(Paul Smith teaches Literary and Cultural Studies at Carnegie Mellon University. He is author of Pound Revised (Croom Helm, 1983), and Discerning the Subject (Univ. of Minnesota, 1988), and co-editor with Alice Jardine of Men in Feminism (Methuen, 1987). His next book, Clint Eastwood: A Cultural Production, will appear from Univ. of Minnesota Press in 1990 and a book of his translations of Jean Louis Schefer's work is due from Cambridge Univ. Press in 1990.)
Anthropology's current discourse is pock-marked with the signs of this encounter. Many have analysed that situation and a large part of this essay will be concerned to sketch out some of the issues that they have brought up. My own point of view touches somewhat upon some thoughts that I have published elsewhere about the discipline of anthropology and its relation to others in the human sciences (Smith).  The need to repeat and perhaps expand on some of those thoughts was provoked by Stephen Tyler's recent book, The Unspeakable : Discourse, Dialogue, and Rhetoric in the Postmodern World, and from my attempt to write some kind of review of that book. 
The title of Tyler's book is a little distracting because it's not actually so much about the postmodern world as it is about Tyler's imaginary of the contemporary status of anthropological and ethnographic discourses. Equally, it's not actually so much about the unspeakable as about what Tyler imagines has yet to be spoken in those discourses--it's a book that is trying to theorise a new method and a new mode of representation for anthropology. Its projective character is one of the reasons for being interested in it, even if the problems which Tyler tackles are probably familiar ones for many contemporary anthropologists in the west-- beginning, indeed, with the central problem for many, namely the whole issue of language and/as representation. Tyler is attempting to offer a sweeping solution to what George Marcus and Michael Fischer, leading the insecure cries of many of their colleagues, have called "a crisis of representation in the human sciences."  But it's a solution which I want to pretty much reject and whose dangers I want to indicate arise as much from its temptation towards a particular kind of easy security as from any of its specific theoretical positions or propositions.
But before entering that particular argument I'd like to offer a kind of reading of the of the epistemological and methodological problems that Tyler is addressing along with other anthropologists, and of which the discipline itself seems to be taking more and more note. Marcus and Fischer designate the current "crisis of representation" as the result of an uneasy interplay of two projects in anthropology: first, ethonography 's commitment to a systematic (if gradual, or partial) description of given cultural and social units; and second, anthropology 's chronic dream (somewhat shattered lately) of discovering an encompassing totality.
The simultaneity of these two apparently disparate aims seems these days not to constitute much of a problem for many who would call themselves scientists. Bourgeois empiricism in our day becomes increasingly confident of its ability to conflate theory and evidence--and, concomitantly, part and whole--even as it produces ever more reductive notions of precisely the inseparability of description and conclusion (this is especially true in areas such as statistics, cognitive psychology, etc.). But for anthropologists (and to their credit, most likely) the relation between part and whole has been made problematic by the nature of the anthropological object itself: that is, the ethnographer clearly has more often than other kinds of observer to recognise that the conception of the whole is a construct of the observer, and that the part is not readily assimilated to that construct: specific evidence still does not contain its own theorisation into a totality.
This is, of course, a fundamental and even quite worn dilemma: one isn't saying anything especially new in pointing to it, nor probably in pointing to anthropology's consequent difficulty with assigning itself the status of a science. But I think it's also true to say that the nature of the dilemma is itself altered, depending on how one describes it. Clifford Geertz, for instance, poses it in terms of self and other, asking in his inimitably grave and humanist manner his own fundamental question: "how is it that other people's creations can be so utterly their own and yet so deeply part of us?"  The self/other dilemma for Geertz points to the necessity of interpretation (Marcus and Fischer indeed dub his brand of anthropology "interpretive"). And this interpretative mould of anthropology has led directly to what I call the "literarisation" of the ethnographic text and to the fairly unmediated importation into anthropology of all the conundrums of western hermeneutics.
Geertz's by now celebrated meditations on the fundamental difficulty of assimilating the native point of view to the anthropologist's gaze seem to me to land up in a kind of paranoia--a paranoia repeated across the text of anthropology itself. What I mean by that is really a number of things, but in part I want to mark the fact that Geertz's way of dealing with the dialectic of other people's creations being their own but also deeply part of us, expressed as a "puzzlement" or a "mystery" in his texts, is in fact just another way for the humanist to stake a claim to a reality which he does not yet own, or owns no longer. Another way of saying that is to foreground the quite generalisable moves and situations of the post-colonialist awakening to the risk of which Trilling had already warned: the inevitable inanation of liberal epistemology if it cannot recognise--and fight for--itself as an ideology.  Geertz, in my view, is merely part of the liberal palace guard, the last to fight for the ideology of liberalism after its claims have been ruthlessly shelled and radically relativised by both left and right.
As the USA enters a post-Reagan time, the bankruptcy (nay, the inanity) of the liberal position on the world-historical stage becomes magnificently exposed, even as it becomes relatively more empowered. Its strategies have any number of analogues in interpretative anthropology, whose fundamental epistemological move here is to claim the empirical otherness and substantiality of anthropology's object and also at the same time to vindicate the innocence of the observer. This is akin to a classically paranoiac position in that paranoiac delusions are delusions of interpretation and constitute a fictionalisation of the external world designed to protect the subject from any alteration or influence which might threaten to betray something unlovable or unwholesome about him. Perhaps the most defended against recognition is that his interpretations are in fact designed for as defences. For Geertz, and for humanist liberalism more widely, the fending off of such threats leads to a potentially endless negotiation, or what Geertz calls a process of "approximating" the cultural meanings which circulate and are exchanged between self and other.
The paranoiac's most secret knowledge, then, is the knowledge that the world is the construction of his own desire to be whole and wholesome. In order for that wholesomeness to be maintained the subject must be able to interpret all of the world, any of it that comes into his purview. It seems to me that it's not a large step from that kind of position to the construction of what I call a "general knowledge," or a knowledge that presignifies or preconceives a totality, a whole, which in function correlates to the subject's own desired wholeness. In modernist anthropology the stress on this endless dialectic--between what is called the irrefutable (i.e. empirical) reality of the other and the internal and defensive coherence of the subject--is perhaps a symptom of the conniptions caused for the subject by the breakdown of the possibility of such a general knowledge.
However, the interpretative rounds can equally be seen as a way of reconstructing that totalism in the shape of a general relativism, brought about not just by the collapse, after modernism, of the possibility of thinking a theoretically whole and complete world into which all the parts will somehow fit and where self and other will join, but equally and simultaneously by the collapse of empire, the postcolonial emergence of an at least potentially autonomous world and its troublesome claims.
Geertz and his interpretative followers in anthropology have of late been suspected and accused--in other ways than my own here--of assuming the load of hermeneutics precisely as a way of defending against the politics of postcolonialism. Geertz's retreat into the hermeneutic circle has also, however, provoked a replay of the battle in other disciplines in the sixties and seventies between hermeneutics on the one hand and the then emerging semiotics and semiology and their later politicised readings on the other. The structuralist and post-structuralist emphasis on questions of representation poses a huge challenge to the sustained circlings of the hermeneut by insisting upon the mediation of language, the unconscious, and ideology in the circuits of self and other. Even where he claims that he wishes to avoid it, the hermeneut is usually open to the charge that he ignores in those circuits the very fact that these are channels of exchange : meaning is the commodity here; what Marx called juridical relations are established between subjects; and so power is always at stake.
It seems to me--and I think Tyler's book confirms this in many ways--that the most effective critiques of interpretative anthropology emerge from work like James Clifford's (even though I would say that much of that work is itself still indefeasibly connected to what I called the literarisation of anthropology). Clifford's work, most of it appearing as a series of articles examining one after another the genres of ethnographic writing from a deliberately alienating post-structuralist perspective,  already assumes the breakdown of the modernist part/whole problematic and pays attention instead to the ways in which different theories, forms and genres of representation might be pressed into service to counter both the paranoiac mode of modernist ethnography and the straightforward imperialist mode of more traditional anthropologists. Clifford's work appears as some of the most important theorisation of the beginnings of what Tyler will come to call "postmodern anthropology and ethnography," and takes its place alongside various experimental efforts at re-aligning the representational modes of ethnographers.
Of these, two strands seem especially important: first what is known as the dialogic mode, and second the collaborative mode (there are others, not the least intriguing of which are the surrealist mode arising out of French ethnography, and what almost in mimicry I call the "whole earth and poetry" school of Jerome and Dianne Rothenberg). In the dialogic mode, for instance, and as in books like Kevin Dwyer's Moroccan Dialogues, the classic representation of the other and the dialectic between observer and observed are offered in the form of lightly edited transcripts of field-work conversations and interviews. In the collaborative mode the attempt is made to give the native voice a co-equal place with that of the anthropologist. In this mode Crapanzano's Tuhami : Portrait of a Morrocan is perhaps the most radical exemplar: here the native informant supplies the text and the anthropologist supplies only questions and challenges to his western audience.  What is demonstrated in either case is the absence of the anthropologist's pretension to general knowledge, or its displacement into general relativism. Instead, the anthropological encounter is offered as highly provisional, specific, and non-generalisable.
The immediate danger in dispensing with the aspiration to a general knowledge is that the desire will be displaced to produce a general relativism instead. But more than that, there is something of an irony, I think, to these kinds of experimental ethnography. Even if it is their aim to prevent the voice of the native becoming subservient to the authoritative voice of an ethnographer speaking to a western audience, it surely cannot be forgotten that it is nonetheless the ethnographer who orchestrates all this; or who, to put it this way, finally has authorship. Here is a kind of classic poststructuralist conundrum, a problem opening out onto a typically deconstructive sort of abyss in speaking of the other. There are few easy pragmatic answers to the problem. One might be the kind that Derrida, for instance, sometimes indulges in as he continually resigns and re-signs authorship. One problem with that strategy is that it's always potentially available to Derrida at any given moment to indeed re-sign or re-assert the legal rights of the signatory to his own advantage. Another solution might be akin to the one offered by Meaghan Morris to men wanting to speak about feminism: to strategically withdraw into silence.  Seen as a political option, this suggestion is difficult: it might have the advantage of releasing the "other" from a certain kind of interference, and it might have the disadvantage of allowing the aggressor and interloper to easily do away with his historical conscience and sense of responsibility, while at the same time abrogating for now potentially useful alliances and articulations.
These kinds of problem are analogous, at the very least, to those facing the new generation of theoretically informed anthropologists at whose hands these experimental ethnographies have undergone some severe critiques. Marcus and Fischer go so far as to dub them all mere remnants of modernist anthropology, while they themselves begin to move towards a different kind of ethnography: one which in their words takes account of "world historical political economy," and/or which repatriates itself to the USA, where--the implicit claim is--the problems of self and other recede somewhat.  To talk about the theory and practice of that kind of move would be to enter a much longer meditation than I want to undertake right here, but I dare say that in my view it too is highly problematical in that it tends to return us to and replicate some of the epistemological problems of holism and objective realism that mark much older ethnographies. Equally, such a move would need to furnish itself with an already much clearer articulation than Marcus and Fischer have so far offered of how the notion of a nation is to be understood. Furthermore, it is probably a world-historical--if not a political-economic--mistake to suggest as they do that "apparent increasing global integration suggests not the elimination of diversity, but rather opportunities for counterposing diverse alternatives that nonetheless share a common world, so that each can be understood better in the other's light."  Another kind of general knowledge, that of a "common world," pulls deceptively into view at that point.
Marcus and Fischer's solution to the yawning abyss opened up by poststructuralist theories of representation and their derived politicised forms is basically to turn away from it, leaving fundamentally intact the central problematic of representation and language. Unfortunately, this leaves Stephen Tyler as one of the few anthropologists to stare deeply and contemplatively into that abyss. Tyler's book does of course have the merit not only of concentrating on the issue of representation but also of trying to come up with a theoretical prescription for what he is pleased to think of as an authentic postmodern anthropology and ethnography. Tyler's way into the question of representation and the problems that it causes is quite a simple one in that it takes seriously both the theoretical and the pragmatic ramifications of a basic fact: that the ethnographer's work is always (or nearly always) literally a transcription . That is, ethnography is an activity which, in his words, attempts "to achieve in writing what speech has created." (p.197) Thus he claims that, in the absence of a thorough retheorisation of the relation between speech and writing (or in an even wider catchment, between orality and literacy), the ethnographer will always be caught up in the dilemmas posed by the trammels of western representational practices, and thus will be confined to merely repeating a belated modernist experimentation with representational forms.
Of course, it is Jacques Derrida who is in our day most likely to be known for the attempt to make the retheorisation of the relationship between writing and speech relationships. But in Tyler's view, Derrida's grammatology is woefully wrong-headed in that it locates speech and its concomitant "presence" as that which has marginalised or oppressed writing. If Derrida--in Of Grammatology, at least --can be understood as fighting against the logocentrism of speech and championing the cause of writing, Tyler sees things differently and indeed in a fundamentally Platonic manner: for him speech is the key and writing is the problem.
A good part of his book is given over to discussing how the traditions of western philosophy--and in particular its recent analytical formations in grammatology and semiotics--are inevitably led to privilege writing because of the western obsession with a certain epistemological structure: that is, the pathway which leads from the priority of human perception, through the secondary nature of mental concepts that are representations of perceptual data, and on into the abstractions of signs and processes of signification in which sensory perceptions are assigned an utterly devolved reality. That epistemological structure is obviously most familiar to us through Kant's triadic line of phenomenon, schema, noumen. Tyler has little difficulty, and less hesitation, in daubing most western epistemologies and philosophies of language with the tar of this Kantian brush. His considered conclusion about this whole epistemological "edifice" is that it is
built in order to surpass itself, to transcend itself, [and] can do nothing more than reproduce--endlessly--allegories of itself. So the imperious urge to symbol and metaphysics makes no metamorphosis after all, and succumbs finally to its own allegory. And that is why Hermes is the tutelary deity of the postmodern age rather than Prometheus. (p.187 n.)
So here for Tyler is the postmodern abyss which his book is designed to help us negotiate. A veritable postmodern discourse, he thinks, will reject this sequential epistemological path, and along with it the western obsession with the separation of perception, signs, and objects. Philosophy will thus close the gap betwen words and action: in short, there will emerge a discourse that has done away with what Peirce establishes as the secondness and thirdness of representation itself. Sticking with Peirce's terms, one can say that the world would thus become an icon where it and language would stand in a relation of pleonasm with each other. 
The thrust and tone of all this can perhaps best be illustrated by quoting a couple of Tyler's specifications for a truly postmodern anthropology:
1) postmodern anthropology asks, 'how did signs and signifieds get pulled apart?,' and identifying writing as the culprit, argues that sign and signified are mutually constituted in 'saying'. (p.194)
2) The chief advantage to eliminating the idea of representation is that it would "show there is no problematic relationship between word and world, for both are mutually present. It is only when there is the possibility of an ABSENCE of either one that the problems arise, and it is of course just this possibility that writing enables." (p.196)
From these kinds of statements--Tyler's contributions to anthropological theory--emerges his adumbration of what kind of ethnography would thereby be required. The lineaments of such a postmodern ethnography (it doesn't yet exist, of course) are thick, but it's probably sufficient here to just point to a couple of its characteristics. First, it is a discourse which doesn't try to make a better representation (the modernist goal) but tries to avoid representation (cf p.205). It could consist as speech or writing, but if it's a writing it must be one that achieves what speech creates without simply imitating speech (to merely imitate speech would be, in Plato's sense and thus in Tyler's too, "rhetoric"). It would be a discourse that evokes (rather than represents) commonality or ethos/ethnos: "if a discourse can be said to evoke, then it need not represent what it evokes." (p.206) The desired signified of such discourse of evocation is described by Tyler as the recognition of the commonplace, the commonsense, or the everyday commonality, of a particular community or social unit. It would thereby, naturally, be a discourse without universals, and without the damaging pretension to "general knowledge." It would be perfectly acceptable, then, if it were simply a discourse of fragments; but at the same time it would mark a "return to the idea of aesthetic integration as subjective and communal therapy." (p.211) It would, finally, be a discourse thoroughly and ritually bathed in the ethos of a given community.
In this rather brief essay I cannot, of course, offer much more than a very vague idea of what Tyler is after. But on the other hand he too is quite vague in his way, as we might tell from the final words of his book:
I call ethnography a meditative vehicle because we come to it not as to a map of knowledge nor as a guide to action, nor even for entertainment. We come to it as the start of a different kind of journey. (p.216)
The reader might well be thinking by now that Tyler should probably just be left to undertake his journey into what amounts to a kind of mystical Platonism, and perhaps wondering why one would bother to bring him to a reader's attention. I think it's probably because his work has struck me as threatening in a certain way; and that, before offering a brief closing remark on a possible alternative approach to the questions of representation that Tyler accosts, I might take the liberty of delivering myself of a small homily in relation to this book.
This work is clearly a determined response to the condition that we already call postmodern. At the level of social life, it is avowedly a response to a perceived loss of authenticity and honesty in social relations; at the level of theory in the human sciences, it is a response to the aforementioned abyss, and to the conundrums of post-structuralism. As such I think it deserves our attention, whatever it looks like. However, there is here, it seems to me, a rather clever and even disguised return to the thematics and ideologies of modernist anthropology and the human sciences: in my own terms, a return to the paranoia of general knowledge. But for Tyler that general knowledge is obviously not the holistic sum of a number of parts (which construct can sometimes at least lay a claim to being materialist, as Fredric Jameson often demonstrates).  Tyler's general knowledge is perhaps even more transcendental than mere traditional western holism. It is in fact a kind of synaesthetic gathering of a phenomenological sense of being into a conflated word/action nexus of speech. This kind of synaesthesia seems to me to be quite analogous in many ways to the ideological and theoretical reductionism and nostalgia of fascism. My worry is that while so many theorists in the western academy are rendered immobile, whirling in a kind of low-grade stasis, faced with the impasses of post-structuralism and indeed of postmodernism itself, Tyler's kind of sense of community (and thus of politics), and its subvention by the authority of a return to a quasi-Platonic mysticism, might begin to look to many rather more attractive than not as a way of emerging from those impasses and leaving them behind.
Whatever the theoretical objections that might be made to Tyler's points, it does seem possible that such work could have an attraction at the fringes of the deconstructive project. While Tyler disagrees somewhat with most of Derrida's "positions," it is clear that he has read and been influenced by deconstruction. The disagreement with Derrida seems to stem largely from differing understandings of Heidegger. Tyler's Heidegger is approvingly contrasted to Derrida at one point (p.47) because he uses "verbalized nouns" where the latter favors "juxtaposed nouns": according to Tyler "the Western lust for the noun, the res " has been properly counteracted by "the verb, verba ." Tyler's Heidegger is perhaps best represented, in On the Way to Language,  by his notion of the "showing of saying," where word and action are once more whole and united. (cf. p.186) This formulation of Heidegger's might be taken to stand as a formulation of Tyler's peculiarised "general knowledge" which is posed against what he might understand as the general relativism of deconstruction. But of course, it is by means of exactly such a disagreement that the propinquity of both sides to a kind of Heraclitean epistemological structure might be discerned. Tyler's virtue, perhaps, is to have allowed a glimpse, to a degree that deconstruction refuses to do, of the political and social implications of the terms of such an epistemology and its dialectically bound terms.
Perhaps Tyler's is the only way to both pass beyond some of the seemingly impenetrable problems of post-structuralism while at the same time recognising how much they constitute a necessary component of current theoretical investigations in the west. On the other hand, if Tyler's attempt to adumbrate a new (old?) logic of the nexus of speech and writing is a response and a rebuttal of Derrida's view of the marginalisation of writing, it may be possible to offer other kinds of rebuttal. One thinks here, for example, of Jean Louis Schefer's efforts to explain writing as itself the agent of an alienation. For Schefer, the investigative beginnings of a history of representation different from Tyler's would be not the philosophy of the Greeks, but rather the epistemological break that occurs with Augustine and the establishment of a specifically Christian practice of writing. The claim would be that writing is scarcely the repressed of western logocentrism, but rather that which suppresses a history of the body. 
Schefer is, it's true, concerned to discover a whole cultural history of the west through an investigation of the relationship between the text and the western subject. However, the claim that writing might be seen to have repressed a history of the body needs very little allegorisation to become of relevance to anthropologists. The body has been replaced by writing: if such a formula seems at some level a little simplistic, it would nonetheless be immensely complicated for a literarised anthropology to emerge from its theoretical shell to produce an ethnography which fully took account of it--complicated, but perhaps useful. In other words, anthropology might begin now to theorise another kind of foreign body.
Carnegie Mellon U.
 Translated as "Language to Infinity" in Michel Foucault, Language, Counter-Memory, Practice (Ithaca: Cornell Univ. Press), 1977.
 See Paul Smith, Discerning the Subject (Minneapolis: Univ. of Minnesota Press), 1988, especially the chapter, "Paranoia." An earlier version of this paper was given as a talk at a conference on anthropology and criticism at Princeton University in April 1988; I thank Chris GoGwilt and Fraser Easton for inviting me to speak there.
 Stephen Tyler, The Unspeakable : Discourse, Dialogue, and Rhetoric in the Postmodern World (Madison: Univ. of Wisconsin Press), 1987. All references are given in parenthesis in the text.
 George Marcus and Michael Fischer, Anthropology as Cultural Critique (Chicago: Univ. of Chicago Press), 1986, chapter 1.
 Clifford Geertz, Local Knowledge (New York: Basic Books), 1983, p.54
 See Lionel Trilling, The Liberal Imagination (New York: Scribner), 1976
 Most of James Clifford's work has now been collected as The Predicament of Culture (Cambridge, Mass.: Harvard University Press), 1988
 Kevin Dwyer, Moroccan Dialogues (Chicago: Univ. of Chicago Press), 1982; Vincent Crapanzano, Tuhami : Portrait of a Moroccan (Chicago: Univ. of Chicago Press), 1980
 Meaghan Morris, "in any event...," in A. Jardine and P. Smith (eds.), Men in Feminism (New York: Methuen Inc.), 1987, p.181
 See Marcus and Fischer, chapters 4, 5, and 6
 Marcus and Fischer, p.136
 Jacques Derrida, Of Grammatology (Chicago: Univ. of Chicago Press), 1976
 Tyler leans quite heavily on Peirce's semiotic. An explanation of why this is feasible for his project might be had by a reading of chapter 3 of Peirce's Elements of Logic (in Collected Papers vol. 2 [Cambridge: Harvard Univ. Press], 1932), specifically on p.156/7. There Peirce allows for the possibility that "there may be Representamens that are not Signs," and gives the hypothetical example of a sunflower which "in turning towards the sun, becomes by that very act fully capable...of reproducing a sunflower" which in its turn reproduces another, and so on. This hypothetical situation, where the sunflower is a Representamen (Sign) without an Interpretant, seems to me to be very much akin to the situation where language and action become the same thing for Tyler's desired community. This is partly to suggest that Tyler's community can be constituted only in and through subjects without agency, or "subjects" utterly at the behest of ethos/ethnos.
 See, for example, his Political Unconscious (Ithaca: Cornell Univ. Press), 1981
 See Martin Heidegger, On the Way to Language (New York: Harper Row), 1971; Heidegger's work in Poetry, Language, Thought (New York: Harper Row), 1971, contains similar suggestions.
 A collection of Jean Louis Schefer's work, entitled The Enigmatic Body, is forthcoming from Cambridge University Press. L' Invention du corps chrétien (Paris: Galilée), 1975, and Espèce de chose mélancolie (Paris: Flammarion), 1979, contain perhaps his most suggestive work in the vein I'm describing here. Two short essays from the latter have been translated into English: "On the Object of Figuration" (trans. D.Judowitz and T.Corrigan), Sub-Stance, no.39 (1983), p.26-3; and "Thanatography, Skiagraphy" (trans. P. Smith), Word and Image, vol.1, no.2 (1985), p.191-94.