1.
Ethnography and Literature:
A History of Their Convergence

1

"For the first time in the long and fruitful relationship between literature and science," David Porush writes, "literature actually has the means to meet science on its own territory in a contest concerning which epistemological activity does a better job of telling the truth" (373). The context for this "contest" is "postmodernism." For Porush, "literary postmodernism and scientific postmodernism [are best seen as] two aspects of a single enterprise" (376–7), an "enterprise," moreover, in which currently the balance has tipped in literature's direction. This is because, as Porush continues,

First, postmodernism places the self-conscious activities of the human observer/scientist/teller—and consequently the making of narratives—in the center of things. Second, postmodernism stresses the paradoxical power of structures of information and codes. That is, while the postmodern position states that codes create reality , postmodernism does not trust codes to tell the whole truth. . . . they do a good job of delivering information, but they are less successful at capturing an underlying inexpressible, inchoate, silent realm where meaning resides. (377)

As Rorty and others have cheerfully submitted philosophy to literature, finding it advantageous for one reason or another to abandon the hopes—as I have said above—of a variety of logics and dialectics to the situation of an all-pervasive person- and occasion-governed rhetoric—so, now, Porush will bring science into (what he understands as) literature's empire. I will only a little worry the question of what "truth" can possibly mean in comments like these.^[1] For if one really does believe that "codes create reality ," it is difficult to see how "truth" would be an operative term at all. How could such a view establish criteria for determining "which epistemological activity does a better job of telling the truth" (373, my emphasis), or for distinguishing "the whole truth" (377, my emphasis) from what James Clifford calls "partial [truths]—committed and incomplete?" (1986 7) "Truth" for Porush predictably elides to "meaning," but even so, it may be worth noting that, so far as literary aspirations in these regards are concerned, secular literature has only occasionally (and mostly from the latter eighteenth century forward) been interested in incursions into "underlying inexpressible, inchoate, silent realm[s]," in the interest of "capturing"—but I must here remark the belligerently imperial nature of Porush's discourse—the transcodal, as it were—for all that literature's capacity successfully to colonize such "realm[s]"—may currently be more attractive to "scientists" than it has heretofore been.

Porush does not question the premise that "meaning resides" in the "underlying . . . silent realm" he posits, sim-

ply taking it as axiomatic that whatever we may mean by meaning does, in point of fact, live there . But his assumption, which I would characterize as a turn to mystico-religious and away from secular perspectives, is no doubt shared by many today in the context of postmodernism. Porush's reference to "literature and science" as "epistemological activit[ies]" (my emphasis) goes a ways toward explaining why it is that many of the social scientific workers called ethnographers have, of late, become particularly interested in literature. (Literary people have become interested in what the ethnographers do, as well—but less so.)

An obvious starting point for the current convergence is Clifford Geertz's self-interrogation of 1973. Asking, what does the ethnographer do? Geertz answered, he writes.^[2] (I assume she does, too.) If it is indeed writing that is the ethnographer's central or preeminent activity, then clearly it is to the practice and product of that activity that we must attend, to its "signature" or individual author-specific nature, as Geertz would further state, and its "discourse" or style-specific nature (1988 9). Thus in 1982 we have the publication of an essay by Dick Cushman and George Marcus called "Ethnographies as Texts." Here, Cushman and Marcus, with all the excitement of Molière's bourgeois gentilhomme discovering that he has been speaking prose all his life, recognized that ethnographies were, indeed, texts, and that therefore "rhetorical analysis is prior to an evaluation of truth claims [the ostensible priority of "science"] because explanation and theory building cannot escape the rhetoric of the language in which they are expressed" (56n).

This implicit acknowledgment of a necessary attention to literary matters came more then fifteen years after Roland

Barthes, Jacques Derrida, and Jacques Lacan, among others, at a Johns Hopkins University conference,^[3] had announced to the Americans that just about everything was texte . And it came nine years after Hayden White's Metahistory had denied the possibility of any "truth claims" in the writing of history, so far as these might be evaluated according to more or less agreed-upon standards of verifiability or falsifiability. For White, there simply could be no objectively valid "truth claims" in a discipline like history, or, for that matter, in a discipline like ethnography "not yet reduced (or elevated) to the status of a genuine science [e.g., with agreed upon definitions and quantifiable data, where] . . . truth remains the captive of the linguistic mode in which it seeks to grasp the outline of objects inhabiting its field of perception" (1973 xi). This subordination of logical and empirical operations to the rhetorical, for White, meant that we were constrained to base our acceptance or rejection of any social scientific account whatever on strictly esthetic or moral grounds, for "there are no apodictically certain theoretical grounds on which one can legitimately claim an authority for any one of the [possible modes of historiography] . . . over the others as being more 'realistic"' (1973 xii), that is to say, more scientific. As White puts it in the epigraph (from Roland Barthes) to his most recent book, "Le fait n'a jamais qu'une existence linguistique" : a fact never has anything but linguistic existence (1987 n.p.). It should be noted here that the epistemology of science all these writers use—perhaps unconsciously, or perhaps strictly for polemical purposes—is essentially a Newtonian or "realistic"/positivistic one. The relativistic

paradigms of Einsteinian/Heisenbergian science of the 1920s, the response to relativist science in the cybernetic revolution of the 1940s, and the various ongoing projects of postmodernist science, all go unmentioned.

Fully aware of the radical, or as the poststructural theorists of the period would (cheerfully enough) say, the abysmal implications of White's confinement of history to the "prison house of language"—a full-blown postmodernist move—Cushman and Marcus referred their readers to "sources of literary criticism that might be useful to a perspective on ethnographic writing," among these the work of Barthes, Derrida, Michel Foucault, and Raymond Williams. (But Williams's insights, it seems to me, would not easily be consistent with those of his cited French contemporaries.) And Cushman and Marcus briefly surveyed what they called "Ethnographic Writing Experiments" (58) as these seemed to mark a dissatisfaction with what they called "realist ethnography."

In 1986, Marcus, now with James Clifford, used the etymology of the word ethnography itself for the title of their coedited book of essays titled Writing Culture: The Poetics and Politics of Ethnography . Although the tone of the introduction and in some of the contributor's essays tends to be one of dismissing or, at least, of passing beyond Geertz, still, the editors and contributors essentially develop and expand upon the implications of Geertz's determination that what is central to ethnographic practice is ultimately the textualization of interpretations deriving from particular experiences/encounters. In 1987, we have the publication of Anthropology as Cultural Critique by Marcus, once again, and Michael Fischer, which, among its many inadequately theorized assertions and recommendations, turns to the Russian Formalist critics of literature of the early part

of this century, and takes their central concept, ostranenie , translated usually as defamiliarization or, simply, making strange , as authorizing experimentation and innovation in the writing of ethnography—a practice that might or might not (Marcus and Fischer are curiously timid radicals, in these regards, as Marcus and Cushman were earlier) turn ethnographic writing from its traditional status as a "document of the occult" to an "occult document." I have referred once again, here, to the title of Stephen Tyler's essay in Writing Culture , one that defines the postmodern project of ethnography as indeed a shift from the production of documents of the occult to the production of occult documents.

Such documents—they would be conformable to Porush's enthusiasms—would subsume all claims to "scientific" realism or "truth" to an essentially "literary" evocativeness. In Tyler's words, a postmodernist ethnography "describes no knowledge and produces no action," instead transcending these "by evoking what cannot be known discursively or performed perfectly." This "Evocation," for Tyler, "is neither presentation nor representation;" it is "beyond truth and immune to the judgment of performance" (123). Occult document to be sure: for if one is to take Tyler seriously, ethnography's postmodern move beyond science (again, a science that is a straw man, one whose positivistic paradigm has long since been abandoned in theory, if not always in practice) here passes beyond all literature except that which specifically attempts to give voice to the mystico-religious "silent realm" of the inexpressible, and inchoate—which, indeed, in David Porush's account, the postmodernist scientist also values.

Of course, by 1973 and 1982, my points of beginning for this subject, a great many other developments had already

occurred to prepare the current convergence of ethnographic and literary concerns. No account of this subject ought leave out the extraordinary conference held at Indiana University in 1958, which brought together anthropologists like the young Dell Hymes and Conrad Voegelin, literary critics of international reputation like René Wellek, I. A. Richards, and William Wimsatt, and, of course, linguists like Roman Jackobson, and Thomas Sebeok of Indiana's Research Center in Anthropology, Folklore, and Linguistics. The volume of papers from the conference edited by Sebeok and published in 1960 contained the influential "Concluding Statement" by Jakobson titled "Linguistics and Poetics." In it, Jakobson called for collaboration between the linguist and the literary critic, a call not for the estheticization of science, but, to the contrary, for certain scientization of literature. Only two years later, Jakobson would collaborate with an old friend, none other than Claude Lévi-Strauss, on an analysis of Charles Baudelaire's sonnet, "Les Chats." The essay was published in the French journal, L'Homme . In introducing their essay, the authors acknowledged that

It will perhaps be found surprising that a review of anthropology should publish a study devoted to a French poem of the XIXth century. Yet the explanation is simple. A linguist and an ethnologist decided to combine their efforts in an attempt to understand the creation of a Baudelairian sonnet, because they had found themselves independently confronted by complementary problems. (In DeGeorge and DeGeorge 125)

These "problems," of course, seemed amenable to solution by the methods of what was called structuralism, methods that, at least for a time, appeared useful equally to linguists,

literary critics, and ethnologists. Lévi-Strauss's use of these methods for the study of myth, and, as well, his apparent willingness to intermix such seemingly disparate genres as autobiography and ethnography in texts like Tristes Tropiques , appealed for obvious enough reasons to literary critics. One reason for the appeal of linguistic and ethnographic structuralism, such as it was, to American literary critics was a certain fatigue with New Critical textual explication, the dominant critical paradigm from the latter twenties into the fifties.

There is no doubt that the New Criticism, whatever else it did or didn't do, taught generations of American students of literature to read carefully, and to find citable bases in the language of a poem (more rarely of a story or novel) for what E. D. Hirsch, by 1960, was calling its "meaning." It did little, however, to help the critic determine the "significance"—I again cite Hirsch, who attributes his understanding of the distinction between "meaning" and "significance" to Husserl—of that meaning.^[4] For help in that regard, prior to the appearance of structuralism (which, indeed, offered no help at all in these regards), the alternatives to New Critical close reading were Northrop Frye's version of myth criticism, announced in The Anatomy of Criticism , in 1957, and a variety of contextualist, historicist approaches, either Leavisite or more homegrown. (I am thinking of the work of Roy Harvey Pearce and a number of Americanists, particularly.) The first of these seemed to lead directly to religion, the second to sociology, or even—dreaded fate—to politics . (This is, recall, the late fifties and

earlier sixties.) No wonder, then, that imported linguistic and anthropological structuralism might be appealing.

So appealing that Susan Sontag, in 1966, could call her essay examining Lévi-Strauss's early work, "The Anthropologist as Hero." What Sontag did not like about Lévi-Strauss's work, interestingly enough, was its refusal to relinquish claims to some kind of scientific status—a status Frye's own system most certainly courted. Throughout the fifties and sixties, I would suggest, there was an ambivalence—reactions, that is, of attraction and repulsion—on the part of literary people to science, but a clearly growing attraction to literature (with a concomitant rejection of science) on the part of social scientists.

Thus, for the student of literature in the sixties there were a number of anthropologist heroes—or at least ethnographers who worked in ways that (themselves influenced by literary study) appeared congenial to literary critics. I will mention only Victor Turner's dramatistic and performative studies of ritual, and Clifford Geertz's exposition, essay by essay, of an interpretive ethnography, anthropology as "thick description," in the phrase Geertz took from the philosopher Gilbert Ryle. By 1973, in "Thick Description: Toward an Interpretive Theory of Culture," Geertz could explicitly announce, "The concept of culture I espouse . . . is essentially a semiotic one . . . and the analysis of it to be therefore not an experimental science in search of law but an interpretive one in search of meaning. It is explication I am after" (5). It is one of the ironies of anthropology's recent history in America, I suppose, that Geertz should specifically criticize, for his espousal of an "ethnoscience, componential analysis, or cognitive anthropology" (1973 11) still "in search of law," none other than

Stephen Tyler, whose attraction to a transcendent and evocative ethnography has all the passion of a recent conversion.

Of course Tyler, in his earlier enthusiasm for truth and law in anthropology, along with a number of other ethnoscientists or philosophers of the social sciences such as May Brodbeck,^[5] was only carrying forward a call to make anthropology a bona fide science that was little more than a half-century old. And as these people did so, in the fifties, they encountered what I have remarked in Geertz, a strong reaction against claims to science and law in anthropology and a turn to semiotics, hermeneutics, and interpretation. This, to repeat, occurred at exactly the moment when literary people themselves became fascinated by semiotics, hermeneutics, and interpretation, most particularly of a structuralist sort—which, paradoxically enough, they partly valued for the scientific authority or at least the scientific aura structuralist methods appeared to bestow on literary studies. To employ Raymond Williams's useful terms once more, we have a curious mix of "residual," "emergent," and "dominant" values, where "science," as I believe, is the dominant disciplinary value of the period, and where (vaguely defined) esthetic values—"thickness" in description serving to indicate what once (as we shall see just below) had been called "atmosphere," and which would in our time be called evocativeness—have both a residual and emergent function. This would help explain the appeal of the Canadian Frye and the Frenchman Lévi-Strauss to Americans, inasmuch as both of them invoke scientific categories (for literary criticism and ethnographic explanation) while privileging the study of myth. It would help explain as well the

attraction to semiotics on the part of literary people, for whom this study appears to be far more scientific in its methodology than anything they are used to, while for social scientists, it seems to open the door to interpretive/explicative rather than explanatory results.

I will only add that these disciplinary developments occur in the context of independence movements in the colonized worlds of Africa and Southeast Asia, movements which present the most direct sorts of challenges to "scientific" models of knower and known as epistemological justification for the maintenance of differential power relations, and in the context of American post-Sputnik anxiety, spurring aggressive hard-scientism in the most traditional mode. More needs to be said about both of these, apparently contradictory, developments. Nonetheless, I will, with all necessary, if inadequate, apologies, turn to a more distant history—and an essentially cultural rather than materiopolitical history.

2

Inasmuch as the term literature, from Latin, littera , letter, for years in the West referred to whatever information a given culture wished to preserve and transmit by the technology of writing, it might be said that ethno-graph -y in its earliest manifestations was inevitably a part of literature. Thus, there was a time when all who were literate, who knew their letters, knew also the Latin of Caesar's account of the tripartite nature of Gaul and his descriptions of the Gauls themselves. It seems reasonable to say that in one form or another, ethnography most certainly existed in the classical period. Indeed, Tacitus's De Germania , a study of Teuton people appearing in the year 98 of the Common Era was, as one writer has called it, a political ethnography, for

Tacitus was concerned about the threat that restlessness among the Germans might pose to the Empire's borders. If one wished to show anthropology's relation—perhaps an inevitable, most certainly a very longstanding relation—to imperialism, one could easily begin with Tacitus. (But Tacitus, like Thucydides, was also in some measure offering a criticism of the direction his own society was taking; so if his name may signal an early instance of imperial anthropology it may also signal an early instance of "anthropology as cultural critique.")

Although all writing, almost until the latter eighteenth century, was, as I have said, nominally literature, there did nonetheless exist a distinction between what was then called poetry—or poesie —the term for imaginative and affective writing (literature) and the discourses, most particularly, of history and philosophy. That epistemological distinctions among these types of writing were, or ought to be, clear and well-defined does not seem to have been nearly so important to ancient thought as it was to become later. In Foucault's useful terms, it was not so much whether statements were true ("scientific"), vrai , but whether they were dans le vrai , whether they met the appropriate conditions, epistemologically but also socially—"discursively"—to be taken seriously as making "truth claims." The operative distinction was not between truth (e.g., real, scientific, actual, empirically verifiable, etc.) and fiction (made up, invented, wished, dreamed, imagined, etc.) but, rather, between truth and error; rudely, to mix Foucault with Lévi-Strauss, between what was thinkable or "good to think" as opposed to impensable . At least until the late Renaissance, the blurring of genres that Geertz has remarked upon as particularly a development of our time was more nearly the rule than the exception.

Thus the "voyage narratives" of the sixteenth century recounting European explorations of the New World, are an extraordinary mix of "ethnography" and "literature," of what probably appears to the modern reader as careful observation and description—the truth—on the one hand, and of the wildest assertions, which, from our present perspective, couldn't possibly be true. Yet what strikes us as inevitably "poetical" or fantastic imaginings in these texts—encounters with people ten feet tall, with four arms, or breathing fire—are also offered on the "ethnographic authority," in James Clifford's justly celebrated phrase, of the author's own experience, an experience which, for all our conviction that it "couldn't really have been that way," by conforming to the epistemic assumptions of its age, may perfectly well present "truth claims." (Hence the logic of the Church's rejection of Galileo's "discovery" of sun spots, and the refusal to accept an invitation to look through the glass and "see for oneself" as in any way relevant to the subject at hand.) There were enough of these narratives by the latter sixteenth century to spur the Reverend Richard Hakluyt—for purposes, it should be noted, that were not strictly those of a disinterested curiosity—to begin editing and collating them, a task taken up, after Hakluyt's death, by the Reverend Samuel Purchas, who published his massive Hakluytus Posthumus, or Purchas His Pilgrimes in 1625. As Mary Louise Pratt has recently shown (1986),^[6] the mix of personal "narration" and cultural "description" established in these narratives continues to shape the presentational strategies of ethnographies from early in the twentieth century up to the present, for all that they operate with different epistemic premises than their predecessors;

for all that, at least until the current postmodern reaction, they might claim to see things "as they really were," scientifically, realistically, and so on.

Responses to the seventeenth-century voyage narratives prompted what may be the earliest, or at least they are the best known, enunciations of what we would now call positions of cultural relativism and cultural absolutism or evolutionism, and both are by authors firmly established as "literary." I refer, first, to Michel de Montaigne, whose celebrated essay on the cannibals rhetorically asks how a people who punish by use of the rack and execute by drawing and quartering (this latter practice seized upon by Foucault for the beginning of his Discipline and Punish ) can denounce as "primitive" those who merely eat their slain enemies. And I refer to William Shakespeare, whose play The Tempest , known to have been performed in 1611, anagrammatizes the term cannibal for the character of Caliban, who is presented as man in the state of nature, a filthy, lustful brute, far inferior to the cultivated human products of civilization.^[7]

Perhaps the first self-conscious attempts to separate literature and ethnography, rigorously to distinguish between art and science, the imagined and the real, fact and fiction occur in the eighteenth century. Broadly speaking, these define science in the ways we continue to know it best, in terms of its commitment to sequential ratiocinative steps, quantitative and methodological documentation, and to procedures that are at least theoretically replicable. This is to say that although the observer's unique personal experience could convey upon her account a certain ontological

authority—the être là , or simple fact of her "being there"—its scientific (epistemological) authority would depend precisely upon the absence of its uniqueness; had others been there, they would have seen and concluded the same. To offer a single, well-known American example, I will mention Thomas Jefferson's celebrated reply, in his notes on the State of Virginia , published in the latter 1780s, to the Comte de Buffon, concerning the "productions mineral, vegetable, and animal" of the New World. In this text, Jefferson refuses to speak of anything he has not seen himself—and he invites other interested parties to come and see for themselves. It is not the richness of his deductive powers that is to be celebrated but, as it were, the richness of the material itself—material, he is proud to imply, that is peculiar to America.

This is hardly an adequate account of Jefferson as naturalist-scientist. I mean my few words to indicate the way in which, from Jefferson forward, America will stand as one of the world's foremost laboratories for anthropological science, a science Americans proceed to establish upon the basis of detailed first-hand study of the Indian. Through the nineteenth century, in the researches of Cadwallader Colden, and, particularly, of Henry Rowe Schoolcraft, a massive amount of information (some of it accurate, some of it not) on the languages and cultures of Native peoples began to be assembled, culminating, from the middle of the nineteenth century forward, in the work of Lewis Henry Morgan.

Morgan was an amateur or freelancer, a retired lawyer inspired by enthusiasm and a growing curiosity; he was not, that is, like the older voyager, or missionary, or entrepreneurial writer, an ethnographer only as an accident of his employment by a church or government, a land sale or fur-

trading company. Morgan might, therefore, claim for his work the sort of disinterestedness that was to become one of the cornerstones of the new commitment to objectivity and science in American ethnography toward the end of the nineteenth century. Thus, those like James Mooney who did work for interested parties—government agencies or the great museums—still could demonstrate their commitment to science by the suppression of subjective commentary of an ethical or esthetic nature, by the assemblage of massive quantities of detail observed firsthand, and by a tone or rhetorical stance in their published texts that confirmed the distance between themselves and the subjects of their researches. It seems to me that it is a failure or a refusal to respect most of these criteria that still makes Frank Hamilton Cushing, for example, a rather enigmatic figure in the history of American ethnography. (David Murray's Forked Tongues has much to say about this—as it does about translation, autobiography, and other matters I discuss. Murray's fine book appeared while this one was in press so that, except for this parenthesis, I do not make reference to it.)

A fuller account of these matters than I can give here is to be found in such volumes as the 1974 Proceedings of the American Ethnological Society edited by John Murra and called American Anthropology: The Early Years ; in the studies of Regna Darnell; in Robert Bieder's Science Encounters the Indian ; and in George Stocking's masterful Victorian Anthropology . ^[8] What cannot be left out, of course, is the advent of Franz Boas, whose name—probably more accurately than not (for all that I interrogate the nature of

Boas's contribution to anthropology as a science in the chapter that follows)—is synonymous with the scientization of anthropology, according to a largely positivistic model intended to parallel if not quite mirror the meaning of science in pre-Einsteinian physics. Of course, in the years of Boas's mature production, Einsteinian relativism (if not Heisenbergian uncertainty) was very much part of the intellectual climate, and cultural relativism was from the first a cornerstone of the Boasian program. But I see not so much as a hint of epistemological relativism among the Boasians. This is yet another subject I can only glance at here. Suffice it to say that Boas's privileging of fieldwork over library work; his insistence upon some fair degree of competence in the language of his or her "people" on the part of the ethnographer; his rejection of the broadly deductive generalizations indulged in by the evolutionist practitioners of the so-called "comparative method," in favor of an inductive, particularist, and rigorously relativist method: all of these were linked to attaching to ethnographic work the authority of the hard sciences.

The scientization of ethnography under Boas paralleled its professionalization as this was directly linked to its academicization, its institutionalization, in the University. Boas, to be sure, began his career with the Berlin museum and had associations in this country with the Field and the American museums, but his enormous influence is directly associated with his almost half-century-long tenure (1896–1942) at Columbia University—during which time the government ethologists (particularly in the Bureau of American Ethnology under Major John Wesley Powell) and the museum ethnologists declined in influence. Only Clark Wissler, so far as I know, was both a Boasian and a "museum

man." By 1926, as George Stocking has noted, every academic department of anthropology in the United States was headed by one of Boas's students.^[9]

Now, for all that was gained by the linkage of anthropological professionalism and anthropological academism in the name, to be sure, of anthropological science, there was one important loss, the consideration of which will bring us back to literature. I mean to point to the fact that the professionalization and academicization of anthropology were achieved at some cost to what I will simply call its public significance and utility. Most of the collections made by the museum anthropologists were, after all, available for public viewing; the ethnographic data gathered by the anthropologists attached to the government bureaus were available for use in making public policy. And, with whatever qualifications we might want to add, Americans did actually go to the museums, did actually feel the government was in some substantial measure "theirs."^[10] But to whom did the data gathered by the Columbia University anthropology department belong? Of what use was it to anyone but the academic anthropologist and his or her students, a number of whom would themselves perhaps become anthropologists, and study and teach, turn and turn again. It is exactly the success of anthropological science in achieving the status of disinterested professionalism that threatens to render it trivial or irrelevant. The issue to be addressed here is not epistemological so much as sociopolitical. Boasian social scientists, as I shall have further occasion to explain in the following essay, are secure, or, at the least, highly optimistic

about the anthropologist's ability to arrive at the scientific truth; what they decently wonder about is the social uses of such truth.

Two sorts of response to the problem of academic anthropology and social significance seem to have developed. One was given in Boas's own citizenly practice. For although Boas pretty thoroughly avoided the statement of theoretical implications or the suggestion of practical applications in the vast majority of his published work, some of that work vigorously deployed scientific findings to combat, for example, racism and anti-Semitism in American society. Boas believed in the public role of the academic, professional scientist; he wrote letters to editors, and spoke out on behalf of his beliefs and principles, never as a scientist—for Boas took for granted, in a way I think we cannot any longer, the value-free nature of science—but always as a citizen informed by science. Academic anthropology, then, could show the way toward a more clear-sighted society, founded upon the best modern scientific information rather than upon ancient prejudice. The career of Margaret Mead, Boas's student, may, in the future, remain most interesting to us precisely as developing this line of response; that is to say, whatever the academic-scientific status of Mead's data from Samoa, the social force of her prescriptions ostensibly based upon that data was substantial.

Another sort of response to the separation of academic science from public significance came in the attempt on the part of some anthropologists to engage the interest of an audience not strictly professional by attempting literary forms of writing, couching some of their observations about Native people and cultures in autobiographies elicited from them, and, most particularly, in fictional narratives about those cultures and people. Whatever the philosophical in-

fluences on Boas's thought and on that of his students, I believe that the view of literature they had would have inclined them toward the traditional capacity of literature to provoke the reader to moral imagination, and thence to moral action. They would not, if I am at all correct, have seen the function of literature as merely to entertain or amuse, not even to put one transcendentally in touch with the "beautiful." If this is so, then the esthetic of the Boasian milieu would have been as old-fashioned as its epistemology, for—as I shall have further occasion to note—it would be based upon romantic-realistic perspectives at just the moment when these were being abandoned for modernist perspectives. To say this, I hope it is clear, is not to say that either the esthetic or the epistemology was hopelessly mistaken. Fuller biographical work on Boas, Radin, and Kroeber is necessary to determine whether the speculative intuitions I have offered are more accurate than not.

As early as 1913, Paul Radin, of the first generation of Boas's students, had published "The Personal Reminiscences of a Winnebago Indian"; this was followed by "The Autobiography of a Winnebago Indian" in 1920. In the introduction to that text, Radin noted that "For a long time most ethnologists have realized that the lack of 'atmosphere' in their descriptions is a very serious and fundamental defect," a defect, according to Radin, that "could only be properly remedied by having a native himself give an account of his particular culture" (1920 1)—which account, to be sure, would be edited and actually published by the anthropologist. But Radin had published both "The Personal Reminiscences" and "The Autobiography" in scholarly journals, not the best way to reach a broad audience. Accordingly, he revised and expanded the 1920 "Autobiography," composed an introduction which began with the acknowledgment that

"the common-sense man, the man in the street, has always been good-naturedly skeptical of the academically trained scholar" (xv), and published it as Crashing Thunder: The Autobiography of an American Indian in 1926 with the commercial publisher D. Appleton and Company of New York.^[11] Although Boas himself, late in his life, was to deny the scientific usefulness of Indian autobiographies, noting them as good only for documenting "the perversion of truth by memory" (1943 335), a great many anthropological life histories continued to be recorded.

The other important linkage of ethnography and literature I want to mention comes through what I will call ethnographic fiction, a literary genre with ostensibly wider public appeal than the (developing) genres of professional ethnography. Here the example of Adolph Bandelier's novel The Delight Makers , published in 1890, serves as an important precursor. In his preface, Bandelier wrote that he

was prompted to perform the work by a conviction that however scientific works may tell the truth about the Indian, they exercise always a limited influence upon the general public; and to that public, in our country as well as abroad, the Indian has remained as good as unknown. By clothing sober facts in the garb of romance I have hoped to make the "Truth about the Pueblo Indians" more accessible to the public in general. (xxi)

We may compare this statement to that of Boas's student Elsie Clews Parsons, in her preface to a volume she edited in 1922 called American Indian Life . In that book, Parsons offered no less than twenty-seven short fictional pieces by professional anthropologists focused on the cultures about which they had formerly published scientific papers. The final text is by Boas himself, his unique attempt, so far as I am aware, at writing fiction. Parsons asks:

Between these forbidding monographs [by the scientific anthropologists] and the legends of James Fenimore Cooper, what is there to read for a girl . . . or, in fact, for anyone who just wants to know more about Indians? (1)

It was "From these considerations," Parsons explains, that "this book was conceived." Like Bandelier before her, Parsons, like Boas's students generally, firmly adhered to a rigorous separation of literary "legends" from ethnographic "science," while being attracted to the possibility that science might use the novel's (apparently predictable and foreknown!) form and manner for its own purposes, purposes which might, indeed, be "just . . . to know more," for all that I would again propose that knowing more implies doing better. The important thing, as Alfred Kroeber, the first to take a Ph.D. in anthropology with Boas at Columbia, remarked in his introduction to Parsons's book, is to adopt "The method of . . . the historical novel, with emphasis on the history rather than the romance" (Parsons 13). In that introduction, Kroeber went on to praise Bandelier's Delight Makers —for all that Bandelier himself referred to his book not as a "historical novel" but as a romance.

What may be interesting to note from the point of view of literary history is that the anthropologists in their ele-

vation of the novel over the "romance" or "legend" as the literary genre most congenial to—although still quite clearly distinguishable from—the realism of their science, are taking positions in a battle that long ago had been won. From somewhere around the time of Henry James's brief study called Hawthorne , in 1879, the realistic novel in this country began to take on an authority vastly superior to the romance in American literature. Bandelier's accommodation of his text to the romance is, in this regard, similarly belated; by 1890, romance means Cooper and Longfellow, if not necessarily Hawthorne, as they represent a tradition thoroughly out of fashion in an age ready to admire William Dean Howells, Henry James, and Edith Wharton. And by 1922, when Kroeber indicates his own approbation of the novel over the romance in his introduction to Parsons's book, even the realistic novel is in eclipse; 1922 is the year James Joyce's Ulysses , that masterpiece of a modernism bent on subverting the pretensions of realist fiction, was published. But it is also the year Bronislaw Malinowski's Argonauts of the Western Pacific was published, a text which has been said to instantiate a "modernist" anthropology, which, here, means a "scientific" anthropology, in line with a largely abandoned realism in the novel (not to say a realist/positivist science that atomic physics with its relativist findings also eclipses). To complicate matters further, we may note that 1922 is also the year T. S. Eliot's "The Waste Land" was published. Eliot's poem owed much to the work of Sir James Frazer—whose particular sort of "armchair anthropology" Malinowskian anthropology supposedly superseded.

This "out-of-phaseness,"^[12] as the anthropologist Edwin Ardener has called it, between literature and ethnography,

even as literature and ethnography importantly influenced one another all through the twentieth century, currently continues in Marcus and Fischer's turn to the Russian Formalists and to their encouragement of what they (Marcus and Fischer) call modernist experimentation in contemporary ethnographic writing. To call for modernism today is quite as belated as Bandelier's turn to the literary romance or Kroeber's to the historical novel, inasmuch as the experiments of modernism have themselves, for many, become old hat, monumentalized and stultified, so that the genuinely new appears to take form in a postmodernism that is itself already more than two decades old—a postmodernism that may be the continuator of a modernism whose revolutionary potential has been more nearly occluded than ended, for all that most of the present-day ethnographers—Tyler, and others—interested in it have embraced only the postmodern break with modernism (see chapter 2).

Not only Parsons, but Ruth Underhill and Ella Cara Deloria, among anthropologists in the Boasian milieu, wrote ethnographic novels, always marking these off, as Boas insisted they could and should be marked, from their truly "scientific" work. Thus in 1939, Underhill published a scientifically proper Papago ethnography, the Social Organization of Papago Indians , and followed it, in 1940, with a novel about these people, Hawk Over Whirlpools . Deloria, a Dakota from the Yankton Reservation, who had worked with Boas in 1915 and 1928, after a great many years devoted to the scientific study of Dakota language and culture, sometime in the 1940s set to work on the novel that would be called Waterlily (1988). According to Raymond DeMallie, in his afterword to the book, it was Boas and Ruth Benedict who suggested she attempt the work. Other of Boas's students, in particular Ruth Benedict and Edward Sapir, also had a strong penchant for literary work, both of

them writing and exchanging poetry (some of which Benedict published under the name of Anne Singleton). Sapir's book of poems, Dreams and Gibes , appeared in 1917.

All of those I have named, whatever their attraction to literary pursuits, did keep their art distinct from their scientific pursuits, pursuits which, for the most part, defined their working lives. Nor is it clear whether—or how much—the injunction to keep literature and ethnography, art and science apart caused these second-generation Boasians pain. This is not the case with the last of Boas's students I must mention here, Zora Neale Hurston. Hurston studied with Boas and with Benedict at Barnard College, from which she graduated in 1928. Although she admired "Papa Franz," whom she referred to as the greatest anthropologist alive, Hurston's major was English, not anthropology; and her major achievement came in the writing of literature, not ethnography. Thus a text like Hurston's novel, Their Eyes Were Watching God , for all that it is rich in ethnographic detail, could not fairly be categorized strictly as ethnographic fiction—an appropriate enough appellative for The Delight Makers, Hawk Over Whirlpools , or Waterlily , all of which, to be sure, are books I admire. Unlike these latter, Hurston's novel is more nearly modernist in its esthetic—its commitment to verbal and linguistic richness and ingenuity, and a "poetics" very different from realist poetics or prosaics—and in its intentionality, at least partly animated by the concern of her friends associated with the Harlem Renaissance, to show that the "negro" may be the subject of "high" literary culture.

To call Their Eyes Were Watching God a "masterpiece" rather than a "mere" example of "ethnographic fiction" would require an excursus on evaluative standards that would take us too far afield just here. Perhaps it may be all right, if not quite adequate, to say that the range of its

possible interest to the reader—its portrait of Florida's rural black culture, its version of a highly elaborated vernacular speech with a determinedly poetic narrative style privileging the spoken over the written, and, of course, its questioning of accepted gender relations—qualify it for comparison to the most interesting American novels, and make a strong claim for its inclusion in the canon of American literature.

Even in the ethnography she published, Hurston regularly threatened to collapse what was still seen as the hard-won distinction between science and art. Her Mules and Men , a study of southern black folklore published in 1935, was legitimated as "science" thanks to a single-page preface provided by Boas—after he had authenticated all the facts. Hurston's Tell My Horse , of 1938, an account of visits to Jamaica and Haiti (where she wrote Their Eyes Were Watching God ), received no such legitimation—indeed, its manner of proceeding did not permit that—and as a result, as late as 1977, Robert Hemenway, Hurston's first (and, to date, her only) biographer could call Tell My Horse Hurston's poorest book. It is exactly this book, however, that seems to me likely to be of special interest to ethnographic and literary theorists. Hurston's persistent discomfort with the stance of detached, scientific, professionally disinterested observation, a stance upon which Boas insisted, now may be seen as contributing to a redefinition of ethnography, one that places it more comfortably in our present moment of ethnographic and literary convergence than in its actual historical moment.

3

The most recent convergence of ethnographic and literary concerns, to return to the point from which I began, does

indeed take place within the context of postmodernism. Scientific claims to truth are threatened not only by rhetoric and ideology but by indeterminacy and interconnectedness, these latter two terms reminding us of David Porush's description of the milieu of postmodern science. The Heisenberg indeterminacy principle,^[13] it would seem, operates not only at the subatomic levels of physics, but everywhere, with the consequence that even as would-be disinterested observers we inevitably enter into relation with and have effect upon whatever it is that we observe. And now, beyond Heisenberg, there is Bell's Interconnectedness Theorem implying that—I know this only in Porush's account—local operations have nonlocal effects: everything we do and say, perhaps everything we think, alters the universe, an exhilarating or a frightening prospect.

The postmodern perception of a world organized in terms of signal/noise or figure/ground relations that are constantly shifting doesn't only blur genres, as Geertz might have it; rather, as I have several times noted, it blurs epistemological distinctions, asserting a kind of cognitive egalitarianism on the part of literature, ethnography, and even—at least at the highest theoretical level—the physical sciences. This particular sort of equivalence between truth and beauty is not at all of the sort apparently sanctioned by Keats's great lines, "'Beauty is truth, truth beauty,'—that is all / Ye know on earth, and all ye need to know." For Keats, as indeed for Shelley in his Defence of Poetry , the identity of truth and beauty resided not in an equivalence

in the type of knowledge/experience each produced, but in a potential equivalence of function ; both truth and beauty were, each in its own way, conducive to the achievement of that more capaciously informed consciousness which—so it was hoped—might result in more just and generous living.

David Porush observes that the postmodern privileging of literary discourse comes as the result of literature's traditional ability to express "a vision of the beauty not only of order but of disorder" (388), something that is more difficult for the physical or social scientist. Porush further notes as "cybernetic laws" that humans 1) find it compulsory to interpret or disambiguate uncertainty—uncertainty I take as synonymous with a perception of disorder—and that 2) one man's noise is another man's signal, as, indeed, one person's figure is another's ground (396). But this latter observation makes problematic Porush's earlier claim: if anyone's figure may be ground for anyone else, how, then, to establish a distinction between "order" and "disorder" stable enough to support "a vision" of one or the other? How to determine that that vision does indeed express the "beauty" of order or disorder, inasmuch as "beauty" would have to be defined differentially in contradistinction to something that could be called ugliness, or some such? In the same way, Porush's overexcited suggestion that "any exchange of information creates a narrative" (379) seems to me also mistaken. For the term "narrative," like "beauty," or "order" represents a determination as to what counts as signal or figure; narrative, beauty, and order are sociolinguistic constructs, which is to say that only those exchanges of information we take as fulfilling the conditions we posit for narrativity can be taken to constitute a narrative.

I say all this by way of instantiating the importance of human agency. This is not, as I have had occasion to remark

earlier, to reinvoke humanistic claims to autonomous individualism or a fully self-present subjectivity. Rather, it is to centralize the observation that inasmuch as figure and ground, signal and noise may, indeed, be constantly shifting; inasmuch as no "apodictically certain grounds exist" (H. White 1973 xii) for our determinations as to which is which, that each of us must bear the responsibility at any given moment for our choices in these matters. (The influence of an older and currently largely disregarded existentialism should be obvious here.) In the same way, although it is no longer possible to believe that we can literally represent "reality," "history," or "truth," it still makes a difference whether one chooses or refuses to take it as axiomatic that there is, nonetheless, an aprioristic material reality, of whose history we can more or less speak, in a manner positing truth as a value.

I will here only repeat what I have said earlier, with reference, however, to anthropological theorists particularly. I believe, that is, with the anthropologist Jacques Macquet,^[14] that if we cannot be objective, we can still be scientific in ways that allow that word coherent meaningfulness—by specifying the methods and procedures we have followed, by indicating the empirical as well as logical components of our arguments, and so on. I believe with Marvin Harris in "the struggle for a science of culture,"^[15] whether that science is defined solely in Harris's terms or not. No one can doubt that such a science will be more modest and very different from what science has heretofore been in the West—but that strikes me as a good thing. I am reluctantly convinced, mostly by the work of Hayden White, that at

the highest levels of interpretation there are no secure epistemological grounds for preferring one account of sociocultural phenomena as more "realistic," hence more nearly "true" or "scientific," than another. For White, therefore, our preferences have the authority only of choices, truth becoming now not a matter of proof but of value, of our own sense of the beautiful or the good. While this may be the case at some ultimate theoretical horizon of these matters, where rhetoric rules with totalized effectivity, and the linguistic circle has no break in its circumference, I return, again and again, to Raymond Williams's sense of the way in which the very nature of cultural hegemony is such that it cannot help but permit breaks, blanks, holes, areas weakly (or un-) colonized, with room, thus, for "residual" and "emergent" elements to have a certain play. I suggest, at least by way of analogy, that a similar situation prevails in the epistemological realm, with the result that there may indeed be a good deal of practically occupiable space—short of the absolute theoretical horizon of cognitive and esthetic equivalences—space in which probabilistic and tentative statements that offer themselves as more nearly true than others, if not absolutely true, might yet be made effectively.^[16]

Curiously, even Porush admits that at a practical, day-to-

day level, scientists operate and, indeed, achieve results in spite of the constraints imposed by the highest theoretical findings about an increasingly slippery and "undecidable" physical universe. In the same way, authors and readers achieve the communication of more or less "decidable" meanings, for all the poststructuralist insistence on the infinite semantic possibilities of the signifier's free play. In all these instances we have—and Anthony Wilden's work of twenty years ago is absolutely essential to an understanding of this matter—an unfortunate but typical Western capitalist insistence on digitalizing what is in fact a matter of the analog.

Let me return, by way of closing, to what I only passingly mentioned at the outset, that it is among the ethnographers more than among the literary people that some of these issues are most hotly being debated. No doubt this is because it matters immediately and materially to ethnographic work whether one believes that—I return to the quotation from David Porush—"meaning resides" in some "underlying [realm of] the inexpressible, inchoate, [and] silent" (377), and wishes to evoke that meaning in one's ethnographic writing; or whether one believes, however tentatively—and here I return to Nancy Hartsock—in "the goal of accurate and systematic knowledge about the world" (205), attempting, by whatever moves or means, to approximate to such knowledge. That these things do matter more to the ethnographer, to offer at least one bit of empirical evidence, I'd mention that the number of sessions devoted to literary theory and practice at the American Anthropological Association Convention of 1988 (the last I attended) was proportionally much greater than the number of sessions devoted to anthropological theory and practice at the 1988 Modern Language Association Convention. "New his-

toricist" literary critics, to judge from a recent collection of essays,^[17] have only gotten as far as Geertz; they seem not yet to have discovered Clifford, Marcus, Tyler, Fischer, Rabinow, and others—although it is the case that these just-named students of anthropology, all of whom are fascinated by postmodern developments, seem (more or less) up on the latest in literary theory. Now that literary studies are enjoying a certain return to history, a return by no means limited to "new historicism," it remains to be seen what the anthropologists will do. As the literary canon opens itself and expands, and critics see the necessity of attending to multiple canons, most particularly what I have referred to elsewhere as local, national, and cosmopolitan canons of literature,^[18] they—we—will need the expertise of ethnographers, not so much, to my mind, as colleagues in the decipherment of, and meditation upon, codes, but as providers of data for the understanding of other worlds.