1One of the most striking features of Weber’s writings on religion is the frequency with which he uses the word “rationality” and its derivatives (Rationalisierung, Durchrationalisierung, etc.). It follows from this that the sociological explanation of religious phenomena and religious beliefs, for him, as with an explanation of any social phenomenon, comes from an interpretive method. According to the fundamental postulate of the metatheory on which this interpretive method is based, the cause of an individual’s beliefs coincides with the meaning these beliefs have for him or her. According to Weber, for the sociologist to explain that a certain category of people adheres to a certain belief is to show that these beliefs make sense to these people.
2Weber’s writings on the sociology of religion remain relevant because of their methodological and theoretical framework. Like those of Durkheim, the data on which they draw has aged. Scholars and theorists of Indian society are critical of what Weber wrote concerning Indian castes, particularly concerning the Untouchables (Deliège 1993; Baechler 1988). In addition, the Dead Sea Scrolls provide information on ancient Judaism and early Christianity unavailable to Weber. His writings are also mostly posthumous, and it is not certain that they have come to us in a form that Weber would have considered definitive. However, the scope of the methodological and theoretical lesson they hold for us (both dimensions of scientific activity standing in reciprocal relation to one another) remains intact. It is this lesson, above all, that is worth emphasizing today, in the case of Weber’s writings on the sociology of religion, as in that of Durkheim’s Elementary forms (1912/1965).  This will be my purpose in the following remarks.
Two Main Categories of Theories of Religion
3Although Weber applies the metatheory of understanding to the analysis of religious beliefs in a particularly conscious and systematic manner, he is not the only one to do so. If we take a broad overview of the ways in which religious beliefs are explained by the social sciences, we can discern two main categories of theories. On the one hand, discontinuist theories regard religious beliefs as fundamentally distinct from other kinds of human thought. They explain these beliefs by assuming that the believer is subject to laws of thought distinct from those governing scientific thought. Lévy-Bruhl’s “primitive mentality,” Levi-Strauss’ “savage mind,” and now Shweder’s “magical thinking” or d’Andrade’s “cognitive anthropology” fit into this framework.  Shweder (1991) later abandons the concept of “magical thinking” that he had previously used, but under the label of “cognitivism,” he develops a discontinuist theory of thought, assuming that different cultures use distinctive patterns of thought.
4The “primitive mentality” and the “savage mind” are distinct concepts, certainly, but both postulate the existence of forms of collective thought that depart from the rules governing the modern human being’s ordinary thought process. According to Horton, this category of concepts mainly reflects the attitude—condescending at the turn of the century, deferential since the mid-century—which generally prevails among anthropologists toward non-Western societies and especially societies without writing. Indeed, in the passage from the “primitive mentality” to the “savage mind,” it seems that there is moral rather than scientific progress (Horton 1993).
5These discontinuist theories repeat, in an erudite tone, notions we find in Comte (the Three Stages—theological, metaphysical and positive thought—stripped of their evolutionary dimension), in Pareto (the “non-logical” as distinct from both logic and the “illogical”), and, reaching even further back in time, in Voltaire (1767/1961).
6Weber’s metatheory, by contrast, is continuist: adherence to any belief, religious or scientific as well as juridical, is explained by the fact that the subject has strong reasons to believe it, and that this belief makes sense to him or her. Therefore, it is up to the sociologist to reconstruct (nachbilden) these reasons. This assumes that the observer and observed share the same rules of thought.
7Weber is not alone in adopting this theoretical framework. Tocqueville before him shares the same theoretical and methodological principles. He shows that the macroscopic differences observed between the United States and France in the content and distribution of religious beliefs are “understandable”: the Americans have reasons that the French do not for remaining attached to their traditional religious beliefs (de Tocqueville 1845/2004). Renan (1863/1955), too, shares these principles, especially in his Life of Jesus. It is the same with Durkheim’s Elementary forms of religious life: we cannot explain religious beliefs as illusions; scientific interpretations stand in a continuum with religious interpretations of the world.
8It may be noted incidentally that the Voltaire of Lettres philosophiques also adopts this framework. To him, the ideas of Penn, the Quaker founder of Pennsylvania, seemed perfectly “understandable”: he wishes to impose tolerance, respect for the dignity of others, and the supremacy of being over having as core values; Voltaire (1734/2003, 16) only reproaches Penn with translating them into a “ridiculous” symbolism; Penn “would have secured them [his ideas] respectability in Europe if men were able to respect virtue when it lies beneath a ridiculous exterior.” In his Tombeau du fanatisme, Voltaire, well informed of the scholarship of his own times on the history of religions (Pomeau 1956), attributes to those whom we would now call “intellectuals” the fact that ideas based in Christianity have given rise to speculations that eventually make them unrecognizable: “It was only when they tried to Platonize that they became lost in chimerical ideas [Voltaire has in mind the doctrine of the Trinity]” (Voltaire 1767/1961).
An Open Conception of Rationality
9We must not understand the word “rationality” as Weber uses it in the narrow sense it is now commonly given in the social sciences. Thus, Rational Choice Theory (a term commonly translated into French as “théorie du choix rationnel” (TCR), but more accurately translated as “modèle de l’utilité espérée” or “expected utility model”) conceives of social actors as individuals driven solely by the concern to satisfy their preferences by the means each deems most suitable. Here rationality includes characteristics (on the one hand, egoism; on the other, consequentialism and instrumentalism) which are by no means obligatory ingredients of rationality. Furthermore, as it assumes that altruistic preferences are always a matter of enlightened self-interest,  the theory of rational choice confuses rationality per se with instrumental rationality (or, as we might also say, with consequentialist rationality). This entirely unnecessary reduction has the disadvantage of producing a metaphysical image of the social actor, of which one can only see what justifies it from a scientific point of view. Moreover, not everyone shares this view. Thus, the philosopher Nicholas Rescher (1995) states:“[…] rationality is in its very nature teleological and ends-oriented,” adding immediately that the teleological is not to be confused with the instrumental. Indeed, he continues: “Cognitive rationality is concerned with achieving true beliefs. Evaluative rationality is concerned with making correct evaluation. Practical rationality is concerned with the effective pursuit of appropriate objectives.” Thus, it is obviously not instrumental rationality but the rationality that Rescher calls cognitive that gives scientific work its sense (Boudon 2002). In the same vein, the Weberian distinction between “instrumental rationality [Zweckrationalität]” and “axiological rationality [Wertrationalität]” implies that rationality should not be confused with “instrumental” rationality and that the social sciences must take this distinction into account when they propose to explain the beliefs and feelings of social subjects. For those who are troubled by the notion of a plurality of types of rationality, we can see that all of these types posit that a subject believes X or Y because, like all those who find themselves in the same situation (in the broad sense of the term “situation”), he has reasons he perceives as valid to believe X or Y.
10In the chapter “Die Wirtschaftsethik der Weltreligionen” in his Gesammelte Aufsätze zur Religionssoziologie, Weber speaks of “rationalization” (Rationalisierung) to describe various cognitive processes. These include the search for coherence—and more generally for credibility—in the explanation of phenomena, but also the search for means tailored to ends themselves inspired by basic needs, the simplification of theories proposed to explain natural or human phenomena, the definition of practices and design techniques derived from these theories, or the codification of these techniques and practices (Weber 1922/1988, 266). Religious thought, like any form of thought, seems to him to be subject to these processes of rationalization. In short, the subject seeks to produce credible explanations (or, according to his or her social “role,” to find them in the marketplace of ideas and take them up) regarding phenomena of interest or concern to him (pain, bountiful harvests, the survival and reproduction of cattle, etc.), so as to discover useful lines of conduct. He then subjects these theoretical explanations to a critical gaze, which of course will be better or worse equipped depending on the state of the knowledge at his disposal. Far from obeying special rules of thought or being enslaved to psychological or social mechanisms operating without his knowledge, such as those highly speculative mechanisms postulated by discontinuist theories, the subject obeys the general rules that characterize ordinary thought, scientific thought, and legal thought.
11Weber is not the only one to observe that the techniques employed by magical practices often have an objective basis; Pareto had seen this as well. There is a contemporary illustration of this thesis: the geophysicist Jelle de Boer of Wesleyan University in Connecticut has demonstrated that the Pythia of Delphi sat above a fault corresponding to the intersection of two tectonic plates from which emanated vapors of methane, ethane and ethylene with effectively hallucinatory properties (de Boer 2000).
12The “rational” conception of religious beliefs that Weber defends does not imply that a person endorses them based on some reasoning on his part. He recognizes that, whatever their nature may be, beliefs are normally transmitted through education and generally by socialization. He notes that the new beliefs are imposed through the mediation of “charisma,” the special authority that subjects accord to the innovators who induce them to accept the new theories and beliefs they offer:
Die Deckung allen über die Anforderungen des ökonomischen Alltags hinaus gehenden Bedarfs dagegen ist, je mehr wir historisch zurücksehen, desto mehr, prinzipiell gänzlich heterogen und zwar : charismatisch, fundiert gewesen. Das bedeutet : die natürlichen Leiter in psychischer, physischer, ökonomischer, ethischer, religiöser, politischer Not waren weder angestellte Amtspersonen, noch Inhaber eines als Fachwissen erlernten und gegen Entgelt geübten Berufs im heutigen Sinn dieses Wortes, sondern Träger spezifischer, als übernatürlich (im Sinne von : nicht jedermann zugänglich) gedachter Gaben des Körpers und Geistes. 
14In short, this important passage tells us that the response to human needs that go beyond simple economic needs is not usually given by experts or officials, but by innovators with special qualities of body and mind thought not to be accessible to all, and in this sense, Weber writes, to be “supernatural.”
15Like Durkheim, however, Weber does not believe that charismatic authority (any more than traditional authority) is sufficient to ensure the establishment and survival of beliefs. They also must appear to have a foundation—a proposition which follows immediately from the notion of understanding. We must therefore consider “charisma” as a medium, and observe that this medium is not absent from modern societies, although it is most characteristic of societies that have not yet undergone the process of disenchantment. In this sense, Weber could subscribe to Durkheim’s (1912/1965, 486) key assertion: “The concept which was first held as true because it was collective tends to be no longer collective except on condition of being held as true: we demand its credentials of it before according it our confidence.”
16Magic provides the context for all the sections in Sociology of Religion and Economy and Society devoted to religion. The theory of magic allows us to grasp its basic principles:
The sparks resulting from twirling the wooden sticks are as much a “magical” effect as the rain evoked by the manipulations of the rainmaker. Thus, religious or magical behavior or thinking must not be set apart from the range of everyday purposive conduct, particularly since even the ends of the religious and magical actions are predominantly economic.
18This passage expresses the continuism of thought to which I referred earlier. It tells us that magic is goal oriented: for instance, it aims to draw rain upon the crops. Why these “manipulations” rather than others? Because the shaman and those who call upon his services are convinced that they are effective. And for them, the belief in the efficacy of this method is founded upon a theory.
19But this theory is false, one might object, while that which informs the action of the fire-starter is based on a proven theory. Of course, Weber replies. But we know that rubbing two pieces of wood produces fire because we know that the kinetic energy is transformed into thermal energy. The “primitive,” however, does not know this, so there is every chance that the theory which justifies in his eyes the action of the fire-starter is just as false for us, that is to say, just as “magical” as that which justifies the action of the rainmaker. In short, because of our knowledge, we make a distinction where the “primitive” does not. The action of the fire-starter is “as much a ‘magical’ effect as the rain evoked by the manipulations of the rainmaker” in the sense that the effect is explained in both cases for the “primitive” by a theory postulating the intervention of the “spirits”:
Only we, judging from the standpoint of our modern views of nature, can distinguish objectively in such behavior those attributions of causality which are “correct” from those which are “fallacious,” and then designate the fallacious attributions of causality as irrational, and the corresponding acts as “magic.”
21“Primitives” are interested in rain and fire, as both are essential to their daily activities. It follows logically that they are interested in the means of producing them. This means that they are aware of the fire, for a part of the experience, but also theories by which they explain natural phenomena. This is why the “primitive” always combines, in the course of his agriculture, hunting or fishing activities, means that we would call “technical” and means that we would consider “magical.” But the distinction between the two is the product of our own frames of thought. To put it another way, the “primitive” distinguishes between magic and technique, but he believes both to be indispensable. The Zande who stumbles over a root attributes a double causality to the event, mechanical and magical, says Evans-Pritchard (1937/1968), but in this, says the anthropologist, they are indistinguishable from the modern person who, tripping over a root, attributes the event to mechanical causes while also declaring that he was unlucky. Weber’s reasoning here exactly overlaps that of Durkheim (Boudon 1999).  Magicians are rational: they have theories, and it is on the basis of these theories that they act. They are cognitively rational. They are also instrumentally rational, since magic is aimed at producing an outcome considered useful. The “primitive” can still be called rational in the sense that, within the limits of his knowledge, he submits the theory in question to criticism. If “primitives” believe in the efficacy of rainmaking rituals, says Durkheim, this is to the extent that the rain actually falls more often in periods when they practice them, thus corroborating their confidence. Social psychology has amply demonstrated that “moderns” are no slower than “primitives” to seek verification of their beliefs in “spurious correlations.”
22The case of religion is like that of magic, says Weber, who always avoided the somewhat pointless discussion—which abounds in the classic anthropological literature—of the relationships between religion and magic: is one anterior to the other? If so, which one? Are they complementary or antithetical? Because, for Weber, religion inevitably contains greater or lesser doses of magic. Moreover, religion and magic both pursue well-defined objectives and try to achieve them by means based on theories that the believer finds acceptable. In the first lines of the chapter on the sociology of religion in Economy and Society, Weber (1979, 399) cites Deuteronomy: we follow the precepts of religion “[t]hat it may go well with thee. .. and that thou mayest prolong thy days upon the earth.” The objectives of Deuteronomy’s prescriptions are no different from those pursued, for example, by religious theories of a Chinese character.
23Like magical theories, religious theories are subject to critique by different categories of social actors. Like Durkheim and Evans-Pritchard, Weber posits that, though the theories to which “primitives” subscribe seem irrational, they are just as concerned as the moderns with verifying the validity of the theories on which they base their beliefs.
24Here, incidentally, let us note a point to which we will return. Weber thinks it is easier to grasp the meaning of religious beliefs in their earliest manifestations, before intellectual speculation takes hold of them. This evolutionism doubtless explains why he is more interested in ancient Judaism, early Christianity or ancient Buddhism than in the later forms of these religions. This is also the perspective of Renan and Durkheim.
25From the opening chapter of Economy and Society on religion, Weber (1979) suggests that beliefs that seem strange to us and that we call “irrational” seem to be such because of what Piaget has called “sociocentrism”: we distinguish between the firestarter and the rainmaker because we interpret our observations through our own knowledge (Piaget 1965/1995). To distance ourselves from this, it will suffice to understand that “primitives” would consider both behaviors legitimate because they enact formulas drawn from theories that they believe have a foundation. There is no need to attribute to them a “primitive mentality” or “savage thought,” nor to assume that they are the victims of some “symbolic violence.” We may attribute to them the same rules of inference as ours and simply acknowledge that they have no access to what took Western societies centuries to build. When we distance ourselves from our “sociocentrism,” their beliefs become “understandable” to us; we can explain them by way of acceptable assumptions, referring to cognitive mechanisms readily observable in other contexts (and especially in our own), and not presenting the obscurity and speculative nature of the “primitive mentality,” the “savage mind,” and all of the other “collective concepts” (Kollektivbegriffe) that, for Weber, must be eliminated from sociology, lest it fall into mere verbiage, forfeiting the status of a settled and established science. 
The Soul, Demons, Spirits, and Gods
26Apart from the theory of magic, we can illustrate the theoretical framework of Weber’s sociology of religion via his analysis of animism. As with magical beliefs, it is easy for us to see animistic theories as so foreign to our own modes of thought that we tend to attribute them to causes such as a “savage thought” or a “primitive soul”—in short, to mobilize the Kollektivbegriffe that Weber intends to eliminate.
27This is not the direction Weber takes. The concept of the soul, he says, was designed to explain a variety of immediately observable and important phenomena. First, death leaves the body intact; only the breath of life disappears. This is why breath is treated as the cause and condition of life. The notion of soul also explains the doubling of the personality in dreams or ecstatic states, as well as the phenomena of fainting or loss of self-control. Thus, it explains a variety of crucial phenomena. Given its explanatory power, it is understandable that it should be present in all religions. Durkheim says (1912/1965, 66) almost the same thing in Elementary Forms. He rejects the assumption that “the idea of the soul was first suggested to men by the badly understood spectacle of the double life they ordinarily lead, on the one hand, when awake, on the other, when asleep.” He does not accept, in fact, that the notion of soul is an illusion suggested to the social subject by the phenomenon of dreaming. Like Weber, he wants to trace it to a theory about phenomena that cannot escape our awareness. Here, it is of no importance to us that the content of this theorization is different in the two authors, and that Weber imagines this theorization proceeding from an ensemble of psychic phenomena, while Durkheim envisions a theorization of the individual’s duality as a being of passions, instincts, and desires who at the same time feels subservient to moral values.
28Disease in its infinite varieties as well as a person’s longevity are likewise complex and crucial phenomena. No society has ignored them. To act on these phenomena, it is necessary to theorize them. In one form or another, the notion of “spirit” is normally referred to in order to account for them. It is very close to the notion of “force,” Weber tells us. These “spirits” are in fact conceptualized as forces that can act on the body and that can be counteracted or propitiated by what are considered to be the appropriate formulas according to certain theories. The immediate character of these theories, in societies where scientific medicine is unknown, explains the fact that they are found in many societies, without any of these having needed to borrow them from the others. The notion of a “demon” is a variation of the concept of “spirit.” It refers to those spirits that produce consequences perceived as negative by the subject.
29Again, Weber offers a simple explanation that refers to cognitive mechanisms readily observable in contemporary societies. These mechanisms simply consist in accounting for a certain phenomenon by creating an explanatory principle that is normally interpreted as corresponding to a reality. In other words, it is easily hypostasized. Any physicist knows for example that “centrifugal forces,” so seemingly familiar to us, do not exist. The expression simply refers to the consequences of the principle of inertia. Yet, as the concept of “centrifugal force” allows us to unify so many easily observable phenomena, it is easy for us to believe that it designates a particular kind of force. The notions of “spirit” or “demon” issue from the same cognitive mechanisms. Arguably, cum grano salis, the Kollektivbegriffe I mentioned above issue from the same analyses. Like the concepts of “demon” or “spirit,” those of the “primitive mentality” or “savage thought” describe many easily observable phenomena and attempt to subsume them under an explanatory principle to which a “reality” is naturally attributed. In support of the continuist theory of knowledge proposed by Weber (and Durkheim), it may be observed that the history of science is replete with “explanatory” principles of the same kind, commonly used and interpreted in a realistic manner, abandoned as soon as we can do without them (cf. the “vortices” abandoned as a result of Newton’s formulation of the principle of inertia, the “horror vacui” attributed to nature and ultimately abandoned in the wake of Pascal’s “great experiment,” the “ether” invalidated by the theory of relativity, etc.).
30Animism’s inductive concepts are thus explained by simple and universally observable cognitive mechanisms. They testify not to a discontinuity between “magical thinking” and “scientific thinking” but, on the contrary, to the uniformity of rules of thought and to the continuity between religious or magical thinking, on the one hand, and scientific thought on the other (which obviously does not mean that science and magic are to be confounded). The impression of discontinuity we feel is easily explained: it is the result of our sociocentrism, and the universal tendency to attribute substantial causes to certain categories of phenomena (“centrifugal forces,” the “primitive mentality,” “demons,” etc.).
31A concept such as “force” in physics has given rise to endless debates. At the beginning of the century, the Vienna Circle itself was not clear on what to think of it, and in view of its persistent opacity, occasionally proposed to eliminate it. The case is the same with the very notion of the “soul,” and of “spirits” or “demons.” Once such explanatory principles have been posited, it is indeed necessary to clarify them. Moreover, they give rise, a posteriori, to further theories, forming networks. Thus, Weber remarks, the phenomena of “ecstasy” are theorized as contact with demons or spirits.
32Charisma itself is a variation on ecstasy. It describes individuals with qualities inaccessible to ordinary mortals. Because they are perceived as such, they can under certain circumstances give others the impression that what they say is true simply because they are the ones who say they enjoy these abilities. Charisma is a concept that Weber restores to its place among the tools of scientific sociology, but he borrows it from both St. Paul and the German romantics (Abbruzzese 2001, 298–9). The latter believed that “genius” allows access to higher truths inaccessible to ordinary mortals. We still believe this, on occasion.
33The fuzziness of concepts such as “soul” or “spirit” thus gives rise to a cognitive dynamics. The “gods” are an evolution of the concept of spirit. They frequently lack names and are designated only by the processes they control (Weber 1979, 402). They give rise to rituals. The content of these rituals is dictated by the theorization of the desires, sensibility, and psychology of a particular god that is proposed and accepted. Naturally, the psychology attributed to the gods is drawn from social observation. Scheler states, in a very Weberian analysis, that “Mohammed’s God has the character of a fanatic and sensuous sheik roaming about the desert,” that “Aristotle’s God has the character of a self-sufficient, contemplative Greek sage who is satisfied with his own wisdom,” and that “the Christian God of the Germanic tribes more or less has blue eyes and the temperament of a Germanic duke to whom one belongs by trust and without a contract […] quite different from the Romans’ regimental religious representation of God” (Scheler 1985, 292–3). Rituals are fixed from the moment they have demonstrated their effectiveness: “Every purely magical act that had proved successful in a naturalistic sense was, of course, repeated in the form once established as effective” (Weber 1979, 405).  Weber sees a process of “rationalization” in the substitution of the gods for the spirits: by this, he means that the “gods” allow for a finer and more complex theorization of the world. This is why the gods always appear after the spirits (Weber 1979, 437). Similarly, the concept of inertia entails a more precise concept of motion, and therefore must emerge later. It should also be noted, however, that the evolution of the spirits into gods does not take place everywhere; the circumstances must favor it (Weber 1979, 437). Weber’s evolutionism is in no way mechanistic.  We will return to this point.
34Durkheim, too, notes that the central concepts around which the religious theorization of the world is organized are fuzzy and “inductive,” that this fuzziness is crucial to their efficacy, that it gives rise to a constant movement of theorization in order to clarify them. He also notes that these processes are at work not only in religious thought, but also in ordinary and scientific thought.
The Dimensions of Religious Rationality
35Like ordinary or scientific thought, religious thought is “critical”: it wants to be sure of the solidity of the theories it presents.
36The rationality of religious thought appears first of all in the fact that the believer is “falsificationist.” A god who does not perform the services expected of it disappears (Weber 1979, 427). The peasant is reluctant to accept monotheism, because the unity of inspiration we expect from a unique divine will is hardly compatible with the vagaries of nature with which he is confronted. One theory, according to which phenomena are in fact wills in competition with each other, seems more consistent with what he sees every day. This is why the word “pagan” comes from the exchange of an insult: those who appeared recalcitrant toward monotheism were called paganus. The saints have become an essential part of Catholicism because they allow religious theory to be reconciled with the real in the peasant’s eyes. In Ancient Judaism, Weber (1952, 310) emphasizes that the interpretation of the prophecies appears to be constantly adapted to reality. A prophecy announced an impending disaster; when this was not forthcoming, its predictions were forgotten for some time (Weber 1952, 308). Having identified cognitive mechanisms similar to those evoked by Weber, Renan says that when the hope of Christ’s imminent return was not realized, John no longer interpreted the resurrection literally at the end of his life, in contradiction to Paul’s two Epistles to the Thessalonians (Renan 1863/1955, 18).
37The rationality of religious thought, then, translates into the fact that the believer turns out to be a “verificationist.” Since Popper, verificationism has had a bad name in the philosophy of science: we cannot verify a theory, we can only “falsify” it (Popper uses falsify in the sense of disproving). But this is only true of theories in the form of universal propositions (such as “all swans are white”), not theories in the form of singular propositions. From a scientific point of view, singular propositions are no less important than universal ones (e.g. “a meteorite impact caused the disappearance of the dinosaurs”; “global warming is taking place”). Verificationism also describes current scientific practice. A scientist convinced of the truth of a theory normally looks for confirmation, seeking out peers and seminars in sympathy with this theory. All histories of science demonstrate this spontaneous verificationism. Thus, a careful study of the history of discussions on the language of bees shows that we have long believed in its existence because those who were in favor of this hypothesis not only tended to pay attention solely to evidence confirming it, but also to associate only with colleagues who shared their opinion (Wenner and Patrick H. Wells 1990). Here we find familiar cognitive mechanisms well known to sociologists of science, as well as analysts of electoral choices (which indirectly confirms Weber’s conjecture that ordinary knowledge and scientific knowledge are based on the same procedures).
38The crystallization of religious beliefs involves a similar mechanism. When a prophecy is “verified,” it tends to acquire an authority that is all the more solid to the extent that its fulfillment had seemed unlikely. Furthermore, tradition tends prioritize prophecies that were fulfilled or that can be expected to be fulfilled, emphasizing those that are fulfilled against all odds: “This applied, first, to Amos’ oracles of disaster concerning the then mighty Northern Kingdom, then to Hosea’s oracles of doom for the dynasty of Jehu and for Samaria. […] Above all, the conquest and destruction of Jerusalem confirmed the frightful oracles of disaster of the youthful Isaiah, of Micah and especially of Jeremiah and Ezekiel” (Weber 1952, 308). 
39Renan developed similar ideas. When a prophecy announces an unlikely and desirable event and it is fulfilled, even if only partially, the result is an effect of authority: the prophet is credited with the power of divination. Fulfillment of the prophecies and miracles causes contemporaries to focus on the supernatural character of the prophet’s mission. To which one must add, Renan says—and Weber and Durkheim concur on this point—that “all antiquity, with the exception of the great scientific schools of Greece and their Roman disciples, accepted miracles” Renan 1863/1955, 189). There was no idea of a rational medical science; that is why Jesus practiced exorcism (the expulsion of demons) (Renan 1863/1955, 191–3). In general, we are hard pressed to distinguish metaphor from reality; the inhabitants of the Middle East, particularly in the time of Jesus, had no concept of the laws of nature and consequently no difficulty in accepting the idea that the world obeys God: “There was no supernatural for him, because there was no nature” (Renan 1863/1955, 181–3). In very similar terms, Durkheim emphasizes that the supernatural and miracles are modern concepts that could appear only after the advent of science; the category of the supernatural could have had no meaning, since the notion of “laws of nature” had not been conceived. Weber and Durkheim, however, distance themselves from Comtean positivism much more clearly than does Renan. For them, metaphor, analogy and symbolism do not belong to a bygone stage of human thought; rather, they are constitutive.
The Search for Coherence
40A third essential characteristic of the rationality of religious beliefs is the search by different categories of social actors for an internal coherence in religious interpretations of the world. Like the scientist or the man in the street described by the “theory of cognitive dissonance,” the believer wants the components of the religious explanation of the world to be compatible with one another. We can illustrate this by the example of theodicy.
41Weber returns repeatedly to the subject of theodicy. When the world is conceived as governed by gods who are in competition or conflict with each other, explaining the imperfections of the world poses little problem. The gods each have their own supporters who fight one another in the name of their god, for their god, or under his influence; natural phenomena are seen as subject to contrary influences and serving conflicting interests. Conversely, theodicy becomes a central concern when the world is conceived as subject to a single will.
42In this regard, Weber says, the historical religions proposed a few solutions. Manichean dualism, the doctrine of the transmigration of souls and the theory of predestination are the three main solutions to the imperfection of the world proposed by the historical religions (Weber 1979, 523–5). Zoroastrianism, Buddhism, and Christianity respectively illustrate these three solutions. Perhaps there are more, but it is likely that the number of solutions is finite, Weber suggests.
43According to the dualist solution, God is responsible for good, demons for evil. This was adopted in the very extensive areas over which Zoroastrianism exercised influence, but it also appears outside these areas of direct influence as a heresy or a threat of heresy within religions that do not adopt this solution (Weber 1979, 523–4). Notably, it inspires Christian Gnosticism.
44The second solution to the problem of theodicy is the Indian solution: the injustices suffered by the subject today will be corrected in a future life, so that the world observes a general equilibrium, if it is considered in the totality of its existence. Like the first solution, the second is required by its explanatory power, which allowed it to expand considerably. That is why it too constitutes a threat of heresy or is reproduced in a form furnished by religions that do not recognize karma. Thus, Weber suggests, the idea of the resurrection of the dead probably comes from the Indian East. It is discussed in ancient Judaism: the Pharisees believe in it, while the Sadducees do not (Weber 1952, 390).
45While it is essential to understand that these theories gain a foothold because of their explanatory power, that is, their ability to account for the imperfections of a world governed by God, their emergence and establishment depended on intellectual contingencies of an historical character. Thus, the solution of the transmigration of souls is grafted onto the common belief that spirits of the dead have the ability to enter into natural objects (Weber 1979, 525). Without this idea, the idea of the transmigration of souls would perhaps not have appeared and would not have been imposed so easily.
46Thus, ideas are brought together through the action of a cognitive dynamic. First, the transmigration of souls is grafted on ideas relating to migration of the spirit of the dead. Later, as transmigration achieves a kind of mechanically assured general equilibrium, it results in the erasure of the idea of God from ancient Buddhism.  The notion of soul itself, ceasing to be related to the physical self, loses its importance (Weber 1979, 525). As a logical consequence of its solution to the problem of theodicy, then, ancient Buddhism presents itself as a religion without a god. The mechanical nature of the world order resulting from the Buddhist solution to the problem of theodicy leads to another significant result, namely that the link between religion and ethics tends to diminish.
47The Indian solution to the problem of theodicy appears to Weber to be superior to the dualist solution: the theory of reincarnation is more complex than Manichaeism, which may appear somewhat facile. It is also more abundant in unexpected effects: it severs religion from ethics, giving religion a decidedly ritualistic character, and ultimately eliminates God from the world order.
48The third solution is the one proposed by Calvinism: since God is all-powerful, his decisions cannot be affected by the actions of men. He has made his decisions for all eternity. If they appear sometimes difficult to understand, if the good often suffer in life and the wicked prosper, this is because God’s decrees are unfathomable. The Calvinist solution of the deus absconditus is that which strikes Weber as the most remarkable, perhaps because it is simpler than the Indian solution, less facile than the Manichean solution, and more compatible with the notion of God’s omnipotence.
49There is one point to which Weber does not give much emphasis: the hypothesis (if we can use that word) of the deus absconditus renders the coherence that must be attributed to the divine compatible with the chaos of the world. It deprives polytheism of its main line of resistance against monotheism and eliminates the need for polytheistic elements (e.g. saints) to be grafted onto monotheism.
50It is because it has an intrinsic logical force, if we can permit ourselves the expression, that this third solution appears, like the other two, not only as a component of Calvinism and of the many religious movements that it inspired, but as an element latent in many others. The idea of predestination is found in St. Augustine, as we know, but it is also in ancient Judaism, Weber tells us. This is the meaning of the book of Job: it reflects the omnipotence of the Creator (Weber 1979, 522). It indicates that God justly imposes incomprehensible trials. Why does Job complain? “The damned might well complain about their sinfulness, imposed by predestination, in the same manner as animals might complain because they had not been created human beings, a notion expressly stated in Calvinism” (Weber 1979, 522). One detects a trace of the idea of the unfathomable character of divine decrees in several other passages of the Old Testament. The prophets suggest that it may be the Lord himself who encourages his people to exhibit an obstinacy that at first sight seems fatal for him in human terms (Weber 1952, 310). Isaiah, although he received his mission from Yahweh, knows that he cannot know him: he appears to Isaiah as a blur; he sees only the hem of his robe (Weber 1952, 310). Weber would certainly have noted with interest that, even today, a rabbi has tried to read the Holocaust as a manifestation of the divine will. Calvinism has thus only made central an idea present in ancient Judaism, one that appears very early on because it is a corollary of the notion of God’s omnipotence, and it offers a solution to the problem of theodicy that best preserves the rights of God and ethics.
51As the second solution to the problem of theodicy, the deus absconditus abounds in unintended consequences: since God is too distant from and in fact inaccessible to believers, they can no longer attempt to contact him. Apart from the “virtuosi,” ordinary believers give up hope of approaching God, contenting themselves with diligently and methodically fulfilling their role in the world. If they succeed in their undertakings, they will tend to think that they belong to the cohort of the elect. When God is all-powerful, one can no longer possess him, one can only be his instrument (Weber 1979, 536). With the solution of the deus absconditus, the Indian magical techniques of self-deification must be left behind (Weber 1979, 536–7 and 544). Magic is discredited: no “technique” could influence an almighty God. The two greatest religious forces in history, the Roman Catholic Church of the West and Confucianism, repressed ecstasy, expelled magic, and established ritual techniques which, like prayer, punctuate daily life, Weber tells us. In the West, Calvinism and its most important offshoot, Puritanism, have contributed mightily to consolidating this trend.
52The world’s imperfection raises not only the theoretical problem of theodicy but also the practical problem of the attitude to adopt towards the world. To this question, two answers are possible. Judaism exhibits no rejection of the world: the goal pursued by its religious practices is “long life and happiness on earth.” But one can also respond to the world’s imperfections with ascetic refusal. This response gave birth to monasticism in particular (which of course can also be explained by the desire of the believer to lead an evangelical life when it seems impossible in the world). But when God is too distant for the virtuoso to attempt to approach him, the monk is out of place, and asceticism must be accomplished not through the refusal of the world, but within the world (innerweltliche Askese).
An Evolutionary Conception
53These processes of “rationalization” explain the appearance of historical irreversibilities, including the “disenchantment of the world” which is at the forefront of Weber’s interests. His analysis of disenchantment alone would justify classifying Weber among evolutionists. In fact, in the early passages of chapter 5 of Economy and Society devoted to religion, Weber shows himself to be an evolutionist when he describes the passage from “pre-animist naturalism” to “symbolism.”
54The primitive rips the heart from an enemy’s chest or severs his genitals on the basis of a theory: the strength of the enemy has its seat in these body parts. By assimilating them, he takes on the power of his enemy. Symbolic theories are then substituted for theories arising from what Weber called “pre-animistic naturalism.” Fire is conceived first of all as a god: this is the phase of “pre-animistic naturalism.” Then a more abstract conception appears in which it is not the fire itself that is treated as a god, but in which the existence of a god presiding over the fire is postulated. Authority over the fire is maintained by the worship dedicated to him (Weber 1979, 406–7). Unlike the fire itself, the god persists and is endowed with an identity: we enter into the phase of “symbolism” (here in its animist variant).
55The reason why Weber is not mentioned in histories of evolutionism may lie in the fact that his conception of evolution is especially cautious and nuanced. Thus, it is more satisfying than the “classic” versions of Comte, Spencer, or John Stuart Mill, or perhaps even the modern versions of Robert Bellah (1970, 211–44) or Friedrich Hayek, which, beyond all the differences between them, all appear much more mechanical and “fatalistic” than his. Weber deserves to be ranked among the evolutionists insofar as the identification and explanation of historical irreversibilities is a central concern for him. The history of ideas, including religious ideas, reveals innumerable such irreversibilities, and all the pages comprising the sociology of religion are dedicated to them. But he is by no means fatalistic, as he expressly declares time and time again, particularly at the end of The Protestant Ethic, in the notes accompanying the famous metaphor of the iron cage (Stahlhartes Gehäuse). In no way does he see the modern world as doomed to produce only Fachmenschen ohne Geist—specialists without spirit—and Genussmenschen ohne Herz—sensualists without heart—even if it seems to encourage these types of personality.
56Ultimately, monotheism has contributed most powerfully to the “disenchantment of the world.” By this phrase, which he borrows from Schiller, Weber refers to the discrediting of magic. It is generally accepted that the elimination of magic that characterizes modernity is mainly due to the success of science. This is not untrue, but religion had already laid the groundwork. The establishment, with Judaism, of a single, all-powerful God tended, indeed, to discredit magical practices: magic only functions when the gods are conceived as open to influence, but an all-powerful God cannot be such. Ancient Judaism repeatedly affirms this attribute of God. Only a God more powerful than Pharaoh could bring the Jews out of Egypt. Yahweh hardened Pharaoh’s heart in order to let the believer witness the extent of his own power (Weber 1952, 341).
57The idea of one God Almighty still appears alongside the notion of providence, which is also inconsistent with magic, and has many other consequences, namely that if we cannot influence God, we can try to please him, through prayer and obedience to the divine law. Religion, then, promotes the development of ethics, which now takes precedence over rituals. Again, Renan anticipates Weber.  “I am full of the burnt-offerings of rams,” Yahweh declares through the voice of Isaiah, for it is obedience to the law, not devotion to ritual, which pleases God (Renan 1863/1955, 89). The substitution of ethics for ritualism, according to Renan, is the characteristic that would ensure Judaism’s impact on the Western world: “The millenarianism gave the impulse, the moralist insured the future” (Renan 1863/1955, 112). This is very close to anticipating Weber: charisma is the medium through which ideas pass, while the power of the ideas is what commands respect for them. This domination of ethics over ritual and social conventions was reaffirmed, according to Renan, by early Christianity. Jesus consents to dine with publicans, members of a profession abhorred by the Jews as a token of Roman domination: by this gesture, he means to emphasize the distinction between the function and the individual, and thus to affirm the dignity of humanity as such (Renan 1863/1955, 168–169). Weber will make a reading similar to the Epistle to the Galatians: in his polemic against Peter, Paul recommends commensality, which expresses the equality of all in dignity. Of course, the differences between Renan and Weber cannot be ignored. Among many points of divergence: for the former, it is primitive Christianity that echoes the universal aspects of Judaism’s message, whereas for the latter it is Puritanism (Weber 1979, 615–6).
A Complex Evolutionism: Symbolism or Realism?
58Disenchantment, itself resulting largely from the evolution of religion, invalidated magic and set science in its place, but it has not eliminated magic. Charisma, prophecy, symbolic interpretations of the world understood in a literal manner, and more generally, theories of the world interpreted without distantiation, have not disappeared with disenchantment. New myths are produced, notably by “proletaroid intellectuals.”
59The persistence of these archaisms results from general cognitive mechanisms, in the first place. Human thought and psychology generally include invariants, according to the metatheory on which the interpretive method is based. One of the most important of them is that it is difficult to identify the symbolic; as such we tend to interpret in a realistic fashion. It is true that, even among those who have been socialized in modern scientific culture, it is difficult to escape this shift. We must be convinced that centrifugal forces do not exist. It is difficult to resist hypostasizing “chance” or to abandon the notion of “luck.” It is difficult to admit that a flip of the coin can give the result “tails twenty times in a row” with the same probability as any “disorderly” series of results, because we want the “orderly” series to be not the product of randomness but of some organizing power. We believe in the existence of the “savage mind” or “primitive mentality”; we attribute an explanatory force to these concepts rather than reading them as mere labels identifying issues to be addressed and mechanisms to be discovered. Animism itself remains present in modern societies. Whoever lets himself be injected with “fresh cells” for a supposed rejuvenation effect—to use an example that I obviously do not borrow from Weber—demonstrates a cognitive mechanism that is indistinguishable from that obeyed by the warrior who eats his enemy’s heart.
60Tocqueville (1845/2004) had likewise seen the importance of the distinction between realism and symbolism. In a remark that anticipates Weber, he quietly suggests that the transmigration of souls and the immortality of the soul are symbolic expressions of the same reality and are thus interchangeable, such that we interpret these notions symbolically.  This interpretation, however, does not enjoy a wide audience:  indeed, since symbolism draws its force from its graphic nature, it is easy to take at face value, which most people do. The result is that they do not understand symbolism for what it is, but rather interpret it in a realist manner.
61Durkheim insists that the concepts of “soul,” “mana,” and “oranda” are many different symbols referring to the same reality. Weber, too, proposes reconciling the notions of “charisma,” of soul and spirit, with the concepts of “mana” and “oranda” (Weber 1952, 140). But these equivalences, while they are perfectly valid, are hardly persuasive to those who take these concepts seriously. They only have a chance of being accepted by persons with sufficient distance from the concepts to perceive their symbolic character.
62The difficulty of grasping the symbolic as symbolic, its typical transposition into the realistic mode, is a fundamental datum for the sociologist. It goes a long way toward explaining the misunderstanding and conflicts that tend to oppose symbolic interpretations of the world, and it explains the persistence of magic and prophecy in the modern world.
63Weberian evolutionism is complex, first of all because it is sensitive to the fragility of the distinction between the symbolic and the realistic modes of interpretation of religious categories. The social subject demonstrates a particular receptivity to the symbolic, but may have difficulty gaining sufficient distance to apprehend it as such. Likewise, in Elementary Forms, Durkheim insists that both religious and scientific theorizations of the world initially take on symbolic forms, and that symbols are only deciphered, if at all, after a rather lengthy process.
64The Weberian conception of evolution can be called complex for a second reason: in his work on the sociology of religion, as elsewhere, Weber gives full consideration to the importance of unexpected effects. We have already identified several of these. A conception that makes God an all-powerful master necessarily tends to discredit magic; the theory of the transmigration of souls tends to eliminate God and promotes ritual at the expense of ethics; a deus absconditus tends to elicit a valorization of the world, and, when it appears in a context where it has been devalued, to convert otherworldly asceticism into “worldly asceticism.” A singular God tends to devalue ritualism in favor of ethics.
65The search for coherence that characterizes the religious theorization of the world therefore leads to a number of diverse, tangled, and unintended cognitive and social effects, which unfold over a long period. The concept of the afterlife is latent in that of the realm of souls: from the moment we attribute to the souls of the dead a home of their own, it is difficult to locate it here, in the visible world, yet the idea of an afterlife does not take on real substance until a God establishes a connection between ethics and religion (Weber 1979, 520). It is only when ethics rises to the rank of an essential dimension of religion that the problem of theodicy arises. We can, then, imagine a system in which the injustices of this world would be mended in the world beyond. Yet it is only later, after additional developments, that the separation of heaven and hell appears (Weber 1979, 522). This “gradualism” is one of the remarkable points of the Weberian theory of innovation and evolution. 
A Nonlinear Evolutionism
The Importance of Innovations
66Moreover, according to Weber, religious and social evolution is largely dependent on genuine innovations, which cause discontinuities. The innovators are in religious affairs, depending on the case, priests, prophets or reformers. Circumstances favor one or the other, so these can partially determine the form or content of the innovation. Ezekiel is uninterested in social reform, for he is a priest and “scarcely” a prophet at all (Weber 1979, 443–4). But the content and impact of innovations are largely unpredictable.
67However, innovators can play a significant role, provoking remarkable historical turns. Saint Paul raises “a stout fence against all intrusions of Greek, especially Gnostic, intellectualism” and makes the “the sacred book of the Jews into one of the sacred books of the Christians” (Weber 1979, 622). In advocating commensality against Peter, who is unsure whether one may lawfully sit in the company of uncircumcised proselytes, he does no less than “announce the birth of citizenship in the West” (“die Konzeptionsstunde des ‘Bürgertums’ im Occident”) (1922/1988, 40). Weber mentions several times the verses of the Epistle of Paul to the Galatians recounting the episode inaugurating this principle: “[T]he first great turning point in the history of Christianity was the communal feast arranged at Antioch between Peter and the uncircumcised proselytes, to which Paul, in his polemic against Peter [Galatians 2:11–14], attributed such decisive importance” (Weber 1979, 435).  This principle had to permeate the history of the West, Weber suggests: it affirms the value of the human being as such and assigns to the organization of the city the ultimate goal of ensuring respect for the dignity of all.
68The evolution of Judaism to Christianity, however, is not linear. Rather, it presents a profound discontinuity. In no way does Judaism argue for the rejection of the world. By contrast, Christianity celebrates the refusal of the world: “Jesus was not at all interested in social reform as such” (Weber 1979, 444). But this dramatic shift was not at all necessary. Calvinism will take up once again the universal teachings of Judaism, but it will retain the Christian refusal of the world (Weber 1979, 615–6). It is an unintended effect of Calvinist theology that the refusal of the world will result in asceticism, not in an otherworldly manner, as in Catholicism, but in a worldly manner. The former is exemplified by monasticism (the Catholic monk withdraws from the world), the latter by the faithful Puritan (who refrains from tasting worldly pleasures).
69It may be noted here that the controversies raised by The Protestant Ethic so focused attention on this work that commentaries often fail to consider it within the wider context of the Sociology of Religion and Economics and Society and fail to see that worldly asceticism is the result of a complex cognitive dynamics: Calvinism takes the idea of the worthlessness of the world from early Christianity (“My kingdom is not of this world”), but since the solution it gives to the problem of theodicy (that of the deus absconditus) denies the believer access to God, its asceticism can only be expressed within the world. 
70At the same time, innovation must satisfy the need for coherence. The emergence of a single all-powerful God entailed a refusal of theogony, since such a God cannot come from another God; at the same time it is pregnant with theology, as a single God evokes the problem of theodicy. This is why Greek polytheism had a theogony and not a theology, while Judaism, conversely, had a theology and not a theogony. If Christian theology, for its part, as Weber points out, underwent an historically unprecedented expansion, this is partly because of problems raised by the notion of a son of an all-powerful God.
71The success of any religious movement depends ultimately upon a multiplicity of factors: the content of its doctrines, the efficacy of the cognitive dynamism they engender, their agreement with circumstances, and a multitude of other considerations affecting its capacity for penetration.
72A religious movement also depends on factors that initially seem secondary, such as the form of religious messages: the literary quality of the texts chosen by tradition and grouped in the Old and New Testaments, the use of the parable and the analogy in these texts offer an explanation of the world order with “understandability […] for everyone.” Ancient Judaism and early Christianity produced sacred texts with “understandability for everybody, even for a child,” said Weber (1952, 397). These are clearly much easier to disseminate. The “popular” character of the texts produced by ancient Judaism and early Christianity was itself made possible because they appeared in the sociopolitical context of Israel. The ease of access they show contrasts with the esoteric religions established in imperial societies, whether of Egypt, Assyria and Persia. Weber speaks of a “monotheistic […] transition” in reference to Akhenaton’s attempt to establish a solar monotheism on the basis of an assimilation between the monarch and the sun (Weber 1979, 419). From Weber’s perspective, this transition probably failed due to the esotericism of the Egyptian religion.
73The form of religious messages is in no way secondary. The parables are effective because a message is more accessible when it is expressed in the form of images than in an abstract form. Let it be noted: a key point of Weber’s theory of knowledge, as well as of Durkheim’s, affirms the efficacy of the symbolic theorization of the world and the difficulty the public usually encounters in understanding the symbolic as symbolic.
The Importance of Contingencies
74The non-linearity of religious evolution (and of the social change that it powerfully impels) is again due to the fact that it is dependent on contingencies and unforeseeable encounters between independent series. The God of Israel is originally the patron deity of a small nation, possibly imported from Phoenicia. But once he is bound to Israel’s particular fate, Yahweh cannot avoid concerning himself with Assyria and Egypt, such that he eventually presides over the destiny of empires and takes on the stature of a universal God, ending up not only involved in international politics but presiding over the very course of history (Weber 1979, 418–9). The conceptual construction of Yahweh is completed as a result of the Jews’ awe with regard to Egypt: only a divine king with a stature more imposing than Pharaoh could bring them out of Egypt (Weber 1979, 450).
75Another crucial example of contingency: world-rejecting stances may be favored when a society undergoes an evolution such that the religious ideals to which the believer subscribes appear to them to be violated (Weber 1979, 578–9).
76The reception of religious messages, innovations, and theories further depends on the social composition of the audience they are addressing. However, it is not a matter of arguing that these sociological variables are decisive. Rather, they should be considered as a few variables among others. Here as elsewhere, Weber’s sociology sets itself apart from all sociologism: it is not a matter of turning the variables of social stratification into a primum movens, and it is essential—the interpretive method demands it—to reconstruct the mediations that permit an affirmation of a causal relationship between social positions and adherence to a particular beliefs.
77Peasants have strong reasons not to accept a strict monotheism. This is why popular Christianity places great emphasis on the worship of saints. In contrast, the Roman centurions who fight on behalf of an emperor reigning over a centralized and hierarchical empire readily accept monotheism in its Mithraic and Christian versions. Their affinity for Mithraism comes from the bureaucratic organization of the cult, which reminds them of the organization of the Roman Empire (Weber 1979, 475–6). The aristocratic Sadducees did not believe in the resurrection of the dead. Above all, the upper classes ask religion to legitimize their position (Weber 1979, 491). They tend only to be interested in religious salvation when they have “lost […] the possibilities of political activity” (Weber 1979, 503). This case aside, it is mainly among the middle and lower classes that the idea of compensation in the afterlife seems attractive.
78Here, Weber encounters the theory of Nietzsche, who attributes the success of Christianity among the lower classes to ressentiment. This is speculative and can in any case have only a rather limited scope (Weber 1979, 494–9). There is no question of denying the existence of ressentiment. Ressentiment against the arrogance of the great effectively explains why the Messiah should have returned to Jerusalem perched on a donkey, the “workhorse of the poor” ((Weber 1979, 510). But we can assume that the poor condition of the lower classes simply renders them receptive to theories promising something better, without having to assume that they are filled with ressentiment. Furthermore, it is unacceptable to assign the cause of the complex theoretical constructs that are religious doctrines to an emotional datum (Weber 1922/1988, 240ff). Certainly the promises of Christianity interested the lower classes. However, its influence also comes from its responding to an intellectual need to explain the world. Therefore, we cannot make the conditions of existence into the primary condition for the expansion of a religious system (Weber 1979, 506).
79It should be added that Weber opposes Nietzsche not only on his interpretation of the origins of Christianity, indeed rather summary compared to his own, but also on other key points: thus, unlike Weber, Nietzsche sees monotheism as a regression, and Weber relies on a rational psychology, whereas Nietzsche heralded depth psychology, etc. These massive differences between the two authors did not deter Nietzschean readings of Weber by conservative authors such as Leo Strauss (1953), Carl Schmitt (1923/1984), or Eric Voegelin (1952), or by proponents of postmodernism. 
80Weber’s objections to Nietzsche are also addressed to Marx, who underestimates the metaphysical need of human beings to give the cosmos meaning (Weber 1922/1988, 240ff). This need is so powerful as to explain the extraordinary influence of intellectualism (Weber 1979, 506). There is, therefore, no question of accepting the principles of historical materialism.
81The theme of intellectualism in Weber deserves an exposition that I cannot undertake here except to identify a few key points. Intellectualism is grafted onto a basic need (to make sense of the cosmos). It is a source of progress, but can also contribute to the degeneration of positive ideas. The strength of intellectualism and the direction in which it advances depend on the circumstances. Again, Weber is not far from Renan. Like Voltaire, he points to the ambiguous effects of Greek intellectualism on the evolution of Christianity.
82Finally, it is a complex web of cyclical and structural factors that determines the audience of a religion. Presenting himself as the savior, Jesus uses an idea that had reappeared at more or less regular intervals since Deutero-Isaiah had introduced a mysterious “Servant of God” (Weber 1952, 458n7). He also incorporates the Jewish idea of compensatory retribution, a backdrop for the idea of the resurrection of the dead. Thus, artisans and the lower middle class tend, by their social position and nature of their activities, to be particularly receptive to the idea of compensation, and consequently, to symbols conveying this idea. This complex combination of causes particularly explains why the Pharisees, who mostly belong to these social strata, accept resurrection more easily than the Sadducees, who are found rather among the aristocratic classes of Jewish society.
The Mille-feuille of Religious Evolution
83The fact that within the system of beliefs comprising a religious tradition, such varied ancient elements can coexist, is in the end due to the diversity of publics, the diversity of social positions, resources, and cognitive postures of these publics, and to a good many other factors. One such factor is the cognitive mechanism that gives the social subject a special receptivity to the symbolic, even if the subject is hard pressed to identify it as such.
84In principle, prayer excludes magic, since it is addressed to an almighty God, whom one can seek to please, but whom one cannot influence. However, it is easily seen, especially among the lower classes, as carrying magical effects: i.e., as having the function or at least the effect of promoting the objectives pursued by the believer. Catholicism is opposed to magic, deeply inconsistent with its representation of God, yet the saints have magical virtues, and the celibacy of priests has a magical origin. Indeed, from the earliest times, asceticism has been considered a means of fostering the emergence of charisma, of the ability to contact the higher powers, and chastity was conceived as a way to get closer to God (Weber 1952, 603). The idea of an omnipotent God excludes theogony. Yet Christianity introduced a son of God.
85Modernity has no more eliminated prophecy than it has magic. It has not eliminated the phenomenon of escaping from the world. The latter attitude is even “characteristic of intellectuals,” as disenchantment has produced in them an “inner need.” This distress can be resolved, as in Rousseau, by “escape into absolute loneliness,” by a romantic escape to an idealized “people” supposed to be free from the pollution afflicting the other classes, by a flight into contemplation, or even by escaping through an act aimed at the ethical or revolutionary transformation of the world as it is (Weber 1952, 506). “Agents of proletaroid intellectualism,” whose theories have the function of legitimizing movements of a demagogic inspiration, are also descendants of the prophets, although, due to “disenchantment,” they are merely pale reflections of them. They nourish the growth of ideologies and have a limited influence, one might add, unless they become ideocrats, as was the case following the Bolshevik or Nazi seizures of power. In alluding to these agents of “proletaroid intellectualism,” Weber has in mind the Russian populists, and probably also the intellectuals of 1848, to whom Flaubert’s Sentimental Education and Tocqueville’s Memoirs devote so many incisive pages. Or he may have been thinking of Lukács, whom he knew personally and whose theories, draped in the mantle of the philosophical tradition, especially expressed his rejection of the world and the promise of a future world (Bell 1988). Perhaps Weber would also have seen representatives of “proletaroid intellectualism” in the intellectuals who yesterday sang of a radiant future or who today preach the rejection of “globalization.”
86Another unexpected effect has favored a stratified evolution of adherents of religions, bringing together elements from quite different ages. The development of the sciences has largely deprived religion of its authority over the explanation of the world. It has also encouraged a revival of mystical forms of thought and religious practice. We might add that it has also favored the emergence of a religiosity characterized by a marked dominance of the moral dimension over the dogmatic dimension, as in the case of American Protestantism, and the popularity of “religions for departing from religion,” such as Buddhism, within today’s Western societies.
Evolution’s Lines of Force
87Despite the complexity of Western religious evolution, we can discern some primary tendencies: disenchantment, the discrediting of magic, and the substitution of ethics for rituals. Yahweh is unmoved by offerings; Calvinism deals the final blow to magic (Weber 1952, 217; 1979, 574). Astrological determinism is marginalized by monotheism as incompatible with the notion of a world governed by God (Weber 1952, 396). Judaism establishes the notion of responsibility. Deuteronomy breaks with collective responsibility and affirms the principle of individual responsibility (Weber 1952, 316). According to Isaiah 3:10, God rewards each according to his merits; earthly happiness is for all to enjoy the “fruit of their doings” (Weber 1952, 284). The ethical notion of just compensation is opposed to priestly ritualism (Weber 1952, 255). The responsibility of each is conceived as going hand in hand with the dignity of the individual, regardless of station; in Deuteronomy 10:17, God “regardeth not persons” (Weber 1952, 255).
88The paradox of an imperfect world and a perfect God has encouraged an organization of the city calculated to promote respect for the dignity of the human being as such. The pursuit of this goal has been facilitated in societies that today we might call “multicultural.” It began to develop among pacified groups, who were ruled by natural laws, trust, the rationalization of social relations, and the consolidation of ethical ties (Weber 1979, 430).
89Calvinism, in distancing the believer from God, facilitated humanity’s intellectual mastery of the world, and created an even greater distance from God by this other indirect effect (Weber 1979, 548). As we know, Merton (1937/1970) devoted a study to probing this Weberian theme. In the field of ethics, Calvinism takes up the universal elements of Jewish law (Weber 1979, 615–6).
90It is ultimately through Judaism and Christianity—early Christianity and Puritanism having a decisive influence here—that the values of individual responsibility, respect for the dignity of the other, and the progressive organization of the city toward taking the interests of all into account were irreversibly implanted. There is no irreversibility of ideas without the selection of ideas, and there can be no selection of ideas unless ideas possess an intrinsic force.
The Metatheory of Understanding: An Effective Theoretical Framework
91From a theoretical and methodological standpoint, what has most interested me here is firstly the sobriety of Weber’s psychology. We detect no influence of depth psychology here, nor any trace of causal psychology (which claims to impute beliefs exclusively from non-rational causes—biological, emotional, social, etc.). Instead, Weber deliberately utilized a “rational” psychology: a psychology based on “classical rationalist categories of volition, will, and individual consciousness” (Nisbet 1966, 110). Lazarsfeld, who, without himself ever using psychoanalysis in his own work, had grown up in Vienna, was intrigued by what he perceived as Weber’s total indifference towards Freud (Lazarsfeld 1993).  Weber’s disdain for psychoanalysis in particular and depth psychology in general, however, is not very difficult to explain. It follows that rational psychology is the only one compatible with the metatheory underlying his “comprehensive sociology.”
92Refusing depth psychology and causalist psychology of any kind, Weber falls entirely outside Nisbet’s critique of the social sciences. According to Nisbet, many sociologists thought they could explain the beliefs of the social subject via unconscious psychological mechanisms of a purely conjectural kind which they thought to be sufficiently legitimized, particularly by the authority of psychoanalysis.  Thereby, we might add, they only demonstrated that “traditional” and “charismatic” authority is widespread not only in religious thought, but also among intellectuals. In other words, Nisbet suggests, as Weber had understood, that sociology and social thought more generally cannot without danger eliminate the criterion of understanding and thereby reify the mental processes responsible for individual beliefs and actions.
93Weber also escapes the pseudo-explanations proposed by the holistic tendency within the sociological tradition (“savage thought,” the “primitive mentality”). His analysis of religious beliefs is persuasive because it contents itself with psychological assumptions that are uncontroversial and attentive to the facts. The warrior consumes the heart of his enemy because he believes he will draw strength from it, just as the modern person takes an injection of fresh cells because he believes it to be an elixir of youth. The warrior follows a theory that we have strong reasons to find false but that he has strong reasons to find true. The theoretical framework for comprehensive sociology appears here with perfect efficiency. The cause of collective beliefs is good in general, on the grounds that the individual placed in a particular context is to believe, and hence in the sense that they have for him.
94Weber also marks, in advance, some decisive points in relation to modern “cognitivist” theories such as those of Needham, Shweder, or d’Andrade, which represent original and refined variants of the holistic tradition: they see different cultures as carriers of specific cognitive systems and schemas (Needham 1972). The theory of understanding suggests, on the contrary, that we can understand any belief, however strange, because it engages universal cognitive mechanisms. For the human mind obeys the same rules among “primitives” and moderns. Everywhere, it meets a number of universal needs (the desire to survive, the desire to understand the world, etc.). Scientific thought, ordinary thought, and religious thought do not rely on different cognitive mechanisms. All that varies from one culture to another, and more generally from one context to another, is what we can agree to call “contextual parameters.” Thus, the “primitive” does not know the laws of energy transformation, but we do. For this reason, we consider the behavior of the firestarter and that of the rainmaker to be of a different nature, while the “primitive” sees them as having the same nature. This misunderstanding ceases when it takes into account the difference in contextual parameters. There is no need to assume that primitive psychology is different from ours, but it is crucial to observe that primitive knowledge is different.
95It must be remembered that Weber is not the only one to adopt this framework: Benjamin Constant, Tocqueville, and Durkheim (if we agree to read the Elementary Forms through a lens other than the neo-Durkheimian kind) also analyze religious beliefs as founded on valid reasons in the subject’s mind. Durkheim’s theory of magic is indistinguishable from Weber’s. His theory of the soul is different, but shares the same inspiration. Closer to us, Evans-Pritchard and, among contemporaries, Robin Horton employ the same framework (Horton 1982 and 1993). It is responsible for the strength of their analysis.
96The cause of adherence to religious beliefs is thus to be found among the reasons that a social subject placed in a particular context has to endorse them. Yet we must see that these reasons belong to a very open range. It is not a question, here, of limiting oneself, as in the case of the proponents of the “rational choice model,” to the Benthamite tradition, nor, as with many “postmodernists,” to the Nietzschean tradition. An affective datum such as ressentiment can help to establish a belief. But above all, religious messages must offer persuasive explanations and useful prescriptions. This is why Weber insists so firmly on the cognitive dimension of rationality: How could a peasant faced with the vagaries of nature easily accept the idea that the universe is the result of a single will? How could an omnipotent God have a need for rituals?
97The bulk of Newton’s work is, as we know, in the field of alchemy. Was he a victim of the “savage mind,” demonstrating a “primitive mentality,” subjected to “symbolic violence” exercised by priests or the tune called by an underground conductor when he posed problems of chemistry, only to become rational when he posed physical problems? This janus bifrons aspect of Newton has long been seen as a mystery, particularly since his case is not isolated—far from it (Roger 1963/1993). This sense of mystery is the effect of our sociocentrism: the alchemist Newton seems “irrational” because we have a sufficient dose of chemistry to measure the distance separating alchemy from modern chemistry.
98Weber’s writings on the sociology of religion may seem to offer a comparative history of religions, one that is well informed but that risks looking a little cavalier to a scholar’s eyes. In any case, this seems to be how Weber is read by some commentators who limit his originality to the creation of a few concepts (“charisma,” “religious virtuosi,” etc.) or to the famous thesis of The Protestant Ethic, or who are content to see Weber as a relativistic and vaguely pessimistic philosopher of history. But of what importance to us are mere opinions and moods concerning historical change, even those of Max Weber? The singular force and precision of his writings comes from the fact that they propose to read the history of religions using an unexpected key, that of rationality, and that this gamble has been successful. This key is highly visible: the word “rationality” and its variants are scattered over hundreds of pages of his Sociology of religion and the rich chapter on religion in Economy and Society. This key cannot be used by those who want to reduce all rationality to instrumental rationality. When we see instead that rationality can also take a cognitive or evaluative form, we understand how Weber was able to achieve this feat: describing the evolution of religious ideas by postulating that it is guided by mechanisms the principles of which are not fundamentally different from those that explain the evolution of legal, economic or scientific ideas. 
99Finally, his approach renders moot in advance several nagging questions that haunt the sociological, anthropological, and philosophical literatures, such as the relationship between magic and religion, and it eliminates the false antinomies that speculative thought so easily produces if it tries to pigeonhole the theory of beliefs (labeled “intellectualism,” “expressivism,” “symbolism,” “emotivism,” “rationalism,” etc.). Any of Weber’s concrete analyses (for example, that which deals with the difference in attitude of the Pharisees and Sadducees with regard to the resurrection of the dead) suffices to demonstrate that we must conceive of collective beliefs as resulting from a system of causes and reasons that are closely articulated with one another. In something like the way that a system of equations must include parameters and variables, any individual or collective belief has a rational dimension and a causal dimension, as well as a “cold” side and a “hot” side. The notion of compensation plays an essential role in the minds of the Pharisees, but not in that of the Sadducees, as a result of their respective social roles. The former believe in the resurrection of the dead because it allows them to reconcile their notion of divine justice with the injustices in terms of compensation that they observe around them; the reasons for their beliefs are rooted in affects and give rise to emotions. That said, the core of the analysis is well established, in accordance with the assumption of understanding, by noting the reasons why the former agree with what the latter reject. To avoid any misunderstanding, it should also be noted that these reasons are often metaconscious. According to Weber, action is most often semi-conscious or unconscious of the reasons (and, equally, of the irrational motivations) that constitute its meaning.  This unconscious, which has nothing to do with Freud’s, is that which can be observed in the mechanical or “instinctive” behaviors characteristic of everyday life. 
In Boudon (1999), I presented an interpretation of the Forms that focused on Durkheim’s persistent refusal to treat religious beliefs as illusions, as well as on his assumption of the continuity between science and religion, and on his idea that religious theories are representations of the world to which believers subscribe because they make sense to them in their own context. He proposes that we see in the sense of the sacred a notion very close to what we call a sense of values. Interpreted literally, Durkheim’s phrase, as well-known as it is unfortunate, according to which homo religiosus unconsciously worships his own society, would appear to be incompatible with the basic assumptions that animate all the analyses of the Elementary forms: this term would simply mean that people are moved by values that are clearly not of their own making but the product of complex processes of interaction, like science or language, that they feel a sense of respect for them, and that they are moved when these values appear to them to be violated.
See Lévy-Bruhl (1938; 1966a; 1966b; 1975), Lévi-Strauss (1966), Shweder (1977), and d’Andrade (1995).
The assumption of egoism is as essential to Rational Choice Theory as the parallel postulate in Euclidean geometry. To those who argue that this deprives RCT of generality, its supporters reply that the assumption of egoism does not preclude any explanation of seemingly altruistic behavior or preference, but that it must be analyzed as arising from enlightened self-interest. RCT is grafted onto the utilitarian tradition, which, like Nietzschean or Marxian traditions, owes part of its success to the fact that it is perceived as a healthy denunciation of the Pharisaism lurking behind altruism by an always apparent principle.
“The provisioning of all demands that go beyond those of everyday routine has had, in principle, an entirely heterogeneous, namely, a charismatic, foundation; the further back we look in history, the more we find this to be the case. This means that the ‘natural’ leaders—in times of psychic, physical, economic, ethical, religious, political distress—have been neither officeholders nor incumbents of an ‘occupation’ in the present sense of the word, that is, men who have acquired expert knowledge and who serve for remuneration. The natural leaders in distress have been holders of specific gifts of the body and spirit; and these gifts have been believed to be supernatural, not accessible to everybody” (Weber 2009, 245).
Obviously, I do not suggest that the theories of Weber and Durkheim can be superimposed on one another: just that both are based on the principle that believers believe what they believe because their belief makes sense to them, because it appears to them to be well-founded. It is this shared assumption that explains the convergence of their views on important issues, such as the explanation of magical beliefs. The convergence is obscured by the abstract turn that Durkheim gives to his analysis, while Weber’s theorization sticks much closer to the facts: a difference that can be traced to the fact that Durkheim’s milieu is that of the philosophers, while Weber is closer to lawyers, historians, and economists.
Letter from Weber to Robert Liefmann, March 9, 1920, quoted in Mommsen (1965).
“[S]uccessful in a naturalistic sense”: appearing to carry real effects.
Nor is it mechanistic in matters of law or economic science: see Steiner (1998).
See also Weber (1979, 427).
According to Weber, two great religions first and foremost are—to borrow M. Gauchet’s term—the “religions for departing from religion.” These are Puritanism and Buddhism, the first animated by a logic that severs the believer from any hope of contact with God, the second rendering God somewhat superfluous. This aspect perhaps explains in part the attraction of Buddhism in contemporary Western societies. As to Confucianism, Weber analyzes it as a minimalist religion in the mode of utilitarianism.
It is interesting to note these points of convergence between Renan and Weber, given that the two theorists differ in many other ways. These convergences result from the fact that both assume that believers can only believe what they believe if they see it as well founded. The same assumption plays a fundamental role in Durkheim’s sociology of religion and explains the convergence between the two authors, in spite of the differences that separate them.
This idea may have been suggested to him by Benjamin Constant (1971). For Constant’s influence on Tocqueville, see Siedentop (1994). “Most religions,” writes Tocqueville (1845/2004, 635–6), “are merely general, simple, and practical means of teaching men the immortality of the soul. […] Metempsychosis is surely not more reasonable than materialism, yet if a democracy were absolutely obliged to choose between the two, I would not hesitate: its citizens, to my mind, would be in less danger of reducing themselves to brutes by thinking that the soul of a man might pass into the body of a pig than by believing that it is nothing.”
Inglehart et al. (1998) reveal that in the United States, belief in transmigration is not widespread (26 percent—probably mainly due to Asian immigrants), while belief in the immortality of the soul is very widespread (78 percent of respondents).
The assumption of a fundamentally gradualist innovation was well defended, on a more general level, by Alfred Vierkandt in Die Stetigkeit im Kulturwandel (1908), an unjustly forgotten book that may well have inspired Weber. See volume 3 of Boudon and Cherkaoui (2000).
See also Weber (1922/1988, 39).
I have attempted to reconstruct it in Boudon (2000, 201–46).
Cf. Turner (1992): “[…] the revival of interest in Nietzsche […] for the development of poststructuralism and postmodernism […] has been parallel to the revival of interest in the shaping of Weberian sociology by Nietzsche.” On the convergence between the nonsense of the conservatives and the postmodernists, see Boudon (2000, 201–46). François Chazel (2000, 43ff) shows that Weber did not consider Nietzsche as a scientific authority in any way.
Lazarsfeld interpreted this indifference as a rejection on Weber’s part inspired by certain psychological problems from which he suffered. In fact, Weber’s skepticism is dictated by scientific considerations: he feels that the psychic phenomena that Freud evokes have been known at all times. Reading one of Freud’s example of “repression,” he declares that it says nothing new: “im Prinzip erfahre ich also keinesfalls etwas Neues.” He sees psychoanalysis as a discipline still in its diapers (Kinderwindeln) and finds some of Freud’s fundamental concepts so equivocal, so garbled, that they are in a state of utter dissolution (“wichtige Begriffe […] sind bis zur völligen Verschwommenheit verstümmelt und verwässert worden”). Weber letter dated September 13, 1907, reproduced in Baumgarten (1964, 644–8).
“Durkheim shares with Freud a large part of the responsibility for turning contemporary social thought from the classic rationalist categories of volition, will, and individual consciousness to aspects which are, in a strict sense, non-volitional and non-rational” (Nisbet 1966, 110). The only weakness of this important text is to charge Durkheim with what is more true of the neo-Durkheimians, as well as structuralists of all persuasions, and to omit the influence of the neo-Marxists. The theory of “false consciousness” was no less disastrous an influence on social thought than that of the Freudian “unconscious” or the “collective consciousness” of the neo-Durkheimian structuralists.
There are a number of editions in German and partial translations in French of Weber’s writings on the sociology of religion. Those to which I refer in the text, listed in the bibliography below, have just one thing in common: they are those I have on hand.
“In the great majority of cases actual action goes on in a state of inarticulate half-consciousness or actual unconsciousness of its subjective meaning. The actor is more likely to ‘be aware’ of it in a vague sense than he is to ‘know’ what he is doing or be explicitly self-conscious about it. In most cases his action is governed by impulse or habit. Only occasionally […] is the subjective meaning of the action, whether rational or irrational, brought clearly into consciousness” (Weber 1979, 21–22).
Here, Weber clarifies in advance the notion of “instinct” in Wittgenstein. See Corbier de Lara (2000).