This long work consists of seventeen case studies of revolutionary religious movements—which range over history from the Maccabean Revolt to the Taiping Rebellion to Catholic radicalism in contemporary Latin America—with an introduction and seven chapters of theoretical conclusions. In his preface, however, Guenter Lewy suggests that readers interested in his general argument skip the case studies and turn directly to the last chapters. That is bad advice. The book really has no general argument, and Lewy’s conclusions are inconclusive. But the cases are fascinating. As an anthology of historical studies, Religion and Revolution is an unquestionable success. Lewy, however, is more ambitious. For him, as for many social scientists these days, it is theory or nothing.

At the same time, he wants his theory to be true, and he knows that there isn’t much one can say about the connection of religious faith and revolutionary activity that will fit all his cases. It’s not entirely facetious to suggest that, for a really good theory, Lewy has sixteen cases too many. What proposition would be true for all of them? He might say, “Religious faith sometimes does and sometimes does not lead to revolutionary action.” Indeed, he does say something like that, but the statement is, unhappily, not yet a theory. He must go on to consider why it does and doesn’t. He thinks he can describe only the necessary, not the sufficient conditions of religious radicalism. He aims at non-statistical probabilities: if A and B and C, then probably D, where “probably” has its ordinary language meaning—more likely than not—and nothing more.

Now, in principle it must be possible to do this, to identify a set of variables which appear in some patterned way, so that it can be said that they probably will or won’t produce some specified outcome, depending on the pattern. But even this is very hard to do, and Lewy is honest enough to pull back from every serious attempt to do it. He is in flight from the theory he is committed to have, and the spectacle is sometimes painful.

Lewy identifies four variables which must be studied if we are to explain the political effect of religious belief. The first is the religious creed itself, the doctrine and dogmas of the believers. Might not belief have an impact on political life like that of Weber’s Calvinism on economic life? But it appears that Weber’s argument works only because we somehow assume that it isn’t terribly peculiar to be a Calvinist; Calvinism itself doesn’t have to be explained. It is the chief advantage of Lewy’s seventeen cases however that some of them are bound to seem peculiar. Consider the bizarre melange of Christian millenarian and traditional Confucian beliefs that swept China during the Taiping Rebellion. We are less inclined to explain the rebellion as an effect of the dogmas proclaimed by that son of God, Hung Hsiu-ch’üan, than to explain the dogmas, or the reception of the dogmas, as the effect of whatever (we think) caused the rebellion. By analogy, we might want to explain the reception of Calvinism by reference to whatever factors “caused” the rise of capitalism.

In any case, the early capitalists were not merely Calvinists, but Calvinists of a particular kind, and the case is the same with Lewy’s revolutionaries: they recombine and reinterpret conventional religious doctrines, and there seem no limits, certainly no limits of logic, consistency, or coherence, to the combinations and interpretations they come up with. And if that is so, we cannot look for explanatory force in the original doctrines. Lewy is right to warn us that “the influence of abstract ideas on human affairs should…not be overstated,” and he is right not to overstate it in his cautious conclusion: “The link between religious doctrine and politics is often tenuous.”

The second variable is ecclesiastical organization. Here Lewy has in mind the argument of the German theologian Ernst Troeltsch about the different behavior of sect members and church members—the sectaries, in Troeltsch’s view, were more likely to be radical. Once again Lewy demonstrates that with seventeen cases, and no reference short of universal history, such arguments don’t work. In both Burma and Ceylon, for example, the Buddhist establishment has played a significant part in militant politics: Lewy studies both cases. I don’t think it is his intention to suggest that Troeltsch should have done likewise (Lewy himself uses the notion of a “sect” in writing about the German reformation). But his cases clearly do suggest that, from a universal standpoint, the politics of a religious institution—whether it will function “as a critic or supporter of the status quo”—“does not depend on the type of ecclesiastical organization involved.”


The third variable is leadership. We are invited to reflect on the role of charismatic figures in religious movements. Lewy believes they have an important, indeed, a vital role, but he is at a loss what more to say. It is a feature of charismatic leaders that their followers don’t join up because of predictions they make about where they will be led. They don’t know in any specific sense where they will be led, and the doctrine the leader proclaims offers no firm clues. His characteristic form of address is, “It is written…but I say unto you….” Apparently scholars cannot make any predictions either. Lewy is entirely agnostic: “Even if we knew what causes certain people to become leaders, we could not foresee the choices they would make in certain critical situations.”

Lewy’s fourth variable is a residual category: situational factors “unrelated to religion”—social structure, economic conditions, political arrangements, and so on. Here most of the explanatory force of this “multi-factoral analysis” seems to lie. We might ask whether the first three variables ever operate independently of the fourth. Lewy seems to assume that they don’t (he never says never). But if that is so, his real subject might just as well be revolution itself. Why separate out religious revolutions if religion has as little independent significance as his treatment of doctrine, organization, and leadership seems to suggest?

With variables so conceived, it’s not likely that much can be explained. If a religious creed lends itself to millenarian interpretations (they all do), and if institutions are flexible enough to permit political activity (they all are), and if a charismatic leader arises who offers the appropriate interpretations and stimulates the appropriate activity, and if social and economic conditions are favorable (which usually means unfavorable to ordinary human life), then it is likely that there will be (one might say, that there is) a radical religious movement. I think this is a fair example of the “tendency statements” to which Lewy aspires, but I should give an example of his own. I take it from a methodological discussion in his final chapter: “we conclude…that revolutionary millenarianism is a rare phenomenon in an increasingly secular society….” Since millenarianism is defined in religious terms, this means that religious movements are unlikely in a nonreligious age. The assertion is true even if we “cannot specify…how few cases amount to a ‘rare phenomenon,”‘ and trivial even if we could.

One theorizes in this way only under compulsion. It is, I suppose, what the profession requires, some gesture toward abstract statement, some small move toward a genuine social science. But the enterprise is terribly misguided, and it makes a sad conclusion to what would otherwise be an exciting if untheoretical book. It would be (it is) exciting precisely because of its “anthological” character, which turns all its readers into laymen and invites us to reflect on a range of (mostly) unfamiliar experience.

But I am afraid that for Lewy the cases exist largely for the sake of the theory. He has no particular interest in any of them; he is not committed to a particular interpretation of any of them. Each of his case studies is balanced and objective, a model of historical scholarship; his understanding is always (to judge now only by the cases I know something about) resolutely conventional. He seems to think that theory can be based on cases discovered and recounted without theory, that theory must stand, at the end, in isolation and abstraction. That seems to me profoundly wrong. Theory needs to be tried out on cases; we value it for its illumination. Lewy offers us a wealth of material, but he sheds no “Lewyan” light on any of it. He does not give us a theoretical account of any particular event.

There is no better statement of Marx’s theory of class struggle than his Eighteenth Brumaire, no better statement of Weber’s theory of religion than his Protestant Ethic. No doubt, these examples set a high standard, but they also suggest what it means to have a theory. One can be more modest than Marx or Weber without forgetting that lesson. A theorist sees things, sees everything, perhaps, in his own peculiar way, and he must live with the risks of doing that. These risks are in the deepest sense concrete. The theorist is not likely to say, or to say only, that variable X and variable Y generally vary together. He insists that they actually did vary together in this case and in that one; he gives hostages to historical scholarship.

But it is not always possible to do that, nor is it always necessary to try. Sometimes we are overwhelmed by the insistent particularity of history and by the uniqueness of the men and events we study. It’s not that a set of variables cannot be singled out; at some level of abstraction that can always be done. But the set may not be enlightening or its members may not vary in any consistent or patterned way. Then a scholar can decide to study something else—or take seriously the idea of comparative history and let the different cases play against one another, directing our attention to their peculiar incompatibilities. Here one aims at insight and sensitivity, not at theory, but those goals are valuable too. Lewy gives them up, however, when he insists that his case studies are “self-contained,” that they can be read in isolation from one another. Having chosen that perspective, he is committed to his last chapters, for only they can tie the studies together.


Lewy is a good historian and a loyal social scientist, so his case studies are theoretically neutral and his conclusions are abstract. His great learning is no protection, it turns out, against the program of contemporary social science: to build models of social interactions and to do this at a great distance from the interactions themselves. Lewy too is led to climb very high in order to escape the constraints of his cases. It is to his credit, I think, that he never does escape. He remains torn between his respect for his material and his commitment to his variables, and so his model is radically incomplete, endlessly hedged. We are left to relish his wonderful stories about cargo cults, political monks, holy wars, and millennial kingdoms, and for these, though not for his “systematic political and social theory,” we should be grateful.

Postscript. In the late 1960s, Lewy tells us, because his research was supported by a Pentagon grant, he “was criticized by radical students for serving the establishment.” At the same time, Senator Fulbright “inserted into the Congressional Record summary descriptions of six of my case studies in order to demonstrate how the Department of Defense wasted the taxpayer’s money on research without utility to the military. I leave it to my readers to evaluate these criticisms.” My own sympathies are with Senator Fulbright. Whether or not it is immoral to “serve the establishment” depends on what one does; I can think of all sorts of useful projects which the Pentagon might or might not pay for. But there is nothing in this book that they should have paid for, nothing relevant to the national defense.

Now, I am sure Guenter Lewy never promised that his research would help the United States save the world from religion and revolution. But clearly someone sold the generals a bill of goods about the utility of this kind of scholarship. Lewy presumably said nothing to enlighten them, but they have so much money they probably never asked for enlightenment. Taxpayer resentment about such expenditures, though clearly misplaced when there is so much else to be resentful about, nevertheless seems to me justified. If serious work like Religion and Revolution deserves public support, as I think it does, the money should come from some agency openly committed to scholarship without regard to its immediate or even its long-run utility. We might all get less money then, but it would also be less degrading to ask for it.

This Issue

October 17, 1974