Religion in Human Evolution: From the Paleolithic to the Axial Age

by Robert N. Bellah

  Poetic Symbolization

  Here I want to talk about symbolic language or linguistic symbolization. I am using the term “poetic” broadly to cover the nondiscursive uses of language, in the sense that Susanne Langer distinguishes between discursive and presentational forms.60 Though we think of iconic symbols as intrinsically related in some way to what they symbolize, we usually think of the relation of words to what they symbolize as perfectly arbitrary: aren’t “dog” and chien equally accurate and equally arbitrary ways of referring to the same thing? Yet we also know that an English poem about a dog doesn’t go at all easily into French and that if the French translation succeeds, it is because the English poem has stimulated a new and genuinely French poem. It is not at all obvious to young children that the relation between words and things is arbitrary. Piaget recounts the following dialogue between an adult and a young child:

  And how did we know that the sun is called “sun”?-Because its in the sky. It’s not on earth. It gives us light in the but how did we know?-Because its a great ball. It has rays. We knew it was called how did we know its name was “sun”? We might have called it something else.-Because it gives us light.-How did the first men know it was called “sun” and not something else?-Because the big ball is yellow and the rays are yellow, and then they just said it was the sun, and it was the how did the first men know it was to be called sun?-Because it’s up in the air. It’s high up … How did the first men know the name of the sun?-Because they had seen the sun.

  On the basis of many answers of this sort, Piaget offers an explanation:

  These answers are very suggestive, for although they press nominal realism to its utmost limit they are not absurd. For indeed, although children may suppose they need only to look at a thing to know its name, it does not in the least follow that they regard the name as in some way written on the thing. It means rather that for these children the name is an essential part of the thing … the name sun implies a yellow ball that shines and has rays, etc. But it must be added that for these children the essence of the thing is not a concept but the thing itself.61

  Wallace Stevens in a late poem rephrases the children’s understanding when he says,

  Schutz clarifies the distinction in poetry between the concept and “the thing itself,” being part of the thing itself rather than about it:

  A poem, for instance, may also have a conceptual content, and this, of course, may be grasped monothetically. I can tell in one or two sentences the story of the ancient mariner, and in fact this is done in the author’s gloss. But in so far as the poetical meaning of Coleridge’s poem surpasses the conceptual meaning-that is to say, in so far as it is poetry-I can only bring it before my mind by reciting or reading it from beginning to end.63

  Archibald MacLeish makes the same point poetically when he says in “Ars Poetica” that a poem should be “mute,” “dumb,” and “motionless,” that is, not talk about something. Rather,

  Susanne Langer explains:

  Artistic symbols … are untranslatable … and cannot be explicated by any interpretation. This is true even of poetry, for though the material of poetry is verbal, its import is not the literal assertion made in the words, but the way the assertion is made, and this involves the sound, the tempo, the aura of associations of the words, the long or short sequences of ideas, the wealth or poverty of transient imagery that contains them, the sudden arrest of fantasy by pure fact, or of familiar fact by sudden fantasy, the suspense of literal meaning by a sustained ambiguity resolved in a long-awaited key-word, and the unifying, all-embracing artifice of rhythm.64

  But it is not just that poetic symbolization, like all symbolization, participates in what it symbolizes. Poetry is a form of action, a way of “doing things with words.“65 As Mandelstam has said, “poetry is power,” and as Rimbaud put it, “metaphor can change the world.” More modestly, Helen Vendler has pointed out the “performative” character of much lyric poetry. Her discussion of Shakespeare’s sonnets suggests that even in those sonnets that are not obviously “dramatic”-that is, addressed to another in praise, demand, or query-there is often an implicit rebuttal or defense against accusation or something of the sort, and if such “speech acts” go undetected, the poem may be seriously misunderstood.66

  Just as images and sounds can “reach right into the body,” so can heightened (or, as we shall see in a moment, “condensed”) language. A doctoral dissertation gives a powerful instance. The author was the pastor of an urban California congregation:

  A woman’s mother was near death, but she didn’t have a church home. She asked her neighbor, a member of the congregation I served, if she belonged to a church. The neighbor gave her my name and I visited her. I met her at the door and the conversation went something like this. “I don’t know why I called you, but my mother is near death and I thought maybe we should have someone from the church invited me into her home and we talked about her mother. I found out that the mother and daughter had been involved with a church many years ago. Then I suggested that we go into her mother’s room and have a prayer. The woman suggested that we have the prayer in the living room because her mother had been in a coma for many days and couldn’t participate. But I urged that we go in the room anyway. I offered a prayer and then asked if she knew the “Lord’s Prayer.” I invited her to join me. We had barely said, “Our Father who art in heaven,” when her mother joined us through the rest of the prayer. She came out of her coma for a few days before she died and the mother and daughter had significant conversation.67

  Here familiar, powerful words reach into the mother’s body and pull her back into consciousness, at least for a few days.

  In the young child the connection between language and action, what the body does, is a close one. As Bruner says,

  The initial structure of language and, indeed, the universal structure of its syntax are extensions of the structure of action. Syntax is not arbitrary; its cases mirror the requirements of signaling about action and representing action: agent, action, object, location, attribution, and direction are among its cases. Whatever the language, the agent-actionobject structure is the form soon realized by the young speaker … For what the child himself shows us is that initial development of language follows and does not lead his development of skill in action and thought.

  Although later language becomes (in part) free from the context of action, as Bruner says, “to understand what a baby is saying, you must see what the baby is doing.“68

  The child learns to use language in contexts, which Bruner describes as “conventionalized into conventional forms and regularized as formats.” What he means by a format is “a routinized and repeated interaction in which an adult and child do things to and with each other.” The peekaboo game is a good example. But much adult language is also contextual and follows conventional formats. Bruner insists that such language is in a sense “constrained” so as to be “cognitively manageable.“69

  Bruner’s talk of “contexts” that are “constrained” and that give rise to “formats” leads one to think of the work of Basil Bernstein, though Bruner does not make the connection. Bernstein, as is well known, distinguished two kinds of language codes, one a “restricted code,” embedded in the concrete social relations of the people using it, and the other an “elaborated code,” largely decontextualized and available for use between people with no intimate or particular “Restricted code” is perhaps an unfortunate choice of terms, for such language can be very rich in overtones and implications and involve considerable virtuosity in use. I would prefer the term “condensed” to “restricted,” and Mary Douglas, whose Natural Symbols is, in her words, “an essay in applying Bernstein’s approach to the analysis of ritual,” speaks of “condensed” versus “diffuse” symbols, as at least partially analogous to Bernstein’s usage.71 Condensed language, and here I would include the intimate language of parent and child, and of lovers, as well as of poetry, requires a world of shared experience that the wor
ds imply and that gives a surplus of meaning to any utterance. Poetry that has been learned by heart and been recalled in a variety of situations during a lifetime certainly has more meaning than a poem that is met on the page by the solitary reader for the first time. But even then, if the poem is to mean anything at all, it will almost certainly be because the reader knows something of the tradition of poetry out of which this particular poem comes and to which it will inevitably allude.

  If condensed language is effective in forming identity in intimate contexts, in families, it also operates at the level of national identity, as Benedict Anderson points out:

  There is a special kind of contemporaneous community which language alone suggests-above all in the form of poetry and songs. Take national anthems, for example, sung on national holidays. No matter how banal the words and mediocre the tunes, there is in this singing an experience of simultaneity. At precisely such moments, people wholly unknown to each other utter the same verses to the same melody. The image: unisonance. Singing the Marseillaise, Waltzing Matilda, and Indonesia Raya provide occasions for unisonality, for the echoed physical realization of the imagined community.

  Indicating how intense the experience of the “mother tongue” can be, Anderson, quoting from Thomas Browne, says that although up to a point the words can be translated, they “can bring goose-flesh to the napes only of English

  Religious language is often condensed, poetic, and, because of its involvement in ritual, performative. No more than any other kind of poetry can it be conceptually paraphrased without significant loss of meaning. Even a poet who has ostensibly left religion behind can catch the intensity of religious language, as when Stevens, noting that “There was a heaven once,” goes on to say:

  Or late in life, after saying, in accordance with the teachings of Romanticism, “God and the imagination are one,” he writes, “How high that highest candle lights the dark,” returning to that symbolism of light and sun that haunts all his poetry.74

  Narrative

  Narrative is in one sense only a part of what I have been calling poetic symbolization. But there are aspects of narrative that make it transitional to my third category of representation, conceptual representation. Conceptual representation is always linguistic (or quasi-linguistic like mathematics), but not all language, as I have argued in the case of lyric poetry and related forms of condensed language, is conceptual. By conceptual I mean something like what Susanne Langer means by discursive. As she explains, “Language in the strict sense is essentially discursive; it has permanent units of meaning which are combinable into larger units; it has fixed equivalences that make definition and translation possible; its connotations are general, so that it requires non-verbal acts, like pointing, looking, or emphatic voice-inflections, to assign specific denotations to its terms.“75 One can imagine narratives that would be discursive in Langer’s sense-that is, literal accounts of what actually happened. But many forms of narrative (including some that claim simply to recount what actually happened) are in fact governed by symbolic modes of organization. As Herbert Richardson says of myth, a significant form of narrative for our purposes, “mythical discourse rises at the level of the total story, the most complex level of linguistic utterance. The linguistic unit appropriate to myth is not the single word nor even the sentence, but the story.“76 Thus, the truth of a narrative in this sense does not arise from the “correspondence” of its words or sentences to “reality,” but from the coherence of the story as a whole. Just as a poem cannot be paraphrased conceptually without irreparable loss, neither can such narratives be.

  That the “total story” is the significant form of myth gives it a presentational quality, in Langer’s sense, that overcomes the temporal sequentiality of its discursiveness. Levi-Strauss nicely makes this point in relating myth to music:

  Myth and music [are both languages] which, in their different ways, transcend articulate expression, while at the same time-like articulate speech, but unlike painting-requiring a temporal dimension in which to unfold. But this relation to time is of a rather special nature: it is as if music and mythology needed time only in order to deny it. Both, indeed are instruments for the obliteration of time … Because of the internal organization of the musical work, the act of listening to it immobilizes passing time; it catches and enfolds it as one catches and enfolds a cloth flapping in the wind. It follows that by listening to music, and while we are listening to it, we enter into a kind of immortality.

  It can now be seen how music resembles myth, since the latter too overcomes the contradiction between historical, enacted time and a permanent

  There are other features of narrative that anchor it in the symbolic or even in the enactive modes of representation. Stories can be acted out and often are in plays, movies, television, ritual. Even when the spectator is only an observer, something like the bodily response to music is also present. This is obvious with adventure movies, where we cringe, gasp, and so on with each turn of the plot even though we are sitting in the audience and the action is on the screen. But even a verbal narrative can do the same thing. Hold a young child on your lap as you tell the story of Little Red Riding Hood and you will feel the bodily response.

  Another feature of narrative that links it to the symbolic is that the distinction between inner and outer, between self and world, is not as clear as in conceptual discourse. We identify with what is going on in the narrative. It is all very well to comfort a child by saying, “It’s only a story,” but at some level the child knows that the story has its own truth that such disclaimers don’t reach. We are even drawn into narratives that purport to recount actual events. When we read in the newspaper or hear on television of someone who has lost an only child in a drive-by shooting, we cannot help, however briefly, feeling the pain, feeling “that could be me.” This is not an aberration of “primitive” or “regressive” thinking. It is a normal human response.

  Human beings are narrative creatures. Narrativity, as we shall see, is at the heart of our identity. One study of conversations at home between mothers and preschool children showed that narratives occurred every seven minutes, three-quarters of them told by the The next time you are on public transportation or in a waiting room and can’t help overhearing conversation, notice how much of it is narrative.

  We not only tell stories to others, we tell them to ourselves, and this begins remarkably early. Bruner reports a study of soliloquies recorded from the crib of a child named Emily between the ages of 18 months and 3 years:

  Listening to the tapes and reading the transcripts repeatedly, we were struck by the constitutive function of her monologic narrative. She was not simply reporting; she was trying to make sense of her everyday life. She seemed to be in search of an integral structure that could encompass what she had done with what she felt with what she believed.79

  From the ubiquity of narrative there emerged rather recently, as Bruner recounts it, the notion of “Self as a storyteller.” Bruner quotes the psychoanalyst Roy Schafer:

  We are forever telling stories to ourselves. In telling these self-stories to others we may, for most purposes, be said to be performing straightforward narrative actions. In saying that we also tell them to ourselves, however, we are enclosing one story within another. This is the story that there is a Self to tell something to, a someone else serving as audience who is oneself or one’s self … On this view, the self is a telling.

  Psychoanalysts have discovered what writers of fiction have always known: that the stories we tell about ourselves, although based on our lives, contain “screen memories,” even fictions, as well as fact. The truth they contain is not historical truth but narrative truth, and the task of the analyst is not at chaeological recovery but assistance with the construction of a new narrative that will help the analysand deal more successfully with whatever trouble brought him to analysis, will help him redescribe his life.80

  The self is a telling. And inevitably a telling about others as well as
the self. Indeed, the self cannot be disentangled from significant others except ideologically, leading Bruner to speak, somewhat awkwardly, of a “distributed Self.“81 A distributed self is made up of the significant relations, “identifications,” of a lifetime. A distributed self will behave differently in the different spheres and relationships in which it is engaged. If the distributed self doesn’t fit too easily with the “deep self” of modern ideology, neither is it merely the shifting play of masks, the “presentations of self in everyday life,” of Erving Goffman.82 Narrative truth is no more secure than any form of truth, but it can be stable, reliable, even profound.

  If personal identity resides in the telling, then so does social identity. Families, nations, religions (but also corporations, universities, departments of sociology) know who they are by the stories they tell. The modern discipline of history is closely related to the emergence of the nation-state. This is a peculiarly interesting example for our purposes. Families and religions have seldom been concerned with “scientific accuracy,” with conceptual discursiveness, in the stories they tell. Modern nations have required national histories that will be, in a claimed objective sense, true. And unquestionably a great deal of accurate fact has been uncovered. But the narrative shape of national history is not more scientific (or less mythical) than the narrative shape of other identity tellings, something that it does not take debunkers to notice. Benedict Anderson in Imagined Communities recounts both the widespread establishment of chairs of history within a generation of the French Revolution and its unleashing of nationalist fervor and of the strange mixture of memory and forgetting that that history produced (not so strange to those familiar with other forms of self-telling).83

  There is one other feature of narrativity that I want to discuss before coming back to the sense in which narrative is transitional to conceptual representation. In Bruner’s account, Kenneth Burke in A Grammar of Motives describes a pentad of elements that can be found in any story: an Actor, an Action, a Goal, a Scene, and an Instrument-plus, according to Bruner, Trouble.84 Trouble results when two or more elements fall out of harmony: the Buddha rejects the preordained Goal when he refuses marriage and the inheritance of his father’s kingdom, or Jesus proclaims a radically new understanding of the Scene: the kingdom of God is at hand. These disharmonies are not fortuitous. There is something inherent in the order of human life, of the fact that it is a normative order, of the way things ought to be, that provokes or reveals disorder. The myths of the Australian Aborigines, as W. E. H. Stanner interprets them, express this “immemorial misdirection” of life.85 Thus the Buddha discovers what his parents seek to keep from him: that human life inevitably involves old age, sickness, and death. And until after the resurrection, Jesus’s disciples are too horrified to accept that Jesus must be crucified. So the denial of time in narrative is possible only by taking up time and its Trouble into the heart of narrative itself. The task of the Self’s telling is to deal with the Trouble that is at the heart of every life, to find a form that makes it possible to go on in spite of suffering. And, as Benedict Anderson points out, national histories are centrally concerned with “our dead,” who must be remembered in order that “they did not die in vain.“86