A history of the secularization issue

David Martin
David A. Martin was born in London in 1929²⁸¹. After teaching for seven years in primary schools, he earned an external degree in sociology from the University of London through private part-time study²⁸². He received his Ph.D. from the London School of Economics in 1964. In 1965, Martin burst into the secularization "debate" with a paper entitled "Towards Eliminating the Concept of Secularization" (Martin [1965] 1969, pp. 9-22). The paper started with a warning: "This is a work of 'demolition'". True to his word, Martin systematically attacked the "utopian elements" of the notion of secularization. Briefly stated, Martin's argument rests on the idea that there can be no yardstick allowing us to measure anything which could be designated as secularization. All the notions - like "this-worldliness", or "religion" - which could be put to this use are much too ambiguous and shifting with regard to different historical circumstances to allow for comparison²⁸³. Secularization theory must therefore be understood as expressing a variety of counter-religious ideologies - most notably rationalism, Marxism, and existentialism. In other papers reprinted in The Religious and the Secular, Martin continued to undermine the secularization thesis, pointing to its diverse "utopian aspects" (1969, pp. 23-36), and underscoring the polysemy of the concept (1969, pp. 48-57).

In the following years, Martin joined the scholarly community vindicating the secularization paradigm. From 1975 through 1983, he acted as president of the CISR. During the same period, he appeared to retreat from his former position. In 1978, he published a book under the title A General Theory of Secularization. The title of the book, however, is rather deceiving. By "general", Martin does not refer to some universal process, but rather to a series of considerations qualifying an unduly universalistic theory: "The general theory is general in that it relates 'universal processes' which are empirically quite well established to a typology of cultural contexts and then specifies the type of refraction[²⁸⁴] which the processes then undergo" (Martin 1978, p. 3).

Martin's position on secularization is not easy to understand. In 1965, he wanted to eradicate "secularization" from the dictionary (although later on, he explained that he wanted to "open a debate rather than to banish a word"; 1978, p. viii), but in 1978, he proposed a secularization theory under the most ambitious title ever. To give another example of apparent inconsistency, the concluding sentence of the 1965 essay read: "Secularization should be erased from the sociological dictionary" (Martin [1965] 1969, p. 22); but the next essay reprinted in the same book (written four years later!) started with these words: "Christianity has experienced at least two major types of secularization" (Martin 1969, p. 23). These apparent contradictions often draw gently ironical remarks from his colleagues²⁸⁵. Martin answers by describing these two analyses as running "along two lines" (Martin 1978, p. vii) which are not contradictory. There is indeed good evidence that Martin did not really shift his views, but held them concomitantly. The first chapter of the General Theory, which constitutes its blueprint, was written as early as 1968 (1978, p. vii) - three years before the publication of an article in which he essentially reasserted the conclusions of his 1965 iconoclast paper (Martin 1971). Martin's position with regard to secularization can best be understood if one considers that he is opposed to the notion of secularization as a unilinear, irreversible process, but agrees with the idea that secularization is a complex and ambiguous process.

The writings of Martin probably provide the most typical example of "mopping-up" work in the framework of the secularization paradigm. At the onset of his General Theory, Martin accepts the pattern of secularization, as laid out, among others, by his friend Bryan Wilson (Martin 1978, p. 12). But he remains rather vague about this underlying process. "Certain broad tendencies towards secularization in industrial society have already been fairly well established. They are of the following kind". What follows is a list of seven quite classical items (like industrialization, urbanization, differentiation, etc.), the seventh of which reads: "And so on. There is a wide variety of such propositions" (1978, p. 3). To my knowledge, the most detailed analysis²⁸⁶ of the underlying process attempted by Martin is to be found on pages 115-17 of The Religious and the Secular (Martin 1969). I would contend that this apparent sloppiness is possible precisely because Martin works in the framework of an established paradigm²⁸⁷, and that his concern is not to revise or re-state the basic features of this paradigm, but to bring it in closer harmony with empirical (i.e., historical) evidence. This, in other words, is exactly what Kuhn calls a "mopping-up operation".

Starting from this "universally" acknowledged process, Martin then proceeds to introduce some qualifications. In other words, Martin contends that the exemplars would apply only if all other things were equal. "But things are not equal - ever - and [...] they are most conspicuously not equal with respect to the particular cultural [...] complex within which they operate" (Martin 1978, p. 3). In light of this fact, Martin introduces very numerous qualifications into secularization theory. In his commitment to a complex view of the secularization process, he often seems to somehow abandon the general thread of his argument to inquire into multi-faceted variations of the pattern. As a result, his typologies are not always consistent, his terminology is shifting, and many things remain implicit. I will not even try to be exhaustive in my presentation of Martin's argumentation, but will rather stick to the most general contours of his theory.

Martin's qualifications bear primarily on two exemplars: pluralization* (Martin speaks only of pluralism), and differentiation. Pluralism is the most important of the two. It is treated in the General Theory as an independent variable, whereas differentiation appears as one of the dependant variables. Martin's central idea is that the degree of pluralism in a given country will have a very powerful effect on the outcome of the secularization process. The countries analyzed can all be situated on a continuum which runs from those where religion is organized on a monopolistic model (Spain and Portugal) to the country where religion is the most pluralistic (the United States). Along this continuum, three broad categories emerge: 1) Catholic Monopoly; 2) Catholic-Protestant Duopoly, or the Mixed Pattern; 3) Protestant Pluralism.

In the case of the Catholic monopoly, the transition to "modernity" is assisted by the Enlightenment (1978, p. 8). The gist of Martin's argument is that, as a result, society splits into two warring halves: Catholicism comes to be identified with the antecedent system of authority. In other words, there appears a bipartition between Catholicism and secularism, and the Church becomes an ally of the political right - so much so that even some religious groups (Protestants, Jews) may find themselves pressed into an alliance with irreligion. A vicious circle ensues, in which religion becomes a bone of political contention. By contrast, in the case of Protestant pluralism, this vicious circle never starts. The transition to "modernity" having been assisted by Calvinism, religion as such never becomes a political issue. There is no inherent conflict between "progressive" secular forces and "reactionary" religious forces, and the political and the religious establishments are from the start clearly differentiated. The plurality of the confessions makes possible the adaptation to the different political attitudes and status levels: "The universalization of dissent permits religion to take on as many images as there are social faces" (1978, p. 30). Religion dilutes itself into civil society, and a vague "civil religion" emerges to legitimize the state. With the intermediate, or "mixed" pattern, the situation is more complex. On the one hand, two quasi-monopolies (in the approximate "60/40" ratio) vie for domination. Generally, Protestantists are the majority, while Catholicism is a powerful minority (the best example being the Netherlands). This situation could lead to a fratricide confrontation. But on the other hand, this very danger forces every side to make compromises in order to maintain unity. Thus, a pattern of tolerance emerges, which crystallizes itself in a federalist structure. Furthermore, the Catholic Church contributes to this "beneficent circle"²⁸⁸ by appearing politically on the left or on the center, thus removing the issue of religion from the political arena.

This basic tripartite typology is rendered much more complex as Martin examines some other factors, like the relation of religion to nationalism (which can affect the Catholic pattern, as in the case of Poland), the existence of large minorities concentrated in a particular geographical area, or the "secular monopolies of the left" (communist countries). I will refrain from discussing these further complications, since they would carry me too far away from my concerns.

The second exemplar where Martin brings major qualifications is differentiation²⁸⁹. Martin's strategy is to "[run] the impact of differentiation [...] through the refracting elements of the various patterns [on the monopoly/pluralism continuum]" (1978, p. 69). Differentiation is thus in a way treated as a dependent, or at least as an intervening variable: what, given a certain pattern of pluralism, is the further impact of differentiation on secularization? Martin analyzes this impact with respect to different dimensions of differentiation: between church and state, in politics, in the unions, in voluntary associations, and with respect to cultural identity. The most interesting for our purposes is the differentiation between the church and the state. In Catholic countries, one finds "enormous pressures toward the separation of church and state, of Christianity and civil religion, simply on account of the immense split over religion per se" (1978, p. 71). As one moves in the direction of pluralism, the extent of the separation becomes less marked: in Sweden, some degree of association between the church and the state obtains - but it is minimal, because Sweden's pluralism is also minimal, in that it is limited to internal dissent (1978, p. 33). In more genuinely pluralistic countries, like for instance England, one finds a strong church-state relationship. But what about the polar case, the United States? In a sense, this case seems exceptional. "However, as was argued earlier, it is just this explicit separation of church from state that enables a pluralistic religion-in-general to buttress the higher-level legitimations of American society" (1978, p. 70). In other words, the American state is not as differentiated from the church as appears on paper, because it is firmly grounded in civil religion.

The upshot of these considerations is that secularization is far from being an universal or unilinear process. What the secularization paradigm has gained through Martin's work is a further refinement and distance from the mechanistic XIX^th century evolutionary assumptions. Another interesting feature of Martin's analysis results from his ambiguous position with regard to secularization, a position which has consequences for one of the most important exemplars: generalization*. As noted above, the acceptance or rejection of this exemplar rests largely on an a priori decision: the definition of religion. People like Luckmann (and, as we will see, Parsons and Bellah), who work with a sufficiently large definition of religion, accept the idea of generalization*. These theorists are also the most inclined to reject "secularization", precisely because they estimate that the idea of generalization* is incompatible with the notion of a decline of religion. On the other hand, people like Berger and Wilson, who work with a much narrower definition of religion and pass a more severe judgment on its fate tend to refuse the idea of generalization*. Martin takes an intermediate position in this context. He accepts both secularization and generalization*. And he shows why the two must be linked: "Any characterization of the United States must emphasize the fact that it represents a very high degree of differentiation in that church is formally separated from the state and even religion from the school, and yet the overall social order is legitimated by a pervasive civil religion. The differentiating impulse presses a sociological retro-rocket, and in America this is exemplified by its civil religion as described by Robert Bellah and William [sic] Herberg" (1978, pp. 28-29). No matter what justified reservations one may have concerning the notion of "sociological retro-rocket", the link between differentiation and generalization* is clearly brought out: If religion takes as many forms as there are social interest groups, it can no longer function as a unifying world view, therefore some kind of national myth has to emerge to foster unity (1978, p. 36).

Richard Fenn
Richard K. Fenn was born in the early 1930s. After an original training in theology, he became a minister of the Episcopal Church. Very soon, however, having read Bonhoeffer and Bultmann, Fenn concluded that they were right, and decided to live in the world "as though God were dead". He left the congregation, and became a sociologist²⁹⁰. Thus, Fenn clearly came to the study of secularization through secular theology. In one of his first papers (Fenn 1968), he analyzed the death of God theology as the symptom of an ideological crisis in American society. From these concerns with theology, he quickly moved to the problem of secularization as posed in sociology. In 1969, he published a paper where he lay the foundations for "attempt[ing] to develop a testable theory of secularization" (Fenn 1969, p. 112). After breaking down the "religious" and the "secular" orientations into different dimensions of normative orientation, Fenn challenged the conventional labeling of certain activities as "religious", and of others as "secular", thus laying the groundwork for his later approach.

But Fenn's first major contribution to the secularization debate came in the form of an attack on Parsons entitled "The Process of Secularization: A Post-Parsonian View" (Fenn 1970). Fenn's dissenting opinion is based on an alternative view on the process of differentiation. Differentiation, Fenn argues, does not result simply in the specialization of different institutions, but also in a separation between the different levels which Parsons considers as congruent: culture, social structure, and personality. The very fact that culture, as well as social structure, have become internally differentiated makes congruence impossible: "How [is] congruence [...] to be achieved if incongruence obtains within the culture itself and if subvalues proliferate on the structural level [?]" (1970, p. 131). In making his intervention, Fenn recognizes that he enters a debate that has been going on for some time, and he explicitly refers to Luckmann: "What [...] distinguishes Parsons is the basic premise of congruence, whereas a writer such as Luckmann begins his analysis with the frank recognition of incongruence in modern society" (1970, p. 129). As a result of this position, modern society cannot be considered to be based on a moral order; the sources of its integration must lie elsewhere²⁹¹. The essay concludes on an outline of a theory based on a totally different model, in which the conflicts, at the cultural as well as at the social level, constitute the theoretical focus.

This first sketch was developed further two years later in a paper in which Fenn, influenced by Daniel Bell²⁹², argued that not only did modern societies depend much more on partial than on total ideologies, but that these partial ideologies were increasingly limited to pragmatic interests and utilitarian norms: "It appears to be characteristic of modern societies to avoid conflict over matters of value and to reduce conflict to the level of instrumental concerns, to matters of fact" (Fenn 1972, p. 20). Modern societies are viewed by Fenn as moving toward a dual model, in which, in view of the fact that a large minority is chronically unemployed, it could be advantageous, rather than deleterious, to decrease the level of motivational commitment: "Modern societies [...] will be able to afford higher levels of motivational noncommitment and may even seek to raise the level of noncommitment in order to 'cool out' overmobilized segments of the population" (1972, p. 29). The general conclusion with regard to secularization is very close to Wilson's²⁹³: "Secularization therefore does not drive religion from modern society, but rather fosters a type of religion which has no major functions for the entire society" (1972, p. 31).

Although these considerations were developed again in a somewhat different form in a later paper (Fenn 1974), this new formulation did not include any dramatic shifts. In 1978, basing himself on these first sketches, Fenn put forward a complete formulation of a secularization theory, under the modest title: Toward a Theory of Secularization. As this piece of writing contains the most explicit and complete treatment of the subject, we can turn directly to it to present the theory.

Fenn's secularization theory comprises five steps. These steps, however, are not strictly chronological. Even though step two, for instance, cannot take place as long as step one has not happened, there is not a clear-cut transition from step one to step two. For one thing, even when step two has started, the processes characteristic of step one are not abolished: The different steps can exist simultaneously. Furthermore, there is no necessary pattern leading from one step to another, but a return to a previous step is always possible. Fenn's secularization theory is the most open-ended of all those analyzed in the present work. "We must [...] be prepared to see the process of secularization in complex patterns: as a cause as well as an effect of religious groups, movements, and institutions; as proceeding according to a variable sequence of events and leading toward no single outcome; and as involving a number of reversals and simultaneous but contradictory developments at different levels of a single society" (Fenn 1978, p. 29). This having been said, it remains true that Fenn's five steps are arranged in a roughly chronological order, and that his theory needs almost no reconstruction to fit my own diachronic scheme.

Fenn does not explicitly discuss the initial situation* with regard to religion before the onset of secularization. But this situation is clearly assumed to be a form of religion in which there are no religious specialists. Step one of the process, which corresponds to the rupture*, is the emergence of the clergy. This first step is described very rapidly: half a paragraph on pp. xvii-xviii, and one paragraph on p. 32 (1978). From this, we gather that a clergy, as clearly distinct from the "cultural or political center of a community" (1978, p. xviii), first appeared "several millenia ago" in the primitive communities, and that this event "lies at the origins of secularization" (1978, p. 32). This process of differentiation, however, does not only lie at the origins of secularization; it constitutes one of the aspects of secularization itself. During the whole process of secularization, well after the next steps have come into effect, distinct religious institutions continue to emerge. Thus, step one of the process remains effective all along. In other words, Fenn's view is based on the idea that the basic phenomenon underlying secularization (both in the sense of having started it, and of constituting its continuous underlying frame) is differentiation. However, the resulting process can be reverted in certain historical circumstances: Just as new religious institutions can appear, so they can disappear, giving rise to a greater homogeneity (1978, p. 32).

The second step in Fenn's process is the "demand for clarification of the boundary between religious and secular issues" (1978, p. 32). This step is also very rapidly described, in less than one page. The demands for clarification represent the correlate, on the ideational level, of differentiation on the social-structural level: "The separation of religious from secular issues often takes longer than the separation of religious from secular structures" (1978, p. 33). In a sense, up to this point, Fenn's theory is exactly the opposite of Luckmann's. As we remember, Luckmann argues that societies became differentiated first on the level of culture, and that from this initial event, a further differentiation on the cultural level arose. In other words, although Fenn uses the same exemplar as Luckmann, he uses it in a theoretical framework that is so singular that the end result is markedly different in some respects²⁹⁴. As the issue of the clarification between the religious and the secular issues lies at the very heart of Fenn's conception of secularization, a few clarifications as to this notion are in order. As an introduction to this approach, we must briefly mention one aspect of Fenn's broader approach.

Fenn's secularization theory is inserted in a sociology of sociology perspective. One of the results of secularization is to "undermine [...] confidence in our belief that a social whole exists which is more than the sum of its [...] parts" (1978, p. xiii). In other words, to speak of "American society", for example, can be no more than a convenient metaphor. In the Durkheimian tradition (as continued most notably by Edward Shils), however, this metaphor is taken at its face value. This tradition of sociological analysis can be understood only as a response to the anxiety about the weakening of the metaphysical status of "society" as a category: "The sociologist, like the priest, takes the fragments of everyday life and displays an underlying, moral order" (1978, p. ix). Fenn refuses this tradition which, following Ernst Cassirer, he calls "mythical sociology", and places himself in the discursive tradition (as continued, most notably, by Daniel Bell): "The tradition of Simmel and Weber [...] represents a far more secularized, demythologized stage in the development of social theory. [...] Just as nineteenth and early twentieth century sociologists demythologized religion, cannot the sociology of the twentieth century demythologize sociology?" (1978, p. 6).

From this general perspective, Fenn draws the conviction that it is not the sociologist's task to give a definition of the sacred, nor of religion. "Religion", or the "sacred" do not exist as such; they are constructed by concrete human agents in a continuing struggle for the legitimation of claims to social authority of different groups. "To establish arbitrary semantic limits for the sake of analytical clarity would make analysis all the more difficult since our terminology would fail to orient us to the highly ambiguous and contestable boundary between the sacred and the secular. That boundary is continually contested, expanded, obscured, and contracted in the process of secularization" (1978, pp. 24-25). "Sociological language may offer premature clarity by defining what is socially ambiguous, thus missing the sociological point" (1978, p. 29). Indeed, as already noted on several occasions, sociologists who give a broad definition of religion tend to refuse the idea of secularization, whereas those who work with a narrower definition tend to accept it. In Fenn's view, this choice is not only a choice faced by the sociologist, but a choice present in actual social life. Therefore, by choosing one or the other of these definitions, the sociologist takes sides in the social contest, and thereby incapacitates himself as an observer.
In other words, Fenn contends that sociologists fail to draw all the conclusions contained in the notion of differentiation. As society has becomes structurally differentiated, it has also lost its cultural homogeneity. As a result, the sacred is defined differently by actors located in different institutional spheres. These different actors all attempt to use "the sacred" to bolster their claims to social authority. Thus, secularization is not caused by "impersonal or abstract forces like technology or education, but [by] living and active human agents" (1978, p. xii). The conflicts between these agents must not be brushed aside by definitional fiat, but rather taken as the very starting point of the theory, placed at its very center: "My own approach is to regard these sources of ambiguity as the essential ingredients of a theory that will locate and order them as elements of a complex, indeterminate process" (1978, p. xvi). The aim of the theory must not be to remove this dilemma, but rather to "articulate" it (1978, pp. 27-28). As a result of this conception, Fenn uses a definition of secularization which is neutral with regard to the definition of the scope of the sacred: "Secularization is a process in which the parts of a society seek to define their relationship to each other and to the larger society" (1978, p. 29). Thus, although he uses the same exemplar as Luckmann, Fenn incorporates it in a very different theory of religion. Religion is no longer a universal anthropological foundation, but the result of a social struggle. As we will see, however, with regard to secularization, the result is very similar: religion becomes at the same time generalized and privatized, and the individual's relation to the social order undergoes a radical transformation.

Step three in Fenn's scheme is the emergence of civil religion (generalization*). This development comes as a response to the process of differentiation. Civil religion is a creation which allows society to transcend the potential conflict between its different constituent parts. In a way, this development constitutes a desecularization, in that one level of society which was hitherto secular becomes encapsulated in religion. However, this desecularization is, in large part, illusory. Fenn is definitely a critic of Bellah's civil religion hypothesis. He considers this type of political religion to be merely "inauthentic rhetoric" (1978, p. 43). Why, then, does he include civil religion in his scheme? Because it constitutes one of the prime examples of the underlying process which lies at the heart of secularization: the conflict over definitions of religion²⁹⁵. With civil religion, political authorities use a religious legitimation as a basis for their claim to be taken seriously. But they can do this only by redefining the boundary between the religious and the secular so as to encompass the hitherto secular state into a religious conception. Thus, Fenn considers that a complete secularization theory can be formulated only by combining insights from the "mythic perspective" - of which he is rather critical - with insights from the "discursive perspective" (1978, p. 10). Indeed, it is one of Fenn's basic aims to disclose the "basic theoretical factors" that lie "beyond differing perspectives" (1978, p. 30). As we will see later on, this concern is characteristic of the mature phase in the life of the paradigm.

Step four synthethizes all that has been said thus far. The struggle over the boundary between the religious and the secular becomes general. Not only the clergy (as in step two) and the state (as in step three), but a wide variety of groups, some of which are very small, enter the contest, and redefine religion in their own terms. This new development in the process of secularization first took place with the advent of the Reformation, which, in giving the "sanctions of the sacred to private religious judgment", started to "blur the line between the sacred and the profane that had been so carefully drawn over centuries of conflict between church and state" (1978, p. 57). As a result, we witness a dispersion²⁹⁶ of the sacred: "The dispersion of the sacred has given legitimacy to a wide range of actors in modern societies: to charismatic individuals, to the people as a whole, to elites with access to scarce or transcendent knowledge, and to large, powerful institutions. [...] There is no single whole that can lend authority to particular parts without being challenged by another part. [...] It is difficult to know whom to take seriously because each part of the society may appeal to sacred or quasi-sacred sources of authority" (1978, p. 37). A good example of such conflicts involving definitions of the sacred is the Jehovah's Witnesses refusal to salute the flag. In this context, decisions over the boundaries of the sacred increasingly tend to befall to the courts.

This step again constitutes at once a move forward, and a move backward in the process of secularization. On the one hand, the sacred becomes dispersed, and thus loses some of its potency; but on the other hand, a growing number of groups, organizations and people resort to the sacred to legitimize their claims at being taken seriously, thus putting the sacred to a new use. As a result, uncertainty over the location of social authority grows dramatically: "Demands to be taken seriously (claims to social authority) therefore multiply as uncertainty increases regarding the sacred bases of such claims. Indeed, the sacred becomes so widely dispersed, and the boundary between the sacred and the profane so uncertain that societies develop acute uncertainties about the nature and location of any social authority" (1978, p. 55).

Step five in Fenn's process defines the present situation*. Modern society is characterized by a deep separation between the individual and the collective spheres. In other words, "the boundary between what is personal and what belongs to the requirements of institutional roles becomes another fluid boundary between the sacred and the secular. Corporate goals and values become increasingly secular, while individuals pursue values which they, at least, hold sacred" (1978, p. 64). Fenn takes from James Coleman the idea that, in modern society, two parallel worlds co-exist: the powerful world of the corporate actors, and the increasingly irrelevant world of the individuals. So much so that "the individual, as an individual, is not part of the relationship among corporate actors. He may fill a role in an institution, but it is the role, not the individual, that is part of the corporate world" (1978, p. 66).

But what is, in this general context, the situation of religion? Roughly speaking, the more a religious group accepts this dichotomy, the better it will be adapted to secular society. More precisely, the position of the different religious groups can be ascertained by the combination of two variables: the scope assigned to the sacred, and the degree of integration between corporate and private value-systems. The four types defined by the combination of these two variables can in turn be arranged on a continuum ranging from the least to the most adapted forms of religion in a secular society. The least well adapted type, which is exemplified by mainline Protestantism as well as by traditional Catholicism, combines a wide scope of the sacred with a high degree of integration between corporate and individual values. At the other end of the continuum, the religious forms best adapted to secular society combine a rather limited scope of the sacred with a low degree of integration between these two types of values²⁹⁷. The best examples of this type are magical and occult religions (1978, p. 69). The two intermediate types are represented by sectarian religions like Mormonism (limited scope of sacred/high integration) and evangelical religions (wide scope of sacred/low integration). The general conclusion reached by Fenn as to the fate of institutional religion in secular society is very close to Wilson's: "The process of secularization, then, is compatible with esoteric or occult religious culture. In fact, secularization may foster these forms of religion as a reaction to pervasive routinization and rationalization in social and political organizations" (1978, p. 72).

To conclude, as the reader may have noticed, while Fenn's secularization theory is firmly grounded in exemplars contained in the secularization paradigm, he also puts forward ideas which definitely go beyond this paradigm. Fenn is one of the major carriers of the paradigm, but he may also be regarded as one of its most brilliant critics.