Religion in Human Evolution:

Where did religion come from?

posted by Robert N. Bellah

The most incomprehensible thing about the universe is that it is comprehensible.
—Albert Einstein

The more the universe seems comprehensible, the more it seems pointless.
—Steven Weinberg

When an interviewer for the Atlantic Monthly blog asked me “What prompted you to write this book?” I apparently replied, “Deep desire to know everything: what the universe is and where we are in it.” I don’t deny that I said it—it’s just that I would have thought I would have given a more pedestrian reply, because I am a sociologist, with a Ph.D. in my discipline and some 40 years experience as a professor at Harvard and Berkeley. And I am quite aware that early in the last century Max Weber, in a famous 1918 talk called “Science as a Vocation,” warned that “science has entered a phase of specialization previously unknown and this will forever remain the case.” It does seem that he didn’t apply this dictum to himself, but he was talking about the future when huge projects like his own would no longer be possible. So what is this “deep desire to know everything” in a world of super-specialization? When I look at books like Robert Wright’s The Evolution of God, Nicholas Wade’s The Faith Instinct, Pascal Boyer’s Religion Explained: The Evolutionary Origins of Religious Thought. and Scott Atran’s In Gods We Trust: The Evolutionary Landscape of Religion, recent books that might seem parallel to my own new book, I can only say Weber was right—these books should not have been written, or, to be charitable, they may be good journalism but they are not serious contributions to understanding.

Weber was certainly right that we are in a world of specialization, and dangerously close to the point where specialized work is only intelligible to other specialists. A few years ago a study found that over half of sociology professors couldn’t understand many articles in the ASR or the AJS. Are we living in a world of ever increasing knowledge and ever declining meaning? In the end all that specialized knowledge has to be put together again if it is to be of use. Yet, as I have suggested many of the books that purport to give the big picture are shockingly shallow, based on tertiary sources that only repeat tired clichés or on novel claims that have not been adequately evaluated. We have an enormous “external memory,” as Merlin Donald calls it. It is potentially part of our very selves if we know how to access it. But therein lies our problem.

I’m sure there will be some who will gladly throw my book on the same heap as those I have criticized, but I will try to show a third way, a way that could possibly overcome the split between knowledge and meaning. This way would be to take Weber seriously about specialization but to follow him in not giving up the search for the big picture. What that means is to try to learn a lot about quite a few things. We have more information available about biological and cultural evolution than anyone has ever had before. We have resources to access that knowledge, but it cannot be done quickly or on the cheap. The resources we now have, and I very much mean the web but also e-mail, and books, ever new books, allow us to become quasi-specialists in at least several fields.

It is now possible not only to find out a lot about many areas, but to find out if the real specialists  think you are crazy or not. Some of these are people in the academic world one happens to know—for example the greatest specialist on Shang China in the world, David Keightley, Professor of Chinese history at Berkeley and an old friend, went over my section on Shang China with a fine-toothed comb and saved me from serious mistakes. I had read Terrence Deacon’s The Symbolic Species when it first came out in 1997 and had been very impressed by it, but when I realized over 10 years later that he actually teaches at Berkeley I went to hear him lecture and got acquainted. He and his group were especially helpful in reading my chapter on religion and evolution, giving me some advice, but telling me I was on the right track.

But when it is a field where you know no congenial specialist, you can make friends on the web. Since I wanted to situate religious evolution in the deep biological past I had to learn a lot of biology—rather late in life to say the least. Stephen Jay Gould’s enormous The Structure of Evolutionary Theory of 2002 was a marvelous introduction to many things for me, but Gould was already dead by the time I got to his book. It turned out that animal play was going to be quite important in my argument and the greatest specialist on that subject, Gordon Burghardt, whose splendid book The Genesis of Animal Play would be invaluable to me, is alive and well at the University of Tennessee at Knoxville. I have never met him in person but to this day we have a continuing e-mail friendship and, as I note in Chapter 2, Religion and Evolution, he made many suggestions to me about what I wanted to say about play. Equally important for my whole argument is the evolutionary psychologist Merlin Donald, whom I have met but who has been especially helpful with e-mail comments. For early Greece Ian Morris, the historian and archaeologist of ancient Greece, but also the author of the stunning and breathtakingly ambitious new book, Why the West Rules—For Now, gave me several pages of single-spaced comments on my chapter on ancient Greece, and since he is at Stanford we did meet for coffee once when he was in Berkeley. Michael Witzel, a Sanskritist at Harvard and a great historian of early India was equally helpful with many pages of comments on my ancient India chapter, where of all the four axial cases I had most to learn, but we have never met in person.

Of course there are people who will turn you down—I have had my share. But what is more surprising is how many busy, productive scholars will help, especially if your questions indicate that you have already prepared yourself in the field. Becoming a quasi-specialist in several fields takes time, but becoming a super-specialist in one field also takes a lot of time. And what are all those juicy monographs waiting for if no one is going to take them seriously enough to show their theoretical and comparative importance? I have 12 case studies in my book, several of the axial age chapters being long enough to be small books in themselves, but I have one case of tribal religion, the Kalapalo of the Amazon Basin, about which there exist exactly two books of only one anthropologist, Ellen Basso of the University of Arizona. I know as much as anyone knows about the Kalapalo, except for Ellen Basso and the members of the tribe itself. But even for my other two tribal cases, the Australian Aborigines (though I did focus on a Central Australian group, the Walbiri), and the Navajo, there are thousands of publications.

So from early on in my book I had to develop strategies that would give me more than superficial knowledge without taking over the rest of my life. Obviously you have to use the best of the most recent books, and if possible, as in the case of the Navajo, consult specialists (and I started out studying the Navajo for my undergraduate honors thesis, Apache Kinship Systems, over 60 years ago). Without any guidance the amount of material available on any one case is overwhelming. Even finding the best recent secondary works often requires help or maybe luck and you also need to look at the classic secondary works. And you can’t just rely on secondary work where good translations are available as they usually are for most cases (Shang China being a notable exception). There you have to find out which are the most reliable, also not an easy task. It is claimed that the Daodejing is the most translated book in the world, but 99% of those translations are worthless. You need to find the reliable ones. When working on ancient China I regularly used five translations of the Confucian Analects because they had different virtues. It also helps to know enough Chinese to check key terms in the translations against the Chinese original. I don’t mean to discourage scholars from pursuing similar studies; I’m just trying to describe what is involved in serious comparative work.

But of course, if you are a sociologist, you are doing more than describing fascinating cases, though you have to do a lot of that as Weber and Durkheim already showed us; you have to have a theory, maybe a multi-stranded theory, since so much is going on in each case and there is no simple one fits all formula. Before discussing my theoretical resources let me give you another example of the kind of work I did in my new book: one of the finest books ever written by an American sociologist is Randall Collins’s The Sociology of Philosophies, a book even longer than mine. And Collins is not just talking about Classical Greek and modern European philosophy—he includes Islamic, Indian, Chinese and even Japanese philosophy. I may think he is sometimes wrong but I never think he is stupid. How he read so much while carrying a full teaching load staggers me. I had to retire to write my book, even though retirement is not so retired as you might think. And Collins’ approach is theoretical all the way through. He uses his micro theory of interaction ritual chains amazingly well to understand the macro development of philosophy in a variety of very different traditions. I might have added Collins’s interaction ritual chains to my theoretical took kit, but I have to admit that I didn’t read every last word of this great book until after I had finished my own.

So let me just list some of my theoretical frameworks and address them as thoroughly as I can here.

I start with Clifford Geertz’s definition of religion in his “Religion as a Cultural System,” which I should give in my abbreviated version to clarify what I mean and don’t mean by religion: “Religion is a system of symbols which, when enacted by human beings, establishes powerful, pervasive, and long-lasting moods and motivations that make sense in terms of an idea of a general order of existence.”  I should point out that neither Cliff nor I use the terms gods or God. What Geertz meant by a cultural system is very dependent on his reading of Alfred Schutz, particularly his paper on multiple realities or multiple worlds, terms which Schutz took from William James. Besides what Schutz called the paramount reality, the world of daily life, what Weber called “the everyday,” Schutz distinguished the world of science, the world of religion, and the world of art.

After describing what kind of multiple reality religion is, I wanted to look at the major forms of religious representation, the ways in which people engage in religious action and religious thought. Here I turned to the field of child development, not to look at the ways in which children become religious, though some have worked on that, but to look at the way infants and then children acquire the various capacities to relate to the world. Here was another big field to master, but one in which I have long been interested—especially the work of Jerome Bruner, one of my teachers in graduate school, who is the most important cultural psychologist still living and whose categories for the cognitive development of the child turned out to be remarkably relevant for my purposes. Bruner, himself adapting ideas from Piaget, sees the child as moving from enactive to symbolic to conceptual representations. I prefaced these with the idea of unitive events rooted in the original unity of mother and child but emerging later as religious experiences, usefully described by Alison Gopnik of UC Berkeley’s psychology department in her recent book The Philosophical Baby. So Piaget, Bruner, and Gopnik were my anchors but I looked at a lot of other things as well, particularly the work that links cognitive development in human children with comparable development in the great apes and other mammals.

The major stages of ontogeny turn out to parallel the major stages of phylogeny as described by Merlin Donald in Origins of the Modern Mind: Three Stages in the Evolution of Culture and Cognition. Donald prefaces his three stages by referring to episodic culture which we share with other higher mammals and that I see as analogous to unitive events in ontogeny.

I should note that in both Bruner and Donald stages are never left behind, but are reconfigured in new contexts when subsequent stages emerge, leading to my general rule that “nothing is ever lost,” by which I don’t mean cultural content which is all too easily lost (most of the plays of Aeschylus, Sophocles and Euripides, for example) but the cultural capacities themselves, which never lose their essential and indispensible nature. Donald’s three stages are mimetic, mythic, and theoretic, paralleling Bruner’s enactive, symbolic, and conceptual.

I want to describe what Merlin Donald means by mimetic culture because it makes intelligible what happened during a long period of human evolution, most likely the period between the appearance of Homo erectus, 1.8 million years ago, and the emergence of our own species, Homo sapiens, during the last two or three hundred thousand years. Mimetic culture involves a kind of bodily communication more elaborate than anything comparable among the other great apes, lacking language but probably involving spoken or sung communication, what some evolutionary musicologists call musilanguage. Mimetic communication almost certainly led to ritual, though as yet without myth, which requires language capacities that were lacking.

In modeling the society itself as well as its constituent roles, mimetic culture provided the necessary resources for moving beyond the rather anarchic chimpanzee band to a larger group capable of controlling in-group aggression such that pair bonding and same-sex solidarity in various contexts could result. In-group solidarity did not mean these mimetic-culture based societies were peaceful. There is every reason to believe that they were not, that there was endemic conflict between groups and probably in-group aggression was only relatively successfully controlled.

The limitations of mimetic culture are evident. Donald writes:

Mimesis is thus a much more limited form of representation than symbolic language; it is slow moving, ambiguous, and very restricted in its subject matter. Episodic event registration continues to serve as the raw material of higher cognition in mimetic culture, but rather than serving as the peak of the cognitive hierarchy, it performs a subsidiary role. The highest level of processing in the mimetically skilled brain is no longer the analysis and breakdown of perceptual events; it is the modeling of these events in self-initiated motor acts. The consequence, on a larger scale, was a culture that could model its episodic predecessors.

It is well to remember that we humans are never very far from basic mammalian episodic consciousness, the awareness of the event we are in. Mimetic culture is an event about an event. Narrative, which is at the heart of linguistic culture is basically an account of a string of events, organized hierarchically into larger event units. But the moment when our predecessors first stepped outside episodic consciousness, looked at it and what was before, around, and would be after it, was a historic moment of the highest possible importance. Other higher mammals, although they are social, are more tightly locked each in their own consciousness. They are, as Donald says, almost solipsists. But humans, once mimetic culture had evolved, could participate in—could share—the contents of other minds. We could learn, be taught, and did not have to discover almost everything for ourselves. Mimetic culture was limited and conservative; it lacked the potential for explosive growth that language would make possible. But it was the indispensable step without which language would never have evolved.

Further, mimesis is, though in many respects less efficient than language, indispensable in its own sphere. As Donald writes, mimesis “serves different functions and is still far more efficient than language in diffusing certain kinds of knowledge; for instance, it is still supreme in the realm of modeling social roles, communicating emotions, and transmitting rudimentary skills.”  Maybe not just rudimentary skills, for mimesis is basic for the teaching of quite complex skills in such fields as athletics, dance, and possibly other arts. Finally mimesis remains indispensable in “the collective modeling and, hence, the structuring” of human society itself. That is what ritual does, and if Randall Collins is right, it is micro-ritual moments that make our lives bearable whenever we interact with others.

So far I have been talking mainly about where religion came from so I must say a little about where it was going. Where it was going was toward language, what Donald calls mythic culture, and beyond that theory, though it would take a long time to get there. But remember we are still in the world of egalitarian foragers. Most of my book deals with hierarchical class societies, yet they all derive from egalitarian forager societies. That’s where it all begins and that is where our most basic capacities were formed.

We are so fascinated with ourselves as language users that we think discovering the origin of language is the key to understanding human evolution. It is one of the great virtues of Merlin Donald’s work that he takes culture—the ability to escape our solipsism and connect with a larger shared consciousness—as the key to what makes us unique. It is in this context that his idea that language “piggybacks” on culture makes sense. Language acquisition in the individual is social: even if there were such a thing as a language module, which neither Donald nor I for a minute believe, it could only become operative in a socially provided linguistic context. Isolated children do not learn spontaneously to speak. Jerome Bruner, as Donald reminds us, has shown convincingly that language learning requires an external support system, a linguistic milieu, to be effective. The question is, what was the “external support system” that made language possible in the first place? My answer would be ritual, which provides the security, intensity, and redundancy without which language would not emerge. Donald writes:

Linguistic universals spring from the context in which real-world languages are learned and, more important, in which they evolved. Like any other set of conventions, linguistic conventions are shaped by the situations in which they originated. They have mimetic origins. Thus, once we change our paradigm, the features of universal grammar emerge smoothly from a close analysis of gesture, mime, and imitative behavior. The “language instinct” exists, but it is a domain-general instinct for mimesis and collectivity, impelled by a deep drive for conceptual clarification.

But why this drive toward conceptual clarification? Donald suggests that there was a need for a more coherent representation of the world than was possible through mimesis. “Therefore,” he writes, “the possibility must be entertained that the primary human adaptation was not language qua language but rather integrative, initially mythical, thought. Modern humans developed language in response to pressure to improve their conceptual apparatus, not vice versa.”  Myth is a profoundly ambiguous word, so it would be well to be clear what Donald means by it:

[M]ythical thought, in our terms, might be regarded as a unified, collectively held system of explanatory and regulatory metaphors. The mind has expanded its reach beyond the episodic perception of events, beyond the mimetic reconstruction of episodes, to a comprehensive modeling of the entire human universe. Causal explanation, prediction, control—myth constitutes an attempt at all three, and every aspect of life is permeated by myth.

It is because of, in a sense, the primacy of myth over language that Donald calls the stage after mimetic culture, mythic culture.

Donald, in emphasizing the cognitive role of myth, approaches the view of Claude Levi-Strauss, the anthropologist who, more than any other, has emphasized the intellectual function of myth. Levi-Strauss, nonetheless, does not think of myth as a form of science, or a primitive precursor of it, but as having a different cognitive function:

To say that a way of thinking [myth] is disinterested and that it is an intellectual way of thinking does not mean at all that it is equal to scientific thinking. . . It remains different because its aim is to reach by the shortest possible means a general understanding of the universe—and not only a general but a total understanding. That is, it is a way of thinking which must imply that if you don’t understand everything, you can’t explain anything.

That is a view of myth that would indeed see it as “impelled by a deep drive for conceptual clarification.”  So Aristotle was not wrong when he wrote the first sentence of his Metaphysics: “All humans by nature desire to know.”  And what did Aristotle want to know? Everything. But for him it wasn’t myth but theory that would get us there, and we can see how well we are doing with that right now by looking at the institution in which most readers of this blog are presently situated:  the university.

Tags: , , , , , , , , ,

Printer-Friendly Version


3 Responses to “Where did religion come from?”

  1. avatar Benson Saler says:

    According to Pliny the Younger, Pliny the Elder held that no book is so bad that some good might not be extracted from it. I regret that Robert Bellah has not adopted that point of view. It might have served him well both in writing his book and in his post-publication pronouncements. Thus, for instance, instead of consigning books by Scott Atran and Pascal Boyer to the flames, he might have discovered something interesting and informative in what they say. Indeed, Bellah might even have achieved an accurate understanding of what they claim. Had he done so, he would not maintain, as he does in Religion in Human Evolution (629, n. 154), that they suppose that religion is “explained by a module for supernatural beings.” Atran and Boyer no more suppose that than they suppose that there is a “god gene.”

  2. “Bob,” my old thesis director of thirty years ago. My heartiest congratulations! Religion in Human Evolution is the best book in my library of over 3,000 books on religion. You have shown the world how religion is as natural as evolution and leads us men and women to stand in awe and wonder of all the ways that it takes us by the hand and leads us beyond nature to the great Mysterious Transcendent.

  3. avatar Edward Manier says:

    I wonder if Professors Saler and Bellah could accept a slight paraphrase of Bellah’s comment on Atran: “I have found particularly unhelpful those who think of the mind as composed of modules and of religion as explained by the absence of a module for supernatural beings.”

    Clearly Bellah prefers the work of scholars less sympathetic to the deep anthropological significance of the “modularity of mind” than Atran’s “In Gods We Trust,” e.g., Merlin Donald and Michael Tomasello.

    Bellah’s real quarrel, in the same note, is with those who indulge a tendency toward “speculative theorizing” because of their “lack of insight into religion as actually lived.”

    My own less sardonic suggestion for Bellah is that he might look to Atran’s impressive work on folk biology and the folk psychology of jihad (see “The Native Mind and the Cultural Construction of Nature” and “Talking to the Enemy: Faith, Brotherhood, and the (Un)Making of Terrorists”) for more common ground.

Leave a Reply

Please note: All comments will be approved by an administrator before they appear on this page.