, it clearly does not refer to social change in general, but only to those aspects of social change that either have their source in religion, or have consequences for religion. In other words, secularization must not be confused with modernization per se (or with rationalization, societalization, democratization, etc.), and attempts like those of Howard Becker (1957), who conceived of secularization as a global process of social change, do not fit into the CISR paradigm. Second, the CISR paradigm is centered on processes occurring at the macro-social level. The church-sect typology, which was so central from Weber and Troeltsch to the beginning of the 60s, is on the margins of the paradigm.

A majority of the persons who work inside the CISR paradigm also share a definition of the object to be studied, that is, religion. This characteristic of the paradigm is less unanimously shared than others. However, I will show that the approaches which are closer to the "core" of the paradigm share a substantive definition of religion, whereas some approaches which can be situated at the periphery of the paradigm tend to rely on a functional definition⁴⁴. But I will also show that where the approach becomes the most relevant to the CISR paradigm, the definition of religion is indeed implicitly substantive - as in the case of Luckmann (1967).

Finally, the CISR paradigm comprises a set of scientific habits, which determine the means which are considered as appropriate for the study of the phenomenon. This means, specifically, that sociologists working in the CISR paradigm typically engage in broad historical analyses and theoretical reflections. These can be supplemented by empirical material - generally collected by others - but the main focus is on conceptualization, historical analysis, and theory. In this respect, the CISR paradigm again stands in sharp contrast to the former scientific habits in the field, which laid stress on an a-theoretical gathering of empirical material. I will show that the conflict between these two sets of scientific habits is closely related to the institutional evolution of the field.
The Infra-Theoretical Level
We now reach the bedrock of the paradigm, the level of the exemplars. As the presentation of these elements is rather complex, I will first give a rapid overview of the overall structure of this level. The precise content of the different exemplars will be explained only after this global picture has been given. The reason for this procedure is not only didactic. In the Kuhnian sense, a paradigm is a model, a relatively simple construction, that can be apprehended fairly easily as a whole. Even though the technical implementation of a paradigm may be very complex, its cognitive basis is simple, and can be spelled out in a simple schema and no more than a few lines. This is what I will do now.

The secularization paradigm is based on a core of three elements: differentiation, rationalization, and worldliness. Each of these elements affects religion in a different way. With differentiation, religion becomes a separate sphere of social life. With rationalization, the non-religious sphere born from the process of differentiation starts to function according to criteria of its own, which are rational⁴⁵, non-religious. With worldliness, the religious sphere born from differentiation loses more and more of what constituted its own interest, to turn to interests peculiar to the non-religious sphere. These three core elements are interrelated in several ways: rationalization and worldliness presuppose differentiation, but differentiation itself proceeds according to criteria of rationality and, in so far as the religious sphere becomes more worldly, it also becomes more rational. The three core elements are then surrounded by a number of subordinate elements, which can best be represented in a schema.

cf. Les theories de la secularisation, Genève: Droz, 1992
As can be seen from this schema, the subordinated elements are themselves interrelated in numerous ways. Generalization and privatization represent two complementary processes which both consist in a relocating of religion in modern societies: religion strikes new roots in the private sphere on the one hand, and in the cultural-political sphere on the other. Decline in practice and unbelief are related for obvious reasons. Scientization and sociologization are both expressions of the more general process of rationalization. Pluralization and rationalization work together to undermine the plausibility of the traditional world view, causing it to collapse. This collapse on the cultural level is closely related to unbelief on the individual level. Finally, all these elements are situated in a field defined by the two introductory statements situated in the upper part of the schema.

Let us now turn to a somewhat more detailed presentation of these exemplars, starting with the two introductory statements situated in the upper part of the diagram.

First Introductory Statement
Some of the roots of the process of modernization can be found in the religious sphere itself. The classic statement of this exemplar is to be found in Weber's Protestant Ethic, and has been thoroughly developed by Peter Berger.
Second Introductory Statement
Secularization does not imply the disappearance of religion. There is not really any classic statement of this element, but it is implied in the whole sociological tradition from Weber onwards, and is closely related to the philosophical assumption mentioned above.

Let us now turn to the exemplars situated within the circle in the diagram.

First Core Element: Differentiation
Functional differentiation is the prerequisite to any process of secularization. It is a process by which the different institutional spheres in society become separated from one another, and perform each a particular function. The classic statements of the idea of functional differentiation abound, from Spencer and Durkheim to Parsons, and have been totally incorporated in standard international sociology through functionalism.
First Correlate of Differentiation: Autonomization
The autonomization of society is one of the possible results of functional differentiation. (It must be noted that functional differentiation does not automatically imply autonomization, but is a prerequisite to it.) It implies a modification in the balance of power between the religious and the non-religious spheres which originated in the process of differentiation. (Seen from the religious point of view, this can also be expressed as "marginalization".) On the social level, the power of social control of the religious institution diminishes; it can no longer impose itself on social actors, but must be "sold" on a free market. On the cultural level, the power of the religious symbols declines, and the moral order is no longer religiously prescribed. This idea, which can be found in various forms in the writings of several sociologists throughout the 20^th century, found its classic expression in the works of Berger and Wilson, and, although it is logically subordinated to differentiation, it is one of the most important elements of the CISR paradigm.
Second Correlate of Differentiation: Privatization
Privatization is another possible consequence of differentiation. It is a process in which religion, being no longer imposed, recedes from the public into the private sphere. In parallel, this religion becomes more subjective, and established belief gives way to personal interpretation. Every individual builds his own sacred cosmos by choosing among the values available on the religious market. This position has found its classic expression in the works of Luckmann and Bellah.
Third Correlate of Differentiation: Generalization
The logical complement of privatization is the generalization of religious world views, beliefs and practices. It is a process by which symbols, values, norms, patterns of behavior and interaction or types of organizations which originated in the religious sphere are taken over in the secular sphere, and become dominant in this sphere after having apparently lost their religious character. The classic statement of this position is of course again to be found in the Protestant Ethic, and has been further developed by Parsons and Bellah and, in a different way, by Luckmann. Very often, proponents of this view use it to show that secularization has "not happened", but this need not concern us here. To justify the incorporation of generalization in the CISR paradigm, it is sufficient to note that it fits perfectly in the paradigm in the sense that it presupposes autonomization, and is compatible with the other exemplars outlined here. Indeed, autonomization, privatization and generalization can all be seen as part of a process of recomposition⁴⁶ of religion in modern society.
Fourth Correlate of Differentiation: Pluralization
Pluralization is another element tied to the idea of differentiation. It is a situation in which cultural uniformity is no longer imposed by political authority, and the "sacred canopy" splits into a multiplicity of competing world views. This applies to the non-religious as well as to the religious sphere. In the religious sphere, it results in a situation in which one religious world-view is set in competition with another religious world-view (and no longer, as in the case of autonomization, only with a secular world-view) in an open market. The classical statement of this position is to be found in all the works which have been devoted to American religious pluralism, and has been incorporated into the paradigm by Berger.
Fifth Correlate of Differentiation: Decline in Practice
Another element closely related to the previous ones, but with a particular position inside the CISR paradigm, is the decline in practice and affiliation. The particularity of its position results from the fact that this element is not at all peculiar to the CISR paradigm, but was already a prominent element in the views previoulsy held by Catholic sociologists. Furthermore, the peculiarity of the position of religion in America makes it difficult to consider this element as a recognized exemplar, although the trend is considered as having been documented without doubt in Europe, in particular by Catholic sociology. One way out of this dilemma is to state, as Luckmann does, that when there is no decline in practice, this is because of internal secularization (1967, p. 37), or, as Herberg does, that it is because religion in reality has come to fulfil another function (1955, p. 261). Another way out is to state that secularization normally results in a decline of practice, but need not have this result. Be this as it may, decline in practice appears as one of the most peripheral and controversial elements of the CISR paradigm, indeed, as a region where it seems to overlap with other paradigms.
Second Core Element: Rationalization
Rationalization is a very general process by which certain areas of social life, which were previously organized on a religious basis, come to be organized according to criteria of means-ends rationality. Another way of expressing this transformation is to speak of the transfer of things and ideas from the sacred realm to the profane. (Means-ends rationality is, in effect, a complete negation of the sacred, in the sense that it reduces things to their empirical functional value.) Although the classic statement of this process is originally to be found in Weber, the way in which it is used in the CISR paradigm (mainly by Wilson) departs from Weber insofar as rationalization includes mainly formal rationality, and neglects substantive rationality (see chap. 8). Finally, it must be stressed that rationalization is not only a transformation in the form of knowledge, but has a direct impact on social life: it expresses itself in a rational economic order and administration, in rational political and legal systems.
First Correlate of Rationalization: Scientization
If we restrict the idea of rationalization to the area of knowledge, it can be seen as scientization. It is then a process by which religion as a form of knowledge is replaced by science. Classic statements of this more narrow process go back to the Enlightenment, and hold a very central role in the less sophisticated approaches situated in the framework of the CISR paradigm. But, even if somewhat relativized by its incorporation in the broader rationalization process in the more complete approaches, scientificization never loses its importance.
Second Correlate of Rationalization: Sociologization
If we combine the rationalization trend with the scientization of knowledge concerning society, we obtain sociologization. This implies that, as a result of rationalization, the determination of the social order is no longer left to chance or to tradition. Sociology more or less takes on the function of religion - which traditionally played a fundamental role in the definition of the social order - and, through social engineering, helps define a rational social order. This element holds a very particular position in the paradigm, for it is almost never explicitly recognized - although Wilson, the most extreme of the secularization theorists working in the modern paradigm, does recognize it.
First Floating Correlate: Collapse of the World View
In a sociology of knowledge perspective, the collapse of the world-view, or sacred canopy, is one of the consequences of pluralism, and must thus be considered a correlate of differentiation. The very competition among different world views in effect drasitically diminishes the plausibility of each of the competing views. This position has been most consistently put forward by Berger and by Luckmann. But in another, partly pre-paradigmatic view that still survives in some theories (for example Wilson's), the collapse of the world view is related to scientization: it is science as a world view that threatens the dominant position of religion and causes it to lose ots plausibility. In this second view, collapse of the world view must be related to rationalization, and not to differentiation - hence its "floating" position.
Second Floating Correlate: Unbelief
The second floating element, which is in some ways related to decline of practice, is unbelief. It is a process situated at the individual level, in which persons cease to believe in a supernatural reality, or lose the sense of the sacred. As such - if unbelief is considered to be an automatic result of secularization which applies to all individuals in a society - it does precisely not belong to the CISR paradigm, but to pre-paradigmatic views. In the framework of the CISR paradigm, this view has been restricted to a statistical sense, namely, that one of the results of secularization is a decline in the proportion of believers - which does not even imply that a majority of the population has ceased to believe. But even so, this position, like the position concerning the decline of practice, is one of the points on which the "mopping-up" work - in the form of numerous surveys - has the most efficiently challenged the paradigm, up to a point where unbelief has become of very minor significance in the paradigm as a whole.
Third Core Element: Worldliness
This last core element is, in a way, subordinated to the two other core elements - for it presupposes a differentiation, and has one of its most important roots in the process of rationalization. It must nevertheless be dealt with separately. On the social level, worldliness is a process (in this sense, I could also use the German term verweltlichung, which conveys the idea of a process) through which societies, groups or organizations turn their attention from the supernatural to this world - for instance in that they allocate a lesser amount of scarce goods or resources to other-worldly purposes. On the cultural level, it is a process by which a distinctively this-worldly ethic takes the place of an other-worldly ethic. (This transformation can comprise the transition from sect to church. But as the CISR paradigm is concerned mainly with developments occurring on the macro-level, this aspect of the question is of secondary importance here.) In contrast to the case of autonomization, there is no reference to power in this element. The classic statement of this position is obviously again to be found in the Protestant Ethic as far as the cultural level is concerned.