Back to top

Chiefs, Scribes, and Ethnographers

Chiefs, Scribes, and Ethnographers
Kuna Culture from Inside and Out

This sweeping study by a noted anthropologist examines the relationship of the indigenous Kuna of Panama with writing and ethnography over the course of the twentieth century.

Series: William and Bettye Nowlin Endowment, The William and Bettye Nowlin Series in Art, History, and Culture of the Western Hemisphere

November 2009
This is a print-on-demand title. Expedited shipping is not available.
360 pages | 6 x 9 | 15 b&w photos |

The Kuna of Panama, today one of the best known indigenous peoples of Latin America, moved over the course of the twentieth century from orality and isolation towards literacy and an active engagement with the nation and the world. Recognizing the fascination their culture has held for many outsiders, Kuna intellectuals and villagers have collaborated actively with foreign anthropologists to counter anti-Indian prejudice with positive accounts of their people, thus becoming the agents as well as subjects of ethnography. One team of chiefs and secretaries, in particular, independently produced a series of historical and cultural texts, later published in Sweden, that today still constitute the foundation of Kuna ethnography.

As a study of the political uses of literacy, of western representation and indigenous counter-representation, and of the ambivalent inter-cultural dialogue at the heart of ethnography, Chiefs, Scribes, and Ethnographers addresses key issues in contemporary anthropology. It is the story of an extended ethnographic encounter, one involving hundreds of active participants on both sides and continuing today.

  • Preface
  • Acknowledgments
  • Chapter 1. Introduction: Literacy, Representation, and Ethnography
  • Chapter 2. A Flock of Birds: The Coming of Schools and Literacy
  • Chapter 3. Letters of Complaint
  • Chapter 4. Representation and Reply
  • Chapter 5. North American Friends
  • Chapter 6. The Swedish Partnership
  • Chapter 7. Collaborative Ethnography
  • Chapter 8. Post-Rebellion Ethnography, 1925-1950
  • Chapter 9. The Ethnographic Boom, 1950-
  • Chapter 10. Native Ethnography
  • Chapter 11. Chapin's Lament
  • Notes
  • Abbreviations
  • Bibliography
  • Index

James Howe is Professor of Anthropology at the Massachusetts Institute of Technology. He is the author of The Kuna Gathering and A People Who Would Not Kneel and has worked closely with the Kuna for more than thirty-five years.


The study offered here deals with one indigenous people, the San Blas Kuna of Panama, and in particular with three closely related aspects of Kuna engagement with the outside world: with writing and literacy; with representations of indigenous character and culture; and, most of all, with anthropology and its own characteristic form of textual representation, ethnography. Of necessity, I write as an outsider and an anthropologist, caught in the assumptions and practices of my discipline, and also as one of the book's subjects, one of many ethnographers who have written about the Kuna during the last century. I have tried, however, to position my listening post to narrate and make sense of this story as best I could from both inside and out: to treat writing as it was imposed, resisted, and ultimately accepted; to hear what outsiders had to say about the Kuna and what the Kuna had to say about themselves; and to situate writing and ethnography as both impositions and tools of self-knowledge and self-defense.

The Kuna, one of the most studied native societies of Latin America, have for centuries inhabited the Darién Isthmus, which today constitutes the eastern third of the Republic of Panama. Of the several regional Kuna populations, by far the largest, which has grown from roughly ten thousand at the beginning of the twentieth century to more than fifty thousand at its end, occupies four dozen villages scattered along the Caribbean shore on the Coast of San Blas (now officially known as Kuna Yala), most of those villages situated on small inshore coral islets. Supporting themselves with a mixture of subsistence agriculture, cash-cropping, wage labor, and artisanry, the Kuna have maintained an active village democracy, with elective chiefs and nightly council meetings (onmaked), which I call "gatherings," as well as an intense ritual life carried on by named specialists known collectively as "knowers of things" (immar wisimalat). At the same time that they have erected barriers between themselves and the world, the Kuna have crossed those barriers repeatedly to participate in the life of the country that claims them.

In November of 1903, when Panama gained its independence, the Kuna were finishing a century of benign inattention, an epoch of relative peace and isolation following two hundred years of war and struggle on the frontiers of the Spanish empire. As nominal citizens of Colombia, the Kuna engaged in cash-cropping and trade with merchant vessels but largely escaped effective national control. Subjects for centuries of commerce, diplomacy, war, missionization, and exploration, the Kuna of 1903 were still known to outsiders in terms of a few stereotyped words and traits, and no more than a handful of Indian men, all of them from a single village, could read or write.

All this changed drastically with Panamanian independence, beginning almost immediately. The struggles and convulsions that ensued, which reached a climax in 1925 with an uprising against Panama, formed the subject of my previous book, A People Who Would Not Kneel (Howe 1998, 2004b). The present study expands the temporal scope of that work to include the years after 1925, while at the same time narrowing its thematic range from indigenous-state politics in general to matters of discourse, representation, and media of communication—if indeed attention to such globally significant matters can be called narrowing.

In this book, the three topics of writing, representation, and ethnography overlap and interpenetrate, but in the world of theory they have generated large and only partly coterminous literatures notable for their contentiousness, which must be addressed. The reader should know what assumptions and studies have guided my work, why I fail to genuflect before certain altars, and how perhaps the Kuna case bears on the questions in play.

The Politics of Literacy

The fundamental nature of writing, its relationship with oral communication, and its impact on thought and social life have engaged scholars across a broad range of disciplines. Recent work in anthropology and related fields has conventionally been sorted into two opposed camps or tendencies, both of which have their anthropological origin in an essay by Jack Goody and Ian Watt (1963), "The Consequences of Literacy," and in a volume edited by Goody, Literacy in Traditional Societies (1968). The so-called "autonomous model" (Street 1984; Collins and Blot 2003), or "literacy thesis" (Halverson 1991, 1992), associated with the work of David Olson (1994), Walter Ong (1982), and especially Goody (1986, 1987, 2000), treats the divide between orality and literacy in global terms, looking for the cognitive as well as organizational consequences of alphabetic writing. Proponents of the "situated," or "contextual," approach, often advanced in adamant opposition to the autonomous model, insist that writing and literacy, far from exerting universal effects, must always be understood in terms of their embedding in particular cultures and speech communities, each with its own complex mix of oral and written communication.

As much as I admire Goody's ambitious research program and regret the sometimes misguided criticism it has inspired, the present study belongs squarely in the contextualist camp. Like the contextualists (and like Goody in his own ethnographic explorations), I see writing as a social practice thoroughly enmeshed in webs of social and political relations. Like them, I will be concerned with the interplay between oral and written communication. And like them, I see writing as a practice and medium freighted with social meanings and implications for the identities of its users—and perhaps even more for its nonusers.

Writing and Struggle

Though many of the richest studies of schooling and literacy have been ethnographic, the present work follows another line of research from ethnohistory, one equally significant if so far less intensely theorized, which targets the role of writing in struggles between indigenous peoples and state and imperial powers. James Axtell's pathbreaking study, The Invasion Within (1985, 102-104), discussed the place of writing in the seventeenth-century evangelization of the Huron in New France largely in terms of Jesuit manipulation and Huron wonder and awe, an interpretation since critiqued by Peter Wogan (1994). More recent work has emphasized the extent to which colonized peoples have engaged with writing and attempted to neutralize or appropriate its power. James Merrell's fascinating study of eighteenth-century emissaries and mediators on the Pennsylvania frontier, for instance, describes extensive diplomatic use of letters by both colonists and Indians, often in conjunction with the beaded belts called wampum (1999, 193-197, 215-221). Recognizing the power of the written word, Indians sometimes carried letters or decrees they could not read, but they were denied access to treaties and land titles, and when peace broke down irrevocably, few had yet learned to read or write.

A very different picture, of skillful, effective use of writing, emerges from Matthew Restall's study of the colonial Yucatec Maya (1997, 229-319). Village scribes produced documents in their own language, probably from public dictation and very likely drawing on preconquest scriptural traditions as well as a sixteenth-century Franciscan orthography. Among scribal documents, the most numerous were wills and other property records, which reveal a "masterful use of the Spanish legal system . . . to defend their lands" (1997, 234). Also significant were petitions combining professions of loyalty with complaints of suffering and expressions of "relentless hostility" against external authorities (1997, 266), and the books of Chilam Balam (1997, 276-292), mytho-historical histories validating land titles and the interests of indigenous elites.

The most dramatic and best-known story of indigenous success with writing is presented in William McLoughlin's magisterial Cherokee Renascence in the New Republic (1986). Early-nineteenth-century Cherokee leaders made unceasing use of letters, memorials, meeting records, and written laws to enlist white allies, denounce rivals and corrupt Indian agents, further indigenous nationalism, and counter proposals for dispossession and removal. Resisting encroachment in part by demonstrating the civilized nature of the Cherokee, a national council uniting conservative nativists with affluent progressives enacted a series of laws "that together constituted a political revolution" (1986, 284)—an example of what Philip Curtin (2000, 128-155) has called "defensive modernization"—and in the process its literate members perused national periodicals and state laws, published council legislation, and founded their own newspaper.

The modernizing program, in the few years before the Cherokee, however completely they had civilized themselves, were removed across the Mississippi, never resolved internal differences. Despite the extensive use of writing, all of it in English, bilinguals never constituted more than 15 percent of the population, literates 10 percent. In the early 1820s, however, after an illiterate man named George Gess, or Sequoyah, devised a syllabic writing system, "thousands of Cherokees were soon writing notes and letters to each other," helping them "tremendously in overcoming feelings of inferiority and self-doubt" (McLoughlin 1986, 351-352).

Elements from all three cases are readily apparent with the Kuna. Writing and schools initially provoked sharply mixed reactions: some illiterate but forward-looking parents embraced schooling for their children, while many conservatives objected strongly to the perceived threat to orality and gerontocracy. Among the handful of literates, some used their skills to gain power and displace their elders, while others were recruited and co-opted by illiterate chiefs. Kuna leaders bombarded the government with letters and petitions, and like the Cherokee, they eventually adopted written laws and meeting records, leading to a more formalized political system, a written constitution, and security for their lands. Most important in the present study, Kuna scribes recorded their own traditions, in an attempt to influence their representation to the outside world.

The Scriptural Economy

In indigenous-state struggles, writing is not just one more powerful technology like vulcanized rubber or repeating rifles, but a key medium, integral to state, colonial, and national systems, as well as to authoritative conceptions of civilization. Of all the different ways in which linguistic variation has been used to mark social difference, the ability to communicate on paper or its lack has had special salience. Concerning Latin America, this understanding has been advanced most influentially by Angel Rama, who in La ciudad letrada (1996) argues that throughout the colonial era, literacy and its masters, the letrados, played a dominant role in the Spanish empire. "Servants of power, in one sense, the letrados became masters of power, in another" (1996, 24, 22). Just as literacy separated the urban minority from everyone else, mastery of elevated Spanish "crystallized a social hierarchy . . . marking a defensive perimeter between [elites and scribes] and the threatening lower classes" (1996, 33).

With independence from Spain, the "scriptural economy" (Collins and Blot 2003, 129), far from diminishing, grew in size and importance, as Latin American governments, like their counterparts elsewhere, called on schools and literacy to further nationalist sentiment and to create homogeneous populations speaking and writing the national language. If, as Hobsbawm argues, "the era from 1870 to 1914 was above all, in most European countries, the age of the primary school," Latin America was only a few years behind. As Hobsbawm also notes, however, state nationalism was "a double-edged strategy," alienating or rejecting "those who did not belong, or wish to belong, to the nation" (1987, 150). For marginal peoples like the Kuna, as the scriptural economy marked the external boundaries and internal hierarchies of the nation, language and literacy constituted not just "the means by which the battle is fought [but] the site of the battle itself" (Collins and Blot 2003, 131).

Writing and Mediation

Indigenous letrados, the first generations to read and write, have fascinated social scientists and historians, in part perhaps because they remind us of ourselves. The literature on these marginal actors does not always treat writing as a defining attribute of their practice, and literates and illiterates have sometimes played similar roles, but it is schooling that places most of them on the borders between social worlds. They have been categorized and conceptualized in a variety of ways—as go-betweens, cultural and political brokers, courtiers, passeurs culturels, or mediators, middlemen, new leaders, and intellectuals—each categorization promoting a different understanding of their nature and actions.

James Merrell's Into the American Woods (1999), mentioned above, shows that a motley collection of unlettered Indians and mostly literate colonists on the Pennsylvania frontier, recruited on an ad hoc basis and thrust into heated and sometimes intractable situations, succeeded nonetheless for years in keeping the frontier from erupting, typically while avidly pursuing their own private interests. F. G. Bailey's very different Stratagems and Spoils (1969, 167-182) analyzes "middlemen" as points of contact, not between clashing worlds, but between small, encapsulated political units (in his examples, peasant villages in India) and the larger polities encompassing them. Like Merrell, Bailey emphasizes the moral ambiguity, deception, and difficult balancing acts inherent in the work of middlemen.

Between Worlds (1994), Frances Karttunen's absorbing portrait of nine indigenous "guides," "civil servants," and "native informants," while by no means neglecting politics or conquest, puts emphasis on translation and cultural mediation. Even more than Bailey and Merrell, Karttunen points to the marginalization and personal cost suffered by mediators, as well as the gratification and support some of them found in the wider world and the need some felt to explain themselves and their people in writing.

Concerning indigenous Latin American letrados of recent years, much recent work conceptualizes them as indigenous intellectuals and/or as a novel kind of leader in the "new politics of identity" (Brown 1993). As young literate men and women have attempted to create new movements and organizations, some of them specific to a particular ethnicity or region, others national or international in scope, anthropologists have watched closely, cheering on these new leaders but all too aware of their uncertain trajectory as they careen through an obstacle course set by multiple powers, constituencies, and discourses.

Indigenous and subaltern letrados, when categorized as intellectuals, are most often analyzed from a perspective associated with Antonio Gramsci, concerned with the way in which so-called "organic intellectuals" produce the plans, demands, propaganda, and theory connected with political and social action, especially with nationalism and indigenous activism. Despite an arguably superficial engagement on the part of his adherents with Gramsci's thought—few seem to have noticed or cared that he located organic intellectuals in the ranks of the Communist Party, or that he denied that peasants could analyze or critique their situation (Feierman 1990, 32, 18-19)—Gramscian perspectives and concerns have inspired notable work on subaltern intellectuals in Colombia (Rappaport 2005b), Guatemala (Warren 1998; Fischer 2001; and others), Mexico (Gutiérrez 1999; Lomnitz 2001), and elsewhere (Feierman 1990).

Throughout the present work, indigenous intellectuals, both literate scribes and the unlettered chiefs they served, are central characters (see also Martínez Mauri 2007). I am concerned with how the parties to this alliance represented themselves and their people in written communications with government functionaries; how, as mediators, culture brokers, and intellectuals, a number of them collaborated with foreign anthropologists; and how some came to write their own ethnography.

The Problematics of Cultural Representation

As a study of representation even more than of literacy, concerned with the ways in which insiders and outsiders have depicted the Kuna, this book enters a large and contentious field dominated by the figures of Said, Gramsci, and Foucault. These three preeminent theorists of our time, by focusing attention on the social and political work of discourse, the problematic nature of cultural representation, and the connections between power and knowledge, have effected a sea change in social thought. They have made it impossible to consider cultural discourses and representations divorced from either their political contexts or their impact on those they represent. Lamentably, they have also encouraged a monolithic, essentialized, and aprioristic understanding of discourse and ideology, in which questions about dominant ideologies and their effects are often prejudged rather than subjected to scrutiny.

Hegemony, Coherence, and Essentialism

One key question, concerning the degree to which dominant discourses are coherent, consistent, and unchallenged, was emphatically answered by Edward Said's epochal work, Orientalism (1978). More than sensitizing a whole generation of students and academics to the dangers and contradictions of alterity, to the ways in which Western discourses have essentialized and orientalized others, Said argued that a pervasive and comprehensive discourse about "The Orient" has proved so persistent and insidious, so invulnerable to experience or skepticism, that it governs "an entire field of study, imagination, and scholarly institutions—in such a way as to make its avoidance an intellectual and historical impossibility" (1978, 13-14, also 96). As numerous critics have pointed out, however, Said's claims for this discursive juggernaut, able to squash all doubt or contradiction in its path, could be maintained only through the same selectivity of which he accused others.

The essentialism, overgeneralization, and inattention to self-contradiction in Orientalism can also be found in many of the works using Gramsci's notion of hegemony—"the 'spontaneous' consent given by the great masses of the population to the general direction imposed on social life by the dominant fundamental group" (Gramsci 1971, 12)—a concept that comes to anthropology primarily through Raymond Williams (1977, 108-120) and John and Jean Comaroff (1991, 19-27). Although the content of hegemony—"signs and practices, relations and distinctions, images and epistemologies" (Comaroff and Comaroff 1991, 23)—is both broad and vague, there is general agreement, first, that, whatever it might consist of, hegemony is concerned with and constitutive of one group's domination of another, and second, that it is, by definition, naturalized, internalized, and taken for granted, even by the group it subordinates. According to the Comaroffs (1991, 24), ideology is voluble and noisy, while "hegemony, at its most effective, is mute," or as Sahlins (2002, 76) puts it less sympathetically, "Hegemony is supposed to determine not only what one thinks but also what one cannot think."

Despite Scott's powerful attack (1985, 1990) on the claim that subordinate groups accept their domination or that silence can be construed as consent, few enthusiasts of hegemony have faced up to the difficulty of demonstrating the universality and freedom from challenge or doubt that should characterize putatively hegemonic traits. The Comaroffs, to their credit, do recognize the problem, but their solution, that hegemony is "intrinsically unstable, always vulnerable" and that "the hegemonic portion of any dominant ideology may be greater or lesser" (1991, 27, 25), is a theoretical Band-Aid, which never makes clear how anything unstable, vulnerable, partial, challenged, and hard to locate can also be powerful and taken for granted.

The difficulties presented by notions like hegemony, orientalism, and "other terms from the same self-confident semantic family" (Burrow 2000, x) are empirical as well as theoretical, since they depend not just on tendentious text-reading but on spotty and selective sampling from whole universes of discourse—in Sahlins's pithy phrase, "hegemony is homogenizing" (Sahlins 2002, 20). Concerning the supposedly "'hegemonic voices of the West,'" J. H. Elliott notes (1995, 399), "in reality there are many voices, among the conquerors and conquered alike," while Sherry Smith (2000, 6), concerning early-twentieth-century popularizing authors, writes of "a cacophony of voices speaking about Indians."

One might also wonder why more comment or concern has not been provoked by hegemony's family connections with false consciousness and delusion, or by its inherent condescension toward the poor and oppressed. As with false consciousness, part of hegemony's conceptual work is to explain why the revolution has not arrived, why subordinated groups fail to recognize their true interest in the way that we recognize it for them. As Kate Crehan makes clear, moreover, if Gramsci did not portray the consciousness of Italian peasants as absolutely false, he did find it incomplete, deficient, and badly in need of shaping by nonpeasant "organic" intellectuals like himself—an attitude understandable in a Marxist revolutionary in Fascist Italy, but much less so today (Crehan 2002; Feierman 1990, 18-20, 26).

In the case considered in this book, I try to show that the dominant ideologies affecting the Kuna were sometimes stable and consistent (notably concerning the value of writing), but often displayed internal variation and disagreement and, over time, significant change. Higher and lower bureaucrats might agree on the inferiority of Indians but not on its implications; anti-indigenous polemics might be conditioned as much by the situation at hand as by enduring ideology; grudging admiration might sneak into philippics; subaltern groups might reject ideologies but not escape their pain; and what might seem like ineffectual gestures or imperial nostalgia could over time facilitate indigenous revindication.

Silenced Subalterns

The reverse side of hegemony is subaltern muting. A question posed by Gayatri Chakravorty Spivak (1988) in a famously opaque essay, "Can the Subaltern Speak?," was answered in a decisive negative—a denial not just of the presence but of the possibility of a subaltern counterdiscourse—which Spivak based almost entirely on theoretical grounds, through a critical reading of Foucault, Deleuze, Marx, and Derrida. Her thesis, which, as Nicholas Thomas notes (1994, 56), reproduces "imperialism's own sense of its pervasive efficacy," has been emphatically and globally reasserted by John Beverley, a prominent member of the Latin American Subaltern Studies Group. "Spivak's notorious claim . . . is meant to underline the fact that if the subaltern could speak in a way that really mattered to us, that we would feel compelled to listen to, it would not be subaltern" (1999, 66), an argument whose circularity escapes him. For both Beverley and Spivak, the ultimate question is really whether a subaltern voice or consciousness can be identified or represented by the analyst, by the Spivaks and Beverleys of this world. A question about the subaltern, about "them," all too often ends up being about us.

The existence of subaltern voices and discourses, theorists to the contrary, is an empirical problem. Whether one hears them or finds them amenable to analysis does not affect whether words were uttered or written. A great deal of recent publication, brought out by subalterns as well as elites, shows that while Latin American peasants and Indians have regularly been silenced or denied access to schooling and the mass media, they have often had something to say, and that a surprising amount of subaltern discourse can be found if only one listens and looks (e.g., Rappaport 2005a, 2005b; Restall et al. 2005).

Alterity, Othering, and Classification

One of Said's most influential ideas has it that orientalist discourses work by creating a spurious binary distinction between the Western self and a radically different and debased non-Western alter, a process now often referred to as "othering." This view, though no means universal (see Thomas 1994, 51-55), has been widely accepted (see Sax 1998; Rapport and Overing 2000). Cultural and ethnic distinctions are often treated as necessarily hierarchical and invidious, at least in the contemporary world. A related but much more nuanced position stresses distinctions, less between self and other than among non-Western others, a classificatory grid imposed on subject populations through such measures as maps, censuses, surveys, and reports.

That colonial and orientalist discourses make invidious distinctions is in a rather obvious way undeniable. That colonial and national regimes control through classification is significant in nontrivial and suggestive ways. But cultural difference—between self and other, and between various sorts of other—is also real, in both superficial and fundamental senses, and its recognition is not in itself a crime (cf. Abu-Lughod 1991). The obvious objection is that the differences posited by ideology are often false, imaginary, or conducive to misinterpretation, exploitation, and abuse—all true. But no amount of guilty chest-beating or apriorist theory can prove that recognition of difference is by itself necessarily invidious, false, or controlling.

Exclusive emphasis on othering also obscures the complexity of national and colonial discourses and the extent to which they combine difference and similarity. Numerous writers have shown how Europeans contemplating New World peoples, while strongly asserting difference, also invoked resemblances with their own past and present (Elliott 1970; Grafton 1992; Pagden 1993): "their concern . . . was [often] directed to the finding not of otherness but of commonality" (Elliott 1995, 398). In a similar vein, Thomas Metcalf argues that in British colonial India there was "an enduring tension between two ideals, one of similarity and one of difference . . . At no time was the British vision of India ever formed by a single coherent set of ideas" (1995, x, 66). As later chapters will show, the same ambivalence and self-contradiction are evident in discourses about Kuna alterity.

It is also misleading to think that imperialism can only be constructed or rationalized on the basis of difference. Dominance and oppression, when they consist of making someone over—civilizing savages, reeducating the politically backward, turning heathens into Christians or Kulaks into collective farmers—depend on an assertion, explicit or implicit, of underlying similarity and mutability; an assertion, in effect, of the insubstantiality or ultimate irrelevance of difference.

In addition to alternate rationales, finally, colonization and domination can also have quite different styles, different culturally and situationally conditioned procedures and tendencies. The highly distinct style of the British Raj, as the very epitome of the classifying, bean-counting, distinction-making regime, had a great deal to do with the overwhelming ethnic and cultural complexity of nineteenth-century India, with the daunting task of dominating a subcontinent on the bureaucratic cheap, and with the "relentless need to count and classify everything . . . [that] defined much Victorian intellectual activity" (Metcalf 1995, 113). As we shall see, the Panamanian bureaucrats who tried to subdue and transform the Kuna, heirs to a radically different tradition, could not be bothered to study native customs, to count Indians accurately, or even to use their names: indigenous identities were to be blurred, ignored, muted, or suppressed rather than exhaustively studied and classified.

Power and Domination

Said insistently connects orientalist discourse to the exercise of power as precursor, rationalization, or constituent of imperialism, a manifestation of the West's sway over the Middle East. "Europe was always in a position of strength, not to say domination" (1978, 40, also 73)—a claim that blithely sweeps aside failed Western crusades, successful Ottoman conquests, and a seven-hundred-year Islamic occupation of Iberia. As Linda Colley points out, moreover (2002, 99-134), by locating the florescence of full-blown orientalism in the late eighteenth century, Said misses its development in the previous two hundred years during an era of Western fear and vulnerability.

Whether written from weakness or strength, comments on the Other, rather than unremittingly deprecatory, are surprisingly often mixed, ambivalent, and contradictory. Colley's close reading of a wide variety of British texts leads her to question "the extent to which Islam was regarded and treated as a uniquely different and degraded 'them', and also the degree to which Britons saw themselves as a unified, superior 'us'" (2002, 103). Karen Kupperman, similarly, finds in early British commentary on North America that "sometimes writers . . . [castigated] the Americans as primitive savages. But then again the very same writers turned and praised the Americans' vigor, simplicity, and primary virtue" (Kupperman 2000, 20; see also Pike 1992, 25-27, 31-34). Emma Teng (2004) finds the same mix of praise and criticism in Qing Chinese accounts of the aboriginal Taiwanese.

Some Westerners have gone beyond ambivalence to laud and defend subaltern peoples. As the example of several self-appointed friends of the Kuna will show, advocacy can be mixed with racism and condescension. As Thomas notes, moreover, "attractive and even sympathetic constructions of colonized peoples may admire or uphold them in a narrow or restrictive way" (1994, 54). But dismissing positive portrayals across the board as mere gestures or conceits—treating attacks as real and defense as illusory—obscures the sometimes positive effects of such representations over time. As Sherry Smith (2000, 14) argues concerning early-twentieth-century North American indigenist writers:

To acknowledge that [popularizers] often failed to grasp the complexities of Indian peoples; that they often failed to transcend their own ethnocentric and even racist assumptions; and that . . . readers [today] might find their works sentimental, romantic, and simple-minded does nothing to negate their cultural power.

Some portrayals, moreover, have little to do with either domination or rescue, and everything to do with the situation of their authors. Philip Deloria's brilliant Playing Indian (1998) depicts generations of white North Americans dressing in feathers as a form of symbolic appropriation of perceived indigenous qualities such as freedom, authenticity, and a truly American identity. In the case of the Kuna, early-twentieth-century North American admirers, few of whom had ever visited San Blas, used them as abstract symbolic vehicles for their own preoccupations with racial boundaries, lost autonomy, and sexual threat.

Most generally, the notion of power itself, so widely invoked in contemporary social analysis as an all-purpose explanatory ether—or as Sahlins (2002, 20) puts it, "the intellectual black hole into which all kinds of cultural contents get sucked"—deserves more scrutiny than it receives. That some groups and individuals have vastly more power than others seems so blindingly obvious, and the concept of power so intuitively realistic and hardheaded, that we forget how theoretically and semantically complex power is (Pitkin 1972, 275-286) and how very difficult it can be to demonstrate its workings. Foucault's "poly-amorphous perverse" model (Sahlins 2002, 2) has in some ways usefully complicated understanding, but the neo-Foucauldian assumption that power and knowledge are so closely intertwined—indeed so nearly identical—that they can be routinely written as "power/knowledge" begs the questions it sets out to answer. We should be open to the possibility that not all of Foucault's epistemes are equally potent or monolithic, that power can be fractured, contradictory, and messy rather than smoothly controlling, and that even in the case of "great powers," you can't always get what you want.

Facts and Imaginaries

Another issue, closely related to those already discussed, concerns the role of "the imaginary," a term that has its origin in Benedict Anderson's Imagined Communities (1983), where it is integral to his argument that modern nationalism is based on units of such large scale that they can exist only as they are conceptualized or imagined by their members. The notion of the imaginary, influenced as much by Said (e.g., 1978, 71-72) as Anderson, has expanded and proliferated in works like Inden's Imagining India (1990) to suggest the creative power of the nationalist, racist, and colonialist imagination—its ability to conjure up and impose identities, essences, and fantasies regardless of uncomfortable or discordant facts, "to construct a world, geographic domain, or ethnic grouping in a comprehensive way, rather than merely express a particular perception of something that already existed" (Thomas 1994, 37). Studies of the eighteenth‑ and nineteenth-century Raj have in particular credited British colonial writers with imagining up a totalizing account of Indian society, a complete orientalist ethnography partly constituting the very institutions, such as caste and Hinduism, that it purported to discover (Inden 1990; Dirks 2001).

The notion of the imaginary has usefully focused attention on the often artificial and constructed nature of shared understandings about Others. But it can also obscure the extent to which ideologists begin and end with the known, with accepted "facts." Sati, foot-binding, the veil and the harem (and yes, the caste system and Hinduism) did not spring full-blown from the orientalist imagination: ideology did its work by interpreting and distorting these practices and inflating them with meaning, not by making them up out of whole cloth. In my own examination of the discourses articulated about the Kuna, I have been struck by how closely both friends and enemies usually stuck to the facts at hand, how much they imposed meanings on the Indians not through flights of fancy but through construal of such well-known characteristics as separatism and ethnic endogamy. Such interpretations can be just as biased, misleading, and mendacious as any fantasy, but they stay closer to the ground, and once elaborated, it is their alleged connections with experience and accepted knowledge that validate and reinvigorate them.

Ambivalence, Contradiction, and Mixed Messages

Social science, happily, is no more monolithic than what it studies. Ann Stoler and Frederick Cooper's influential introduction to Tensions of Empire (1997) shook many prevailing "Manichean" colonial dichotomies, arguing that "the otherness of colonized persons was neither inherent nor stable" (1997, 7), and more generally, that "colonial states were often in the business of defining an order of things according to untenable principles that themselves undermined their ability to rule" (1997, 8). Nicholas Thomas, in the same vein, describes colonialism not as a "unitary project but a fractured one, riddled with contradictions and exhausted as much by its own internal debates as by the resistance of the colonized" (1994, 51). Such rifts and fractures are just as prevalent, moreover, in the programs and discourses of nationalism and internal colonialism in countries like Panama as they are in the high colonialism of the great nineteenth-century empires, and for that matter, in almost every cultural and social form.

The Reflexive Study of Ethnography

Among different kinds of cultural representation, this book is concerned most of all with ethnography, sociocultural anthropology's endlessly problematic method of choice and, especially since the 1980s, the subject of its own agonized critical literature. I am concerned with ethnographic studies of a small-scale, non-Western society, a traditional, even old-fashioned, sort of subject population. At the turn of the twenty-first century, ethnographers are as likely to study financial traders, marine biologists, or drug dealers as they are peasants or indigenous peoples. Not so long ago, however, those peasants and "primitives" constituted anthropology's assigned subject matter, its raison d'être, and if we are to make sense of ethnography, whether as a process, a mode of inquiry, or a kind of social representation, we must attend to that past. Thus for the purposes of this book ethnography means accounts of the lifeways of non-Western peoples, mostly but not always written by Westerners, and the messy, problematic dialogues and collaborations from which they have emerged.

In terms of the scope of ethnography as written form, I have cast my net widely, drawing in travel books, magazine and newspaper articles, even letters, reports, fiction, poetry, and oral testimony—anything, so long as it attempts to portray a people's culture or character. The greatest attention goes to hard-core ethnography—professional and amateur nonfiction works based on co-residence, participant-observation, texts, conversations, and interviews—but I also attend to quasi-ethnographic forms of cultural representation at the borders of the genre, some of it by authors who never even saw their subjects.

Dialogue and Agency

Ethnography is in one way or another a collaborative enterprise, entirely dependent on conversations between investigator and investigated. The relationship between ethnographer and informant/consultant/assistant, typically extending across cultural and ethnic lines, has in recent years increasingly been seen as both significant and problematic (see, in particular, Roger Sanjek's fierce scolding of 1993). In response to such concerns, ethnographers have in recent decades searched for more overtly collaborative research practices and for new forms of writing aimed at preserving the dialogic or polyphonic nature of their research.

These laudable changes in the ethnographic present, which in a few short years have become the "disciplinary norm" (Spencer 2001, 450), have been paralleled by attempts, still incipient, to understand past ethnographic dialogues, with the goal not just of offering belated credit but of rethinking the whole enterprise. Some of the need is met by biographies of consultants and native intellectuals such as those by Karttunen mentioned above (1994), but to my knowledge, the call for "a disciplinary history that paid . . . greater attention . . . to the role our informants play in the development of our ideas" (Herzfeld 2001, 9) has so far seldom been met (though see Darnell 2001).

One reason these issues deserve attention is that "we begin to realize . . . that informants are engaged in theoretical practices" (Herzfeld 2001, 7; see also Clifford 1986, 117; Whiteley 1998, 13), that they bring their own perspectives, preoccupations, and agendas to the table, and that they join the ethnographic conversation in no small part because they understand "the connection between knowledge, ideas, and truth . . . and agency, power, and practice" (S. Smith 2000, 11). Here I examine one ethnographic tradition from two sides—assessing the agency and intellectual contribution of native informants and facilitators as well as their Western counterparts—and especially the articulation of the two.

Indeed, I argue that Kuna involvement in ethnography has been collective as well as individual, that in this case and probably others, communities, leadership networks, or whole polities can arrive at a rough consensus or even a conscious policy concerning ethnography and how to deal with it (see Murphy and Dingwall 2001, 344). Paul Sullivan's Unfinished Conversations (1989) recounts negotiations of Yucatec Maya with Sebastian Morley and Alfonso Villa Rojas concerning Villa's studies in their villages, though Maya leaders showed great interest in obtaining guns and next to none in revealing themselves to outsiders. At the opposite extreme, the Kwaio of Malaita long encouraged Roger Keesing's fieldwork in hopes that by recording their kastom he would bring them the power they perceived in writing and codified law (Keesing 1992). As Wellin and Fine put it (2001, 329), peoples like the Kwaio may "see the fieldworker as a tangible embodiment of the more abstract promise implicit in ethnography: that empathic understandings can matter in exposing and shaping realities." In the present day, indigenous groups in North and Latin America often have strong opinions and policies, positive and negative, about non-Indian scholars and their representations; quite a few, moreover, have embarked on their own ethnographic projects.

I am not the first, it should be said, to notice Kuna interest in their own representation. In Mimesis and Alterity (1993), a brilliant but exasperating work by Michael Taussig, he portrays the Kuna derisively as masterful colonial toadies, engaged in "a mimetic contract" with foreign explorers and anthropologists, "a set of largely unconscious complicities between the whites from the north and the Indians they are studying," and even more than that, "a positive connivance in Cuna men being mimetic with white men, and Cuna woman being alter" (1993, 162, 154). This characterization, strangely hostile as well as grossly unfair, nonetheless hits close to the mark: Kuna men and women have better things to do than connive at mimicry, and until the tourist boom of recent years, nothing they did could fairly be called self-marketing. But they have, for a much longer time, taken pains, not to sell but to represent themselves, mounting a sustained public-relations campaign to counter national anti-Indian prejudice. And the medium they seized on was ethnography.

Exclusion and Identification

Ethnographic representation, according to the politicized, reductionist claims that have taken hold in anthropology and critical theory, has throughout been an exercise in domination and deprecation, the creation of "a fantasized other easily digestible for Western colonialist and scientific consumption" (Rapport and Overing 2000, 10). As a key element in a general "process of exclusion" and a "strategy through which to disempower others" (2000, 13), ethnography has created its own object and placed it in an evolutionary past, thus denying the "coevality," modernity, and full humanity of non-Western others (Fabian 1983).

This widely accepted caricature, with just enough grounding in anthropology's colonial past to lend it superficial plausibility, depends on a relentless teleology and essentialism, and on an insistently ahistorical perspective from which no significant change is truly taken into account between the discipline's earliest beginnings and the moment (somewhere between the mid-1960s and the 1990s) when we all saw the light and repented. Like testimonials in an evangelical church, these confessions are filled with pride as well as guilt, and in their "wildly disproportionate sense of the efficacy" of theory and representation, such claims—characterized by Thomas (1997, 335) as "the megalomaniac pretensions of politicized scholarship and theory"—mask our deepest fear, which is not orientalism but irrelevance. The truth lies somewhere between our fears and hopes: ethnography can have real-world effects, some of them negative, but not all-powerful ones, and not necessarily those we predict or intend.

Among the ethnographic representations considered in this book, two of them do indeed place the Kuna in a retrograde barbaric past, but both were written by amateurs who took their cues from nineteenth-century diffusionism and evolutionism. Another set of representations attacked the Indians relentlessly and rationalized their domination, but these came from nonanthropologist bureaucrats with no interest in the lives of the dominated. Of the others, both amateur and professional, who wrote at length about the Kuna, almost all tried not just to describe their customs but to characterize them as a people in some fundamental way and to situate them in some larger narrative. These attempts inevitably reflected the dominant assumptions and ideas of their times, some of them racist, imperialist, assimilationist, ethnocentric, or naïvely romantic. But almost all of them showed sympathy and identification with their subjects, and far from remaining static, ethnographic narratives and portraits changed radically over the course of the twentieth century.

Indeed, I would argue that for many or most of the ethnographers considered here, like many others elsewhere, empathy, identification, and advocacy—in some cases a "possessive identification . . . analogous to the therapeutic relationship in psychoanalysis" (Wellin and Fine 2001, 326)—are more apparent than exclusion and othering. As Nicholas Thomas points out (1997, 334), the intimacy of fieldwork, the sense of indebtedness to those who help us, and a propensity to see oneself in others have "prompted ethnographers to adopt an affirmative attitude toward the people studied, even to write accounts of their culture . . . to some degree complicit in dominant local understandings."

For all the ethnographers, amateur and professional, considered in this book (myself included), their identification with the Kuna loomed large, not just in the field and, for the professionals, in their careers, but in the life histories and personae they presented to friends, family, and professional peers. They were not merely any old expatriate, priest, or academic, but Markham, our local expert on the San Blas; Gassó of the famous mission to the Caribe-Kunas; or Stout (Wassén, Sherzer, Howe), who wrote that book on those Indians in Panama. For all of them, "identity work and the (re)construction of the self [were] part and parcel of the ethnographic endeavour" (Atkinson et al. 2001, 324).

That being said, there can be no doubt that, here and elsewhere, the propensity to assume for oneself "the authority to define the essential elements and boundaries of [other] cultures" (S. Smith 2000, 9, 13), to characterize another in writing without chance of reply, necessarily creates a fundamental inequality. No matter the complicity in local understandings or the desire to present "the native point of view," it is an outsider's version of that point of view that is presented, and indeed, the more complete the attempt at ethnographic ventriloquism, the more problematic the representation. As Murphy and Dingwall note (2001, 341, 344), the greatest risk and the least power for those represented occur at the moment of publication, because whatever influence they can exert over the process, if any, has passed.

Here again, however, the story presented in this book is of change and indigenous agency, not stasis. As the Kuna recognized the potential impact of outsiders' characterizations, they latched onto friendly foreigners precisely to counter hostile misrepresentations already in circulation. Actively feeding material to visitors and anthropologists, and to the extent possible reviewing and correcting their work, when it came to publication they still had to trust to luck and their interlocutors' goodwill. Over time, however, they were able to control outsiders' access more systematically, to review what they had written, and by the end of the century to develop a cadre of their own anthropologists—one version of a story that has been repeated all over the world.

Single Subjects, Multiple Ethnographies

The present work follows in a short but useful analytical tradition devoted to the reflexive historical reexamination, not just of single classic ethnographies, but of a whole corpus of work on one society or region. This tradition, which can be traced back as far as James Boon's Anthropological Romance of Bali (1977), is as diverse as the scholars who have embraced it, though a strong tendency toward indicting as well as critiquing ethnographic predecessors is apparent (Gordon 1992; Apter 1999; Salemink 2003).

The approach adopted here has been shaped by several works in this tradition, including, not least, Boon's supremely humanistic account and its dense, complex, sympathetic engagement with ethnographic texts and traditions, and by Jan Rus's "Rereading Tzotzil Ethnography" (2004), which deftly mixes appreciation for a collective ethnographic project with sharp criticism and close attention to the way the project was shaped and limited by its political context.

Also influential is the voluminous critical literature on the late-nineteenth‑ and early-twentieth-century ethnography of the North American Southwest and especially of the "Pueblo" Indians—archetypal primitives who were the destination and target for dozens of anthropologists and popularizers, the subject of every theoretical paradigm and every yearning for gender equality and communitarian wholeness (Bennett 1946; Hinsley 1989; Whiteley 1998; Dilworth 1996; S. Smith 2000; McFeely 2001; Jacobs 1999; Lavender 2006). Critical studies of Southwestern ethnography focus attention on amateur as well as professional work, and on the porous boundaries between the two. They show students of the Hopi and Zuni involved in complex cultural transfers and exchanges, in which ancient oral tradition was to be re-created in writing, indigenous identities could be transmitted to white audiences, and anthropologists and advocates could flirt with alterity in a liminal, intersocietal space, fashioning roles for themselves as feminists, anticonformist rebels, or heroic cultural mediators. Their writing about the Pueblos—self-referential, "drenched in antimodernism" (S. Smith 2000, 8), and pervaded by anxiety about industrialism, modernity, gender, and national identity—was also fractured by disagreement, self-contradiction, and a reluctance to acknowledge change or complexity. If San Blas and the Kuna never attracted quite such obsessive attention, many of the same themes and preoccupations reappear in the dozens of books and the hundreds of articles, poems, dramas, letters, reports, films, posters, museum exhibits, court documents, and advertisements dedicated to their representation.

Chapters to Follow

In tracing the extended social and intellectual encounter between writing, anthropology, and the Kuna from the early 1900s to the present, this book begins in Chapter 2 with the first introduction of Western education, placing emphasis on the mixed feelings and conflicts schooling and writing inspired and the roles they promoted for young literate men. Chapter 3 considers the most important early use of writing, letters to government functionaries and other powerful outsiders. The considerable success indigenous writers enjoyed in conveying their grievances is contrasted with their failure to overcome stigmatized Indian identity.

Chapter 4 is concerned with representations and counterrepresentations of Indian identity and character in the early twentieth century. I argue that government functionaries could not be bothered to develop a systematic orientalist discourse, and that Kuna leaders, in answering attacks on their character, were limited by the dominant assumptions of the time. Chapter 5 deals with another and much friendlier sort of representation by Anglophone writers, who used Indians as symbolic vehicles for their own preoccupations concerning modernity and sexuality. As a loose alliance developed with sympathetic outsiders, and as government policies drove the Kuna to rebellion, two self-appointed advocates intervened on their behalf, in no small part through a kind of ethnography.

Chapter 6 is the first of two concerned with a remarkable collaboration between a Kuna chief and his secretaries and Swedish anthropologists led by Baron Erland Nordenskiöld. The narrative of a 1927 expedition to Panama and an extended trip to Sweden in 1931 by a young Kuna scribe focuses on the dialogue and symbiotic relationship that developed between Western and native partners. The second of the two chapters examines the published results of that collaboration, a massive ethnographic compendium and series of native-language sacred texts, showing how this corpus was shaped by the predispositions of the two sides, by their working methods, and by the largely textual nature of the material they generated.

Chapter 8 is devoted to other ethnography from the period 1925-1950, including a remarkable folkloric study situating Kuna culture in a national mosaic; Catholic ethnographies staking claims to wardship of the Indians; and popular works by Anglophone visitors carrying on a special relationship with the Kuna. Chapter 9 carries the story into the second half of the twentieth century, an era in which the growth of anthropology as a field and the increasing accessibility of the Kuna to both professional and amateur ethnographers led to proliferation of research and publication.

The tenth, penultimate chapter returns to the subject of auto-ethnography, work by the Kuna about themselves—first by a cohort of native scribes who worked as archivists, chroniclers, and ethnographers, and much later, by university-trained scholars and activists. I try to account for the shape of the corpus they generated and to come to terms with persistent questions about indigenous self-representation and ideology. A brief concluding chapter discusses questions raised by the book as a whole.