A history of the secularization issue

CHAPTER 5

ON THE METHODOLOGY OF THE HISTORY OF IDEAS
Unfortunately, we are not quite done yet with methodological considerations. Before embarking upon our journey through history, it is necessary to briefly discuss what I take to be the proper way to conduct an investigation in the history of ideas. In effect, the way I have proceeded up to this point raises an important question. By presenting the CISR paradigm first, and then looking backwards in time for its genesis, I run the risk of reducing the problems addressed by past sociologists to those considered legitimate in our own time. The danger is quite real, and I am determined to do all I can to avoid this mistake. To achieve this purpose, there seems to be no better way than to discuss the views of one of the most recent and successful adversaries of this type of reductionism: the historian of political science Quentin Skinner.
Quentin Skinner's intentionalism
Since the publication of Skinner's Meaning and Understanding in the History of Ideas in 1969, a debate has been going on among historians of science, and also among sociologists, as to the proper methodology for the understanding of past texts. This debate has led to the development of a new approach in the history of sociology, which was introduced mainly by Robert Alun Jones (1977; 1983). The main point raised by the proponents of this new approach is that ideas must be understood in the context in which they were produced, and that failure to do so results in what has been dubbed a "presentist" position, that is, a position where ideas situated in the past are illegitimately reduced to present concerns.

As stated above, this debate has far-reaching implications for my own work. How, for instance, must Weber's position on secularization be understood? As is well-known, Weber did not use the term "secularization" very often. In spite of this, he is often considered the father of secularization theory, and many modern sociologists read a secularization theory into his writings. Is this procedure legitimate? This question cannot be answered on the basis of mere expediency. To ground our answer in a sound and systematic methodological perspective, we must, for a moment, step out of the framework provided by the secularization debate itself.

The two opposed views on the way the history of ideas ought to be practiced can best be exemplified by Parsons on the one hand, and by a certain reading of Skinner on the other. In the Structure of Social Action, Parsons seeks to demonstrate that there exists "an 'immanent' development within the body of social theory and knowledge of empirical fact itself" (Parsons 1937, p. 5). To demonstrate this, Parsons chose to analyze four authors who had worked in very different contexts: "It would scarcely be possible to choose four men who had important ideas in common who were less likely to have been influenced in developing this common body of ideas by factors others than the immanent development of the logic of theoretical systems in relation to empirical fact" (1937, p. 14, emphasis in original). Indeed, Parsons explicitly stated that factors such as the "influence of the general 'climate of opinion'" were "irrelevant to the purpose of [his] study" (1937, p. 13). As has been pointed out by Steven Seidman, this approach implies a positivist philosophy of science, which views social science as a cumulative enterprise (Seidman 1983, p. 80).

If I chose to follow the perspective exemplified by Parsons, my investigation would need to be designed to uncover the "immanent logic" underlying the development of "secularization theory", from the first "incomplete attempts" of Weber to the complete, "stable" theoretical system to be found in the CISR paradigm.

The most radical alternative to this approach can be exemplified by a certain way of reading Skinner. According to Seidman, for instance, "the historicist maintains that each generation of scientists faces a unique subject matter and set of problems that reflect a particular sociohistorical constitution of society" (1983, p. 85). Likewise, Joseph Femia presents Skinner as stating that "there are no universal truths or perennial questions [...]. Past thought must be completely dissolved in its precise context [...]; it possesses no capacity for independent life" (Femia 1981, p. 157). In the framework provided by this reading of Skinner's position, my investigation would have to be set up to enable me to show that there is no such thing as a history of "secularization theory", and that the CISR paradigm was created ex nihilo in the 60s. All the previous uses of the term should thus be shown to be unrelated to each other, and the "immanent logic" of the development of secularization theory would crumble into a heap of independent statements. (Of course, this is not really Skinner's position.)

What methodology, then, will I implement in my own research? I do not want to engage in an elaborate discussion of the merits and drawbacks of the two perspectives presented. This would take me too far away from my concerns. What I will do, instead, is simply state explicitly which course of action I intend to follow, without much critical elaboration. The advantages offered by the methodology I am going to apply will not be demonstrated in theory, but in practice, by showing that it allows me to shed more light on the material concerning the history of the idea of secularization.

I will first present my method in very general terms. Subsequently, I will discuss more in detail some of the technical procedures it implies. The first step I will take in my analysis will be to implement a radically historicist approach. It seems obvious to me that, if I want to answer the question "What did Weber say with regard to what we today call secularization?", I must be very careful to understand what he said in his own terms, not in ours. The technical problems involved in this - most notably those which revolve around the question of intentionality - will be taken up at a further stage of this discussion.

As we will see, this position will lead me to the conclusion that Weber's use of the term secularization was purely descriptive, and that whatever theory (if any) he put forth concerning what we today call secularization did not involve a systematic use of this term. Furthermore, I will be led to show that, when he developed his "theories" of rationalization and/or of disenchantment, Weber's concerns were different from the concerns of modern sociologists of religion.

But what next? In spite of all this, Weber's impact on modern sociology of religion cannot be ignored; to take just one example, Berger's secularization theory owes much to Weber's writings. Does this mean that we must, after all, revert to an "immanentist" or "presentist" position, and consider that Weber's attempt was a step "in the right direction", but only a first step? Shall we assume that Weber failed to fully grasp the implications of the idea of secularization, which was somehow "out there"? I think not. I think that we can account for Weber's impact on modern sociology of religion by continuing to use an historicist method. As a second step in the implementation of this method, I will concentrate on Berger's text, and ask two different questions. The first is the same we already posed in relation to Weber's text: "What did Berger say with regard to what we today call secularization?". The second, which will enable us to establish the link between Weber's and Berger's approaches, is: "What did Berger say that Weber had said with regard to what we today call secularization?".

Only this two-fold procedure can enable us to understand the genesis of the secularization paradigm. To say that Berger found the idea of secularization in Weber presupposes neither that Weber had put it into his text, nor that the contemporary reader would have been able to find it there. To reformulate this important point: One the one hand, I will be analyzing the history of the secularization issue in terms of the genesis of a paradigm - that is, by superimposing on the texts I am analyzing a classification and a series of categories that are usually not used in them⁴⁷. But on the other hand, I propose to apply a Skinnerian methodology - that is, a methodology in principle precisely opposed to this kind of "presentist" reading. In other words, Skinner would probably argue that Max Weber, for instance, cannot be considered to have contributed to the modern secularization paradigm, since he naturally knew nothing of it. Does this mean that there an insuperable contradiction in my methodology? I think not. I would admit that there is a tension in it. But I would also argue that this tension, if correctly controlled, should prove to be beneficial to my endeavor: It will compel me to straddle the ridge, always watching my step in order to avoid falling on either side, thus avoiding any simplistic appraisal. To reformulate in less poetic terminology: My aim is to determine just to what extent secularization as it is discussed today is a pure produce of the 60s, and to what extent it is heir to a longer tradition. To do this, I cannot rely exclusively on any single methodology: I am compelled to choose a difficult, uncomfortable approach.

This procedure clearly implies that the text as such does not have a specific meaning. A text can be meaningful only in relation to a given context. Which of these contexts - the present or the past - are to be considered legitimate for the interpretation of the text is a question of point of view or, in other words, of the question the researcher asks. All that needs to be carefully distinguished is the past meaning of a text, and its present meaning.

What consequences does this procedure have with regard to the question of the "immanent logic" of ideas? It seems to me that this procedure remains neutral or, in other words, allows us to answer the question of the immanent quality of a theoretical development on an empirical basis. And the answer may vary from case to case: even if some theoretical developments are not immanent, others may be. By carefully distinguishing between past and present possible meanings, this procedure allows us to assess which exactly were the ideas that were developed through a concrete chain of influence, and which - if any - were developed independently by different writers, that is, through the influence of an immanent logic. Furthermore, by allowing us to take into account the differential impact of different historical contexts, this procedure allows us to take into account a third factor beside influence and immanence, that is, context.

To summarize, the method I will use is based on relativist presuppositions in the sense that the history of an idea is to a large extent considered to be the history of the transformations of the meanings of different texts in the eyes of different readers. It is relativist in the sense that a text can never be said to have a fixed meaning. But the method is also based on non-relativist presuppositions in the sense that, at any given period, the context in which a text was written does determine its meaning if we take this period as a frame of reference. Thus, every reader is trapped in a particular historical setting: even if it is possible for him, through the application of a careful methodology, to familiarize himself with past historical settings, he can never get ahead of his time. Hence, the later historian of ideas has one superiority over his predecessors: he possesses one more context through which he can read a past text.

I must now turn to a more detailed discussion of the implementation of Skinner's method. I will discuss four points: the different pitfalls that must be avoided, the rules that must be followed to establish that a writer influenced another writer, what is needed to "understand an idea in its context", and the more difficult and general question of what it means to recover the meaning of a text.

The two first points are quite straightforward, and I will accept them without any discussion. First, Skinner has presented the different pitfalls which must be avoided as a list of different "mythologies". First, the "mythology of doctrines" corresponds to "the expectation that each classic writer [...] will be found to enunciate some doctrine on each of the topics regarded as constitutive of the subject" (1969, p. 32). This kind of mythology is very present in the discussion on secularization: modern sociologists of religion often turn to Weber with the intent of finding out what exactly Weber's theory was with regard to secularization. A second form of this mythology Skinner mentions (1969, p. 36) is also present in discussions on secularization: Durkheim is sometimes blamed for "not having a secularization theory". The second type of mythology is the "mythology of coherence" (1969, p. 39). Thus, to revert to the example of Weber, some sociologists seem to find it unfortunate that Weber's position with regard to disenchantment or rationalization was not sufficiently clearly stated, and endeavor to "recover" the coherence they think "must somehow be there". The next mythology is the "mythology of prolepsis", which arises "when the historian is more interested - as he may legitimately be - in the retrospective historical significance of a given work or action than in its meaning for the agent himself" (1969, p. 44). This, again, is what happens when Weber is hailed as the "father of secularization theory". The last mythology is the "mythology of parochialism" (1969, p. 45), which arises, for instance, when "the historian may conceptualize an argument in such a way that its alien elements are redescribed into an apparent but misleading familiarity" (1969, p. 47).

Skinner completes this negative advice with some positive rules concerning "the necessary conditions [...] for helping to explain the appearance in any writer B of any doctrine, by invoking the 'influence' of some earlier given writer, A" (1969, p. 46). The conditions, as set forth by Skinner, are: "(a) that there should be a genuine similarity between the doctrines of A and B; (b) that B could not have found the relevant doctrine in any writer other than A; (c) that the probability of a similarity being random should be very low" (1969, p. 46). In fact, it could seem at first sight that this test is most clearly passed in one precise case: when a writer explicitly acknowledges his debt to his predecessor. And this is much more often the case in the heavily institutionalized framework of the modern social sciences than it was in the period Skinner is used to studying. However, this method can be very misleading: a writer can acknowledge one of his predecessors for a number of reasons, some of which may not be intrinsically scientific - thus, there is an implicit rule that, in any good sociological research, one ought to cite Weber - or result from a misunderstanding of the author who is cited.

The next point is designed to prevent any misunderstanding of what is implied in the "contextualist approach". For what Skinner means by context is different from what is usually understood under this term. Indeed, in the 1969 article, the contextualist approach was one of the two positions Skinner criticized (1969, p. 59). He charged that, by ignoring the intentions of the concrete human agent who performed a linguistic act, this approach, even in the cases where it enabled us to explain an act, did not allow us to understand it. In order to understand an act, we must replace the utterance in its linguistic context: "The essential aim, in any attempt to understand the utterances themselves, must be to recover this complex intention on the part of the author. And it follows from this that the appropriate methodology for the history of ideas must be concerned, first of all, to delineate the whole range of communications which could have been conventionally performed on the given occasion by the utterance of the given utterance, and, next, to trace the relations between the given utterance and this wider linguistic context as a means of decoding the central intention of the given writer" (1969, pp. 63-64, emphasis in original).

The last, and much more difficult question concerns the method which must be applied in order to recover the meaning of a text. I will start with Skinner's well-known and very convenient example: "Consider the case of a policeman who sees a skater on a dangerous pond and issues to the skater the following serious utterance: 'The ice over there is very thin'" (Skinner 1972, p. 83). The question Skinner asks in order to understand the meaning of this utterance is: What was the policeman doing in issuing this utterance? As this example shows, the answer to this question cannot be given simply by analyzing the meaning of the utterance as such. The context tells us that, in issuing this utterance, the policeman was probably warning the skater. This act is what, following J. L. Austin (1962), Skinner calls an illocutionary act⁴⁸, that is, not simply the issuing of an utterance, but an action performed in issuing this utterance. To illustrate, consider briefly an example drawn from How to Do Things With Words, the book on which the whole speech act theory on which Skinner relies is founded. In pronouncing the utterance "I name this ship the Queen Elizabeth", I do much more than issue an utterance: I perform the act of naming the ship (Austin 1962, p. 5). To quote Skinner again, "any agent, in issuing any serious utterance, will be doing something as well as merely saying something, and will be doing something in saying what he says, and not merely as a consequence of what is said" (1972, p. 83). Skinner's point is that the meaning of an utterance cannot be understood by itself, if detached from its context, but can be understood only if we recover the intention of the person who issued the utterance.

These considerations led Skinner to conclude that a writer must always be granted a "special authority" over his own intentions (Skinner 1969, p. 48). But, some of his critics countered, this equivalence between meaning and intention is wrong: Surely, in some instances, the meaning of what he says escapes a given agent, or, conversely, a given agent may be unable to make himself understood (Graham 1981, p. 151; Femia 1981, p. 172).

In a later text (1988), Skinner grudgingly accepted this rather obvious caveat - which he had tended to overlook before, although he had mentioned it casually time and again - and set himself to the task of fully developing the consequences of the possibility of unintentional illocutionary force. The complete reconstruction of Skinner's causal chain can be presented as follows. The first cause of a given statement - and hence, the ultimate element in the light of which it should be understood - are the reasons an agent has for possessing a motive (1972, p. 89). If we revert to the example of the policeman, we could say that the policeman's professional culture had stamped on him a sense of duty⁴⁹, which led him to warn the skater. But Skinner does not fully develop this important point. Next in the causal chain are the agent's motives. As already mentioned, the policeman's motive could be professional duty, but it could equally well simply be fear - a genuine, brotherly fear that the skater might drown. Any of these motives (situated at the causal level of analysis, in the positivist tradition) might prompt the policeman to act - that is to say, to issue his utterance, to perform an illocutionary act. But in order to understand this act, we must no longer situate the analysis at the causal level, but at the level of comprehension - in the verstehende tradition: We now understand that the policeman had an intention, which was to warn the skater.

Depending on the situation, the policeman's intention may be conveyed to the skater, or may fail to be conveyed. We can easily imagine a number of reasons for failure. In any case, the conveyed meaning might be different from the intended meaning. This conveyed meaning, Skinnner calls the illocutionary force (1988, pp. 265-66).

Now, the question is, what is it that must be recovered if we want to understand the meaning of an utterance? The illocutionary act, or the illocutionary force? Skinner's answer is the following: "I have tried to distinguish two questions about the meaning and understanding of texts. One is the question of what the text means, the other the question of what its author may have meant. I have argued that, if we are to understand a text, both questions must be answered" (1988, p. 271).

On the basis of the relativist position exposed earlier, I would like to add a further distinction to Skinner's. In reality, we must ask at least three questions: 1) what the author may have meant; 2) what the text may have meant in the original context; 3) what the text may mean in a more recent context.

Which of these questions we are going to answer depends on the original question we are posing. If our intent is to understand the author as a person, on a psycho-analytical level, then surely we must try to discover what he may have meant. But if we want to understand the impact an idea had in its historical context, we will want to understand what it meant for a contemporary reader. And if we want to understand the impact an idea had on subsequent generations, we will be interested in understanding what the text may have meant in this more recent context.

Now, what is the question I am myself asking in this study on the history of the secularization issue? My main interest is in the genesis of the paradigm. I will therefore concentrate on the different meanings a text may have spawned as it was passed from one generation of readers to the next. It might therefore seem as though Skinner's approach could be of no help to me, for to understand this, I probably do not need to recover the original meaning of a text, let alone its author's intentions. But this is not so. For, to understand the genesis of a paradigm, I must be able to understand whence an idea came from. As already mentioned, we can imagine at least three sources: influence, immanence, and historical context. For instance, whenever a modern reader has superimposed a new reading on an old text, we must conclude that he did not develop his idea so much through influence - although he might have gained an intellectual stimulus through this text - as through another factor, which might as well be the immanent theoretical development of the field as the historical context. But, in order to be able to tell when exactly a reader superimposed an alien reading upon a text, we must be able to recover its original meaning. I will therefore have to answer at least the last two questions.

But what about the first question? At first sight, the history of ideas is concerned with ideas that had a certain amount of public currency. Weber's private thoughts - no matter how great a man he was - are certainly of no great direct historical significance. But again, we must refuse this too easy dismissal. The history of ideas is not concerned solely with the transmission and the transformation of ideas, but also with their production. And ideas are produced by individuals. This is not to say that we must fall back on a naive psychologistic position. For individuals can only produce the ideas that are available to them in a given time. Aristotle could never have produced relativity theory - nor, for that matter, Parson's systems theory. To fully understand the genesis of an idea, we must take into consideration the dynamic interplay between the individual author and his, or her historical context. And in order to grasp this dynamic nexus in which new ideas are produced, we must do all that is in our power to recover the author's intentions.

To summarize: In my research, I propose to apply as systematically as possible Skinner's methodology every time I am able to do so. This implies, first, that I will try to recover - beyond the meaning of a text - the author's intention, that is, the illocutionary act he performed. Unfortunately, this will not always be possible. Second, I will do all I can to grasp the meaning a text had in its historical context - that is, the contemporary illocutionary force. Finally, I will also endeavor to understand how this original meaning has been modified through subsequent readings - that is, the anachronistic illocutionary force.

As will become evident, in the following analysis, I am not going to quote Skinner every time I provide an interpretation of a past text. Explicit references to the Skinnerian method will be rare - the matter at hand is sufficiently complex to avoid superimposing to is a scholastically contrived methodology. The only point where I will explicitly invoke Skinner at some length is in the anlysis of a recent period - the "secularization debate" that occurred in the 60s and 70s. But Skinner's methodology will remain present throughout this study as an implicit standard, against which the plausibility of my interpretations can always be tested.