Semiotics, Self, and Society 9783110859225, 9783110119787

Citation preview

Semiotics, Self, and Society Edited by

Benjamin Lee Greg Urban

Mouton de Gruyter Berlin · New York

1989

Mouton de Gruyter (formerly Mouton, The Hague) is a Division of Walter de Gruyter & Co., Berlin.

Library of Congress Cataloging in Publication Data

Semiotics, self, and society / edited by Benjamin Lee, Greg Urban. p. cm. — (Approaches to semiotics ; 84) ISBN 0-89925-560-4 1. Semiotics. 2. Self (Philosophy) I. Lee, Benjamin, 1948. II. Urban, Greg, 1 9 4 9 . III. Series. P99.S394 1989 302.2-dc20 89-8239 CIP

Deutsche Bibliothek Cataloging in Publication Data

Semiotics, self, and society / ed. by Benjamin Lee ; Greg Urban. — Berlin ; New York : Mouton de Gruyter, 1989 (Approaches to semiotics ; 84) ISBN 3-11-011978-1 NE: Lee, Benjamin [Hrsg.]; GT

® Printed on acid free paper

© Copyright 1989 by Walter de Gruyter & Co., Berlin, Federal Republic of Germany. All rights reserved, including those of translation into foreign languages. No part of this book may be reproduced in any form — by photoprint, microfilm, or any other means — or transmitted, or translated into a machine language without written permission from Mouton de Gruyter, A Division of Walter de Gruyter & Co. Printing: Ratzlow-Druck, Berlin. — Binding: Dieter Mikolai, Berlin. Printed in Germany.

For Milton Singer From his students, colleagues and friends

Preface Thomas A. Sebeok

In this insidiously sinuous sibilant celebration of Singer's accomplishments — punning can be infectious — the pivotal substantive is "Self' — specifically that manifestation of it localized as the "semiotic self." For this sentimental reader, the title, and some of the contents too, reverberate with the sound of another from a bygone age: Symbols and Society, a nowadays all too seldom revisited volume, published in 1955. Graced with a remarkable, lengthy paper on "Symbol, Reality and Society," contributed by the "Continental Phenomenologist" Alfred Schutz, his oral presentation was directly followed by a concise comment on the part of the "American Pragmatist" Charles Morris (Schutz 1955: 202), who enthusiastically welcomed this "addition to the literature of contemporary semiotic," since, as he noted, "there are few basic discussions in this field written from the standpoint of phenomenology." Singer cites Schutz's later, more developed phenomenological notion of "wereality," that is, the notion of the priority of the We to the I, or, in common parlance, the overriding importance of human sociality. Schutz argued along two lines: first, that "I" is born into the world of others, who raised "me" and bequeathed to "me" patterns of signification ("knowing") and of communication ("behaving"); and second, that "I" is able to "stop and think" — the expression is John Dewey's — that is, to become conscious of "my" concealed individual self (Schutz 1962: 169-172). This of course echoes what Coleridge — whose philosophy of Ich-heit Peirce himself cited circa 1902 — called, as far back as 1817, in his Biographia Literaria (Ch. XIII), "the primary Imagination." By this Coleridge meant "a repetition in the finite mind of the eternal act of creation in the infinite I AM," offering the means of escape from the prison-house of the self by engagement with others, or the ability to distinguish subject ("I am") from object ("you are," "it is"). However, by 1908, as Peirce wrote (on December 14th) to Victoria Lady Welby, he had come to realize that the putative contrast between Subject and Object — "in any of the varieties of German senses" — was misleading, and "led to a lot of bad philosophy...," these terms requiring a subtler sort of semiotic analysis than had been accorded them theretofore. (Some passages in this book, too, independently attempt to grapple with this unresolved issue).

vi Thomas A. Sebeok One can be sure that Morris would have embraced Singer's sympathetic, clarifying augmentation here of this perpetual, complicated dialectic, the synthesis of which was assigned in some circles to a hybrid field later dubbed "Social Phenomenology." (Cf. Morris 1970: 149. He appears however to have been unaware of Schutz's ideas about sign systems, for The Phenomenology of the Social World, where Schutz faced such matters most candidly, came out in English only in 1967). Just how prodigious this dialectic interplay between self vs. other can become in its ramifications into countless categories and directions is beautifully exemplified, on a vast geopolitical canvas, by Todorov's (1982) insightful meditations on the discovery and conquest of America, or the encounter between "us," the Europeans colonizers, and "them," the colonized Indians. The closest link of the self in nature as well as in culture is with memory, both as a feature of a physical repository and as a social construct. The reasons for this are quite straightforward: each organism requires information — I use "information" casually here to mean the representation of sets of prior events embodied in a code — about certain experiences in its past to enable that individual to steer with reasonable certitude of survival in its specific current Umwelt. Memory in "man, proud man" makes up, as it were, a multi-sensory private documentary archive, severally composed of nonverbal signs with a verbal overlay. It is the articulatio secunda, or the syntactic aspect of language, which provides the machinery whereby memory organizes, continually remodels as a child playing with a tinkertoy, and finally imposes a coherent and personal narrative schema upon each of us. Since writing tends to conserve the semiotic self far beyond any oral tradition, literate peoples have invented the diary or intimate journal (and, later, the family photo album, home movies, and comparable technological accoutrements), to delineate for themselves, in the form of supplementary aides mimoire, a kind of dramatic "I" to furnish, in Peirce's memorable phrases (MS 318-355, 1907), their theatres of consciousness, to turn its activity from the stage "of the internal, to that of the external experience," or to "come down to the footlight of consciousness" (MS 339C; 505, 1905). This blueprint, too, is what Jacob envisioned when, in the concrete titular and key metaphor of his recent autobiography, he fantasized carrying within himself a kind of statue intdrieure, "sculpted since childhood, that gives my life a continuity and is the most intimate part of me, the hardest kernel of my character. I have been shaping this statue all my life. I have been constantly retouching, polishing,

Preface vii refining it... Not a gesture, not a word, but has been imposed by the statue within" (Jacob 1988: 19). Along with Popper and Eccles (1977: 129), we may say of the self that, "like any living organism, it extends through a stretch of time, from birth to death." The semiotic self is by no means identical with "the consciousness that binds our life together" (Peirce 1.381). While even the consciousness of synthesis is interrupted by periods of sleep, continuity of our semiotic persona (presumably even by those claiming to have been Born Again) is normally taken for granted. Again, Jacob's (1988: 14) question is stirring: "Why doesn't the system slip so that, after sleep has disassembled the mind, its memory and will, the mechanism is not reassembled somewhat differently, to form a different person, a different me?" Memory agglutinates "man's glassy essence," a phrase which, thanks to Shakespeare, Peirce, and Singer has become a condensed emblem for specular semiosis, the process by which man converts his Umwelt into a unitary system of signs, a configuration, and, as Lecky had insisted in his theory of personality, SelfConsistency (1945), into an enduring, more or less singularly consolidated autobiographical identity (excepting, arguably, in that one per cent or so of the population designated, for that very reason, as "schizophrenic"). Memory also creates the illusion — "most ignorant of what he's most assured" — that all our acts are performed by the selfsame person, labeling the array of "fantastic tricks" that happened to us in our past as incidents which, notwithstanding that they may "make the angels weep," do compose a coherent sequence of experiences. Memory itself is, of course, continually refigured to insure the maintainance of positive self-esteem. The twin functions of memory in our lives — the archival and the amalgamative — have two coupled sources: genetic and semiotic. There are, as well, as I have previously argued elsewhere, two intertwined notions of individuation: the immunologic, or biochemical, "with semiotic overtones," and the semiotic, or social, "with biological anchoring" (Sebeok 1988: 263-267). An apprehension of how the immune system creates the capacity for distinguishing between self and non-self by the immunoglobulin molecules expressed in Β cells and antigenspecific receptors expressed in Τ cells is a fascinating problem at the very frontiers of research, with momentous implications for the diagnosis and treatment of critical human ailments. So also the human self is passed on by a fusion of two means of inheritance, which may — though imprecisely — be termed the "Darwinian" and the superimposed "Lamarckian" avenue (Sebeok 1989).

viii Thomas A. Sebeok The notion "semiotic self registers and emphasizes the fact that an animate is capable of absorbing information from its environment if and only if it possesses the corresponding key, or code. There must exist an internalized system of signposts to provide a map to the actual configuration of events. Therefore, "self' can be adequately grasped only with the concepts and terminology of the doctrine of signs. Another way of formulating this fact is that while living entities are, in one commonly recognized sense, open systems, their permeable boundaries permitting certain sorts of energy-matter flow or information transmissions to penetrate them, they are at the same time closed systems, in the sense that they make choices and evaluate inputs, that is to say, in their semantic aspect. For this reason, a Turing Machine necessarily lacks a "semiotic self," its Umwelt forever merely mirroring its creators' Innenwelt (cf. Uexkiill and Wesiack 1988: 188,196). It would be instructive to reflect at leisure (not afforded here) on three predominant styles of ethnographic discourse. It is the hallmark of great (or neargreat) bourgeois ethnographies that they depict and interpret the semiotic self of their subjects by their possessions rather than, as was the custom in other literary works, by their actions. In the bygone era of "culture-and-personality" studies, dramas of imaginary inner conflicts were juxtaposed as sharply at odds with the outer milieu; in other words, the "unconscious" was enthroned as the "other" within "one-self." Ethnography melds with the fruits of bioanthropology when the focus of inquiry is narrowed from its looser accessories and belongings to the human body (L'Homme nu) in the strict sense. This body has — or rather consists of — a veritable armamentarium of more or less palpable indexical markers of unique selfhood (save again perhaps for identical twins). Fingerprints, and what Alphonse Bertillon in such proto-semiotic works of his as Service de signalements (1888) and Identification anthropomitrique: Instructions signalitiques (1893) called "the professional signs" (cf. Rhodes 1956: 143), and later biometric refinements in Sir Francis ("When you can, count") Galton's anthropological contributions in the field of measurement, exhibit one such array. Ginzburg's brilliant essay (1983) on the interpretation of a variety of phenotypic clues musters other striking manifestations — by art historians (Morelli), psychoanalysts (Freud), and detectives (Sherlock Holmes) — of this same epistemological model. On the genotypic plane, so-called "DNA fingerprinting" can in fact now identify, with absolute certainty, on a level of discrimination far above anything available before, every person (again excepting an identical twin),

Preface ix even by a single hair root, on a small piece of film displaying his or her unique sequence of DNA molecules. The odors — of which Peirce remarked (1.313) that they "are signs in more than one way," especially by contiguous association — and the subjacent chemical composition of every human being differ from that of every other. Some consequences of these stark facts were horrifyingly depicted in Patrick Siiskind's gripping and (in John Updike's words) "beautifully researched" 1985 novel, Das Parfüm (Ch. 31): "the odor of human being did not exist, any more than the human countenance. Every human being smelled different, no one knew better than [the book's antihero] Grenouille, who recognized thousands upon thousands of individual odors and could sniff out the difference of each human being from birth on." Above the basic theme of human odor "each individual's aura [hovers] as a small cloud of more refined particularity," the highly complex, unmistakable semiotic code of a personal odor. The study of the distinctive pheromonal function, subsumed under the new scientific rubric "semiochemistry," of human chemical signatures is in fact comparable with individual fingerprints (e.g., Albone 1984; on the comparison with fingerprints, see p. 65). It is, by the way, well known that human infants can sort out their mother's peculiar odor, and possibly even her idiosyncratic breast odor, from those of all other women. The immune system utilizes approximately as large a number of cells dispersed throughout our bodies as the number of cells that composes our brains. The mammalian nervous system, consisting of an endosymbiotic (better: endosemiotic) aggregation of spirochetal remnants, provides the hardware for the functioning of the no less important but even less clearly understood self, the semiotic capacity of which is based on electrochemical signing function and kindred processes begun already to be utilized by bacteria some 3,400 million years ago. Half a century ago, Jakob von Uexkiill introduced a musical metaphor of "Zell-Subjekten" being endowed each with a quasi-melodic specific "Ich-ton." Concrete applications from brain research may be found in, for instance, Vernon B. Mountcastle's and John C. Eccles important series of studies of neuronal impulses (Uexkiill and Wesiack 1988: 217f.). Some geneticists, like Waddington (1961: 121), have argued that even selfawareness "evolved from simple forms which are experienced by non-human [comprehending inanimate] things." Wherever this line of reasoning, which smacks of panpsychism but which was once very influential all the same, may lead, it is certain that sexually reproducing organisms, "since complex animals

χ Thomas Α. Sebeok undergo a complex program of development from their conception in the form of a fused egg and sperm cell, ...have special aptitudes for identity and death" (Margulis and Sagan 1986:162). Are all speechless creatures capable of internal self-representation? D.E. Koshland claims that E. coli, a mostly friendly bacterium living in our gut, has no need of representation, memory, or mentation of any sort. This is an empirical question, though not an easy one to investigate. The "special aptitude for identity" in animals, their semiotic selfhood, has been most fruitfully examined so far by that nonpareil explorer of animal behavior, Heini Hediger (e.g., 1980: 38-62). Does an animal's self-appraisal include an awareness of its detachable appendages (such as spontaneous tail amputation in the course of evasive maneuvres of many Sauria)? Of its exclusive body-odor? Of its image reflected in water? Of its shadow upon the ground? Hediger affirms, for instance, that there is evidence to show that animals evolved relatively early an ability to recognize their own shadow, impermanent as this may be, as a spatial extension of, or image isomorphic to, their self (the model). The genesis of the image, the nature of the physical processes which come into play in such examples, was newly analyzed in a pathbreaking (though Peirce-inspired) study by the topologist Rend Thom (1973). In the identification of the semiotic self, as in all domains of semiosis, context is determining: a live snail's shell is a part of that mollusk's integumentary self, whereas the hermit crab's shell, its temporary dwelling, equally vital though it may be in securing that creature's survival, remains but a foreign object. What of body-size? As long as sixty years ago, R. Oeser (cf. Hediger 1980: 42) propounded this intriguing and empirically testable principle: that an animal tends to know the dimensions of its own body (although not, say, that it is white). Such knowledge, both logical and biological, must be unambiguous. (It incidentally concerns zoo-planners directly, as for example in the design of enclosures for Bovids and Cervids of both sexes and different ages). Singular proper names (in whatever channel, inclusive of chemical pheromones) tend inherently to be found in those vertebrates which form personal societies. One's name is therefore deeply conjoined with one's self (Hediger 1980: 63-84; Sebeok 1986: Ch. 7). The self assumes a degree of "corpo-reality" through the semiotic act of denomination; one's name becomes thereby a quasi-iconic index which denotes it In baptism, it is the "within" that motivates the "without." Some people, like Alice when she passed into the wood "where things have no names" in Through the Looking-Glass, feel that their name should embody the

Preface xi exceptional status of their semiotic self. So Todorov (1982), relying on contemporaries like the Dominican "Apostle of the Indians" Bartolomi de las Casas, showed that the obsessive preoccupation of Chistopher Columbus with both his name (viz., his signature) and his emblematic siglum indeed bordered on the fetishistic. Dante's formula, Nomina sunt consequentia rerum, does appear to find application at least in this limited, privy domain of our vocabulary. And what of social position? "The self," it has been suggested (Chance and Larsen 1976: 205), "is intimately tied to primate social dominance..." The salience of this link in human affairs — the iconic spatial expression of social hierarchy — is generally clear (Sebeok 1988: 118), but let me adduce one familial example. In my father's household, in interwar Hungary, he, the "head" of the family, was seated during meals, as you would expect, at the "head" of the table. On Sundays and most other holidays, the midday (main, often extended) family meal of the day began almost invariably with chicken soup. Cooked in the broth, there floated the severed head of a rooster, cockscomb prominently displayed. The maid always served my father first, ladling out of the tureen the rooster's head into his bowl, beak facing my father. After he had finished his soup, he would pick up the head, munch on the comb, then, with a special instrument, trephine the skull and suck the brain out. When my father was absent, protocol demanded that the elder son, myself, be served the head, with the same tasty rite ensuing. Leaving aside the psychoanalytic innuendoes (if any) of this recurrently observed ceremony of my youth, it will be observed that status in our domestic establishment was being reinforced by the manner in which we all behaved around the table. Iconic signs were used both to remind of and fortify status, in order to warrant the perpetuation of the prevailing system of distinctions (cf. Sommer 1969: 17). When guests were present, as was often the case, they learned at a glance who was the alpha and who the beta male, and all the servants knew as well who had maximum access to amenities and information. The mechanisms at work here were incorporated, indeed, drummed into my self, which, as Peirce however wisely observed (5.462), "is only inferred," for "your neighbors are, in a measure, yourself, and in far greater measure than...you would believe" (7.571), and because "the Self of the man is...included within a larger Self of the community" (8.123), as if these Selves formed a continuum. Singer has even suggested, how seriously I can't tell, that perhaps the ceremony recounted here "represented a ritual enactment of a Hungarian family's iconic diagram of I, it, and Thou."

xii Thomas A. Sebeok Some of the topics touched on in this book, notably those that consider aspects of personal identity, were substantially refined by the late Erving Goffman, especially in Stigma, the most semiotic among all of his writings. He was one of the first to explicitly recognize positive indexical marks of personal identity, which he named "identity pegs" (roughly corresponding to Morris's family of signs called "identifiors"; cf. Sebeok 1985: 138-139). Goffman, moreover, duly credited (1963: 57) the role of memory, in the idea of "the unique combination of life history items that comes to be attached to the individual with the help of these pegs for his identity." The trope he used, in its way every bit as homely yet as original as Jacob's statue, was that of candy floss, which he envisioned as a sticky substance to which the "facts" of one's biography gradually adhere. According to Buddhist precepts (as who among us knows them better than Professor Singer?), the "self is an illusion, or, more accurately, an infinity of interlocking illusions — it is in fact a set of dharmas and relations among dharmas. From this it follows that the "other" must likewise be an illusion. Neither exists. This being so, the semiotic self cannot, in that tradition, ever be understood: what is understood is that there is nothing and that there is noone to understand it (cf. Kolm 1982: Ch. 23). But in Western ontology, we prefer to heed the admonition inscribed upon the Delphic oracle: Gnothi seauton — Know thyself. As the contributors to this book, inspired by Singer, grapple, each in his way, with the oracle's injunction, they all do so within a social science frame. During the same decade, in the domain of the neurosciences, by connecting Darwinian selection, that most basic of biological principles, with events in the brain, Gerald M. Edelman has, perhaps even more provocatively and decisively, moved individuality back to center stage, convincingly showing that the semiotic self, far from being an epiphenomenon, plays the starring role. Such events, that eventually result in behavior, thought, and memory (Rosenfield 1988) all argue for the supreme importance of individuality. The standard man/computer (or man/logic-machine) analogies, leave novelty out of the equation that links the Innenwelt with the Umwelt. The beauty of Edelman's theory of Neural Darwinism derives precisely from its inability to predict how the brain will evolve. Each human brain develops its own distinct way of ordering reality in the course of a ceaseless, searching progression of sign interpretation; or, as we might say in semiotic parlance, every mind has an inherent capacity to launch ever more developed, enriched interpretants, in their three Peircean varieties (corresponding respectively to sense, meaning, and signification) on their journey towards infinity.

Preface xiii Singer's 1984 book marked not just the companionship, or even the cohabitation, but the legitimate, happy marriage of an older semiotics with a younger anthropology. Let this book be but the first offspring of their union.

References Albone, Eric S. (1984). Mammalian Semiochemistry: The Investigation of Chemical Signals Between Mammals. Chichester: John Wiley & Sons. Chance, Michael R.A., and Ray R. Larsen, eds. (1976). The Social Structure of Attention. London: Wiley. Ginzburg, Carlo (1983). Clues: Morelli, Freud, and Sherlock Holmes. In The Sign of Three, Umberto Eco and Thomas A. Sebeok (eds.), Chapter 4. Bloomington: Indiana University Press. Goffman, Erving (1963). Stigma: Notes on the Management of Spoiled Identity. Englewood Cliffs: Prentice-Hall. Hediger, Heini (1980). Tiere Verstehen: Erkenntnisse eines Tierpsychologen. Munich: Kindler Verlag. Holm, Serge-Christophe (1982). Le Bonheur-liberti. Paris: Presses Universitaires de France. Jacob, Francis (1988). The Statue Within. New York: Basic Books. Lecky, Prescott (1945). Self-Consistency: A Theory of Personality. New York: Island Press. Margulis, Lynn and Dorion Sagan (1986). Microcosmos. New York: Summit Bodes. Morris, Charles (1970). The Pragmatic Movement in American Philosophy. New Yoric George Braziller. Peirce, Charles S. (1931-58). Collected Papers of Charles Sanders Peirce, 8 vols., C. Hartshorne; P. Weiss; andA.W. Burks (eds.). Cambridge: Harvard University Press. Popper, Karl R. and John C. Eccles (1977). The Self and Its Brain. Berlin: Springer-Verlag. Rhodes, Henry T.F. (1956). Alphonse Bertillon: Father of Scientific Detection. New York: Abelard Schuman. Rosenfield, Israel (1988). The Invention of Memory: A New View of the Brain. New York: Basic Books.

xiv Thomas A. Sebeok Schutz, Alfred (1955). Symbol, reality and society. In Symbols and Society, Lyman Bryson; Louis Finkelstein; Hudson Hoagland; and R.M. Maclver (eds.),. 135-203. New York: Harper & Brothers. Schutz, Alfred (1962). Collected Papaers, vol. 1, The Problem of Social Reality. The Hague: Martinus Nijhoff. Sebeok, Thomas A. (1985). Contributions to the Doctrine of Signs. Lanham: University Press of America. (1986). I Think I Am a Verb. New York: Plenum Press. (1988) The Sign & Its Masters. Lanham: University Press of America. (1989). Darwinian and Lamarckian evolution of semiosis. Lecture prepared for delivery at the International Colloquium on the Evolution of Culture, September 22,1988, at (he Villa Vigoni. Bochum: Norbert Brockmeyer. Summer, Robert (1969). Personal Space. Englewood Cliffs: Prentice-Hall. Thom, Ren6 (1973). De l'icone au Symbole: esquisse d'une thöorie du symbolisme. Cahiers Internationaux de Symbolisme 22/23: 85-106. Todorov, Tzvetan (1982). La Conquete de l'Amärique. Paris: Editions de Seuil. Uexküll, Thure von and Wolfgang Wesiack (1988). Theorie der Humanmedizin. Grundlagen ärtzlichen Denkens und Handelns. Baltimore: Urban & Schwarzenberg. Waddington, C.H. (1961). The Nature of Life. London: Allen & Unwin.

Imatra, Finland July 18, 1988

Contents

Preface Thomas Α. Sebeok

ν

Introduction Greg Urban and Benjamin Lee

1

The Self in Psychoanalytic Self Psychology Ernest S. Wolf, MD.

15

The Τ of Discourse Greg Urban

27

The Self in Northern Cheyenne Language and Culture Terry Straus

53

The Semeiosis of Suicide in Sri Lanka E. Valentine Daniel

69

Responsibility and Compensatory Justice in Arab Culture and Law Lawrence Rosen

101

A Confusion of Signs: The Semiosis of Anthropological Ireland Hervi Varenne

121

Against Coping across Cultures: The Semiotics of Self-Help Rebuffed James A. Boon

153

Singer, Kant, and the Semiotic Self Murray J. Leaf

171

Semiotic Origins of the Mind-Body Dualism Benjamin Lee

193

xvi Contents Pronouns, Persons, and the Semiotic Self Milton Singer

229

Afterthoughts Benjamin Lee and Greg Urban

297

Index

305

Introduction Greg Urban and Benjamin Lee

These essays are concerned with the philosophical "category of person," viewed from anthropological and semiotic perspectives. In one sense, the essays continue the Αηηέβ Sociologique tradition, and the work of Marcel Mauss (1985[1938]), whose classic study charted a comparative approach to the cultural construction of the self concept. And in this same sense they continue also the work of Irving Hallowell (1955a,b) and his students (see Fogelson 1982), who have probed empirically the problem of how different cultures differentially encode understandings of what it means to be a self, with relative boundedness with respect to other selves and with respect to the world of non-selves. But these essays also represent a departure from the main lines of classical social and cultural research, this due in large measure to the influence of Charles S. Peirce and of Milton Singer, in whose honor this volume was originally conceived.1 The departure is twofold: (1) the opposition set up in the Maussian tradition between a universal self, associated with the biological person, and the culturally-variable social construction of the self is gradually broken down to show the thorough interpenetration between these levels; and (2) the idea of a social construction of the self is explicitly associated with signs and sign phenomena. Both of these departures have to do in turn with the reconceptualization of the cultural-encoding of the self as occurring on a number of analytically distinct levels, which may be thought of on analogy with a system of concentric rings, conically arrayed, with pronominal usage at the core, lexical items and belief systems about the self occupying a more encompassing domain, and the broader struggle between voices occurring in the public sphere being at the furthest remove. This model provides the organization for the papers in this volume. One may think of the levels in terms of the semiotic complexity of the reflective processes involved. At the lowest level are the basic indexical processes of selfawareness; at a higher level of remove, the referentially encoded awareness of what it means to be a self; at a still higher level the struggle to define the referential code of the self itself. But the model has its limitations as well. In the case of philosophy, we are dealing with the highest reflective levels, which are also semiotically most

2 Greg Urban and Benjamin Lee complex. Philosophers are part of the constellation of voices which seek to define the referential code of the self. At the same time, the philosophical reflective processes work their way down to the most basic levels of subjective awareness, for example, in the case of Descartes, to the personal pronouns.2 So there is a danger in conceptualizing the levels as unconnected. In fact, as these papers amply demonstrate, the levels are in constant interaction with one another, this interaction forming part of the continuing processes of cultural constitution of the self. Mauss regarded as a universal the subjective sense of being a self, distinguishable from a surrounding world: "there has never existed a human being who has not been aware, not only of his body, but also at the same time of his individuality, both spiritual and physical" (Mauss 1985: 3). This view of self in terms of self-awareness of discreteness or boundedness with respect to a surrounding cosmos, however, is first and foremost a semiotic phenomenon. Despite the fact that Mauss distinguished between this universal self-awareness and the cultural construction of self concepts in different societies, the present papers view self-awareness as one pole of a continuum in the semiotic construction of self, a pole which depends upon culturally constituted, and especially, linguistic signs, such as the first person singular pronoun, as Peirce, and after him Singer, and other papers in the present volume, notably those by Lee and Urban, have stressed. But Wolfs paper, which opens the volume, takes us to perhaps an even deeper semiotic layer, to a sense of self not as cognitive self-awareness, constituted through indexical pronouns, but as a feeling of wholeness, an experience of completeness. Here we are at a level where we do not have a conscious encoding of self in linguistic signs, but a peculiar feeling which is only subsequently interpreted, within the self psychological theoretical discourse, as a "sense of self." The feeling is, to be sure, one of which individuals can become aware, especially in the course of psychoanalytic treatment. But the push to formulate the sense of self in language, whether in internal dialogue or in the analytic session, is more often occasioned by its absence, by the disruption of normal feelings of wholeness or integration. Intriguingly, in the self psychology framework originally propounded by Kohut (1977,1978), and here developed by Wolf, the feeling of wholeness may require the presence of what to an observer appears to be an external object, another individual or a thing, which in this theoretical terminology is called a "selfobject." The selfobject is traceable ultimately to the original parental objects with whom the individual was once merged. But the important point is that the sense of self is

Introduction 3 not necessarily associated just with the limits of the body, as in Mauss's formulation of the universal of self awareness. Depth psychology takes us to a level where the feeling of selfhood may depend upon the involvement of the individual with other objects. That the constitution of possible selfobjects may be culturally variable is an hypothesis that has yet to receive adequate investigation. One can readily imagine, however, that the sense of selfhood at this emotional plane may be produced by quite different circumstances in different cultures. The candidate for a universal under this view would be the experience of selfhood, with the specific cluster of circumstances capable of producing it — involving a possibly culturally specific set of selfobjects — being variable. While this may violate Mauss's universal in its specific formulation, however, it actually supports the more general argument. Mauss saw the Western individualistic conception of the self, as formulated on the cultural plane, as distinct from the more widespread identification of the person with the social role. In the latter case, distinct individuals with discrete bodies may be interchangeable with respect to the socially constituted conception of a self in a given society. From the self psychological point of view, because the experience of selfhood may depend upon the presence of selfobjects, a cultural conception in which the self is seen as more encompassing than the individual would, correspondingly, appear more natural. It would correctly interpret the emotional plane, in which selfhood is a product of involvement with others. The inability of the Western individualistic conception of self to interpret the underlying emotional experience of self as more encompassing than the individual recalls the work of Takeo Doi (1973), who found that the Japanese concept of amae, an emotional experience of involvement with others, could not be readily translated into English, despite the fact that Americans could understand, and were even struggling to formulate, the cluster of experiences which amae comprehends. Daniel's paper in the present volume represents an attempt to isolate a similar discrepancy between cultural interpretant and underlying emotion. Daniel shows that in Sri Lanka there is a differential ability of the different cultures to comprehend certain underlying emotional experiences, which the Sinhalese capture in their concept of "aloneness-illness." Here it is a question of the culture providing, or failing to provide, a set of semiotic linkages which make emotions satisfyingly interpretable within conscious understanding. But what is therelationshipbetween the sense of selfhood as formulated by Wolf and awareness encoded by "I" and "me"? The nature of the awareness associated

4 Greg Urban and Benjamin Lee with first person singular pronominal usage is something that must be carefully investigated cross-culturally, as we do not know the extent to which Western awareness may be suffused by the broader cultural construction of an individualistic self. In philosophy, of course, the pronominal categories have been intensely reflected upon, as in the Cartesian cogito and in the work of Peirce himself. But one of the issues confronted by comparative research is the extent to which such reflection — and the content of consciousness more generally — is a culturespecific product. In fact, in the present volume Lee argues that the Cartesian system is essentially a product of the language structures that the reflection process necessarily employs, and specifically, of the grammatical parallelism between verbs of speaking and verbs of thinking. In comparative terms, it may be that research on language as decontextualized structure will reveal a first person singular pronominal category, which everywhere functions indexically to point to the "speaker of the present utterance," and hence seems to suggest a self-awareness of the individual as individual. But the key question for empirical research is the nature of usage of first person forms. Here Urban's paper points to a way in which the "I" transcends the here-and-now unique individual body uttering the words. Urban shows that in fact some "I" usages, specifically, those within direct quotations, are actually anaphorical devices, maintaining the reference of third person forms. But in using the first person forms, the speaker is semiotically entering into the subjectivity of the other through a kind of role-playing. The locus of subjectivity transcends the individual body, and the subjective discourse of that individual body is itself suffused with multiple voices. When it is understood that there is a continuum between such quotative usages and various forms of de-quotative, theatrical, and projective "I"s, it becomes more immediately comprehensible how the seemingly ordinary first person singular form may involve an awareness of a self that is greater than the individual. It even suggests the extent to which the ordinary "I" may bring into play numerous voices, without doing so explicitly, and so point to what is in fact a cultural self, quite distinct from the individual body. The work by Urban, like that of Lee in the present volume, thus effaces the boundary between subjective and objective, and further erodes the contrast set up by Mauss between the universal awareness of individuality and the cultural representation of self. Culture intrudes even into awareness — in the form of pronominal play and in the relationship between first person and third person forms. Indeed, it is the interplay between deictic and non-deictic pronouns that makes possible the conversion of objective into subjective and vice versa.

Introduction 5 At the same time, there is a question of the interaction between first and second person forms, which Peirce and Singer have repeatedly stressed. If the boundary between subjective and objective, between self and other, is eroded in the first person forms, the boundaries between the self and other selves are affirmed by the use of the second person. Whereas the Western individualistic psychological tradition conceptualizes selves as isolated and independent monads,3 the semiotic point of view suggests that the self may be a precipitate of the dialogue with other selves, conceived as second persons. While this dialogical view of the self may not underlie the principal cultural constructions of self in the West, it nevertheless appears to be the basis for such constructions among the Northern Cheyenne, as argued by Straus in the present volume. The semiotic studies thus tend to suggest that there are two distinct processes at work at this bedrock level of indexical self-awareness. There is the process of selfto-self interrelationships. In this process, which is associated with the interplay between first and second person, the world is a world of selves before it is a world of objects. The process seems to be one primarily of differentiation in which use of second person in relationship to first person affirms the boundaries between self and other. But there is also, and contrastively, the process of self-to-object interrelationships. Here the issue is that of the boundary between self and thing, conceptualized as a specifically non-self. Where the boundary is sharply defined, as in the Western individualistic cultural representation, the process of self-to-object interrelation is also one of differentiation, of mechanization and deanimation of the world, in which the world consists primarily of things. But at the level of indexical self-awareness through pronouns, the boundary between self and object is in fact permeable, with the processes of incorporation and merger — transformation of third person into first, as described in Urban's paper — occurring alongside differentiation and objectification — transformation of first person into third, as described in Lee's paper. Straus's paper also raises the question of what might be termed the "agency of the self," whether the self is viewed primarily as the source or originator of action and control, or more as the addressee in a dialogue between first and second persons, and the recipient of actions initiated elsewhere. Here Straus, like Rosen in his work on the image of person embodied in Arab law, is concerned with higher semiotic planes than those of simple indexical self-awareness. She is concerned with how various semiotic representations within Northern Cheyenne culture portray the self, for example, the mythological image of a monster who cuts off the ears of humans, thereby depriving them of what is most central to the

6 Greg Urban and Benjamin Lee human self — the ability to participate in dialogue and to be addressed by another. But the issue here is that of how the semiotic planes intersect, and, especially, how the plane of indexical self-awareness is infused with a specific cultural understanding, namely, that the self is primarily a recipient of actions and powers that lie outside itself. This cultural construction conflicts with the Western individualistic notion of the body as the locus of agency, but it may be, again, that the Northern Cheyenne representation correctly interprets certain emotional experiences of selfhood that the individualistic understanding obscures. In this case, we are reminded of Wolfs formulation of the four types of selfobject experience a child needs, if it is to develop a healthy self. One is the feeling "that another person possesses special qualities of strength and knowledge, and exercises these powers benevolently, the so-called idealizing selfobject experience." It is clear that what Straus calls the "mistical" conception of self found in Northern Cheyenne culture makes sense of this feeling. If a culture represents the locus of agency and power as lying, in considerable measure, outside the boundaries of the self, then the feeling that another person possesses special powers is transparently interpretable within that culture. Correspondingly, within an individualistic cultural representation of agency, the feeling that other selves possess special powers is at best puzzling; at worst it is denied altogether, bringing about the need for specialized discourses, such as those of psychoanalysis, to acknowledge it We are reminded in this case of the statement by Daniel's New Tamil informant: "even I did not know that I had it then." The informant is referring specifically to the mysterious "aloneness-disorder" (tanimai töSam), which for the New Tamils of Vavuniya is only a nascent cultural category, one not yet adequately articulated with a broader cultural construction of the self. The same informant remarked that "nobody knows what tanimai töSam is." In the first instance, there is the recognition of a previous experiential state that could not be lexically labelled. The labelling process begins to make sense of the experience by drawing it into the web of language. But in the second instance it becomes apparent that labelling is not enough to fully interpret it, since the label itself has not been adequately integrated into a set of statements or beliefs that prove generally satisfying to a community. In the case of "aloneness-disorder," the statements connected with it are designed to interpret an underlying experience of the self, one which, in this case, may again be related to the self psychological perspective, with aloneness translated as the failure to experience a sense of wholeness or completeness. Even though the New

Introduction 7 Tamil statements have not yet evolved into a system that is satisfying to the community, it is important to note that to be satisfying they must interpret these lower levels of semiosis. Put differently, the goal of such statements is to become transparent to self experience. We can readily imagine that in almost any culture, if not every culture, some significant sets of statements will have such goals, especially those related to medical and therapeutic practices, but also others in the realm of religion or social practices. It would be a mistake, however, to assume that all cultural statements that could be seen as pertaining to the self or person have this interpretation of self experience as their goal. Of course, Mauss and Hallowell both used evidence from many different realms to support their arguments regarding the cultural image of the self. This they did under the assumption that in each culture there was one such image, a question taken up in this volume by both Leaf and Singer, who conclude that there may be multiple culturally constructed selves within a given community. But the important point here is that the goal of transparency to underlying levels of self experience may pertain to some areas of discourse more than others. In particular, the realm of legal discourse, addressed in this volume by Rosen, is less likely to rely for its communally satisfying qualities on its ability to interpret underlying self experience. This is so because legal discourse is concerned at least in part with solving or avoiding problems that have to do with others rather than with the self. If the other could never be perceived as aggrieving the self, then all discourse could be essentially therapeutic. But legal discourse is first and foremost a discourse stemming from the experience of otherness, where the other is explicitly outside the boundary of the self. Consequently, it tends to frame the person as other, rather than as a self, and its satisfying qualities for the community will more likely pertain to its ability to make sense of the experience of the other. When we refer to differences in discourse here, we have in mind genre differences, ways of talking about the individual that are relatively more oriented to the self and self experience versus those that are more organized around the other. The distinction is related to but distinct from that developed by G.H. Mead (1934) between the "I" and "me," which referred to the idea of the self as subject or as object. Here we are referring to legal discourse as oriented not to the self — whether as subject or object — but to the other. It is also related to but distinct from what is sometimes referred to as the contrast between "self" and "person," where the latter is regarded as the "actual, concrete participant" in social relations (Smith 1985: 61). In suggesting the possibility of a genre distinction between relatively more therapeutic and relatively more legalistic discourse, we are not

8 Greg Urban and Benjamin Lee referring to participants in social relations generally, but specifically to points of view relative to those participants, points of view that are encoded in discourse. The main differentiating feature is the basis on which the discourse achieves its communally satisfying status, whether as transparently interpreting self experiences or as comprehending political realities. If there are such genre differences as at least incipient tendencies in many different societies, it becomes possible to question the assumption that in each culture there is a single category of self or person. Where this genre difference is pertinent, there will be different ways of talking about individuals. The relative coherence versus lack of coherence between these different ways becomes an empirical question. This is not to deny Mauss's claim that there is such coherence, only to suggest that it must be demonstrated empirically. In Mauss's own data, we can see an array of genres at work, from religious beliefs and practices, to law, to philosophy. In what measure is it safe to assume that each of these discourses constitutes a common object — a cultural self? Of course, it is possible to point to cultural differences between analogous genres in different societies. Rosen does just this, showing that there are differences between the Arab and Western legal traditions as regards the construction of the individual. Particularly intriguing is the insight that a "person constantly displays what lies inside him when he speaks or acts," so that there is less emphasis than in Western law on the conception of inner intention as distinct from outer consequence. The law does not seek to penetrate into the mind of the individual, but rather to evaluate the person in terms of the web of social relations of which he is part and of his history of actions. But this does not answer the question of culture-internal coherence among the genres. In regard to the latter, it is possible to search for points of convergence between distinct genre-encoded images of self and person within a given culture. One such point might be perceived in the concept of a self that is not coterminous with the limits of the individual body. Such a convergence would be especially significant in light of Mauss's original claims. In the case of Arab law, the notion of responsibility connects the individual to a web of persons, a social network, which must be taken into account in each case. It is as if the individual per se were not right or wrong, guilty or not guilty, independently of this web. Therefore, although legal discourse would ultimately rely for its communally satisfying qualities on criteria other than those of self-experience, it may share with that other discourse a view of the person as extending beyond the limits of the physical

Introduction 9 body. In this sense, it might be possible to speak of an Arab cultural category of person or self distinct from the individualistic Western conception. In this regard, we should note that discourse oriented to the other — as distinct from the self and as a potential threat to it — would naturally tend to constitute the other as greater than the individual body. The reason for this is not to be found in the self psychological sense of wholeness, which is what ultimately determines the communally satisfying character of therapeutic discourse, but instead in the practical experience of politics and alignments, where it is difficult to dislodge any action of the other from its political ramifications. If all acts can potentially be politicized, then it would be quite erroneous to constitute them as individual. The insight that emerges from this view of the self, therefore, has to do with the potential convergence between legally and therapeutically oriented discourses, a convergence that involves the relationship between certain self-experiences, as they are understood in the self psychological framework, on the one side, and an appreciation of the potentially politicizable character of actions, on the other side. This natural convergence between discourses, which may be largely fortuitous, nevertheless casts into sharp relief the central question of how and why a Western individualistic conception of the self, as coterminous with the body, might have arisen. One potential answer to this latter question emerges as we continue to break down the notion of a "cultural category" of the self or person. We have spoken of a "point of convergence" between different discourses. But this convergence is not necessarily itself something that becomes a cultural fact, represented in signs. In the case of legal and therapeutic discourses, it is possible to imagine societies in which the two can peacefully coexist side by side without entering into contact with one another. They may simply be discourses that relate to different communal needs, the needs to interpret politics and self experience, respectively. The convergence is something that is discovered by an outside observer. Another plane of semiotic problem arises, however, when we consider the ways in which discourses enter into competition with one another, each laying claim to being the best representation of reality. This jostling about of discourses is the theme of papers in the present volume by Varenne and Boon. Varenne in particular explores the varying images of Ireland encoded in different voices, each seeking to define the "true" Irish self—what language that self uses, in what part of Ireland it resides, what its customs and lifestyle are. At the same time, Varenne tries to situate his own voice among these voices. The anthropological description of the Irish self jostles about and competes with other projected images.

10 Greg Urban and Benjamin Lee Immersion into the fray of discourse is the central theme in Boon's paper, which endeavors to play several voices off against one another, and thereby to situate itself. It is the operation here that is of present concern. The voices in this case are all of a single genre, what Boon labels "self-help" anthropology. Boon does not deal explicitly with how distinct genres come into relationship with one another. But the underlying operation provides some insight in this regard. Boon's paper is meta-discursive, in the sense of taking discourse itself as an object, and we can conjecture that the active interconnection of genres must similarly involve a meta-discursive moment. In such a moment, the self appears not as an image embedded within discourse, which must be painstakingly culled by the analyst, but rather as an explicit category constituted by reflection. It is not coincidental, therefore, that when Mauss (1985: 20-22) comes to consider the greatest flowering of the Western individualistic conception of the self, he turns to philosophy, and to the writings of Spinoza, Descartes, Hume, Berkeley, and Kant. Here we are dealing with a discourse genre that is, on the one hand, merely one genre among many. On the other hand, insofar as it is constituted by reflection, its basis for evaluation is quite distinct from that of other genres. It becomes in effect a master genre, which can claim legitimacy insofar as it is able to successfully articulate understandings from any other genre, to provide commonalities and shared insights, which it can do successfully only insofar as it is able to guard against objections that come from any domain of culture. The best example in the realm of the self is perhaps the work of Descartes, whose cogito argument is explored in this volume by Lee. For Descartes, the selfawareness encoded in the proposition "I think therefore I exist" became the only undeniable claim in a scheme based upon systematic doubt. It thus became what in the present terms is the least common denominator in the competition between voices. Of course, other philosophers have not centralized the self in this way. But when they do turn their attention to the self — as did Kant, discussed in the present volume by Leaf — the basis of the constitution of the category becomes its ability to provide a point of agreement for different discourses. It is a product of the collision between voices and even whole genres. Is it the case that the individual body becomes the locus of the self when there is a generalized competition between genres, which is largely a Western phenomenon? Is the self-conscious being, localized in the material of a human form, a kind of least common denominator in a culture where points of view systematically clash? These are some of the questions that the papers in the present volume suggest but leave unanswered, awaiting further research. Based on

Introduction 11 the earlier discussion, one might conclude that the least common denominator between discourses, judging from the comparison of legal and therapeutic discourses, would be a view of the self as greater than the individual. But this does not seem to be the case in the West. It is true that the situation here is more complex than we have portrayed, since there are many more genres than the legal and therapeutic. Among these is science, a genre from which the personal pronouns have been in some measure banished, and which has elevated the third person and material substance to the status of undeniability. But even focusing just on the interaction between therapeutic and legal discourses alone, it is possible to see grounds for the emergence of an individualized self concept. This is so because the concept of a "least common denominator" necessarily involves the interaction between points of view, and in this sense it is different from the notion of a "point of convergence," which is an element of commonality perceived from an external point of view. From the point of view of the self experience of wholeness, the other is sharply distinct from the self object. The selfobject is associated with a feeling of completeness, the other with a sense of boundary, limit, and externality to the self. Consequently, when the two discourses come into contact, they necessarily conflict. What makes sense of the actions of the other calls into question a merger of the self with its surroundings; what makes sense of the feeling of unity with the world calls into question the externality of the other. One hypothesis in this regard is that the notion of self and other as localized in the individual body and psyche serves to mediate this conflict. It sets the limits of selfhood, but it also keeps the experience of the other from contradicting the understanding of self. This is in any case one hypothesis. A full analysis remains to be done of whether and how the individualistic psychological conception of self may have emerged from the interplay of many different discourses in the complex cultural history of the West (cf. Carrithers 1985, and also Dumont 1986). The present volume in this respect contains only some of the sign posts along the way to a semiotically-informed cultural understanding of the self. The volume is part of an ongoing attempt to reassess what is meant by a "cultural understanding," localizing cultural phenomena in the diverse kinds of sign functioning to which they are appropriate. It is fitting that it should end with a contribution by Milton Singer, who has done so much to stimulate thinking in this area, and whose own contributions continue to span the range of sign phenomena from pronouns to literature to philosophy, and back again.

12 Greg Urban and Benjamin Lee Notes 1.

2. 3.

Many of the papers in this volume were first presented at a symposium entitled "Semiotics, Self, and Society: A Session in Honor of Milton Singer," held at the 85th annual meeting of the American Anthropological Association in Philadelphia, Pennsylvania, on December 4,1986. However, the papers proved to be much more than a homage, reflecting instead a complex conversation over issues of semiosis, self and person, and the nature of culture — a conversation in which Milton Singer's voice is an important one, but only one among several. Of course, in his Latin version of the cogito, the freestanding pronoun was not used; person was expressed as an inflectional category of the verb. There are other traditions, of course, wherein the self is conceptualized dialogically, Peirce and Singer being exemplars of one such, and we would include here also G.H. Mead (1934). Among the parallel dialogical traditions we would also single out that associated with M. Bakhtin (1981, 1984).

References Bakhtin, Mikhail (1981). The Dialogical Imagination, C. Emerson and M. Holquist (trans.). (=University of Texas Slavic Series 1). Austin: University of Texas Press. (1984). Problems of Dostoevsky's Poetics, C. Emerson (trans.). (=Theory and History of Literature 8). Minneapolis: University of Minnesota Press. Carrithers, Michael (1985). An alternative social history of the self. In The Category of the Person, M. Carrithers; S. Collins; and S. Lukes (eds.), 234-256. New York: Cambridge University Press. Doi, Takeo (1973). The Anatomy of Dependence. New York: Kodansha International. Dumont, L. (1986). Essays on Individualism. Chicago: University of Chicago Press. Fogelson, Raymond D. (1982). Person, self, and identity: some anthropological retrospects, circumspects, and prospects. In Psychosocial Theories of the Self, B. Lee (ed.), 67-109. New York: Plenum Press.

Introduction 13 Hallowell, Irving (1955a). The Ojibwa self and its behavioral environment. In Culture and Experience, 172-182. Philadelphia: University of Pennsylvania Press. (1955b). The self and its behavioral environment. In Culture and Experience, 75-110. Philadelphia: University of Pennsylvania Press. Kohut, Heinz (1977). The Restoration of the Self. New York: International University Press. (1978). The Search for the Self, 2 vols. New York: International University Press. Mauss, Marcel (1985[1938]). A category of the human mind: the notion of person; the notion of self. In The Category of the Person, M. Carrithers; S. Collins; and S. Lukes (eds.), 1-25. New York: Cambridge University Press. Mead, G.H. (1934). Mind, Self and Society from the Standpoint of a Social Behaviorist, C.W. Morris (ed.). Chicago: The University of Chicago Press. Smith, M. Brewster (1985). The metaphorical basis of selfhood. In Culture and Self: Asian and Western Perspectives, A.J. Marsella; G. Devos; and F.L.K. Hsu (eds.), 56-88. New York: Tavistock Publications.

The Self in Psychoanalytic Self Psychology Ernest S. Wolf, M.D.

Ego and Self The term self has had only a peripheral role in psychoanalysis until recent years. Not until the advent of Self Psychology in the late sixties has the term self assumed increasing importance in conceptualizing the clinical analytic process and the theories stemming from that In the Indexes and Bibliographies of the Standard Edition of the Complete Psychological Works of Sigmund Freud (Volume 24 of the 24 volume Standard Edition) the term self does not appear at all as an independently defined concept To be sure, there are references to consciousness of self and to turning round of instinct upon self; and there are references to self- as prefix in terms such as self-analysis, self-destruction, self-betrayal, etc., but one looks in vain for a definition of self. Turning to the Gesamtregister of the Gesammelte Werke of Sigmund Freud one finds a reference to Selbst s. Ich ("self see ego"), equating self and ego. The same index also gives a reference to Ego s. Ich, equating the term for the first person singular pronoun with ego. In Sigmund Freud's prose Ich represents not only the first person singular pronoun and the self but also the system ego, one of the three agencies that together with the id and the super-ego make up the mental apparatus. It is interesting to note that this use of Ich as a noun is not just a Freudian idiosyncrasy. Gerhard Wahrig's Deutsches Wörterbuch (1968) gives the following definitions for the noun Ich: die eigene Person, das eigene Wesen, das eigene Innere und Äussere, Einheit des Selbstbewusstseins ("one's own person, one's own being, one's own inner and outer being, the unity of one's self-consciousness"). From these definitions as well as from actual usage during the early years of the emergence of psychoanalysis it is evident that the term Ich could mean either the self or the system ego when it was used as a noun. Gradually, and especially after the introduction of the tripartite mental apparatus (ego, id, super-ego) in the 1920s, the term Ich referred increasingly to the system ego only. Ego is the predominant term in the translation by James Strachey that we know as the Standard Edition (published beginning in the 1950s more than a decade after Freud's death) and it refers to the

16 Ernest S. Wolf system ego, the executive part of the tri-partite mental apparatus made up of id, ego and super-ego. However, the term self had not completely disappeared from psychoanalysis. In A Glossary of Psychoanalytic Terms and Concepts by B. Moore and B. Fine, self is defined as: ...the total person of an individual in reality, including his body and psychic organization; one's 'own person' as contrasted with 'other persons' and objects outside one's self. The 'self is a common-sense concept; its clinical and metapsychological aspects are treated under self image, self representation, etc. See ego, identity, narcissism. That same year (1968), Charles Rycroft in A Critical Dictionary of Psychoanalysis made an effort to distinguish between self and ego. He defined self: 1. When used by itself: the subject regarded as an agent, as being aware of his own identity and of his role as subject and agent. 2. As part of a hyphenated word: the subject regarded as the object of his own activity. The self differs from the ego of psychoanalytic theory in that a) the self refers to the subject as he experiences himself while the ego refers to his personality as a structure about which impersonal generalizations can be made; and that b) the ego, as defined by Freud, contains repressed, unconscious parts which cannot be recognized by the self as parts of itself. One of the existential criticisms of classical analytical technique is that its theory, particularly its metapsychology, leaves no room for the self. It should therefore not come as surprise to find that the highly praised exposition of psychoanalytic concepts The Language of Psychoanalysis by J. Laplanche and J.-B. Pontalis devotes 13 pages to the ego while giving no mention to the self. Similarly, Otto Fenichel's masterful and comprehensive The Psychoanalytic Theory of Neurosis had not given any reference to the self in its generally quite thorough index. I have cited the above references not to point up any deficiencies in either indexing or lexicographic skills revealed by the strong emphasis on the use of the

The Self in Psychoanalytic Self Psychology 17 term ego and the relative absence of interest in using the term self. The reason for the preference of ego over self is to be found in the structure and evolution of psychoanalytic theory. In the tradition of Helmholtz and in keeping with the contemporaneous trends in late 19th century thinking in the medical sciences, Freud detested any taint of the animistic or mystical and insisted on a strict adherence to materialistic-objective criteria for data collection in psychology just as in the more traditional natural sciences. Yet the investigation of the neuroses confronted him with behavioral disorders that he had the genius to recognize as resulting from thoughts and feelings, i.e., psychological experiences and memories, and not from demonstrable changes in the structure or chemistry of the brain and nervous system. This was — and is — a revolutionary discovery that has found only grudging acceptance by the medical profession. Many psychiatrists are still looking for the chemical that will cure the ills imposed on the personality by fear and shame. Freud invented a method for systematically studying the thoughts and feelings of his patients, the psychoanalytic method, and crystallized his discoveries into a psychoanalytic theory. As a by-product of his psychoanalytic explorations he noted that his patients benefited from the psychoanalytic investigative procedure and he ascribed this desirable outcome to the insights into their unconscious conflicts that his patients had gained. Thus psychoanalysis was discovered to be a treatment for neurosis. Though the data that Freud, and, initially, his colleague Breuer, collected by listening to patients intently, hour after hour, daily, for weeks and months — an unprecedented clinical procedure — were leavened with perceptions obtained by Einfühlung (to feel oneself into, empathy) Freud could not afford to emphasize this subjective aspect of his new psychological science. Agosta (1984:44) states: Although the paths of Husserl and Freud never crossed...they were both in a similar position with regard to the word 'empathy' itself. Neither of them could use the word in the full sense of interhuman understanding with which we use it today... Freud could not apply the word 'empathy' directly in discussing his clinical technique without evoking allusions to Lipps' metaphysical psychology of beauty. In fact, Freud used the terms Einfühlung, einfühlen or sichhineinversetzen fifteen times without, however, any major discussion of this activity. A few quotations will illustrate. Stressing the role of empathy in education he wrote: "only

18 Ernest S.Wolf someone who can feel his way into the minds of children can be capable of educating them" (Freud 1913: 189). One of his papers on psychoanalytic technique called attention to the need for empathy in clinical analysis: "It is certainly possible to forfeit this first success if from the start one takes up any standpoint other than one of empathic understanding such as moralizing" (Freud 1911-1915: 140).1 The last mention of empathy came in 1921: "we are faced by the process which psychology calls empathy and which plays the largest part in our understanding of what is ego-alien [Ichfremde] in other people" (Freud 1921:108).2 And finally, "a path leads from identification by way of imitation to empathy, that is, to the comprehension of the mechanism by means of which we are enabled to express any opinion at all towards another mental life" (Freud 1921: 110, note 2).3 I believe it is significant that after 1921 there was no further mention of empathy by Freud. His pivotal work, "The Ego and the Id," was published in 1923 and the era of ego psychology was launched. Empathy seemed to have been forgotten and, according to Margulies (1984: 1025), the process itself remained a neglected and unsolved theoretical problem. Its rediscovery must be credited to Kohut for whom it became the sine qua non for both clinical and theoretical conceptualizations in psychoanalysis: Empathy is not a tool in the sense in which the patient's reclining position, the use of free associations, the employment of the structural model, or of the concepts of drive and defense are tools. Empathy does indeed in essence define the field of our observations. Empathy is not just a useful way by which we have access to the inner life of man — the idea itself of an inner life of man, and thus of psychology of complex mental states, is unthinkable without our ability to know via vicarious introspection — my definition of empathy — what the inner life of man is, what we ourselves and what others think and feel (Kohut 1977: 306). The emphasis on the ego in psychoanalytic theory and practice together with the relative absence of interest in the self was an expression of a trend in psychoanalytic theorizing that took psychoanalysis, from the 1920s until very recently, increasingly further away from the data of clinical psychoanalytic practice. To be sure, there were exceptions, especially in Britain where W.R.D. Fairbairn, H. Guntrip, D.W. Winnicott, M. Klein and J. Bowlby rejected the excessive

The Self in Psychoanalytic Self Psychology 19 preoccupation with ego psychology that was characteristic especially of post-World War II American psychoanalysis (Bacal 1987). Their collective conceptualization of a psychoanalytic object relations theory achieved in Britain the kind of revitalization of psychoanalytic theory and practice that H. Kohut has accomplished in the United States with self psychology. Whether a psychoanalytic theorist uses ego or self is of crucial significance for both his theorizing as well as his clinical practice. The study of the ego is in essence an investigation of the external world with our sense organs, while the study of the self is in essence an investigation of the inner world (Omstein 1978: 28; Wolf 1982: 23-5). This statement may strike one at first as both an oversimplification and as very astounding because it seems to imply that psychoanalytic ego psychology observes not intra-psychic conflicts, as it claims, but psychological behavior on the couch. And, indeed, the patient's behavior in the form o f f r e e associations, other speech, postures and gestures and other actions are the data from which the ego psychologist infers, with the help of his theories, what he surmises to be the patient's intra-psychic conflicts. By contrast, the psychoanalytic self psychologist "can observe, strictly speaking, two psychoanalytic processes simultaneously — within the analyst and within the analysand — as well as the relation between these two processes. This observation follows from the definition of psychoanalysis as a psychology of intrapsychic processes; a correlate is the definition of specifically psychoanalytic data as those obtained by introspection and empathy..." (Wolf 1976: 111); "Kohut has shown that free associations and the analysis of resistances are not the basic tools of psychoanalysis, but are to be considered auxiliary instruments employed in the service of the introspective and empathic mode of observation" (Wolf 1982:25). What is the essence of this shift from the ego to the self? The ego, as we saw above, is conceptualized as part of a psychic or mental apparatus. The psychoanalyst as ego psychologist observes and studies a psychological structure, the ego of his analysand, that exists outside of his, i.e, the analyst's, mental apparatus. In principle, the approach is that of the natural scientist who attempts to study the world outside himself. The ego psychologist's data result from the manifestations of the ego's actions. Most prominent among these are the analysand's behavior, particularly the verbal behavior on the analyst's couch. Guided by psychoanalytic theory the ego psychologist infers from the analysand's speech (and silences) as well as from other behaviors he might observe — e.g., a tense and anxious restlessness, or an angrily raised tone of voice, or a depressed mood, etc. — something about the structure and intensity of the conflicts between

20 EmestS. Wolf ego and id or between ego and super-ego or between ego and environment. The insights gained by the analyst are conveyed to the analysand through verbal interpretations and other explanations. It is an attempt to objectively observe the manifestations of psychic structures interacting with each other, those within the analysand's mental apparatus as well as the interactions of analyst and analysand. The psychoanalytic self psychologist defines his science and its field of observation differently. For the self psychologist, psychoanalysis is the science of subjective experience. Kohut's (1977: 302) most concise definition states "that psychoanalysis is a psychology of complex mental states which, with the aid of the persevering empathic-introspective immersion of the observer into the inner life of man, gathers its data in order to explain them." Atwood and Stolorow (1984:41) have elaborated Kohut's definition of psychoanalysis as follows: ...that psychoanalysis seeks to illuminate phenomena that emerge within a specific psychological field constituted by the intersection of two subjectivities — that of the patient and that of the analyst. In this conceptualization, psychoanalysis is not seen as a science of the intrapsychic, focused on events presumed to occur within one isolated 'mental apparatus.' Nor is it conceived as a social science, investigating the 'behavioral facts' of the therapeutic interaction as seen from a point of observation outside the field under study. Rather, psychoanalysis is pictured here as a science of the intersubjective, focused on the interplay between differently organized subjective worlds of the observer and the observed. The observational stance is always one within, rather than outside, the intersubjective field...being observed, a fact that guarantees the centrality of introspection and empathy as the method of observation.

The Subjective Self The outline of the changing conceptualizations of ego and self given above will serve to give some context for a more detailed discussion of the self as seen by self psychology. The self may be defined as an organization of a person's subjective experiences that results in his/her having the conviction of being an individual person. The subjective experiences thus organized consist of conscious or

The Self in Psychoanalytic Self Psychology 21 unconscious knowledge resulting from what the person has actually felt and endured, for experience is knowledge resulting from actual observation or from what one has undergone. A number of conditions are required in order for an infant to attain a sense of selfhood, i.e., for a subjective experience of self to come into existence. Critical are the perceptions that the infant has of its significant caregivers. Usually those are primarily the mother during the earliest weeks, and then, later, also other family members or their surrogates. The needed perception is that of the presence of another person functioning as a selfobject and thus providing a selfobject experience. A selfobject experience is the result of certain functions performed by an object (e.g., the mother) — the selfobject — that evoke in the subject the organization of its experiences into a sense of selfhood, that is, into a self. Four types of these functions, the selfobject functions, have been investigated. Apparently, a developing child needs to experience all four kinds of selfobject functions for a healthy, whole and cohesive self to develop: 1) the child needs to feel that another person acknowledges that it is a valued person, the so-called mirroring selfobject experience which thus affirms the child; 2) the child needs to feel that another person possesses special qualities of strength and knowledge, and exercises these powers benevolently, the so-called idealizing selfobject experience during which the child experiences itself as merged with the idealized selfobject; 3) the child needs to feel its essential likeness to another person, the so-called alterego selfobject experience which the child experiences as an approval of its being different from still other selfobjects; 4) the child needs to feel it can assert itself against another person's benevolently opposing intent, the so-called adversarial selfobject experience which the child experiences as a potential for self-expression (Wolf 1980: 125-6; Lachmann 1986: 341-55). Reconstructions of a person's developmental history as it emerges in the course of a psychoanalysis allows some plausible inferences about the origin of the self. Such retrograde constructions from psychoanalytic data present evidence that a "nuclear self' normally has arisen out of the various selfobject experiences during the child's second year, probably around 18 months. However, recent findings resulting from research with infants suggest that a "core self' is "likely to form during the first half year of life as a primary social task..." (Stern 1985: 69-123). Stern tentatively lists four types of experiences available to the infant that are needed to form an organized sense of a core self. These include self-agency, selfcoherence, self-affectivity and self-history: "these four self-experiences, taken together, constitute a sense of a core self. This sense of a core self is thus an

22 Ernest S.Wolf experiential (my italics) sense of events" (Stern 1985: 71). It is noteworthy that the very different data of psychoanalytic self psychology and of infant research lead to conclusions that, though they are not at all identical, do not significantly contradict each other (Lichtenberg 1983). Both speak about the same basic unit of observation and study, the subjective self, even while the infant observations seem to discern a self earlier than psychoanalytic observations. Depending on the quality and content of the selfs constitutive selfobject experiences, it will emerge in a more or less healthy state. Indeed, under certain conditions that are inimical to self development no completely structured self will be organized. In these cases the crippled remnants of disturbed and disorganized — essentially incomplete — self development are likely to be manifested in later life as the severe psychopathology associated with major psychotic illness such as schizophrenia. But even if a complete self is evoked during early development it may come into existence in a very fragile state as a result offaulty developmental selfobject experiences. Such a fragile but still cohesive self will have a tendency to regress into a state of fragmentation. The latter will be manifested as either a Borderline State with violent and mainly uncontrollable affect storms or, if the disturbance in development was less severe, a personality disorder (=neurotic character structure) will ensue. However, a self that has not been exposed to a series of major psychological traumas and that, in general, had the good luck to be able to grow and flourish in an ambience of feeling protected, cared for, valued, and, perhaps most important, responded to as someone important, accepted and worthy of the parents and other caregivers — such a self will experience itself in a state of well-being, able to function and exercise its faculties and able to initiate and maintain satisfying relations with others. We have designated this a cohesive self. The fragile self, easily regressing into fragmentation, can be recognized in a number of editions: a self that experiences itself as inherently significantly impaired in its feeling of well-being or functioning, or a self that experiences itself as having lost some of its former cohesion and as having returned to a precohesive, more or less fragmented state. These selfs whose distress is usually classified as Narcissistic Personality Disorders are the best candidates for psychoanalytic treatment. The structure of their subjective distress (I am trying to avoid the use of the term psychopathology because their symptoms are evidence for active self-protective measures) results from the inimical experiences of early life that forced them at that time to adopt defensive measures, among which one finds denial, or erotization, or excessive submission, or excessive assertiveness, all of

The Self in Psychoanalytic Self Psychology 23 which have the effect, singly or in their complex aggregate forms, of interfering with smooth and close relationships to others. As a consequence, they loose the easy intimacy and affective closeness that is experienced as a positive selfobject ambience and that is so essential for the strengthening and sustenance of a healthy self. The built-in fragility and the resulting defensive reactiveness of the self — beyond interfering with social relationships — combine to produce a heightened subjective sense of vulnerability to all sorts of psychological dangers, leading to a chronic level of increased anxiety and a readiness to panic, or to retreat into depression or to explosion into aggression. These symptoms may look to the outside observer as a kind of sick driveness from within and one can understand how Freud came to think of them as the manifestations of instinctual drives in conflict. We think that they are better understood as desperate attempts to maintain some semblance of the crumbling structure of the self behind the screening (and often screaming) barrier of the symptomatology.

Summary Psychoanalysis, as a practice for the investigation of the human psyche, approaches its task by studying subjective experience with the method of introspection and empathy. However, classical Freudian psychoanalysis, in reporting its findings, simultaneously translates them into the terms of objective observations in keeping with the traditional observational stance and language of science. In so doing the resulting conceptualizations — e.g. about ego, id and super-ego and their interactions — become increasingly experience-distant, i.e., distant from the data of observation. Psychoanalytic Self Psychology attempts to conceptualize its data of subjective experience in a language of subjective experience. The resulting conceptualizations — e.g., about the varieties of the experience of self and of so-called selfobjects — remain experience-near, i.e., close to the data of observation.

Notes 1.

Strachey's translation in this sentence of Einfühlung understanding" is subtly but significantly in error.

as "sympathetic

24 Ernest S.Wolf 2.

3.

Freud's original phrase is ambiguous (Freud 1968: vol. 13, 119) and was translated by Strachey as "understanding of what is inherently foreign to our ego in other people." Strachey's translation of Stellungnahme as "take up any attitude" seems less precise than what I believe Freud intended here.

Rtferences Agosta, L. (1984). Empathy and intersubjectivity. In Empathy, I, J.Lichtenberg, M. Bornstein, D. Silver (eds.), 43-62. Hillsdale, N.J.: The Analytic Press. Atwood,G. and Stolorow, R.(1984). Structures of Subjectivity. Hillsdale, N.J.: The Analytic Press. Bacal, H. (1987). British object relations theorists and self psychology: some critical reflections. International Journal of Psychoanalysis 68:81-98. Fenichel, O. (1945). The Psychoanalytic Theory of Neurosis. New York: W.W. Norton. Freud, S. (1911-13). On beginning the treatment. Standard Edition 12:121-144 (1913). The claims of psychoanalysis to scientific interest. Standard Edition 13:165-190. (1921). Group psychology and the analysis of the ego. Standard Edition 18:69-143. (1968). Gesammelte Werke: Gesamtregister, compiled by L. VeszyWagner. Frankfurt am Main: S.Fischer Verlag. (1974). The Standard Edition of the Complete Psychological Works of Sigmund Freud; Indexes and Bibliographies, compiled by A. Richards. London: The Hogarth Press. Kohut, H. (1977). The Restoration of the Self. New York: International University Press. (1978). The Search For the Self, 2 vols. New York: International University Press. Lachmann, F. (1986). Interpretation of psychic conflict and adversarial relationships: a self-psychological perspective. Psychoanalytic Psychology 3(4):341-355. LaPlanche, J., and Pontalis, J.-B. (1973). The Language of Psycho-Analysis. New York: W.W. Norton.

The Self in Psychoanalytic Self Psychology 25 Lichtenberg, J. (1983). Psychoanalysis and Infant Research. Hillsdale, NJ.: The Analytic Press. Margulies, A. (1984). Toward empathy: the uses of wonder. American Journal of Psychiatry 141:9,1025-1033. Moore, B., and Fine, B. (1968). A Glossary of Psychoanalytic Terms and Concepts. New York: The American Psychoanalytic Association. Ornstein, Paul H. (1978). The evolution of Heinz Kohut's psychoanalytic psychology of the self. Introduction to The Search For the Self, 2 vols., by Heinz Kohut, 1-106. New York: International University Press. Rycroft, C. (1968). A Critical Dictionary of Psycho-Analysis. New York: Basic Books. Stern, D. (1985). The Interpersonal World of the Infant. New York: Basic Books. Wahrig, G. (1968). Deutsches Wörterbuch. Gütersloh: Bertelsmann. Wolf,Ε. (1976). Ambience and abstinence. Annual of Psychoanalysis 4:101-115. (1980). On the developmental line of selfobject relations. In Advances in Self Psychology, A.Goldberg (ed.), 117-132. New York: International University Press. (1982). Comments on Heinz Kohut's conceptualization of a bipolar self. In Psychosocial Theories of the Self, Β I x e (ed.), 23-42. New York: Plenum.

The "I" Of Discourse Greg Urban

The analysis of the first person singular pronoun ("I"), as outlined by Benveniste in his now classic articles "The Nature of Pronouns" (1971a [1956]) and "Subjectivity in Language" (1971b [1958]), has supplied a framework for conceptualizing the relationship between self, language, and ultimately culture. This analysis has been taken up in theorizing about the self that is of a semioticphilosophical character, especially in the work of Ricoeur (1974) and Singer (1984), but also in the semiotic-linguistic framework developed by Silverstein (1976). I wish to argue in this paper that the analysis proposed by Benveniste is only partially adequate, and that, in fact, an empirical investigation of the use of the personal pronoun "I" in actual discourse reveals a much richer picture of the semiotic functioning of that pronoun. Ultimately, this enriched picture leads to a modified conceptualization of the relationship between self and culture. Specifically, I propose that in narrative discourse "I" occurs predominately within quotation marks, and there it acts as an anaphoric device, analogous to the third person anaphoric pronouns (in Hnglish, he, she, it, and they). This "I" conforms only apparently to the Benveniste analysis, its character as a "referential index," to use the semiotic terminology, arising only metaphorically through the semantically described situation, and being considerably removed from true tokenlevel indexicality. There is also, in some instances at least, a kind of "de-quotative T," where the metaphorical "I" of quotation, through a kind of theatrical substitution, becomes again a referential index, but this time pointing to the speaker not with respect to the speaker's everyday identity or self, but rather with respect to an identity the speaker assumes through the text I will argue that this substitution of de-quotative "I" for referential indexical "I" is at the heart of the cultural construction of self.

Indexical-Referential "I" Benveniste's analysis of the personal (first and second person) pronouns focuses on their referential character. Benveniste recognized that, from the point of view of reference, personal pronouns appear to be wholly distinct from the common noun

28 Greg Urban phrases, and even from the third person pronouns, with which they might naturally be compared. Common nouns as abstract types have a more or less definite referential value. In English, the word table, taken in the abstract, has associated with it a stereotypical class of objects, a notion of what a "table" is. As Benveniste (1971a: 218) writes: "each instance of use of a noun is referred to a fixed and 'objective' notion, capable of remaining potential or of being actualized in a particular object and always identical with the mental image it awakens." In the case of personal pronouns, however, there is no such abstract class. Personal pronouns are " 'empty' signs that are non-referential with respect to 'reality'" (1971a: 219). "Each /has its own reference and corresponds each time to a unique being who is set up as such" (1971a: 218). This referentially "empty" character of "I" arises from the fact that the first person pronoun is indexical. Reference is achieved at the token level through an actual continguity between the instance of language use in which /occurs and the utterer of that instance. Consequently, each time /is used it has a singular and definite reference, but as an abstract type, apart from this instance, it has no associated class of objects. Of course, normally one thinks of utterers as human. However, the use of / in quotations, especially in myth narrative, leads to the realization that virtually anything can be construed as a speaker, and, as a result, can be taken as the singular definite referent of /. The class of referents is left unspecified at the level of type. The shifting referential character of I led Benveniste to formulate a tokenreflexive definition of the form: "lis 'the individual who utters the present instance of discourse containing the linguistic instance of /'" (1971a: 218). The definition thus differs in an important way from the kinds of glosses that can be given of common nouns, e.g., "featherless biped" for man, where the gloss can be achieved abstractly, at the level of type, without any direct reference to concrete instances (cf. the concept of "stereotype"). From a broader perspective on pronouns, "you" is similar to "I" in being indexically referential, but both first and second persons differ sharply from third person, which is in Benveniste's terms a "non-person." In its method of achieving reference, third person is more like common nouns than like the true personal pronouns, in that it is non-indexical and can be glossed purely at the level of type, e.g., he is an entity conceptualized as of male gender. Even in a decontextualized sentence, it retains some referential value, however more definite that value may become in a given context of use. So the principal split in languages is between the indexical referential pronouns and the semantically referential pronouns and nouns.

The"!" of Discourse 29 This view of "I" is, I believe, only partially adequate. It provides an intuitively satisfying characterization of / a s part of the language code, if one simply reflects upon the nature of language and how it is used. But an actual empirical study of texts reveals phenomena, such as the use of I within quotation, that require an amplified conception of how this pronoun is actually functioning. At the same time, Benveniste's analysis has been taken up by theoreticians and philosophers concerned with the "self." As Singer (1984: 61) suggests, Ricoeur believes that Benveniste's analysis of pronouns provides a "new basis for a 'hermeneutics of the I am' as well as for a reconciliation between structuralism, psychoanalysis, and philosophy." Essentially, "I" is one of the hinge points between language, as an abstract Saussurean structure of oppositions, and discourse, as a specific instance of language use. Consequently, it represents a kind of socialization of the self, as it is brought into a culture-specific structure. The personal pronouns make possible an expression of "subjectivity." The analysis is taken much further by Singer (1984: 74-104), who in his chapter on "Personal and Social Identity in Dialogue," building upon Peirce and Benveniste, discusses the specifically dialogical character of the self constituted through language. The opposition is not between self and other, but, since '"all thought is addressed to a second person'" (1984:85), between an already interacting "I"-"you" dyad and a third thing. In these passages, Singer seems to echo the dialogism of Bakhtin. But the "I" of discourse is not only an actual in-the-world subject, indexically referred to by means of the first person form. The discourse "I" can also be any being or entity, imaginary or not, capable of being reported as a speaker. The central question here concerns the relationship between this reported "I" and the indexical referential "I" that points to a subject. I wish to argue that the true hinge point between self and "culture," the point at which the self becomes a socialized subjectivity, is not to be found in the relationship between indexically referred to "I" and the abstract Saussurean system. It is be found in the relationship between the quoted "I" of discourse and the indexical referential "I" of the language code.

Anaphoric "I" Benveniste (1971b: 225) noted that "it is a remarkable fact...that the 'personal pronouns' are never missing from among the signs of a language, no matter what

30 Greg Urban its type, epoch, or region may be." Whether this stands up under comparative scrutiny — as it apparently would — it is equally remarkable that languages also have a distinction between direct and indirect quotation, as shown in the following contrastive pair from Shokleng, an Amerindian language of Brazil: (1) e tei) yk tä ke mü he-coref. go+stat. that he say active "Hej said that hej was going." (2)

eft ceo w ä ke tä I go Stative say he "Hej said, 1j am going.'"

mü active

[indirect quotation]

[direct quotation]

Grammatically, there are important distinctions between the indirect and direct forms of reporting the speech of another. These parallel the distinction between indirect and direct discourse in English. The most important of these, for present purposes, concerns the use of pronouns, especially the first person forms. In both forms of reporting speech in examples (1) and (2) the main clause has as its subject the third person "he"(tä= nominative form). The semantic information encoded in this main clause is simply that the third person subject has engaged in an act of saying, the specific contents of which are recorded in the embedded clause. The embedded clause in each case has as its subject the same entity that is subject of the main clause. It is in the pronominal shape of that embedded clause subject that the key difference between the indirect and direct forms emerges. The indirect form uses a third person pronoun, which specifically indicates coreference (β). In the direct form of quotation, however, in Shokleng as in English, the embedded clause has as its subject the first person pronoun en CI1)· The decoder of this sentence infers from the use of "I" that a co-referential relationship obtains between the subject of the main and embedded clauses. From the point of view of the earlier discussion of "I" as referential index, the central problem posed by this usage of "I" is that it is not indexical. It does not point to the speaker of example (2) as the subject of the embedded clause. In such cases, the pronominal shape M, understood intuitively at the code level as a referential index, functions instead as an anaphoric device, indicating the coreferential relationship between the subjects of the two clauses. For this reason, I will refer to the "I" of an embedded clause in a direct quotation as an anaphoric urn

The "I" of Discourse 31 This anaphoric first person is functionally equivalent to an anaphoric third person, whatever the differences between these two as non-anaphoric forms may be. When it comes to establishing coreference, neither form is functioning "referentially." Consequently, the distinction between indexical and semantic reference, as discussed by Benveniste, is neutralized. First and third person, as anaphoric devices, are both tools for establishing coreference, and they both do so by signalling that the semantic content of the noun phrase for which they are substituting is to be understood as identical with the semantic content of some other contiguous noun phrase. Because anaphoric devices rely upon contiguity with other noun phrases, they are indexical. But in this regard they differ sharply from the indexical referentiality discussed by Benveniste. The indexicality of an anaphoric device is (1) purely discourse internal, involving proximity in discourse terms with the noun phrase for which it substitutes; and (2) substitutive, not pointing to or supplying information about the other token as a physical object, for which it stands as a sign, but rather acting as a kind of replacement for that other object. Since the other object is itself a sign, the anaphoric device is a substitute sign. This analysis, which distinguishes an indexical-referential "I" from an anaphoric 7", based upon their distinct characters as signs, reveals immediately the functional motivation for a distinction between direct and indirect quotation. Direct and indirect quotations provide a means for managing coreference that cannot be achieved through one or the other construction alone. Specifically, the use of an indirect quotation form signals that the personal pronouns are to be taken in terms of their ordinary indexical-referential value. Thus, the "I" of the English "he said that I..." is understood, by virtue of the indirect construction, as the ordinary indexical referential "I" discussed by Benveniste and others — "the utterer of this token" — just as the "you" in the English "he said that you ..." is understood as addressee of this token. So the indirect quotation is a way of preserving ordinary indexical-referential values within an embedded clause, insofar as personal pronouns are concerned. In contrast, direct quotation is a mechanism for signalling that personal pronouns in an embedded clause are to be understood as freed from their normal indexical-referential value and, consequently, available for coreferential use in a situation where it is necessary to manage cohesion relations between a number of discourse subjects. This becomes problematic, from the point of view of direct discourse, even when there are only two participants in the main clause that are coreferenced in the embedded clause. When the problem of multiple coreference

32 Greg Urban between main clauses in extended discourse is taken into account, the extent of the coreferencing problem becomes obvious. A simple example (from a Shokleng myth), with two participants in the main clause, will help to clarify the problem: (3)

kuyankätj te name def.

yuouQ tö falcon ergative

wQ e nOQfien nom he-coref. brother

köflQäq te öq man def. they

Ii eft cö ke kQ thus I erg. do conj. ke tä e say he he-coref.

wun get

tö erg.

mö dative

tä he

kü ÖQ ko rjfcke fiä conj. they eat habitual cont.

mä Sil kukö te you I bone def.

nflgfien brother

tß def.

klaflmär) name

tu carry te def.

yt purposive

tapä go up

mö dative

"kuyankäg said to his brother, 'when the falcon that has been carrying off men and eating them does this to me, you will go up to fetch my bones,' so he said to his brother klaflmäq." The coreference relations are perfectly explicit and unmistakable here, thanks to use of anaphoric first and second persons. The "I" of the embedded clause can only refer to the subject of the main clause, kuyankäq. The "you" can only refer to the dative object of the main clause. If one were forced to rely upon indirect quotation only, the resulting sentence would be confusingly ambiguous, as the English translation with subscripts in (4) indicates: (4) "kuyankk) said to his brother, klaflmäg, that when the falcon that had been carrying off men and eating them did this to hin^ j, hey will go up to fetch his: ; bones." This indirect quotation could easily be mistaken for: (5) "kuyankäg said to his brother, 'when the falcon that has been carrying off men and eating them does this to you, I will go up to fetch your bones,' so he said to his brother klaflmäi)."

The"!" of Discourse 33 In actuality, Shokleng has more resources than English for managing multiple coreference relations, since it has both third person coreferential (S) and noncoreferential (ti) forms. In example (1) above, the coreferential form was employed. In (6), below, the non-coreferential form is employed, the anaphoric difference in English being reflected in the subscripts: (6)

ti he

teg yb tä ke mfl go+stat. that he say active "Hfei said that hej was going."

The patterning of the coreferential and noncoreferential anaphoric pronouns is linked to the case role of the noun phrase in main and subordinate clauses, and is discussed further in Urban (1985a: 180-183). Even the relatively more powerful anaphoric capabilities of Shokleng are easily overwhelmed by the needs of coreference in actual discourse situations. While the principal verbs of speaking usually take only two noun phrases in the main clause, which then require coreference in the subordinate clause, the various possibilities with respect to case lead to hopeless confusions. Use of the anaphoric personal pronouns, however, allows for perfect clarity regarding coreference. This is because of the invariant relationship between the personal pronouns and case role in the main clause. The two rules governing coreference are: Rule 1: Within the clause containing the direct quote, first person is always coreferential with the subject (transitive agent or intransitive subject) of the main clause; Rule 2: Within the clause containing the direct quote, second person is always coreferential with the (dative) object of the main clause. While these rules are formulated specifically for the English and Shokleng cases, they may be proposed, based upon a number of additional languages studied, as universals. Specifically, I propose that every language, in addition to having personal pronouns that function indexical-referentially, will also have some means of indicating direct quotations, and in these direct quotative constructions the above rules of anaphora will apply.

34 Greg Urban If the requirements of coreference suggest a cross-linguistic functional motivation for the direct quotative form, there are also functional motivations for the indirect form. These motivations are implicit in the observation made above that in indirect quotation the indexical referential value of the personal pronouns is preserved. Indeed, indirect quotation is nothing other than a means of signalling that the personal pronouns within the quote are to be taken as having their usual values, i.e., as pointing to the actual speaker and addressee(s) of the utterance. Because direct quotation suspends these values, in favor of the anaphoric values, it has no way of referring to the actual speaker and hearer of the utterance within the quote, unless the speaker and hearer happen also to be the subject and object of the main clause, in which case the anaphora point back to them. There is no direct construction in Shokleng or English analogous to (2), for example, that would yield the meaning of the indirect sentence in (7):

(7) efl he

eft) yfe tä ke mQ go+stat. that he say active "He, said that Ij was going."

Indirect quotation continues to make possible reference to the participants in the actually occurring speech event even within the quoted clause.

The Metaphorical Basis of Anaphoric "I" At the same time as anaphoric "I" must be sharply differentiated from indexical referential "I" because of the differing mechanisms by means of which they function, it is also obvious that there is some connection between these functions. The connection is a metaphorical one, and, indeed, the metaphor is so powerful that its status as such is ignored. We look through the metaphor, so to speak, to the situation that is represented. In a sentence such as "He said, Ί am going"' the /refers anaphorically back to he, and the speak» of this sentence is not indexed. Nevertheless, the words inside the quote purport to be like the words in some original but now non-present utterance. If and insofar as they actually resemble some original utterance, the words inside the quote are iconic with that utterance. Nevertheless, they are never identical with the original words in semiotic function, because the quoted statement

The "I" of Discourse 35 is not a statement being made by the actual speaker of the sentence. The quoted portion of the sentence, "I am going," is linked metaphorically to the sentence "he was going." In effect, the speaker is suggesting that he is to be understood as like an /. The notion that the anaphoric "I" is a metaphorical "he" is somewhat more difficult to grasp than the ordinary notion of metaphor. This is because ordinary metaphor typically deals with two terms that are both purely semantic, lacking an indexical component entirely. Consequently, the two terms of the metaphor are seen as distinct but simultaneously as being related in some way, as in the expression "a sea of troubles," where the words a sea are taken as related to many, or some other expression of extent, metaphorically. When it comes to /being taken as related to he, however, the basis of the similarity is not a semantic, but a pragmatic, one. It is a question of how we are to regard "he." The anaphoric "I" tells us to regard "he" from "he's" point of view, as if "he" were a kind of "I." The anaphoric "I" entails a kind of play acting on the part of the speaker of the utterance, who regards himself as momentarily taking on the role of the third person referent Simultaneously, the hearer of the utterance is invited to regard the speaker as engaging in a kind of role playing. The metaphorical basis of the anaphora is to be found in this role playing. Hearers understand the preferential value of anaphoric "I" precisely because they know that the main clause refers semantically to an act of speaking, and they know that the speaker of the main clause is to be regarded as playing the role of speaker in the subordinate clause. It is only the first person that provides this metaphorical pivot. Second person anaphora is not based upon the hearer of the main clause playing the role of hearer in the subordinate clause. Rath«- the actual hearer is required to imagine the actual speaker — who is playing the role of another speaker — as addressing some non-present individual, as, for example, in "Hei said, 'Ii already gave it to you,'" where the actual audience need not evince any response to the discourse within the quoted clause, contrary to the speaker, who must actually utter the words of the subordinate clause speaker, and who may even take on other of the speaker's behavioral attributes. It is thus the imagination that provides the foundation for a stable system of cross-clause coreference, the ability of speakers to regard themselves metaphorically as others, and of hearers to similarly participate in this imaginary system. I want to argue that the role playing involved in anaphoric "I" is crucial to culture, conceived for present purposes as a socially-transmitted system of discourse. On the one hand, discourse may be directly transmitted through a

36 Greg Urban process of imitation. Speakers need merely repeat the words of others, without any awareness or indication that their words are borrowed. The words are nevertheless in some sense cultural. On the other hand, to use a Freudian model, simply imitated discourse will always be at the behest of instinct. The imitation can be abandoned whenever it is out of accord with desire. Consequently, it is not cultural in the further sense of being regulative of the individual. In the case of anaphoric "I," however, there is an awareness that the discourse of another has been assumed. This awareness, achieved through the metaphorization of "he," and hence fundamentally metapragmatic, places an additional constraint on the speaker. The imitated discourse of the other is no longer simply subject to whim. It is also subject to the control that the imitated other exercises over the speaker, since modifying or overturning the words of another is understood with the awareness that they are the words of another. The anaphoric "I," now also understood as a metaphorical or theatrical "I," brings into one's discourse the real control that the imitated others have over one. It creates, so to speak, the "weight of tradition." This is the essence of what might be called the regulative aspect of culture, the sway that culture exercises over individual whim. It is possible to see in this metapragmatic awareness the subjective experience that psychoanalysts discuss in connection with the term "superego." The "Γ that one assumes metaphorically is also a "not-I." It is not another addressing the actual "I," but rather an expectation or ideal that the actual Τ must live up to, thanks to the recognition that one has assumed the role of another. For this reason, it is probably better understood not with reference to the Freudian model, but rather with reference to the self psychology model developed by Heinz Kohut (1985; cf. Wolf 1986). This metaphorical "I" can become a kind of "ideal" of the self.

De-quotative "I" The anaphoric "I," with its metaphorically indexical properties, may be distinguished from what can be called the de-quotative"I." I use this term to describe the "I" of a quotation wherein the matrix clause has disappeared. A further extention of the de-quotative "I" is the theatrical "I," wherein there is virtually no trace of quote-framing, the individual speaking through the character that he or she represents. This is the "I" of theater and similar representational performances, wherein speakers use the first person pronoun to point to themselves, but not as

The "I" of Discourse 37 the individuals they are outside the performance context. The "I" they use points instead to themselves as the concrete representation of a character in discourse. Even further along this continuum is the "I" of trance, possession, and similar states, which I will refer to as the projective "I." From a purely linguistic point of view, the de-quotative "I" is identical with the ordinary indexical referential "I" discussed by Benveniste. There are no obvious grammatical properties that distinguish it, as in the case of anaphoric "I." In the elementary form of de-quotation, it is the surrounding narrative discourse that indicates how the "I" is to be understood. Typically, there are preceding main clause frames that set up the subsequent interpretation of the de-quotative "I," which is thus, in this sense, closely related to anaphoric "I" and distinguished from the ordinary indexical referential type. De-quotation is, according to this hypothesis, a special and derivative form of language use, spun off of the more widespread and frequent direct quotation, as a means of talking about the speech of others. While it can, and probably does, occur in some measure in all cultural traditions whose languages have the direct quotation form, its further elaboration is a contingent and culture-specific phenomenon. The indexical signs (voice quality, etc.), first employed together with the quoting main clauses, can later come to stand on their own as signals to the hearer that the "I" employed is to be taken as a replacement for some other noun phrase that has occurred earlier in the narration. In culture-specific elaborations of this form, it is often possible to do away with the main clause quotations altogether. When there is still a single narrator, as in the puppet theater tradition (Gross 1983), for example, great emphasis must be placed on the indexical cues that allow different de-quotative or de-anaphoric Ts" to be kept distinct so that overall discourse cohesion and coherence is maintained. The indexical cues are thus functioning metapragmatically as if they were the quoting clauses of narrative discourse. They, so to speak, presuppose the semiotic capacity to understand the quotative speech of narration, which they come to replace by convention. The actual theatrical tradition that has evolved in the West may be seen as an even further refinement of the de-quotative "I" pattern, in which the anaphorically different Ts" are given physical embodiment in the form of different individuals. The individuals become part of the indexical cues, and it is in this sense that we can understand actors striving for characteristic indices of the theatrical "I," much the way the puppeteer strives for indexical distinctiveness among the various Ts." While each actor in theater may play the role of only a single anaphoric "I," this "I" must be kept distinct from the indexical referential "I" of the individual playing

38 Greg Urban the role. Like the quoting clauses in narratives, these indices are what allow actor and audience alike to gain awareness of the two Ts." According to the present interpretation, theatrical traditions should be much less widely distributed than the direct/indirect quotation contrast, which forms the foundation for the theatrical form. In fact, there is no direct counterpart in Shokleng to this tradition. Graham (1983) describes a theater-like performance among the related Shavante Indians of Central Brazil. On one occasion, the Shavante actually used different individuals to perform a myth. Graham's account is particularly interesting, because it makes clear that the bulk of the speaking is done by a narrator, who establishes the anaphoric relations among the different actors. A study of Graham's transcriptions shows that, for whatever reasons, the first person is nearly absent from the de-quotative speaking, second person being much more frequent. In any case, the overall performance is somewhere in between a single-narrator myth-telling and our Western theatrical tradition. There is no evidence that the Shokleng have ever had any sort of theatrical production. The elementary narrative de-quotation, however, is fairly common. Indeed, close study of a number of tapes shows that there is a veritable continuum between quotation and de-quotation, as the quote-framing clause is articulated with greater or lesser intensity. In some instances, it is strongly articulated as: (8)

... ke tä ti say he-nom. he-obj. " '...,' he said to him."

mö dative

mfl active

In some, it is weakly articulated, there being in many instances only a barely audible ke. In still others there is no trace of the ke whatsoever. Consideration of the semiotic functions of de-quotation makes it clear that theatrical "I" is linked to the anaphoric "I" of narration. The theatrical event, like the narrative event, becomes a mechanism for separating out the two different kinds of "I." It allows actor and audience alike a kind of metapragmatic lever through which awareness of the social "I," the "I" of self ideals and role playing, can be achieved. But theatrical "I" goes even further down the road of involvement in another self than does the typical anaphoric "I," which may function in nearly pure anaphoric fashion, or which may entail considerable role playing on the part of the speaker. In theater, the actor can become so immersed in another "I" that that other "I" becomes once again virtually indexical referential. For this process of total immersion in another "I," there is a corollary in Shokleng. Indeed, it was that peculiar phenomenon that first enabled me to

The "I" of Discourse 39 comprehend the different kinds of "I." In the telling of the origin myth of the tribe (the wäfleklöri), the narrator frequently lapses into the first person pronoun even in presumably non-quoted portions of the narration, as in (9): (9)

efl hä cö töl ki yun w&n mü kan I focus erg. inside in emerge be first active egressive "It was I who emerged first."

Since this is a non-quoted portion of the text, one would expect the third person to have been used. It is as if, however, the narrator becomes identified with the original ancestor, the one who was presumably first to narrate this story. Through the theatrical or projective "I," the present-day narrator assumes the persona of his historical antecedent There is evidence that this shift to the first person, which is by no means consistent, as will be demonstrated below, is accompanied by subjective changes. In the one telling where shifting to the first person occurred most frequently, the narrator seemed especially bound up in the narration, paying little attention to those around him. He seemed to enter at points into what might be described as a trance, having what Americans sometimes call a "far-away look," as if he were at that moment focused intently on some wholly internal reality. In tellings where the first person shifting was less prominent, the narrator seemed more focused on the audience, and on the kind of responses he was being given. Here third person reference tended to dominate. The evidence collected thus far is actually insufficient to determine whether the projective "I" of this myth represents a lapse on the part of the narrator, or whether, in fact, it is the norm, the lapse being into third person. This is something I will discuss further below. My opinion at this time is that the projective "I" may indeed be the primary performance form, the shift to third person occurring in less formal tellings. In any case, the differences between tellings as regards pronominal usage and subjective state can be studied particularly readily in connection with the Shokleng origin myth because, like Western theatrical performances, this myth should be memorized and recited verbatim, syllable by syllable (Urban 1985b, 1986a,b). Indeed, this fact may be correlated with the tendency of the narrator to turn inward, as he searches his memory for the precise wording. Memorization may in turn be one factor favoring the theatrical "I," such as has been institutionalized in the Western theatrical tradition.

40 Greg Urban A study of different tellings shows that, in fact, whichever is the base form, "I" and "he" represent or encode alternative subjective orientations to the same material. The following examples show the same line of the myth recited in different tellings. Example (10) is from a telling of the origin myth, performed by Nil of the Macuco, that took place in 1975: (10)

wägyö tö zägpope tö patfe Bfl yo katfele relative erg. name erg. name I in front of arrive descending "Relative zäpope pate arrived in front of me."

The italics on the "me" stresses the first person orientation of the narrator, as if he were the original narrator who had undergone the experiences recounted here, which presumably took place in the earliest phases of human history. This particular line is taken from an episode in which the first jaguar is made by the character here referred to by means of "I." The narrator has become the ancestral hero. In a more didactically oriented telling from 1981 by the same narrator — in which he is trying to teach me the myth — this line is told using a deleted third person: (11)

wägyö tö zägpope tö patfe no relative erg. name erg. name in front of "Relative zäpope pate arrived in front of him."

katfele arrive descending

In a variant of this same line in a 1982 telling by a different narrator, the full third person forms are employed: (12)

ti köfikahä tö his relative erg.

zägpope name

tö patfe erg. name

ti no he in front of

katfele arrive descending "Relative zäpope pate arrived in front of him." This last telling was by a younger speaker who was much more attuned to his audience and their responses, and less mentally involuted, than the other speaker. The evidence seems to suggest that this use of first person, as a means of referring to a mythico-historical figure, is not a fully institutionalized pattern, or at

The "I" of Discourse 41 least is subject to functional forces that occasionally cause it to be substituted by third person. When it is used, it is functionally analogous to theatrical "I," in which the speaker has taken on the role of the character he might otherwise describe by means of third person. Semiotically, it involves a complicated sign process in which speaker/narrator becomes in effect the substitutive ostensive referent of a third person form, and consequently capable of referring to that referent by means of the first person form. From a psychological point of view, there is clearly a kind of maximal projection here of the speaker/narrator into another self, in this case, a character in a narrative. The projection is similar to that talked about by actors. However, the trance-like quality that is involved suggests a possible comparison with actual trance behavior and possession. It may be that in this phenomenon, as well, the first person pronominal usage, arises from a shift from the third person forms in certain kinds of narrative. As in the origin myth-telling case, first person usage may represent the speaker/narrator becoming an anaphoric substitute for a third person form. If the present suggestions should be borne out by empirical research, it will be possible to link together trance and theater, as well as other phenomena, such as the Shokleng origin myth-telling, in a single scheme, in which the discourse "I" of these forms grows out of the basic anaphoric "I" of direct quotation.

Cultural

Functions

The different types of "I" discussed thus far are shown in Figure 1. They are ranged along a formal/functional continuum, explication of which can shed light on the general problem of what cultural functions are served by the different types of "I." On the left side of the figure, discourse "I" (indexically) refers to the everyday self — the self with which the speaker is associated on a normal basis. The indexical cues of the speech tokens accompanying this "I" are those cues through which the individual wants to be known. As one moves to the right along the figure, indexical cues come into play which purport to point to a self other than the everyday self. The important point here is the signalling of a distinction between the two types of "selves." In the cases of indexical-referential versus anaphoric "I," the signalling of this distinctiveness is present in the discourse as sign vehicle itself. There is the

42 Greg Urban twofold opposition between main and subordinate clause, on the one hand, and indirect and direct quotation on the other. Anaphoric "I" is distinguished by virtue of its presence in the subordinate clause of a direct quotation. As one moves further to the right in this figure, the discourse internal signalling of the distinctiveness of selves gives way to contextual signalling. In the case of de-quotative "I," there are abundant traces in the discourse of the anaphoric character of the "I." The use of indexical cues such as intonation, voice quality, characteristic grammatical constructions, etc. first occurs in conjunction with a main or quoting clause in which the "I" is used anaphorically. At the same time, this kind of de-quotation is already set up, and its possible occurrence signalled, by the marked speech situation — telling of a story — in which it occurs. The context in some measure acts as signal. In the case of theatrical "I," context takes over as the primary signal by means of which the two selves are distinguished. Here as well there is evidence that a direct continuum from de-quotative "I," correlated with the degree to which a narrator, who sets up the third person referents of the various Ts," is present. In the limiting case of no narrator, it is nevertheless critical that the context prepare the audience to interpret the various Ts" as different characters within a story-line. The context of speech here is highly marked with respect to that of everyday language use. At the far right of this continuum — the projective "I" of trance and similar states — the context of language use must be highly marked and salient with respect to everyday contexts, if the "I" is to be interpretable as referring to a nonordinary self. This is the norm in trance and possession behavior. If the speech situation is not marked, there is no signal to differentiate the different Ts," and the speaker is susceptible to being labeled, in the Western cultural tradition at least, mentally unstable, stricken with a multiple personality disorder. From a discourse point of view, it is important that in the limiting case of projective "I" there is no script or story that is accessible to the audience apart from its manifestation in the present speech. This is the key distinction from theatrical "I," where the self is usually part of a well-known story. The projective "I" tells an emergent story, and in this regard is distinguished only with difficulty, and by virtue of a highly marked context, including non-linguistic behaviors, from the ordinary indexical referential "I." It is as if the far right of the continuum of Fig. 1 were connected back up with the left hand side. With respect specifically to the Shokleng case, there are two types of nonindexical referential usage that are salient. One is the de-quotative type used in the

The"!" of Discourse 43

>8 I •s•5s*3=9s 3 ? " §2B a a 3 0 > 2 μ c e•υ 3M &M ο

•3 e Π

I •8

oΜ sS > U > a9 .ο. sο88Uh,®a ι i m s i imiS esi o Biolslt s1

υ > -β I

1

1

1 I Β 1

£28 i 8J _ •S Β " 2 9 ©2 3 «Α U CS Μ

2» is. iΚ* nο 73 3 -a 3 Ε S ·3 δ βΉ .ö8 ο1.8 a Uä Μ

8l •ο5 Sa υ .s r

44 Greg Urban telling of myths (other than the origin myth) and in other narratives. The other is somewhere between theatrical and fully projective "I," involving the speaker in the virtual assumption of the self of an original narrator of the story. It is not fully projective because there is a fixed, memorized text However, unlike the usual theatrical "I," the "I" of origin myth discourse involves the speaker/narrator as well as the characters quoted in the narration. It is perhaps most akin to those stories in the Western tradition that involvefirstperson narrative reminiscence. Regarding ordinary de-quotative "I" in Shokleng, the principal cultural function would seem to be reality enhancement for the audience. De-quotative "I" usage tends to occur when the speaker/narrator is finely attuned to audience response. This involves an elementary type of role playing through the use of indexical cues (voice quality, intonation, etc.), which can also be present in greater or lesser measure in anaphoric "I" generally. These indexical cues, accompanied by the dropping of the quote framing clause, make it seem as if the speech were occurring in the present situation of telling. The speech situation is brought into the here and now. The audience is made to feel as if it is in the presence of a non-present dialogue. The speech portraying a reality referentially comes to (purport to) resemble that reality iconically. The more iconic the sign — the less purely referential — the more it assumes an air of reality for the audience. De-quotative "Γ verges on theatrical "I." With the origin myth, the situation is distinct in one important regard. In this case, in contrast with ordinary narrative, the more involved with the audience the speaker/narrator becomes, the greater will be the tendency to switch from first person to third person. However, this shifting occurs not within the quoted clause, i.e., direct quotation becoming indirect quotation, but rather in the main quoting and narrative clauses. When the origin myth is actually being performed, the tendency is for the speaker/narrator to use first person in this clause. When the myth is being told, the tendency is to use the third person. "Performance," however, in this case is quite unusual and is definitely untheatrical, in the sense of theater in which an actor gesturally plays a part The characteristics of the origin myth performance (wäflklön) have been discussed elsewhere (Urban 1985b, 1986a,b) and can only be summarized here. A performance involves two speaker/narrators who sit facing one another. The performer shouts the first syllable of the myth and this is echoed by the second performer. The first then shouts the second syllable, which is again echoed by the second, and the two go back and forth in this way in rapid-fire exchange through the entire myth. The effect on the audience is achieved as much by the acoustics

The "I" of Discourse 45 and body movements as by the semantic content of the story itself. The actual narrative line is understandable, in greater or lesser degree, by members of the audience, but it is the sound effect that is especially stunning. Unlike theater, therefore, gestural performance in this case is disconnected from the story line. The actual rhythms of the syllable exchange are accompanied by a rocking back and forth of the two speakers. The entire effect is hypnotic, especially for the actual participants. Being involved in a performance of wSflkltn means turning away from any audience and concentrating intently on the immediate interaction, with its emphasis on the precise repetition of a myth syllable by syllable. The mesmerizing effect of the rhythm, coupled with this intense inward focusing and shutting out of the audience, produces a discourse "I" that it is, consequently, closely related to the projective "I" of trance. The argument has been made elsewhere (Urban 1985b, 1986a,b) that, from a semiotic point of view, the accurate re-performance of the origin myth simultaneously symbolizes and is the transmission of culture. However, in this previous work it was not possible to link the phenomenon of transmission to the subjective experience of the speaker/narrator. The present work on discourse "I" makes such a linkage possible. In fact, the projective "I" of origin myth telling is an "I" that aligns speaker/narrators with past "I's," and, in particular, with the original (mythical) narrator. In the period of the actual performance, therefore, the speaker/narrator assumes the "self' of the original ancestor. He comes as close as is possible, through this projective means, to assuming an identity as "I" of an ancestral figure, and of thus subjectively embodying the continuity of culture. It can be seen from this analysis why the simple narration of the origin myth — where it is not dyadically performed — involves a switch to the third person in the main clause. The speaker/narrator is more attuned to audience response and is, consequently, less inwardly focused. The subjective experience of the assumption of another "I" is less important than the assumption of an interactive role as narrator, with its attendant emphasis on reality enhancement for the audience. The narrator is the real here and now speaker of the utterance, and the quoted voices of the narrative are the ones that are brought into the present through de-quotation. Audience focus is thus particularly relevant to the anaphoric and de-quotative portions of the continuum. In the theatrical to projective range, the subjective experience of the speaker/narrator becomes progressively more central. As one moves closer to the trance-like "I," the function of the discourse "I" shifts to facilitating the assumption of a different self by the speaker.

46 Greg Urban It is not clear that the projective "I" will be associated in all cultures with the assumption of an ancestral-historical self, as among the Shokleng. This is something that requires further comparative research. The straightforward emphasis on cultural transmission is probably a function which, while not unique to Shokleng, is nevertheless far from universal. Still, one can predict at a more general level a connection between the right half of the continuum represented in Fig. 1 and sociability, construed as the internal regulation of one's discourse by others. Such constraint is already present in the anaphoric "I," as was argued earlier. Here the constraint is a communicative one, however, not a subjective one. The speaker/ narrator signals the representation of the words of others to an audience, and is bound by audience recognition of the representational quality of his or her assumption of the words of others to whatever standard of faithfulness they hold a speaker. However, there is no question here of an internal experience of the other. In the case of projective "I," emphasis is shifted in the direction of a subjective experience of otherness. The speaker/narrator must attempt in greater or lesser measure to appear to be another. The appearance becomes more convincing outwardly in proportion as it is more convincing inwardly. Sociability becomes the subjective experience of another point of view.

Referential and Iconic Otherness The behavioral, including discourse, patterns of others may be assumed without the behavior or discourse simultaneously representing them referentially as assumed. This kind of assumption or imitation I will call "iconic otherness" of the self. This is the basic stuff of culture — the participation of individuals in socially transmitted patterns of action and representation of the world, which are adopted "unconsciously" and without reflection. At the same time, the adoption of such a pattern is a sign, in particular, an icon of the adopted pattern. Insofar as the imitation is faithful, the behavior, including speech, of the imitating actor is a sign vehicle capable of being read by others. It points to the conformity of the actor to the cultural patterns. It is thus in some measure meta-cultural, simultaneously as it is cultural. But it is meta-cultural in a peculiar sense, viz., in that the similarity or dissimilarity of the behavior to its imitated counterpart need not be taken as such. There is nothing intrinsic to it that requires that it be taken as meta-cultural.

The "I" of Discourse 47 In the case of direct quotation, the very sign itself encodes referentially its own status as meta-cultural, necessitating that it be interpreted as representing other speech. The "I" of discourse is explicitly referred to as the "I" of another. Insofar as the quoted portion of the utterance is an icon of the utterance it purports to represent, speakers are also engaged in "iconic otherness," but it is an otherness from which they are distanced by the quote-framing clause. That distance is progressively shortened as one moves to the right in Fig. 1, but it is never obliterated, except possibly at the extreme right, where projective "I" merges with indexical referential "I." In other words, the discourse and/or the context in which it is embedded explicitly signals that the "I," and, indeed, the entire imitation, is to be understood as such. This type of otherness, in which the "I" is explicitly signalled as other than the everyday indexical referential "I," I will call "referential otherness." From this perspective, it is extremely interesting that different points of the continuum in Fig. 1 are associated with differing degrees of spontaneity in speech. While any speech can involve iconic otherness, de-quotation and theatrical "I" especially tend to be associated with more or less fixed texts, such as myths and traditional stories. If the Shokleng case and Western theater are indicative, it would appear that, where theatrical and projective "I" are involved, the texts display extreme rigidity. The speaker/narrator/actor must with considerable precision replicate earlier tokens of a fixed discourse type. Hence, the entire discourse itself is an instance of genuine iconic otherness, simultaneously as it contains instances of referential otherness. Fig. 2 is an attempt to represent this condition of double or two-tiered otherness of the self. The instances of narrating/performing (A), insofar as they are tokens of a fixed discourse type, involve iconic otherness. In the recitation, the main plus subordinate clauses are icons of the main and subordinate clauses of previous narrations/performances (B). At the same time, this iconic otherness contains within itself instances of referential otherness, in which the subordinate clause of a direct quotation is represented as iconic with some previous (imagined or real) main clause (C). Where de-quotative and theatrical "I" are involved, the actual grammatical trace of the representation of quoting disappears, and the context carries the signal of referential otherness. The discourse thus operates simultaneously at two levels. At the upper level, it signals itself as referential otherness, and thus as not to be mistaken for that otherness. The "I" of discourse is not the everyday indexical referential "I," and the discourse does not actually involve an embodied iconic otherness. At the lower level, however, insofar as it replicates a fixed type, the discourse does in fact

48 Greg Urban involve an embodied iconic otherness. It is another telling or performance of a fixed type. It thus really is an imitative speech behavior on the part of the narrator/performer.

Instances of speech reported: [referential otherness]

Instances of narrating/performing: main + subordinate [iconic otherness]

C main

main + subordinate

Fig. 2: Two-tiered Otherness

The interpretation of this phenomenon proposed here is that the upper level is designed to achieve a kind of metapragmatic awareness of the lower level, simultaneously as it allows that lower level to be a fully embodied replication. In other words, the discourse becomes a sign vehicle capable of bringing about an awareness of the social character of indexical referential "I." The speaker/narrator is represented as assuming a non-everyday "I," but an "I" simultaneously that is identical to one that other speaker/narrators before him have assumed. With ordinary indexical referential "I," there may be an embodied iconic otherness, with the discourse of the "I" an imitation of the discourse of others. However, that embodied iconicity is not understood as such. The "I" of de-quotation and theater is capable of being represented as an assumed "I," but it is also one that is preeminently social, i.e., it is also assumed by others. De-quotative and theatrical T s " thus become levers for bringing about an awareness of a genuinely social "I." If the present interpretation is correct, de-quotative and theatrical "I" should go hand in hand with fixity of the underlying text. Fixity ensures the genuine social or shared character of the "I." Explicit signalling of the Τ as the "I" of another ensures an awareness of the assumed character of the "I," but simultaneously does not interfere with its occurrence as actual embodied iconic otherness.

Conclusions The indexical referential analysis of "I" originally proposed by Benveniste, and later taken up in theoretical-philosophical accounts of the relationship between self,

T h e ! " of Discourse 49 language, and culture by Singer and others, is an essentially correct one. However, if the present arguments obtain, the analysis is incomplete. The personal pronouns, when studied within their discourse contexts, are considerably more complex in terms of semiotic functioning than the indexical referential analysis alone suggests. Simultaneously, these added complexities necessitate an amplified conception of the relationship between self, language, and culture. Specifically, I have proposed that there are functionally distinct kinds of "I," with indexical referential "I" representing merely the end point of a continuum. The fundamental distinction is between an "I" pointing to an everyday self and an "I" pointing to an imaginary or assumed self. In the latter case, the individual speakers to whom the "I" points are in fact anaphoric substitutes for characters in a narrative text. As one moves along the continuum, the individual speakers come more and more to resemble (outwardly and inwardly) the imaginary self for whom they act as discourse substitutes, taking on the indexical cues of that other self. There are thus in discourse always two (or more) kinds of "I" — one pointing to an everyday self, one pointing to an assumed or "other" self. In the Benveniste model, it is the first person pronoun, by virtue of its participation simultaneously in indexical reference and in the Saussurean distributions of language, that makes possible the linkage between self and culture. In the present model, there are two levels. At one level, the anaphoric self that is a substitute for a discourse character allows an individual to fit into a culture-specific text. At another level, that discourse or textual self functions as a blueprint for the everyday self. In other words, texts or specific instances of discourse, which contain the imaginary "I" as well as the everyday "I," make possible a truly cultural self. Indexical referential "I" alone is essentially unsocialized. Discourse or context, however, signals that an anaphoric "I" is imaginary, that it points to an other than "I." Speaker and audience alike grasp that the "I" of anaphora, de-quotation, theater, and projection is to be understood as distinct from everyday "I." At the same time, by virtue of fixing the text, as in the wäflkldn or in Western theater, even the narrative "I" becomes a genuinely cultural one. The same point of view, so to speak, can be taken up by different speaker/actors. This is the essence of the cultural self— a point of view on the world that is shared and socially transmitted. The role of metapragmatic awareness of the two T s " deserves further investigation. It may be proposed that this awareness makes possible individual manipulation of the two Ts." With the ability to grasp the two T s " and to select the specific anaphoric substitutions, the individual is not entirely subject to the reign of culture through received texts. Furthermore, new texts can be created, old

50 Greg Urban ones modified. Metapragmatic awareness is the motor of cultural change in the discourse constitution of self. At the same time, metapragmatic awareness of the two "I's" may be linked to the depth psychological phenomena studied by psychoanalysts — to the interplay between superego and id, in the classic framework, or between self and ideal, in more recent terminology. While the two "I's" make possible the cultural constitution of self, and while metapragmatic awareness of the distinction liberates the individual in some measure from the tyranny of culture, the semiotic capacity to grasp the two 'Ts" also opens a dialogue within the individual over everyday "I" and the "I" of discourse. It creates the ground for apprehension of a possible discrepancy, and consequently for representable internal affective processes that might otherwise never exist. The present paper has attempted to demonstrate the complexity of "I," when viewed from within its discourse context. It is discourse — specific instances of speaking and the types to which they are related — that is the fulcrum between self and culture, between individual and society. If the indexical referential character of the first person pronoun makes possible discrete reference to different speakers by means of the same form, it is nevertheless the relationship between indexical referential and anaphoric "I" that supplies the ground for sociability.

References Benveniste, Emile (1971a [1956]). The nature of pronouns. In Problems in General Linguistics, M.E. Meek (trans.), 217-222. Miami: University of Miami Press. (1971b [1958]). Subjectivity in language. In Problems in General Linguistics, M.E. Meek (trans.), 223-230. Miami: University of Miami Press. Graham, Laura R. (1983). Performance Dynamics and Social Dimensions in Xavante Narrative: Hoimana'u'ö wasu'u. M.A. Thesis. Department of Anthropology, University of Texas at Austin. Gross, Joan (1983). Creative use of language in a Lifege puppet theater. Semiotica 47 (1/4): 281-315. Kohut, Heinz (1985). Self Psychology and the Humanities, C. Strozier (ed.). New York: W.W. Norton.

The "Γ of Discourse 51 Ricoeur, Paul (1974). The question of the subject. In The Conflict of Interpretations: Essays in Hermeneutics, D. Ihde (ed.). Evanston, IL: Northwestern University Press. Silverstein, Michael (1976). Shifters, linguistic categories, and cultural description. In Meaning in Anthropology, K. Basso and H. Selby (eds.), 11-55. Albuquerque: University of New Mexico Press. Singer, Milton (1984). Man's Glassy Essence: Explorations in Semiotic Anthropology. Bloomington: Indiana University Press. Urban, Greg (1985a). Ergativity and accusativity in Shokleng (GS). International Journal of American Linguistics 51(2): 164-187. (1985b). The semiotics of two speech styles in Shokleng. In Semiotic Mediation, E. Mertz and R. Parmentier (eds.), 311-329. New York: Academic Press. (1986a). Ceremonial dialogue in South America. American Anthropologist 88(2): 371-386. (1986b). Semiotic functions of macro-parallelism in the Shokleng origin myth. In Native South American Discourse, J. Sherzer and G. Urban (eds.), 15-57. Berlin: Mouton de Gruyter. Wolf, Ernest S. (1986). Basic concepts of self psychology. Working Papers and Proceedings of the Center for Psychosocial Studies, No. 6. Chicago, IL.

The Self in Northern Cheyenne Language and Culture Terry Straus

Singer (1984: 85-87), following Peirce, described "tuism," in the context of developing his dialogical view of the self, as a philosophical understanding in which primary emphasis is placed on the second person, or "tu," rather than the first person, or ego. The ego or self is not seen as given, prior to its engagement with others, but rather as precipitating or solidifying out of interactions with others who are there from the start. This is a view in many ways compatible with that developed by Bakhtin (1981,1984) in various of his writings. My purpose in the present paper is not to investigate tuism at this philosophical plane, as applying to an understanding of the self in all cultures at all times. Rather, I wish to suggest that certain cultures and certain languages may be more mistical than others, and, in particular, I wish to marshall evidence in support of the view that Northern Cheyenne culture is tuistical, or, at least, that it is relatively more tuistical than the ego-focused cultures that have grown up in Europe. I will do this by discussing the role of pronouns, and especially second person pronouns, within Northern Cheyenne language, and by relating this to evidence from Northern Cheyenne culture more generally.1 The Northern Cheyenne people live today on a small reservation in southeastern Montana. They call themselves Tsetsihestahase, "those who share the same heart," and distinguish themselves from Hevetaneo?o, their Southern Cheyenne relatives. Tsetsehestahase define themselves as bound together by common substance, common land, and by the teachings associated with the Sacred Arrows (kept today by the Southern Cheyenne in Oklahoma) and the Sacred Hat (kept today by Elmer Fighting Bear in Montana). The tribal bond establishes a mutual sensitivity and interdependence among tribal members which is fundamental to the meaning and motivation of their behavior. Individual behavior is invariably interpreted as dependent upon the compelling nature of personal relationships. This acknowledged interdependence is the context for tuism in Northern Cheyenne culture. The web of interdependent "persons" is not to be identified with an objective group of living individuals. The Northern Cheyenne people live today, as they always have, in a web of social relations of "persons," only some of whom are

54 Terry Straus living human beings (cf. Hallowell 1976). Deceased Cheyenne and other spirit beings also participate in the tribal community. Within the community, the individual is first identified by who his past and present relatives are: his social identity, and ultimately his self, derives from his connection to others within the tribe. The web of "persons" thus refers to culturally-defined entities, and it is through interaction with these persons that the Northern Cheyenne self is constituted. In order to understand the linguistic basis for the cultural construction of self, personal pronouns must be considered. Northern Cheyenne is an Algonquian language. It differs slightly from Southern Cheyenne in pronunciation and in certain vocabulary items. The verb is the core of every Algonquian sentence. A sentence often consists of a single, highly inflected verb, and each verb must be modified by a personal pronoun. In contrast to English, Algonquian languages are "pronominal argument languages" (Jelinek 1985). In Algonquian languages, a pronoun must be specified in every utterance. In English, it is typical for a sentence, such as "the man went," to have a free standing noun phrase and a verb, with no pronominal cross-referencing. In Algonquian sentences, however, the requisite structure consists of a verb plus pronoun, with the full noun phrase optional, hence: "he went (man)" or "(man) he went." Here the pronoun must always occur and cannot be replaced by its nominal counterpart. The mandatory incorporation of the personal pronoun in the verb has the effect of grounding the sentence in its dialogic context. The obligatory pronoun "he," while cross-referencing a third person form, points to the dialogically prior first and second person; "he" refers to a particular other specified by the dialogue but not involved in it. Whereas English allows for abstraction and distance from the dialogic context, since there is no obligatory pronominal cross-referencing of noun phrases, Cheyenne and other Algonquian languages require continual reference to iL This continual reference forms the linguistic basis for Northern Cheyenne dialogicality. While the verb in Algonquian sentences may be inflected for mode, tense, aspect, and other features, it must also be inflected for gender. Gender is expressed in verbal forms, nominative plurals, and demonstrative pronouns. Northern Cheyenne divides the behavioral environment into two genders commonly called "animate" and "inanimate." The animate is the unmarked gender in that, when in doubt, a speaker will use the animate form. This is relevant to the general tendency of the language to elevate entities to the status of potential "persons," and hence potential interlocutors, participants in the dialogical community.

The Self in Northern Cheyenne Language and Culture 55 Inanimate nouns label objects generally understood as non-living. However, the meaning associated with this category in Cheyenne and in Cheyenne-English is not the same as the meaning assumed in Standard American English. Inanimate objects are immobile, unresponsive, non-reproductive, and neither inherently dangerous nor helpful. They are things as opposed to beings: they exist but are not actively connected, not part of the interdependent, dynamic community. Animate nouns label various classes of living or powerful objects: trees, large animals, small animals, birds, humans, spirit beings, and certain natural phenomena. But it is as a specific item, not as a member of a class, that a particular object is assessed as animate (see also Brightman 1983). The two genders are neither closed nor set. Certain nouns have both animate and inanimate forms. An example of such a noun is mo'eiko, which means "ring" in the inanimate form (plural, mo'eikonötse) and "finger" in the animate form (plural, mo'eSkono). Furthermore, nouns which do not have this double form, and are habitually either animate or inanimate, may shift in gender according to the context of their use. Finally, speaker variation in gender assignation is clear: a particular place or a particular object which has special meaning to one speaker may be assigned to the animate gender by that speaker, while the same item may be assessed as inanimate by another speaker with different experiences. Hence, the assignment of an object to animate status, and so also to potential participation in the community of persons, is at least in part a function of the dialogic context in which speaker and object are mutually constituted. In this regard, the speaker's role is significant in another way as well. Whatever the habitual status of a particular noun, if a specific item in that class of objects is addressed by a Cheyenne speaker, it necessarily becomes animate. By dint of being addressed, even the inanimate object becomes a person, another self with which the speaker himself is dialogically engaged. An individual Cheyenne might converse with spirit beings, with deceased human beings, with a particular animal which serves as his guardian, or with a sacred object or place which embodies a spirit. To an outside observer, the dialogue in these instances may seem to be a monologue; the speaking seems to emanate entirely from the human person. But to a participant in this community of persons, the addressee in such instances understands the speaker and responds with other kinds of signs which are interpretable by the speaker. The addressee is thus affirmed as second person, "you." Hence, the dialogue is an exchange of signs between mutually constituted selves and it is the relationship between these selves, not the physical form, that is significant.

56 Terry Straus Various Cheyenne traditions and beliefs support the linguistic evidence for the critical role of dialogue in the creation of the self. Speech is defined as necessary and sufficient for personhood, and only man among tseametanenesso ("those with physical life") retains the generic capacity for verbal speech, although dialogue through other kinds of signs, as described above, is possible with a range of entities. In the distant past, animals and birds also spoke, but the Great Race changed the relationship among living things: it confirmed the generic way of each species; it gave man the right to kill and eat buffalo; and it established man as capable of speech, and thus of prayer and blessing, of seeking and acquiring power. In the recent past, some men have been able to understand certain other modes of communication as well, receiving warnings and guidance, for example, from coyotes. Certain specially powerful animals, also, can still speak and understand Cheyenne; but, as a class, each species has its own way of communicating, and that way is not fully comprehensible to non-members. Speech and understanding come to man through omotome, which means "breath," "word," and also "consciousness." Through it, man both separates from and connects with those around him: it gives him individuality unknown in other species (Straus 1976). Simultaneously, it enables him to participate in dialogue. Cheyenne cultural traditions emphasize the role of addressee, of the person spoken to in a dialogue. Children are symbolically "opened" to teaching in the earpiercing ceremony. Listening and understanding connect the child with his tribal traditions and establish him as a person. It is as addressee, second person, that the child's identity is affirmed: opening his ears to that possibility is the symbolic intent of the ear-piercing ceremony. Those who ignore tribal teachings are commonly described as having "hard" or "closed" ears. There is even a tradition of a monster with no heart which chops off and collects in a chain he wears around his neck the ears of human beings. In so doing, he renders them deaf to their tribal traditions and their status as persons within the tribe. When the failure to listen results in criminal behavior, it was and sometimes still is punished by exile and/or shunning. Without communication, the personhood of the criminal is denied. It can be restored only by the effort of a relative of the one punished, by re-educating and by re-integrating the individual into the tribal community. A Cheyenne, by speaking to another, creates that other as a person, simultaneously as he is himself created. First and second person in any dialogue are necessarily animate in gender. Inanimate gender occurs only in reference to third persons. Gender assignation is best viewed as an index of the status of a particular item in a hierarchy of significance to the speaker (Brightman 1983).

The Self in Northern Cheyenne Language and Culture 57 Animate gender implies "respect" on the part of the speaker and activates certain behavioral norms. Disrespect or deprecation may be indexed by the use of inanimate forms for habitually animate items. Hence the question of relative animacy, and so also of personhood, is linked inextricably to dialogue. Consistent with these observations, the Northern Cheyenne pronominal system distinguishes first and second person from third persons and specifies a person-order or "person hierarchy" (Lehman 1979) within the animate class. As discussed earlier, Cheyenne pronouns do not stand alone: they occur only as verbal affixes. For intransitive verbs, the pronominal system in Cheyenne is: lnä2 n63 6-

11 nä-.. . 12 ηέ-.. . 22 n6-. . . 33 6- . . .

-me ("exclusive") -ma ("inclusive") -me -o'o

For example, with the verb stem meaning "to eat" (-mesehe-), the paradigm is: nämesehe ndm&sehe 6m&sehe

nämSseheme n£mesehema n6m£seheme dmfeseheo'o

The basic forms are found in the verbal prefix: ηά-, ηέ-, έ-. These are bound forms which can appear only as the prefix of a verb and never separate from it. While it is possible to give linguistic evidence for the distinctiveness of first and second persons with respect to third person, the saliency of this distinction in Northern Cheyenne awareness is emphasized here. For example, the first and second person pronouns are related in segmental sound shape to the possessive pronouns which modify nouns. One linguist (Lehman 1979) has argued that the basic pronouns differ from the possessive forms in a having relatively higher pitch on their vowels. However, this does not concur with native assessment: native speakers group the two sets of forms together, recognizing them as the same. The third person possessive form, he-, is recognized by those same speakers as different from the personal pronoun έ-. So native speakers see first and second as alike in having the same basic and possessive forms, the third person as distinct from these in having a basic form that is distinct from the possessive form.

58 Terry Straus There is other, and even more dramatic evidence of native-speaker recognition of the distinctiveness of first and second person with respect to third person. Asked to consider and comment on personal pronouns in his language, one consultant developed a 45 minute tape in which he mentioned only first and second persons. In a written commentary, he noted, "I think that in the Cheyenne language there are no personal pronouns (i.e., a system specifying "I," "you," "and "he/she/it" such as he had been asked to address), but that there are pronouns that refer to the person talk(ing) or to the person or persons being addressed. Like na- would be that "I" referring to myself or me as a person, ne- would be the person or persons beings addressed as in yourself, you people, you guys." In native awareness, therefore, there is a sharp break between these local pronouns, which refer to the participants in the dialogic speech situation, and the non-local third person forms. This goes along with the general cultural emphasis on the role of dialogue in the mutual constitution of selves within a community of interacting persons. And there is a sharp break between this community, symbolized in language by means of the first and second person pronouns, and everything else, symbolized in language by the third person form. The third person form in Algonquian languages is often translated "he/she/it," as in imesehe, "he/she/it ate." This translation, however, imposes a specificity, an immediacy, not there in the Cheyenne. While first and second persons are immediate, specific, indexical, the third person is vague and distant by comparison. There is no sexual identity at all associated with the third person in Cheyenne, so it is misleading to render it as a composite of possible sexual identities (he/she/it). The third person form is better translated as "someone" or "something." Again, this distinction is one that is salient in native speaker consciousness. One consultant suggests translating the third person as "a person" or "a thing," assiduously correcting all translations in the form "he/she/it...." The third person in Cheyenne is defined by reference to the dialogic context, but is itself outside of the dialogue. While we have evidence here of the distinctiveness of the dialogic participants with respect to non-participants, however, we still do not have here evidence of "tuism" per se, that is, of the relative primacy of the second person within the dialogue. Evidence in this regard can be found in the "person hierarchy" in transitive sentences. Other evidence for the primacy of "tu" is found in the plural forms themselves. Final suffixes are used to pluralize the basic pronominal forms. There are two first person plural forms in Cheyenne, exclusive and inclusive. The exclusive

The Self in Northern Cheyenne Language and Culture 59 form signifies "us, but not you", or, in Cheyenne-English, "us guys": the second person is specifically excluded. The inclusive signifies "you and me (and perhaps others)." In each case, of course, the speaker is included. In the first case, the plural form is constructed from the basic first person prefix, ηά-, Interestingly, however, where the second person is included as well as the first, it is the second person pronoun, ne·, which appears in the basic form in the prefix. In the absence of second person, the first person is used; but where the second person is included it takes precedence over the first person. The primary linguistic evidence for the primacy of "tu" is to be found in the phenomenon of transitivity. In Cheyenne, both the subject and the object of a transitive verb are always coded by means of pronominal affixes. However, the choice of pronominal form is not determined solely by the referent of the form. Rather, it is dependent on the relative positions of the subject and object within a hierarchy of animacy. Figure 1, from Alford (1976), gives the pronominal affixes used with transitive animate verbs in Cheyenne. The chart reveals that there is a set order among persons, which reflects a "person hierarchy" (Lehman 1979). The hierarchy is reflected in which of the two participants in a transitive sentence — subject or object — is indicated by the prefix and which by the suffix. For example, if a first person subject is acting on a third person object, it is the subject (ηά- = "I") that appears as the prefix, as in nävoomo ("1 sees 3"), and the object (-ο = 3rd person) that appears as the suffix. However, if it is a third person acting on a first person, it is not the subject but rather the object that appears as prefix, as in ndvooma ("3 sees 1"). In other words, first person always preempts third person as the prefix, regardless of which is subject and which object. A careful inspection of the chart will reveal that there is an order in this regard. Whenever 1st or 2nd person has 3rd person as goal, the 1st or 2nd person pronoun (πα-, ne-) appears as the prefix of the verb. When the 3rd or 4th person has 1st or 2nd person as goal, the local index — 1st or 2nd person pronoun — still forms the verbal prefix. The goal and direction of action are marked in the suffix. Thus, whatever their participant role in a sentence — subject or object — the 1st and 2nd person pronouns retain the prefixal position, and the more distant 3rd person adjusts to the priority of the local forms. Among "persons," then, first and second person may be said to be "higher" with respect to third person in the hierarchy. Insofar as language is a symbol of culture, therefore, this again confirms the dominance of a dialogical model, in which the participants in a conversation have take precedence over the non-participants.

?

έ

ύ ν

ts

co

^ "-1

Aloneness-disorders

The square brackets containing "disorder" is intended not only to convey its "always-already-there" aspect, but also its contained or hidden aspect. The distinction is analogous to our current understanding of ARC and AIDS. The hyphen bonding is intended to denote a more integral union. The NTV case would be represented thus, where the subscript "t" stands for the general Tamil pattern: Alonenesst

> Aloneness-disorder,,

Typically, and in general for both Tamils and Sinhalese, "aloneness" provides the ideal and inviting precondition under which an individual will be possessed by a wandering spirit and thereby become a victim of "aloneness disorder." Such a precondition of "aloneness" is conceptualized in physical terms. For instance, a young woman is vulnerably "alone" thus, if she is found in at the wrong time, in the wrong state, in the wrong place. To list but a few, the wrong time could be during a particularly malefic planetary configuration; the wrong state, during

78 Ε. Valentine Daniel menstruation or pregnancy; and the wrong place, when facing an inauspicious cardinal direction, at the cross roads through which a corpse had recently been carried, or where two streams meet. Paradoxically, "aloneness" is not being alone, in the strict sense. It is being disconnected from other human beings with whom one ought to be connected. The corollary of such disconnectedness is finding oneself in the company of undesirable entities, abnormal persons and powers. The social bonds and one's bondedness with the social prevents the intrusion into society and the socialized self of such "alien" persons and powers. The symptoms in question may include, among others, madness (pissu), the running of a high fever (una), or recurrent miscarriages and still births (baf/a gässvlma and lamäy nätivelä hambuvlma). The same conjunctions of time, state and place that result in these symptoms in the case of the Sinhalese may also result in similar symptoms among the Tamils. But there are significant differences. In order to appreciate these differences between the Sinhala and non-Sinhala cases we need to examine the symptoms in their fuller semeiotic context.

The Variations Semeiotically

Conceived

Let us begin with one of Peirce's most general and often quoted definitions of a sign: A Sign, or Representamen, is a First which stands in such a genuine triadic relation to a Second, called its Object, as to be capable of determining a Third, called its Interpretant, to assume the same triadic relation to its object in which it stands itself to the same Object (CP 2.274). The "First," "Second," and "Third," in this definition, is meant to correspond to Peirce's Phenomenological or Cenopythagorean 4 categories of Firstness, Secondness and Thirdness, categories we shall find reason to expand upon further on in this essay. Eight years later, in 1910, in what is one of his clearest explications, he defines the sign as follows: By a Sign I mean anything whatever, real or factile, which is capable of a sensible form, is applicable to something other than

The Semeiosis of Suicide in Sri Lanka 79 itself, that is already known, and that is capable of being so interpreted in another sign which I call its Interpretant as to communicate something that may not have been previously known about its Object. There is thus a triadic relation between any Sign, and Object, and an Interpretant (MS 654.7,1910; cited in R. Parmentier 1985: 26). And in a third definition, he brings out the infinitely open process of semeiosis within which the sign is to be found: [A sign is] anything which determines something else (its interpretant) to refer to an object to which itself refers (its object) in the same way, the interpretant becoming in tum a sign and so on ad infinitum. (CP 2.203) All three definitions5 may be diagrammed as in Fig. 3. The lines opening outward from {R}, {O}, and {I}, in this diagram, are intended to indicate the openness of semeiosis. Also, the small case 'r's, 'o's and 'i's indicate that any single correlate in a given triad which functions in one way with respect to the other two correlates in

OBJECT {O} i

{1} Ο

r

{R} Γ

Figure 3: General Structure of the Sign

80 Ε. Valentine Daniel that same triad — say as an 'object' — may, with respect to a different triad function differently — say, as a "representamen" or as an "interpretant." If we were to attempt to understand "aloneness" in terms of the general structure of the sign, we shall arrive at figures 4, and S, representing the general Sinhala pattern and the general (non-NTV) Tamil pattern respectively. The outward branching lines of Figure 3 have been omitted in these and in Figure 6 in the interest of simplicity but not to deny the openness of the semeiotic process. "Aloneness [disorder]" {0}

"Aloneness-disorder" {1}

Symptoms [R] 1. Miscarriages 2. Stillbirths 3. Fevers 4. Madness, etc. 5. "aloneness"

Figure 4: General Sinhala Pattern

Illness-type {O}

"Aloneness" {O}

•

"Aloneness" {1}

Illness-token {R} (Symptom) 1. Miscarriages 2. Stillbirths 3. Fevers 4. Madness, etc.

1. Miscarriages 2. Stillbirths 3. Fevers 4. Madness, etc.

Illness-type {1} 1. Miscarriages 2. Stillbirths 3. Fevers 4. Madness, etc.

Figure 5: General (non-NTV) Tamil Pattern

The Semeiosis of Suicide in Sri Lanka 81 What these figures, especially Figure 5, do not make clear is the distinction between what seems to be identical lists that appear under Illness-type {0} and Illness-type {I}. The difference is significant For instance, a miscarriage as a semeiotic object describes a cultural (or natural, for those who insist) fact, to be discovered or understood, and, thereby, to be interpreted. Miscarriage as a semeiotic interpietant describes the actual or potential effect, the action taken or to be taken and the disposition arrived at, in light of the revelation of the semeiotic object. In both Tamil and Sinhala constructions, the objective precondition for subsequent interpretation is the state of "aloneness" — with the qualifications, for the Sinhala case, as discussed above. The symptoms, as representamens, mediate — function as means and media — "between that which "they are capable of revealing something about and what they are capable of revealing about it" (Ransdell 1986: 53). But here is the difference. In the case of the Tamils the mediating symptoms are not presumed to be as transparent of the object ("aloneness") they represent as these same symptoms are for the Sinhalese in their representation of "aloneness [disorders]. In other words, for the Tamils, symptoms such as miscarriages, fevers, madness, etc., may, on the one hand, index "aloneness," but on the other may be iconic tokens of types of disorders known as miscarriages, fevers, madness, etc., respectively. As indexes what is picked up by interpretive interests is not some shared quality inherent in the object and the sign but the fact of contiguity or coincidence of sign and object. To the extent that these signs are iconic, they are such because the object and sign are linked by a shared quality; they are tokens (or sinsigns) because they are actually occurring signs. Not only then are these symptoms capable of representing different objects, i.e., polysemously, but they also represent by means of different modes, i.e., polymodally. This polymodal and polysemous representation, while enhancing the representational possibilities of the signs and symptoms in question, also makes them sufficiently ambiguous so that no single significant object or mode dominates. At this point it is worth remembering, not only with respect to the coordinate conceptions of icon, index and symbol, but also with respect to the classes of interpretants and Peirce's phenomenological categories, both of which will be considered in some detail shortly, that these classes and categories: function in application only to isolate dimensions of the significance in things, ...not to classify or sort out things into

82 Ε. Valentine Daniel distinct groups in the way that, say, the conceptions of an orange and an apple and a pear can be used to sort a batch of mixed fruit into separate piles. Thus when we identify some sign as being iconic, for example, this only means that the iconicity of that sign happens to be of peculiar importance to us for some reason or other implicit in the situation and purpose of that analysis, but there is no implication to the effect that it is therefore nonsymbolic or non-indexical (Ransdell 1986: 56-57). From an ethnographical point of view then, the methodological issue is to identify the dominant mode of signification with respect to the interpretive interests and propensities of those with and among and about whom we study. And that is what we are engaged in in the current interpretive exploration among the Tamils and Sinhalese of Sri Lanka. For the NTV, symptoms that are associated with the "aloneness-disorder" have what I have characterized as a polythetic quality wherein more than one dimension of signification carries more or less the same freight of significance, unlike in the Sinhala and non-NTV Tamil cases, where one dimension or mode is foregrounded at the expense of the others. This built-in ambiguity or polythetic quality lessens the determining force of the object, "aloneness," allowing for the possibility of other determinants to play a part. For the Sinhalese, the symptoms in question present themselves as far more transparent media or means by which the object "aloneness" may be perceived. The result is that despite the variability of the symptom, it is deemed to be a sign, so comprehensively adequate in its revelatory power that what it reveals to the Sinhalese about its object, aloneness [disorders], appears to be almost all there is to be revealed. Such a revelation, which brings into identity what the object is and what the sign could reveal about it, is the total proper interpretant of the sign (Ransdell 1986: 53). For the Tamils, the symptoms in question present themselves as possible signs of "aloneness," but also as possible signs [iconic sinsigns] of other objects. The Sinhalese see these symptoms as "transparencies" through which an identity is established between something considered in itself (the object) and something considered under the aspect of being known (the interpretant). If viewed as a relationship between the symptom and the "dynamical interpretant" — the latter defined as the actual significant effect of the sign — the relationship in the Tamil case is suggestive whereas in the Sinhala case it is imperative. One of the psychological consequences of this difference is that "aloneness" looms larger in the

The Semeiosis of Suicide in Sri Lanka 83 consciousness of the Sinhalese than in that of the Tamils. To put the matter in yet another way, "aloneness" is far more immediately and immanently present in certain symptomatic afflictions among the Sinhalese than it is among and for the Tamils. These subtle but significant differences, their cultural groundings and implications, merit further study by medical and psychological anthropologists working in these two very closely related cultures and societies.

The Place of "Aloneness" in NTV Semeiosis The signification of "aloneness" among the NTV consists in the coming together of the Non-NTV Tamil Pattern and the Sinhala Pattern, giving us three triads of six correlates, as in Fig. 6: (O) "Aloneness"

Illness-types {O} (1 to 4)

Figure 6: The NTV Pattern Apart from the conjoining of the non-NTV Tamil and Sinhala-Patterns, there is also a modification in the Sinhala structure (indicated by the triad of lines going into the page, as it were). No longer is "aloneness-fdisorder]" — that latent disease — a semeiotic object. "Aloneness" serves as object to two interpretants, "aloneness" and "aloneness-disordern," and one representamen, symptoms 1 through 4. There is yet another difference. This difference returns us to the issue of the nascentness of tanimai töiam among the NTV. This difference can be best appreciated by analyzing "aloneness-disorder" as an interpretant, and contrasting this with the interpretants in the semeiotic constitution of "aloneness" and

84 Ε. Valentine Daniel "aloneness-disorder" among the Non-NTV Tamil and Sinhalese respectively. In order to do this, let us parenthetically and briefly review Peirce's classification of interpretants, remembering all the while that such a review is not intended to provide a classificatory scheme for sorting out entities into distinct groups, but rather to help us isolate dimensions of signification as interpreted in and experienced by different subcultural practices.

Peirce's Interpretants Peirce left us with several notes of caution regarding the risks of failing to understand the larger picture of semeiosis inherent in reducing the interpretant to and identifying it with the interpreter. He characterizes his own use of psychologistic terminology as a sop to Cerberus (Hardwick 1977: 81). In his attempt to rescue Peirce's broader, non-mentalistic, non-individualistic, theory of interpretants from the narrower psychologistic simulacrum, John Dewey (1946: 85) sees the latter as a veritable ersatz. My own use of Peirce's ideas may be read as an instance of throwing such caution to the winds. But this is not the way I intend it. To be fair, despite Peirce's own diatribe against psychologism he did frequently (and I feel, not only as a sop to Cerberus) employ psychologistic formulations. This is especially true, and in my opinion quite appropriate, with respect to discussions pertaining to the emotional and energetic interpretants, where the interpretive effects of the sign are to be located in the recipient of the sign's message. With the logical interpretant, the larger, purely logical and even cosmogonic process of semeiosis becomes both clearer and more appropriate. My caveat emptor may be stated as follows: the interpretant is not the same as the interpreter; it is a larger and an impersonal process. To quote T.L. Short's (1986: 98) general definition of Peirce's interpretant·. "[an] interpretant is not an interpreter. Instead, it is the particular thought, action, or feeling which interprets the sign. The formation of interpretants constitute an interpreter, which in some cases is a person." This is more or less the case with regards to "aloneness-disorder" among the NTV. Wojciech Kalaga (1986: 45-46) said it well when he wrote of the interpretant thus: "while capable of encompassing the concrete psychosomatic effect of the sign in the interpreter, the concept also refers to Thought construed in the non-mentalistic sense, to Thought broadly defined as a sign-relating process." I agree. With that caveat out of the

The Semeiosis of Suicide in Sri Lanka 85 way, let us move on to look at Peirce's types of interpretants more closely, as they pertain to our understanding of "alonenesst"aloneness-disorder,," and "alonenessdisordern." Interpretants are either Immediate, Dynamic, or Final. The immediate interpretant refers to the capacity of a sign to transmit a certain kind of information to the interpretant or interpreter. The kind of interpretant involved here is a potential interpretant. The dynamic interpretant, as already indicated, is the actual semeiotic effect of a sign. And "the final interpretant is the interpretant that would be the best interpretant of a given sign, given a goal of interpretation" (T.L. Short 1986: 107).6 Each one of these interpretants are either Emotional, Energetic, or Logical. The differences among these three can be understood most clearly in their dynamic or actualized aspects. The emotional interpretant is, in Savan's (1976:40) words, "the qualitative semeiotic effect" of a sign or, as Peirce describes it, "a mere quality of feeling" (cited in Shapiro 1983: 51). An energetic interpretant is an action that interprets the sign by engaging the outer physical world in muscular effort or the inner world in dialogic thought. It is constituted by an interpretive act in which some energy is expended. A logical interpretant, even when it is dynamic, i.e., is actually instantiated, is an expression of a habit of thought or action that is rule-governed (CP 5.476). Even though the intersection of the two trichotomies yields nine possible classes of interpretants, some pairs are more "at home together," so to speak, than others (See Figure 4). Thus the emotional interpretant is more "naturally" paired with immediate, the energetic with the dynamic, and the logical with the final/normal? These interpretants are represented in bold face in squares with heavy boundaries and lie along the major diagonal of Figure 7. These are the ones onto which I intend to map the ethnographic data of the aloneness-complex (if you would allow me this blanket term to include the varieties of "aloneness" and "alonenessdisorders" being analyzed). I do this partly to simplify and partly because these interpretants are ideal-typical representations of Peirce's phenomenological categories in the domain of interpretants. This second point will become clearer when we examine Peirce's categories in greater detail below. Suffice it to note, however, that the nine classes of interpretants constitute a hierarchy, with the Immediate/Emotional Interpretant presupposing nothing but itself on the one extreme, and, on the other, the Final/Normal Logical Interpretant presupposing and containing within it every other class of interpretant.

86 Ε. Valentine Daniel IMMEDIATE

DYNAMICAL

EMOTIONAL

IMMEDIATE EMOTIONAL

Dynamical Emotional

ENERGETIC

Immediate Energetic

LOGICAL

Immediate Logical

DYNAMICAL ENERGETIC Dynamical Logical

FINAL/NORMAL Final/Normal Emotional Final/Normal Energetic FINAL/NORMAL LOGICAL

Figure 7 Interpreting the Aloneness-Complex Among both the Tamils and the Sinhalese, "aloneness" and "aloneness-[disorder,]" respectively, and their symptomatic expressions, have become completely "naturalized." That is, they have gone beyond being mere dynamic interpretants, simple, singular, actual, events, and have become fully realized final logical interpretants. As logical interpretants they conform to a general rule, embodied in habits of thought and action (CP 5.476), and as final or normal interpretants they do "not consist in the way in which any mind does act but in the way every mind would act" (CP 8.315) in the culture in question. By contrast the phenomenon of "aloneness-disorder," which the New Tamils of Vavuniya associate with the increase in suicide among them, fails to conform to either of the above, broadly similar, constructions. It is to this interesting distinctiveness that I wish now to turn, and, to anticipate, I shall argue that "aloneness," in the experience of the New Tamils of Vavuniya, has not claimed for itself, in this subculture's consciousness, the status of a final or normal logical interpretant. In its "nascent" state it is to be located as oscillating between an immediate interpretant and a dynamic interpretant, between potentiality and actuality. And even when actualized as a dynamic interpretant, it does not fully evolve into a dynamic logical interpretant, i.e., into an interpretant that has entered into a stable relationship with a set of reasonable and meaningful regularities or familiar patterns but remains in the dynamic/energetic phase. Let us look at some ethnographic examples in greater detail. Among the NTV, the most conspicuous signs of "aloneness-disorder" is the withdrawal from all

The Semeiosis of Suicide in Sri Lanka 87 manner of transactions. To quote Sangili, "there is no giving nor receiving, no going nor coming, no words nor assertions, no saying nor listening, and, finally, there is even no eating" (kofukkal vänkal ilia, pöratu varatu ilia, pSccu väittai ilia, colluratu kekkiratu ilia, kafaiciila cäppifaratumilla). In the words of Sinnamuthu the diviner (ufukku pücäii), "for those who have the affliction of aloneness, there is neither joy nor sorrow, satisfaction nor dissatisfaction, anger nor compassion, fear nor courage."8 (Tanimai tOSam vantä cantöiamum ilia tukkamum ilia, tiruptiyum ilia atiruptiyum ilia, kOvamumilla kanmaiyumilla, payamumilla, taihyamumilla). Viewing "aloneness" from within the context of Hindu society, I was struck by the similarity between the one afflicted with "aloneness-disordern," on the one hand, and the Hindu renouncer, the sanyasi, on the other. I asked my informants about this likeness. There were many answers. But one stands out. "The sanyasi is like a rock inside. The one who has aloneness is like a vast, deep, howling ocean inside. His confusion is great as that" (Canniyäsi, ulla kallupOla. Tanimaiullavan eppafiinu kSffinkannä, Onkira penya kafal pOla. A van kalankal awalavu perucunka). This was Sangili again, who told me that he knew this because his best friend had hung himself after suffering from"aloneness-illness,,."" Tanimai tOiam is like a dark empty cave in the middle of the night" (tanimai töiam vantu nafu rätmyila irunfa verunkukai pölä). And compare this with the words of Angalayi as she ended her narrative, horribile dictu, of how she found her twenty year old son's body on the railway tracks. Angalayi had not noticed anything unusual about her son's behavior except that he had stopped "asking and receiving food. If someone placed food before him he would eat. Otherwise he starved." (cäppäfu keffu vänkirata niruttlfßn. YStO, yärö cäppäfu vaccä, appafi munnäla, tinnuvän. Illäffi paffmiyäkkefappän). There were some specialists who considered "aloneness-disordern" far less sympathetically. The same diviner, cited earlier, also said, "there is nothing to this tanimai töiam. It is mere fat" (tanimai tOSamenkiratu onnumillinka cummä kofuppu). Literally, "kopippu" means "fat." It also refers to the kind of arrogance resulting from a surfeit of comforts, and to being spoiled thereby.) Another, more somatically oriented individual described "aloneness-illnessn" as a "chill in the brain;" "chill," being his own English word in an otherwise Tamil statement. A third was quite clear that the label was "learned from the Sinhala Veddas," and was quite contemptuous about such knowledge. The specialists — be they astrologers, diviners or healers of various kinds — are sanguine about being able to tell what the signs of "aloneness-disorder„" are. One young man, let us call him Bala — who survived a suicide attempt — disagrees.

88 £. Valentine Daniel

No one knew that I had tanimai töSam. I went to work, I ate, I drank, I played cards, I even went to the movies. I would be talking to you like this. You will think, "He is talking to me." But I will not be talking to you. Nobody knows what tanimai tOSam is. Even I did not know that I had it then. Only after I was cured did I know what it was. Even that (still) one cannot say. My tanimai tOSam can be quite different from your tanimai tOiam and his tanimai tOiam.(Enakku tanimai toiam irunticciinu oruttarukkume teriyätu. VÜaikku pönSn, cäppiffSn, kuficcen, clffu aßccüi, pafanküfapätten. Ippafi pScikkiffu iruppen onkalöfa. NInka nenaipplnka, "nammaJOfa pScurün," appifiinu. Ana nän onkalöfa pScamäffen.

Tanimai töiam ennännu oruttarukkume teriyätu.

Enakke atu irunücciinu appa teriyätu. Atu cokamänatukkappurantän atu ennannu teriyum. Atukküfa colla mufiyätu. En tanimai tOSattukkum, onka tanimai tOiattukkum, ivan tanimai toSattukkum, periya vittiyäcamirukkaläm). "Nobody knows what tanimai tOSam is." This statement seems to sum it up best, even if imperfectly. In short, it is as if "aloneness-disordern" has not made up its mind what it wants to be or rather the NTV are not quite sure what it is. One thing is clear: a fully acculturated Final/Normal Logical Interpretant it is not. In other words, "aloneness-disordern" does not belong to this class of interpretants that is a relative Third of Thirds.9 This last allusion is to Peirce's phenomenological categories upon which Peirce's Semeiotic and the trichotomies of interpretants it entails is based. A review of these categories will not only help the non-Peircean reader understand the allusion, but will also bring the three main classes of interpretants being focused on within the purview of Peirce's categorical triads, thereby leading us towards a better understanding of the Tamil person, especially the one affected by "aloneness-disordern." Peirce named his categories, with disarming simplicity:

"Firstness,"

"Secondness," and "Thirdness." Firstness is a monadic property, a singularly intrinsic property of an event or experience, a mere potentiality. This means that: in regarding a thing as having such a property, one is making no implicit reference to any second thing. Insofar as we regard something in this way, we are regarding it neither as existent or

The Semeiosis of Suicide in Sri Lanka 89 non-existent, as real or unreal, since, to regard a thing only in respect to its monadic properties, is to regard it as if it had no relationship whatsoever to anything else (including oneself), whereas the ideas of existence or reality pertain to things in their relationships to one another (Ransdell 1986: 57). Secondness is Peirce's category that describes those properties of things and events that are dyadic or of two-term relations. "Brute" is the word used repeatedly by Peirce to characterize this property of secondness. When the effect of two entities upon each other is a matter of the hie et nunc, ungeneralized (even though not ungeneralizable), and unmindful of reason, the effect may be called "brute," and deemed exemplary of secondness, and describable as the abrupt interruption of continuity. In his taxonomy of consciousness, Peirce calls the kind of consciousness that falls under the category of secondness altersense: The self and the not-self are separated in this sort of consciousness. The sense of reaction or struggle between self and another is just what this consciousness consists in (CP 7.543). In the broadest sense, thirdness is the name for the property of three-place or triadic relationships. One of its governing concepts is mediation (CP 5.66; 5.104; 5.121; 6.32; 8.332). In exchange, which constitutes a genuine triadic relation, giving "does not consist in A's putting Β away from him and C's subsequently picking it up. ...There must be some kind of law before there can be any kind of giving..." (Hardwick 1977: 29). If giving is to be more than a mere mechanical act of physical transfer of B, it must conform to or be mediated by cultural habit or lawfulness. It is the absence of this mediating Third element, what Peirce calls the mental element, that is remarked upon when Sangili says of "aloneness-disorder": "there is no giving nor receiving," or when Angalayi said of her son that he stopped "asking and receiving food. If someone placed food before him he would eat. Otherwise he starved." The triadic structure of the sign-relationship (representamen-object-interpretant) constitutes thirdness in its generic form (Ransdell 1986: 58). Of the three the interpretant is the third because it is that which "brings a First (a representamen) into relation to a Second (the object)" (CP 8.332, the parenthetical insertions are mine; see also Kalaga 1986: 49-50). By extension the interpretive mood that mediates all interpretants is the Final/Normal

90 Ε. Valentine Daniel Logical Interpretant (see Figure S), and hence my characterization of this class of interpretants as the Third of Thirds. What is of importance, however, is the triadicity not of the sign as structure, but rather as process, the process of semeiosis, forever open and forever continuous: "the thread of life is a third; the fate that snips it, its second... Continuity represents Thirdness almost to perfection" (CP 1: 337). The type of sign-object relationship in which this semeiotic feature of openness and continuity is maximally realized is the symbol, again the "third" in the triad of signs in which icon and index are "first" and "second," respectively, and the conveyance of which is thought. And to live as a human qua human is to live in thought. Peirce puts it this way: Every state of consciousness (is) an inference; so that life is but a sequence of inferences or a train of thought. At any instant then man is a thought, and as thought is a species of a symbol, the general answer to the question what is man? is that he is a symbol. (CP 7.583) Milton Singer (1984), McKim Marriott (1976) and I (1984a and 1984b) have argued, each in our own way, that the person in South Asia, unlike the one idealized in the West, is loosely and openly constituted, a "dividual" embedded in the flux of transactions and process. To be bounded, to be static, and to be individuated is to be dead:10 "a symbol, once in being, spreads among the peoples. In use and in experience, its meaning grows" (CP 2.302); it does not flounder. Symbols, like icons and indexes, are stricto sensu dyadic relations and, therefore, not full-fledged, triadic signs. They are characterizations of the nature of the bond that holds representamen and object together. In the symbol this bond is characterized by convention, in the index by contiguity of object and representamen and in the icon by similarity between object and representamen. In fact, however, the very act of imagining these dyadic relationships is made possible only by an artificial exercise of the mind in which the interpretant is prescinded out (or imagined as if it were not there) for the sake of analysis. That is what I meant by the "strict" sense, in which icon, index and symbol are dyadic and not completed triads. However, having accepted the inevitable part played by the interpretant in making a symbol a symbol, an index an index, and so on, we may go on to specify the kind of interpretant entailed in the triangulating or sign completing

The Semeiosis of Suicide in Sri Lanka 91 function. Any one of the nine interpretants in our taxonomy of interpretants can be thus entailed. However, once again, it is the Final/Normal Logical interpretant that is capable of allowing the symbol its greatest freedom, its greatest chances for continuity and generativity, the least chaos and the least caprice. Thus to the question, what is man? or even, what is essentially man? The answer will be a symbol enjoined by a Final/Normal Logical interpretant. This does not mean that men and women as humans qua humans are to be emotionless or incapable of unreflective action. It means that a fully human person has emotions and actions that are already contained or evolving towards containment by the mediating acceptance of a Third, viz., culture, with its complex web of meanings and habits of thought and action. If culture as a Third, as a Final/Normal logical interpretant, is also the locus of reason, lawfulness, habit and meaningfulness, then an event, an object or an encounter is meaningful to the extent that it is not completely discontinuous (or discontinuable) with expectation, habit or belief. When belief is frustrated or interrupted, we call it doubt. If meaning has the property of thirdness, information, in the technical sense, is characteristic of secondness. It takes us by surprise, and it remains so until it is interpretively triangulated into meaning. If secondness describes a world with a surfeit of information and an absence of meaning, it also describes an existence fraught with doubt, shock and caprice. Existentially speaking, then, our New Tamil victim of "aloneness-disordern" is a second rather than a third. As will be shown more fully, the disorder lacks the meaningful mediation of culture, that is, culture's immanent presence within the experience of "aloneness," capable of constituting the latter as a disorder with a culturally comprehensible name. To be a second is to be a mere individual, not part of a community of shared experiences and shared meanings, and, therefore, not wholly human. Recall Bala's words: "Nobody knows what tanimai tOSam is... My tanimai tOSam can be quite different from your tanimai tö&am and his tanimai tOSam." We need to add to this what Peirce had to say about individuality in the strict sense: We know that man is not whole so long as he is single, that he is essentially a possible member of society. Especially, one man's experience is nothing, if it stands alone. If he sees what others cannot, we call it hallucination. It is not "my" experience, but "our" experience that has to be thought of; and this "us" has indefinite possibilities (CP 5.402n).

92 Ε. Valentine Daniel To our NTV victim of "aloneness-disorder," two other famous quotes from Peirce apply. He may be understood as that"[Individual man, [who], since his separate existence is manifested only by [what he is not and what others cannot be], is only a negation" (CP 5.317), whose "existence lies in opposition merely" (CP 1.434). We may see him as a locus of imperfection and error, aberrant and set apart from the rest. Pushing his experience back even further, one may say that it tends, in its sheer virtuality and vagueness, as opposed to its actuality and sharpness, towards firstness. Again recall Bala's words: "Even 1 did not know 1 had it then. Only after I was cured did I know what it was." Compare this with Peirce's attempt to describe Firstness: "it has no unity and no parts. It cannot be articulately thought. Assert it and it has already lost its characteristic innocence;...Stop to think of it and it has flown " (CP 1.357). And again, in Stearns paraphrasing of Peirce, Firstness "is identifiable with feeling before the latter has thereby undergone change" (Steams 1952:199). If the kind of consciousness in which ego and non-ego are precipitated against each other is called altersense (the Second in Peirce's taxonomy of consciousness), the taxonomic First is called primisense. Primisense, as pure feeling, is not felt if "to feel" means to be aware (cf. CP 7.530). "It is in fact, a quality of feeling that 'we cannot attend to...for want of a background' (8.294)" (Houser 1983: 333). Is this state an emotion or a feeling? For that matter, is there a difference between the two? The answer to the second question, I believe is, yes, there is a difference. Emotions such as the ones mentioned by the specialists and Bala — joy, sorrow, anger, fear, courage, satisfaction and dissatisfaction — are all conceptually and culturally mediated. They are thirds, triadically constituted signs of object, representamen and interpretant. One is angry at something, afraid of somebody, and satisfied with something; all for some reason. Furthermore, anger, fear, satisfaction, etc., objectify feelings and determine appropriate interpretants. When feelings, sensations, forces and facts — internal or external — remain (though interpretable) uninterpreted seconds, when their semeiotic continuity is arrested, one is stricken or seized with "aloneness-disorder." As David Savan (1981: 321) writes: "It is characteristic of the strongest passions that while we are in their grip we are least aware of our feelings. It is only after the first surge has passed that we reflect upon our feelings and identify them as feelings of, say, terror." In the case of our victim of "aloneness," the surge continues "like an ocean," with nothing to give it shape or form, with no dynamic interpretant to deliver it into the

The Semeiosis of Suicide in Sri Lanka 93 flow of semiosis and meaning, into the cultural continuity of logical-normal interpretants, conceptually shared or sharable by others in his community. Signs which are specifiable as symptoms — miscarriages, fevers, madness, etc., — or as emotions with cultural labels — anger, sadness, fear, etc., — by virtue of their actuality, generate dynamic interpretants, which are in turn "normalized." "Aloneness-disorder„" may be a sign, but only in that its interpretant is more immediate rather than dynamic or normal, virtual rather than actual or general. But, as we already know, immediate interpretants are signs with potential interpretants — what Peirce called the sign's peculiar interpretability (CP 8.315). And it is this very potentiality

that the optimistic specialists referred to above

choose to emphasize with hope, and thereby attempt to actualize in a dynamic interpretant, and then deliver it to the normalizing function of culture, i.e., to the final or normal logical interpretant (CP 8.315). But Bala, the young man who almost killed himself because of "aloneness-disorder„," with retrospective pessimism, emphasizes the non-actualization — even the non-actualizability — of the same immediate interpretant. For the former, "aloneness-disordern" is a third, or at least a potential third, whereas for the latter it is a second at best, but receding towards the firstness and immediacy of an entropic black hole. We may look at the matter at hand from a complementary perspective. On the one hand, by admitting its difference from fear, anger, joy and other emotions, and by likening it to inchoate feelings and sensations, the specialists are acknowledging the actuality of "aloneness-disordern" as experienced by our young informant. On the other hand, by attempting to normalize it as a concept called "aloneness-disorder n ," they are seeking to turn it in the direction of emotions. Why? Because, as Peirce suggested (CP 5.308 and 5.292), emotions are the hypotheses of affect, and hypothesis is the first step in the construction of reality. To quote Savan (1981: 325): Peirce pointed out that an emotion begins with a situation of unexpected confusion and disorder. We are puzzled over the causes of some new situation, and aware that our normal control over events is disrupted. The future is suddenly uncertain. Our usual assurance has lost its support. We are caught in cross currents of conflicting purposes and conflicting feelings. Into this chaotic situation the immediate interpretant introduces the emotion as a simplifying hypothesis.

94 Ε. Valentine Daniel The easy equation of emotion with the immediate interpretant by Savan is problematical, and I hold with T. L. Short (1986) "that emotions are dynamic interpretants." However, I am in the main in agreement with Savan that emotions are like hypotheses. In suggesting a namable emotion, these specialists proffer a hypothesis, hoping that it would, even as a magnet draws a fugitive iron filing into its field of influence and subjects it to the normalizing forces of the laws of magnetism, draw this fugitive experience into the discourse of culture and "naturalize" it there. The NTV strive to leam the signs of "aloneness-disordern," and do seek help from specialists such as astrologers diviners and healers of various kinds. Details of these specialists' diagnostic and therapeutic techniques vary, but they do have one thing in common. They attempt to provide, even impose upon those afflicted by "aloneness-disordern," an emotion as an interpretant of feelings experienced. Since the actual emotion experienced is the dynamic interpretant, the specialist's task then is to try to convert the immediate into a dynamic interpretant. Once actualized, a dynamic interpretant is exposed to the influence of "collateral observation" and experience, defined by Peirce (CP 3.841) as "previous acquaintance with what the sign denotes," and other factors in addition to the sign it interprets. These other factors or signs may be, among other things, the position of planets as revealed by the astrologer or an evil charm as detected by the diviner or a humoral imbalance diagnosed by the physician. These factors are normalizing forces; forces of Thirdness. Actual emotions themselves, by virtue of their being dynamic interpretants, may be said to further have emotional, energetic and logical instantiations. As emotional-dynamic-interpretants, anger and fright have distinctive qualitative effects. When these effects encounter no resistance, they evoke further sympathetic responses of anger or fright, as the case may be, from moment to moment of the evolving self or from others with whom one lives or works. Usually the latter, that is, the replication outside the self, is likely to be short-lived, meeting resistance, deflection and transformation. But this is not always the case, as could be attested to by the various crowd phenomena we have record of in the modem (and even ancient) world: the liberating crowds of Bastille, the rhythmic crowd of death-chanters, the crystallized crowd of Nazi Germany, the flight crowd of the "tiger day,"11 the refusing crowd of non-violent protesters, the pack-crowds of the 1983 anti-Tamil pogroms of Sri Lanka, and the baiting crowd of the lynch mob. 12 When emotional dynamic interpretants such as anger or fear are self directed, and if these emotions are negative, they can reach a point where their only check becomes

The Semeiosis of Suicide in Sri Lanka 95 an energetic act of self-destruction, giving Peirce's expression (CP 1.337), "the fate that snips the thread of life," special force. The specialists, therefore, are not only concerned with the conversion of an immediate interpretant into a dynamical one but then in proffering a resistance to the unchecked replication of a dynamical emotion, by throwing it into redemptive conflict with the interrupting forces of the culture's interpretive interests, interrupting thereby its downward spiral. In this respect, the massively interruptive intrusion of the Sri Lankan armed forces into the lives of these Tamils has succeeded better than the symbolic domain of their culture and its traditional wisdom. The initial result of such an interruption is a further emotion which Peirce would have called shocking or percussive (CP 8.370). This interruption, I think, may have unwittingly effected a drop in the suicide rate. It has precipitated an energetic interpretant of armed resistance in some cases. Initially in such responses, these energetic interpretants manifest themselves as spontaneous, unhabitualized, events. More correctly, such responses were governed by relatively primitive, self-defensive habits. Recently these primitive habits have undergone a "habit-change." It is this actual formation of a new habit or the change in an old one that is a logical dynamic interpretant, Peirce's third grade of dynamic interpretants. As a logical interpretant it provides "good reasons" to kill the enemy and not one's self. Anger, frustration and fear that were once inner-directed, and lost in "aloneness" and to suicide, have now begun to spread outward, in manifestations of "reasoned" armed violence. The outward spread may appear to sustain semeiosis, connecting, mediating, exchanging and creating thirds. However, to the extent that the replication of interpretants is faithfully iconic, numbed by redundancy, we may justifiably say that genuine growth characteristic of symbolic freedom has ceased and the quintessential creativity of semeiosis is "dead." In this instance, death, as the extinction of a community's semeiotic life, is a second; it is merely "the final result of the complete induration of habit reducing the free play of feeling and the brute irrationality of effort to complete death" (CP 6.201). The question that needs to be asked is whether this sad state of affairs among the New Tamils of Vavuniya in particular, and the island in general, is the final logical interpretant. Be the sign in question aloneness, suicide or murder, the final interpretant is the interpretant that would be the best interpretant of that sign, given a goal of interpretation. At least as I see it, the encouraging fact is that there is a widely shared sentiment in South Asia that killing, in whatever form, is evil.

96 Ε. Valentine Daniel It is my hope that this will provide a final interpretant to whose gentle persuasion all else will, some day soon, submit.

Afterword Admittedly, the selective focus on the "ecology of affect" may appear to slight the analysis of those wider socio-economic forces that have made the NTV their victims. A semeiotic analysis, unlike many other analytical modes, does not claim for itself the power to totalize, to complete, or to close the inquiry. It is by definition partial and must be so. This is not to admit that those others' claims at totalization are any more warranted. On the contrary, it interrogates all totalizing strategies. A semeiotic analysis captures but a moment in the flow of events. But it also does more. It points to those areas which inquiry and analysis have yet to penetrate; those areas that remain in the dark, undescribed and uninterpreted. In the case of Sri Lanka, the provenance of the socio-economic forces alluded to lies in struggle, in the exploitation of the unprivileged, the accumulation of capital by the few, the history of colonialism, racism, and more recently militarism. As socially lived experience, the significant effects of these forces are likely to find their accounting largely in that phenomenological domain characterized by Peirce as brute secondness. And in so doing it invites such more specifically sociological thinkers as Dürkheim, Weber, and Marx, especially Marx — whom I read as the greatest analyst of sociological "secondness" — to fill the lacunae, and to shed more light. Alas, an apology must be made for the tedium that the Peircean semeiotic terminology is likely to have created in the short run. Given the relative novelty of its application in interpretive anthropology, this is inevitable. And yet it is hoped that the use of Peircean semeiotic is seen as a gentle nudge to traditional interpretive anthropology to grow up and out of its footloose and fancy-free days. But most of all, it is also hoped that the time will come when the "peculiar terminology" will not be needed as declarative indexes any longer, and instead the powerful perspective of semeiotic, with its capacity for providing finer and finer insights into ever shifting realities, will be as commonplace and second nature as Cartesianism. Has not the reign of the latter long outlasted its usefulness?

The Semeiosis of Suicide in Sri Lanka 97 Notes 1.

2. 3. 4.

5.

6.

7.

8.

A very special word of acknowledgement is hereby extended to Ben Lee, Richard Parmentier and Greg Urban whose critical reading of an earlier draft of this paper has saved the present version from several embarrassing errors and helped make it a better essay in general. Thanks are also owed to H.L. Seneviratne and Sam Sudanandhar. Needless to say, the responsibility for the final version is mine. I am also grateful to the National Endowment for the Humanities and to the Institute for Advanced Study at Princeton for providing me an ideal environment in which to prepare the present manuscript. For further details see Mick Moore (1985:65-84), N. Sanderatne (1972), R.D. Wanigaratne, et al. (1980). For instance, this tendency among the Sinhalese, to conflate cause and effect, is seen in Lorna Amarasingham Rhode's paper (1984). '"Cenopythagorean,' because like the Pythagorean, they are essentially numbers, yet are neither Pythagorean nor Neopythagorean, but fresh, καινό Pythagorean" (Houser 1983: 352 fn. 10). Peirce has left us many formal definitions of the sign. The examination of the differences, similarities, and contexts in which these definitions occurred are interesting and important but may be overlooked for the present purpose. The three definitions chosen happen to provide us with the least confusion and the greatest clarity available and called for by the purpose of this essay. The expressions, "best" and "goal of interpretation," deserve unpacking, a task which extends beyond the scope of this essay. For the curious reader I recommend David Savan's (1977) attempt at clarifying Peirce's signclassifications and T.L. Short's (1986) response to this attempt. For now, suffice it to note that the goals of interpretation are culturally constituted, their ends are directed towards the expectations that have become habituated within the culture in question. The reasoning behind such pairing should become clear when we consider the two trichotomies in terms of Peirce's three phenomenological categories of Firstness, Secondness, Thirdness — to be discussed below — that generate the two three-fold interpretants. Many of these symptoms match up with what biomedical psychiatry might check as symptoms of clinical depression. See the Diagnostic and Statistical Manual-Ill (American Psychiatric Association 1980: 213-215), and Arthur Kleinman (1986:36-41 and f.n. 1 on pp. 228-29). However, any such

98 Ε. Valentine Daniel equation merits examination under a semeiotic analysis before being made definitively. 9. By a similar logic it may be seen that the Dynamical Energetic Interpretern is a Second of Seconds and the Immediate Emotional Interpretant is a First of Firsts. 10. To emphasize the openness and fluidity of the South Asian person is not to deemphasize the concern that South Asians do have about the uncontrolled and uncontrollable fluidity and the unmitigated vulnerability of personal boundaries. Recognition of the dividuating reality is counterbalanced, to varying degrees in different contexts and by different kinds of persons, by attempts at individuating and rebounding the person against an uncontrolled flux. 11. "Tiger Day," is Friday, August 1, 1983, in Sri Lanka. This was the day when rumor spread that "Tamil Tigers," members of a militant separatist group, had landed in the heart of Colombo. Panic-stricken citizens of Colombo started fleeing south in their cars, bicycles and on foot. When it was realized that there had been no basis for the rumor, Sinhala mobs went on a rampage, massacring Tamils, in an attempt to redeem their honor. 12. See Elias Canetti's delightful essay on Crowds and Power, where several of these and other crowd phenomena have been analyzed.

References American Psychiatric Association (1980). Diagnostic and Statistical Manual of Mental Disorders-Ill . Washington, D.C.: The American Psychiatric Association. Cannetti, Elias (1972). Crowds and Power. New York: Farrar Straus Giroux. Daniel, E. Valentine (1984a). The pulse as an icon. In South Asian Systems of Healing (=Special Issue of South Asian Studies 18), E.V. Daniel and J. Pugh (eds.), 115-126. Leiden: Brill. (1984b). Fluid Signs: Being a Person the Tamil Way. Berkeley: University of California Press. Dewey, John (1946). Peirce's theory of linguistic signs. The Journal of Philosophy 43:85-95. Dürkheim, Emile (1952). Suicide: A Study in Sociology. Glencoe: Free Press. Giddens, Anthony (1978). Dürkheim. Sussex: The Harvester Press.

The Semeiosis of Suicide in Sri Lanka 99 Hardwick, Charles S., ed. (1977). Semiotic and Signifies. The Correspondence between Charles S. Peirce and Victoria Lady Welby. Bloomington: Indiana University Press. Houser, Nathan (1983). Peirce's general taxonomy of consciousness. Transactions of the Charles S. Peirce Society 19(4): 331-359. Kalaga, Wojciech (1986). The concept of interpretant in literary semiotics. Transactions of the Charles S. Peirce Society 22(1): 43-58. Kearney, Robert N. and Barbara D. Miller (1985). The spiral of suicide and social change in Sri Lanka. Journal of Asian Studies 45(1): 81-101. Kleinman, Arthur (1986). Social Origins of Distress and Disease: Depression, Neurasthenia, and Pain in Modern China. London: Yale University Press. Lee, Benjamin (1985). Peirce, Frege, Saussure, and Whorf: the semiotic mediation of ontology. In Semiotic Mediation, E. Mertz and R. J. Parmentier (eds.), 99-128. New York: Academic Press. Marriott, McKim (1976). Diversity without dualism. In Transaction and Meaning, B. Kapferer (ed.), 109-142. Philadelphia: ISHI. Moore, Mick (1985). The State and Peasant Politics in Sri Lanka. Cambridge, England: Cambridge University Press. Parmentier, Richard J. (1985). Signs' place in medias res: Peirce's concept of semiotic mediation. In Semiotic Mediation, E. Mertz and R. J. Parmentier (eds.), 23-48. New York: Academic Press. Peirce, C. S. (1931-1958). Collected Papers of Charles Sanders Peirce, 8 vols., C. Hartshorne; P. Weiss; and A. Burks (eds.). Cambridge: Harvard University Press. Ramanujan, A. K. (1985). Poems of Love and War. New York: Columbia University Press. Ransdell, Joseph (1986). On Peirce's conception of the iconic sign. In Iconicity: Essays on the Nature of Culture, P. Bouissac; M. Herzfeld; and R. Posner (eds.), 51-74. Tübingen: Stauffenberg-Verlag. Rhodes, Lorna Amarasingham (1984). Time and the process of diagnosis in Sinhalese ritual treatment. In South Asian Systems of Healing (=Special Issue of South Asian Studies 18), E.V. Daniel and J. Pugh (eds.), 46-59. Leiden: Brill. Sanderatne, N. (1972). Sri Lanka's new land reform. South Asian Review 6(1): 719. Savan, David (1976). An Introduction to C.S. Peirce's Semiotics. Toronto, Ontario: Prepublication of the Toronto Semiotic Circle.

100 Ε. Valentine Daniel (1977). Questions concerning certain classifications claimed for signs. Semiotica 19: 171-195. (1981). Peirce's semiotic conception of emotion. In Proceedings of the C. S. Peirce Bicentennial International Congress, K.L. Ketner, et al. (eds.), 319-334. Lubbock, Texas: Texas Tech. Press. (1986). Response to T. L. Short. Transaction of the Charles S. Peirce Society 22(2): 125-143. Shapiro, Michael (1983). Language as Semeiotic. Bloomington: Indiana University Press. Short, T. L. (1986). David Savan's Peirce studies. Transactions of the Charles S. Peirce Society 22(2): 89-124. Singer, Milton (1984). Man's Glassy Essence: Explorations in Semiotic Anthropology. Bloomington: Indiana University Press. Stearns, Isabel S. (1952). Firstness, secondness, and thirdness. In Studies in the Philosophy of Charles S. Peirce , P.P. Wiener and F. H. Young (eds.), 195-208. Cambridge: Harvard University Press. Wanigaratne, R. D., et al. (1980). Implementation of Land Reform in Selected Villages of Sri Lanka. Kuala Lumpur: Asian and Pacific Development Administration Center. Whorf, Benjamin Lee (1956). Language, Thought and Reality. Cambridge, Mass: M.I.T. Press.

Responsibility and Compensatory Justice in Arab Culture and Law Lawrence Rosen

The attribution of responsibility is one of the most problematic and revealing aspects of any culture. Whether it is in Cain's audacious challenge to his creator or in the Quran's assertion that "no man bears the burden of another," the allocation of responsibility taps into a host of assumptions about what human beings are really like and against what expectations their relations to others may be viewed and appraised. To address the issue of responsibility is, therefore, to consider a broad range of attendant concepts — justice, causation, identity, will — whose particular manifestations partake of both the regularizing structures and the deepest uncertainties to which each culture falls heir. The present paper will concentrate on the conceptual and relational features that cohere around the idea of responsibility in contemporary Islamic life and law.1 I will try to show how ideas of the person and of time, of narrative style and truth give shape and direction to those actions and perceptions that we may broadly treat under the rubric of responsibility. By extending the particular interpretation offered here to specific aspects of Islamic law I will also try to show how these ideas bear on concrete issues that come before the Muslim courts, particularly those bearing on the question of compensation for harm done. Throughout, my comments will largely refer to research I have conducted over the course of a number of years in Morocco although comparisons will also be drawn to other parts of the Middle East. Although I certainly do not mean to imply by my general use of terms like "Islam" and "Arabs" that there is a single Muslim culture or that Arabs are exactly the same throughout the Middle East I do believe that many of the features described here have validity, in a theme and variation sense, for a very wide range of countries in the Arabic-speaking world and may even prove suggestive for some 2 other Islamic countries as well.

Person and Relationship in Arab Culture In a series of recent publications I have suggested that in thinking about cultural life in the Middle East an analogy that may prove useful to keep in mind is that of the marketplace, the bazaar (see, e.g., Rosen 1984).3 It is a domain in which

102 Lawrence Rosen hawking and haggling are rampant, where prices are not fixed and considerable uncertainty may exist over the quantity and quality of goods, where attachments of a patron-client nature often prove most effective as economic strategies, and the overall well-being of a merchant or consumer can change with striking rapidity. A similar image also applies to the realm of social relations. For just as the free play of individuals in the marketplace seems barely containable by the precincts of their endeavors or the limits of human audacity, so, too, in the formation of social relationships an intense process of interpersonal negotiation, framed by the conventions of language, etiquette, and social constraint, informs virtually every act. It is an image of contract and negotiation that is replicated in numerous domains: in the Quran, where mankind is envisioned as free to enter relationships of any sort so long as they do not overstep that limited number of substantive rules known as "the bounds of God"; in popular discourse, where the vision of each person as the center of a network of obligations becomes the impetus for a constant quest for knowledge about others' ties and the ways they are most used to forming them; in literature and popular culture, where stories center on the ways others have maneuvered their social ties as they seek to carve out a safe haven in a world of premonitory chaos; and in law, where the image of humanity as proprietary creatures gives point to a host of procedures, presumptions, and prescriptions that seek to order this defining quality. Whether it be in the relations among total strangers looking for a basis of mutual comprehension and engagement or in the heart of the family itself, the sense of the negotiable relatedness of all persons runs as a constant theme in Arab cultural life. The implications of this orientation for moral and legal approaches to the ideas of responsibility and justice are, of course, profound. Franz Rosenthal (1983:36) accurately captured a central feature of Islam when he said that since its earliest times it has been a religion in which "man was seen...as the center of action in this world." In religious doctrine and popular views alike human beings are seen as so endowed with the capacity for reason ('aqel) that they may acquire control over their passions (nafs) and thus avoid the chaos (fitna) that their own forgetfulness and urges might otherwise engender. For reason to develop, however, it is important to place oneself in association with those teachers or leaders who, by the development of their own reason, can provide the context for the enlargement of one's own self-control. Thus knowledge of Holy Writ and of worldly affairs both serve to develop the capacity to act wisely and responsibly. And since, as the Prophet himself said, there is no distinction among believers except as to knowledge there is even strong support in religion for the

Responsibility and Compensatory Justice in Arab Culture and Law 103 legal proposition that the more knowledgeable a person is the higher the standard of responsibility to which he should be held. The image of mankind as reasoning creatures is further supported in a number of other domains of social and cultural life. Consider, for example, the concept of time. In the west time has, at least since the period of ancient Greece, been envisioned, in no small part, in terms of such metaphors as growth, development, and unilinear direction. We say that time flies like an arrow, and we think of the person as possessing a life history that accords with such an image. In the Arab world, however, time does not reveal the truth of persons: it is not by tracing the course of a person's life over chronological time that one can most thoroughly apprehend a person's distinctive qualities. Rather, for the Arabs time is envisioned as a series of discrete packets of relationship, moments that encapsulate particular networks of obligation individuals have personally negotiated. To know a person what one needs to know is not some distinctive self that perdures across time and is made evident in the course of growth and development: instead, one wants to know the variety of relationships one has participated in for it is such relations that show the most essential feature about a person, his ties to others and his ways of establishing them. The same image is manifest in narrative style. In the west stories are generally told in chronological order — again, because time reveals certain truths about persons. But westerners who approach the Quran or listen to Arabs relating popular stories or accounts often find the recitation confusing and disjointed. Instead of moving in a fairly clear chronological order the story often jumps about in time: instead of giving a clear picture of events by referring to the sequence of their occurrence central characters are referred to in numerous situations whose precise chronological order is not necessarily given. Yet if it is not time but contexts of relationship that are central to this vision of the person the relative avoidance of strict chronology becomes more comprehensible. Thus the Quran shows us the figure of the Prophet by showing the various contexts of his utterances and acts, while popular stories show us individual lives, like the facets of a gem, by turning them round in the light of a series of relationships whose precise sequencing is comparatively unimportant. The overall image of the person, then, is of a proprietary creature whose exercise of reason plays a significant role in the exploitation of negotiated bonds and whose totality is knowable to others to the extent that they can gain information about the variety of his ties to others. That each person has a high degree of freedom of choice is sustained in religion and lore: it is directly assured in Quranic statements

104 Lawrence Rosen emphasizing the contractual and covenantal quality of the ties between God and man and among men themselves; it is demonstrated in ordinary narratives, for example, where repetition is used as a device to demonstrate that at each moment in time a reasoning man can choose what to do and that no two moments are ever precisely the same because the multiplicity of relationships that form the surround are forever in a state of flux. Men do not fashion their own selves, in the sense of creating moral visions of their own, nor do they possess an internal psychological structure separate and distinct from their overt acts. Instead, men fabricate the contexts of their relationships, their webs of indebtedness, and in doing so create islands of relative certainty in a world of potential chaos. A constant quest for information about others' networks animates everyday life. It is a vision of the world that is well expressed by a Muslim trader in one of Joseph Conrad's (1976: 113) novels when he says: "In the variety of knowledge lies safety." The preservation of one's freedom to negotiate relationships wherever they may prove most advantageous is also reflected in the Arabs' idea of truth. In the west we tend to regard assertions about relationship as capable of being tested for their truth value. To say that someone is your friend or has certain obligations to you may be true or false but is at least susceptible to evaluation in terms of its veracity. By contrast, for Arabs it is well understood that an assertion of relationship, standing as a bare utterance, is not necessarily subject to evaluation in truth terms at all. It is like a price mentioned in the marketplace which does not take on the capacity for truth assessment until something more happens — until it is validated. This validation itself may occur by the use of a holy oath, by agreement among the parties, by the formation of additional ties created in reliance of another's utterance, or by having someone who is himself regarded as a reliable witness acknowledge the statement Once an utterance has been validated truth and lying become extremely important. Thus even in the conceptualization of truth the image of a marketplace of possible arrangements which men must be free to attempt without being marked as liars should the attempt fail reinforces the image of negotiation that runs through so many domains of Arab cultural life. If it is the contexts of bargained for relationships that is central to the definition of the person in Arab culture then it also makes sense that for the Arabs the idea of intentionality should also look somewhat different than that found in the west. The Arabic word for intent, niya, not only means "purpose," "design," and "will," but inasmuch as these features manifest themselves in direct faith niya also means "simple," "naive," and "sincere." It suggests in its semantic range what informants readily acknowledge in actual use, that a person constantly displays what lies inside

Responsibility and Compensatory Justice in Arab Culture and Law 105 him when he speaks or acts. Indeed, most Arabs do not recognize the idea of a distinct inner self that could exist apart from action, only of overt expressions which must of necessity conform to what a person must carry inside himself. To the extent that a person can be conduced to speak or act, or to the extent that one can learn of the variety of another's past utterances and deeds by careful investigation, one acquires direct access to that person's intentions. In ordinary social life this means that people readily presume to know others' states of mind and do not hesitate to elicit behavior or exert pressure by characterizing words and deeds in such a way that their attendant motivations may also be adduced (see Mills 1940, Burke 1962). In law, as we shall see, this means that judges usually presume they can discern intent simply from what a person has said or done and that intent can, therefore, remain a distinct element of legal consideration even when, in terms of formal doctrine, it does not necessarily appear as a constituent feature. For the Arabs it also follows that what matters most in evaluating actions is not their connection to a series of abstract propositions that lie behind them but to the consequences that actions have in the world, their impact on those networks of relationships, those webs of obligation, that are constitutive of reality itself. Thus in ordinary life one constantly encounters reference to the effect that words and deeds have in the world as the means for assessing what they really are. And in law it is the consequences of an act, rather than its antecedent precepts, in terms of which a logic of remedy will be fashioned. Indeed, because the harm that people can do varies with their knowledge and hence their connections to others the evaluation of harm itself turns in no small part on the assessment of a person's situated ties. When, therefore, we tum directly to the realm of Islamic law as it is actually practiced in contemporary courts the proceedings become far more comprehensible for being placed in their larger cultural context. Western scholars have often said of Islamic law that it lacks the rigorous logic of doctrinal development found in a number of the other great systems of law. No general concept of contract exists in Islamic law, it is said, only specific forms of permissible or impermissible contracts; no idea of good faith is present, only concrete practices that are or are not enforcible (Schacht 1964). But such an analysis misses a key aspect of Islamic law. Although it is true that the legal scholar often appears in Islamic history to have played a more prestigious role than the court personnel themselves and even the scholars did not strive for logical consistency as the preeminent goal of their work. Islamic law displays in its practice features that, by completing an otherwise

106 Lawrence Rosen truncated view, demonstrate the particular logic that informs its overall shape. This order is most evident when one looks at the ways in which facts are shaped for decision and evaluated for their believability. In its modes of proof Islamic law gives effect to the underlying cultural assumptions of the society at large. For just as Arabs assess events and persons by concentrating on the individual person as a situated actor in a variety of particular encounters, so, too, in court it is by focusing on the socially constituted person that the judge can shape and weigh the facts. Thus all evidence is regarded as essentially oral in nature. Witness testimony is the predominant form of proof: the greater the number of witnesses, the more consistent their story, and the greater their direct contact with the parties in question the more their version will appear to be credible. Even documents are regarded as the reduction to writing of oral testimony, and the techniques for assessing their believability are not mainly those one thinks of for documentary proof — the correctness of form or the indication of forgery — but for oral believability. Thus the main questions asked are: Who is this person who is testifying or whose words are inscribed here? What is his social background and to whom is he attached? In the past, each witness actually had to be certified as credible by the judge and notaries of the court before being allowed to testify, credibility itself being a function of a course of dealings with people in the community suggestive of sufficient contact that a person would not readily lie lest it affect his network of obligations and the willingness of others to form ties with him. Although at present this type of certification process no longer exists, the criteria for assessing believability remain those which are based on the face-toface assessment of oral testimony. Similarly, the court continues to have attached to it certain personnel who are, in a sense, institutionalized reliable witnesses. For any document to be valid it must first be witnessed by two of the court notaries (pi. 'adul) who by inscribing their names lend their voice to that of the party in certifying a given transaction. When necessary, the court will also turn to one or more of the experts who are attached to the court and who are regarded by virtue of their experience to possess vital knowledge about such subjects as building practices, costs of living, or the nature of medical issues. When a matter remains in doubt notwithstanding the use of witnesses the court may also turn to the use of a decisory oath. This practice allows the court to assign one party the opportunity to take a holy oath and thereby win his claim. If the party who is offered this opportunity refuses he may, however, require his opponent to take the oath and if the latter does so it will be the latter who wins.

Responsibility and Compensatory Justice in Arab Culture and Law 107 What makes the system highly distinctive is that it is for the judge to assign the oath first to one or another party not on the basis of who is the plaintiff or defendant in the case — this structural feature being present in the version of the decisory oath still found in some European jurisdictions — but on the basis of who it is, in the court's opinion, who is most likely to know what is true about the particular matter at hand. Thus what comes to play a great role in the oath is actually the same feature one finds present in the use of witness evaluation, judicial questioning, and the use of reliable witnesses attached to the court — namely, a constant reference back to local custom, local conventions, and local people who are familiar with local contexts of relatedness. And it is here, too, that the consistency that appears lost at the doctrinal level shows itself most clearly. For Islamic law is, as it were, organized vertically, not horizontally: its referent is not other doctrinal propositions logically related to one another to form a coherent body but the relation of general propositions to the local circumstances that give them content. Seen in this light Islamic law is highly organized and developed. Custom need not be a formal ingredient of the law: it is the baseline to which each issue ultimately has reference and hence the support for a system that finds in localized persons, practices, and presumptions the content for its overall form. It is a system which can be highly effective in big cities or in rural locales if what we understand by effectiveness embraces two vital aspects: first, that the predominant goal of the law is not simply to resolve differences but to put people back into a position where they can, with the least adverse implications for the social order, continue to negotiate their own arrangements with one another; and, second, that even though the specific content of a court's knowledge about particular individuals may be both limited and stereotypical, the terms by which the courts proceed, the concepts they employ, the styles of speech by which testimony is shaped, and the forms of remedy they apply are broadly similar to those that people use in their everyday lives and possess little of the strange formality or professionalized distortions found in some other systems of law. The result, then, is a legal system that remains relatively close to the terms and perceptions found in a host of other domains of social life. To many Muslims courts can no more be expected to preserve individual rights than might be expected in life at large, nor be any less corrupt or any more wise. For them courts are familiar realms that gain in legitimacy from the very familiarity of their style and underlying assumptions; they are part of life and not some extraordinarily refined version of it. And it is in these terms — of seeking out the manifold connections

108 Lawrence Rosen between cultural conceptualizations and their legal articulation — that we can approach the particular issues of responsibility and compensatory justice.

Compensatory Justice In many legal systems of the world the conceptualization of wrongful actions tends to be organized around such categories as private versus public or criminal versus civil. Classical Islamic law, like some other religious systems, begins from a slightly different base. It distinguishes those acts for which punishment is prescribed (Aadd) from those in which judicial discretion may be exercised (ta'zir). The former includes unlawful intercourse, drinking alcoholic beverages, theft, highway robbery, and the false accusation of unlawful intercourse. Because they are regarded as acts that violate the bounds, rights, and claims of God it is not for human beings to interpose their own judgment as to appropriate remedies. By contrast, discretionary remedies apply to all those unlawful acts for which mankind may set the appropriate level since they infringe on those "claims of man" which the Quran specifically leaves men free to negotiate among themselves. Thus killing — to choose only the most notable example — is a matter to be composed by human judgment rather than prescribed punishment. Although, as we shall see, discretionary remedies are themselves shaped by a variety of conventions their assessment and resolution are within the realm of human control. The logic of the hadd-ta'zir distinction, however, goes well beyond a simple division of religious versus secular. It implies a set of cultural expectations and assumptions that are deeply intertwined with those we have already discussed. Acts that are subject to discretionary remedies have in common that they break the continuity of a relationship forged by human beings as proprietary creatures in such a way that it is nevertheless possible to reconstitute ongoing relationships without having to consider their impact on society at large. The harm these acts cause is, in a sense, limited by the break in continuity of relationship which a properly chosen remedy could help to reconstruct. Hadd-type wrongs go beyond this boundary inasmuch as, in the Islamic concept, they not only intrude on a domain marked out by God but because that domain is perceived as fraught with vast potential for social chaos. It is a realm of acts in which the reconnecting of bonds is regarded as particularly difficult and in which the systemic repercussions for many other networks of obligation are difficult to foresee and control.

Responsibility and Compensatory Justice in Arab Culture and Law 109 Both types of action possess significant procedural safeguards that establish when an act is of one type or another, safeguards that are aimed, in part, at preventing mankind from mistaking one form of action for the other. Yet even these safeguards turn on the question of consequence and harm. Thus, in the case of unlawful intercourse classical Islamic law requires four eye witnesses to the act itself, an evidentiary requirement that is not only difficult to achieve but which, when it is, demonstrates that the act must have been so open and notorious that its social repercussions are vast indeed (Yamani 1968). Once established, hadd wrongs also take the offender out of the ordinary way of assessing acts: where normally Islamic law, like Arab culture, stresses the person in all his situated obligations and characteristic modes of interaction, the establishment of a hadd wrong removes the individual from further consideration as a person (in the Arab sense) since reintegration is not possible. By contrast, ta'zir wrongs retain the full panoply of personalistic considerations found in all domains of Arab social life. Similarly, justice is rendered in the former case by the very act of de-personalization where in the latter it can only be rendered by the fullest possible consideration of the person. Or, put somewhat differently, justice for a hadd wrong consists in the withdrawal of sociality through de-personalization while ta'zir wrongs stress reintegration through retention of the normal modes of person perception. This pattern may also help to account for the seemingly inconsistent way in which intentionality and causality enter Islamic law. It is generally said of Islamic law that it is a system that pays no attention to an actor's state of mind when a Aadd-type wrong is involved. Here, the harm is regarded as so great, and the infraction so deeply involved with religious proscription, that a principle of strict liability may apply. When, however, one turns to the realm of ta'zir injuries the matter appears to become more complex. Thus in classical Islamic law the remedy for the killing of another varies, in part, with the intent of the perpetrator. But this intent is imputed predominantly from the weapon employed: if the weapon is the sort that is normally used for killing the intent to kill will be assumed; if something normally less lethal was employed it will not. Moreover, the recourse — retaliation by the deceased's kinsmen or the payment of bloodmoney — turns on the characterization of the intent of the actor.6 However, when non-lethal injury to a person or damage to property occurs Islamic law has traditionally focused on several distinct elements — the unlawfulness of the act itself and the range of damage caused. Thus it is predominantly the fact that an act is illegal that makes liability attach — hence even an infant may be held personally liable for the harm he does if his act was itself unlawful. Nor can the

110 Lawrence Rosen victim's own consent ever justify the wrong. Liability exists only for those damages which the facts show were directly related to the act, the mental state of the actor being relevant only if it assists in establishing the causal connection rather than as a vehicle for introducing fault (Chehata 1970: 140, Limpens, et al. 1983). Thus classical Islamic law appears to stress that if an act is deemed wrongful no separate inquiry of mental state is necessary although a showing of mental state may be sufficient to draw the causal connection to direct damage which an assessment of the ordinary course of events could as well reveal. This apparent interchangeability of strict liability with certain considerations of intent — what to some appears an inconsistent use, or poor logical development of, the concept of intentionality — makes sense, however, when seen against the cultural conceptualization of others' minds to which we referred earlier. For if it is believed that a man's acts are necessarily connected to his state of mind and that such a set of acts can be deciphered in terms of social background, connections, and modes of negotiating obligations, then it follows that for the Arabs another's intentions are regarded as readily available to discernment and do not constitute a separate domain hidden from human view. This attitude is clearly articulated by judges who say that they can always tell what is in a person's mind — by reviewing the course of his utterances and actions. A Saudi judge put the matter in the following way: "The judge has to have an acute sense of observation; for example, just by looking at a suspect he should be able to tell what the man had concealed in his testimony." Or as a Moroccan judge put it: "If I question people, if I find out who they are and what they have done, I can always tell if they are lying, I can always tell their intent (niya)." Similarly, one sees in the legal ideas associated with liability the constant stress on consequence so common in other domains of Arab cultural life. Direct damages are compensable in law, including the costs that relatives of an injured party may have to incur in visiting and caring for the person. The way in which the chain of damages is itself constructed has certain distinct features. First, the damage must be concrete: what European lawyers call "moral damages" and Americans "compensation for pain and suffering" are not included. Second, the chain of causation does not extend to everything which might be foreseeable but to those injuries that disrupt the injured party's capacity to operate within his web of obligations. Thus, thirdly, even involuntary injury possesses the same quality as voluntary acts since it is the harm it does, not the meaning that lay behind the deed, that is central. Moreover, the category of wrongs includes not only acts that

Responsibility and Compensatory Justice in Arab Culture and Law 111 are clearly unlawful but those undertaken in disregard of a person's security ([shart as- salama) — a category which allows a significant degree of discretion within the general framework of focusing on the impact on the injured person's obligational network. It is the rupture to what is most distinctive to human beings — their proprietary associations with persons and things — that shapes what the law seeks most to protect, whether it be the killing of a kinsman or the usurpation (ghasb) of another's property. It is this emphasis, too, that is central to the remedies that Islamic law has fashioned From this perspective we can now recapture and elaborate the Islamic ideas of responsibility and justice. We have already seen that, in a sense, responsibility is connected not to the question of what lies behind a deed so much as to what it leads to. To this day in Saudi Arabia, where the system of compensation for unlawful killing still applies, it matters less why a killing has occurred than that compensation must be paid for its repercussions. What one is responsible for — who one is responsible to — is the schema of social obligations by which any individual may possess identity and by which society as a whole may stave off the threat of chaos. Responsibility is a personal matter even though compensation may have to be acquired by utilizing one's collective attachments. Indeed the cultural conceptualization of responsibility is closely linked to ideas g about the nature of causality itself. Briefly, Arab social and legal thought tends to draw causal chains quite narrowly: a person or thing is said to have caused something if no other entity, sentient or material, intervenes to displace the commonly expected results. Two distinct lines of analysis have followed from this orientation in the course of Islamic legal thought. On the one hand is the line of direct causation. This is exemplified by such instances as the following: If a man opens the door to a bird's cage and the bird flies away it is the bird and not the man who, by the exercise of its natural propensity, has caused the escape; if X unlawfully places a stone on a public path and Y, in stumbling over it, injures Z, only Y is said to be the cause of Z's injury. Such examples, which appear not only in classical texts of an earlier age but in some modem jurisdictions, suggest that causation, and with it responsibility, are to be limited to only the most direct and "natural" consequences of an act. By thus limiting causality the system reinforces the cultural emphasis on direct and personal contacts, and places a significant limitation on arbitrary judicial action and the potential for socially and politically disruptive retaliation. A second line of analysis, however, looks to the idea of fault in the sense of overstepping the permissible limits of action. It holds that if one acts in a way

112 Lawrence Rosen that is not normally regarded as appropriate to a given endeavor the consequences of this negligence may be attributed to oneself despite intervening factors. Thus a man who loads a pack animal improperly is liable if the pack later falls and injures another, and a man who lets too much water run out on the road to settle the dust is regarded as the cause of a horse slipping and falling on a passerby. Although these two approaches have had a long and varied relationship in different periods and jurisdictions they share certain distinctive cultural emphases. For both imply that there are culturally recognizable propensities possessed by all creatures and things and that the normal course of causality is a function of their innate natures. Humans are not responsible if some other entity, with a nature of its own, displaces their's unless, by failing to recognize the nature of the artifacts and creatures that fall und« their control, they allow them to do harm to others. It is a way of thought that can, therefore, hold children strictly liable for their injurious acts or blame an animal for doing what comes natural to it because human beings must simultaneously take responsibility for what comes within their proprietorship and be free from responsibility for that which gives expression to its own intrinsic nature. It is here, too, that the ideas of excuse and accountability enter. For most westerners, responsibility incorporates elements of mental state and morality. One need hardly subscribe to the moral imagination of a Richard Nixon who, caught in the tightening web of Watergate, said (as did Ronald Reagan during the Iran-Contra affair) that he was prepared to accept the responsibility but not the blame in order to understand that for us the line between self and society may sometimes be made to coincide with the line between the moral and the legal. By contrast, for Arabs the extent of actual liability has very little if anything to do with the willfulness of one's acts; fault is largely displaced by repercussion, itself an amalgam of one's own situation and that of the injured party. And since one is responsible for the network one creates one is necessarily responsible for the harm one might engender. Blame, in the sense of fault, is seldom an issue in moral or legal discourse; responsibility, in the sense of repercussions for the network of others' obligations, carries its own moral imperative inasmuch as it renders worldly position vulnerable. As responsibility focuses on the situated person so, too, does the idea of justice. For justice in Islamic life resides not simply in each event but in the distribution of events over the range of obligational ties. Asked whether they would subscribe to a principle like that of American law that says that similar cases ought to be decided similarly, Islamic judges consistently respond by saying that since no two

Responsibility and Compensatory Justice in Arab Culture and Law 113 individuals are the same no two cases are really the same, and that even the same person in a later situation has accumulated a changed set of relationships. Regularity, they say, lies in using the same modes of reasoning to a conclusion, but two judges might reasonably draw different results even though they use similar forms of reasoning. Justice lies not in the similarity of result but in the assessment of the totality of one's personhood. That is why, as a sixteenth century Moroccan poet (see Scelles-Millie and Khelifa 1966 and also Khadduri 1984) could say: "When the times are just, one day is for me and one day is against me." Justice and responsibility are also, no doubt, connected to views about the nature of power. Briefly, it can be argued that in the Arab view power stems from an extremely wide range of sources. Where one man may use wealth as a vehicle for acquiring power another may use his capacity to capture the terms of discourse through his masterful rhetorical abilities. Power, being diverse, is very difficult to hold on to, and others may rapidly challenge existing pretensions from highly diverse bases. Arabs frequently use the word that is often translated as "rights" (haqq), but its implications are not the same as our idea of a right. Haqq means "obligation" and thus the distribution of bonds of indebtedness among sentient beings. It does not convey our sense of an indubitably supportable claim. One only has rights, the Arabs say, to the extent that one can enforce them. To be away from your "rights" — i.e., to be away from your land, the people who support your claims to that land — is to have no right. One does not, therefore, look to the law or to political figures to enforce an entitlement but to add themselves to your network of interlocked associates whose particular resources can then be brought to bear on your behalf. And it is in this context, finally, that we may be able to consider some of the features of compensatory justice as it exists in the contemporary Arab world. Modem commentators frequently say that in third world nations, where formal tort law was never elaborately developed, one is often confronted either with the breakdown of traditional approaches or their replacement by colonial models, and that the quest for an authentic, indigenous way of handling injury is, as in so many other domains of these nations' lives, a quest to recapture something that never really existed. Thus it will be said, for example, that collective liability, or at least collective support for individual liability, has broken down along with family solidarity and that given so fundamental a shift in social arrangements earlier remedies have little present effect. Similarly, it is said, the nature of wrongs has changed: where once there may have been individual injuries now the harm done may extend to hundreds or thousands of people, and if the wrongdoer is not of such

114 Lawrence Rosen stature as to be untouchable by ordinary litigants he may, by the scope of his 9

injurious act, be essentially insolvent (Tunc 1983). As they apply to the Arab world these points have quite mixed applicability. It is not, I think, true that individual litigants can rely less on familial ties than previously for support in their legal claims. Indeed, there is reason to believe that as the use of lawyers has increased and as litigation has expanded many people find themselves having to renew and rework their kin-based obligational bonds as they seek to use them for purposes of legal disputes. Just as political party politics has had the effect, in many areas, of reinforcing the modes of interpersonal negotiation, so, too, legal disputes have come to constitute an additional forum within which similar intensification of obligations becomes necessary. At the same time, the impact of the colonial experience on the legal systems of Arab countries has been profound. In many instances Islamic law has now been restricted to matters of personal status, and even here many countries have sharply modified traditional Islamic rules, procedures, and judicial organization. In the realm of tort law, for example, the two major influences have been the Ottoman Civil Code, known as the majalla, and French law. The Turkish code, which dates from the mid-nineteenth century, was clearly intended to be a modernist break with traditional law, and although it does not constitute a simple adoption of European models it moved sharply away from certain Islamic practices. The majalla remained close to Islamic practice in its deemphasis on formal considerations of fault or intent, concentrating instead on the distinction between direct and indirect damage. The majalla remains the basis for the law only in Jordan, but its effect in opening the way for systems of codification in the Middle East was extremely important. A far wider range of Arab countries has been influenced by French law. French tort law is often said to use the single-rule technique: the basic proposition, taken from article 1382 of the Napoleonic Code, says that every act that causes damage makes the person by whose fault the damage occurred liable for compensation. Some countries (Egypt, Syria) have retained the unlawfulness of the act as the only condition for liability; others (Tunisia) have gone further in introducing the concepts of negligence and moral damage which were quite absent from classical Islamic law. Indeed, under the influence of French law a significant number of Middle Eastern countries now uses fault along with illegality as a test for liability. In some instances these changes reflect attempts at the amalgamation of Islamic and western ideas. Iraq and Kuwait, in particular, are countries in which the codes, drawn up after the Second World War by the Egyptian jurist Abd al-Razzaq al-

Responsibility and Compensatory Justice in Arab Culture and Law 115 Sanhuri, reflected extensions of Islamic categories to previously unaffected domains. Thus the categories of usurpation (ghasb) and destruction of property came to include injuries that had previously been dealt with by bloodmoney payments. This was particularly significant because bloodmoney payments depended for their operation in no small part on the willingness of the injured party or his survivors to accept the compensation, whereas inclusion of these wrongs under the category of disruption to continuity of possession places far more control in the hands of the state. And it is this latter element — state control — which has, of course, become increasingly important. Not only have newly independent nations, like their colonial predecessors, seen fit to assert power through jurisdiction but they have also sought to do so through the promulgation of codes and compensatory schemes that include a much more elaborate list of specific torts than was previously susceptible to judicial consideration. Moreover, the large scale of injury, brought about by the use of modern technology and marketing, has necessitated greater state involvement. Although there have been few cases of massive injury in the Middle East to compare with the Bhopal disaster in India, the Chernobyl disaster in Russia, or the various environmental cases in Japan, Europe, and the United States, it is clear that in many less noted cases Arab governments have had no other choice than to compensate victims in the certain knowledge that little chance for recovery against the actual perpetrator would be possible. Indeed, in some countries, like Algeria, no-fault compensation schemes have been introduced to handle road accident cases in such a way that damages for personal injury are entirely eliminated from tort actions and it is possible to sue only for property damage (Tunc 1974). And yet the question remains whether, notwithstanding foreign influence, the actual course of adjudication for injuries will retain a distinctively Arab design in those countries that have not thoroughly abandoned the use of Islamic law. From what little we know of those countries that have attempted to reinstate Islamic law — Pakistan, Sudan, and Iran — it is primarily in the realms of criminal and family law that reform efforts have been directed. However, one could imagine that the course of tort law, too, could be affected by broader cultural changes that may be underway. These changes take us right back to the fundamental cultural aspects of Arab life with which we began. Briefly, the possibility exists that we are seeing for the first time in the Arab world a shift in such basic ideas as probability, causality, and intentionality. For example, in the past people spoke of events as either occurring or not occurring:

116 Lawrence Rosen there was no real discussion in terms of whether one could increase the likelihood of one or another result. Now, in agriculture, social arrangements, and political events one hears from time to time a language of probability, of events being assessed for the features that will affect the frequency of their occurrence. Just as in the West where, until the seventeenth century, probability referred to authority rather than evidence, so, too, in the Arab world it may be that these apparently probabilistic statements still have as their referent the opinion of reliable authorities rather than the existence of data which everyone could perceive equally. 11 If so the personalistic element will have remained strong and the evidentiary element weak notwithstanding some change in the terms of discussion. If, however, a real conceptual shift is occurring it is possible that a wide range of matters — the legal implications of compensatory justice not least among them — could be swept up in a reorientation of basic assumptions. Similarly, there are linguistic hints that causality is coming to be spoken of not just as strings of events made comprehensible because some reasoning creature intercedes at virtually every point in the chain but as sets of occurrences that possess ordering principles that are not dependent on human agency. Indeed, both intentionality and responsibility as fundamental cultural concepts could also be undergoing change, as the ability to discern intent from overt connections is disrupted by increased mobility and responsibility moves inexorably toward strict liability in the face of uncertain networks of obligation. Thus far, in my opinion, such fundamental cultural change has not, in fact, taken hold, and given the flexibility and malleability of Arab negotiable relations it may never occur. For just as chess became the consummately correct Muslim game — whereby mankind demonstrates its capacity for reasoning its way toward safety — and games of chance became theologically incorrect — because they suggest that man is at God's will but without the reasoning ability to make the best of the realm that has been assigned to his control (Rosenthal 1975: 167) — so, too, in life and law Arab culture demonstrates a highly personalistic vision of responsibility and justice which may, for a long time to come, affect the actual course of Islamic legal development. Notes

1.

An earlier version of this paper was delivered at the University of Virginia's Center for Advanced Study and was incorporated in a pamphlet published by

Responsibility and Compensatory Justice in Arab Culture and Law 117

2.

3. 4. 5. 6. 7.

8. 9. 10. 11.

the Center. I am grateful to Professor Ravindra Khare and his colleagues for their comments on the paper and for permission to reprint portions of iL Although the literature on Islamic legal history and doctrine is extensive there are very few studies of the actual operation of contemporary courts of law anywhere in the Arabic-speaking world, a fact that makes direct comparison between my own observations of the Islamic courts of Morocco and those of other places extremely difficult. Among the most useful discussions are Antoun (1980), Hill (1979), Messick (1983a, 1983b), and Mayer (1985). For a fuller account of the Moroccan data, see Rosen (198081 and 1989a). For a full analysis of the marketplace proper, see Geertz (1979). On the distinction between the logic of antecedents and the logic of consequence, see Dewey (1924). For analyses of legal reasoning, intentionality, and probability drawn in similar terms, see Rosen (1985 and 1989b). On the Islamic law of homicide, see Schacht (1964: 178-187), Kennett (1968), Tyan (1965), and El-Hakim (1971). Sheikh Abdulla Qadir Shaybat al-Hamd, an Arabian legal scholar, is here being quoted with approval by Sheikh Saleh Ibn Mohammad al-Laheidan (1976: 162), a member of the council of the Supreme Court of Saudi Arabia. The following discussion draws on examples found in Tyan (1926). See also Pacha (1944) and Ruxton (1916). For an anthropological approach to issues of strict liability see Moore (1972). For a selection of majalla principles as they relate to tort law, see Liebesny (1975: 218-20). This comparison between the history of the west and the modern Arab world is heightened by a reading of Hacking (1975) and Shapiro (1983).

References Antoun, Richard T. (1980). The Islamic court, the Islamic judge, and the accomodation of traditions: a Jordanian case study. International Journal of Middle East Studies 12:456-67.

118 Lawrence Rosen Burke, Kenneth (1962). The Rhetoric of Motives. Berkeley: University of California Press. Chehata, Chafik (1970). Islamic law. In International Encyclopedia of Comparative Law, vol. 2, R. David (ed.), 138-142. Paris: Mouton. Conrad, Joseph (1976[1896]>. An Outcast of the Islands. London: Penguin Books. Dewey, John (1924). Logical method and law. Cornell Law Quarterly 10: 17-27. El-Hakim, Jacques (1971). Le Dommage de Source Delictuelle en Droit Musulman. Paris: Librairie General de Droit et de Jurisprudence. Geertz, Clifford (1979). The bazaar. In Meaning and Order in Moroccan Society, C. Geertz, H. Geertz, and L. Rosen, 123-313. Cambridge: Cambridge University Press Hacking, Ian (1975). The Emergence of Probability. Cambridge: Cambridge University Press. Hill, Enid (1979). Mahkama! Studies in the Egyptian Legal System . London: Ithaca Press. Kennett, Austin (1968). Bedouin Justice. London: Frank Cass & Co. Khadduri, Majid (1984). The Islamic Conception of Justice. Baltimore: Johns Hopkins Press. al-Laheidan, Sheikh Saleh Ibn Mohammad (1976). Means of evidence in Islamic law. In The Effects of Islamic Legislation on Crime Prevention in Saudi Arabia. Rome: United Nations Social Defense Research Institute. Liebesny, Herbert (1975). The Law of the Near and Middle East. Albany: State University of New York Press. Limpens, Jean; Kruithof, Robert M.; and Meinertzhagen-Limpens, Anne (1983). Liability for one's own acts. International Encyclopedia of Comparative Law, vol. XI, Torts, ch. 2, s. 39. Tübingen: J.C.B. Mohr. Messick, Brinkley (1983a). Legal documents and the concept of "restricted literacy." International Journal of the Sociology of Language 4: 41-52. (1983b). Prosecution in Yemen: the introduction of the Niyaba. International Journal of Middle East Studies 15: 507-18. Mayer, Ann E., ed. (1985). Property, Social Structure and Law in the Modern Middle East. Albany: State University of New York Press. Mills, C. Wright (1940). The vocabulary of emotions. The American Journal of Sociology 5: 904-13. Moore, Sally Falk (1972). Legal liability and evolutionary interpretation: some aspects of strict liability, self-help and collective responsibility. In The

Responsibility and Compensatory Justice in Arab Culture and Law 119 Allocation of Responsibility, Max Gluckman (ed.), 51-108. Manchester: Manchester University Press. Pacha, A. Ibrahim (1944). De la Responsabiliti Pinole en Droit Islamique d'apris la Doctrine Hanafite. Paris: TJE.P.A.C. Rosen, Lawrence (1980-81). Equity and discretion in a modern Islamic legal system. Law and Society Review 15: 217-45. (1984). Bargaining For Reality: The Construction of Social Relations in a Muslim Community . Chicago: The University of Chicago Press. (1985). Intentionality and the concept of the person. In Criminal Justice (= NOMOS 17), J. RPennock and J. W. Chapman (eds.), 52-77. New York: New York University Press. (1989a). The Anthropology of Justice: Law as Culture in Islamic Society. Cambridge University Press. (1989b). Islamic "case-law" and the logic of consequence. In History and Power in the Study of Law, J.Collier and J.Stan, eds. Ithaca: Cornell University Press. Rosenthal, Franz (1975). Gambling In Islam. Leiden: Ε J.Brill. (1983). "Sweeter Than Hope": Complaint and Hope In Medieval Islam. Leiden: E.J. Brill. Ruxton, F.H. (1916). Maliki Law. London: Luzac & Company. Scelles-Millie, J. and Khelifa, B. (1966). Les Quatrains de Medjoub le Sarcastique: Poite Maghribine du XVIi Siicle. Paris: G.-P. Maisonneuve et Larose. Schacht, Joseph (1964). An Introduction to Islamic Law. Oxford: Clarendon Press. Shapiro, Barbara (1983). Probability and Certainty in the Seventeenth-Century. Princeton: Princeton University Press. Tunc, Andr6 (1974). La riforme du droit des accidents dela circulation — l'ordonnance Algdrienne du 30 janvier 1974. Rivue Internationale de Droit Compart, 1974:345-47. (1983). Introduction. International Encyclopedia of Comparative Law, v. XI, Torts ch. 1, s. 2. Paris: Mouton. Tyan, Emile (1965). Diya. In The Encyclopedia of Islam, B. Lewis, C. Pellat, and J. Schacht (eds.), vol. 2, 340-343. Leiden: Ε J . Brill. (1926). Le Systime de Responsibiliti Dilictuelle en Droit Musulman. Beirut: Imprimerie Catholique.

120 Lawrence Rosen Yamani, Ahmad Zaki (1968). Islamic Law and Contemporary Issues. Jidda: The Saudi Publishing House.

A Confusion of Signs: The Semiosis of Anthropological Ireland Herv6 Varenne

The direct word, as traditional stylistics understands it, encounters in its orientation toward the object only the resistance of the object itself (the impossibility of its being exhausted by a word, the impossibility of saying it all), but it does not encounter in its path toward the object the fundamental and richly varied opposition of another word (Bakhtin 1981:276). Who writes anthropology?1 Who, in particular writes the anthropology of Ireland — a genre framed by a multiplicity of "words," "voices" texts and ideologies all straining to offer the final statement of this "object" of study, the people who live in one of the islands off the West Coast of Europe? Inevitably, the cultural convention will have to be adopted and it will be written that the author of this paper is "I," not a person but rather a pronoun, a deictic moving a reader towards somebody ( — > Herv6 Varenne), a real, singular human being, administratively, and uniquely, designated by a sequence of letters placed in an appropriate typographic location.2 But how seriously can the grammatical implication be taken that "I," the designated material entity, is alone in writing this paper? Precisely because "I" is not alone, because this mode of referring to a complex act has been "adopted on the collective mode," to use L6vi-Strauss's criteria for what makes a text a myth (1971: 560), because identities are social and only become "individual" if the culture so constructs them, for all these reasons, I must, from now on, use the first person pronoun without quote marks. To do otherwise would be to choose meaninglessness, radical withdrawal. It would have no critical impact as the collectivity would never notice the absence. That we, as anthropologists, are limited in our freedom to construct alternate ethnographic texts, just as I am limited in my freedom not to use first person

122 Herv6 Varenne pronouns, is the main point of this paper. Ethnographic writing is not controlled by anthropologists but rather by the ensemble of texts that compete with it for authority and, in the process, create the anthropological voice in its specificity. We have begun to learn this, most starkly perhaps in the recent writing on the doing of ethnography in the United States (Marcus and Fisher 1986; Varenne 1986). We know that, in the long run, the voices of our "natives" have to be taken as more than mere "data" for analysis. They reveal themselves as practically powerful alternate analyses with which we struggle. What this may entail is well illustrated by the conditions one finds oneself in when one is working on Irish matters. In the United States, a large portion of the population has a keen interest in things Irish for any statement about the country or its people challenges their legitimacy. In Ireland, a casual look at the "Irish interest" section of even the most suburban of bookshops is enough to make one realize that one's voice will be but one among many and must earn the right to be heard by, in some way, inserting itself in a position recognizable by some future interpretant. To be heard, a voice must place itself.

The Voice of Anthropology

What is then this ethnographer's magic, by which he is able to evoke the real spirit of the natives, the true picture of tribal life? As usual, success can only be obtained by a patient and systematic application of a number of rules of common sense and well-known scientific principles... (Malinowski 1961 [1922]). My voice is that of an anthropologist, a type of social "scientist" As such, I lay claim to a particular knowledge of the way things are in Ireland. This claim is intelligible only if the audience recognizes the social scientific voice for what it is, whether or not it accepts its legitimacy. Anthropologists know certain things in certain ways. Like other scientists they are entitled to claim a more direct route to "reality" than others — philosophers, politicians, journalists, social critics, artists, novelists — who also try to transform their experience into a statement supposed to reveal something to some audience. To lay such a claim to "reality," social

The Semiosis of Anthropological Ireland 123 scientists must adopt a certain style of research and reporting which constitutes the external signs that "science" is being conducted. Anthropology, however, has resisted these stylistic and rhetorical constraints, sometimes in the name of humanism, but often also in the name of science itself — and this is what I do here. Anthropology argues that there are other routes to systematic knowledge than the ones which sociology and psychology have demarcated. Against their surveys and statistical correlations, anthropology offers "participant-observation," a technique of systematic critique and analysis of the kind of experiences that people have in the course of their everyday life. The strength of the anthropological argument lies in the fact that it makes sense to say that human action is intelligible only in the context of its occurrence. To understand any human behavior one must understand all the behaviors that occur around it — before it and after it, and in parallel to it — to the extent that it can be shown that these other behaviors somehow impinge on iL

The Stream of Anthropological

Consciousness

Peirce confronted [...] the problem of finding the locus, identity, unity, and continuity of the self among the rapidly changing "phenomena" of a stream of consciousness. His solution to the problem was to look in the sign-processes themselves for the answer. For Peirce, the locus, identity, and continuity of the self was not to be found in the individual organism [...]. It was, rather, an "outreaching identity," which connected the feelings, thoughts, and actions of one individual with those of others through the processes of semiotic communication. The self was thus both a product and an agent of semiotic communication, and therefore social and public (Singer 1984: 57). For reasons that will become clearer further on, I have just presented one of the not so simple-minded versions of anthropological work, the version which, most

124 Herv6 Varenne significantly, we present to the interested outsider, a prototypical reader, but also sometimes a curious informant, one who has been schooled, by teachers, journalists and the like, in the difference between literature and "science," and whom we are still trying to school into recognizing "science," or at least a voice of special authority, in a text that does not use the paraphernalia of objectivism. In this paper, however, the reader I am constructing is someone who is already quite well versed in anthropology itself and the perennial debates about the epistemology of our trade: How do we know? What must we take into account as we try to know? What is there to know anyway? The quote marks I earlier placed around words like "science" and "reality" are the obvious signals of my ultimate design. Another signal would be a reference to the set of works which, in recent years, have analyzed the rhetorical implications of traditional ethnographic writing and have called for new forms (Marcus 1982; Fabian 1983; Clifford and Marcus 1986; Marcus and Fisher 1986). Still, even if we are not addressing the unschooled reader, his ultimate presence, like the original presence of our informants, must be acknowledged. This, I feel, is something which recent work may not have emphasized enough as it has analyzed the conditions of ethnographic writing and called for "transformations both in the way ethnography is written, and in the ethnographer's awareness of for whom it is written" (Marcus and Fisher 1986:164). The experience of those who are not "us," anthropologists, as they are confronted by the many texts of authority which they, and we, encounter, is our experience also. If we are to take seriously what such recent commentaries on Peirce as Singer's Man's Glassy Essence (1984) have taught us, then we must draw the consequences of the fact that the "self' of the anthropologist, the self of anthropology itself, is a dialogical self. Indeed it is a "multilogical" self, one that exists at the intersection of, in the case of an anthropology of Ireland, a reader, all other writers of Ireland, and all experiences of Ireland or Irishness that one may have had, in the United States, in Ireland, and elsewhere. The ethnographer, as author, is not a free agent. What I would like to do here is sketch my education into Ireland and justify the anthropological tradition in the semiotic terms which, I hope to show, do help us to understand our collective situation. The need is not for another "new" anthropology — we are not gods that can create such beasts (only society can, or is it God?). The need, rather, is for a better understanding of the tradition that defines anthropology, so that we can fulfill the responsibility that is given to us to keep the tradition (and, possibly, in the process, to contribute to its evolution). In an Irish context this means that we must look at anthropological texts that present

The Semiosis of Anthropological Ireland 125 themselves as pointing to Ireland in the context of other texts of reference. We must step out of anthropology, become an observer of anthropology in order, eventually, to return to the traditional anthropological task which we have in fact never left.

Anthropological Signs to Its Self I had gone into anthropology in search of Otherness. Meeting it on an experiential level was a shock which caused me to begin fundamental reconceptualization about social and cultural categories (Rabinow 1977:29). Anthropology, we know, is rhetorically constructed as a task where a representative of an archetypical Us (the Ethnographer) encounters an archetypical Other (the Native) who is a puzzle to be solved. The anthropologist, as hero, takes himself to a place that is most exotic to a privileged audience (other anthropologists and those assimilated to them) and reports back what he has found. We have been told repeatedly, perhaps most forcefully by Fabian (1983), that this image is not representative of our situation as workers producing ourselves in the language we use, that is, in the texts which we write. For anthropology to produce proper knowledge (to be "scientific"?) it must not take a rhetorical stance that separates it from what it constructs as an Other. This is true as far as it goes. I take this position as my starting point It is also true that to highlight the way texts specifically marked for "anthropology" so construct themselves is not to free oneself from the underlying mechanisms. Fabian, who is so conscious of the need for an acknowledgement of coparticipation, still subtitles his book "How Anthropology Makes Its Object," thereby implying both that anthropologists have not been aware of their situation — which is probably unfair — and that the task of transformation can be authored by anthropologists as independent, indeed heroic, agents. In fact we must investigate more systematically the source of the linguistic tools which constitute us, particularly the rhetorical forms that we are accountable to use and the texts which compete with ours. I have dealt elsewhere with the issue of the implicit analytical power of rhetorical forms which are given to us by the American tradition.3 Here, I want to

126 Herv6Varenne deal with the "natives'" own voices as they speak about themselves (and us as anthropologists). What did these voices do to us? What are we to do to them? The Other which traditional anthropological writing constructs is, of course, a particular type of other. This Other is not able to understand the anthropological task — except perhaps as a vague "making them (Us) aware of who we are." The Other is also unknown to Us — or at least "badly" known (for we typically reject the authority of reports by missionaries, explorers, development experts, journalists, tourist guides, etc., that we have not reconstructed). The Other is the result of a "discovery" of some sort. On his return "home," the anthropologist confronts, at most, a generalized feeling of either repulsion or romantic attraction that We supposedly attach to the kind of people about whom we know nothing, except that they are "not at all like us." To deal with this, he first has made general statements about the fact that They are not savages (whether degenerate or innocent) and about the fact that other texts from Us are improper for various reasons having to do with the social and political position of the author. The anthropologist can then proceed to construct a picture that is unencumbered by any competing image that his readers may carry and use to evaluate what he is painting. The anthropologist is the privileged reporter of voices. He is not the hawker of a line in a marketplace of ideas. I will not escape here a construction of anthropology that may have been useful at the end of the 19th century but which anthropologists rejected in the practice of their craft, if not in their presentation of self. From the earliest they visited places, from Mexico (Redfield) to Indiana, U.S.A (Lynd), to Ireland (Warner, Arensberg and Kimball) that were not terrae incognitae, blanks on Western maps. Anthropologists have always insisted that they have something to say about such places, and they continue even as they are challenged by the other social sciences that have staked the field for themselves. This indeed is the position I have taken in my work on America and the position I will take in my work on Ireland. I will, inevitably, be reporting on voices and I will not pretend that these voices are speaking through me. At best I can signal their existence for those who want to go and listen and need an initial map to the marketplace and its various booths. The problem here is to understand how these other voices (who are not simply "voices of the Other") participate in constituting my own voice. Before the anthropologist starts working in the country, and after he has written his report, Ireland exists as a complex discourse, an ensemble of statements and counter-statements in a multitude of media from high literature to cheap novels, polemical historical analyses and political speeches, powerful movies and the snap

The Semiosis of Anthropological Ireland 127 shots which a tourist brings back home. Such images, cultural images in the strongest sense of the term, are not the unitary expression of a single author carefully crafting an univocal statement. They are the collective voice of a people. They are continually being produced and reproduced by a multitude of authors with various, often conflicting, interests who must take each other into account In this sense, this image has a life of its own. What must concern us is not the product of the historical process. It is the process of expression itself. As we produce our own text, we become part of this process. In the case of Ireland, such a statement is not a pro forma bow to current intellectual fashions in America. Minimally, anthropology is an intelligible task in Ireland. All major ethnographic texts have been sold, read, reviewed and criticized in dominant publications. Maximally, it is also the case that anthropology was invented in Ireland at about the same time as it was invented in the rest of Europe and for the same political purposes: curiosity, a demonstration of the superiority of the urban middle classes and, more explosively, for the legitimization of nationalistic movements, in France, Italy, Germany, the Balkans and Central Europe, which produced the Europe that bloodied itself in a series of major civil ("World") wars, the final ambers of which are still glowing in the north of Ireland.

Signs to Ireland Signs ofIreland Peter. Did you see an old woman going down the path? Patrick: I did not, but I saw a young girl and she had the walk of a queen. (Yeats [1902] 1952: 88) Ireland is not an island off the coast of Western Europe. It is the "Emerald Isle," "the land of Saints and Scholars." It has the "terrible beauty" of Cathleen ni Houlihan (unless it is Kathleen Ni Hoolihan), the old woman who walks like a young queen once the first son has been offered in blood sacrifice.4 It's color is green; its plant is the shamrock; its ancestors are St. Patrick, St. Brendan, Brian Born (and O'Connell, Parnell, Pearse). The Irish, one is told before leaving, "are wonderful people, you will just love them." "You will have a wonderful time!"

128 Herv6 Varenne They are to be depicted as old, drunk, and courageous in the face of misfortune for the tourists (Bord an Failte),5 and for school children learning about themselves in Peig: The autobiography ofPeig Sayers (1974). They are also to be depicted as young, eager, and well educated for potential investors.6 With each of these signs comes a complex discourse in which it is endlessly debated: - whether the Irish are, or are not really, Celt; - whether the true civilization of Ireland was eclipsed for 700 years and must now be recaptured (the official nationalist stance); - whether it is to be invented anew (Kiberd 1984); - whether contemporary Ireland is what has been constructed with the various elements which successive invasions and dominant administrations have brought to the shores of the island (Sean OTaolain 1980[1947]); - whether — as is the consensus in much contemporary social science — the state of Ireland must be understood as the product of the economic upheavals which followed the Famine, the Land Reform, Independence, the Economic Revival and entrance into the Common Market (Brown 1985; Peillon 1982; Wilson 1984). Whatever the cause, there is more agreement as to what was produced: A conservative, Catholic, small farmer who steadfastly works his own land, passes it on to one of his sons, and, in recent years, overeducates his other children to give them an edge as they emigrate to England or the United States.7 For an anthropologist, the initial temptation is to see all this as "data," a corpus of myth and symbols in relation to which the scientist takes the stance of the coroner at the beginning of an autopsy. This, it would seem, is the "native's point of view," to be taken into account, indeed to be reconstructed for an audience that knows little about it. In classical cultural anthropology, however, the native's point of view is made into something alien in which the anthropologist himself does not participate and to which he will be accountable when he releases his own statement for public consumption. When an anthropologist lands in a place like Ireland, the situation is different as several have found, sometimes in a very hard way.

The Semiosis of Anthropological Ireland 129 The issue is not simply an ethical one. It is an analytic one as it becomes obvious that anthropological concerns are native concerns: whether it is overprotective mothers or a peculiar land tenure system that produces late marrying men, such issues are Irish issues before they are American, scientific, anthropological issues. In their everyday life, in their newspapers and novels, in the intellectual work that is done for local consumption, the kind of reality which people in Ireland see is, essentially, the reality which the anthropologist also sees. This is not to say that the anthropologist will not be recognized as a peculiar voice within the general conversation. It is to say, rather, that powerful persons will understand his statements and recognize his interests. They will, directly and indirectly, suggest concerns to him that he will then appropriate and come to think of as his own. Whatever an anthropologist may say, it will eventually be taken as no more than the statement of another voice within the chorus that constitutes the mythical corpus of Ireland. Even if it is rejected as "wrong," "irrelevant," "romantic," "improper," or whatever, it will not be treated as alien. It will be intelligible.

The Anglo-Irish Signs of Ireland The great convulsion which society of all grades here as lately experienced, the failure of the [...], [...], [...], and a most unparalleled extent of emigration, together with [...], pauperizing [,..]-laws, grinding officials, and decimating [...], have broken up the very foundations of social intercourse, have swept away the established theories of political economists, and uprooted many of our longcherished opinions. (Wilde 1979[1852]) It is the case that the anthropology of Ireland as a location for a special mode of being European predates an official American anthropology which presents itself as having "discovered" the island sometimes in 1935. "Ireland," in fact, was first discovered by another group of "foreigners"9: Protestant Anglo-Irish who, starting in the middle of the 19th century, began to listen to the voices of their tenant farmers and wondered at the paradox of their own identity. Over 70 years they

130 Herv6 Varenne produced a body of work, the evolution of which seems to prefigure the evolution of ethnography itself as it evolved from an antiquarian concern with collecting dying myths and rituals (Wilde 1979[1852]), to trying to understand the exotic Irish as they understood themselves (Synge 1979[1907]), to experimenting with new ethnographic forms (Joyce 1961 [1921]).10 The difference was that, for them, Ireland was not "the Other," it was, ambiguously, "Us." Yeats, the prototypical figure of this generation, once wrote: "Gaelic is my national language, but it is not my mother tongue" (in Brown 1979: 157), thereby revealing his understanding of his position and the nature of this position. Yeats spoke and wrote English. He, and all the other figures of the various movements of which he was a part, had the political will of identifying with an Ireland "independent" of England. It was a basic premise of these movements that this Ireland had to be "revived." For Ireland was a cultural substance, an entity to which a rhetoric of life and death could be applied in non-metaphorical ways. The existence of this entity, the fact that it was "alive" was the only justification for the call to independence and indeed the only argument that could be used against England. And yet it seemed obvious that Ireland was "dead" in the political centers of the island, but was still gasping at its periphery, in those areas where "Gaelic" was still spoken. Men, and some very famous women, went to the West coast and its islands to learn about something that could not be found in the drawing rooms of Dublin where the movement itself was most manifest. What was sought was something that was not in Dublin, a Gaelic Other that had to be brought forward and translated for the benefit of an Us that was Irish.11 The Gaelic, language and culture, was not a curiosity. It was not a museum piece to be laid alongside "answers that others, guarding other sheep in other valleys, have given, and thus to include them in the consultable record of what man has said" (Geertz 1973: 30). Neither was it to be used to criticize the culture of turn of the century Dublin. The Gaelic was the reality of Ireland, if not of the selves of those who sought to revive it and, to a certain extent, succeeded in reviving it, albeit not all in the form they expected: one hundred years later, after more than 50 years of formal independence, English is still the mother tongue of the political forces in Ireland but Gaelic as been enshrined as the "Irish." The Constitution says "Irish is the national language..." Irish, not Gaelic. Re-written in modern political English, Yeats' statement would have to read "Irish is my national language, but it is not my mother tongue." And if, as Yeats and his contemporaries would have done, we equate language, culture and political legitimacy, Yeats' recognition that he did not speak what is now known as "Irish"

The Semiosis of Anthropological Ireland 131 places him in a peculiar position. The voices that became dominant after formal independence made it meaningful to say of him, and of all the Anglo-Irish of the period, that he was "English" and not "Irish." The Gaelic Other whom Yeats had tried to recapture had rebelled and fought an apparently successful political struggle to establish itself as indeed the "real" Irish Us that controls social identities on the island. The battle about the definition and political place of the Gaelic/Irish was first fought publicly in the famous riots which, in 1907, greeted the first staging of Synge's Playboy of the Western World. He presented himself as having stilled a personal artistic voice to let the Gaelic speak through the poor farmers of the West coast Most literary critics of his work have agreed with Yeats that he succeeded (Kiberd 1984; McDermott 1985). Many other critics have disagreed. For Synge's play was the focal point of a fundamental debate: What should these farmers say? And what does what they are presented as saying itself say to those who overhear their reported speech? What will they say? What can one fear that they will say? What are the best methods to be used to control the use of the speech by the significant Other of a nationalistic movement — the "colonial" power? What, in the case of Ireland, should England overhear of what is said in the island to its west? The voices of nationalism that came to power gave a consistent answer: they specifically rejected the voices that had spoken through the Anglo-Irish authors who most specifically claimed to represent the current conditions of Ireland. Institutionalized Ireland stilled Yeats, Synge, Joyce and their peers in the institutions it controlled. In the national schools the children of Dublin were steeped in a minor literature that was specifically "Irish" and remained essentially ignorant of the voices that made Ireland famous. The actual language which the West coast farmers spoke, "Irish-English," the dialect of English heavily influenced by Irish (Gaelic), was deliberately ignored in an attempt to revive "real" Irish. In the process Irish-English dies, Irish ceases to be a medium of every day exchange anywhere in the island, and basic North Atlantic English thrives.12 The paradox goes even further. Those who challenged Synge's right to speak for the small farmers of the West were by all accounts not themselves small farmers. For them, too, Ireland was the Other that was really Us. They, too, had not learned Irish from their mothers and the language they actually spoke was, by their own account, "foreign." As Joyce (1969[1916]: 189) has Stephen Dedalus reflect during a conversation with a British Dean of Studies: "The language we are speaking is his before it is mine." The dominant nationalistic voice thus presented itself as belonging to someone else — as indeed it was once the Act of Union with England

132 Hervd Varenne was declared illegitimate. The only solution to recapturing an Irish self passed through a reconstruction of the speaker as, if not literally a small farmer himself, at least the legitimate representative and defender of such farmers and, through them, of a truly Irish inheritance. In any event, there was something real and alive (though perhaps barely) on the island, something to own or to be owned by. As Brown notes: The writers of the Irish Literary Revival..., for all their individual quarrels and disagreements with Irish Nationalism and with its most vigorous representatives, the writers and thinkers of the Irish Ireland movement, had accepted the fundamental tenets of that faith. Ireland was an historic nation with its sources in pre-history. Gaelic civilization had been a glorious flowering of the Irish spirit, reborn in the Rising of 1916; and a modern imagination drinking at the well-springs of that world, even in translations of its literature, could enjoy a refreshment so revivifying that the splendors of that old spirituality might be born anew in a Europe grown weak and infertile in an old age of rationalism, science and economic utilitarianism (Brown 1979: 157). By 1916 G.W. Russell (AE) had a long tradition to lean upon as he sought "to discover...the ideas that lie at the root of [our] national character" (Russell 1982[1916]: 124). His definition of this character will be familiar to anthropologists: "In every nation which has been allowed free development, while it has qualities common to all humanity, it will be found that some one idea was predominant, and in its predominance regrouped about itself other ideas" (1982[1916]: 123). For Russell, as for all the theoreticians of European nationalisms, a population stably settled in a particular location develops a particular way of being human which becomes institutionalized if the political relations of this population to its neighbors allow. Let an oppressed people free, and its spirit will flower again. Such ideas, of course were not original to Russell or to the movement which he represents here, anymore than they were to Benedict or Mead (or Dewey who prefigured them in the United States). The roots of the idea are much deeper in European social thought and may have been more fully explored, in their political consequences as well as their scientific plausibility, by the German philosophers of the 19th century.

The Semiosis of Anthropological Ireland 133 In the 1920s, once formal emancipation had been won, the Ireland that emerged "in fact" had more in common with the Ireland in which the nationalists had been borne (the Ireland of petty English provincial bureaucrats) than it had with the heroic Ireland for which Pearse had died. Or so at least it seemed to the sensitive authors of the period, such men as O'Faolain or Kavanagh. As Brown says of them: "Knitted with the common life of Ireland, the petit-bourgeois life of the towns, closer to the actual life of the small farms of rural Ireland than Yeats ever was, they knew of the drab, unadventurous, unromantic, puritanically Catholic, English-speaking, economically prudent reality" (Brown 1979: 157). Another voice, another "reality," but certainly not the voice of power as de Valera, who had lived for the other Ireland, established his place in the country and, indeed, the world.

Anthropological Signs of Ireland Looking out from the right-hand window, there below was the bare hungry countryside of the Rosses and Gweedore; Bloody Foreland yonder and Tory Island far away out, swimming like a great ship where the sky dips into the sea. Looking out of the door, you could see the West of County Galway with a good portion of the rocks of Connemara, Aranmore in the ocean out from from you with the small bright houses of Kilronan, clear and visible, if your eyesight were good and the Summer had come. From the window on the left you could see the Great Blasket, bare and forbidding as a horrible otherworldly eel, lying languidly on the wavetops; over yonder was Dingle with its houses close together... I have never heard it said that there was any house as well situated as this on the face of the earth (O'Brien [1941] 1986: 21). Let us look at what is generally known in the discipline as "the anthropology of Ireland." It is rather easy to caricature this work and perhaps I should refrain from

134 Herv6 Varenne it. Caricature is always unfair and never descriptive. It is also powerfully evocative and, to the extent that my intent here is not to review in detail the anthropological literature, the above quote from Flann O'Brien's famous satire of another kind of work on Ireland is not inappropriate. Caricaturally, then, what anthropologists know of Ireland, they know of "a place called Corkadoragha in the townland named Lisnabrawshkeen." Of course, anything that is mentioned in this literature "concerns only Corkadoragha and it is not be understood that any reference is intended to the Gaeltacht areas in general: Corkadoragha is a distinctive place and the people who live there are without compare" (O'Brien 1973[1941]: 7). Anthropologists have conducted their participant-observation in the house where Bonaparte O'Coonassa was bom, a house that has "two windows with a door between them." The passage I use as an epigraph to this part of the paper is a description of the view from O'Coonassa's house. Through the right hand window one can see the field sites of Fox (1978) and Taylor (1980,1981) in Donegal. Through the door one can see the island where Messenger (1969) did his work. From the window to the left one can see Scheper-Hughes (1979). The house, clearly, is facing West, with Connemara and Donegal on its North and Kerry on its South. This locates it in the County Clare of Arensberg and Kimball (1940), Brody (1973), Cresswell (1969), the Co. Clare that remains the touchstone of Irishness.13 O'Brien wrote his satire as a caricature of the Ireland of Gaelic (Irish) enthusiasts. It was a satire of the Ireland that a generation of scholars and revolutionaries, and then a generation of politicians had constructed for themselves, first, and, second, as an external justification for an act of power. That it could also have been written about the Ireland of anthropologists is no simple coincidence. The Ireland that was to be useful in Dublin and London was the same Ireland that was useful in the hallowed halls of Boston, New York, Chicago or San Francisco. One might say that the purveyors of anthropological Ireland were co-opted both by the politicians of Dublin, and by those of Queens or the South Side of Boston. W. Lloyd Warner describes as follows the process which led him to chose Co. Clare for Arensberg: In the summer of 1931 I made a preliminary survey of the twenty-six counties of the Irish Free State and chose County Clare as most likely to fulfill our needs — a county in which there was a blending of older Gaelic and modern British

The Semiosis of Anthropological Ireland 135 influences, and one that was neither entirely Gaelic nor entirely English in speech (in Arensberg and Kimball 1940: ix). Who, but nationalistic politicians, could have told Warner that he should pay attention to "Gaelic influences" in an anthropological study? Certainly, it is not chance that focused the attention of Warner on Clare. But it is not sure either that the only factors which operated are theoretical ones (structural-functionalism and the accompanying need for bounded communities). The recent critical literature has made much of these factors (Brody 1973; Gibbon 1973; Wilson 1984) and, indeed, it has become a kind of leitmotiv. But anthropologists keep returning to Lisnabrawshkeen. There is something else that focuses the anthropological imagination of, by now, three generations of scholars and it must be found in their interaction with the natives, of course, but perhaps not those natives that appear in the public work. In the 1930s certainly, all political signs that might be seen in Dublin pointed to the West of the island, and, in the West, to small farmers (rather than large farmers or towns people) as the place where Ireland was to be found. 1932 had seen the accession to power of de Valera, the American born, Dublin educated, native English speaking, hero of the 1916 Rising who had been, since 1919, member of the Irish parliament for, precisely, Co. Clare. 14 De Valera's party, Fianna Fail, presented itself as the populist supporter of the small farmer who owns his land, lives a precisely "parochial" life in the confines of a parish that is the center of his social life, and eventually passes the farm freely to his appointed son. This small farmer had the absolute reality of a cultural model. To challenge it was to be both blind and a traitor. For the nationalists of the time, on both sides of the Atlantic, the small farmer of anthropologists, de Valera and, behind him, of Yeats, is what Ireland was, in the strongest sense of the verb "to be," and what it would be, once it had recaptured institutionally what it had always been. Co. Clare was the present of Ireland. It was also its past and, the revolutionaries turned statesmen hoped, its future. Aiensberg and Kimball, from a certain point of view, "fell for" something which, at about the same time, various outsiders like Sean OTFaolain were describing as a dangerous myth. For all we know, Arensberg and Kimball fell for it with their eyes wide open, perhaps hoping to contribute to the international justification for Irish independence: that Ireland was, "really," a live Gaelic, and above all not British culture. Either way they expressed something that was true to the Irish imagination of the time, if not to its pragmatic "reality." They gave something

136 Herv6 Varenne that can now easily be read as a version of the national myth of Ireland as an unchanging traditional, rural and Catholic society that is on the verge of a great transformation that is a revival.15 They powerfully captured something that spoke to and reflected a native point of view. They told us Ireland as Ireland was telling itself. That they may have missed other voices may not be surprising when we now realize how powerful were the forces that were stilling the modernist statements of Joyce, the socialist writings of Connolly or O'Casey, the anguish of OTaolain or Kavanagh. The discourse that framed Aiensberg and Kimball is a discourse that is independent of them though, as some have noted (Scheper-Hughes 1979:41), what they said could be incorporated into it. Anthropologists can be shown to be blinded by its verisimilitude so that they are never aware of the sign which led them down a particular road. They rarely are in a position to ask: were there other roads? where would they have led me? Clearly, the analytic task must now be to produce an account of the signs that point the way and of the mechanisms which blind us to the signs and to the roads not taken. But we cannot either expect any such account to make the signs vanish. Such signs and symbols as I am talking about now cannot be "deconstructed" by an individual, or even a group, as long they have political currency within a society. While they have currency any contribution that can be related to the dominant discourse will be so related. Any contribution to our knowledge of Ireland is necessarily part of this discourse, even if, or indeed particularly if, it presents itself as a critique of the discourse. Any one who is interested in the discourse will want to know the critique and then either dismiss this critique, or else modify the discourse.

Signs in Ireland

I was living in a three-bedroomed house in Meadowbrook. I have three children. I needed extra space. You can think about the problem in two ways. You say "well, I either extend my house and put up an extra bedroom, or move." It was less bother to move, and anyway moving was better for tax and mortgage reasons. So you look around... But it's an awful pity that people do look at a

The Semiosis of Anthropological Ireland 137 particular place as being a particular class. I can see outsiders saying "oh well, such and such lives in Oakpark..." (from an interview with a welfare administrator). Anthropologists, once they have presented themselves as analysts of the Irish way of life, are thus pointed in a particular way. In the suburbs of Dublin, when I mentioned that I was "interested in Irish life," I was told several times "if you want to see the real Ireland, you must go into the country, to the West." Where else could they send me, if they wanted to make sense? My interest, however, was not in what was being pointed at but in the people who were doing the pointing. How do they do the pointing, what do they speak about when they are not pointing at Ireland, what are the tools at their disposal and whom did they get them from? 16 Most foreigners will now enter the geographical space (Ireland?) that is supposed to be one of the referents of all these discourses in Shannon airport I crossed the frontier in Rosslare, where the ferry from France lands and I drove on to Dublin. In the distance were low cliffs liberally sprinkled with white salt boxes: tourist cottages. A parking lot. A hotel in the international Howard Johnson style. The road to Dublin. More tourist cottages. "Look, lode, a thatched cottage!" Endless "Bed and Breakfast" signs. A huge truck passes by at a roar. A litter of garbage around a few derelict house trailers: the resting place of 'travellers' as the old 'tinkers,' I will soon learn, are now known. Miles of gently rolling fields. "Dublin, 10 miles," the sign by the road says. "Dual carriage way." A construction area with a sign specifying that the improvements are funded by Common Market development funds. The traffic picks up. Red lights, packs of cars. "Dunnes stores, next left" proclaims a huge billboard. Further on, Quinnsworth, Superquinn, Roches Stores: the local and international purveyors of mass consumable goods. Arrows pointing to various suburbs: Dundrum, Rathfarnham, Dun Laoghaire, Blackrock. Gas stations: Esso, Shell, Texaco. Dublin below the tortuous road which, I will find out later, the civil engineers call the "Southeastern bypass": a sea of housing developments, "estates" as they are called in the British Isles, thousands of houses, all the same it seems at first — it will take me quite a while to see the differences that the people do see. At the bottom of the basin, where I already suspect that the center of the city lies, an indistinct blur broken by the spires of a few churches and some modern buildings. At the other edge of the basin, on the horizon, mysterious towers — the infamous

138 Herv6 Varenne high rise "council flats" (low income housing) of Ballymun, the product of 1960s slum clearance enthusiasm now turned to slums themselves. Later, in the center of Dublin, one will soon recognize the sights to which postcards, picture book renderings and literary evocations point: the Georgian squares, the Liffey and its quays, O'Connell street and the General Post Office, the Martello tower on Sandymount strand. In Dun Laoghaire, an actual street sign points to "James Joyce's Tower," the other Martello tower given eternal fame in the opening pages of Ulysses. There a museum teaches one to see Dublin, not as it may be now, but as it was for Joyce. "7 Eccles Street," Bloom's home, has been demolished to make room for a hospital: *'Do not look at the hospital',* says the subtext. "The door to the house that has been preserved in a nearby pub," says the text: ""Go there to place yourself in Leopold Bloom's shoes,' is a subtext acted out every day by tourists who religiously follow the route Bloom took on June 16,1904. Thus the best ethnographer of daily life in Dublin is made into an antiquarian, another purveyor of a "real Ireland" that is only a small part of the Ireland one can see and touch. The Joyce museum will indeed direct one to experiences that one can have in Dublin. The Dublin it constructs is no more imaginary than the Ireland that Yeats, de Valera, or Arensberg and Kimball constructed. What the museum will not do is what one can think a new James Joyce might do, and that is make one see, as only a great novelist can, those aspects of everyday Dublin that may now be most real to a sensitive adolescent — and that would probably include... McDonald's (not the local pub of course, but the international "American" fast food chain). Be that as it may, one can move through Dublin without following the signs planted by the Joycean tradition. One may not know they exist, or one may just not be in the mood to construct the appropriate Dublin (after all the children are restless, you are tired from all the driving and you wonder whether you will find the obscure street in Drumcondra you are looking for). Then what you may experience above all is a traffic jam, double decker buses belching black smoke, indistinct crowds and heavy coal smog in the air, hamburger joints and their attendant trash, boarded up shells of half demolished houses, all the signs of modernity gone slightly sour. Have you really come all this way to something that, you shudder to think, could just as well be Trenton, New Jersey? Three months later, comfortably installed in one "very nice" house in one of the best estates of your suburb (or so is everyone telling you — particularly the voices whom you are most willing to trust), you realize that all the Irelands to which

The Semiosis of Anthropological Ireland 139 point the many signs you have encountered can indeed be found on that big island to the west of Great Britain. Even in the most "modern" of suburbs there are echoes of Arensberg and Kimball, Scheper-Hughes, Brody, and also of course, of all those who have written about modernity. Co. Clare is right here, and so is London. Everywhere there is the Catholic Church. Everywhere, too, there is talk of matters sexual. AIDS and abortion are major topics. Strong families, broken homes, migration, drug addiction, delight in conversation, the pub and the singing of ballads (and of American "golden oldies"), saints of all persuasions, and even some scholars. Who is going to inherit the house? Who is to take care of the parents? How will the mortgage be paid? Should we send our children to private schools? Will the county council really let some travellers settle in the field across the road from us and lower our property values? How can we stop them? The signs to all Irelands are there, and so are experiences that, as seen from the point of view framed by the signs, justify their existence: if a widow can be found who talks about her 21-year-old, unemployed son as if he were 13, and if this son, living calmly with her, for all practical purposes, behaves like a 13-year-old, then there is justification in the signs that make of "over protective mothers" something which an ethnography of Ireland must deal with (even if only to argue, generally on dubious methodological grounds, that "things are changing as more people move away from the country and come into contact with modern attitudes").

The Semiosis of Ireland Writing Ireland in America Que ce soit dans l'ordre du discours parl£ ou du discours 6crit, aucun 616ment ne peut fonctionner comme signe sans renvoyer ä un autre 616ment qui lui-m&me n'est pas simplement prdsent... Cet enchainement, ce tissu, est le texte qui ne se produit que dans la tranformation d'un autre texte (Derrida in Hollier 1972: 583). Writing Ireland, I stated at the outset, is not a lonely task. Writing the anthropology of a culture that is one of those from which anthropology has been generated, is a dialogical, multilogical, social act. All descriptive texts cany, or are

140 Herv6 Varenne given to carry, an ideological baggage. They place themselves, they are placed, they use and they are used, inside and outside the domains to which they may claim to belong. "Anthropology," wrote J.P. Dumont (1986: 363), "must be literature;" it must produce carefully crafted texts. The trick now, as he explains, is not to deconstruct its mechanisms so as to lay the pieces on the table. We have excellent examples of the yield of such work (Clifford 1982; Fabian 1983). The challenge, and I expand on Dumont's point, is to construct something that will, at the same time, be recognizable as ethnography and aware of itself as a social act, a voice participating in a multilogue. There are no better ethnographies of a provincial capital then the three Joyce wrote about Dublin, but nothing would be sillier than an attempt to write Bloom Redux (except perhaps an attempt to write about Dublin as if Ulysses had never been written). An anthropological text, by virtue of its place within the social sciences, claims (or is made to claim) a voice of authority in the domain of the "real" (though perhaps not "the truth" — a term from which even theologians recoil). Science worries appearances (symptoms, facts, data) until they yield the real, structuring mechanisms that can then be written up, inscribed on our knowledge. Anthropologists, as scientists, have learned, and taught each other, that, in the midsts of the cacophony of signs all claiming to reveal something of a living culture, the "real Ireland" will not stand up. Neither can it be "dug up," in a process that would combine archaeology with private eye sleuthing: it is not a matter of "scratching the surface" or "scooping up the dirt" to reveal the inner substance. The anthropologist, in his attempt to bring something back from his sojourn in the island that will interest a member of his privileged audience, or even of his peripheral audiences, cannot rely upon the flash of insight that will suddenly make Ireland intelligible. Neither can he rely on a patient work of "data collection," of measurements and systematic observations — however necessary it is that these be also conducted. The more he does so, and the more careful he is in contextualizing what he has seen, then the less he will be sure that he has something, some THING that would be Ireland. Eventually, if he does his ethnographical task well, he will, like his informants, find himself bound tight in a web of signifying relationships linking signs with experiences, grasping one strand, and then another as he attempts to navigate the waters of his own everyday life in the country. What I would write in Ireland, in answer to a year of editorials in the Irish Times and direct or indirect conversations with local intellectuals is not what I would write — am here writing — in the United States, in answers to years of participation in American anthropology.

The Semiosis of Anthropological Ireland 141 What, then, am I doing? Hie answer to this question is double. I have to answer it in terms of the two roles which I must take, first as the producer of an anthropological text, and, second, as the observer of this process of production in the midst of the various social groups that can claim me (from my informants eager to know what I wrote about them to a skeptical reader wondering whether he should read further). Until now, this paper has mostly been written from the point of the observer, observer of earlier observers, observer of various participants, observer of my own activity. Underlying all this however is my conviction that the observer stance cannot be institutionalized. Eventually the deconstruction of a text becomes the construction of another text As human beings we can be anything but silent. The act of laying bare the signs of Ireland, the mechanisms of their production and the range of their power is a creative, constructing act which, for an anthropologist, must be anthropological.

Towards a Modern Anthropology ofIreland Culture...is that complex whole which includes knowledge, belief, art, law, moral, custom, and any other capabilities and habits acquired by man as a member of society (Tylor 1871: 1). From the point of view presented earlier, the institutional voice of anthropology is that of the scientific observer, that of the observer who has learned not to privilege one set of symptoms or signs over any other. The Ireland of American anthropologists will have to be built with the bits and pieces, if not shreds and patches, that come to the observer loosely bundled under the index entry "Ireland." Kathleen ni Hoolihan and Celtic nationalism, Superquinn and the Common Market, thatched cottages and the tourist office, Yeats, de Valera, pastoralism and its critics, Charlie Haughey and Margaret Thatcher, all will have, somehow to be dealt with. To isolate any is to collapse anthropology into, at one extreme, bad sociology as one contravenes the principle of contextualization which drives the discipline or, at the other extreme, bad literature as one contravenes the principle of observation. The Ireland we are shown is not a matter of simple signs but rather of complex interpretations, traditions, that constitute the reality of Ireland for major

142 Hervi Varenne groups in and out of the country. The Ireland of Irish politicians is one where large majorities of voters defeat referendums proposing that divorce be allowed. The Ireland of Irish feminists is one where deserted wives have no legal resources and abortion counseling is forbidden. For the Pope, Ireland is the last Catholic country in Western Europe. For the Prime Minister of the United Kingdom (or is it "England"?), Ireland is a possible strategic risk and a drain on the treasury that may or may not be compensated by the markets it offers to British industries. The wife of the plumber struggling to pay the mortgage on a house they can barely afford constructs an Ireland that is not the Ireland of the traveller woman who begs at her door. Neither is it the Ireland of the managing director of Digital Business Computers as he flies home from a visit to corporate headquarters in Connecticut. There is more, as the history of the earlier anthropologies of Ireland can teach us. The Ireland that was constructed by the first nationalists, an Ireland of glorious Celts and noble islanders, was later challenged by the Ireland of disappointed patriots, with its landscape of conservative small towns systematically putting out all creative sparks. The Ireland of the first anthropologists, the Ireland of local structural-functionalism, was challenged by the economic functionalism of much recent work. None of that work can be ignored. It is both "data" for future work and the frame which will limit what it can accomplish. To the extent that this is true, that is to the extent that I have data from which an analyses of the differences between these Ireland could be written, then a modern ethnography of Ireland must be an ethnography of these voices and of their echoes in everyday life. It must be a modem version of what was called the "ethnography of speaking." We must report the formal qualities of various voices. We must also report the political relationships between these voices. As I began to show earlier, what is now most interesting in the work of the early discoverers of Ireland was not only the quality of their rendering of forgotten peasant voices, it was the political use to which these renderings were put. There is revealed what Bakhtin referred to as the centripetal force of language, what anthropologists generally refer to as "culture" — that is the "whole" that gives historical import to the part — the national language which places the various voices that one can hear.18 Ireland, I would say in my anthropological voice, is not "behind" the voices one can hear there. Ireland is not hidden by these voices. It is revealed in them. Anthropologists cannot take upon themselves the political tasks of deciding whether Ireland is "really" Celtic, Catholic or modern. Even less can they take explicit sides within such native controversies about whether it would have been "better" for the country to follow Synge, Connolly and Joyce in forging a national

The Semiosis of Anthropological Ireland 143 spirit than to follow Yeats (Kiberd 1984). In other words the discipline cannot take upon itself the settling of controversies and hypotheses that are its "data," the signs and symptoms which it must use in the construction of its own analyses. What it can do is show how Ireland is, and in fact uniquely so, a place where Celtism is institutionalized in a Catholic context in a modern environment. Anthropologists do not have to decide whether Ireland is or is not a country of dying small farmers, unemployed blue collar workers, or petty bureaucrats. Anthropologists can show how these people are made to deal with each other, in the Dail, and in the imagination of the country and of those who like to think about it. Terms like "computers," "feminists," "mortgages," "Popes," "structuralfunctionalism" have currency in Irish discourse, whether intellectual or not They provide a foundation that must shape the building that will rise above it, however baroque its decorations.

Anthropology in Dialogue By maintaining that all thinking is by means of signs, that it takes the form of an inner dialogue structurally similar to and continuous with the outer dialogue of conversations with others, Peirce was able to develop a concept of personal identity that is not confined to the individual organism but that extends as far as his social and cultural consciousness and his circle of society. The individual's consciousness of self and others is a "double consciousness," in which the consciousness of others may precede the consciousness of self, and, in any case, develops with the individual's interactions with others and with the world as selves emerge from these interactions (Singer 1984: 83). A building is not only the product of an interaction between an architect and the technology and economy that give him the means to express his authorship, and that limit what he can create. A building, is, also, an interaction between itself, as

144 Herv6 Varenne constructed, and what people will make of it. As Peirce said, and I paraphrase slightly, "the interpretant [of a building, a word, a discourse, an anthropological monograph] is the future memory of this cognition [building, word, discourse, monograph], his future self, or another person he addresses, or a sentence he writes or a child he gets" (CP 7.591 in Singer 1984: 56). A semiotic construction is not simply an object that reveals its determining historical production, it is also a subject that is already failing to contain its indetermining use in future objects. The point here, for one who is getting ready to embark upon the construction of a text on "Ireland," is that it is not enough to understand the constraints under which one works. One must also understand the constraints under which the work will be put, once it has escaped the writer. After twenty years of deconstruction of the claim to special, "scientific," knowledge, one must know that the signs that mark a work as "anthropological" are "arbitrary," that is, they reveal something else than the signs that would mark the work as, for example, the work of a government agency "selling" Ireland to American companies. Each work can use the other but, to the extent that they cannot "be" the other without ceasing to be themselves, the most they can do is point at the other to help the future reader not be taken by the apparent authority of the text. Neither text is "closer" to the reality of Ireland. Yet both texts construct an Irish experience. The very arbitrariness of the anthropological means that it should be constructed in such a way as to point to itself as, precisely, "anthropological" and, indeed, "scientific." To deny the possibility of science to an activity that has become aware of itself as a semiosis, a process of signifying the world, is to prove oneself still the prisoner of the old objectivism that assumes that the world reveals itself to he who can free himself from the myths of tradition. It is only when the social conditions that have produced anthropology have ceased to be active that the arbitrariness of the activity will become relevant. Anthropologists do not control anthropology. The ones who fund us, who buy us and read us, do produce us. If we, in fact, want to say something to them that they do not already know, we must catch them by cloaking our statements under the signs that will, hopefully, be recognized as "anthropology," and, in this process, become, unwittingly perhaps, the natives of a discipline that may not really be "our own." In the mean time, as De Valera said on St. Patrick's Day in 1943 : Bail έ Dhia oraibh agus bail go gcuire S6 ar an obair atä romhainn. Go gcumhdai Dia sinn agus gur fiü sinn choiche, mar näisitln, na txolocai a thugh Pädraig chugainn. Go dtuga an

The Semiosis of Anthropological Ireland 145 tUilechumhachtach, A thug slän sinn goo dti seo όη anachain is όη mi-ädh atä ar oiread sin näisiun eile de bharr an chogaidh seo, scäth agus didean duinn go dti an derireadh, agus go ndeonai S6 gur flu sinn cion uasal a dhdanamh sa saol nua atä romhainn.19

Notes

1.

I must acknowledge here the contribution R. P. McDermott, Milton Singer, Larry Taylor and Tom Wilson have made to this paper. I apologize for the many ways in which I have not taken their suggestions into account. "I" am responsible. Very important to such a paper are also the people I met in Ballinteer. For their welcome, my family and I thank them from the bottom of our heart As they told us many time, I tell them "God bless!"

2.

For more on the place of pronouns in semiotic analysis, see some of the other papers in this volume, particularly the discussion by Singer. In my work on rhetoric in American schools, I investigated the relationship between the various tellings of the school and the construction of social relations through the differentiated use of pronouns (1978, 1982, 1983, 1984b). I also showed the theoretical impact of the anthropological use of third person plural pronouns to refer to our objects — as in "the Irish, they ..." (Varenne 1984a, 1986)). The two lines I used as an epigraph to this part of the paper are the last two in a play which some credit with inspiring the imagination if not the actuality of the 1916 insurrection and the subsequent struggles (Thompson 1982[1967]). Ireland's "terrible beauty" was first so described by Yeats also in a poem about the 1916 insurrection:

3.

4.

I write it out in a verse — MacDonagh and MacBride Connolly and Pearse Now and in time to be Wherever green is worn, Are changed, changed utterly, A terrible beauty is bom. (Yeats 1962[1921]: 87)

146 Herv6 Varenne

5. 6.

(Note that these quotes are presented less as "proof' of an analysis than as an invocation of ancestors through the recitation of myths). For an analysis of the structure of current tourist oriented depictions of Ireland, see Torode (1984). As may be seen in the advertisement which the Industrial Development Authority places regularly in European and American newspapers. The I.D.A. is generally considered by the Irish press the most successful agency of the Irish government Some vocal critics disagree. The I.D.A., however, is very good at advertising its own work. It describes itself in the following manner: The Industrial Development Authority (IDA) is the organisation responsible for promoting industrial development in Ireland. It encourages the development of manufacturing and international service industries both home and overseas, and administers the financial incentives available to industry in Ireland (from a pamphlet).

7.

An analysis of this rhetoric remains to be conducted. Significantly, the historical "facts" (poverty, migration, stagnation, development) are less in dispute than the interpretations, that is, hypotheses as to the "causes" of all these facts in various periods. Only recently has there been serious debates about the "facts" themselves — particularly the relevance of the small farmer and his fate to an understanding of Irish history. As for the Protestants, their place is, to say the least, ambiguous. In the mythical history they are the "foreigners," the "English" not to be considered when discussing "Irishness."

8.

Irish ethnography was shaken by two causes calibres, books by Messenger (1969) and Scheper-Hughes (1979) which were, we are told, badly received in the respective locales where Messenger and Scheper-Hughes worked. I do not know that the people themselves actually spoke in their own voices either for or against these books. Intellectuals and journalists in Dublin, and anthropologists (American mostly), have, however, spoken extensively and often very harshly (Kane 1982).

9.

The issue of who is or is not a foreigner in Ireland is of course a profoundly political one. The significant Anglo-Irish about whom I am writing here thought of themselves as "Irish." Most American anthropologists of Ireland

The Semiosis of Anthropological Ireland 147

10.

11.

12.

13.

14.

are often classifiable as "Irish-Americans." In Ireland, both groups now find it difficult to make much of this "Irishness." Joyce, of course, was Catholic. His position, as one who withdrew from Ireland at the height of the political expression of Irish independence, is ambiguous. His fate, in Ireland, is not so far removed from the fate of the Anglo-Irish. Most, in this generation of discoverers of Ireland, were Protestant They did not however look for Irishness in the Northeastern counties which, even before partition, seem to have been treated as the foreign land which they became. The Gaelic, and the Irish, was Catholic and, in the collective imagination of those who continued to make Ireland, it has remained Catholic. That Ireland could be as much Protestant as it is Catholic is something that cannot, apparently, quite be handled. The situation, of course is more complex and the field is excellent hunting ground for those interested in linguistic variation and change (Trudgill 1984; Harris 1985). The issue, however, is not purely linguistic. There is no way anymore to measure "objectively" how many people in Ireland speak Irish as it has become a specifically political issue to which major economic resources are attached. Whatever is spoken in the official Gaeltacht (regions were Irish is officially spoken), it cannot be English without the central government loosing legitimacy and the local settlements loosing significant tax privileges. As for the Irish-English which Synge attempted to record, it was refused legitimacy. What the Irish "really" speak in Ireland, what is usually labelled "Hiberno-Irish," is as much a mystery as ever and no sociolinguistic map of regional and class dialect will affect this. Such descriptions will remain academic curiosities, as long at least as the current culture remains in power. Some work has been done at some distance east from the coast, particularly in the North (Bax 1943, Harris 1972, Glassie 1982, Vincent 1983) but the emphasis remains rural. For the best recent review see Wilson (1984). There is almost nothing on Dublin, except for the work by Humphreys (1966) which had little impact on subsequent research. There are signs that things are changing (Curtin and Wilson in press; Kane 1986). De Valera's childhood constituency was South East Dublin where the AngloIrish and the upwardly mobile Catholics lived and still live. He never tried to get elected there and, at the beginning of his political career, conducted a

148 Herv6 Varenne search for a likely constituency, a search which took him both to Co. Clare and to the Falls in Belfast where he failed to get elected. 15. The best version of this myth was rendered by de Valera, in a famous radio address on St. Patrick's day 1943. For a discussion of various other versions of this myth, those told by Flaherty and Arensberg and Kimball, see Wilson (1987). It is still the case that a reading of Arensberg and Kimball as a myth is a reading produced by a collective appropriation of the authorial text (LiviStrauss 1971: 560). It is not to be assigned to the authors as agent. Arensberg and Kimball knew that they were not writing about all Irish people, or even all Irish farmers. The first generation of users of their text often forgot that. The next generation, of critics mostly, were not dealing anymore with Arensberg and Kimball as such, as they also knew, they were dealing with a tradition of interpretation, a culture which, through their criticisms, they established as a cultural fact 16. These questions summarize the main goals of the research I conducted in Ballinteer, a southern suburb of Dublin. I spent ten months there with my family. The core data were the result of "participant observation," more or less focussed interviews, tape recordings of conversations and public representations in newspapers, television and governmental publications. 17. I place asterisks before a statement to indicate a statement that is not attested in my fieldwork. 18. I tried to show the yield of this conception of culture in an earlier article on the interpretation of everyday language in settings controlled by America (Varenne 1987). 19. 'God bless you and bless the work that lies before us. May God protect us, and may we always, as a nation, be worthy of the gifts that St. Patrick brought us. May the Almighty, Who has brought us safe until now from the calamity and misfortune that have befallen so many other nations in consequence of this war, grant us shelter and protection to the end and make us worthy to play a noble part in the new world of the future.'

References Arensberg, Conrad and Kimball, Solon (1940). Family and Community in Ireland. Cambridge, Mass.: Harvard University Press.

The Semiosis of Anthropological Ireland 149 Bakhtin, M.M. (1981). The Dialogic Imagination, C. Emerson and M. Holquist (trans.). Austin: University of Texas Press. Bax, Mart (1973). Harpstrings and Confessions. Amsterdam: University of Amsterdam Press. Brody, Hugh (1973). Inishkillane: Change and Decline in the West of Ireland. London: Norman and Hobhouse. Brown, Terence (1979). After the revival: the problem of adequacy and genre. In The Genres of the Irish Literary Revival, R. Schleiger (ed.), 153-177. Dublin: Wolfhound Press. (1985). Ireland: A Social and Cultural History, 1922 to the Present. Ithaca: Cornell University Press. Clifford, James (1982). Person and Myth: Maurice Leenhardt in the Melanesian world. Berkeley: University of California Press. Clifford, James and Marcus, George, eds. (1986). Writing Culture: The Poetics and the Politics of Ethnography. Berkeley: University of California Press. Cresswell, Robert (1969). Une Communauti Rurale d'Irlande. Paris: Institut dlithnographie. Curtin, Chris and Wilson, Tom (in press). Ireland from Below. Galway: Galway University Press. Dumont, Jean-Paul (1986). Prologue to ethnography or prolegomena to anthropography. Ethos 14: 344-67. Fabian, Johannes (1983). Time and the Other: How Anthropology Makes Its Object. New York: Columbia University Press. Fox, Robin (1978). The Tory Islanders: A People of the Celtic Fringe. Cambridge: Cambridge University Press. Geertz, Clifford (1973). The Interpretation of Cultures. New York: Basic Books. Gibbon, Peter (1973). Arensberg and Kimbal revisited. Economy and Society 4: 479-498. Glassie, Henry (1982). Passing the Time in Ballymenone: Culture and History of an Ulster Community. Philadelphia: University of Pennsylvania Press. Harris, John (1985). Phonological Variation and Change: Studies in HibernoEnglish. Cambridge: Cambridge University Press. Harris, Rosemary (1972). Prejudice and Tolerance in Ulster: A Study of Neighbours and 'Strangers' in a Border Community. Totowa, N.J.: Manchester University Press. Hollier, Denis, ed. (1973). Panorama des Sciences Humaines. Paris: Gallimard.

150 Herv6Varenne Humphreys, A.J. (1966). New Dubliners: Urbanisation and the Irish Family. London: Routledge and Kegan Paul. Joyce, James (1961[1921]). Ulysses. New York: Vintage. (1967[1914]). Dubliners. London: Jonathan Cape. (1969[1916]). A Portrait of the Artist as a Young Man. New York: The Viking Press. Kane, Eileen (1982). Cui Bono? Do Aon Duine. Royal Anthropological Institute Newsletter 50:2-3. (1986). Stereotypes and Irish identity: mental illness as a cultural frame. Studies 75: 539-551. Kiberd, Declan (1984). Inventing Irelands. The Crane Bag 8:11-23. L£vi-Strauss, Claude (1971). L'Homme nu. Paris: Plön. McDermott, R.P. (1985). Wisdom from the periphery: talk, thought, and politics in the ethnographic theater of John Millington Synge. Quarterly Newsletter of the Laboratory of Comparative Human Cognition 7: 1-6. Malinowski, Bronislaw (1961[1922]). Argonauts of the Western Pacific. New York: E.P. Dutton. Marcus, George and Cushman, Dick (1982). Ethnographies as texts. Annual Review of Anthropology 11: 25-69. Marcus, George and Fisher, Michael (1986). Anthropology as cultural critique: An experimental moment in the human sciences. Chicago: University of Chicago Press. Messenger, J. (1969). Inis Beag, Isle of Ireland. New York: Holt, Rinehart and Winston. O'Brien, Hann (1986[1941]>. The Poor Mouth. London: Grafton Books. O'Faolain, Sean (1980[1947]). The Irish: A Character Study. Rev. ed.. London: Penguin Books. Peillon, Michel (1982). Contemporary Irish Society: An Introduction. Dublin: Gill and Macmillan. Rabinow, Paul (1977). Reflections on Fieldwork in Morocco. Berkeley: University of California Press. Russell (AE), George (1982[1916]). The National Being: Some Thoughts on an Irish Polity. Dublin: Irish Academic Press. Scheper-Hughes, Nancy (1979). Saints, Scholars, and Schizophrenics: Mental Illness in Rural Ireland. Berkeley: University of California Press. Singer, Milton (1984). Man's Glassy Essence. Explorations in Semiotic Anthropology. Bloomington: Indiana University Press.

The Semiosis of Anthropological Ireland 151 Synge, J.M. (1971[1907]). The Aran Islands. Oxford: Oxford University Press. Taylor, Lawrence (1980). Colonialism and community structure in Western Ireland. Ethnohistory 27: 169-181. (1981). "Man the fisher": salmon fishing and the expression of community in a rural Irish settlement American Ethnologist 8: 774-788. Thompson, William (1982[1967]). The Imagination of an Insurrection: Dublin, Easter 1916. West Stockbridge, Mass.: The Lindisfame Press. Torode, Brian (1984). Ireland the Terrible. In Culture and Ideology in Ireland, C. Curtin; M. Kelly; and L. O'Dowd (ed.), 20-29. Galway: Galway University Press. Trudgill, Peter, ed. (1984). Language in the British Isles. Cambridge: Cambridge University Press. Tylor, E.B. (1958 [1871]). The origins of culture. New York: Harper and Row. Varenne, Herv6 (1978). Culture as rhetoric: the patterning of the verbal interpretation of interaction between teachers and administrators in an American high school. American Ethnologist 5: 635-650. (1982). Jocks and freaks: the symbolic structure of the expression of social interaction among American senior high school students. In Doing the Ethnography of Schooling, G. Spindler (ed.), 210-235. New York: Holt, Rinehart and Winston. (1983). American School Language: Culturally Patterned Conflicts in a Suburban High School. New York: Irvington Publishers. (1984a). Collective representation in American anthropological conversations about culture: culture and the individual. Current Anthropology 25: 281-300. (1984b). The interpretation of pronominal paradigms: speech situation, pragmatic meaning and cultural structure. Semiotica 50:221-248. (1986). Symbolizing America. Lincoln, Nebraska: The University of Nebraska Press. (1987). Talk and real talk: the voices of silence and the voices of power in American family life. Cultural Anthropology 2(3):369-394. Vincent, Joan (1983). "Marriage, religion and class in South Fermanagh, Ireland, 1846-1920." in Emergent structures and the family, ed. by 0. Lynch, 175193. Delhi: Hindustan Publishing Corporation. Wilde, William (1979 [1852]). Irish Popular Superstitions. Dublin: Irish Academic Press.

152 Hervd Varenne Wilson, Thomas (1984). From Clare to the Common Market: perspectives in Irish ethnography. Anthropological Quarterly 57:1-15. (1987). Mythic images of the Irish family in the works of Flaherty, de Valera, and Arensberg and Kimball. Working Papers in Irish Studies 872/3: 15-31. Yeats, W.B. (1962[1921]>. Easter 1916. In Selected Poems, M.L. Rosenthal (ed.). New York: Collier Books. (1952 [1902]). Cathleen ni Houlihan. In Collected Plays. New York: Macmillan.

Against Coping Across Cultures: The Semiotics of Self-Help Rebuffed James A. Boon

In Man's Glassy Essence Milton Singer poses the self as object and subject of semiotic systems. He thus broadens and advances his life's project of comparing different value complexes — from entrepreneurship to symbolic constructions of personal identity — in India and the West, or more precisely (I wink) greater Madras and greater "Yankee City." With that characteristic combination of profound dignity and equally profound play, both essential to his erudition, Milton Singer has turned Americanized semiotics (from Peirce to Sebeok) and Indianized structures (from traditional to modern) towards each other, thereby crossilluminating their respective civilizations. I first had the privilege of playing jester to Milton Singer's philosopher-king when helping him teach "Comparison of Cultures" in Chicago in 1972 (wasn't it?). No experience in the classroom has ever matched it; it was pure pedagogical pleasure. I here resume my previous role of antic side-kick hoping to complement exemplary learning. Milton Singer has charted as no one else the highepistemological side of comparative studies, Indic/Euro-American semiotics, and philosophies of social form. This paper pokes fun at the low-therapeutic side of some semiotics of self-help advanced in recent popularized accounts of presumed encounters between Anglo-American culture and a range of other "others," including India. •

*

*

Two extreme entries in the self-help sweepstakes in anthropology and Indie studies respectively are Colin Tumbull's tepid The Human Cycle (1983) and Gita Mehta's scalding Karma Cola: Marketing the Mystic East (1981). I open with a cheeky critique criss-crossing these two books. I then turn to an apparently unrelated yet oddly parallel work, a quaint confessional by the long late Charles Hose called Fifty Years of Romance and Research, or a Jungle Wallah at Large (1927). My critical tactics include countering subjectivity with subjectivity, fighting froth with froth, undercutting current clichds with platitudes passis, and

154 James A. Boon comparing the psychobabble of now with colonialist pomposities of yore. I try to strike sparks between Turnbull's first-person avowals and Mehta's hip slogans, and between each of them and Hose's earlier pieties and my own somewhat dated doggerel. Toward rapidly raising the semiotics of self-help into self-conscious doubt, in this collision of readings I adopt the rhetoric I am contesting in order to wear it out — with apologies to all (honest). Now-a-Donahue-days in U.S. medialand, spokespersons for academic disciplines, if "disciplines" we be, popularize findings by converting intricate motives of research into psychotherapeutic objectives and a journalistic moral or "message." Self-help-style anthropologists often suggest that Providence scattered diverse languages and customs over an otherwise whole Mother Earth to provide lessons in tolerance, to teach all peoples mutually, cross-culturally to cope. My skepticism, which I would rather call irony, about this comforting assumption persists, despite Colin Turnbull's thusly persuaded Human Cycle. Basically this book converts human (includes women and children) existence — the bad times and the good, the thin with the thick — into a tidy succession of arts: becoming, transformation, reason, doing, and being. These stages represent what many would call childhood, adolescence, youth, adulthood, and old age. Born British, now "being" in America, Turnbull has come to feel free to confess his personal preference for any puberty other than the one he suffered: I cannot think of a single culture I know that handles this crucial stage of life more abysmally than we do [an ethnologist as snide as Gita Mehta (see below) might immediately wonder whether he has considered subincisionists]... It should be a criminal offense that this natural, wholesome, and utterly wonderful signal of the transformation of a young body into something else should be... associated with dirt and impurity, if not with sin (p. 81). Positively no footnotes and a single reference intrude upon Turnbull's two hundred and eighty three (283) pages of earnest, resolve squarely and forthrightly to parallel his tribulations with the hardships and triumphs of other souls and selves, be they Pygmy novice, Buddhist abbot, or Hindu sannyasi. Although The Human Cycle is dedicated to Arnold van Gennep's classic Rites of Passage ("Rites de passage," the aforementioned reference) ,alert readers may find more echoes of Gail Sheehy's Passages ("Paaeehsages") fleshed out with Turnbull's very-own-true-life-crises and fieldwork experience, and marketed accordingly.

Against Coping Across Cultures 155 "Supportiveness" spills over every page. "Saints are few and far between," observes Turnbull, "but we do not have to be Tibetans or abbots to give something of ourselves — and we all have something worthwhile to give..." (p. 246). That's a relief: Even without renunciation (not much fun) or going Tibetan, we can all get into sharing, giving, learning, being, being in touch with ourselves individually, and helping others be in touch with themselves individually too. Confronted by this nonstop conspicuous kindheartedness, even readers who are fairly "nice guys" may fail to suppress retorts of the kind perfected in Karma Cola's style of sarcasm. (Again, see below. Fortunately for Turnbull, his book was published after Mehta's; unfortunately for us, it was published before Walker Percy's knowingly unprophetic Lost in the Cosmos: The Last Self-help Book). Accordingly, I confess: Turnbull's text brings out the "Mehta" in me. It's Mister Rogers across cultures; Sesame Street gone native; global "free to be," total "you and me." The sweetest part of Turnbull's scheme is that everything winds up with not death but "The Art of Living." A not-so-tiny hint of immortality is tucked in as the last word, the benediction: We are like the piglets who like it that way, and have quite forgotten the ecstasy of divinity. Like them, for us the human cycle has become so bogged down in the muddy business of survival rather than being a soaring flight of rich fulfillment (p. 283). Really, the book ends right there, soaringly. Turnbull's earlier popular works include the much-taught Forest People on the affable Mbuti, hunter-gatherers of northeastern Zaire (still affable in The Human Cycle) and the much-condemned Mountain People on the tragically unaffable Ik whose historical and environmental woes are omitted from The Human Cycle. Some readers now find The Forest People more sanctimonious than they once realized. Some observers doubt whether an earlier, clearer alarm sounded about the plight underlying The Mountain People would have helped victims of starvation anyway, given Uganda's political circumstances. In 1984 the Human Cycle might have renewed such controversies;however, a brief correspondence in the New York Times Book Review was bumped by bigger news: Derek Freeman's attack on Margaret Mead. Even posthumously, Mead alone seems able to garner sustained media coverage for anthropological issues, always difficult to convert to headlines or even special features.

156 James A. Boon In the London Review of Books Edmund Leach (1984: 22) entered the non-fray about Tumbull, declaring his utter exasperation with The Human Cycle: "That a professor of anthropology should cast himself in the role of Old Testament prophet calling his fellow sinners to repentance is more than I can take." Quick on the draw as always, but careless of aim as usual, Leach then attempted to implicate longtime rival L6vi-Strauss in Turnbull's preacherly ploys. Leach mistakenly declared Turnbull's rhapsodies about spirit "a mode of expression...very LiviStraussian" and confused Turnbull's first person reminiscences with the prose devices of Tristes Tropiques. L6vi-Strauss's subjective narrative constructs a sampler of confessionals, travelogues, philosophies, and ethnographies; his is a formulaic quest that "crosses over" into both the world of the tropics and, as important, the world of tropes. The patent irony of the "je" in Tristes Tropiques is reiterated by L6vi-Strauss in Mythologiques (and by his English translators John and Doreen Weightman.) A pastiche of styles of discovery of self, Tristes Tropiques' contrastive comparisons seek "to find a language" for depicting the limits of cross-cultural discourse. Whatever one may think of L6vi-Strauss, a project could hardly be more remote from Turnbull's. Leach possibly found Turnbull's casual, colloquial, self-invoking experiment in informal exposition a bit close for comfort. Leach's A Runaway World based on his Reith lectures brimmed with personal and professional counsel for a troubled globe. Rather than revisiting (or rethinking) his own confessions of being trained as an engineer, Leach simply brands The Human Cycle "reflexive anthropology" (sniff), a presumably passing genre he deems "highly fashionable in the United States." Leach, I think, is actually reacting against the confessional anthropology of self-help. As persons more profoundly read in American schools of semiotics than Leach know (sniff), "reflexive anthropology" betokens an intensifying recognition by ethnographers that their descriptive and interpretive devices are themselves culturally and historically specific products of the semiotics of crosscultural representations. Leach's major works have, of course, helped advance this development, but his testy reviews of late lose sight of this fact. Enough about Sir Edmund. Whether or not Turnbull's style of self-help is one's cup of tea (and in this instance I obviously share Leach's distaste), its distinguishing features repay careful scrutiny. The book's hyperbole illustrates recent sales pitches for marketing popular ethnography. A comical disparity separates the book's contents from its cover's cant "Astonishing, extraordinary, passionate and illuminating, inspiring...destined to become a classic...one of the world's major contemporary social scientists." A string of suggestive but modest

Against Coping Across Cultures 157 memoirs covering a bumpy career thus enters what Cannel and Macklin call The Human Nature Industry. The inflated packaging of Turnbull's loosely woven, highly selective ethnography and autobiography seems designed to fill a genuine void left by the loss of Margaret Mead. Mead's accomplished arts of they/me cross-illumination seldom even heuristically separated the strands of other- and selfinvestigation. But her descriptions and confessions were geared to popular issues that she helped formulate: from generation gaps to various modes of liberation. Mead, then, helped make the times. She wrote a cross-cultural Passages before Gail Sheehy, not after her. Mead earned the vast audience she attracted. Not so Turnbull. Long after The Forest People, The Human Cycle discloses his gradual awakening to playful Mbuti eroticism and metaphors of copulating with the forest. He can now digress into Mbuti sex, and into unexpurgated Westminster schoolboy sex as well. (Possible movie material?) It is predictably revealed that Turnbull, no stranger to Weltschmerz, has brushed with evil: his Nazi nanny Irene was succeeded by Roman Catholic nanny Feldman, "a lovely German woman from Wipperfurth," who introduced him "by mail to her younger brother, Hans, who was an active member of the Hitler Youth" (p. 62). (Definitely, movie material.) Turnbull laces the melodrama with repeated avowals of a comparative "I'm O.K. You're O.K.," uptight repressives excepted. Here is an example from "Childhood: The Art of Becoming": Earth, water, air, and fire: the forest itself. Through these symbols the Mbuti are constantly reminded of Spirit, for wherever they are, whatever they are doing, those symbols surround them and even permeate their whole being. In the more artificial world we have built for ourselves we are not so fortunate; such effective symbols are harder to come by, particularly if we have never learned to employ our whole being as a tool of awareness. But that does not mean that Spirit is inaccessible to us. I am sure that many like myself groped their way through childhood aware that there was something lacking and found their own Great-uncle Willie, their own Rule Water, their own Arthur Poyser, and their own counterpart of what music was for me. All I am saying is that our form of social organization merely allows it to happen as an accident, if at all, whereas that of the Mbuti writes it into the charter from the outset, at conception (p. 77).

158 James A. Boon There now, doesn't that make us feel better? We too can learn from Uncle Willie to hush, grope, listen to running water, and see our own reflection in a Rule Valley stream, or, for that matter, just about everywhere. Primed early on for sentimentalized responses to "natural symbols," Turnbull had a ready-made empathy for forest people, or so he tell us. The rest, according to his book, is history. It would be child's play, but clearly "unbecoming," to go on caustically exposing, as Gita Mehta might, the transparent semiotics of The Human Cycle; its too-conversational style and format; its too-casual first person and pretentious colloquialisms; its studied elimination of any and all scholarly apparatus, irksome jargon, or cumbersome kinship diagrams deemed to alienate readers from books anthropological. To counteract likely suspicions that I am misrepresenting Turnbull's text or exaggerating the degree it condescends to a readership imagined by those marketing it, I now cite the climax in this cross-cultural therapy without barricades: I used to love it when two trees were close enough near the top for me to jump from one to the other in the bland rather than blind conviction that the lower branches would break my fall. I really should have been born an Mbuti [Italics added]. Rooftops were irresistible, only dangerous when you got too close to the edge and could see down... There were only two friends with whom I could share my play life, and they were both what were called "tomgirls." They shrank from nothing until one of them asked me if I knew what boys did to girls in deserted homes, and all I could think of was urinating. I was then made aware that urinating, or whatever boys did to girls,was improper [Italics in original] (pp. 68-9). Do you and I, readers, want or need to know this? Must Turnbull, too, tell it like it is and was? Dreaming that incomparably impossible dream of being born Mbuti — of what cut is left conveniently unspecified — he beseeches us to resist artifice, get into nature and supemature, and be ourselves being with others being themselves too. This kind of passage is all that this book is.

Against Coping Across Cultures 159 First Comparison (dyadic) Gita Mehta's Karma Cola allows me to frame Turnbull's excesses against ones diametrically opposed. Again, it is Mehta's voice, a polar extreme, that my quickie put-downs of The Human Cycle have thus far been mimicking. Mehta sounds bitter-bitter to Turnbull's saccharine-sweet, spleen to his ϊάέαΐ. The author — an Indian educated in the West, her bookjacket announces — strips all the goody-goodyness off going Hindu. She compares such spiritual quests to an extremist passage well beyond the pale of Turnbull's gingerly metaphors: "To go from the monomania of the West to the multimania of the East is a painful business. Like a sex change" (p. 36). Ugly, funny, unfair, insightful chapters portray her experience in decadent ashrams (compare Boccaccio on monasteries) besieged by guru-groupies who merely add touches of tackiness to her homeland's "implacable Eastern cruelty that lays the blame on the doer" (p. 39). Mehta is often facile, but it still hurts. No question of or room for benevolence here. Any advocate of "Eastward Ho!" is fair game — Huxley, Yates, the Beatles, and Ginsberg — during, after, and before the Hippie era: Those were the days when everything was in flux. There were rumors about Tab Hunter. Elvis Presley had just made a movie with a scene set in a whorehouse. Now a famous, published, avant-garde American poet had looked upon India and pronounced it free. Naturally Calcutta thought his reference was to the carnal. Before you could say snap, the Beat prophet was encircled by vers libres satyrs (p. 70). The book's barbs include preposterous slogans of East/West transgressions, including Joycean hybrids of Sanskrit/Cockney: "Om is where the art is" (p. 197). Relentlessly skewering the half-bakedness of West/East encounters, Mehta's diatribe cuts both ways: "Hindu thought is without dogma and dogged by Dharma;" "no one teems like we do" (pp. 192,86). Readers may end convinced, as I was beforehand, that hyperbole and hard-sell can be important factors when cultures contact Indeed, such rhetorical forces may be as vital as, say, religious rationalization in the history of cross-cultural discourse — a point I developed in Other Tribes, Other Scribes (1983). The figurative style includes both ultra-denigration and ultra-prettification of others; it stretches from missionary propaganda to touristic come-ons, both designed in part to convince the

160 James A. Boon gullible to book passage — either to save them or to join them. Simply to denounce the exaggerated style is to capitulate to it; nor can we ethnographers completely avoid commercializing "the other." We may as well admit it and factor this fact into our interpretations. Mehta deigns explain her super tide: It would appear that when East meets West all you get is the neo-Sanyasi, the instant Nirvana. Coming at the problem from separate directions, both parties have chanced upon the same conclusion, namely, that the most effective weapon against irony is to reduce everything to the banal. You have the Karma, we'll take the Coca-Cola, a metaphysical soft drink for a physical one (p. 107). She provocatively lists linguistic influences at work in Indian English today, including her own parodic prose. The post-war availability of Dell Comics is credited with moving Anglo-Indian rhythms from orotund elocutions of colonial days to the "fractured prose of America." Turnabout, as usual, became fair play: By the Sixties, modulation had given way to acceleration. The explosive shorthand of America seemed infinitely preferable to the dilatory obliqueness of England...On the other side of the planet the world's fastest speech looked for new words for slowing down. For twenty years we had burrowed in their vocabulary, now they scavenged in ours. Together with their own "laid backs" and "mellowed outs" went our Karmas, Sadhanas, Nirvanas, Tantras, and Sanyas. With language as with goods you take what you need. The British took from us jodhpurs and bungalows, riding breeches and colonial cottages, words for a more settled times. We had taken the idiom of modern America because it seemed to have no discernible provenance, a spontaneous verbalism that embraced the immediate as well as the immediate future. But now that America has taken our most complicated philosophical concepts as part of its everyday slang, things are getting sticky (p. 103).

Against Coping Across Cultures 161 As we have seen, British Tumbull, gone American, compounds certain offenses that Mehta exposes. Even when criticizing Westerners who consider India essentially "spiritual," Turnbull avows that his own experience in Banares taught him how to have everything all ways: At the university I practiced the art of learning; at the ashram, where I lived, I learned the art of being ecstatic, and at the age of twenty-six went through my second adolescence, properly this time. Between the two of them I began to learn the art of reason, the proper application of knowledge (p. 147). Again, Tumbull's probably well-intentioned pages idolize his own quest in the gurukula. Mehta travesties any such devotion to renouncers' "first miraculous micturition" (p. 83). Countering "religious Esperanto" (p.54) with traditionally caustic free-thinking (Western-style), Mehta signals the paradox of visitors who discover rebirth and learning-to-live in the land of karma, "a bad choice for narcosis." And she wonders how travellers' self-projections can possibly find fulfillment among Indian sages, when Eastern wisdom's primary concern is "the annihilation of narcissism" (p. 106). No visitors escape unscathed: Canadians, Australians, Germans, the British, "still conscious of the lines of Imperial Vision," and Americans, the Grand Optimists "trying for the big one — the vault from solitary into nothing. Well, they have the money and we have the time..."(p.79). The semiotic point of my criss-crossed reading is this: The vivid contrast between Karma Cola and The Human Cycle is a shallow one; the seeming polarity conceals a deep resemblance. Mehta is actually playing a game similar to Turnbull's (and in mimicking her my essay partly succumbs as well, although reflexively.) Mehta's glitzy irreverences, hilariously studded with pop tune doggerel and commercial clichi, are as sales-conscious as the prepackaged guruspeak and exotic psychobabble she slickly derides. She displays all those emotions common among outraged, deracinated critics: anger, anger, and anger. She markets the marketing of the Mystic East. As Turnbull hitches the crossing of cultures to the star of self-help, Mehta lodges West-meets-East in tabloidlike exposes, dressed out both against and as Madison Avenue hype. What I hope my comparison shows is this: Each rhetorical simplification is valuable mainly for debunking the other's pufferies. Turnbull and Mehta represent equivalent halftruths in opposite directions: rhapsody and bile. Readers might read both Mehta

162 James A. Boon and Turnbull so that the two can cancel each other out. But if you read only one, let it be neither, let it be... Naipaul. Both V.S. Naipaul and the late Shiva Naipaul (his younger, less sanguine brother) reveal disturbingly mottled qualities in cross-cultural encounters: both opportunistic and generous, both self-effacing and self-serving, on both sides. Turnbull soars across cultures, Mehta scathes across cultures. V.S. Naipaul — along with other informed travellers and, I think, many hard working, languagelearning, context documenting, career-making ethnographers and semioticians — limps. Naipaul, however, in such travelogues as An Area of Darkness, limps (via customs agents, visa stampers, taxi drivers, and the entire demimonde of civilizations' go-betweens) across cultures exquisitely, with disturbingly surer a sense of self than most of us ever muster. Certainly Naipaul's donnish selfpossession in impeccable prose disturbs critics awaiting the Trinidad-born, England-educated Indian to emerge as champion of or spokesman for the Third World, identifying exclusively with it or some narrower sector — the Creole Caribbean, wounded India, etc. But Naipaul's world (and his brother's, tragically more so), consists of refugees, only refugees — himself (ironically at-home only in his utterly articulate paragraphs) included. This essay, however, is not about Naipaul.

Last Comparison (triadic) My second companion volume for Turnbull is another confessional penned some sixty years ago by Charles Hose, distinguished official, naturalist, and ethnographer. Hose's Pagan Tribes of Borneo (1910) and Natural Man (1926) provided an ethnological foundation for the then-exemplary Sarawak administration of the durable Charles Brooke. In 1927 Hose indulged in looking back via a selftestimonial called Fifty Years of Romance and Research, or A Jungle Wallah at Large. The work's lingering Victorian confidence in civilization's superiority and steady progress leads today's readers readily and willingly to hoot at Hose (much as I indulged above in Mehtaesque pot-shots at Turnbull's text). Hose was perhaps asking for it, even in 1927 (an astonishing five years after Malinowski's Argonauts of the Western Pacific). Hose stipulates at the outset that, unlike most romancers, he prefers "the solid to the sensational" (p. 15). Yet he proceeds to detail his eerie prescience of earthquakes, to cry "Eastward Ho!," and to call his elaborate salute to

Against Coping Across Cultures 163 Aban Jau, Chief of the Tinjar Sebops, "A Rob Roy of Sarawak" (p.48). "A pagan and a savage, he was one of the finest gentlemen I have ever known," Hose concludes in commemorating this late "loyal supporter of the Government (p. 58)": Relentless foe, yet firmest friend, When friendship we did vow. From memory's land, I clasp thy hand And hail thee, Aban Jau! (p. 59). Still, however florid, Hose's autobiography on the whole was nothing if not solid. Today in fact, it appears comically solid. His reminiscence shuttles between contradictory tendencies toward modesty and fanfare: I remember how in my time at Cambridge a saying was current concerning a distinguished Don of one of the colleges, "I am the great Ί am,' none so fine as I." The phrase became a catchword for undue self-exaltation, which I have no desire to emulate... A propos of the great Frenchman [Clemenceau], I may mention that in 1898 I was honored by France in being made an Officier de l'Acad6mie Frangaise. From Germany in 1893 I had received the Order of the White Falcon (Saxe-Weimar), which is, I understand, bestowed on Zoologists only. I am proud of this decoration, as I obtained it on the recommendation of the great Haeckel. Another German decoration was that of the Knight of the Prussian Crown, sent to me by the ex-Kaiser through the German Consul-General at Singapore...[and so on] (pp. 15,2612).

Hose's interim passages include (I adopt Turnbull's scheme): "Transforming" on the Cam; "reasoning" in the footsteps of his kinsmen before him; "doing" with Poppy, his lifetime better half; and living to see his off-spring "becoming" and getting into "transforming" at Cambridge too. Charles Hose thrived during those claustrophobic times when all really was "in the family," both extended and descended:

164 James A. Boon The motto adopted by a member of the Hose Family is "Omne solum viro patria est;" and our family seems to have recognized its truth practically... Thus it happened that when my uncle [father's brother], Bishop Hose [of Singapore, Labuan, and Sarawak] wrote to my family offering to get me, if possible, a Cadetship in Sarawak, of all places, and under no less a man than Rajah Sir Charles Brooke, my delighted acceptance was conveyed in the single word (since become a family joke) "Rather!" (pp. 17-18, 33). Hose's career owed much to the Mrs. as well, whom he dutifully acknowledges: "More than a helpmate, she has been a guide and an inspiration, a companion in my voyages and a sharer in my dangers" (p. 18). Poppy's portrait serves as frontispiece; she is saluted as his editor-secretary and commended for certain original contributions. Once for example, during her husband's absence from the Sarawak residency at Sibu, she single-handedly pickled yet another orangutan killed by Ibans for stealing fruit; her technique faulty, the giant carcass required decapitation, so that its bloated head (a photograph is provided) could be expedited to Dr. Duckworth at Jesus College, Cambridge. Poppy's own account of this episode is inserted into Fifty Years (pp. 90-92). Hose's encomium concludes: "A good all round sportswoman, she has been worthily followed by my two children, who have already distinguished themselves, the one as a Lawn Tennis player [readers can check an action shot of Violet's backswing], the other as an Oar, who has twice gained his oars in Jesus College boats, and rowed at Henley" (p. 18). Who could ask for anything more? Like Turnbull sixty years later, Hose had some none-too-rosy tales to tell of Public School, including a prankish theft of eggs that, "but for the kindly discretion of the Headmaster might have involved expulsion": "A Public Schoolboy may rob nests and break bounds, but he does not lie; we owned up, received our caning, and the incident was closed" (p. 26). (Potential Masterpiece Theater material?) Unlike Turnbull, Hose stayed the course, conformed: "In point of fact relations between boys and masters at Felsted were never than the best...the Headmaster is the school; the school is what the Headmaster makes it. My gratitude I have tried to express in a practical form by sending my son there..." (p. 27). Hose's lifetime achievements included captaining Football at Felsted, playing against Oxford for the Gentlemen of Essex, leading the school bowling averages,

Against Coping Across Cultures 165 and serving on the Cricket First XI; these experiences taught him "common sense, self-reliance, and tact" (p. 27). His profession of faith in athleticism culminates in an anecdote cm, if one may coin a British-American hybrid, "Cambridge crew": This year (1927) once more I was present when the Jesus first boat went Head of the River, both in the "Lents" and "Mays," and I also had the pleasure of seeing my son rowing in the Jesus second boat. Tommy Watt, as he is familiarly known to his many friends, that famous old enthusiastic Don of Jesus College,...said...as I sat next to him at the high table: "Hose, I have seen my last cricket match, and my last football and Rugger match. I have lived, thank God, to see our first boat go Head of the River once again. I can now sing Nunc dimittis" — then he added [coy old Don that he was] — "But I hope the boat club will not get drunk until after tomorrow night" (p. 31). (Definitely, Masterpiece Theater material.) Contrary to the impression these passages may convey. Hose is not guilty simply of padding memoires with details unrelated to his career in Sarawak. Oaring on the Cam had provided the model for annual boat races that he instituted on Borneo's Baram River, a proud accomplishment mentioned in all three major works: "When in 1900 the vice-Chancellor of Cambridge University conferred on me the honorary degree of Doctor in Science, the Public Orator, Dr. Sandys, specifically mentioned in his speech this event, which has so effectually turned the rivalry of head-hunters to peaceful competition in sport" (p. 259). Worse policies are conceivable. Hose's engrossing chapters relate his life among sportsmen, among headhunters, then among sportsmen again, ex-headhunters now included. He recollects his life's passages as a cyclic cruise: from the river Cam to the river Baram and home again, Baram now mirroring Cam. Hose is a sitting duck for campy retorts from us postmodern readers. His work, however, may deserve better. After Sarawak Hose helped eradicate beriberi in Malaysia and Japan; an account of this project reveals that his quaint hopes for civilization's progress (undeflected by World War I) cannot quite be reduced to outright xenophobia. In artful twists of rhetoric, Hose praises Japanese friends and colleagues; although still smugly favoring British schoolways, he concludes with a telling image of joint superiority:

166 James A. Boon Another Japanese personal friend has been Mr. T. Kuga, formerly manager of the Mitsubishi Trading Company, and now Paris manager. His son, Taro Kuga, is shortly to enter a preparatory school in England, and, on my recommendation, will, I hope, eventually proceed to Felsted and Jesus. It has been a great pleasure to me to be associated with representative men of an Island Race like our own, and a people, like us, with a World Mission (pp. 262-3). The usually archaic-seeming Hose (even in 1927) here sounds, unfortunately, almost ahead of his time. Any island that pulled its oar could join the team and shoulder civilization's self-proclaimed burdens. Further hints of underestimated complexities in colonialism's shadowy values glimmer through Sir Arthur Keith's preface to Fifty Years which commends Hose's enlightened administration based on ethnographic evidence: That spirit of boyish adventure and youthful outlook [Hose] carried with him to Borneo, and it abode with him there. It was because he had the power of becoming a child again that it was possible for him to enter into the native mind and see the world as Nature's savages see it — minds which swarm with spirits of all kinds — spirits which have to be obeyed or propitiated. How stupid we white men often are! We have to know each other very intimately for many years to discover the motives of conduct, and yet we can persuade ourselves that in half an hour, by a few questions, framed in imperfectly understood words, we can fathom the secrets of a native people whom we wish to rule. Charles Hose never made this mistake; he approached the tribes as one boy approaches another, and in the course of time came an exchange of secrets, and this exchange gave Dr. Hose the key to successful government. And at the same time Dr. Hose was able to enlarge the common store of that kind of knowledge which is valued and collected by anthropologists (pp. 8-9).

Judging from Keith's preface, so-called colonialist paternalism could more accurately be described as a doubled juvenilization: both They and We.

Against Coping Across Cultures 167 Headmasterly Hose, the consummate colonial official, related to his subjects less in lordly fashion than boy-to-boy. He probably truly did admire the "Rob Roy of Sarawak." Hose linked subordinate populations to the dominant power by locking both sides into displays of perpetual schoolday pleasures. Dreaming of a history that would never need transforming again, Hose governed by exchanging secrets, so to convert every headhunter to an Oar. My brief attempt here to read Hose cunningly aims not to exonerate colonialism or the anthropology sometimes in cahoots with it (but sometimes not). I aim rather to help disclose more convoluted, mythlike undercurrents in political and historical arrangements that appear straightforward. One begins to suspect significantly contradictory sides to Charles Hose (not to mention Aban Jau!). Keith's 1927 tribute can make Hose seem almost relativist, "convinced that the difference between East and West is not so great as is usually believed; under our diverse creeds, codes, customs, tongues, and skin-tint is hid the same human nature — the same reactions" (p. 6). But exceptions doubtless marked such convictions. Keith-on-Hose even recalls moments cited above of Turnbull-by-the-Rule: "As [Hose] was learning in his boyhood the habits of plants, birds, and beasts among the marshes of the Waveney, he was equipping himself for the jungles of the East" (p. 6). But this parallel between Hose and Turnbull is ultimately misleading. Hose imposed other-government instrumentally; Turnbull seeks self-help presumably innocently. Still, who could say whether a colonialist superiority complex ä la Hose or psychologistic self-identification ä la Turnbull is more imbued with so-called "cultural imperialism?" Both, as Mehta's Karma Cola can remind us, manage to suffocate differences. For purposes of this paper, the crucial feature shared by Hose in 1927 and Tumbull in 1983 is this: Each felt compelled eventually to look back, self-regardingly, on his exotic adventures — one in a celebration of his colonialist career, the other in a searching of his personalized soul, both unremittingly, simplistically earnest. For this reader, like Hose's anthropology in its day, Turnbull's anthropology conforms too closely to its era's status quo: ultra class conscious or school snobbish in Hose's case; pop democratic in our own. Still, as Hose's Fifty Years demonstrates, bad books can make provocative cultural and historical documents. Too near in time (and perhaps, like Leach, close for comfort) to judge Turnbull fairly or even interpret him cunningly, I daresay that sixty years hence The Human Cycle may reveal to future readers richer dimensions of cross-cultural self-help than I can now perceive. But not yet — "Then all say in their hundred voices, "No, not yet,' and the sky said "No, not there.'"

168 James A. Boon Critical Conclusions (opposing all three) Between the historical will to dominate "the other" and the personal hope to help "the self" — that is, between the extremes represented by Hose and Tumbull — there exist more subtly inflected motives of and for anthropology and area studies. But it is difficult to get them across. Ardent voices of late — many of them "Mehtaesque" — have decried establishment ethnography and comparative research, charged with insufficient awareness of their collusion in colonial and post-colonial power structures. True enough, but only partly: the headline isn't that simple. Equally insistent voices have proclaimed from exotic locales their self-discovery, their breakthrough into personal coping. This essay has attempted to help these voices undermine each other. Increasingly, even professionals appear to cross cultures either unctuously to And themselves or angrily to denounce old-guard scholarship that supposedly inhibited anybody (including "the natives") from so doing. In anthropology, area studies, and sister disciplines, motives beyond moralistic self-help or ideologized predecessor-denouncing may themselves become closeted. This development too should be resisted, and not just stuffily. Halting, variegated, uncocksure, responsibly and/or subversively doubting semiotics seldom make catchy headlines, book blurbs, interviews, -isms, or blueprints (agendas) for political or psychological improvement. But formats must be preserved for these less marketable stories, too. For they are truer. *

*

»

It is, indeed, this truer kind of story depicting the semiotic complexities of any and all passages between India, U.S. (us), and elsewhere — all these scratches on each other's minds and hearts — that Milton Singer has persisted over many books innovatively in reporting. The comparative semiotics of self that he articulates transcends any and all individualistic merchandizing of self-help. Yet Milton Singer's sense of semiotics is so full that he would doubtless require comparative theory to comprehend even the self-therapeutic motives and slogans that my essay provisionally has rebuffed. Beckoning a future scholarship that would neither simply reject self-help nor privilege it, but rather seek to understand it as one semiotics in the Peircean profusion of human signs, I conclude confessionally with a parodic moral and ironic guidelines for would-be readers of "selves" — theirs and others (including mine):

Against Coping Across Cultures 169 I'm not O.K. nor, I suspect, are you O.K. Nor is my culture or any culture I know anything about O.K., particularly (pace Colin Tumbull) in their puberty rites. Nor will getting to know them, even getting to know all about them, help make either of us O.K. Hence, my in-depth advice to the sexually, spiritually, or otherwise hassled "Westerner", with whom I identify and sympathize, as we all should: Skip fieldwork, avoid ashrams, read pornography and/or scripture (both more therapeutic than ethnography); and hire an analyst, not an informant, a guide, or a guru. My concluding counsel will probably sound "Charles Kuraulty," although I would prefer to echo a voice whose sparkling dignity corresponds more to the semiotic self of Milton Singer and to his teachings in the philosophy of cultural comparisons. I thus profess: Any culture-crosser, whether personally growing, transforming, soaring, scathing, coping, or not coping at the time, should be interested not just in "himself" — this complex semiotic construction — being interested in, for example, Mbuti or Banares. He or she should be interested in Mbuti or Banares, or Britain, or U.S., or anywhere, anyway. Why? Because peoples in all their peculiar social, historical, ecological, economic, political, and ritual circumstances (what Milton Singer has called their "cultural performances") are there: becoming, transforming, reasoning, marrying and not marrying, oppressing, being oppressed, entrepreneuring, semioticizing, liberating, warring, doing, and dying — whether you or I ("me," babe) get ourselves into being there too, or not And that's, semiotically, the way it is.

170 James A. Boon References Boon, James A. (1983). Other Tribes, Other Scribes: Symbolic Anthropology in the Comparative Study of Cultures, Histories, Religions, and Texts. Cambridge: Cambridge University Press. Cannel, Ward and Macklin, June (1973). The Human Nature Industry: How Human Nature Is Manufactured, Distributed, Advertised and Consumed in the United States and Parts of Canada. Garden City, New Jersey: Anchor Press. Hose, Charles (1910). Pagan Tribes of Borneo: A Description of Their Physical, Moral and Intellectual Condition, with Some Discussion of Their Ethnic Relations. London: Macmillan and Co. (1926). Natural Man: A Recordfrom Borneo. London: Macmillan and Co. (1927). Fifty Years of Romance and Research, or a Jungle Wallah at Large. London: Hutchinson and Co. Leach, Edmund (1968). A Runaway World? New York: Oxford University Press. (1984). Intolerance. London Review of Books, May 16, 1984, pp. 22-23. Livi-Strauss, Claude (1974). Tristes Tropiques, J. Weightman and D. Weightman (trans.). New York: Atheneum. Malinowski, Bronislaw (1922). The Argonauts of the Western Pacific: An Account of Native Enterprise and Adventure in the Archipelagoes of Melanesian New Guinea. London: G. Routledge and Sons. Mehta, Gita (1981). Karma Cola: Marketing the Mystic East. New York: Simon and Schuster. Percy, Walker (1983). Lost in the Cosmos: The Last Self-help Book. New York: Farrar, Straus and Giroux. Sheehy, Gail (1976). Passages: Predictable Crises of Adult Life. New York: Dutton. Singer, Milton (1984). Man's Glassy Essence: Explorations in Semiotic Anthropology. Bloomington: Indiana University Press. Turnbull, Colin M. (1961). The Forest People. New York: Simon and Schuster. (1983). The Human Cycle. New York: Simon and Schuster. van Gennep, Arnold (1960). The Rites of Passage. M.B. Vizedom and G.L. Caffee (trans.). Chicago: University of Chicago Press.

Singer, Kant, and the Semiotic Self Murray J. Leaf

I am concerned with two questions. The first is empirical. Given that the self is in some sense semiotic, how many selves are there for a person or a community? Is there one, or are there many? The second is philosophical and historical. Milton Singer (1984) attributes his semiotic conception of the self to Peirce. At the same time, taking Peirce's scattered but persistent criticism of Kant at face value, Singer sees a sharp difference between the basic assumptions of his own position and Peirce's, on the one hand, and Kant's, on the other. This is important to question for two reasons. First, it concerns not only the relation of Kant to Singer but, necessarily, the length and scope of the lineage of the semiotic perspective in anthropology. Second, the issues at stake are prerequisite to dealing with the question of the number of selves. My argument is that Peirce was much closer to Kant than his criticism suggested, and that his criticism itself is misleading. Yet Singer is closer yet to Kant than Peirce was, with the result that his position is empirically stronger than Peirce's, and at the same time more clearly consistent with and pertinent to the recognition of multiple selves. If we accept the Peirce-Singer conception of the uniqueness of Peirce, the intellectual lineage of Singer's semiotic of the self is necessarily short and far removed from the intellectual mainstream of social science as it has developed over the last two centuries. It is also ambiguous to the extent that it uses ideas and assumptions that neither Singer nor Peirce have defined precisely and exemplified empirically. On the other hand, if we see the roots of both Singer and Peirce in Kant, and understand the way in which Peirce weakened the empirical side of Kant's position, then we are prepared to see Singer's work as a movement from one part of a broad series of Kantian co-traditions to another, a movement in the direction of a system of analyses that has defined much of the mainstream of social science for the last two centuries. These include experimental psychology, several major traditions in linguistics, including Sapir's approach to language and meaning and Hymes' approach to the ethnography of speech, and what I have elsewhere

172 Murray J. Leaf characterized as monistic traditions in cultural, social and symbolic anthropology (Leaf 1979).

Singer, Peirce and Kant: The Problem There are two basic components to Singer's semiotic self. The first is the centrality to cultural and social processes of the I-Thou-It relationship, including the idea that each defines the other and cannot be taken apart from the other. The other is the idea of semiotics, more specifically that this relationship is a sign relationship. It is, however, not so much linguistic as part of the logic that language necessarily presumes — an aspect of the structure of conceptions that language use depends on. Peirce's idea of the I-Thou-It relationship went through three interpretations in three separate schemes, each of which was to some extent what Rorty (1982: 161) calls "foundational" — an attempt to provide "an all-embracing ahistorical context in which every species of discourse could be assigned its proper place and rank."1 The first scheme was idealistic. The categories I, Thou, and It were simply held to be the ideal basis of all knowledge as though, in part, Kant had proved that they were. But in fact Kant had not intended to and had not done so. The Kantian categories, from which Peirce's I, Thou, and It derived, did form a system, which Kant did say was complete, but it was only a system of categories of one specific kind, which Kant called "noumena," in a much larger scheme that recognized many other categories of many other kinds. Kant (1950: 329-330) had said that the exclusiveness of these three types of noumenal categories was based on the "formal distinction of syllogisms," but did not elaborate. When Peirce realized that his own scheme provided no inherent reason for I, Thou and It to be interconnected, this hint evidently led him to develop his interest in logic, and this led in turn to the formulation of his second scheme, in "On a New List of Categories" (1867). This introduced the sign relationship as that which connected I, Thou, and It to each other by mutual and specifically semiotic implication (as relations established in communication), and this is the scheme Singer (1984: 60,101) makes his own. In one way, the question that concerns us is how great the difference was between these two schemes. In all its versions, Peirce's I-Thou-It was a summarizing paraphrase of the contents of Kant's "dialectic of pure reason," given in the Critique of Pure Reason and explained more straightforwardly in the Prolegomena to Any Future

Singer, Kant, and the Semiotic Self 173 Metaphysics. This was a part of what Kant called "the transcendental problem." It was, specifically, the part concerned with describing what Kant had previously argued were the three and only three types of transcendental objects, the noumena, which were also transcendental ideas. All such ideas were, according to Kant, either psychological, cosmological, or theological. They were either souls or minds, things or series of things, or absolute or perfectly abstract beings, entirely removed from sense perception. Peirce's "I" is based on the psychological idea, the presumption of the mental world. His "Thou" refers to the abstract world (the theological idea). His "It" refers to the sensory world, Kant's cosmological idea. Exactly why ideas were described as transcendental objects is a key point to understanding Kant, and to the various misunderstandings of him, including those repeated by Peirce. It may sound like idealism when taken out of context, but it wasn't. The most important misstatement of the Kantian position that Singer accepts from Peirce concerns the empirical or epistemological status of the transcendental ideas or noumena. Peirce repeatedly suggests that Kant held them to be actually existent things, but unknowable or not fully knowable. That is, the transcendental objects were actual things, beyond perception. Singer (1984: 46,69,87) accepts this view of Kant and repeats it, arguing therefore that Peirce's own formulations, which do not postulate such actual entities, moved beyond Kant's in more directly recognizing the importance of symbolic processes (Singer 1984: 57,87).

Singer and Peirce After Peirce, the symbolic interactionisms of James, Cooley, and Mead each had a version of the I-Thou-It relation (Singer 1984: 89), as did the distinctly nonpragmatic cosmology of Martin Buber, which Singer (Singer 1984: 101, f.n. 11) is also aware of. Buber, like Peirce, connected it to the relationship between the self and God and made it the origin and prototype of all social relations — a formulation which Simmel in turn borrowed for his idea of social psychology, substituting society for the deity. Singer also recognizes that Peirce's idea is historically derived from Kant, but at this point Peirce's distortion comes into play: "having rejected, with Peirce, both a Cartesian self and a Kantian transcendental ego, the symbolic interactionists turned to empirical distinctions and observable processes" (Singer 1984: 87).

174 Murray J. Leaf While Peirce clearly predates the others, the concern with "empirical distinctions and observable processes" did not begin with him. Kant's actual position was that noumena, transcendental objects, were postulations, made to organize thought and communication. The clarity of his analysis leads well beyond a statement of the importance of symbolic and communicative process into a quite concrete analysis of what can only be called their pragmatic (Kant's own word, alternating with "practical") bases in human life. Kant used the term "object" as the complement of "subject." Anything the subject perceives is an object of perception; anything the subject thinks is an object of thought. An object could be a physical thing, and some were. But, he argued, transcendental objects, noumena, could only be ideas. Further, his description of the way in which these postulations are inherently paradoxical and can never be legitimately thought of as arising from observation itself raised in turn the question of the purpose of such postulations. This further led Kant into a most interesting analysis of the nature of purpose itself, and the way it too was both individual and social. Finally, this in its own turn led to his analysis of the way purpose and learned conceptions interact in perception and cognition — in his Anthropology from a Pragmatic Point of View. This marked a specific mixture of introspective and objective analyses of individual thought processes that led eventually to Wundt and Bastian and through them to Boas and Malinowski. An important aspect of Singer's own position is that the importance of the IThou-It relationship is observable as a matter of ethnographic experience, as is the use of the relationship in communicative activity. A "semiotic" analysis is an analysis of cultural communication that shows such a structure, and as he applies this to India it is further an analysis which shows that there are in fact indigenous analyses of communication which recognize fundamentally the same basic conceptions, and themselves apply those conceptions to significant meaningful behaviors (Hindu ceremonials) — in order not just to analyze an action in a part of the world but to create, maintain, and project an order for the whole of it. Thus: "the interpretations of these symbolic representations are not merely syntactic or semantic but are fully semiotic in the sense that they try to take full account of the contexts of use and of the dialogues between the utterers and interpreters of the signs" (Singer 1984: 182), and: When I first became aware of such selected and enacted selfimages, I called them 'cultural performances.' With the help of a Peircean semiotic analysis, we can see that these cultural

Singer, Kant, and the Semiotic Self 175 performances provide the names and forms, the images and diagrams, the songs, dances, and stories in a moving tableau of a self, not just a self burdened with an ancient past, but of an ideal self just coming into being. Indian identity is an 'outreaching identity' of a self that reaches out to other people through personal devotion to a personal deity, that aspires to the general welfare through disinterested action, and, finally, that seeks a vision of ultimate cosmic reality in meditation (Singer 1984: 182-183). Thus semiotic analysis is at one and the same time an analysis of communication, culture and social structure; the I-thou-it paradigm is the basic framework of communication, the conceptual basis of cultural traditions, and the building block of social structure. It is not just a logical analysis or elaboration of a general view of what the presumptions of communication must be.

The Links Back to Kant When Singer first gave the nub of the argument of Man's Glassy Essence as the Distinguished Lecturer of the American Anthropological Association in 1978, most of those present seemed fairly sure they had heard something important, but they weren't sure what it was. Singer's argument is more philosophical than most, and his claims are stated more circumspectly and documented more painstakingly than anthropologists are generally prepared for. More than that, perhaps, anthropologists have become used to a kind of argument that assumes that the only way to change our general conceptions is to offer a new characterization of everything at once. Singer, by contrast, proposes to change our general views by locating a key phenomenon, focusing on it, and changing our conception of it. Singer assumes that the conception of the "self" is such a crucial phenomenon, on the evident ground that all social relations, and perhaps all communications, assume such a conception. But this is neither self-evident nor unambiguous, and much turns on how it is supported. If it is indeed offered as an empirical generalization, what is the crucial evidence and how does it make the case? Singer's own experience, as he describes it, is logically insufficient to establish the generality of the phenomenon in relation to the full range of communicative acts, whatever that range may be. If it is offered as true by definition, how can a scheme

176 Murray J. Leaf that rests on this kind of proof be regarded as truly pragmatic — especially in the sense of the experimental model of proof that provides the standards for pragmatic argument as developed particularly by Mead and Dewey. The ambiguity of this aspect of Singer's argument directly follows Peirce. Coincidentally, this is an area where Peirce's analysis agreed with Redfield's positivistic tradition and departed from the other pragmatists and their philosophical and psychological predecessors and counterparts. Peirce's essay on "Consciousness and Language" (Peirce 1866) introduced the idea that man not only depended on signs but actually was a sign (cf. Harrison 1986:174). In "Some Consequences of Four Incapacities" (1966c[1868]: 52-53), Peirce insisted on the universal importance of I-Thou-It in all thought and linked it to the idea of the continuity of all thought, meaning that ideas have reference only to other ideas in an unbroken stream, a notion much like James' "stream of consciousness." The point is also argued in "Questions Concerning Certain Faculties Claimed for Man" (1966b[1868]: 34), which can provide a short quotation that shows how close he comes to arguing simply by definition: "from the proposition that every thought is a sign, it follows that every thought must address itself to some other, must determine some other, since that is the essence of a sign."

All of this strongly suggests that there is one I-Thou-It, or rather one "I" for each flow — which is each flow: It is hard for man to understand this, because he persists in identifying himself with his will, his power over the animal organism, with brute force. Now the organism is only an instrument of thought But the identity of a man consists in the consistency of what he does and thinks, and consistency is the intellectual character of a thing, that is, is its expressing something (Peirce 1966c[1868]: 71-72). The view that ideas only ultimately refer to other ideas leads Peirce to go on to hold that the individual flow of ideas refers to and cannot be realized apart from the ideas of the community in which he lives. This is the argument which concludes with the "glassy essence" quotation that Singer makes his own. Yet as an empirical conception, Peirce's formulation leaves much to be desired. It concerns,

Singer, Kant, and the Semiotic Self 177 plainly, not the thought of the community as it might be described at the time of the individual's action or any other reasonably relevant time, but at some time in a seemingly ever-receding future when it will be fully realized: ...so thought is what it is only by virtue of its addressing a future thought which is in its value as thought identical with it, though more developed. In this way, the existence of thought now depends on what is to be hereafter, so that it has only a potential existence, dependent on the future thought of the community. The individual man, since his separate existence is manifested only by ignorance and error, so far as he is anything apart from his fellows, and from what he and they are to be, is only a negation. This is man, '...proud man, Most ignorant of what he's most assured, His Glassy Essence* (Peirce 1966c[1868]: 72). The closing quote clearly suggests that it is this communal basis of the self-as-idea of which man is "most ignorant." Is this empirical, or is it an echo of Hegel and a precursor of the kind of socialized idealism we see in Dürkheim? Singer's paraphrase of this position emphasizes the points that Peirce can be seen to share with the later symbolic interactionists and empirical social science, but neglects his distinctive reification of community thought: For Peirce the self is a "bundle of habits," with the unity of selfconsciousness. Its identity is neither physiological nor psychological but consists in the logical consistency of its feelings, actions, and thoughts, which is also a consistency of symbolization: 'My language is the sum total of my self (Singer 1984: 186). The contents of Peirce's argument as well as this generalized way of expressing it ("man persists...", "the self...") make it at least difficult to ask, with respect to any particular example of a thought or utterance, which I-Thou-It it utilizes, or whose, or under what circumstances it exists or is used. If these questions cannot be asked, neither can "how many?"

178 Murray J. Leaf Singer himself does not pose or answer the question directly, but it is not difficult to pose in his language, and I believe the assumption in his Indian analyses is that the answer is that there can be more than one conception of a self in a community, although there may not be. Kant's own answer was that there were several kinds of self conception; the way he constructed it implied that there may be, in any community, several conceptions akin to Singer's "self."

Kant's

Position

Kant was concerned with knowledge in relation to perception. He was specifically concerned with one type of knowledge, which he characterized as synthetic a priori, as distinct from knowledge which was synthetic a posteriori and knowledge which was analytic a priori. Synthetic meant that the predicate of a statement conveying such knowledge did not merely repeat the subject — it somehow expanded or explicated it. A priori meant that such knowledge was prior to experience; that experience depended on such knowledge. Hence synthetic a priori knowledge was knowledge that was somehow prior to experience although relevant to it, and was not merely tautological or definitional. His method was descriptive, even 2

experimental, in a particular sense appropriate to his subject matter. He coldly and rigorously examined the traditional concepts of philosophy in the light of the operations of his own conscious thought as he saw it introspectively and the conscious thought of others as he saw it in action and communication, and he invited his reader to do the same along with him. Kant formulated the idea of the synthetic a priori in order to rephrase, in an answerable way, a question he took over from Hume: how do we know that our thoughts correspond to real existences? As in all other cases, the question was literal and not rhetorical. Kant was not denying that we know; he was recognizing that we do. The question was howl His answer involved a fundamental change of perspective that must be understood if anything else Kant said is to be understood. He (following Hume) rejected the idea that existences are extrinsic to thought — the idea of a gap between perception and reality — that had been the starting point of speculative philosophy since Plato. For Kant, perceptions were not, as an observable matter of fact, internal. The world of our perceptions is not in our head or mind, but all around us, and the structure of that world is therefore not beyond the structure of our understanding but within i t Since the structure includes time and space, which mathematics represents, time and space must have the same

Singer, Kant, and the Semiotic Self 179 status as other principles that structure our experience — which is apparently why he rejected Hume's separation of mathematics from other basic categories of experience: Sensibility, the form of which is the basis of geometry, is that upon which the possibility of external appearance depends... It would be quite otherwise if the senses were so constituted as to represent objects as they are in themselves. For then... The space of the geometer would be considered a mere fiction, and it would not be credited with objective validity because we cannot see how things must of necessity agree with an image of them which we make spontaneously and previous to our acquaintance with them. But if this image, or rather this formal intuition, is the essential property of our sensibility by means of which alone objects are given to us, and if this sensibility represents not things in themselves but their appearances, then we shall easily comprehend, and at the same time indisputably prove, that all external objects of our world of sense must necessarily coincide in the most rigorous way with the propositions of geometry; because sensibility, by means of its form of external intuition, namely by space, with which the geometer is occupied, makes those objects possible as mere appearances. ...thought space renders possible the physical space, that is, the extension of matter itself; that this pure space is not at all a quality of things in themselves, but a form or our sensuous faculty of representation; and that all objects in space are mere appearances, that is, not things in themselves but representations of our sensuous intuition (Kant 1950: 287-288). This provides a concrete epistemological and psychological alternative to the correspondence theory of truth. It is the foundation of the rest of Kant's analysis of perception, and in the longer run the basis of modern pragmatic theories of the social construction of reality, including Singer's, in the sense that such theories recognize that socially acquired categories of thought are constitutive of all perception. Kant is not, here, suggesting that geometry, either Euclidean or Newtonian, represented an absolute and unchanging system of innate ideas, as some have said.

180 Murray J. Leaf He knew that formal geometry has evolved over time. He was rather saying that we have an "intuition" of space and time, a general sense of it, which specific geometries give expression to and "formalize." We learn our formal mathematics, which then becomes the categories of our world, but we have a more basic human faculty to do so, which is also the faculty, the "sensory faculty," of the human mode of perception we employ. This faculty, and the physical space it renders possible, allows us to order our perceptions, judgements, first of all around the relation between subject and object, perceiver and perceived. Kant's treatment of the subject-object relationship subsumed and resolved one aspect of Hume's problem of determining whether our perceptions corresponded to real existences. The answer was yes and no: All of our judgments are at first merely judgements of perception; they hold good only for us (that is, for our subject), and we do not till afterward give them a new reference (to an object) and desire that they shall always hold good for us and in the same way for everybody else; for when a judgement agrees with an object, all judgments concerning the same object must likewise agree among themselves, and thus the objective validity of the judgment of the experience signifies nothing else than its necessary universal validity. And conversely when we have ground for considering a judgment as necessarily having universal validity (which never depends on perception, but upon the pure concept of the understanding under which the perception is subsumed), we must consider that it is objective also — that is, that it expresses not merely a reference of our perception to a subject, but a characteristic of the object. For there would be no reason for the judgments of other men necessarily to agree with one another. Therefore objective validity and necessary universality (for everybody) are equivalent terms, and though we do not know the object in itself, yet when we consider a judgement as universal, and hence necessary, we thereby understand it to have objective validity (Kant 1950: 298-299). That is, objectivity — the judgement that our personal perception holds good "objectively" — is the judgement that others will share that perception,

Singer, Kant, and the Semiotic Self 181 universally. This judgement, however, is not based on the perception as such, but rather on the categories under which we classify it. We consider a judgement to be objective, as distinct from merely our own personal judgement, when we expect all others to accept the categories under which we make it, and when we know that those categories will have the same significance for all of them. But how can we know this? We cannot, of course, have experience of everyone else's perceptions, or of the future — which Kant is well aware of. Rather, when we learn concepts we learn that some of them are considered universal as part of learning the concepts themselves, while others are not. We learn that time and space are universal when we learn what they are; we learn that beauty is more personal and relative in the same way. A judgment of objectivity is an imputation; it is justified by our experience (of learning and using the concepts) for concepts like space and time, but not for concepts like beauty. Kant's argument inserted a social and consensual component, an awareness of the social source and communicative uses of our conceptual categories, in what many previous thinkers had considered our most interior mental processes. At the same time it provided an empirically observable individualistic mechanism for the historical transmission and development of the most abstract and impersonal systems of cultural ideas, from mathematics to theology. His system was thus social psychology in a strict and rich sense. Thought was human but not private. To perceive was to think, to think was to communicate, to communicate required training, to be trained required life in a social group. This was, however, only the beginning of Kant's analysis. It is his analysis of phenomena. The analysis of "noumena" was built upon iL Noumena are Hume's things which have "real existence" in a second sense. They are not just the side of the mountain I presume to exist when I do not see it; they are the mountain which I presume to have sides, the mountain that exists when I am not looking at it and will exist when I am dead. Noumena are objects which are presumed to lie beyond perceptions, and in important cases to organize them. They are "transcendental" in this very specific and literal sense. To transcend is to start in but go beyond. They transcend perception in the sense that we see them as manifested in particular sensory experiences, but going beyond them. Such a thing, by definition, cannot be embodied in experience, and can not arise from it. It is, in Kant's analysis, therefore a pure creature of the intellect. It arises from the nature of the understanding, its need for organizing principles of a certain kind, and is in that precise sense a pure idea, synthetic a priori. Perceptions of ourselves as subject (in relation to objects, of course) we can order with the transcendental idea of a self or

182 Murray J. Leaf a consciousness. Perceptions of what surrounds us may be attributed to a universe, but "attributed to" is all that is being claimed for it. Apart from that attribution and the uses it serves, there is nothing: Hence if even the pure concepts of the understanding are thought to go beyond objects of experience to things in themselves (noumena), they have no meaning whatever. They serve, as it were, only to decipher appearances, that we may be able to read them as experience. The principles which arise from their reference to the sensible world only serve our understanding for empirical use. Beyond this they are arbitrary combinations without objective reality, and we can neither know their possibility a priori nor verify...their reference to objects (Kant 1950: 312-313). The passage speaks for itself, but, with respect to Peirce's (and Rorty's) failure to acknowledge the Kantian basis of modern pragmatism, it is important to stress that this is precisely what Peirce, Rorty, and others have said Kant was not saying.2 The rest of Kant's argument is the description of noumenal categories that Peirce first drew upon. In the section dealing specifically with psychological objects, the self was construed as both subject and object. As a subject, it is the idea of a "consciousness" which we presume to exist as the subjective locus of our perceptions that always itself remains beyond perceptions, the unperceived perceiver. This he describes as "only the reference of the inner phenomena to their unknown subject" (Kant 1950: 333-334). As an object, it is something we might somehow see in ourselves, and attribute our perceptions to. If there were any part of Kant's analysis where we might think he would be tempted to go beyond strict empiricism to the idealism he is accused of, this ought to be it. Yet he does not. The immediate concern is with the "soul." It was, after all, two centuries ago. But the parallels with the general argument and implications for more modern concepts are clear If, therefore, from the concept of the soul as a substance we would infer its permanence, this can hold good as regards possible experience only, not of the soul as a thing in itself and beyond all possible experience. Life is the subjective condition of all our possible experience; consequently we can only infer the

Singer, Kant, and the Semiotic Self 183 permanence of the soul in life, for the death of a man is the end of all experience which concerns the soul as an object of possible experience, except the contrary be proved — which is the very question in hand. The permanence of the soul can therefore only be proved (and no one cares to do that) during the life of man, but not, as we desire to do, after death. The reason for this is that the concept of substance, so far as it is to be considered necessarily combined with the concept of permanence, can be so combined only according to the principles of possible experience, and therefore for the purposes of experience only (Kant 1950: 334-335). That is, the idea of the soul only has meaning as a concept living people use to order their experience. The conventional religious and Cartesian view of the soul as something that can outlive the person is thus, quietly, held to lack any possible foundation in experience, although it is not therefore held to be irrelevant to experience. Similarly, the main forms of historical cosmological ideas or presumptions were all shown to involve "antinomies," whereby any logical way in which the idea could be held to be inherently related to experience could be contradicted by an opposite but equally logical alternative. The idea of God was held to be absolutely unrelated to experience on its face.3 Thus, in effect, no noumenal postulation can be held to originate in perception or be explained wholly by it: "...the transcendent cognitions of reason cannot either, as Ideas, appear in experience or, as propositions, ever be confirmed or refuted by it" (Kant 1950: 328-329). This raises an obvious question: then why do they exist? The answer is a new form of the importance of action, which Hume had also recognized. Kant argues that the major types of noumenal imputations (selves, totalities, and series) are not merely descriptive and do not refer only to present and past experience. They refer to all possible experience, and insofar as they do that they are also normative: "judgements, when considered merely as the condition of the union of given representations in a consciousness, are rules" (Kant 1950: 305). They are not just attributed, they are imposed. They do not just arise from life, they order it. They are not something we have, they are something we do. Thus on the physical side, Kant's analysis ran from the smallest distinguishable perception of a moment through formal systems of mathematics to the universe. On the social side it ran from the learned habits imposed in an instant of judgment

184 Murray J. Leaf to the most general ideas that have been established over time to frame our histories and our futures, although it did not include the way such consensus was settled. Both sides of the analysis are interdependent at all levels. The physical world depends on social learning, social learning depends on and refers to the physical world. Kant's description of transcendental psychological objects, by itself, does not translate directly into an answer to the question of whether a person or a community can present multiple selves. Yet in the larger context of Kant's system the answer becomes an obvious yes, after he chides us to the effect that the concept of self needs a good deal more specification than the question first suggested. Self as perceiving subject is not the same as self as the object of a specific perception (my image in a mirror, my hands before me on the keyboard) which is itself not the same thing as a "consciousness" or mind I may impute to my perceptions as that which has or does them, either as absolute subject or as object of my reflection about such an absolute subject. When one understands how profoundly different they are — united only by a presumption and in no way by perception — the remaining question becomes "can people use different conceptions of a consciousness or a mind or a soul or an ego (all being phrases Kant himself used) as the presumed organizing basis of their perceptions?" The question virtually answers itself. It demonstrates that the answer is "yes," or at the very least "why not?" Peirce was not greatly concerned with the relation between psychological subject and conceptualized self, and generally wrote as though a person had one self, both subject and object, and one world. In a parallel way, sociologists such as Simmel and Redfield assumed that a culture or civilization had a world view which it imposed on its members; this naturally entailed a view of the self as singular and monolithic. Singer accepts Peirce's phrasing. He fairly consistently speaks of "the" self rather than, for example, "a" self or "one's selves," even though technically there is nothing in his scheme that would preclude the latter. The issue is not whether there is a "world view" or not, or even whether such a world view contains a conception of I-Thou-It. The question is whether there is only one per person or only one in each community, culture or civilization. Is there an Indian concept of self, or are there Indian concepts of self? Or still more precisely are there current among the people of India numerous self-identifications some of which are phrased as distinctively Indian identities, while others may be phrased in entirely different terms (such as professionals, or kinsmen, or members of ethnic or linguistic groups). A further and equally important question is

Singer, Kant, and the Semiotic Self 185 whether all communications actually do assume or manifest at least one of them. The questions are simple. Peirce's insistence on the absolute fundamentality of the I-Thou-It may make asking them seem gratuitous, but really it is not — because the request is for no more than the basic "what-where-when" information that should accompany any ethnological claim. Singer's cultural performances were rituals of a certain type, self-defined as representing a quite particular tradition — Hindu and even then of a certain sort. Would he have found the same basic concept of the self with other ceremonies of other religious traditions (Muslim, for example) or more strongly of non-religious traditions? Would the ideas of the self in Indian kinship systems be the same as in the religious ideas? It seems to me that in the case of India the simple and clear empirical answer is that there are multiple distinct traditions carrying distinct world views, with distinct concepts of a self. Moreover, there are many reasons for believing this to be true in all communities. In addition, and still more strongly relevant to the Peircean and sociological legacy, there are many reasons for believing (with Kant) that some communications do not in fact use such a self conception at all. Where is the "I-thou" in 2+2=4?4 One can, of course, impute an I and a Thou to such a communication by adding ideas it does not on its face contain. But then, if we confine ourselves to what is empirically evident, the real locus of the I-Thou would be in the statement that conveys that imputation. It would be difficult to think of a way to avoid recognizing this that would not reduce the general semiotic analysis to a tautology, making it apply to any thought or communication anywhere in any way.

Kantian Co-traditions Kant's distinction between subject and conceptualized self was present in Wundt's argument for folk-psychology (which would describe evolved concepts of selves and others) as a necessary extension of experimental (individual) psychology (Wundt 1916: 3), and was directly reflected in both Peirce's and G. H. Mead's (1964: 227) distinction between consciousness and self-consciousness. Kant's portrait of the disparate shapes and occasions of conscious perception, however they occurred and without trying to impose his own unity on them, was the basis of Wundt's (1904) conception of "the manifold of consciousness" which he offered in place of traditional ideas of "mind," and his treatment of the way this manifold arose in behavior was in turn the basis of the "functional" psychology that was

186 Murray J. Leaf even more important to pragmatism than the self-other relation. Kant's concept that categories of perception were also norms, and the idea that laws of perception were laws of the objects themselves reappeared in almost exactly the same words in Wundt's (1897-1901: vol. 1,10-11) argument that ethics were prior to logic in the development of human mentality and thought. Wundt was also, of course, a major source of Boas's concern with the diversity of cultural forms within as well as across communities and of Malinowski's functionalism (cf. Leaf 1979: 180ff.). In linguistics, and of perhaps even more direct relevance to Singer's analyses, there is also an important and reasonably clear line of development from Kant's conception of intuition and its formalization through Herder's conception of "concentration" to Sapir's and Whorf s approaches to the relationship between the language forms and basic cultural categories of a community (Leaf 1979: 304ff.). Negatively, Singer's dissociation of the semiotic I-Thou-It from grammatical person agrees with Whorf s finding that basic concepts of a cultural world view do not neatly parallel the phenotypes of a language. But there is much more to be said. Why, after all, is grammatical person so ubiquitous? Can there really be no connection at all? If one takes noumena clearly as imputations, one is naturally driven to look more closely at specific interactions between linguistic forms and ideas they are used to convey, than if one takes noumena as simply ubiquitous. By the same token, if one takes noumena as imputations one has a clearer reason to ask what the relationship is between I-Thou-It and other basic cosmological categories such as space and time, however conceived. Kant held that the transcendental ideas presumed the intuition of space and time, and were built upon it. Whorf, seemingly with this in mind, concentrated on space and time and not on person. Peirce does not greatly concern himself with the question (although it is present by implication) and again Singer seems to follow Peirce — but it could have a great deal to do, for example, with the fact that such basic categories as I-Thou-It are given their most elaborate and prominent cultural expression in ceremonies, such as Singer described, wherein actual people embody them in sequences of action. This necessarily places the ideas in space and time; presents them in or as a spatial metaphor. So, in only a slightly different way, does their use in "dialogues" in all the senses Singer describes. Is such imagery only a kind of accident? Or is it a kind of logical necessity, because the concepts themselves will not be completely conveyed without it? Is culture a text (or series of texts), or is it a play (or series of plays)? Can a play be reduced to a text? If we really want to understand the inner structure of our thought and behavior, such questions must be answered with factual precision.

Singer, Kant, and the Semiotic Self 187 Big Traditions or Small Traditions? Peirce was steeped in Kant, and did not conceal the influence. In "The Place of Our Age in the History of Civilization," Peirce (1966a[1863]: 7) calls The Critique of Pure Reason "perhaps, the greatest work of the human intellect." His formulations generally follow Kant much more closely than Peirce himself acknowledges. His famous essays in Popular Science were virtual paraphrases of the opening sections of the Critique of Practical Reason. Even when Peirce gave his ideas an overall idealist caste, it was an idealistic version of a basically Kantian set of issues and answers, in an evident attempt to make them more self-contained and "logical." The effect on Singer's position of accepting Peirce's criticism of Kant without placing it in the context of Peirce's underlying debt to Kant is a kind of self-mutilation. It shortens the reach of Singer's formulations, cutting them off from a host of allied insights that lead from Peirce's semiotics to formulations and topics much beyond Peirce's schemes into the truly empirical (and not merely programmatic or conjectural) work that has shaped the mainstream of the social and behavioral sciences for the last two centuries. Does this mean Singer has done less than he has claimed? No, actually it means he has done more. His formulations, precisely because they are tied to ethnological practices and to a constructive assessment of the major trends in ethnological theory over the past three decades, have moved Peirce's genuinely original and seminal concentration on communication as the locus of thought back from the shadow of idealism to the larger Kantian tradition that Peirce drew on.5 This yields much more than a formulation of the self-other-it relation and a promising but sketchy general approach to symbolic processes. It also provides an underlying psychology and philosophy of science that elaborates it and gives it at once both greater specificity and greater descriptive power. It provides both a characterization of a key type of social relation (that which involves an ego and alter in normative structured interaction) and a way to differentiate these relations from other types of communications (which do not impose such moral relations). Kant's social psychology was the most important single source of pragmatic philosophy, academic ethnology, and a system of allied and mutually supportive empirical disciplines. Singer's semiotic self demonstrates its continuing vitality and provides modern anthropology with an important key to its future as well as its past.

188 Murray J. Leaf Notes 1.

2.

While this characterization of Peirce is apt, I should not be understood as endorsing either Rorty's overall representation of Peirce's argument or his broader characterization of pragmatism and what it is to be distinguished from. Rather, I take Rorty as another example of the problems that arise when one takes pragmatic criticism of Kant outside the context of the pragmatic debt to him. The contrast Rorty draws between his own position and Peirce's neglects precisely Peirce's concern with communication, his semiotics, which makes Peirce's orientation toward language much closer to Rorty's than Rorty acknowledged. The principle differences are that in place of Peirce's communicative triad Rorty has* a much less well developed idea of a "conversational" dyad (Harrison 1986: 177), and in place of Peirce's view of the world as constituted through language, Rorty, despite his pragmatic selfdefinition, is actually offering a kind of nominalism that includes "a radical disconnection between language and the world" (Harrison 1986:179). With respect to larger philosophical alignments, Rorty applies the "foundationalist" label equally to Kant and Descartes among many others. There is ironic justice in this in that it rests on characterizing Peirce the same way that Peirce characterized Kant, but it is still not correct as a characterization of Kant himself. Although Kant was explicitly attempting to "found" what he called a "new science," the way he intended to do so was not foundational in Rorty's sense. Rorty does not recognize a difference between attempting to invent or provide foundations as Descartes was doing, and as Peirce did not quite avoid, and attempting to describe them, as Kant was doing. Kant's description does, as will be clear, exactly what Rorty says Kant does not do — imbeds such concepts as the self and mind in the ongoing contexts in which they emerge and are employed. Rorty rests his argument that Kant's assumptions and aims were not pragmatic primarily on Dewey, particularly on The Quest for Certainty. The nub of Dewey's argument in that context, however, was not based on a defense of a position like Rorty's or on an attack on foundationalism. Campbell (1984: 183) has described the general differences between Rorty's position and Dewey's, revolving around Rorty's "omission...of the three central Deweyan themes of democracy, education, and experience." The criticism Rorty paraphrases was built on a part of the last: the importance of

Singer, Kant, and the Semiotic Self 189 experiment as the most important or decisive pragmatic basis of knowledge (or basis of pragmatically clear and precise ideas). Dewey recognized that Kant agreed in this valuation of experimental knowledge, but argued that he failed to implement it in fact: There is accordingly opposition rather than agreement between the Kantian determination of objects by thought and the determination by thought that takes place in experimentation. There is nothing hypothetical or conditional about Kant's forms of perception and conception. They work uniformly and triumphantly; they need no differential testing by consequences. The reason Kant postulates them is to secure universality and necessity instead of the hypothetical and the probable. Nor is there anything overt, observable and temporal or historical in the Kantian machinery. Only the result is observed, and only an elaborate process of dialectic inference enables Kant to assert the existence of his apparatus of forms and categories (Dewey 1929: 289). This criticism, and even more strongly Rorty's intemperate expansion of it (cf. Rorty 1982: 72-88), lacks a foundation that it logically requires. The problem lies not so much in Dewey's understanding of Kant as in his understanding of requirements of his subject matter. Kant's subject matter is the corpus of metaphysical knowledge, or at least that part of it that compels attention as knowledge that is genuinely synthetic a priori. Given this, Dewey ought to say what an experimental approach to it could be before he can show that Kant's method is not one. Experiments must be suited to what they are experiments on; they must use the material they are intended to investigate. To say that Kant is not using an experimental method because his conclusions are stated firmly and not hypothetically is trivial. Galileo's conclusions, once he reached them, were stated at least as firmly as Kant's. On the other hand, Wundt's experimental psychology is shot through with important approaches to problems taken directly from Kant. Would this have been possible if the original were not experimental in some important sense to begin with? In fact, Kant offers hundreds of hypotheses which he confirms, eliminates, or modifies in the course of building his argument, as some of the quotes that

190 Murray J. Leaf

3.

4.

will be given should indicate. If anything, this is precisely why Kant is so hard to read. To say that Kant was not aware of the place of his own work in history, and did not recognize it in that work, ought to strike anyone who has read a good sample of the Kantian corpus as unfair at best. History and history of thought were major preoccupations for Kant, and the critiques which Dewey is paraphrasing literally began and ended with it. They began with a historical sketch of the source of the problems and ended with a preview of work to follow, for which his own merely paved the way. This included what Kant called "anthropology," the empirical investigation of the development of thought in history. In my own view, Kant was raising, in a far more systematic way than anyone before him, precisely the question of what an experimental investigation of the bases of thought would be. Unquestionably he was feeling his way as he went, and the work was incomplete and unpolished — as he himself said. It is a question that has not been answered yet. But the conception of an experiment suggested in Dewey's criticism of Kant is no help at all. An experiment is a structure of decision, not a manner of expression. Rorty does not pick up the theme. In fact, as Brodsky (1982: 330-333) notes, he ignores it, which is the major reason he can thus equate Kant, Descartes, and Peirce (and also pragmatism with his own brand of nominalism). This section is very brief, and does not convey a clear idea of what Kant takes this idea to be, but presumably it is the idea of a "perfect" and immaterial being which, in itself, separates it from an imperfect and material universe. Kant did not elaborate the implications of this, but it of course has the effect of rejecting all naturalistic proofs of the existence of God. Peirce's argument for the universality of the I-Thou-It relationship in the second scheme rested on the idea that all (synthetic) knowledge was propositional. Following traditional logic, as Kant had, Peirce held that every proposition contained a copula (the "is") that connected subject to predicate, and it was the copula in particular that he saw as embodying the signing relationship, the I-Thou-It, by asserting that the predicate of the sentence applied to the subject by agreeing in its reference to an object. Even if we accept such an argument for statements that are synthetic a priori, which, remember, is all the Kantian analysis applied to, it is not of immediate or clear applicability to statements wherein the "is" represents a

Singer, Kant, and the Semiotic Self 191

5.

mathematical or definitional equality as such, as in this case. Nor, strictly speaking, is it applicable to such non-propositions as "look out!" unless one wants to join the logical positivists in the dreary game of rewriting all such definitions or statements to show that they "really are" propositions after all. For additional, specifically American, aspects of this broader background see Moreno (1985).

References Brodsky, Garry (1982). Rorty's interpretation of pragmatism. Transactions of the Charles S. Peirce Society 18(4): 311-333. Campbell, James (1984). Rorty's use of Dewey. Southern Journal of Philosophy 22(2): 175-187. Dewey, John (1929). The Quest for Certainty. New York: Minton, Balch and Company. Harrison, Stanley (1986). Our glassy essence: a Peircean response to Richard Rorty. International Philosophical Quarterly 26(102): 164-181. Kant, Immanuel (1974 [1798]). Anthropology from a Pragmatic Point of View, M.J. Gregor (trans.). The Hague: Martinus Nijhoff. (1950 [1783]). Prolegomena to Any Future Metaphysics, L.W. Beck (ed.). New York: Bobbs-Merrill. Page number references in the text are to the 1911 Preussische Akademie der Wissenschaft edition — the standard edition — which are indicated in the Beck edition by numbers in the running heads. Leaf, Murray J. (1979). Man, Mind and Science. New York: Columbia University Press. Mead, George Herbert (1964). George Herbert Mead on Social Psychology, A. Strauss (ed.). Chicago: University of Chicago Press. Moreno, Jonathan D. (1985). Pragmatists and pluralists: an American way of metaphysics. Metaphilosophy 16(2/3): 178-189. Peirce, Charles S. (1866). Consciousness and language. In Collected Papers of Charles S. Peirce, vol. 7, C. Hartshome, P. Weiss, and A. Burks (eds.), 7.579-7.596. Cambridge: Harvard University Press. (1867). On a new list of categories. In Collected Papers of Charles S. Peirce, vol. 1, C. Hartshorne, P. Weiss, and A. Burks (eds.), 1.5451.567. Cambridge: Harvard University Press.

192 Murray J. Leaf (1931-1958). Collected Papers of Charles S. Peirce, 8 vols., C. Hartshome, P. Weiss, and A. Burks (eds.). Cambridge: Harvard University Press. (1966a[1863]). The place of our age in the history of civilization. In Charles S. Peirce: Selected Writings, P.P. Wiener (ed.), 1-14. New York: Dover Publications. (1966b[1868]). Questions concerning certain faculties claimed for man. In Charles S. Peirce: Selected Writings, P P . Wiener, ed. New York: Dover Publications. Also CP 5.213-5.263. (1966c[1868]). Some consequences of four incapacities. In Charles S. Peirce: Selected Writings, P.P. Wiener (ed.). New York: Dover Publications. Also CP 5.264-5.317. Rorty, Richard (1982). Consequences of Pragmatism. Minneapolis: University of Minnesota Press. Singer, Milton (1984). Man's Glassy Essence: Explorations in Semiotic Anthropology. Bloomington: Indiana University Press. Wundt, Wilhelm (1897-1901). Ethics, 3 vols, English trans, from 2nd edition (1892) by Ε. B. Titchener, J. H. Gulliver, and M. F. Washburn. London: Swan Sonnenschein. (1904). Principles of Physiological Psychology, English trans, from the fifth edition (1902) by Ε. B. Titchener. Facsimile reprint 1969. New York: Kraus Reprint. (1916). Elements of Folk Psychology, E. L. Schaub (trans.). London: George Allen and Unwin.

Semiotic Origins of the Mind-Body Dualism Benjamin Lee

One of the more interesting and fruitful developments in the anthropological study of the self has been the introduction of semiotic ideas and approaches to a field which for a long time was locked in a moribund debate between cultural relativism and psychoanalytic determinism. Milton Singer's "Signs of the Self and subsequent articles brought an unusual vigor and rigor to the discussion by their expansion of Charles Peirce's semiotic and its application to such issues as the dialogic constitution of the self and the semiotic roots of personal and social identity. Using Peirce's anti-Cartesian starting point, Professor Singer began exploring the anthropological implications of Peirce's dictum that all our "knowledge of the world is derived by hypothetical reasoning from our knowledge of external facts" (CP 5.265) which always involves inferences from signs. Peirce's anti-Cartesianism is most carefully formulated in his "Questions Concerning Certain Faculties Claimed for Man" (CP 5.213-263) and "Some Consequences of Four Incapacities" (CP 5.264-317). In the former essay, he acknowledges that although we feel that we have an intuitive faculty which allows us to recognize a "premise not itself a conclusion" (CP 5.213), he claims that "there is no evidence that we have this faculty, except that we seem to feel that we have it" (CP 5.214); the particular premise concerned is the intuitive selfconsciousness of the Cartesian cogito. Instead, Peirce argues that this feeling is itself a product of inferences made from previous cognitions. In the latter essay, he objects to the Cartesian epistemological criterion which he claims amounts to "Whatever I am clearly convinced of is true" (CP 5.265), and instead postulates to a community-based inferential process in which "all knowledge of the internal world is derived by hypothetical reasoning from our knowledge of external facts" (CP 5.265) which is itself mediated by signs. If we assume the validity of Peirce's critique, the question arises, from what external signs do we infer the existence of an indubitable internal subjectivity? Although Peirce never answers this particular question, another anti-Cartesianist, Ludwig Wittgenstein (who, ironically, was also the topic of Professor Singer's

194 Benjamin Lee Ph.D. dissertation), provides several points which are critical for understanding why, if such a dualism is untenable as he and Peirce have argued, it constantly recurs not only in such disciplines as philosophy, but also in folk psychology (See D'Andrade (1987) for an overview): It seems that there are certain definite mental processes bound up with the working of language, processes through which alone language can function. I mean the processes of understanding and meaning. The signs of our language seem dead without these mental processes... We are tempted to think that the action of language consists of two parts; an inorganic part, the handling of signs, and an organic part, which we may call understanding these signs, meaning them, interpreting them, thinking. These latter activities seem to take place in a queer kind of medium, the mind; and the mechanism of the mind, the nature of which, it seems, we don't quite understand, can bring about effects which no material mechanism could (Wittgenstein 1960: 3). Perhaps the main reason why we are so strongly inclined to talk of the head as the locality of our thoughts is this: the existence of the words "thinking" and "thought" alongside of the words denoting (bodily) activities, such as writing, speaking, etc., makes us look for an activity, different from these but analogous to them, corresponding to the word "thinking." When words in our ordinary language have prima facie analogous grammars we are inclined to try to make the analogy hold throughout. We say, "The thought is not the same as the sentence; for an English and a French sentence, which are utterly different, can express the same thought." And now, as the sentences are somewhere, we look for a place for the thought. (It is as though we looked for the place of the king of which the rules of chess treat, as opposed to the places of the various bits of wood, the kings of the various sets.) We say, "surely the thought is something·, it is not nothing;" and all one can answer to this is, that the word "thought" has its use, which is of a totally different kind from the use of the word "sentence" (Wittgenstein 1960: 7).

Semiotic Origins of the Mind-Body Dualism 195 These two quotations bring together several points. The first one links language with mental processes which provides the basis for bifurcating the "processing" of language into "inorganic" and "organic" components, i.e., body and mind. The second quotation extends the initial sketch by combining two steps. First, Wittgenstein indicates the importance of the role of grammatical analogy; second, the grammatical analogy interacts with a notion that there exists a realm of thoughts independent of sentences and their uses. This independence of thought and language is linked by Wittgenstein to what could be called a 'propositional regimentation' of language. Wittgenstein's target is Frege's separation of a realm of 'sense-relations' from the uses of sentences, and the concomitant identification of a language independent realm of 'thoughts' (the sense of a sentence) which both determine reference and fix meaning. Since Frege's distinction between sense and reference was motivated by his attempt to apply first order quantification theory to language, Wittgenstein's comments can be seen as part of his larger critique concerning the distorting influence of logic in the analysis of language which, in this particular instance, when combined with certain grammatical analogies, produces the mind-body dualism. The distorting influence of logic and mathematics also appears in Descartes' work in which he explicitly uses them as guiding ideals for his work. Wittgenstein's comments on grammatical analogy, however, do not seem to have a direct application, especially since his remarks do not point out any of the particular linguistic mechanisms involved. Benjamin Lee Whorfs work supplements Wittgenstein's insights, however, by showing how grammatical categories might interlock to produce a "fashion of speaking". The particular fashion of speaking Whorf analyzes is that of the "objectification" of time in SAE (Standard Average European). We will extend his analysis to look at a "fashion of speaking" in SAE which, when combined with a propositional regimentation of language, produces the Cartesian dualism.

Whorf In his article "The Relation of Habitual Thought and Behavior to Language" (1956) Whorf analyzes a wide ranging "fashion of speaking" found in SAE. A "fashion of speaking" is formed by an interlocking of lexical, morphological, and syntactic classifications into a "certain frame of consistency" (Whorf 1956: 158). Furthermore, these configurations may have a large scale effect upon cultural

196 Benjamin Lee behavioral patterns. Whorf believed that the "objectification" of time which involves "imaginatively spatializing qualities and potentials that are quite nonspatial" (1956: 145) is integrally involved with the development in Western Europe of: Records, diaries, bookkeeping, accounting, mathematics stimulated by accounting. Interest in exact sequence, dating, calendars, chronology, clocks, time wages, time graphs, time as used in physics. Annals, histories, the historical attitude, interest in the past, archaeology, attitudes of introjection toward past periods, e.g., classicism, romanticism (1956: 153). Whorfs analysis of objectification involves several steps. The first is the treatment in SAE (Whorfs generic term for English and other related IndoEuropean languages) of time. According to Whorf, the basic subjective experience of time is that of "becoming later and later." SAE speakers overlay this experience and treat time as they treat length: denumerable, quantifiable, and measurable. Whorfs explanation for this "objectification" is that in SAE, plurality and cardinal numbers are applied to both real and imaginary plurals. The latter category includes time words such as 'minutes', 'hours', and 'days'. Real plurals refer to aggregates each unit of which is objectively perceptible (in Peirce's terminology, they are capable of being put into an indexical relation with the speaker), while imaginary plurals are not, but are treated by a process of analogical projection as if they were measurable. Ten men, unlike ten days, can be directly perceived (ten men in one group), but SAE has only one pattern to treat both cases and 'ten days' becomes regarded as a group via a mental reconstruction from 'ten men'. Their similar grammatical treatment causes temporal cyclicity to be likened to physical aggregates. Whorfs second example concerns the origin of the form-substance dichotomy in the structure of the maximally expanded noun-phrase of SAE. SAE contains two types of nouns denoting physical things. Count nouns denote bodies with definite outlines and are directly denumerable by cardinal numbers without the use of classifiers, i.e., 'a tree', 'three trees', 'a man', 'three men'. Mass nouns denote "homogenous continua without implied boundaries (1956: 140)." They are marked by the absence of plural forms and in English drop articles and in French take 'du', 'de la', and 'des'. Examples are 'water', 'milk', 'air', 'wine', 'meat', and 'sand'. When speakers need to individualize mass nouns, they use such body-type classifiers as 'stick of wood', 'piece of cloth', or container types such as 'glass of water' or *bag

Semiotic Origins of the Mind-Body Dualism 197 of flour'. In the contain» types there is a perceptible distinction between container and contents, indicated by the word 'of; for example, in 'a glass of water', both the glass and the water are objectively perceivable as separate and distinct. In the cases where there is no clearcut container/contents distinction, such as 'piece of cloth', there is an analogy created by the '-of plus contents' formula, so that 'lumps', 'pieces', etc., seem to contain a distinct substance or matter like the 'water' or 'flour' in the container-type noun phrases. Since noun-phrases define the referential universe, SAE speakers thus see the world divided into forms ("substanceless forms") and formless items ("formless substances"). For both of the above cases, Whorf presents contrasting Hopi data. In the case of time, Hopi applies plurality and cardinal numbers only to objective plurals, and there is no basis for the kind of analogical projection and linguistic objectification as in the SAE example. In the case of nouns, there is no formal subclass of mass nouns in Hopi, and generalness is conveyed through verbs or predication. Every noun implies a suitable container, and there is no analogical base to construct a concept of existence as involving a duality of form and substance. Whorf then shows how these first two processes, when applied to phase words in English, lead to an abstract notion of "objectifiable" time. Phase words such as "summer, winter, September, morning, noon, sunset" (1956: 142) are all nouns which differ very little from other nouns in their grammatical reactances. Like nouns referring to physical objects, they can take prepositions, i.e., 'at sunset', 'in winter' and plurals, i.e., 'five days', 'six hours', etc. Whorf views phase terms as referring to an experience of time which is basically cyclic, but since they are nouns, phase terms are also subject to the pattern of individual and mass nouns and its concomitant binomial formula of formless item plus form. For phase nouns, the formless item is 'time' (corresponding to 'substance' or 'matter' for physical objects). Thus 'a summer' or 'an hour' is imagined to consist of such-and-such a quantity of 'time' — an hour of time, a year o/time. Finally Whorf also views SAE as having a three tense system of verbs which allows us to imagine the objectified time units as standing in a row. Past, present, and future allow us to think of points along a temporally determined line. This spatial metaphor also extends to our expressions of duration, intensity, and tendency. For example, we speak of long' versus 'short' periods of time, a 'fast' or 'slow'decline. Whorf concludes: The SAE microcosm has analyzed reality largely in terms of what it calls "things" (bodies and quasibodies) plus modes of

198 Benjamin Lee extensional but formless existence that it calls "substances" of "matter." It tends to see existence through a binomial formula that expresses any existent as a spatial form plus a spatial formless continuum related to the form, as contents is related to the outlines of its container. Nonspatial existents are imaginatively spatialized and charged with similar implications of form and continuum (Whorf 1956: 147). Although Whorf suggests that there is a primacy in the direction of these analogies, he also implies that there is the possibility of diachronic feedback, ultimately affecting Western cultural practices. Nonspatial experience has one well-organized sense, HEARING — for smell and taste are but little organized. Nonspatial consciousness is a realm chiefly of thought, feeling and SOUND. Spatial consciousness is a realm of light, color, sight, and touch, and presents shapes and dimensions. Our metaphorical system, by naming nonspatial experiences after spatial ones, imputes to sounds, smells, tastes, emotions, and thoughts qualities like the colors, luminosities, shapes, angles, textures, and motions of spatial experience. And to some extent the reverse transference occurs; for, after much talking about tones as high, low, sharp, dull, heavy, brilliant, slow, the talker finds it easy to think of some factors in spatial experience as like factors of tone. Thus we speak of "tones" of color, a gray "monotone," a "loud" necktie, a "taste" in a dress: all spatial metaphor in reverse. Now European art is distinctive in the way it seeks deliberately to play with synesthesia. Music tries to suggest scenes, color, movement, geometric design; painting and sculpture are often consciously guided by the analogies of music's rhythm; colors are conjoined with feeling for the analogy to concords and discords. The European theater and opera seek a synthesis of many arts. It may be that in this way our metaphorical language that is in some sense a confusion of thought is producing, through art, a result of far-reaching value — a deeper esthetic sense leading toward a more direct

Semiotic Origins of the Mind-Body Dualism 199 apprehension of underlying unity behind the phenomena so variously reported by our sense channels (Whorf 1956:155-156). Wittgenstein's critique of the Cartesian and Fregean models of mind does not account for why we think that words such as "believe', 'think', and 'intend', refer to psychological processes that go on inside that thinking substance we call mind. He argues that philosophical analyses driven by the ideal of logic objectify our normal uses of these words, which have no such commitments, into philosophical positions which do. Wittgenstein himself suggests that "when words in our ordinary language have prima facie analogous grammars we are inclined to interpret them analogously" (Wittgenstein 1960: 7). As we saw earlier, Whorf suggested that SAE contains a configuration of grammatical categories which constitute a "fashion of speaking" whereby we "imaginatively spatialize" things that are nonspatial, leading to an objectified conception of time as an infinitely segmentable abstract and formless substance. In this section, we will look at another fashion of speaking in SAE in which speaking and thinking are equated through an intricate set of grammatical linkages. The discussion will focus principally on English, but will include examples taken from French and Latin. The goal is to show that there exists a fashion of speaking whereby the ways in which speech and thinking are represented are grammatically similar. For example, in all these languages the grammatical forms used to represent speech in indirect discourse can also be used to represent thought, and in some constructions can introduce the problems of referential transparency and opacity that philosophers have seized upon. Since the ways of representing speech and thinking are grammatically equivalent across these different forms, there exists the possibility of an analogical transfer from speaking to thinking and vice versa. This configuration of interlocking grammatical patterns accounts for a host of analogies, such as "inner speech" or speaking to oneself as the analogical basis for the idea of thinking to oneself, and the concomitant belief of the internality of mental activity, thereby describing the grammatical analogies Wittgenstein hints at Evidence for this analogical transfer between speaking and thinking is the development of what Ann Banfield (1982) has called "represented speech and thought" (also called by Cohn "narrated monologue," and in French "style indirect libre"), which develops first in French in the early seventeenth century and occurs primarily in literary texts and seems to be associated with the rise of the novel. In general, the constructions of represented speech and thought can be introduced by

200 Benjamin Lee both verbs of speaking and thinking and have grammatical properties which seem to lie 'in-between' those of direct quotation and indirect speech and thought; it possesses the grammatical persons and tenses of indirect discourse and the expressive aspects of direct speech. All these forms can be used to refer to and characterize speech but quotation is the most 'transparent' in that it maintains a word for word correspondence (and also sometimes matching intonation). Indirect discourse is the least transparent, as its lack of indexically presented expressive elements and its pronomial and tense shifts indicate. Represented speech and thought can thus be considered as the analogical product, both historically and syntactically, of an interaction between direct and indirect speech and thought forms. A critical aspect of the following argument is that in SAE the complement constructions of verbs of speaking and thinking can have both what philosophers have called de dicto (or 'opaque') and de re (or 'transparent1) readings. We will see that in English, French, and Latin, there are formally identical complement constructions for verbs of speaking and thinking (i.e. I say that he will come, I believe that he will come) which can have both de dicto and de re readings. This suggests that there is a slope from direct speech to a de dicto referential portrayal of a contextualized act of speaking or thinking, to a de re objective proposition, which encourages the belief that speech is the expression of decontextualized propositions that are contextualized in the moment of utterance. Under such an analogy, speech is the expression of underlying mental activity, and thus Tjehind' the speaking subject there stands its thinking subject counterpart. In both Frege and Descartes the vision of the mind operating in its own medium independent of the outside world is the product of the analogical identification of thinking with speaking which yields the metaphor of the internality of mental activity, and the abstract or decontextualized interpretation of the complement clause which breaks any indexical linkage between thought and the world. The internality of mind is the product of the grammatical analogies; the idea of a "thinking substance" is the product of the logical objectification of these analogies.

Direct and Indirect Speech and Thought If we compare the following sentences, we shall see several grammatical differences between direct and indirect speech that hold across English and French and are

Semiotic Origins of the Mind-Body Dualism 201 representative of SAE: (1)

Direct: Mary said to me yesterday at the station, "I will meet you here tomorrow." Indirect: Mary said to me yesterday at the station that she would meet me there today.

(2)

Direct: Marie m'a dit hier ä la gare, "Je retrouve te ici demain." or "Je retrouverai te ici demain." Indirect: Marie m'a dit hier ä la gare qu'-elle me retrouverait lä (-bas) aujounThui.

One can transform the direct form into the indirect by: 1)

Removing the quotation marks (or the introductory pause between reporting and reported clauses in speech) and inserting subordinating conjunctions such as 'that', 'if, or 'whether' in English, and 'si' or 'que' in French.

2)

Shifting the grammatical persons (personal or possessive pronouns, determiners, and verbal person) from first or second in direct speech to the third person in indirect discourse. (T —> 'she', 'je' —>'elle' in our examples).

3)

The verbs of indirect discourse are subject to concordance of tense rules which affect verbs in many types of subordinate clauses; typically, there is a backshifting of tenses. (This will be discussed in more detail later).

4)

Conversion of deictic elements such as that of the spatial deictics 'here' —> 'there', 'ici' —> 'lä(-bas)', and the temporal deictics 'tomorrow' —> 'today', and 'demain' —> 'aujordTiui'.

202 Benjamin Lee The conversion of direct questions to their indirect counterparts involves the creation of a subordinate clause, the inversion of the auxiliary + subject word order, and the use of the proper subordinating conjunction. Are you going? —> He asked if I was going. Irez-vous? —> Π m'a demandi SJ j'irais. Direct imperatives can usually be converted into a modalized subordinate clause or an infinitival clause in French and English. Leave! —> They were told to leave; They were told that they should leave. Prenez-le! —> II ordonna qu'on le prenne; II ordonna de le prendre. The conversion of tense rules are similar in Latin, English and French. Of particular interest for later discussion is that the sequence of tense rules in indirect discourse are also characteristic of indirect thought. The information on Latin (taken from Woodcock (1959)) is different from that of the other languages since the only available information is that found in written texts. In Latin, one can directly quote the words or thoughts of someone else by using a parenthetical subject and a verb of speaking, as in the following example: Animus aeger, ut ait Ennius, semper errat — "A sick mind," as Ennius says, "always errs" (Woodcock 1959: 214). Another possibility is to treat the words or thoughts as objects of a verb of speaking — "Animus aeger," inquit Ennius, "semper errat." The verb which usually introduces direct quotation is the defective verb 'inquam' which can take a dative of the person to whom the quoted words are addressed. "En," inquit mihi, "haec ego patior quotidie."— "But she, within our hearing, said "I am only a stranger here" (Woodcock 1959: 215). Other verbs such as 'loquor' and 'dico' do not usually directly introduce direct discourse, but may have a pronoun such as 'haec' or 'talia' as an object to which the

Semiotic Origins of the Mind-Body Dualism 203 quoted words are placed in apposition, or may use a form such as 'ita' meaning 'so', 'thus' or 'as follows'. tum T. Manlius Torquatus ita locutus fertur: "..." Then T. Manlius Torquatus is said to have spoken as follows: "..." (Woodcock 1959:215). Latin differs from English and French in that indirect statements are formed by placing the subject of the subordinate clause of indirect discourse in the accusative and placing the main verb of that clause in the appropriate form of the infinitive (subordinate clauses in the represented sentence go into the appropriate form of the subjunctive). (Is) iuvit earn. — He helped her. Dicunt eum iuvisse earn. — They say that he helped her. The rules for tense are that, regardless of the particular tense of the main verb: 1) 2) 3)

The present infinitive indicates the same time as the of the main verb. The perfect infinitive indicates time before that of the main verb. The future infinitive indicates time after that of the main verb.

The following sentences exemplify these rules: Dicunt iuvare earn. — They say that he is helping her. Dicunt eum iuvisse earn. — They say that he helped her. Dicunt eum iuturum esse earn. — They say that he will help her. Dixerunt eum iuvare earn. — They said that he was helping her. Dixerunt eum iuvisse earn. — They said that he had helped her. Dixerunt eum iuturum esse earn. — They said that he would help her. Dicent eum iuvare earn. — They will say that he is helping her.

204 Benjamin Lee Dicent eum iuvisse earn. — They will say that he helped her. Dicent eum iuturum esse earn. — They will say that he will help her. In English and French there are regular tense shifts in indirect discourse which obey a general sequence of tense rules for subordinate clauses in each of these languages. Thus in English, corresponding to the following sentences in direct speech (taken from Jespersen 1965:292): 1) 2) 3) 4) 5)

I am ill. I saw her the other day. I have not yet seen her. I shall soon see her, and then everything will be all right I shall have finished by noon.

Indirect discourse has these shifted tenses due to the past tense of the introductory verb: He said that: 1) he was ill. 2) he had seen her the other day. 3) he had not seen her yet. 4) he should soon see her, and then everything would be all right. 5) he should be finished by noon. Similar rules hold for French in which verbal time is calculated from the point of view of the narrator. If the introductory verb is in the present or future, there is no change in tense: Π declare: "Je t'aiderai." — Π declare qu'il m'aidera. If the introductory verb is in the past tense, then the present tense shifts to the imperfect, the passe compose to the plus-que-parfait, and the future to the conditional:

Semiotic Origins of the Mind-Body Dualism 205 II declara: "Je te vois." — II declara qu'il me voyait." II declara: "Je t'ai vu." — II declara qu'il m'avait vu. II declara: "Je te verrai." — II declara qu'il me verraiL If the quoted speech is in the subjunctive, then the present becomes imperfect and the past changes over to the plus-que-parfait in the more stylized forms: II declara: "J'irai ä Paris avant ton depart." — II declara qu'il irait ä Paris avant mon depart. II declara: "J'irai ä Paris avant que tu ne sois parti." — II declara qu'il irait ä Paris avant que je ne sois parti. Besides these general conversion rules, indirect discourse does not allow forms not allowed in subordinate clauses in general. These include the following English and French examples (taken from Banfield 1982: 28-34; an asterisk indicates that the form is ungrammatical): (1)

Subject-auxiliary inversion "Can you tell me the way to Regent's Park Tube Station?" inquired Maisie Johnson. •Maisie Johnson inquired whether could they tell her the way... Maisie Johnson inquired whether they could tell her the way...

(2)

Fronting of topics for emphasis "Absurd, she is," Clarissa insisted. Clarissa insisted that she was absurd. •Clarissa insisted that absurd, she was.

(3)

Right dislocation She exclaimed, "How awful they are, women\" *She exclaimed that they were awful, women.

(4)

Proposing of directional adverbs He shouted, "Away I must go." *He shouted that away he must go. He shouted that he must go away.

206 Benjamin Lee Certain constructions, while acceptable in direct discourse are unacceptable in the subordinate clauses of indirect discourse: (1)

Exclamations "Yes, this is love," Constance sighed. ""Constance sighed that yes, that was love. •Charles s'exclama qu' (*ah!) elle serait jolie.

(2)

Verbless exclamatory constructions Miss Brill laughed out loud: "No wonder!" •"Miss Brill laughed that no wonder.

(3)

Exclamatory sentences Laura blurted out: "How very nice workmen are!" •"Laura blurted out that how very nice workmen were.

(4)

Repetitions and hesitations Egbert protested, "But, but — I am almost the unnecessary party." •Egbert protested that but, but — he was almost the unnecessary party.

(5)

Incomplete sentences "Not for the Queen of England," said Mrs. Ramsay emphatically. •Mrs. Ramsay said emphatically that not for the Queen of England!

(6)

Subjectless imperatives Mr. Chubbs repeated: "Excuse me." •Mr. Chubbs repeated that to excuse him.

(7)

Direct Address The private answered, "Sir, I cannot carry out these orders." The private answered that (•Sir) he couldn't carry out these orders.

Semiotic Origins of the Mind-Body Dualism 207 (8)

Addressee-oriented adverbials: *between you and me', 'frankly', 'if you ask me', 'candidly', 'confidentially' John said, "Between you and me, she is lying." •John said that (""between him and her) she was lying.

(9)

There are also some contextual differences between direct and indirect speech. Some communication verbs cannot introduce indirect speech — 'query', 'intone', 'think aloud' — while others can take only indirect discourse — 'recommend', 'reveal', 'mention'. Of particular interest is the fact that the verb 'think' in English can take quotative complements and thus seem to behave like communication verbs. They also can take reflexive dative indirect objects, and 'think', like 'say', can appear in the progressive; in such contexts, both expressions might be better conceived of as verbs of self-communication. I thought (to myselfX "I am tired." Reinforcing this finding is that in French there is no form 'se penser' corresponding to the English forms, and thinking as self-communication would probably be translated as 'se dire'.

(10)

Only direct speech can be introduced by sentences containing a deictic expression which is preferential with the quoted speech. John made this comment: "The wine is overpriced." John made the (*this) comment that the wine was overpriced.

A major difference between direct and indirect speech is their treatment of expressive elements, deictics, and tense. In the case of direct speech, all deictics, expressive elements, and tense elements refer to the moment of speaking of the quoted speaker (the 'source'), as demonstrated by the first example, "Mary said to me yesterday at the station, Ί will meet you here tomorrow.'" In indirect discourse, these elements refer to the speech event of the quoting speaker. Even those cases where there are expressive elements in the subordinate clause of indirect

208 Benjamin Lee discourse are interpreted as expressing the quoting speaker's attitudes, as in these examples taken from Banfield. John said that the idiot of a doctor was a genius. Oedipus said that Momma was beautiful. Of particular importance for the later arguments is that direct speech, when interpreted propositionally, has what philosophers have called a de dicto reading, while indirect discourse has both a de dicto and a de re reading. a. Oedipus said that his mother was beautiful. b. Oedipus said, "My mother is beautiful." Sentence (a) has two possible interpretations. In one case, the philosophers' de re case, the speaker has identified someone as Oedipus' mother and reported that Oedipus said of this woman that she was beautiful; it would be possible that the speaker is representing a situation in which Oedipus did not know that the woman of whom he was speaking was his mother even though the speaker did. The other de dicto possibility is that Oedipus said something similar to the quotational form, "My mother is beautiful." Whereas sentence (a) has both interpretative possibilities, sentence (b) has only the latter interpretation. The difference between direct and indirect speech also indicates a relationship between the de dicto and de re distinction and types of reports about speech. In the first, de re, interpretation of sentence (a), thereporteris committed to (presupposes) the existence of someone he identifies as Oedipus' mother — the sentence might be interpreted as "Concerning Oedipus' mother, Oedipus said of her that she was beautiful." In the second (de dicto) interpretation the reporter merely reports what Oedipus said without indicating any commitment to the existence of Oedipus' mother. The de re and de dicto interpretation of indirect discourse also apply to the interpretation of sentences involving intentional state verbs and other verbs of consciousness. As John Searle points out (1983: 216-217), similar considerations apply to propositional attitudes. For example, if Ralph believes that the man in the brown hat is a spy, we can report this belief in two ways. The first, corresponding to the de re interpretation, might be "About the man in the brown hat Ralph believes he is a spy," which commits the representing speaker to the existence of the man in the brown hat The second case would be akin to the de

Semiotic Origins of the Mind-Body Dualism 209 dicto reading — "Ralph believes that the man in the brown hat is a spy" — in which the representing speaker is committed to representing what Ralph believes, but without any commitment to whether there is a man in the brown hat. For Ralph, however, the distinction between de re and de dicto collapses. There does not seem to be a truth functional difference between Ralph uttering either the sentence "About the man in the brown hat I believe that he is a spy" or "I believe that the man in the brown hat is a spy". In reviewing the data so far, we can see that in English, French, and Latin there exists a "cline" of interpretations of speech and thought which range from the most contextualized (direct speech) to the most "decontextualized" (the de re reading of indirect discourse). In all three languages (the findings hold for other European languages as well), the indirect forms have the possibility of both de dicto and de re readings in which the de dicto interpretation is used to represent a propositionalized interpretation of what the source said or thought while the de re interpretation indicates the speaker's representation of some prepositional truth about the source which can be independent of anything the sourse has said or thinks. The contrast between these two readings suggests the existence of a realm of objective propositions which can be subjectively grasped by individuals. Their relations to quotation and direct speech (reinforced by the forms of represented speech and thought that develop in European narrative discourses) produce a "fashion of speaking" in which speech can be seen as the embodiment of a subjective grasp of what are ultimately timeless objective propositions. Zeno Vendler has extended these findings by his reanalysis of Austin's work on performativity. Although he focuses on English, his results also apply to French and Latin, both of which have performatives (on French, see Benveniste (1971) and on Latin, Woodcock (1959)). Vendler also explicitly uses his work to explicate and support Descartes' dualism, showing how it is a product of these grammatical relations. However, he does not take a Whorfian stance, but instead argues for the universality of the dualism. Although we shall later use Vendler's analysis of performativity to analyze the linguistic origins of Descartes' cogito argument, we will first look at his grammatical analyses of the origin of the Cartesian dualism.

Performativity and the Cartesian Dualism In his posthumous How to Do Things with Words (1962) and an essay entitled "Performative-Constative" (1971), Austin contrasted a performative, the utterance

210 Benjamin Lee of which is the performance of some kind of speech act, with constatives or statements which have the property of being true or false. According to Austin, in saying "I promise to..." in the appropriate circumstances, one has made a promise, and is not merely describing oneself as promising. Such utterances are not true or false, but rather "happy" or not. He hoped that the distinction between performative and constative, although simplistic in many ways and upon close examination hard to maintain, might lead to a "new doctrine; both complete and general, of what one is doing in saying something" (Austin 1971: 22). In How to Do Things with Words, Austin distinguished between implicit and explicit performatives. Although almost any speech act may also be the doing of something, certain constructions make explicit what act is being accomplished at the moment of speaking. In his search for grammatical criteria for isolating these explicit performatives, Austin discovered that they seem to have a reluctance in using the present progressive tense and aspect in expressing their true present. Most verbs use the present progressive, e.g., "I am hitting the ball", to describe what is happening at the moment of speaking. The present non-progressive, "I hit the ball", usually describes some habitual activity. However, the form "I am promising to come" is distinctly odd, and it is the present non-progressive "I promise to come" which indicates what is occurring at the moment of speaking. Although Austin eventually abandoned the search for grammatical criteria for explicit performatives, Vendler felt that Austin's rejection was premature. Vendler notes that performative verbs shared their reluctance for the present progressive and their affinity for the present non-progressive with such mental state verbs as "know', *believe', 'intend', etc. When used in the present non-progressive, they indicate an indefinite time-span which includes the moment of speaking. Explicit performatives have punctual and telic aspectual properties, while their intentional state counterparts are statives. More importantly, Vendler suggests that in addition to their reluctance for the {»«sent progressive, performatives and mental state verbs can both take sentence nominalizations or noun-clauses as complements, the most familiar of which are 'that'-clauses which can introduce referentially opaque contexts. Vendler also identifies another group of verbs which are generally not performative and do not usually refer to speech events but also take prepositional complements. 'Decide', 'realize', 'discover', and 'identify' are examples of what Vendler calls "mental activity verbs" which are all telic, but unlike performatives, they do not take the present nonprogressive except when expressing habitual activity (*"I decide to hit you" versus "I always decide when to let go"). If they do

Semiotic Origins of the Mind-Body Dualism 211 take the present nonprogressive, they lose their habitual meaning and instead are performative — "I realize that you are worried", "I hereby identify him as", "I recognize him as". Vendler then reanalyzes Austin's typology of illocutionary acts to show the different types of sentence complements they can take. For each class of performatives so identified (the different complement clauses are a special kind of grammatical "reactance" in Whorf s terminology), he tries to find corresponding mental state and activity verbs that have the same sentence nominalization complements. For example, Austin's class of expositives includes such performatives as 'state', 'declare', and 'claim'. Their nominalization complement pattern is that of "Nj Vexp that NV+", e.g., "I state that he ran away". The mental state verbs include "know', 'hold', 'think', and "believe', e.g., "I believe that he ran away". After making an exhaustive comparison, Vendler finds only two performative classes with no mental act or state counterparts, one of which Austin called 'exercitives' and the other which Vendler calls "operatives". The exercitives include 'order', 'request', and 'permit', all of which can take an infinitive construction where the subject of the nominalized sentence can appear as the direct object of the performative and the full interpretation of the complement usually reveals the auxiliary 'should': I order you to go = I order you that you should go. The operatives include 'appoint', 'nominate', 'condemn', and 'promote'. Except for the exercitives and operatives, the overlap is complete — for every performative class there is a corresponding class of mental activity or state verbs which takes identical complement patterns. What one wishes, intends, or wants to do, one can also promise, vow or pledge in words. Whatever can be stated, declared, or claimed can also be known, held, or believed. Almost anything that can be said can be thought, while anything that can be thought can be said. Vendler proceeds to use these grammatical relations between thinking and speaking to argue that a Cartesian view of mind is actually our folk theory: If I am accused at this point of holding an essentially Cartesian view of ideas, thought, and speech, and in general, of the human mind, then I must plead guilty to this charge. I do not feel guilty, since, if Professor Ryle is right — and he is in this

212 Benjamin Lee matter — this puts me in the good company of those who profess "the prevalent theory," the "official doctrine." He attributes the prevalence of this doctrine to the enormous influence of Descartes. Here I differ with him, and claim that the "official doctrine is nothing but the commonsense view, and is Cartesian only inasmuch as it has found its clearest philosophical expression in Descartes' works (Vendler 1972: 145). Vendler uses the contrast between performative verbs and their mental state counterparts to provide the framework for the mind-body dualism, with the mental act verbs acting as a bridge — they share telic aspect with the performatives, but generally are not performative when they are put into the first person present nonprogressive. Vendler develops this contrast by examining the auxiliaries used to reverbalize a nominalized verb. One can 'make' a promise, 'hold' a belief, or 'reach' a decision. The auxiliaries most associated with performatives are 'make', 'give', and 'issue', whereas the mental state verbs share "have' (to have a belief) along with a host of interesting alternates. One can 'nourish' a hope, 'cherish' a belief, and 'hide' an intention. What we 'hold' in some mental state, we can 'issue' in some illocutionary act Mental act verbs, as might be expected, lie in between. Some, like 'decide', 'choose', and 'identify', share 'make' with the performatives; but not 'issue', except in 'issuing a decision' (in which case it functions like a performative). Other auxiliaries, such as 'reaching' a decision stress the telic or achievement aspect of these verbs. As Vendler puts it: The total import of the emerging picture is clear enough. Man lives in two environments, in two worlds: as a "body," and "extended thing," he is among objects and events in the physical, spatio-temporal universe: as a "mind," a "thinking thing," he lives and communes with objects of a different kind, which he also perceives, acquires, holds, and offers in various ways to other citizens, to other minds (Vendler 1972: 34). Vendler then proceeds to establish the nature of these "objects of a different kind" which the mind "communes with." These objects turn out to be thoughts or propositions.

Semiotic Origins of the Mind-Body Dualism 213 What we say, in the full sense of this word, is a thought expressed in words, couched in words; whereas the same thought, unexpressed and not coded in words, may be the object of a mental state or a mental act, say of a belief or a realization (Vendler 1972:52). The weak sense of saying is "roughly equivalent to uttering, mouthing, or pronouncing" (Vendler 1972: 25) much as a child might in imitating his parents; the "full sense" of saying is to perform an illocutionary act which, according to Vendler, involves the speaker understanding and meaning what he says. At this point, Vendler invokes Grice's theory of meaning, to show how meaning, intention, and illocution interact. The exact nature of the speaker's intention determines the particular illocutionary force of his utterance. If, for instance, in saying "I'll be there" my intention is to cause you to believe or to expect, by means of your recognition of my intention from these words, that I shall be there, then what I say is intended to have the force of a statement or a forecast. If, however, my intention in saying those words is to entitle you to rely on my going there, then it will have the force of a promise. If, finally, those words are intended to make you fear my going there (say, in order to deter you from doing something against me), then it will be a threat. Accordingly, if the circumstances are not clear, I will attach a sign, usually a performative verb (or an intonation pattern for the threat) to explicitly mark the force. And you, of course, if you fully understand what I say, will know, first, that it is my going there that I, by using these words, intend you to come to expect, rely upon, or fear, and, second, which one of these "illocutionary aims" I wanted to achieve. Whether you, in fact, will come to expect, rely upon, or fear my going there or not is as irrelevant to the uptake of the message as the overt action, if any, you might perform as a result of my statement, promise, or threat (Vendler 1972:62-63). Vendler's argument that what a person says or thinks is a proposition relies upon his interpretation of indirect quotation. When compared to direct quotation, indirect

214 Benjamin Lee quotation minimally involves a tense shift if the speaker is reporting his own speech. If we start with what Vendler calls the "ideal" form in which the utterance contains a performative verb in the present perfect (i.e., non-progressive) with the subject in the first person (Vendler 1972: 56), then the report of such an utterance by the speaker himself will involve putting the performative in the past tense: I promise that... —> I report that I promised that... If we compare the general schemas of which the above is an instance, indirect quotation shares with its target utterance the noun clause, nom (NV+): I V p nom (NV+) —> I Vp nom (Nn+(Vp) nom (NV+)) p=performative h=human t=tense-mark (past). Of course, reporting the speech of someone else shifts N„ from 'I' to some other appropriate human nominal category. Thus the indirect quotation form and the reproduced utterance share a noun clause as complement nom (NV+). Vendler then argues that the noun-clause that appears as complement is not what is said in the "strong" sense, but rather expresses or encodes what is said, that is, a thought or proposition. One set of supporting reasons derives from the various ways one can report speech, changing the forms in the reporting speech while still maintaining referential equivalence. Among nominal categories, such shifts would include the pronominal shifts from Ί' to third person forms. For example, if Joe, a suspect in a robbery case, says to the police, "I stole the watch," the police can claim that Joe had admitted that he stole the watch. If what is said is simply the words that make up the noun-clause, then Vendler claims that if what Joe said was the sentence "I stole the watch," then the police would have no right to claim that what Joe said was that he stole the watch. Similar changes between reported speech and reporting speech also indicate that it cannot be merely words or sentences that one says. For example, the police might use a variety of "paraphrastic transforms" such as "he admitted having stolen the watch", ..."stealing the watch", "that it was stolen by him", ..."that it was he who stole it", etc., and the grammatical freedom in such transforms indicate some extra-linguistic point of equivalence. Vendler's final argument is one from translation: Now surely, if I can contradict in English what Descartes said in Latin, or if I can contradict in English what Descartes said in

Semiotic Origins of the Mind-Body Dualism 215 French, then what Descartes said, and what I say, cannot be a string of words, English, Latin, or French (Vendler 1972: 61). Vendler concludes that: We have to distinguish, therefore, both in the speaker's intention and in the hearer's understanding, the message itself — in our case, that I shall be there — and the illocutionary force with which this message is issued. Moreover, the same content, the same proposition — that I shall be there — may (if I am honest) or may not (if I am not) be the object of my belief, and may become yours if you believe me. Obviously, the same distinction between the message and the force is mirrored in the distinction between the devices of the language that serve to encode the content, and the devices that serve to mark the illocutionary force. To say something, therefore, is to issue a message encoded in a language and marked (explicitly or implicitly) by a performative (Vendler 1972:61). The issue of referential opacity puts a limit on the degree to which the reporting speaker can alter the referential material of the first speaker and still attribute that material to him. For example, Joe might say "Mary is a good cook," where Mary is his wife. It seems legitimate for someone to report that Joe boasted that his wife was a good cook. However, if unbeknownst to Joe, Mary is John's mistress, it seems wrong to report what Joe said was that John's mistress was a good cook. According to Vendler, these points about reported speech and indexicality indicate that the referential factor cannot be entirely abstracted from the speaker's words, mind, and circumstances of utterance. In reporting another's speech, one takes into consideration: ...what he knows, how he thinks about this or that, or at least what he is likely to know, what are the ways he is likely to think about this or that. Thus it is not his actual words that really matter those can be, and in many cases must be overriden; the restriction comes from the limitations of his mind, from the fact that an individual is known to a person only under some

216 Benjamin Lee aspects — be they a series of spatiotemporal appearances or a set of descriptions (Vendler 1972:72-73). According to Vendler, the problems of referential opacity in verbs of speaking, and mental act and state verbs indicate the existence of two types of propositions. Transparent (de re) readings, where intersubstitutability of co-referring terms preserves truth value, point to a realm of 'objective' prepositions; opaque (de dicto) interpretations indicate the subjective limits of our ability to comprehend such entities. This subjectivity is reinforced by egocentric modes of perception and representation of the world ultimately anchored by the indexical specification of objects in space and time. These forms of specification link the Τ to a world outside of thought. Man thus lives in two worlds, that of mind and that of the body. As a body, he enters into causal and perceptual relationships with the material world; as a thinking subject, he communes with a world of facts and possibilities which are reflected in his mind through his subjective apprehension of propositions. At the same time, the thinking subject must be tied to a body and thereby anchored in space and time since the subjectivity of his thought is due not only to ignorance, but limitations imposed by the egocentrism of his own spatiotemporal location (Vendler 1972:85). Vendler's work seeks to establish the validity of a modified Cartesian position by showing how it is compatible with certain linguistic arguments based upon grammatical structures, particularly those relating verbs of speaking and thinking. However, he specifically eschews a Whorfian position: What I have to consider, however, is an objection arising from the Sapir-Whorf hypothesis of linguistic relativism. Partisans of this view might argue at this point that even the world of facts and possibilities does not depend merely upon the nature of the physical world, but also on the quality of the particular language in which the world is reflected to the group of humans speaking that language. For reasons that ought to be obvious to the reader, I do not this accept this theory (Vendler 1972: 88). However Vendler's whole exegesis could itself be seen as a demonstration of Whorf s basic points, suggesting that our standard folk-theory of mind and body results from an interaction between Whorfian-type linguistic considerations, and the use of a propositional model to analyze speech and thought. The latter

Semiotic Origins of the Mind-Body Dualism 217 assumption may be the basis for Vendler's anti-Whorfianism because it seems that Vendler uses his linguistic arguments mainly to support and revise Descartes' argument, rather than to relativize it. The Whorfian considerations buttress his theory of the relation between speech and thought, but they are not used to critically establish how propositionality itself is encoded in English and that the particularities of that encoding partially determine the theory of knowledge he wishes to establish. The arguments which Vendler uses to establish his contrast between a subjective and objective view of propositions derive from his arguments about referential opacity. The de re (also referential and transparent) interpretation supports Vendler's objective reading of propositions, while the subjective reading derives from the de dicto interpretation. The contrast between these two readings provides the linguistic grounds for the Vendler's belief that as a subject, an Ί,' man moves in two worlds. As a subject with a body, he is a physical object in a physical world, subject to all the laws of nature; as a thinking subject, his existence follows from his subjective apprehension of a world of abstract, timeless, and objective propositions. Vendler's grammatical analyses supply a critical step in understanding how the standard Cartesian folktheory is created. He shows that there is a grammatically based fashion of speaking which relates mental state and act verbs and verbs of speaking such that you can say whatever you think and think almost whatever you can say. The development of the contrast between external and internal speech and its interaction with these grammatical relations provides the basis for locating speech as an event in the external world and seeing thought as focused upon an internal world of propositions. Furthermore, the relative ease of transferring propositional complements between verbs of speaking and thinking leads to the inference that external speech is the expression of underlying internal thought, which is in turn the subjective understanding of what can be ultimately translated as a timeless, objective proposition.

Descartes' Cogito Descartes is, of course, the figure who first formulates the modern version of the mind-body problem; the second edition of the Meditations is titled "Meditations on the first philosophy in which the existence of God and the distinction between Mind and Body are demonstrated." The radical separation of mind from body and

218 Benjamin Lee the identification of the ego or "I" with a "thinking substance" has become part of the standard philosophical discussion, and was later explicitly criticized by both Wittgenstein in his arguments against private languages in the Investigations and Peirce (in his 1868 article "Some Consequences of Four Incapacities"). At the same time, Descartes' arguments are also based upon his method of analysis; although the cogito argument is usually identified with his Meditations, it first occurs in published form in his Discourse on Method in which he tries to give the basic conditions that any analysis must follow if it is to establish the truth of propositions we believe in. The model for this "method" is taken from mathematics, particularly geometry, and the Meditations is an explicit application of these rules. Descartes' interest in formal methods for establishing philosophical truths dates back at least to his early adulthood when, at the age of 22 he dreamt that God had destined him to develop a unified science of nature based on mathematics. His first substantial work written, Rules for the Direction of the Mind (1628), already shows a concern with the development of a method of inquiry which will be the basis for all philosophical and scientific reflection — a method which, if strictly adhered to, would yield indubitable knowledge. Already present is the search for "true and evident cognition," "what is completely known and incapable of being doubted," free from any "probability" of error (Descartes 1911: 3). The two certain routes to knowledge are intuition — "the conception which an unclouded mind gives us so readily and distinctly that we are wholly freed from doubt about that which we understand" (Descartes 1911: 7) — and deduction which is "necessary inference from other facts that are known with certainty" (Descartes 1911: 8). The core of the method: ...consists entirely in the order and disposition of the objects towards which our mental vision must be directed if we would find out any truth. We shall comply with it exactly if we reduce involved and obscure propositions step by step to those that are simpler, and then starting with the intuitive apprehension of all those that are absolutely simple, attempt to ascend to the knowledge of all others by precisely similar steps (Descartes 1911: 14). One searches first for "that absolute which contains within itself the pure and simple essence of which we ate in quest" (Descartes 1911:15) and then deduces the

Semiotic Origins of the Mind-Body Dualism 219 other more relative aspects from these absolutes. He applies this method to geometry, the truths of which he claims can be deductively generated from the absolute essence of physical bodies, which is that they possess extension. These are ideas that are considerably refined in his 1637 woik, The Discourse on Method, which seeks to establish indubitable foundations for knowledge. The method to be used is one in which any problem is analyzed into a set of simpler component ideas, a procedure derived from his discovery of Cartesian coordinates which allowed one to analyze complex curvilinear motions as composed of simpler ones. He summarizes his method with four rules: The first of these was to accept nothing as true which I did not clearly recognize to be so: that is to say, carefully to avoid precipitation and prejudice in judgments, and to accept in them nothing more than what was presented to my mind so clearly and distinctly that I could have no occasion to doubt it. The second was to divide up each of the difficulties which I examined into as many parts as possible, and as seemed requisite in order that it might be resolved in the best manner possible. The third was to carry on my reflections in due order, commencing with objects that were the most simple and easy to understand, in order to rise little by little, or by degrees, to knowledge of the most complex, assuming an order, even if a fictitious one, among those which do not follow a natural sequence relatively to one another. The last was in all cases to make enumerations so complete and reviews so general that I should be certain of having omitted nothing (Descartes 1911: 92). Williams (1978) points out that Descartes' search for an indubitable starting point stems from the epistemological question, "what do I know?" This involves analyzing his beliefs and sifting out those that are unlikely. If the sifting-out process is purely reflective (as Descartes insisted), there seems to be a potential paradox. Beliefs can be knowledge only if they are true, but reflections about beliefs would seem to be unable to discriminate just those beliefs that are true since for something to be a belief it must involve a proposition already held to be true. In such circumstances, the only beliefs that could be guaranteed to be true would be those whose truth are guaranteed by the fact that they are believed.

220 Benjamin Lee Genuine, indubitable knowledge must be self-guaranteeing; the question of whether one knows that one knows something (i.e., doubt) would make no sense for an indubitable proposition. Descartes therefore suspends belief in anything which he can in the least way doubt — the whole physical universe, his body, God, the past. He then discovers the indubitable truth of the cogito (the founding intuition) from which he can then deduce the existence of God and the natural world. First he systematically sets aside anything which he can doubt: But because in this case 1 wished to give myself entirely to the search after Truth, I thought that it was necessary for me to take an apparently opposite course, and to reject as absolutely false everything as to which I could imagine the least ground of doubt, in order to see if afterwards there remained anything in my belief that was entirely certain. Thus, because our senses sometimes deceive us, I wished to suppose that nothing is just as they cause us to imagine it to be; and because there are men who deceive themselves in their reasoning and fall into paralogisms, even concerning the simplest matters of geometry, and judging that I was as subject to error as was any other, I rejected as false all the reasons formerly accepted by me as demonstrations. And since all the same thoughts and conceptions which we have while awake may also come to us in sleep, without any of them being at that time true, I resolved to assume that everything that ever entered into my mind was no more true than the illusions of my dreams (Descartes 1911:1,101-102). He is then left with the indubitable cogito: But immediately afterwards I noticed that whilst I thus wished to think all things false, it was absolutely essential that the 'Γ who thought this should be somewhat, and remarking that this truth Ί think, therefore I am' was so certain and so assured that all the most extravagant suppositions brought forward by the sceptics were incapable of shaking it, I came to the conclusion that I could receive it without scruple as the first principle of the Philosophy for which I was seeking (Descartes 1911:1,101).

Semiotic Origins of the Mind-Body Dualism 221 From this he concludes that although he might conceive of having no body or world or place where he might be, he could not conceive that he, the thinking subject, was not: And then, examining attentively that which I was, I saw that I could conceive that I had no body, and that there was no world nor place where I might be; but yet that I could not for all that conceive that I was not. On the contrary, I saw from the very fact that I thought of doubting the truth of other things, it very evidently and certainly followed that I was; on the other hand if I had only ceased from thinking, even if all the rest of what I had ever imagined had really existed, I should have no reason for thinking that I had existed (Descartes 1911:1,101). He then locates the essence of the "I": From that I knew that I was a substance the whole essence or nature of which is to think, and that for its existence there is no need of any place, nor does it depend on any material thing; so that this 'me', that is to say, the soul by which I am what I am, is entirely distinct from body, and is even more easy to know than is the latter; and even if body w o e not, the soul would not cease to be what it is (Descartes 1911:1,101). In the Discourse on Method, we therefore see two of the premises which have become the hallmark of Cartesianism. First, the T is a substance whose essence is to think, and second, that this thinking substance is distinct from any physical body since its existence is guaranteed only by thought. As Williams (1978) points out, this dualism presupposes Descartes' belief that there are only two essential attributes, thought and extension, and all substances must be explained by one of those attributes. All secondary properties of any substance will merely be a mode of its essential attribute, and if the 'I' is a thinking substance, it cannot have any physical properties. In his later work (from the Meditations on), Descartes refines his cogito argument and also makes clearer his commitment to what we would now call a prepositional regimentation of subjectivity. For Descartes, "thought constitutes the nature (essence) of thinking substance" (Descartes 1911: I, 101) and that

222 Benjamin Lee thinking substance "is a thing that doubts, understands (conceives), affirms, denies, wills, refuses, that imagines also, and perceives" (Descartes 1911: I, 153). In Meditation II, Descartes concludes: We must come to the definite conclusion that this proposition: I am, I exist (ego sum, ego existo), is necessarily true each time that I pronounce [my emphasis] it, or that I mentally conceive it (Descartes 1911:1,150). From these and other similar remarks, the cogito argument uses "think" (penser, cogitare) as a generic mental thought verb, so that: I (think)... ρ ..., therefore I am includes the full complement of mental state and action verbs, and could also be extended to verbs of speaking, as the above quote indicates. I (doubt)... ρ ..., therefore I am. I (judge) that p, therefore I am. I (state) that p, therefore I am...etc. As Vendler puts it, Descartes seems to be arguing that one's existence as a thinking substance "follows from the thought or issuance of any proposition whatsoever" (Vendler 1972: 199), even if the proposition is false: If one wants to conclude to one's existence from the sentiment that one breaths, ...even if this opinion is not true...one concludes very well; because this thought of breathing appears to our mind before the thought of our existence, and because we cannot doubt that we have it. And in this sense to say, I breathe therefore I am is no other thing than to say I think, therefore I am (Descartes 1911:11,207). If we take the formula: I think that p, therefore I am

Semiotic Origins of the Mind-Body Dualism 223 and replace 'p' with Ί am', we get the particular form of the cogito argument in Meditation II. This is a "transparent" version of the cogito argument in which the very proposition thought of, i.e., that I exist, is also the conclusion. Normally, when thinking any proposition, the existence of the thinking subject is presupposed and the focus is on what is thought. In the breathing example quoted earlier, "the thought of breathing appears to our mind before the thought of our existence, and because we cannot doubt that we have it" (Descartes 1911: II, 207). In concluding that therefore I exist, the proposition thought about is presupposed and the focus is upon the thinking subject. When the proposition thought about is that I exist, the two moments unite in a quasi-performative moment. In thinking that I exist, I mentally assert the existence of a subjective perspective upon a temporally ordered system of thoughts, the present slot of which is occupied by that very assertion. Furthermore, in asserting the proposition "that I exist," I bring about the very state of affairs it represents, i.e., I bring about my existence as a thinking subject at that moment. Self consciousness arises through my consciousness of my propositional thoughts; by shifting my awareness from their issuance to their issuer, I become "aware that I am, I think, I am a thinking thing" (Descartes 1911: I, 327). Since his basic criterion for the mental is immediate awareness, he also includes sensations, perceptions and the imagination as aspects of the mental, but as the previous discussion makes clear, his prototype for thinking is a propositional attribute; imagination and sensation are treated as mixed modes, being part mind and part body: Imagination...sensation...belong to the soul, because they are kinds of thought [espices de penices]; nevertheless they belong to the soul only insofar as it is joined to the body, and for this reason. They are kinds of thought without which one can conceive the soul (Descartes, trans, in Vendler 1972: 151). Descartes distinguishes three levels in the mixed mode of sensation, the first of which apply to the body and the other two belong to mind. The first level is that of "the immediate affection of the bodily organ by external objects" and is the result of "the motion of particles of the sense organs" (Descartes 1911: 251) and seems to correspond to what we would now call purely causal and physiological processes. The second level is that of "the mind's union with the corporeal organ affected" (Descartes 1911: II, 251), and includes perceptions of such things as pleasure, pain, color, and hunger. This intermediate level seems to correspond to

224 Benjamin Lee what later would be called sense-data which possess a certain level of intentionality in being about something but are not fully propositional. The last level consists of "all those judgments which...we have...been accustomed to pass about things external to us" (Descartes 1911: Π, 251). These judgments correspond to the types of mental acts and states which form the basis of the pure "cogito" arguments. The further one moves from the propositional version of the cogito, the more one moves toward the realm of mixed modes and the body. This is made clear in the following passage: And that is why not alone understanding, willing, imagining, but also feeling, are here the same thing as thought For if I say I see, or I walk, I therefore am, and if by seeing and walking I mean the action of my eyes or my legs, which is the work of my body, my conclusion is not absolutely certain; because it may be that, as often happens in sleep, I think I see or I walk, although 1 never open my eyes or move from my place, and the same thing perhaps might occur if 1 had not a body at all. But if I mean only to talk of my sensation, or my consciously seeming to see or to walk, it becomes quite true because my assertion now refers only to my mind, which alone is concerned with my feeling or thinking that I see andlwalk [my emphasis] (Descartes 1911:1, 222). The interpretation that Descartes views thought as basically propositional is reinforced by his scattered comments on language. According to Descartes, animals do not speak because they have no thoughts to express; if they had thoughts, they would communicate them just as they communicate such feelings as anger or fear. Man is distinguished from the beasts by speech, and his soul is established by his thoughts which are expressed by his speech. However, speech without thought (for example, reading something one doesn't understand) is not true speech, and thought itself is not necessarily verbal. This leads Descartes to state that true speech expresses thoughts which can exist independently of their manifestation in any particular language: For who doubts whether a Frenchman and a German are able to reason in exactly the same way about the same things, though

Semiotic Origins of the Mind-Body Dualism 225 they yet conceive the words in an entirely diverse way (Descartes 1911:11,60). We are now in a position to re-examine Descartes' cogito argument and show how it is based on an analogy with the performativity of the 'dico' verbs. This analogy depends upon the grammatical structuring of verbs of speaking and mental act and state verbs and its resultant objectification. Since, as Vendler has shown, performatives, mental act and mental state verbs share a range of similar grammatical reactances, the basis for an analogical projection from the relatively transparent, creatively self-referential and metapragmatic performatives to their nonindexical counterparts exists. The event-like properties of the performatives are projected upon their 'internal' mental state counterparts, producing an inner world of mental processes and events. The cogito argument is thus a microcosm of the interactions between a propositionally driven mode of analysis, objectification, and linguistic structure; its felt indubitability arises from the total transparency of these moments to one another. Descartes' cogito argument can now be understood as a 'cogito' analogy based on the performativity of 'dico' verbs. If I say "I say that...", the very utterance creates the state of affairs it refers to, thereby guaranteeing its truth. The truth of the utterance entails the truth of the proposition "I exist" and also the existence of the T to which it refers and which it creates. The peculiarity of the argument, which Descartes recognizes when he writes that a person "does not deduce existence from thought by a syllogism, but, by a simple act of mental vision, recognizes it as if it were a thing that is known per se" (Descartes, translated in Vendler 1972: 20), is due to the fact that the premises in most arguments are not creatively selfreferential (they do not create what they describe). Descartes' cogito argument asserts that in merely entertaining any proposition whatsoever, the thinking subject creates the situation it describes, i.e., I think that..., thereby guaranteeing its truth value and allowing one to conclude that the thinking subject exists. If the above argument is correct, the question arises why Descartes did not give a 'dico' argument. First, as mentioned earlier, he seems to come fairly close when he writes in Meditation II "I am, I exist, is necessarily true each time that I pronounce it [my emphasis], or that I mentally conceive it" (Descartes 1911: 150). Second, Descartes seeks an argument for the existence of the thinking subject whose validity cannot be doubted, and the verb 'to doubt', dubito, is a cogito, not a dico verb. Finally, Descartes believed that speech merely encodes thought, so that any dico argument would presuppose its cogito counterpart.

226 Benjamin Lee If Descartes' cogito argument is valid, then it does seem to follow that the thinking subject is created by his thinking about any proposition whatsoever. From this indubitable starting point, one could conclude, as Vendler does: Accordingly, the I may be taken, first, as nothing but the sum total of all of one's thoughts. In this sense I am identical with my mind: l a m a mind. Second, it may be taken to denote the ordering or "generating" principle of the unity of a consciousness, which distinguishes it from all others and is itself not a thought or a sequence of thoughts but their "transcendental" subject. In this sense I have a mind. These two I's correspond, of course, to Kant's phenomenal self and transcendental self. Finally, if we recall that the "function" that gives the individuality of a particular mind itself depends upon the spatiotemporal continuity of a sentient body (not to mention the body's role in action and speech), then we see the reason for yet another sense of the /, the one denotating the body. In this sense I am a body (Vendler 1972: 204). From the indubitable cogito we can thus derive the mind-body dualism.

Conclusion The Cartesian mind-body dualism and the cogito argument have fascinated and perplexed thinkers ever since Descartes first presented them. The appearance of certainty stems from Descartes' apparent discovery of an indubitable proposition and his attempt to deduce from it further indubitable conclusions. This logical regimentation forms the basis for a discursive argument form which focuses upon the referential and propositional functions of language, thereby foregrounding the "cline" of forms ranging from direct speech to the de re interpretation of indirect discourse. The grammatical linkages among these forms allow a propositional content to remain invariant across the various complement constructions; this in turn grounds the dichotomy between an objective world of propositions which the mind can grasp, and a physical world from which it is separated. The cogito argument also involves logical regimentation since it purports to find an indubitable propositon from which others can be inferred. The grammatical

Semiotic Origins of the Mind-Body Dualism 227 linkages between verbs of speaking and thinking allows the "performativity" of verbs of speaking to be transferred to their "cogito" counterparts, thereby allowing thought to be represented as creatively self-referential in the same way speech is. In the very act of thinking I exist, I make true the proposition that I exist. The above arguments do not disprove Descartes' and Vendler's non-Whorfian claims about the universality of their arguments; that would require showing that other cultures lacked these Cartesian premises because they lacked a similar fashion of speaking. On the other hand, they shed doubt on those universalistic claims by suggesting they rest upon what are SAE specific peculiarities in the way SAE grammatically encodes propositionality in speech and thought. The indubitability of internal self-conciousness, rather than being an unmediated premise of all other inferences, may itself be only an inference derived from a semiotic mediation which is co-extensive with thought itself.

References Austin, J. (1962). How to Do Things with Words, J. O. Urmson (ed.). Oxford: Oxford University Press. (1971). Performative-Constative. In The Philosophy of Language, J. Searle (ed.). London: Oxford University Press. Banfield, A. (1982). Unspeakable Sentences. London: Routledge and Kegan Paul, Ltd. Benveniste, E. (1971). Problems in General Linguistics, Μ. E. Meek (trans.). Miami: University of Miami Press. Cohn, D. (1972). Transparent Minds. Princeton: Princeton University Press. D'Andrade, R. (1987). A Folk Theory of Mind. In Cultural Models in Language and Thought, Dorothy Holland and Naomi Quinn (eds.). Cambridge: Cambridge University Press. Descartes, R. (1911). The Philosophical Works of Descartes, S. Haldane and G. R. T. Ross (trans.). Cambridge: Cambridge University Press. Jespersen, O. (1965). The Philosophy of Grammar. New York: Norton. Peirce, C. S. (1931-1958). The Collected Papers of Charles Sanders Peirce, 8 vols., C. Hartshome; P.Weiss; and A. Burks (eds.). Cambridge, Mass.: Harvard University Press. Searle, J. (1983). Intentionality. Cambridge: Cambridge University Press. Vendler, Z. (1972). Res Cogitans. Ithaca: Cornell University Press.

228 Benjamin Lee Whorf, B. L. (1956). Language, Thought, and Reality. Cambridge, ΜΑ: M.I.T. Press. Williams, B. (1978). Descartes: The Project of Pure Enquiry. New York: Pelican Bodes. Wittgenstein, L. (1960). The Blue and Brown Books. New York: The MacMillan Co. Woodcock, Ε. C. (1959). A New Latin Syntax. London: Methuen and Co., Ltd.

Pronouns, Persons, and the Semiotic Self Milton Singer

One Self or Many Selves? Reviews of and commentators on Man's Glassy Essence, friendly and unfriendly, have, almost without exception, made the criticism that a semiotic conception of the self is defective because it neglects or underestimates the multiplicity and diversity of self concepts of individual persons and of social groups.1 The most challenging and clear-cut statement of this criticism will be found in Murray Leafs papers. I quote several lines: "the issue is not whether there is a 'world view' or not, or even whether such a world view contains a conception of I - Thou - iL The question is whether there is only one per person or only one in each community, culture or civilization. Is there an Indian concept of self, or are there Indian concepts of self?" (Leaf in press). To Leaf, the answer to this and the related questions is unequivocal: "In the case of India the simple and clear empirical answer is that there are multiple distinct traditions carrying distinct world views, with distinct concepts of a self. Moreover, there are many reasons for believing this to be true in all communities" (Leaf in press). Leaf softens his criticism by attributing the sources of my position to Simmel and Redfield, and its phrasing in terms of "the self' to Peirce. The underlying difficulty, he suggests, stems from my acceptance of Peirce's criticism of Kant. If only I abandoned that criticism, he believes, my position would fall into line with the great empirical social science tradition which descended from Kant. While it is flattering to be associated with Kant, I have been puzzled by Leafs criticism and the similar line of criticism expressed by Leach, Fernandez, Saberwal and other commentators. I should have thought that a semiotic conception of the self influenced by Peirce's anti-Cartesianism would confront a problem of too many self concepts rather than too few. From this perspective, Peirce belongs with the empirical critiques of the Cartesian concept of the self as a simple permanent mental substance launched by Berkeley, Hume, Kant, William James,

230 Milton Singer and the later symbolic and social interactionists such as Mead, Cooley and Baldwin. Stimulated by Leafs critical questions, I reread some of Kant and Peirce and am ready to concede that the interpretation of Kantianism in Man's Glassy Essence as consisting of empirical and transcendental egos is a bit oversimplified. Leaf distinguishes three conceptions of self in Kant — the self as perceiving subject, the self as the object of a specific perception, and the self as a "consciousness" which is imputed to the perceptions as their unifier. The first two belong, presumably, to the empirical ego, the third is a transcendental concept which organizes possible experience, but does not derive from it. Leaf admits that this concept "does not translate directly into an answer to the question of whether a person or community can present multiple selves. Yet in the larger context of Kant's system the answer becomes an obvious yes..."(Leaf 1986: 20). In his book of popular lectures, Anthropology from a Pragmatic Point of View, Kant (1974 [1798]: 22) distinguishes between the formal element of consciousness, the "I" of reflection or discourse, and the empirical perception, or "I" of apprehension. Since the former is neither noumenal nor phenomenal, but analytic, it represents a fourth conception of the self. In Man's Glassy Essence I did not explicitly differentiate subject and object within Kant's transcendental self or within his empirical self. On checking this point in Peirce, I found that his references to Kant are as complex and differentiated as Leafs. The oversimplification of Kant is mine, not Peirce's (e.g. Peirce 1986: 50-52). Peirce was admittedly indebted to Kant in several respects, as Leaf notes. Kant's analysis of the universality and validity of judgements indicates an important source for Peirce's pragmatic and social theory of truth as that opinion which will be arrived at by an unlimited community of investigators. In Peirce this is closely connected with his social theory of logic and of reality as the object of those opinions destined to be arrived at by the community of scientific investigators. Another positive debt Peirce owes to Kant is the distinction between a representation and its object, as when Kant says that the transcendental self cannot be known as an object, but can be conceived as a representation (Vorstellung). In his earlier writings Peirce used "representation" and "representamen" for "sign" in the definition of semiosis as a triadic relation of sign, object and interpretant. The Kantian source of this usage is not always recognized because Vorstellung is often translated as "idea" (Kant 1965: 344-352; Peirce 1977a: 192-194). In Anthropology from a Pragmatic Point of View, Kant devotes five pages to the power of using signs. This may not appear to be very much in a work of

Pronouns, Persons and the Semiotic Self 231 about 200 pages, yet his analysis of signs enters significantly into the rest of the work in his distinctions between symbolic and conceptual representations; between demonstrative, rememorative, and prognostic signs; and in the definition of thinking as talking with ourselves and listening to ourselves inwardly (by reproductive imagination) (Kant 1974: 65). Not only do such distinctions and definitions anticipate Peirce, but they bring a substantial portion of Kant's anthropology and the Critique within the domain of a semiotic analysis. Kant, however, did not explicitly generalize the conception of the study of signs as a general discipline or method, as Peirce and Locke and the Stoics did (Yngve 1986). Kant seems to have used logical analysis and introspective observation of his own mental processes as his "method." He admonished people, however, not to "eavesdrop" on their own mental processes. Such excessive introspection, he observed, tends to produce "fanaticism" or madness (Kant 1974: IS). Just as Murray Leafs papers sent me back to reread Kant, so Benjamin Lee's papers sent me back to Descartes and to read some of the revisionist interpretations of Descartes such as Zeno Vendler (Lee 1985, 1986; Vendler 1972). And just as the first exercise led me to recognize that Kant distinguished four T s rather than two, so the second increased the number of Cartesian T s from two to six. I was of course familiar with Descartes' dualism of mind and body as distinct mental and material substances, and with his parallel formulations of alleged indubitable intuitions, in Latin as cogito, ergo sum and in French as je pense, done je suis. Drawing on modern analytic philosophy and J.L. Austin's theory of performatives for his subtle analysis of Descartes, Vendler finds three T's: The I may be taken, first, as nothing but the sum total of one's thoughts. In this sense I am identical with my mind: I am a mind. Second, it may be taken to denote the ordering or "generating" principle of the unity of consciousness... In this sense I have a mind. Finally, if we recall that the "function" that gives the individuality of a particular mind itself depends upon the spatio-temporal continuity of a sentient body (not to mention the body's role in action and speech), then we see the reason for yet another sense of the I, the one denoting the body. In this sense I am a body (Vendler 1972:204). Significantly, this updated interpretation of the Cartesian cogito consciously incorporates Kant's distinction between the phenomenal and transcendental egos and adds an explicit parallelism between thinking and saying. Contrary, therefore, to

232 Milton Singer Vendler's announced objective of restoring the "ghostly" mind in the bodily machine, Vendler ends up by letting the "Cartesian cat," as a stream of thoughts, escape from the "Scholastic bag" of a permanent and distinct mental substance. The escapee resembles Humean, Kantian, and even Jamesian and Peircean conceptions of the self more than the conventional interpretations of Descartes. Perceiving that Vendler's Austinian analysis brings the Cartesian dualism close to Whorf s parallelism between verbs of thinking and verbs of speaking, Lee presses the linguistic analysis of Descartes to its logical conclusion. He argues that both the epistemological indubitability of the Cartesian cogito and the dualism of mental and material substances derive from the grammatical parallelism of verbs of saying and verbs of thinking, and from a performative interpretation of the latter. The belief that there is a distinct thinking mental substance doing the thinking is, according to Lee, a "projection" or "objectification" based on the grammatical parallelism between the two kinds of verbs. In his analysis, Lee not only develops a Whorfian interpretation of Descartes but has obviously also been influenced by Wittgenstein, Silverstein and other linguistic philosophers. The interpretation of the Cartesian cogito as a performance was noted by the logician Hintikka, without reference to Austin or to Whorf (Hintikka 1962). Lee's analysis, in any case, increases the number of Cartesian "I"s to six, the three enumerated by Vendler and their three doubles in verbs of saying. Although in his Latin formulations, Descartes usually does not use separate pronouns, occasionally for emphasis he does add them. Lee's semiotic interpretation of Descartes' cogito and Leafs interpretation of the Kantian transcendental self as a postulated concept do not invalidate my interpretation of Peirce's anti-Cartesian, semiotic and social conception of self and person. Man's Glassy Essence may have understated the number of "I"s in Descartes and Kant, but it does not reduce the Peircean concept of self to a singularity, in spite of the use of such phrases as "the self." On the contrary, a good deal of emphasis is placed in Man's Glassy Essence on Peirce's references to the "extended identity" of a person and to the concept of "a loosely compacted person" which embraces several individuals. Peirce rejected the Cartesian individual introspecting thinker and refused to subject his thoughts to the test of universal and systematic doubt. He suggested instead, a developmental theory of consciousness and self-consciousness which depends on social interaction and the acquisition of language. His cryptic dicta "man is a sign" and "my language is the sum total of my self' are related in Man's Glassy Essence to his analysis of signification in terms of an irreducible triadic relation of sign, object, interpretant,

Pronouns, Persons and the Semiotic Self 233 and to the process of communication between a first person ("I"), a second person ("thou"), and a third person ("it") (Singer 1984a: 83-6; 1986). While these analyses and concepts do not necessarily imply a concept of a multiple self, neither do they imply that a person has only one self, both subject and object, and lives in a world in which there is only one conception of a person. Those reviewers and commentators who criticized Man's Glassy Essence for neglecting the empirical plurality and diversity of selves, have also, curiously enough, given excellent descriptions and summaries in their reviews and comments of my semiotic interpretations of Peirce's concept of the self (Fernandez 1986b, Mosko 1984, Saberwal 1985, Watt 1986). Their acceptance of this interpretation is well justified by its many fruitful consequences as I shall try to show in the sections to follow.

"I-It-Thou" as a Unified Structure and Process of Communication Perhaps a consideration of the semiotic self as a unified structure and process of communication will take the sting out of the criticism that it is insufficiently multiple. The phrase "a unified structure of communication" was introduced by L£vi-Strauss to express his vision of a theory that will include the exchange of women, the exchange of words, and the exchange of goods and services within a particular culture, and across different cultures. Mindful of Livi-Strauss's leanings towards a Saussurean interpretation of linguistic structures, Jakobson qualified the hope for a unified structure of communication by pointing out the need for adding the semiotics of sheer messages and subsidiary disciplines (L£vi-Strauss 1976; cf. Singer 1984a; 70,78). While L6vi-Strauss's vision of a grand theory of communication that will integrate social anthropology, linguistics and economics is challenging, it may be possible to realize something of the spirit of such a vision by analyzing the conception of a self as a semiotic structure and process of communication. Peirce himself has left several important clues for such analysis: his definition of signification in terms of a triadic relation of sign, object and interpretant; of communication in terms of a pentadic relation, which would include the interpreter and utterer of a sign together with the previous triadic relation; and of the "extended identity" of the self as a "loosely compacted person" embracing several individuals. By singling out the triadic relation of the pronouns "I," "it," and "thou," Peirce identified a nuclear structure and process of communication with a potential for

234 Milton Singer being interpreted in terms of linguistics, kinship, economics, and psychology, and much else. I shall briefly outline how such interpretations of the pronominal triad, "I, it and thou" can be developed (cf. Singer 1984a: 83-86, 1986; Fisch 1982: xxvii-xxx; Boon and Schneider 1974). Max Fisch (1982) has connected Peirce's 1859 and 1861 manuscripts on I, it, and thou with the search for a new set of categories. Fisch's observations on these manuscripts provide a useful point of departure for any attempt to interpret these three pronouns as a unified structure of communication and also as a structure of a semiotic self. Fisch observes that Peirce used I, it and thou as the names for the three categories until 1867, when he introduced the less colloquial names Firstness, Secondness and Thirdness. Peirce's three categories, Fisch also notes, were the results of an intensive effort by Peirce to reduce Aristotle's and Kant's tables of categories. Peirce's 1861 manuscript is titled I, It, and Thou, A Book Giving Instruction in some of the Elements of Thought and refers to I, It and Thou as "simple conceptions" incapable of definition in terms of one another, yet related to one another. Fisch adds that "if...we expect the categories to manifest themselves in language as well as in thought, it may strike us that in the language we speak there is nothing more prominent than the three persons of the verb and the corresponding pronouns" (Fisch 1982: xxviii). Several linguists and linguistic anthropologists have observed that the triad of the first, second and third persons forms a linguistic and semiotic universal structure, with the first person as speaker, the second person as addressee, and the third person as object spoken of (cf. p. 265 below). Not all languages, however, symbolize the three persons by three separate pronouns as English does. Some languages may symbolize the category of person by inflecting the verb as Latin does, although for emphasis Latin may also use the separate pronouns. When Peirce replaced I, it and thou as names for his three categories by the more technical names Firstness, Secondness and Thirdness, he did not lose interest in the three pronouns. As Fisch points out, Peirce wrote the 1891 Century Dictionary definition of Tuism as the doctrine that "all thought is addressed to a second person or to one's self as a second person" (cf. Singer 1984a: 85,86). In other words, Peirce's youthful use of I, it and thou for his three categories did not exhaust his interest in the three pronouns as kinds of words which also expressed kinds of persons, thoughts, impulses, actions, and feelings. Fisch's citation from an 1862 entry in one of William James' notebooks suggests some of the wide-ranging implications of Peirce's early interest in the pronouns and presages James' later division of the self into an "I" and a "me":

Pronouns, Persons and the Semiotic Self 235 The thou idea, as Peirce called it, dominates an entire realm of mental phenomena, embracing poetry, all direct intuition of nature, scientific instincts, relations of man to man, morality, etc. All analysis must be into a triad: me and it require the complement of thou (Fisch 1982: xxix). In order to show how I, it and thou can constitute a "unified structure of communication" and a structure and process of a semiotic self, it is necessary to make some reference to Peirce's general theory of signs or semiotics. For Peirce, semiotics was a classificatory science like chemistry and biology, as Fisch (1982: xxviii-xxix) reminds us: "we are starting with words, and, among words, with those associated with the three persons of the verbs, and with the names /, Thou, It for those persons." In constructing a Peircean semiotic interpretations of the self, I do not intend to reconstruct the chronological development of Peirce's thought. All I can claim for the following construction is that it draws on some of Peirce's classifications and definitions of signs, and on some of his colloquial illustrations, to bring together an interpretation that is generally consistent with his semiotic approach, although not always explicitly formulated by him. Peirce calculated that there are sixty-six different kinds of signs according to his definitions and criteria. He reduced this number to ten by eliminating unrealistic possibilities. Of these ten kinds we shall discuss only two which are most relevant and important for our problem. These two are the classification of signs according to their respective relation to their objects as iconic, indexical, and symbolic signs, and the classification of signs as they are in themselves as rhemes, dicents, and arguments. These two classifications are well known from ancient times, although not in Peirce's odd technical vocabulary or as part of a comprehensive, explicit and systematic theory of signs. The iconic, indexical, symbolic classification is more familiar as mimetic, denotative, symbolic. For Peirce the difference between these three kinds of signs depends on whether the sign resembles its object, as in a picture or diagram (an iconic sign), or points to its object as in a pointing finger (an indexical sign), or is conventionally associated with its object as the word "dog" (a symbolic sign). Peirce's usage in the latter case departs from the more popular usage of "symbol" as a figurative or iconic sign. He justifies this deviation in usage by claiming that it follows the classical Greek usage of "symbol."

236 Milton Singer The triad of rheme, dicent and argument may sound a bit arcane but it is Peirce's semiotic translation of the triad long familiar in logic and rhetoric as term, proposition and argument, and in grammar as word, sentence and paragraph. Peirce's classification of terms, propositions, and arguments as kinds of signs enables him to incorporate all of traditional and modem logic as part of semiotics. The most important feature of this incorporation is his classification of arguments and their relations to modes of inference as deductive, inductive, and abductive. Deductive inference concludes what must be, given a particular set of premisses; inductive inference concludes what probably will be, given premisses which describe what actually is; and abductive inference concludes what may be on the assumption of a particular hypothesis which explains the facts. The general result of all this is a kind of semiotic redefinition of the trivium into speculative grammar, "the formal conditions of symbols having meaning," logic, "the formal conditions of the truth of symbols," and speculative rhetoric, "the formal conditions of the force of symbols or their power of appealing to a mind" (CP 2.227-29; 1978). Charles Morris' tripartition of semiotics into syntactics, semantics, and pragmatics was probably suggested by Peirce's trichotomy but is not quite identical with it (Singer 1984a: 21-31,101,108). When Peirce's two triadic classifications of signs are combined, we can begin to see the fruitfulness and some of the subtlety of Peirce's analysis. Iconic signs, for example, taken by themselves, do not assert anything and do not form propositions or make statements. When combined with indexical and symbolic signs, however, they assert properties and relations of particular subjects, and thus constitute grammatical "texts." Peirce's example of a portrait is a colloquial illustration of how a nonverbal icon can become part of a sentence: Now the object of an Icon is entirely indefinite, equivalent to "something." ...A pure picture without a legend only says "something is like this." ...To attach a legend to the picture makes a sentence...analogous to a portrait we will say of Leopardi with Leopardi written below iL It conveys its information to a person who knows who Leopardi was, and to anybody else it only says something called Leopardi looked like this (CP 8: 183; cf. Singer 1984a: 107). Proper names like "Leopardi," pronouns and definite descriptions are classified by Peirce as indexical signs which point to the subjects and objects of a proposition. In another of his colloquial illustrations Peirce shows how such indexical signs can

Pronouns, Persons and the Semiotic Self 237 combine with iconic and symbolic signs to designate a complex human relationship in propositional form: in the sentence Έζβίαεί loveth Huldah1 Peirce writes that "the effect of the word loveth' is that the pair of objects denoted by the pair of indices Ezekiel and Huldah is represented by the icon, or the image we have in our minds of a lover and his beloveth" (Singer 1984a: 60). The indexical and denotative function of proper names does not, in Peircean semiotics, prevent such signs from also having iconic and symbolic functions, as Peirce makes quite explicit in the following statement from one of his late papers: The pragmaticist grants that a proper name (although it is not customary to say that it has a meaning) has a certain denotative function peculiar, in each case, to that name and its equivalents; and...he grants that every assertion contains such a denotative or pointing-out function. In its peculiar individuality, the pragmaticist excludes this from the rational purport of the assertion, although the like of it, being common to all assertions, and so, being general and not individual, may enter into the pragmaticistic purport (Peirce 1955: 263; cf. Singer 1984a: 102). Returning to the pronominal triad of I, it and thou, we can now begin to analyze it as a structure of communication. As separate signs, the three pronouns point to particular subjects and objects but do not constitute a structure, or a sentence. Only when the pronouns are combined with icons and symbols will they make an assertion of some kind. If, for example, we add the verb "give," we can form the sentence "I give it to you." Similarly, we can form other sentences from these three pronouns by adding other appropriate verbs. Suppose we start to think of the three pronouns as designating the terms or subjects of a potential three-term relation; we can then say that the three pronouns taken together constitute an iconic diagram which will become a statement when the appropriate relation is specified. Such a diagram with the three pronouns and the relation to be filled in will be called an "iconic diagram." By adding the appropriate verbs to the pronouns I, it and thou, we can convert the iconic diagram into sentences in economics ("I gave it to you in exchange for a copper"), social anthropology ("I will give her to you in exchange for the bride price"), linguistics ("I will translate it for you"), psychology and poetry ("I will give it up for you"), and into sentences in other disciplines. In other words, to consider I, it and thou as an iconic diagram is to consider the three pronouns as an ordered triplet of terms, each of which designates or denotes a

238 Milton Singer person, the three persons being related to one another by a triadic relation otherwise unspecified. Interpretations of the triadic relation can come from the domains of kinship, economics, linguistics and other domains. The linguistic universal of speaker, addressee, and person and object spoken of constitutes one of the interpretations of the triadic relation. Instead of thinking of the triad of pronouns as a potential set of sentences, it is perhaps more realistic to say that the sentences are converted into iconic diagrams when their predicates are "erased" and only the subjects remain with the blanks. At least this is a mode of phrasing Peirce uses in describing his system of existential graphs (Peirce 1977a: 197-201). My interpretation of pronoun clusters as iconic diagrams has also been suggested by Peirce's explicit conception of an icon as a diagram: A concept is not a mere jumble of particulars; — that is only its crudest species. A concept is the living influence upon us of a diagram, or icon, with whose several parts are connected in thought an equal number of feelings or ideas. The law of mind is that feelings or ideas attach themselves in thought so as to form systems (CP 7.467; Singer 1984a: 101).

I f / , it, and you are considered as a triad of interrelated signs and not simply as separate pronouns, then they can be analyzed as iconic diagrams of three subjects or terms of a triadic relation with the relation or predicate unspecified, e.g. "I — it — you." When the relation in such an iconic diagram is specified, the diagram is converted into a sentence, say, "I gave it to you." The difference between the diagram and the sentence is that the sentence makes a statement about the existence of a specific relation between actual subjects, while the diagram represents only the possible existence of a relation between the subjects, as in Peirce's theory of existential graphs. In Peirce's classification of signs, the sentence includes symbolic signs which depend on conventional associations between the signs and their objects, as well as indexical and iconic signs whose relations to their objects is "natural." The explicit analysis of /, it, thou, as an iconic diagram is not frequently encountered in Peirce's writings. The only explicit reference known to me is the one Max Fisch sent me in 1979 together with the photocopied page from Peirce's 1859 listing of English pronouns under I, It, and Thou, and several 1861 manuscript pages for a projected book, I, It and Thou, A Book Giving Instruction

Pronouns, Persons and the Semiotic Self 239 in some of the Elements of Thought (cf. Singer 1984a: 82-86; 1986: 359-60). Fisch (1982: xxvii-xxxii) and Sebeok (1983) are among the few commentators who have discussed the relation of the three pronouns to Peirce's three categories. An implicit use of I, it, and thou as an iconic diagram does occur when the triadic relation of the three pronouns is variously interpreted as a relation of three categories, of signification and communication, of kinship and personality structure, of historical ages. In making such interpretations explicit we not only extend some of Peirce's suggestions, but also indicate how Livi-Strauss's aspiration to unify social anthropology, linguistics and economics into a single structure of communication can be partly realized through an analysis of the use of pronouns in discourse. The interpretation of I - it - you, and of other pronoun clusters, as iconic diagrams is a natural extension of Peirce's classification of signs. Diagrams are classified by Peirce as one of three kinds of icons; images or pictures and metaphors are the other two kinds (cf. Singer 1984a: 101,108). Underlying Peirce's classification of diagrams as icons is his idea that the relations of the parts of a diagram to one another are homologous with the relations of the parts of the object represented by the icon (cf. Livi-Strauss's postulate of homology between two systems of differences). A geometrical diagram, a map, and an architectural drawing are some of the illustrations cited by Peirce of iconic diagrams. With respect to the use of diagrams in mathematics and logic, Peirce emphasized that the observation of such diagrams can lead to the discovery of truths previously unknown (Peirce 1955: 106-107, 135-149; 1977b). In this respect his contributions recall the efforts of Leibnitz and Bishop Wilkins and also anticipate, especially in Peirce's theory of existential graphs, the modern movement of Peano, Frege, Russell and Whitehead, Hilbert, Tarski, Carnap, Goedel and Quine to formalize mathematics and logic.2 Russell's emphasis on contextual definitions and Tarski's on the semantic primacy of sentences over terms, and of the concept of "true sentence," as well as Goedel's demonstrations of undecidable sentences in formal systems, have revealed important limitations to the formalization program (cf. Quine 1981: 143-55; Singer 1940).

Some Egotistical, Idistical, and Tuistical Americans If I, it, and thou and their corresponding pronouns and verbs in other languages are a linguistic and culturally universal expression of a nuclear act of communication

240 Milton Singer and signification, a diversity in usage, norms and emphases may nevertheless exist in different languages and cultures and in different historical periods. The belief that the generation of the 1960s and immediately following has been the "me" generation is still widely prevalent and has not been made obsolete by the selection of Lionel Richie's song "Say You, Say Me" as the top movie song of 1986. Observers of popular usage, on the other hand, have begun to comment on what seems to be a trend towards the increasing use of the third person, instead of the first or second. Reflecting on some of these trends in popular usage, a thoughtful columnist links them to an increasing depersonalization and dehumanization of machine civilization. I suppose there's a reason for our reluctance to focus on the human factors. During recent decades, we have all become more conscious of the centralization of danger. We know that more lives hinge on fewer "things": on nuclear missiles and plants, on chemicals and computers. It may be easier to think of "systems" that can be perfected than on people who aren't perfectible. Again and again, in the wake of a catastrophe, we look for solutions that will correct "it" rather than "us" (Ellen Goodman, Chicago Tribune 1/28/87). Before we conclude that an "it" generation has succeeded the "me" generation, we should consider the question in a broader historical and cross-cultural perspective. In a youthful 1863 oration on "The Place of Our Age in the History of Civilization," Peirce quotes Emerson's "things are in the saddle, and ride mankind" to support the generalization that the age of scientific material progress is an "idistical" one which has succeeded an "egotistical" age, and will be followed by a "tuistical" age (Peirce 1982: 113). These three adjectives were not coined by Peirce, but were used by Coleridge, according to Fisch, to signify an excessive use of the three pronouns (see illeism in the Oxford English Dictionary and Fisch 1982: xxix). For Peirce, the three adjectives are also clearly associated with the use of the three pronouns "I," "it" and "thou," but now used as names for three historical ages instead of names for his three categories of being and consciousness. The use of the three pronominal adjectives to characterize historical eras does not depend simply on a statistical count of the relative frequency with which the first, second and third person pronouns or verbs are used. The characterization of historical eras depends rather on the cultural norms and standards governing

Pronouns, Persons and the Semiotic Self 241 grammatical usage, and the meanings and ideologies symbolized by that usage. The Royal Society's 17th Century resolutions for "a plain and clear style" set new standards for speech and rhetoric just as Quaker religious and democratic ideology prescribed the use of thou/thee for the second person singular instead of ye/you (Silverstein 1985: 242-252). Coleridge's tuism was probably registering a protest against the excessive use of "thou" and "thee," and Byron's flip lines called it a Truism: To hail her with the apostrophe — "O thou!" Of amatory egotism the Tuism, Which further to explain would be a truism. (Don Juan, XIII) For Peirce, by the end of the nineteenth century, "tuism" had become an expression for an evolutionary ideal of love. In the useful concordance to the Beatles' songs of 1962-1970 by Colin Campbell and Allan Murphy (1980), it is interesting to find that the most frequently used pronouns are you (1,394) and me (1,268). She is in third place with a frequency of 955 and I is fourth with 786. It is in fifth place (446) and he is a poor sixth (83). Among the plurals we has a frequency of 279 and they of 140. The significance of these relative frequencies is not transparent. While advocates of a "me" generation can claim support for the high frequency of mes, advocates of a mistical "you" generation can claim even stronger support, and feminists might hail the considerable predominance of shes over hes. Do the 446 its perhaps signal the emergence of an idistical trend? These implications and questions obviously cannot be answered by reference only to the relative statistical frequencies of the different pronouns alone. Consideration must also be given to such other relevant factors as the popularity of the songs and their interpretations, the conventions of a particular genre such as the lyric, historical changes in mood and philosophy of the singers, the influence of producers, promoters, and other significant persons. Campbell and Murphy, for example, point out, as have other commentators, that in 1965-67 the Beatles changed from popular love songs in which "you" and "me" predominated to more sophisticated songs in which nature, complex musical arrangements, and more reflective attitudes toward life became evident. This change is associated, according to Campbell and Murphy, with the Beatles' overwhelming popular success, the

242 Milton Singer death of their manager Brian Epstein, their travels to India and to Mahesh Maharishi's ashram, among other experiences. The relation of such changes to changes in the relative frequencies of particular pronouns is not only a matter of relative frequency counts or of direct correlations. The Beatles' song "Let it be," for example, which was released on May 8,1970 in general fits the trend of changes described by Campbell and Murphy — more reflective and impersonal than the earlier love songs. The words of the song, however, say that "Mother Mary," not Maharishi Mahesh Yogi, "comes to me speaking words of wisdom, let it be." On the same day, moreover, George Harrison's song "I Me Mine" was also released. While hardly a simple love lyric, this song would be difficult to interpret as an impersonal, nature song, as would "All You Need Is Love," which was released in 1969. In order to draw inferences about persons and periods from the statistical relative frequencies of pronouns and other words, it would obviously not only be necessary to consider such factors as popularity, genre, historical changes and other singers, but also to assess the net result of their combined operation. Campbell and Murphy refer to the generalization that the Beatles' songs are oriented to the present and express the feelings of a "now" generation. The relative frequency of now (165) as compared with the frequency of yesterday (13) and tomorrow (12); be (426) and is (317) as contrasted with was (112) and will (131) — all these comparisons would seem to confirm such a generalization. At the same time, these authors point out the strong nostalgic strains in such songs as "Penny Lane" and "Yesterday." They do not explain, either, why "Yesterday" is the most popular Beatles' song. Statistical concordances of pronouns and other words used in songs, poetry, and other literary genres could benefit greatly from Peirce's emphasis on triadic, dyadic and other relations between words. From this point of view, it is not the relative frequency of occurrence of isolated pronouns that is significant, but the occurrence of such dyads as "I - you," "you - me," "we - they," or triads such as "I - it - you." These dyads, triads and n-ads constructed from them, and their iconic diagrams, may help to reveal the underlying structures of communication and signification which form and express the semiotic self (cf. Varenne 1983: 186-8,194-5). Consideration of linguistic, religious, political and philosophical ideologies associated with pronominal and person — deictics is certainly relevant for the interpretation of such usages in other cultures. Among the Northern Cheyenne,

Pronouns, Persons and the Semiotic Self 243 Straus finds an emphasis given to second person constructions and to the "I - you" contrast which sounds somewhat similar to Walt Whitman's Leaves of Grass. One of the most interesting examples of "egotistical" and "mistical" uses of pronouns in American English is Walt Whitman's Leaves of Grass (1855). Even so discerning a critic as Henry James, in his youth saw nothing more in Whitman's poems than "brag," aggressive egotism, and a "blowing of his own trumpet." This interpretation was suggested to James and other readers by Whitman's frequent use of the first person pronoun "I" (1865 in Miller 1969). Henry's brother, William, took a kindlier view of Whitman's poems and said they expressed "a religion of healthy mindedness" (in Miller 1969: 121): The only sentiments he allowed himself to express were of the expansive order; and he expressed these in the first person, not as your mere monstrously conceited individual might so express them but vicariously for all men, so that a passionate and mystic ontological emotion suffuses his words, and ends by persuading the reader that men and women, life and death, and all things are divinely good (in Miller 1969: 121). William James' suggestion that Whitman's use of the first person pronoun expresses a deep religious and democratic philosophy was also affirmed soon after the publication of Leaves of Grass by Emerson (1855), Thoreau (1856) and others. Emerson described Leaves of Grass as "a remarkable mixture of the Bhagavad Geeta and the New York Herald," and Thoreau called Whitman "the greatest democrat the world has seen" (See Miller 1969). Although a continuing stream of devotees beginning with his physician and biographer, Richard Maurice Bucke, attributed a "cosmic consciousness" to Whitman and compared him to Socrates, Jesus and Buddha, Whitman himself was a bit more modest: "I do not teach a definite philosophy — I have no cocked and primed system — I but outline, suggest, hint — tell what I see then each may make up the rest for himself...including all philosophies as I do, how could I nail my self to any one, or single specimen" (in Miller 1969: xiii). On another occasion Whitman gave a simple and lucid description of the personal orientation of his poetry: Most of the great poets are impersonal....I am personal....In my poems, all revolves round, concentrates in, radiates from myself. I have but one central figure, the general human personality typified in myself. But my

244 Milton Singer book compels, absolutely necessitates, every reader to transpose himself or herself into the central position, and become the living fountain, actor, experience· himself or herself, of every page, every aspiration, every line (in Holloway 1949: 583). Many features of Leaves of Grass are explained by this description — the predominance of the first person pronoun "I," followed by a frequent addressing of "you," with the third person, "he, she and it," trailing far behind, and an equally rare use of "we" and "they". The expansiveness noted by William James, and the concrete descriptions of a great variety of people and places are also there, testifying to an omnivorous, democratic curiosity and zest for experience. As for the expression of a cosmic sense in the poetry, that is a more problematic matter. The word "kosmos" does occur several times, as the title of a poem and also in the assertion "I am a cosmos." Bucke's claim that beginning in 1855 with Leaves of Grass, Whitman was trying to describe several experiences of a cosmic mystical illumination, although repeated by some later Whitman followers, was more cautiously interpreted by Whitman himself. He once said to Horace Träubel, one of his disciples: "They tell me I am a mystic myself: may be that's the reason I don't like mystics" (in Miller 1969: xxi). Whitman was also consistently noncommittal about his acquaintance with Oriental mysticism. When Thoreau remarked on some affinities, Whitman said he did not know these and added "tell me about them." There is in these expressions of ambivalence toward the acknowledging of foreign sources a certain tension between Whitman's desire to project a persona as a native American poet speaking a colloquial folk idiom, on the one hand, and a cosmopolitan and sophisticated journalist on the other, who, as editor of the Brooklyn Eagle, introduced his readers to Carlyle and Coleridge, Goethe and Schlegel, among other European writers (Kaplan 1980: 128). As F.O. Matthiessen pointed out in the 1920s, Whitman self-consciously experimented with the use of words and referred to Leaves of Grass as "only a language experiment" (in Miller 1969). In his notebooks Whitman wrote himself such instructions as to use folk idioms and slang and to avoid learned language. Matthiessen said that Whitman's language is "deeply ingrained with the educational habits of a middle class...who put a fierce emphasis on the importance of the written word," and reflects the experience of a Brooklyn journalist of the 1840s, a former country school teacher, and a carpenter's helper (in Miller 1969: 187).

Pronouns, Persons and the Semiotic Self 245 Matthiessen's essay does not fully document his interesting suggestion about the 1840s middle class speech as a source of Whitman's language in Leaves of Grass — but, nevertheless, calls attention to Whitman's self-conscious and serious interest in words and in "experiments with language." At least one of Whitman's biographers has pointed out that the poet probably collaborated with a contemporary writer and professor of philology, William Swinton, on a popular series of articles on etymology, published in 18S9 as a book under the title Rambles among Words (Kaplan 1980). This little book takes a distinctly evolutionary approach to etymology and historical linguistics, refers to the "archeology of words" and to words as "fossils." It contrasts a "natural" with an "arbitrary" relation between "words" and "things," and says that "words are emblematic because things are emblematic" (Kaplan 1980: 34), and that "even the most startling metaphor and the wildest poetic usage has a law and logic of its own." A chapter on "the idealism of words" not only illustrates a Platonic theory of language as "the mirror of the most inner consciousness," but also cites the "original" meanings of some of Whitman's favorite words — kosmos, cemetery, democracy. In a passage which may have referred to Whitman, Swinton wrote: The poet is by divine right the proper namer. Through sympathy with the grand substantial words of the world he imports into human speech the utterance of orphic Nature. Material forms — ocean, air, soil, fire, stars, life, growths — these are sublime primeval words. These the expressive passion dissolves into plastic symbols, and the poet gives voice to mankind (Swinton 1859: 56-57). There is some evidence to indicate that Whitman shared Swinton's idealistic theory of language in the notes for a lecture on "An American Primer" and the many entries he collected for a combined dictionary and phrase book called "Words": "a perfect writer would make words sing, dance, kiss, do the male and female act, bear children...or do any thing, that man or woman or the natural powers can do" (quoted from Whitman's An American Primer in Kaplan 1980: 229). Whitman also declared a preference in the Primer for certain words because they were the correlatives of his identity: "robust, bravery, athletic, muscular, resistance, bracing, rude, rugged, rough, shaggy, bearded, arrogant, haughty. These are alive and sinewy — they walk, look, step with an air of command" (in Kaplan 1980: 229).

246 Milton Singer Whitman's interest in folk idiom, slang and in American, Spanish, and French words was obviously underpinned by a broader curiosity about the evolution of language, linguistic change, English grammar and usage, as well as European and Oriental mythology and religion. To some extent, he consulted William Swinton about these large questions, but his own readings in "Rousseau and Voltaire, Plato, Dante, Shakespeare, Cervantes, and Goethe, ancient cycles, epics and sagas, including the Hindu" (Kaplan 1980: 229-230) were sufficiently broad to belie the image of the author of Leaves of Grass as a "primitive" American folk poet uncontaminated by foreign and erudite influences. Mark Twain, a younger contemporary of Walt Whitman, shared the poet's interest in people and places and in their folk idioms. In prose, rather than verse, Mark Twain wrote his famous books which captured the atmosphere and regional dialects of Nevada gold miners, Mississippi river boat men, Missouri teenagers, black slaves, and Washington politicians. Much of the material in these books was based on his personal experience and his careers as journeyman printer, journalist, and river pilot. Yet he managed, as did Whitman, to read and travel widely, and to acquire a broad knowledge of language and literature. In his Autobiography, Mark Twain records that he was "diligently reading up" for writing the story of The Prince and the Pauper. I was reading ancient English books with the purpose of saturating myself with archaic English to a degree which would enable me to do plausible imitations of it in a fairly easy and unlabored way. In one of these old books I came across a brief conversation which powerfully impressed me...with the frank indelicacies of speech permissible among ladies and gentlemen in that ancient time....I was immediately full of a desire to practice my archaics and contrive one of those stirring conversations out of my own head (Twain 1959: 268). This is Mark Twain's account of how he came to write 1601, but it no doubt also explains the kind of research he must have done for The Connecticut Yankee in King Arthur's Court. In that work there is a generous sprinkling of conversations in which both the Yankee and the Arthurians use "I" as as well as such archaic statements as "me thinketh ye are to blame." Whether Mark Twain knew that before Chaucer the first person singular form was ic is not clear. In any case, Mark Twain's books in the 1870s, 1880s and 1890s continued Walt Whitman's practice of using colloquial English and free use of the first and second

Pronouns, Persons and the Semiotic Self 247 person singular pronouns, without implicating a transcendental conception of self. It is interesting that President U.S. Grant published his Personal Memoirs with Mark Twain's publishing house in 188S. Grant's Memoirs use the first person singular almost as frequently as Walt Whitman's and Mark Twain's work did and rarely use the first person plural. Grant's usage presents an intriguing puzzle since the President's last words written about 1885 were "I think I am a verb rather than a personal pronoun" (Sebeok 1986[1985]; Singer 1986). Perhaps the "I"s in the Memoirs represented a Presidential "I," or perhaps by 1885 the use of the first person singular in personal memoirs had become an accepted conventional usage due in part to the influence of Walt Whitman and Mark Twain and to the individualistic ideology of the time (Sewell 1987).

The "We-ing" of the "I": L£vi-Strauss in and out of the Savage Mind. In the symposium on "Semiotics, Self and Society" held in Philadelphia, Michael Silverstein commented that the papers illustrated two contemporary complementary trends; an "it-ing" of the "I" which analyzes the decline and depersonalizing of subjectivity, as in Benjamin Lee's paper on the semiotics of Descartes' cogito and dualism, and a reverse trend of "I-ing" the "it" which analyzes the return of the subject, as in the paper by Greg Urban, which analyzed the uses of the first person singular pronoun in Shokleng. These two trends toward objectification and subjectification, respectively, are certainly significant. Silverstein's distinction in terms of the use of the pronouns "I" and "it" reminded me of a third trend, the "weing" of the "I". L6vi-Strauss's statement in a recent interview "I have absolutely no feeling of my personal identity" is an astonishing example of this trend (in Bücher 1982: 360-361). Speculating about the meaning of the statement (see the postscript below), I wondered whether his structuralism or his sympathetic interest in Buddhism might explain it. By coincidence I happened to be reading the "Finale" of Naked Man and found there the surprising answer in the following passage: If there is a point at which the self can reappear, it is only after the completion of the work which excluded it throughout (since, contrary to what might be supposed, it was not so much the case that the self was the author as that the work, during the process of composition, became the

248 Milton Singer creator of an executant who lived only by and through it), then, it can and must take an overall view of the whole, in the same way as the readers who will peruse the text without having found themselves in the dangerous situation of feeling prompted to write it. This finale, headed by an epigraph which is unlike the preceding ones and is meant as a commentary on them, is itself in the nature of a commentary on a completed work, from which the writer is trying to deduce his own conclusions, now that his mission is over and he is again free to speak in the first person singular (L6vi-Strauss 1981: 630). The translator of L'Homme nu further supports L6vi-Strauss's explanation by pointing out that "from this point onwards, the author uses je instead of nous in the original" (L6vi-Strauss 1951: 630). The translator, however, seems to misunderstand L6vi-Strauss's explanation by attributing the use of nous only to modesty. For L6vi-Strauss, however, nous expressed his feeling of complete identification with the Indians and their myths. "I do not think the myths. The myths think themselves in me and without my knowing it." Only after he finished telling the myths, did the anthropologist feel conscious of himself as a je, as an individual separate from the nous of the myths. In the context of such an explanation, personal identity expressed by the pronoun "I" becomes a feeling of alienation from the collective identity of the anonymous tellers and hearers of the myths. From the parallel argument, one would draw the conclusion that the use of the pronoun "we" expresses something like a "participation mystique" in a collective identity. The above interpretation of L6vi-Strauss's statement about personal identity and the myths getting thought in him and unbeknownst to him is also supported by his Introduction to the Massey lectures, Myth and Meaning: Although I am going to talk about what I have written, my books and papers and so on, unfortunately I forget what I have written practically as soon as it is finished. There is probably going to be some trouble about that. But nevertheless I think there is also something significant about it, in that I don't have the feeling that I write my books. I have the feeling that my books get written through me and that once they have got across me I feel empty and nothing is left. You may remember that I have written that myths get thought in man unbeknownst to him. This has been much discussed and even criticized

Pronouns, Persons and the Semiotic Self 249 by my English-speaking colleagues, because their feeling is that, from an empirical point of view, it is an utterly meaningless sentence. But for me it describes a lived experience, because it says exactly how I perceive my own relationship to my work. That is, my work gets thought in me unbeknown to be. I never had, and still do not have, the perception of feeling my personal identity. I appear to myself as the place where something is going on, but there is no "I", no "me." Each of us is a kind of crossroads where things happen. The cross roads is purely passive; something happens there. A different thing, equally valid, happens elsewhere. There is no choice, it is just a matter of chance. (L£vi-Strauss 1979: 3-4). L6vi-Strauss's forgetting of his personal identity while working on his Mythologiques may not have been so idiosyncratic an experience as he has suggested, considering that artists, writers, scientists, and religious mystics have reported similar experiences. In any case, social theories which try to explain a loss or decline of individual self-consciousness are familiar enough, from Durkheim's theory of mechanical and organic solidarity to Redfield's folk to urban continuum of societies and cultures. These theories, however, have not generally included a linguistic or semiotic analysis of communication which associates the use of particular forms of expression, such as pronouns, with particular kinds of social structure. In this context, L6vi-Strauss's shiftfromje to nous and back to je may have signaled for him a shift from a modern urban society to a primitive or folk society and return, at least in imagination. Analogously, Peirce's evolutionary sequence of egotistical, idistical and tuistical historical eras can be interpreted as a counter-current and backlash to the laissezfaire industrialism and social Darwinism of post Civil War America. Those thinkers who participated in this counter-current, Royce, G.H. Mead, and Cooley, for example, also urged the ontogenetic and phylogenetic primacy of social consciousness over individual consciousness, and drew the logical conclusion of Peirce's anti-Cartesianism — that "we think (cogitamos), therefore we are." This trend was reinforced in Royce by his ethical concept of loyalty and the highly charged emotions of the first world war: "as I write you these words, Frenchmen are writing the meaning of these words... "Ma Mire, c'est la Republique" in their blood about Verdun. The Mother which is a republic, is a community, which is also a person, — and not merely an aggregate, and not merely by metaphor a person" (in Singer 1984a: 94).

250 Milton Singer Royce's passionately sentimental identification with France in 1916 as a community, a family, and a person owed much to Peirce's semiotic interpretation and in fact was anticipated by Peirce: The constant use of the word "we" — as when we speak of our possessions in the Pacific — our destiny as a republic — in which no personal interests at all are involved, show conclusively that men do not make their personal interests their only ones, and therefore may, at least, subordinate them to the interests of the community (CP 5.355; cf. Singer 1984a: 91). The link between such sentiments and war-time loyalties and social divisions is fairly obvious and may have been a factor in the formation of Durkheim's concept of "collective representations" and "effervescence," as well as in Schutz' Phenomenological theory of intersubjectivity as a "we-reality" (Schutz 1962). That such theories have their historical and cultural contexts was effectively demonstrated by Geertz's study of Person, Time and Conduct in Bali (1966) in which the Balinese "we-reality" was found to be culturally flattened, stereotyped and somewhat depersonalized (Geertz 1973 [1966]; Singer 1984a: 99-100). Victor Turner, on the other hand, tried to universalize Durkheim's collective effervescence by locating it in the "liminal" phase of the ritual process when "communitas" prevails (Turner 1969). President Reagan's 1987 State of the Union speech closed with a civics lesson on the United States Constitution addressed as well to the young people: I have read the constitutions of a number of countries including the Soviet Union's....Many countries have written into their constitutions provisions for freedom of speech and freedom of assembly. If this is true, why is the Constitution of the United States so exceptional? The difference is so small, it almost escapes you, but it's so great it tells you the whole story in just three words: We the People (quoted in the New York Times, January 28, 1987). The President's interpretation of the meaning of those three words is not a surprising one given that they refer to a democratic form of government:

Pronouns, Persons and the Semiotic Self 251 In those other constitutions, the government tells the people what they are allowed to do. In our Constitution, we the people tell the government what it can do and that it can do only those things listed in that document and no other things. Virtually every other revolution in history just exchanged one set of rulers for another. Our revolution is the first to say the people are the masters and government is their servant (quoted in the New York Times, January 28,1987). Whether or not the Attorney General, constitutional lawyers, and the Supreme Court agree with this interpretation of "We the People" in the Constitution, it is an interpretation which introduces a third kind of "we" alongside the royal "we" and the editorial or authorial "we". In the last two cases, as Benveniste has pointed out, the "we" is not a quantified or multiplied "I"; it is an "I" "expanded beyond the strict limits of the person, enlarged and at the same time amorphous" (Benveniste 1971: 203). Benveniste applies this characterization to both the authorial and the royal "we," but it also applies to the democratic "we." A possible differentiation among the three kinds of "we" is that the editorial "we" generally includes the speaker but not the addressee, the royal "we" includes the speaker and, ambiguously, the addressee, and the democratic "we" includes both speaker and addressee (see Singer 1984a: 96,193 fn. 5; for further discussion and illustrations see Levinson 1983:69). Whitman, the poet of democracy, rarely used "we" and he used "I" and "you" frequently. Perhaps these usages reflect the individualism of late nineteenth century America. Whitman's individualism may be idiosyncratic in being highly "mistical" as well as "egotistical." The particular literary genre, the psychological and social setting, and the prevailing linguistic conventions, as well as the "spirit" of an age obviously influence and are influenced by the kinds of pronouns used. President Reagan's speech to explain the Iran arms affair, for example, was short on references to "We, the People" and long on the first person singular pronoun.

"We" and "They" — Corporate Personality and Collective Identity The use of the plural first person "we" raises the question of whether a social consciousness and a collective identity is being expressed analogous to the expression of an individual consciousness and a personal identity by the use of "I."

252 Milton Singer The answer given to this question by Peirce, Royce, G.H. Mead and Cooley, among others, is an unequivocal "yes." William James's answer is more qualified by his individualism and pluralism, but does not rule out the possibility (see Singer 1984a: 89-96). In any case, the discussion of this problem reveals that the use of "we" is explicitly or implicitly associated with a use of "they." The contrastive pronoun pair "we" and "they" (or "us" and "them," "ours" and "theirs") form binary oppositions, to some extent analogous to the "I - you" pair and the dialogue generated by their use. Even President Reagan's civics lesson on the Constitution's "We, the People" introduces a contrast with "They, the Government." Other illustrations of "we - they" binary contrasts are familiar enough — between Americans and Russians, whites and blacks, women and men, yuppies and the elderly, students and the faculty, Latinos and Yankees, Christians and Muslims, etc. In all these cases the speaker uses "we" to refer to the group with which he identifies and "they" to refer to the group with which he disidentifies. The speaker's group identification does not necessarily follow objective criteria of membership in a group. Objectively, President Reagan is a member of the Government, but in his contrastive use of "we, the people" he disassociates himself from the government or at least implies that he is only a "servant" of the people. Although the use of the "we - they" contrast suggests an apparent parallel to the use of the "I - you" contrast, the two contrastive pairs differ in at least one important respect. As Benveniste, Lyons and others have pointed out, individual speakers change their roles in a conversation as they change their use of the first person pronoun "I," and become the "you" of another speaker's reference. Their consciousness of their personal identity presumably does not change with these changes in roles from subject to object, and back again to subject, with the use of pronouns "I" and "you" respectively. In the case of a collective identity expressed by "we - they" contrasts, there are problems both about the location of a corresponding social consciousness and its degree of permanence and capacity for change. Another problem is whether the use of the "we - they" contrast is determined by some preexisting social and cultural distinctions or whether the use of "we" and "they" helps to create and maintain the social and cultural distinctions. The recent tendency of social theory has been to emphasize indeterminism and the creative influence of ideology and linguistic usage. Even the Marxist theory of dialectical and historical materialism is being reinterpreted along these lines. Pierre Bourdieu's formulation — "the class struggle is a classification struggle" — is a characteristic illustration of this trend (Bourdieu 1984; Rosen 1984).

Pronouns, Persons and the Semiotic Self 253 In theory, switching from "we" to "they" and from "they" to "we" in a conversation (or from "us" to "them" and the converse) should be as frequent and as easy as switching from "I" to "you," and the converse. In practice, the "we - they" switching seems less frequent and more difficult than the switching in the corresponding singular pronouns or between singulars and plurals, at least in English conversation. A probable explanation for such differences is that in the case of the singular switching, e.g. from "I" to "you" and the converse, the shift in pronouns is assumed not to change the identity of the speaker or the hearer. My personal identity does not change because I refer to myself as "I" and another person refers to me as "you." Similarly, my interlocutor's identity does not change because I refer to him as "you" and he refers to himself as "I." In the case of the plurals, however, there seems to be a much greater difference in the group identity to which I refer as "we" or "us" and to which another person refers to as "they" or "them," and analogously for my reference to "they" and "them" and his reference to "we" and "us." "We, the people of the United States" uttered by an American does not seem to have the same reference or sense as "they, the people of the United States" uttered by a Russian. Similarly, "They, the Soviets" uttered by an American does not have the same reference or sense as "We, the citizens of the Soviet Union" uttered by a Russian. The difference between the above "we"s and "they"s springs not only from the differences in pronouns and in language, assuming that different languages are used. There are also differences in knowledge and sentiments, values and ideologies, beliefs and philosophies. To a Russian, "they, the Americans" are capitalists and bourgeois and "we, the Soviets" are socialists and workers. For an American, "We, the people" are free individuals and "they the Russians are godless communists." With such broad differences in the interpretations of "Americans" and "Russians," it is not surprising if individual Americans and Russians should find it difficult to negotiate the switch from "we" to "they" whether these pronouns are applied to themselves or to others, applied by themselves or others. Where collective identities are so stereotyped and polarized the pronominal references to and expressions of these identities also become their emblems. "We" and "they" — and "us" and "them" — as emblems of bipolar group identities are so widespread and familiar that sociologists, anthropologists, and psychologists have tried to formulate general laws governing these terms' behavior. Radcliffe-Brown's theory of "structural opposition" and "dual opposition" is probably one of the best known of these generalizations. Applied to totemism, joking and avoidance relations, and to kinship, economics and political relations,

254 Milton Singer this theory seemed to provide a generalized explanation for all human relations (Radcliffe-Brown 1951; 1952; 1958). It may also have been a source for L6viStrauss's structural anthropology (see Singer 1984b). Quite apart from the many criticisms of Radcliffe-Brown's structuralfunctionalism and his conception of a "natural science of society," the theory of "we" and "they" as emblems of bipolar group identities has yet to deal adequately with several important questions. In a complex society with overlapping and cross-cutting social classifications such as race, religion, age, gender, ethnicity, occupation, social class, education, region, etc., do the "we - they" emblems of identity also vary with the social classifications, do they crystallize around some hierarchy of the classifications, or are they subject to the negotiation and arbitration of third parties? Bourdieu's reference to the class-struggle as a "classification struggle" transfers the conflicts of social classes to a conflict of class symbols. This kind of symbolic relationship is already included in RadcliffeBrown's (and Durkheim's) theory of dual opposition, but it needs elaboration in terms of a concept of "symbolic force" (Thompson 1984; Daniel 1984,1987). Another important question that needs investigation is what happens to the cross-cutting and overlapping group identities in encounters and conversations between people from different cultures? Do they leam to communicate and take each other's roles, or do they simply continue to reenact their indigenous rituals and myths? Or do they form more inclusive identities with the help of new emblems or the reinterpretation of old emblems? (Sahlins 1985; Fernandez 1985, 1986a; Redfield 1963[1953]; Singer 1984a; Kipling 1926). One of the most puzzling problems is the persistence and evocative power of ancient class symbols in "modern" situations in which they are presumably anachronistic and inappropriate. The effectiveness of terrorist tactics usually depends on evoking such ancient class symbols and mobilizing group identifications under their banner. Anti-terrorist strategies often resort to similar tactics and so perpetuate cycles of violence between the groups defined by the ancient class symbols (cf. Tambiah (1986) on Sri Lanka). The apparent smooth shifts of the wes in the rhetoric of a war chiefs speech described by Urban (1986) do not provide a counter example to the above generalizations since that speech represents only one voice and one point of view. A dialogue between several speakers representing different government agencies, administrations, legislators, political parties, and the like would undoubtedly reveal frictions and disagreements in the shifting of wes, even within a single cultural domain. In a more recent paper on "The Pronominal Pragmatics of Nuclear War

Pronouns, Persons and the Semiotic Self 255 Discourse," Urban (1988) shows that the use of we for the United States and for the human species has a mutually exclusive complementary distribution in the article by Weinberger and the book The Fate of the Earth by Schell (cf. Wertsch 1987).

Some Classical Greek and Old Testament Sources of Peirce's Anti-Cartesianism I first became dramatically aware of some of the classical Greek sources of Peirce's anti-Cartesianism when I listened to a lecture on "The private man within the citystate" given by Jean-Pierre Vernant at the University of Chicago in 1986. In this lecture Professor Vernant examined how things worked in the Greece of the citystate, between the eighth and fourth centuries before Christ, in order to test Louis Dumont's theory of the origin of the modem individual from the ideal of the individual as a renouncer outside the world. Towards the end of his lecture, Professor Vernant described the archaic and classical Greek ego in distinctly antiCartesian language — neither bounded nor unified, without introspection, extroverted rather than introverted, with an existential rather than a reflexive selfconsciousness. As has been many times stated, the cogito ergo sum, I think therefore I am, had no meaning for a Greek. I exist because I have hands, feet and feelings; I exist because I walk and run, because I see and feel. I do all these things and I know that I do them. I never think my existence through the consciousness of them. My consciousness is always dependent on the outside; I am conscious of such and such object, of a certain sound, or of a kind of pain. The individual's world has not taken the form of self-consciousness, of a limited, inner universe in its interior, radical individuality. Bernard Groethuysen sums up the special nature of the ancient individual in a brutal but provocative formula by saying that self-consciousness is the apprehension of the self in a "He" and not yet in an "I" (Vernant 1986: 12-13). Peirce's anti-Cartesianism of course, included an "I" and a "thou," as well as an "it" Vernant also mentioned the use of the first person by the lyric poets to "grant that indecisive secret, personal subjectivity within us a precise verbal form and a

256 Milton Singer firmer consistency." The subjectivity of the poet qualified the radical objectivity of the classical Greek ego and served as "the touchstone for individual values....The final criterion [of communal values] is the individual subject, what he feels personally, which is the substance of his song" (Vemant 1986: 11). Another apparent exception to the radical objectification of the classical Greek ego mentioned in Vernant's lecture was Plato's conception of the soul (psuchi), separable from the body, "a daemon in us, a divine being, a supernatural force whose place and function in the universe goes beyond our single person" (Vernant 1986: 14). The Platonic soul and the personal ascetic practices associated with it did not, in any case, join with the "ferocious" introspection of Christian selfconsciousness which emerged in St. Augustine several centuries later: The Augustinian man who, in the dialog with God, can say "I" is of course far removed from the citizen of the classical city-state and from the homo equalis of pagan antiquity. But his distance is much greater, and even more profound, from the renouncer and the homo hierarchicus of Indian civilization (Vernant 1986: 17). When I asked Professor Vernant after his lecture whether he was familiar with Peirce's anti-Cartesian papers, he said he was not, but that one of his colleagues at the College de France was working on Peirce. He also told me that his teacher had been Louis Gernet, who had been influenced by Dürkheim and that he himself had listened to Marcel Mauss's lectures. The Duikheimian affiliations no doubt help to explain why Vernant's lecture placed the individual in the strict sense well within the city state and the social group. They may also explain why the "interior dimension" of the ego and the person appears as an objectified projection in mirrors and shadows. Durkheim's influence is more difficult to discern, however, in Vernant's perception of the first person singular subjectivity of the lyric poets who feel and express themselves in personal terms and personal time "as regret and nostalgia, as anticipation and hope, as suffering and memory of lost joys and past events" (Vemant 1986: 11). Professor Vernant promised to send me when he returned to Paris some of the classical references he had collected for his lecture. Among the bibliographic items he kindly sent were articles on Greek language and logic by Luce Giard, on the origins of grammatical theory in classical antiquity by Jean Lallot, and on ancient grammar and the originality of Varron, all printed together in the collection THTL (Travaux d'Histoire des Thiories Linguistiques, Universitd Paris, 1983). The

Pronouns, Persons and the Semiotic Self 257 photocopy of an article by Jacques Brunschwig ("Aristote et 1'effet Perrichon"), with some bibliographic indications concerning the person in Aristotle, was also included. Brunschwig's article directly supports Vernant's anti-Cartesian interpretation of classical Greek sources. In fact, Brunschwig writes that there is "a sort of epistemological obstacle...say Cartesian" to understanding ancient Greek thought. Not only is there no "cogito, ergo sum" in the classical texts, according to Brunschwig, but there is a "paradoxical" objectified version which could be formulated as follows: "I see my self (in my work or in some other projections of my self which have been enumerated above) then I exist; and I am there where I see myself: I am that projection of myself which I see" (Brunschwig 1983: 375). The need for a Cartesian cogito kind of assurance for one's existence is obviated, writes another of Valiant's references, by the fact that classical Greek "seems to be spontaneously objective; to specify the subjective moment, the philosopher must introduce a terminology of his own" (Kahn 1979: 24-25). Interpreting a famous passage in Aristotle's Nichomachian Ethics, Kahn adds that "when one is perceiving oneself or anything else, it is impossible to be unaware of one's own existence at the same time" (Kahn 1979: 29 fn. 86): "Thus Aristotle speaks of our perceiving that we are seeing and hearing just as Locke describes consciousness in reflexive terms as the perception of what passes in a man's own mind" (Kahn 1979: 24-25). This formulation is practically identical with Peirce's definition of "phenomenology" and with his analysis of sensation as distinct from its relation to an ego, which depends on inference: "a sensation as first given does not appear in its subjective aspect but as an object" (CP 3.53). The connection between Peirce's anti-Cartesianism and ancient Greek conceptions goes well beyond those two significant resemblances. It includes also the lack of emphasis on introspection and intuitive indubitability as a source of knowledge of the self; the lack of a sharp dualism of body and mind, and of emotion and cognition; the "projection" of the self in external objects such as mirrors, shadows, signs and symbols. Peirce's definition of personal identity in terms of name, story and character, and his conception of an extended identity which can include "a loosely compacted person" and all sorts of groups is practically Homeric, as Simon and Weiner (1966: 303-314) describe that model. Not only is there a tendency to describe mental activity as a story or narrative, but inner mental states are represented in terms of concrete, observable, behavioral aspects, and "inner" dialogues are presented as outer dialogues with others.

258 Milton Singer These resemblances are not mere coincidences or linguistic and cultural universale. They are probably an expression of Peirce's knowledge of and admiration for ancient Greek philosophy, especially Plato, Aristotle, Epicurus and the cosmologists. As Max Fisch (1986: 244) has noted, Peirce called the place where he lived the last 27 years of life "Arisbe" after a colony of Miletus, the home of Thales, Anaximander, and Anaximenes. If we set aside the first person singular subject of the Greek lyric poets, then Vemant's interpretation of the classical Greek ordinary individual in his social roles, and of the objectified ego as reflected in the projections of the external world, is quite consistent with Durkheim's emphasis on "collective consciousness" and "collective representations." It is also consistent with Aristotle's description of the spontaneous objectivity of perception and thought, and with Peirce. Additional evidence to support Profess«· Vemant's anti-Cartesian interpretation of the classical Greek individual has recently been indirectly provided in the study by Jos£ Faur's Golden Doves with Silver Dots. A student of Hebrew and the Rabbinic tradition, Faur has collated the work of Benveniste on Aristotle's categories, Jakobson on visual and auditory signs, Neugebauer on Greek geometry and Babylonian algebra, and Boman on Greek and Hebrew thought, to conclude that "Greek thinking is organized in terms of simultaneous synthesis and visual experience. Hebrew thought is mainly auditory and therefore organized in terms of successive syntheses" (Faur 1986: 32). Drawing on Saussure's and Benveniste's linguistic analysis, Faur interprets the several binary oppositions as expressions of a general contrast between an "ontological Greek metaphysics" and a Hebrew "semiotics." The distinction between the subject and the object of the verb is carefully analyzed in Hebrew pronouns, an analysis which concludes with an auditory theory of the self that seems to echo George Herbert Mead as well as the Old Testament: Unlike sight which is perceived as in the "world-out-there" and the purely internal feelings that are experienced in the "world-within," hearing is an auditory phenomenon in the "world-out-there" experienced in the "worldwithin." Hearing functions as the nexus between human interiority and exteriority. The self is not only totally vulnerable to sound, but through auditory speech, reflexive experience is realized. Upon hearing oneself talk, the interiority/exteriority of the "I" is simultaneously experienced. Presence is thus realized...the very idea of the universe as the dynamic speech of God excludes the possibility of God's representation (Faur 1986: 47-48).

Pronouns, Persons and the Semiotic Self 259 Faur applies his linguistic interpretation to the obligatory daily prayer "Hear oh Israel, the Lord our God, the Lord is One" which is generally recited in Hebrew and out loud, although the Rabbinic law has permitted its recitation in the language of the speaker or silently (Faur 1985:49). The corresponding linguistic analysis of the subjectivity and objectivity of classical Greek pronouns is not, unfortunately, discussed by Faur, although there is some relevant material in Benveniste and in Friedrich (1978: 117-123; and 1979: 402-405; 412-422; 432-433) and in Redfield (1975). The ambiguous status of first and second person subjectivity within the Durkheimian tradition is conspicuous in Marcel Mauss's (1985[1938]) famous essay on "The Category of the Person." Although Mauss explains that his essay concentrates on the social and legal aspects of the person, and will exclude a discussion of the linguistic and psychological aspects of a personal sense of identity, he does manage to include several important deviations from a Durkheimian collective consciousness. One of these deviations is Mauss's statement that the sense of personal identity and of self is a human universal: "it is plain...that there has never existed a human being who has not been aware, not only of his body, but also at the same time of his individuality, both spiritual and physical" (Mauss 1985: 3). Mauss (1985: 3) says that the subject of his essay is not the sense of self (moi) but "the notion or concept that men in different ages have formed of it...in different societies according to their system of law, religion, customs, social structures and mentality." In his recent commentary on Mauss's essay, Stephen Collins (1985: 68) interprets this statement as asserting that "what is universal...is the sense of self, the immediate 'here-and-now' consciousness expressed in human language in the first person singular." Collins also clarifies this interpretation by pointing out that Mauss, and Mauss's teacher Dürkheim, were probably influenced by two French neo-Kantians — Renouvier and Hamelin — who turned Kant's conception of the universal "I" — das Ich — into a concrete, individual consciousness. Along with Durkheim's dual conception of self, as both a biological individual body and a social consciousness, Collins believes that this neo-Kantian interpretation of the self established a universal form of the person which was a priori for any particular individual. Durkheim's belief in the sacredness of the individual, Collins (1985: 65-66) argues, was itself a social fact and not a conclusion of a methodological individualism. A second deviation in Mauss's essay is that he does manage to introduce into his discussion the major linguistic distinctions, at least in French, which mark the

260 Milton Singer historical and cultural evolution of just that sense of self (le sens du moi) which is universal and serves to distinguish the concept of the person {personne) from social role (personnage, persona) and personality (personnalite). For Mauss, as for Vemant and other scholars, the notion of a moral human conscious person — a sense of being conscious, independent, autonomous, free and responsible — is the critical emergent which is associated with early Christianity and Roman law (cf. Dumont 1985, Momigliano 1985). Mauss quotes Paul's Epistle to the Galatians: "you are, with respect to the one, neither Jew nor Greek, slave nor freeman, male nor female, for you are all one person, eis, in Christ Jesus (3:28)." Mauss's quotation from Cassiodorus — "the person is a rational substance, indivisible and individual" — brings the notion of the person within reach of Augustine and of Descartes (Mauss 1985:19-20). Mauss's abstention from a linguistic analysis of the category of person is not explained by a disinterest in languages. On the contrary, he does make a rather broad linguistic generalization: A very large number of languages are conspicuous for their use of many "positional" suffixes, which deal for the most part with the relationships existing in time and space between the speak«· (the subject) and the object about which he is speaking. Here the "self' is everywhere present, but is not expressed by "me" (moi) or "I" (je)..., although those languages also possess the pronoun (Mauss 1985: 3). Because he did know Sanskrit and Indology, Mauss's essay pointed out that the notion of the individual, of his consciousness of the "self' (moi) was known in India, "the most ancient of civilizations," as ahamkara, where aham can be glossed as "I" (je) or "ego." Mauss also cites "the maxim of the seers" — tat tvam asi — "that thou art" — which asserts the identity of the individual self with the universe, a starting point for all the Brahmanic schools of the upanishads (Mauss 1985: 13; cf. Singer 1984a, ch. 6). Mauss stopped short in his pioneering 1938 essay on "The Category of the Person" of discussing the relations of the diversity of linguistic forms of pronouns and positional suffixes for first, second, and third persons to notions of self and the category of person. This problem was not intensively explored until the development of linguistic and semiotic anthropology.

Pronouns, Persons and the Semiotic Self 261 Three Pronouns in Search of the "Real Self' Dürkheim's homo duplex which distinguished between a social consciousness whose content varies in different societies and cultures, and a biological individual which is a human universal, quickly became the hallmark of French and British anthropology. Radcliffe-Brown's version of the duplex as a distinction between "a social personality," the subject of social anthropology and sociology, and a biological individual, the subject of physiology and individual psychology, left the interaction of the two kinds of individuals in a limbo of uncertainty, as is already evident in his attempt to explain Andaman beliefs and behavior in the use of body paints (Radcliffe-Brown 1948[1922]: 121,315). An interesting expression of the persistence of Durkheim's dichotomy in modern social anthropology is Victor Turner's last essays, in which neurobiology and theories of hemisphere dominance are invoked to account for social and cultural phenomena. What is especially interesting about these essays of Turner's is not the persistence of Durkheim's and Radcliffe-Brown's homo duplex, but rather Turner's explicit introduction of a phenomenological level of analysis, in his references to Dewey and Dilthey and to Freud and Jung. At this phenomenological level of discussion, Turner transcends both the biological and social poles of Radcliffe-Brown's and Durkheim's dichotomy and attains a level of observation and reflection accessible to ordinary experience (Turner 1983,1986). Victor Turner was by no means the first anthropologist to use a phenomenological approach to an analysis of the self. In the personality and culture movement of the 1920s - 1940s, some of Boas' students — especially Ruth Benedict, Margaret Mead, Paul Radin and Edward Sapir — started to develop a theory of personality which took account of the work of Freud and Jung, and of Dewey and G.H. Mead. By the 1950s and 1960s some anthropologists, notably Redfield, Hallowell, Warner, Opler, and Geertz, were turning to earlier phenomenological conceptions of the self in G.H. Mead, Dilthey, Schutz and Wittgenstein in order to avoid both Durkheim's homo duplex and Descartes' cogito (Singer 1961,1984a). When I sketched a genealogy for this phenomenological and symbolic interactionist theory of the self as deriving from Peirce by way of William James, Josiah Royce, John Dewey, G.H. Mead, W.H. Cooley, and Charles Morris, one British social anthropologist called my genealogy a "mirage." Adding insult to injury, this critic claimed that a phenomenological interpretation of Peirce overlooks Peirce's realism and idealism (Hobart 1986). That a Peircean ancestry

262 Milton Singer for a phenomenological and symbolic interactionist conception of the self is no "mirage" is easily established by considering the biographical histories and interrelations of Peirce, James, Royce, Baldwin, Mead, Dewey, Cooley, and Morris at Harvard, Hopkins, Chicago, Ann Arbor and Columbia, as well as their writings, and especially such critical commentaries on mutual influences as G.H. Mead's "Cooley's Contributions to American Social Thought" (1930), his "The Philosophy of James, Royce and Dewey in their American Setting" (1930), and Morris' The Pragmatic Movement in American Philosophy (1970). Ironically, another British social anthropologist still devoted to Durkheim's homo duplex, not only did not think that my Peircean genealogy was a "mirage," but extrapolated from it support for an alleged "well-established" American identification of culturally defined personality types with the "people of one nation state" (Leach 1984: 158). The criticism that a phenomenological interpretation of Peirce's semiotic conception of the self overlooks and is inconsistent with Peirce's realism and idealism is more serious than the "mirage" criticism. This kind of criticism reflects a prevailing popular belief that the philosophy and history of science should provide a comprehensive and complete epistemology and ontology which can then be applied to solve specific problems in special fields of science or scholarship. Peirce did not share such a belief, either as a practicing scientist or as a philosopher. Neither do most practicing scientists and philosophers today. Albert Einstein's reply to physicists and philosophers who expected him to adopt a complete and systematic epistemological theory of physical reality seems to me very much in the spirit in which Peirce himself might have replied. Let us look first at Einstein's reply and then consider a possible Peircean reply to the problem of the reality of a semiotic self: The scientist, however, cannot afford to carry his striving for epistemological systematic that far. He accepts gratefully the epistemological conceptual analysis; but the external conditions, which are set for him by the facts of experience, do not permit him to let himself be too much restricted in the construction of his conceptual world by the adherence to an epistemological system. He therefore must appear to the systematic epistemologist as a type of unscrupulous opportunist: he appears as realist insofar as he seeks to describe a world independent of the acts of perception; as idealist insofar as he looks upon the concepts and theories as the free inventions of the human spirit (not logically

Pronouns, Persons and the Semiotic Self 263 derivable from what is empirically given); as positivist insofar as he considers his concepts and theories justified only to the extent to which they furnish a logical representation of relations among sensory experiences. He may even appear as Platonist or Pythagorean insofar as he considers the viewpoint of logical simplicity as an indispensable and effective tool of his research (Einstein 1982[1949]: 683-684). Why should anthropologists feel that they should be any less opportunistic towards philosophy than physicists? The combination of positivism (or empiricism), realism, and idealism which Einstein describes seems adequate as a foundation for a philosophy of science. Wheeler's summary description of essentially the same synthesis balances Einstein's well-known emphasis on realism with the physicist's participant observation, more congenial to Bohr's Copenhagen interpretation of quantum physics: When we do our observations in the here and the now on photons, quanta of light, hunks of energy coming from distant astrophysical sources, we ourselves have an irretrievable part in bringing about that which appears to be happening. We can put it this way: that reality is, in a certain sense, made up of a few iron posts of definite observation between which we fill in, by an elaborate work of imagination and theory, all the rest of the construction that we call reality. In other words, we are wrong to think of the past as having a definite existence "out there." The past only exists insofar as it is present in the records of today. And what those records are is determined by what questions we ask (Wheeler 1981:24). What Hobart calls the "mirage" of an intellectual genealogy in Man's Glassy Essence does not have a definite existence "out there" in the past, but is a construction based on today's questions and records. The attractiveness of Peirce as a founding ancestor is precisely in his combining empiricism, realism, and idealism in a unique and productive manner. To find all three of these in the person of one scientist and philosopher makes that scientist a resource not to be spumed. My impression from reading Peirce is that he was all of the above — a phenomenologist, an idealist and a realist How else could one make sense of his interpretation of his three categories as "categories of consciousness and being," of his pragmatic theories of meaning, truth and reality;

264 Milton Singer of his triadic analysis of signification as a relation of object, sign, and interpretant; or of his semiotic conception of the self? As early as 1872 Peirce defined "reality" in a manner that seemed to him to conform to the practice of physicists and to encompass an empirical, realistic and idealistic doctrine of metaphysics: "this doctrine is that observation and reasoning are perpetually leading us towards certain opinions and that the fact of such a perpetual tendency is otherwise expressed by saying that the objects of those final opinions have a real existence" (Peirce 1986: 59). Peirce's anti-Cartesian doctrine that knowledge of the self depends on external observation rather than on private introspection and on indubitable intuitions of a thinking substance has much in common with the "spontaneous objectivity" of classical Greek. The affinity is especially close at the level of the phenomenological descriptions of sensation which, for Peirce, were descriptions of what is present to the mind without respect to judgements of truth or falsity. Once questions of truth value are raised, the situation becomes more complicated since Peirce distinguishes different kinds and degrees of reality and different kinds of inference to arrive at them. To illustrate some of these complications we shall refer to our previous discussion of Peirce's three pronouns I, it and thou as a nuclear structure which underlies the formation of selves and communication between them. We shall first deal with these pronouns within the framework of Peirce's theory of signs and then consider some cross-cultural and cross-linguistic applications. Of the three pronouns, the first and second persons, "I" and "you," are clearly indexical signs, in both Peirce's and Benveniste's sense, as pointing respectively to the speaker who utters "I" in a present instance of discourse and to an addressee in the same instance of discourse who is addressed as "you." In both of these cases the speaker and the person spoken to are assumed to be present in a face-to-face situation. The third person, however — "he," "she," or "it" — is not assumed to be present; in fact, the third person is generally assumed to be an object of discourse who is absent. For that reason, Benveniste (1971) has called the third person a "nonperson." In Peirce's analysis, a more appropriate description would be to say that a third person pronoun is not generally an indexical sign which points to a physically contiguous object. That there may be different degrees of physical contiguity between indexical signs and their objects in Peirce's discussion of indexical signs was pointed out by W.C. Watt (1986; cf. Urban 1987). If someone were asked that perennial question about personal identity "who are you?" a Peirce-Benveniste kind of answer would probably not be considered very

Pronouns, Persons and the Semiotic Self 265 informative: "I am the person who said I in the present instance of discourse." Even an iconic answer would also sound enigmatic: "I am the first term in a threeterm relation." Perhaps a full colloquial sentence would begin to supply some of the kind of expected information: "I am the person who gave you the book yesterday." Peirce would freely acknowledge that these three different kinds of answer do not convey the same kinds of information about the same kinds of reality. The first, indexical, answer illustrates his category of "secondness" — physical existence and interaction; the second, iconic, answer illustrates his category of "firstness" or possibility; and the third, symbolic, answer, his category of thirdness, or symbolic mediation. Signs, objects, and their interpretants are all differentiated according to their different kinds and degrees of reality. As we pass from the iconic "I" to the indexical "I" and then to the symbolic "I," we not only traverse Peirce's three realms of being — possibility, actuality, and social reality — we also seem to recapitulate Kant's pure conceptual ego, the empirical body, and the opinions of future selves destined to be confirmed by the community of rational investigators. If a questioner should persist and ask which of the three Is — the iconic, the indexical, or the symbolic — is the "real" /, a Peircean answer, I believe, would simply ask the questioner for a definition of "reality." In other words, beyond a given definition of "reality" and the methods for discovering it, a Peircean would probably argue that the question "which is the real /?" or "which is the real you?" has no logical meaning. Such an answer would not exclude the possibility that there may be disagreement about the criteria defining "reality" or the criteria defining "personal identity." Peirce's definition of personal identity, for example, in terms of a name, a story, and a character would probably be considered inadequate by psychoanalysts who would want to add descriptions of "inner" drives, repressions, repetition compulsions, oedipal complexes, and the like. Those analysts would probably also insist that they have methods for observing the symptoms of such "inner" states and processes in dreams, verbal associations, and in all sorts of neurotic and psychotic behavior. I am not raising the question here of the scientific validity of psychoanalysis, but only the question of whether a Peircean semiotic conception of the self has any place in it for the psychoanalytic theory of "inner" personality structures and processes, in contrast, say, with the "self psychology" developed by Kohut, Wolf, and others. Whether Freud's ontological and epistemological theories are also compatible with Peirce's is a question beyond the immediate scope of the present paper. Nevertheless, there is a fairly obvious comparison to be made between the

266 Milton Singer Freud-Groddeck theory of the ego, id and superego as a differentiated structure of the individual personality and Peirce's analysis of the three personal pronouns I, it, and thou. The next section will deal with this comparison (cf. Singer 1986). Urban's paper on "The I of Discourse" (1987) not only provides a typology of five different kinds of Is and a continuum for scaling different degrees of proximity to the speaker, but also includes the different degrees and kinds of reality represented by the five Is. This last observation, although not explicitly discussed by Urban, follows from his use of Peirce's triadic classification of signs as indexical, iconic, and symbolic. Manley Thompson was the first commentator, I believe, to suggest that Peirce's theory of the self could be interpreted in terms of this threefold classification of signs, but Thompson did not apply the suggestion to the use of pronouns (Thompson 1953:48-50,229-33,255-56,266).

Freud and Peirce on I, It and Thou and "the Inner Self" / , it and thou considered as first, third and second person pronouns, respectively, in English represent a grammatical paradigm which has also been described as a linguistic and cultural universal: the speaker, the addressee, and the person or object spoken about (see, e.g. Boas 1911; Cherry 1966; Greenberg 1963; Ingram 1978; Levinson 1983; Sebeok 1983). In spite of the existence of such a universal, there are many problems of intercultural, interdisciplinary, and cross-linguistic communication. These problems arise because communications are almost never simply a matter of pure grammar, but also involve different linguistic ideologies (Silverstein 1979), as well as different cultural cosmologies and epistemologies. A good example of these problems is the comparison of Peirce's I, it, and thou with Freud's ego, id, and superego. In 1979 at a conference at the Center for Psychosocial Studies in Chicago, I presented a paper on "Personal and Social Identity in Dialogue" in which I suggested that Peirce's dialogical interpretation of personal pronouns was the ancestral source for the symbolic interactionist theory of the self developed by William James, Josiah Royce, James Mark Baldwin, George Herbert Mead, John Dewey, Charles Horton Cooley, and Charles W. Morris. In the same paper I noted a parallel between Peirce's I, it and thou and the Freud-Groddeck Psychoanalytic theory of the ego, id and superego (cf. Singer 1984a: 86). This comparison did not assume any direct contact between Peirce and Freud, but was suggested by several convergent and somewhat coincidental events. One of them was Melford Spire's 1979 article "Whatever Happened to the Id?" in

Pronouns, Persons and the Semiotic Self 267 which he asked why social scientists had neglected the Freudian theory of the id as an unconscious source of sexuality and violence. Bruno Bettelheim's (1983) pointing out that the standard English translation of Freud's works used the professional technical terms of ego, id and superego instead of Freud's colloquial German das ich, das es and das über ich, and instead of the colloquial English pronouns I, it and you, made the resemblance between Peirce and Freud more transparent.3 Max Fisch also sent me in 1979 some comments on my American Anthropological Association 1978 lecture "Signs of the Self' and included a photocopy of Peirce's pages on I, it and thou. These three converging events practically made my perception of the parallel between Peirce and the FreudGroddeck theory inescapable, and also pointed to Fredrick Schiller's Letters on the Esthetic Education of Man as a likely common source for both Freud and Peirce (cf. Fisch 1982: xxvii; and Singer 1986: 360-363). Another confirmation of the Peirce-Freud pronominal parallel appeared in 1981 in an article by the psychoanalyst Erik Erikson (1981) on "The Galilean Sayings and the Sense of Ί'". This article did not come to my attention until 1986, but contains a most dramatic support for the Peirce-Freud parallel, without mentioning Peirce explicitly. Since Erikson has published psychohistorical biographies of Luther (1958), Gandhi (1969), and Jefferson (1974), it is not surprising that he should find some anticipations of Freud in Christ's Galilean sayings and parables. As interesting as these anticipations may be for the early history of Christianity, in the context of the present discussion Erikson's developmental theory of the personal pronouns is the more relevant contribution. In two pages he sketches a developmental map that shows how the sense of "I" emerges, not as a primordial given or posit, but in a sequence of social interactions and dialogues with others, or "counterplayers." Erikson's description of the developmental map of these I-you and we-they dialogues and their changing social boundaries and cultural ideologies is practically a restatement of his theory of identity from the perspective of the personal pronouns involved, and as such deserves to be quoted in full: Writers who take the sense of I seriously will first of all ask what is the / s counterplayer. They may indeed begin with the second pronoun, you, and end with the soul's sense of divine Thou. To my developmental orientation the most telling "map" depicting the development of the sense of I would be the whole list of personal pronouns, from I to They, as each one first gets to be pronounced and understood correctly in childhood, and then as it is meaningfully experienced and reexperienced throughout life.

268 Milton Singer The beginnings of the sense oil itself, one should think, can only emerge in a newborn out of the counterplay with a sensed You in the maternal caretaker — whom we shall call the Primal Other; and it seems of vital importance that this Other, and, indeed, related Others, in turn experience the new being as a presence that heightens their sense of I. It is this interplay, I think, that helps the original sense of I gradually face another fundamental counterplayer, namely, my Self — almost an Inner Other. But the original interplay of You and I remains the model for a mutual recognition throughout life, up to a finite expectation to which St. Paul gave the explicitly religious form of an ultimate meeting now only vaguely sensed beyond "a glass, darkly" (the Ultimate Other, then). Now, one glance at the list of all personal pronouns reveals a whole developmental program in their sequence: for while I and you form the original dyad, this dyad soon turns into a number of triads as a series of hes or shes (and, indeed, a world of its) become additional counterplayers within varied connotations: paternal, fraternal, sororal, and so on. And as this happens, the plural concepts we, you, and they become both verbal necessities and the bearers of important emotional involvements. Thus, the system of pronouns, beginning with I and you, is built into a ground plan ready to unfold in stages; and one can well see how each of them, once learned, serves a widening experience as it includes, on every stage, new counterplayers. Take, for example, the necessity — especially in any patriarchal and monotheistic system — to transfer some of the earliest forms of a sense oil from their maternal origin to strongly paternal and eventually theistic relationships. Or consider the crisis of adolescence as a transfer of the identity elements formed in childhood and youth to the productive milieu in which one expects to find one's psychosocial identity. Or, again, how the sense of we acquired in one's family of origin ("my kind") must be extended to the family and the community one marries into — and, indeed, to one's own new family in which one must help to generate new beings with their own sense of I. Throughout this establishment of new boundaries of We, Ourselves, dictated as it is by the realities and the ideology of work and production, there also emerges a gradual demarcation of the decisive borderlines beyond which live those definitely other Others — those theys and thems whom one has learned to repudiate or to exclude as foreign, if not nonhuman altogether. These habitual rejections, in turn, have helped to

Pronouns, Persons and the Semiotic Self 269 give a clearer outline to one's own "true SelF or to those variant "selves" which are either proudly or fatalistically accepted as a self-description within the contemporary world of roles. And yet, throughout all these critical stages with all their involvements, there remains for the / a certain existential solitariness which in these pages, we depict as seeking love, liberation, salvation. So we now return to such geographic and historical conditions as provide in any given period the basis for an ideological orientation in the widest sense of a Way of Life — the essential bridge between all is and their We's (Erikson 1981: 330-332). Erikson links the third person singular pronoun it with the Freudian id, "an inner cauldron of drives and passions," "the darkness behind consciousness," and with Einstein's scientific curiosity and joy in scientific observation and discovery which "turned him from the I and the We to the It — and its revelatory power." This also placed Einstein on the road to understanding, wrote Erikson, "the human unconscious as the source of the most destructive and self-destructive drives — an aspect of human nature which Einstein like Freud, all too soon found associated with the most sublime scientific and technical inventiveness" (Erikson 1981: 360361). This double association in Erikson's essay of the it with Freud and with Einstein recalls, at least implicitly, the famous 1930s exchange between Einstein and Freud on "Why war?" and in some measure updates that exchange when one of the sayings attributed to Jesus is identified by Erikson as an anticipation of the Freudian id: 'Do you not see that whatever goes into a man from outside cannot defile him, since it enters, not his heart, but his stomach, and so passes on?' (Thus he declared all foods clean.) And he said, "What comes out of a man is what defiles a man. For from within, out of the heart of man come evil thoughts, fornication, theft, murder, adultery, coveting, wickedness, deceit, licentiousness, envy, slander, pride, foolishness' (Mark 7: 18-22; Erikson 1981: 345). Erikson also goes beyond Freud's answer to Einstein by suggesting that Gandhi's application of the golden rule to the Ahmedabad nonviolent strike — namely, "a conscious and determined projection of the best in us, on the (seemingly) worst enemy, and a willingness to face him nonviolently" — may offer a possible

270 Milton Singer solution in human terms to today's biggest threats to humanity — threats that were already apparent to Freud and to Einstein in the 1930s (Erikson 1981: 361-62; Freud 1930:143-144; Einstein and Freud on "Why War?" 1933). "You are That, but the Ego is Always There" Erikson's comprehensive map of English personal pronouns and his FreudianGalilean interpretation of Gandhi's militant nonviolence suggests the possibility of an intercultural and cross-linguistic analysis of the semiotic self and its pronominal expressions. For this purpose, it becomes necessary to take account of the indigenous semiotic theories of another culture as well as of its pronouns and other forms of person deixis. To illustrate, I close with a semiotic comparison of the individual's relation to the world in India and the West (cf. Singer 1984a: ch. 6; Staal 1966; Gerow 1982, 1984; Dumont 1985; Marriott 1976; Kakar 1978). In making such a comparison it is important to note that Erikson's comprehensive developmental map of the pronouns is not only consistent with Peirce's semiotic and dialogical conceptions of personal and social identity, but is also consistent with the symbolic interactionist theory of the self that was associated with George Herbert Mead. The relation of Mead's theory of "role-taking" and "the significant others" to Cooley, Sullivan, Schutz, Buber and other contemporary theories is lucidly discussed by the social psychologist R.S. Perinbanayagam (1981). In this book Mead's analysis of "significant symbols" as a function of social acts is elaborated as a central contribution to social and cultural psychology. Unfortunately, the linguistic and psychological dimensions of the analysis remain somewhat vague, although Sullivan's "interpersonal psychiatry" is briefly summarized. The sanskritic Indian tradition asserts an identity of the individual self and the universe in the famous "you are that" or "you are it" (tat tvam asi) which has long been considered a characteristic expression of Indian mysticism and idealism. While in popular contemporary Indian interpretations, I have heard this maxim invoked to justify the belief that "everyone has a spark of the divine," and is identical with the spirit of the universe, I have also heard the same people add such realistic comments as "the ego is always there, claiming I did this and I did that" (Singer 1980: 337). But I never heard an Indian make Walt Whitman's claim "I am a cosmos." More characteristically Indian is the claim of seeing the universe in a divinity, whether made by Krishna's mother as she looked into his mouth, or by the ninth century Tamil saint Nammalvar as in the following bhakti poem:

Pronouns, Persons and the Semiotic Self 271 We here and that man, this man, and that other in-between, and that woman, this woman, and that other, whoever, those people, and these, and these others in-between, this thing, that thing, and this other in-between, whichever, all things dying, these things, those things, those others in-between, good things, bad things, things that were, that will be, being all of them, he stands there. (Transl. A.K. Ramanujan 1981: 122). As Professor A.K. Ramanujan comments, each Tamil pronoun in this poem behaves like an indexical or "shifter," refers to a generalized person yet points to actual particular persons out there. The grammatical paradigm of the pronouns, however, does not exhaust the poem's significance. Professor Ramanujan adds: "grammar becomes poetry, and poetry becomes theology....Conceptions of God are enacted by word and syntax; furthermore, God's one-and-manyness becomes the living word to be uttered, danced to, sung and chanted in temples as these poems are to this day" (Ramanujan 1981:126). This interpretation of Nammalvar's poem recalls and continues the classical semiotic interpretation of dancing in an ancient manual: Having made the prayers, etc. the dancing may begin. The song should be sustained in the throat; its meaning must be shown by the hands, the mood (bhava) must be shown by the glances; rhythm (tala) is marked by the feet. For wherever the hand moves, there the glances follow; where the glances go, the mind follows; where the mind goes, the mood follows; where the mood goes there is the flavour (rasa) (Mirror of Gesture 1970: 17).

272 Milton Singer Ramanujan also makes an interesting contrast between the poetics of classical Sanskrit (and Tamil) and the poetics of Nammalvar and the bhakti poetry: In the course of these transpositions, the poetry and the poetics become strikingly different from anything in earlier Tamil or Sanskrit, in the following ways: The speaker, the worshiper, is always in sight. This is never the case in classical love poetry, which is an impersonal poetry of personae. The classical poems are framed as dramatic monologues. In the alvar poems, even the classical personae (lover, beloved, etc.) are rendered personal; they express the alvar's relation to his god-lover. The poet and speaker are one (Ramanujan 1981: 161). The first person pronoun is used by the devotee to address the deity. In Tamil the pronoun used means "I, at your feet" (atiyen). Ramanujan's semiotic comment on the Tamil pronouns in Nammalvar's poem, while not identical with Peirce's discussion of the first, third and second person pronouns I, it, and thou, reminds us once more of the power and fruitfulness of Peirce's semiotic analysis, an analysis which not only includes grammatical, poetic, theological and performative interpretations of pronouns mentioned by Ramanujan, but several other interpretations as well. To recapitulate the course of our previous discussion, Peirce interpreted I, it and thou at several different levels: as an iconic grammatical diagram of first, second and third person pronouns; as indexical pointers to the subjects and objects of complete sentences; and as representing the three terms of signification in the sign process — the interpretant CO, the object (it) and the sign (thou). In addition, Peirce first used these three pronouns as names for three kinship relations: son (7), father (it), and mother (thou); for three historical ages of civilization (egotistical, idistical, and tuistical); as well as for the Christian trinity. His three metaphysical categories of being, Firstness (/), Secondness (it), and Thirdness (thou), were also named by the three pronouns in his early papers. Peirce's anti-Cartesianism raised some doubts whether subjectivity finds expression in the pronouns, especially in the first and second person singular. Granting that he rejected Descartes' cogito ergo sum, as a direct, self-evident intuition of a thinking substance, Peirce did not deny an immediate consciousness of objects and subjects which he called phenomenology. Nor did he deny the possibility of inferring the relationship of such consciousness to an individual ego or to "a loosely compacted person" consisting of several individuals. The

Pronouns, Persons and the Semiotic Self 273 attribution of psychological motives to individuals and social groups is embedded in Peirce's pronominal analysis by way of his adoption of Fredrick Schiller's three impulses (or triebe): the I impulse and faculty (the form trieb), the it impulse (the stoff trieb), and the thou impulse and faculty (the spiel trieb). Peirce's use of Schiller's three impulses probably explains the close parallel between the Freud-Groddeck differentiation of the / into an ego, id, and superego and Peirce's I, it and thou. This may also explain why Freud's therapeutic slogan was "where id was, there shall ego be," while Peirce's youthful prophecy for perfection was of a coming mistical age that would succeed the egotistical and idistical ages of civilization (Singer 1986: 361-362). Peirce's use of Schiller's Letters and of quotations from Shakespeare, Emerson and other poets and writers testifies to the fact that his semiotic conception of the self was based on deep insight into the unified structures and processes of communication, and on a definition of reality that included creative contributions by the members of an unlimited community of investigators. The striking parallel between Peirce's triadic semiotic conception of the self as articulated in I, it and thou, and the Freud-Groddeck ego, id and superego seems to break down in one important respect — that the psychoanalytic theory of personality seems to use a method of introspection to reach the true "inner self," while Peirce resolutely rejects introspection as a source of self-knowledge and depends on inferences from observation of external signs. Access to the "inner self' would then seem to be closed off by Peirce's anti-Cartesianism. A similar conclusion would seem to be implied by the linguistic-cultural universal of the triad — speaker, addressee, and person or object spoken about — and by G.H. Mead's analysis of "significant communication" and of an internalization of "a conversation of gestures." Habermas' distinction between an "objective" and "intersubjective" phase of significant communication in Mead's theory does not, it seems to me, affect the negative conclusion about individual subjectivity (Habermas 1987). That the parallel between Peirce and Freud cannot be so quickly dismantled is suggested by several considerations. One of these is Erikson's developmental map of the pronouns in which "you" represents a "maternal caretaker" and "a primal other," interplay with whom heightens the child's sense of I and creates another counterplayer — "my self — almost an Inner Other." An introspectionist might argue that Erikson has abandoned Freud for an objective interactionist approach, but Erikson would probably reply that he has remained faithful to Freudian methods and the theory of psychoanalysis.

274 Milton Singer Another consideration in support of Erikson and the interactionists is that psychoanalytic practice uses the methods of free association and interpretation of dreams that are essentially methods of semiotic mediation between patient and analyst. In Western culture, that mediation is primarily verbal — "a talking cure." In other cultures, such as those of India and Japan, the analyst may depend more on tactile, visual, and "body language." Which of these media gives the most direct access to the "inner self' may itself be a culturally specific question as the papers by Daniel, Rosen, Straus, and Wolf suggest In recent discussions of this question several psychoanalysts and cultural anthropologists have observed distinctive cultural differences in both the media of semiotic mediation through which a sense of the "inn«- self' is communicated, as well as in the cultural definitions of such a self and its relation to other selves and objects. These observations have not led them, however, to deny Mauss's generalization that "there has never existed a human being who has not been aware...not only of his body, but also at the same time of his spiritual and physical individuality" (in Collins 198S: 68). L6vi-Strauss himself had to qualify his denial of any feeling of personal identity by conceding that such a feeling came back to him when he finished his writing on the "savage mind." He marked the return by shifting from nous to je (cf. Roland 1982,1983; Kakar 1978, 1982; Spiro 1979,1982; Geertz 1973,1983; Doi 1986). The parallels between Peirce's triad of pronouns and the Freudian triad, especially as interpreted by Erikson, do not quite settle the issue of subjectivity. An important difference emerges when the Freudian id is compared with the Indian that in the advaitin identity of "that thou art," or the more colloquial versions "you are that" and "I am that." The sanskritist Betty Heimann has explained that the demonstrative pronoun "that" in the identity can also be replaced by "it," and refers to the totality of everything, "the all," as well as to the cosmic spirit, which is three quarters unconscious (Heimann 1964: 56,107,110,111,117-18,133,160; cf. Singer 1984a: 183-84). Heimann does not regard the "I-thou" relation of the later bhakti movements as an exception; she subsumes that relation in Indian thought under the "It" experience: "For all times and all movements, the Indian axiom remains that the ultimate Summum is supra-personal....All T- and 'Thou'relationships are only steps towards the last aim of the It'-experience and of the final vision and of the total re-absorption into the 'It'" (Heimann 1964: 133). Since the Freudian id is also an unconscious and pervasive force there would seem to be a ground for expecting to find a parallel between Freud and advaita philosophy. Indeed, there are apparent parallels between such identities as "we are the world" and "I am that," not to mention the parallel between Whitman's "I am a

Pronouns, Persons and the Semiotic Self 275 cosmos" and "that I am." Yet an important difference lurks in these parallels. Although both kinds of identities refer to an unconscious and impersonal force, the relation of the individual person to this force is very different in the two cases. For Freud and Groddeck the id, or das es, is an unconscious and impersonal force within each individual, an "inner cauldron," as Erikson calls it. For advaitins the it of "you are it," or "I am it," is a vast impersonal force which transcends and absorbs the individual. Curiously, the oceanic metaphor is used by both sides, but for advaitins the individual is a stream or rivulet flowing into the ocean, while for Freudians the individual has an "oceanic feeling" of merging with the ocean (cf. Masson 1980; Obeyesekere 1987; Marriott 1976,1988). From the point of view of traditional epistemology, the Freudian id is the more subjective and represents an "I-ing of the it," while the advaitin it is the more objective and represents an "it-ing of the I." Freud's therapeutic "where id was, there shall ego be" is then to be contrasted with the advaitin's "where ego is, there you, we and it also are." The greater objectivity of the advaita philosophy is closer to Homer and Aristotle than it is to Descartes and Freud: "Homer is pre-Cartesian and, for that matter, pre-Socratic....Homeric man is objective because Homeric narrative is objective....[I]t is a poetry of character, for character is transparently visible in speech and action" (J. Redfield 1975:26,175). Whether this objectivity of Homeric man is to be explained by archaic and classical Greek grammar or by Greek rhetoric and philosophy is a moot question. Edwin Gerow's conclusion from his analysis of the advaitin grammar of /f-ness suggests a reasonable answer: "we speak our philosophies along with our grammars" (Gerow 1982: 116; cf. Staal 1966). Peirce's anti-Cartesianism, in any case, returning as it did to the "spontaneous objectivity" of classical Greek and to an Old Testament faith in words, emboldened him to announce that "my language is the sum total of my self."

Postscript Professor L6vi-Strauss has kindly sent some comments on a draft of my paper which further clarify the issues raised earlier and indirectly, at least, lend some support to the interpretation of the meaning of L6vi-Strauss's shift from nous to je in L'Homme nu. Professor L6vi-Strauss comments that "consciously at least" he used nous as the "pluriel de modestie" in exactly the sense that the Grand Robert — the "best modern French dictionary" — defines it: "pluriel vague [qui] efface

276 Milton Singer dijd la personaliti et vous replonge dans la foule." But, as L£vi-Strauss also points out, his consistent use of the plural forms from Le Cru et le cuit to L'Homme nu is "obfuscated" because "the translator of these works decided to use the first person singular throughout because, he explained, the 'we' would not sound right in English." The translation of the French nous by English "I" has created considerable confusion and controversy in the discussion of L£vi-Strauss's views. Given such a translation, it is difficult, for example, to make sense of the opening statement in the "Finale" of The Naked Man: "throughout these pages, the 'we' the author has deliberately adhered to has not been meant simply as an expression of diffidence." The translator's footnote (L6vi-Strauss 1981: 625), explaining that "I" is used because the English "we" tends "to sound more stilted," only accentuates the puzzle. In fact, structural anthropology as a method of reducing or eliminating the subject — "a Kantism without a transcendental subject," as Ricouer labeled it — has somehow restored the subject in these English translations. L6vi-Strauss's letter points to the translator's decision as the critical factor in the transformation. Perhaps there is also a difference here in linguistic usage and ideology of first person singular and plural pronouns as between French and English (American style) — a kind of Whorfian difference. There are several bits of evidence pointing in this direction. Livi-Strauss's letter, for example, contrasts the English version of his famous statement in Le Cru et le cuit — "nous ne prdtendons pas montrer comment les hommes pensent dans les mythes, mais comment les mythes se pensent dans les hommes et ä leur insu" (Livi-Strauss 1964: 20) — with the English translation quoted in my paper. This translation, however, is not identical with that in The Raw and the Cooked, which reads as follows: "I therefore claim to show, not how men think in myths, but how myths operate in men's minds without their being aware of the fact" (Livi-Strauss 1970: 12). I do not know where the English translation of the sentence in my paper comes from, perhaps from a kind of unconsciously remembered synthesis of several discussions I have read about L6vi-Strauss's formulation. It is in any case an even more individualistic translation of the French sentence than that in The Raw and the Cooked: "I do not think the myths. The myths think themselves in me without my knowing it." A second, more striking bit of evidence for a systematic difference between French and English usage of first-person pronouns is suggested by the Grand Robert definition of nous. The source for this definition cited is Gautier, Histoire

Pronouns, Persons and the Semiotic Self 277 du romantisme .... p. 98, and Brunot and Buneau, Pricis de grammaire historique, 3rd ed., p. 283. When I looked up these references, I found several interesting things — that the Gautier reference, for example, includes the exact same lines of the Grand Robert definition as part of the last paragraph of a newspaper piece he wrote around 1870 recalling the spectacular effect he produced by wearing a red waistcoat and long hair to the first performance of Victor Hugo's "Hernani" in 1830. Gautier's costume did not lose him in the crowd at the theater then, but made him stand out so dramatically that he thereby acquired an unshakeable and unwanted reputation as the "young fellow with the red waistcoat and the long hair." In the last paragraph Gautier apologizes to the reader for doing violence to "his natural modesty" and talking so much about himself. At the same time he writes that this is not an habitual fault of his and that if he knew how to disappear from his work he would do so. At this point he writes that "je so repels him that his expressive formula is nous, whose vague plural effaces the personality and plunges you into the crowd." This of course is the "nous de modestie" in the Grand Robert. It so happens that in an English translation of Gautier's piece published in 1902, nous and all of the plural first person forms are translated by "I" and other singular first person forms. Moreover the translator, listed as a Professor of French at Harvard University, omits completely any direct translation of Gautier's words quoted in the Grand Robert definition of nous. Unlike the translator of Livi-Strauss's Mythologiques, Gautier's translator did not think it was necessary to append a translator's note to call attention to the omission or to explain the difference in usage between French and English first person pronouns. The Brunot Historical Grammar (Brunot 1899: 298-299) cited in the Grand Robert definition of nous adds that the use of nous in place of je is derived historically from the usage in Roman chancelleries to enhance the dignity of the speaker. This practice was transmitted to modern times in administrative acts, and spread from the highest state functionaries to the humblest mayors. It was associated with the practice of saying vous instead of tu, toi to everyone who is not family. Why the French first person plural pronoun, "a plural of modesty," historically descended from a "plural of majesty" in the Roman chancelleries, should come to be preferred to the first person singular pronoun is an intriguing puzzle which social and cultural historians of French are better qualified to resolve than I. Perhaps Gautier's story of the red waistcoat is a special case; it does explain why he would have wanted to be lost in the crowd after that experience. It does not

278 Milton Singer explain, however, why Le Grand Robert, "the best French modern dictionary," would enshrine Gautier's formulation into a definition of standard French usage. Nor does it explain why L6vi-Strauss selected this definition as expressing exactly what he had in mind by using nous in his volumes on mythology. His suggestion in some of his books that the method of structural analysis reduces the subject, and even disregards it completely, provides an important clue, as does his suggestion of an affiliation between structural anthropology and Saussure's structural linguistics (L6vi-Strauss 1963: 86-91; 1970; 1976[1960]: 10-14; cf. Singer 1984a). But why a method of structural analysis should be consistently translated into English with the first person singular pronoun "I" requires further explanation. The question is a bit more complex than it appears at first sight because English also has a first person plural that is fairly close to the French plural of modesty, in the editorial or authorial "we." The OJE.D. identifies and defines this "we" as one of the two kinds: "used by a single person to denote himself, by a speaker or writer in order to secure an impersonal style and tone, or to avoid the obtrusive repetition of T" (O.E.D. 1962[1933]: 213-14). The question is further complicated by the fact that in some of the English translations of L6vi-Strauss's works, e.g. Totemism, Structural Anthropology, The Savage Mind, "we" is used without appearing to be unnatural or stilted in style. It is possible that English usage of the first-person singular and plural pronouns has been changing in the last hundred years or so. My paper suggests two diverse and competing trends: on the one hand a predominance of "I," "me" and "you" over "we" and "they" in Walt Whitman, Mark Twain, William James, and the Beatles' songs into the 1960s; on the other hand, a slowly increasing use of "we" instead of "I" and "me" in Peirce and Royce, the later Beatles, and President Reagan's "We, the people" speech. Phil Donahue's — "The King of the Afternoon Talk Shows" — comment on Oprah Winfrey — "The Queen of the Morning Talk Shows" — probably marks the arrival of a combined royal and editorial "we" in popular American speech: In an interview last week, Mr. Donahue admitted: "The biggest emotional adjustment has been to our ego. After all these years of enjoying this premier place all by ourselves, we looked up to see somebody doing what we do and getting light-years' more publicity. 'The King is Dead' is a very attractive thought."

Pronouns, Persons and the Semiotic Self 279 Oprah Winfrey's graceful reply indicated that the singular first person subject is not about to be eliminated: Ms. Winfrey responded: "If there never had been a Phil, there never would have been a me. I can talk about things now that I never could have talked about before he came on the air. There's room for both of us" CN.Y. Times 2/1/88, p. 17). Peirce's anti-Cartesian papers were published in the late 1860s. While the following 100 years did not witness a fulfillment of his prophecy that the egotistical and idistical eras would be succeeded by a tuistical era, the usages in the United States, and in France, do seem to indicate a trend towards a "nouisticaT'era. Explanations of such trends of change in the linguistic use and style of pronouns call for historical studies, not only in the domain of language and culture, but also in the fields of political and religious movements and ideologies. Wars and revolutions, cultural and otherwise, have something to do with the ways people talk and write and need to be studied from a semiotic perspective.4 It is interesting that Peirce's paradigm of /, It, and Thou for his book on words begun in 1859 (see Fig. 1) gives mostly the singular forms. "They" and "them" are included, but not "we" and other plurals. Kant's and Schiller's individualistic philosophy of the self were probably uppermost in his mind at the time. With the anti-Cartesian papers soon after the Civil War (1868) and "The Grounds of Validity of the Laws of Logic" (1869), the social principle of logic is emphasized. His references to "we" and "our" are directly associated with "our possessions in the Pacific." While there may not be many passages in Peirce's writings such as Royce's 1916 identification of "person" with "community" and "nation" or Cooley's cogitamos, I believe these extensions are consistent with Peirce's anti-Cartesianism and his semiotic conception of the self. The predominant English translations of L6vi-Strauss's nous into "I" raises several paradoxical questions. Not only do these translations seem to restore the singular subject to L^vi-Strauss's structuralism; in the process, they also evoke a deeper philosophical side to American individualism. When I asked my colleague Robert Streeter, a student of American literature, why the use of "we" in 19th century writings was relatively infrequent compared to the use of "I," he replied that such writers as Thoreau, Emerson, Whitman had a double interpretation of "I," once as the ordinary individual person and then as a transcendental self. He cited Quentin Anderson's book The Imperial Self{ 1971), as good documentation of this

280 Milton Singer

THl

TIKSONS.

I.

IV,f

% A^'tf yUw ,lLey>.lL$'LL·., uiffcJr, sJ{ 1

/ny , Avüvt

TS IT /4 vLu füc ^ ίΐ Äi^ ~t!L·-. jit^d- jicfn^ /Lfi^ j THi^Jj , TO 2V, . htMJy^-O^.

I

THOU

ScCJrvyif- JubtOrr*. ,

Figure 1: The first manuscript page of Peirce's book The Natural History of Words begun in 1859 (Fisch 1982: xxviii).

Pronouns, Persons and the Semiotic Self 281 theme. Anderson's book adds Henry James to the list and also discusses Hawthorne as a contrasting case. My colleague A.K. Ramanujan had already mentioned his personal discovery in Whitman's "Song of Myself of a "double self' which reminded him of the Bhagavad Gita. Ramanujan also recalled a statement by Thoreau that "we" is used by royalty, editors, pregnant women, and people who eat worms. Thoreau's apology to the reader at the beginning of Waiden explains why he departs from the practice of most books in using I for first person: "I should not talk so much about myself if there were anybody else whom I knew so well" (lWaiden, p. 3). For two American English professors to propound seriously a transcendental interpretation of American individualism was somewhat surprising. Ramanujan is of course right in being reminded of Indie texts by Whitman's "Song of Myself," but he acknowledges that the transcendence in Whitman's verse comes from an incorporation of the cosmos by the individual, rather than from the absorption of the individual in the cosmos as in most of the Indie texts. Anderson (1971: ix, 228), as well, while having less to say about Indian sources of American transcendentalism, interprets the "transcendence" as an act of "imperial" expansion of the individual self. Peirce's anti-Cartesian papers of the late 1860s called a halt to this runaway American individualism which looked for roots in Descartes, Kant, and Eastern mysticism, by denying the individual as an ultimate physical and psychological isolated reality. Replacing the individual with a triadic, semiotic self, with an extended social identity, Peirce was able to focus attention on the structures of signs and symbols which relate individual selves to culture, society, and nature. The deconstruction of the transcendental individual self was already well begun before Peirce, by the British empiricists — Locke, Berkeley, Hume, J.S. Mill — but was not systematically and explicitly articulated in semiotic terms until Peirce. It is notable that William James, who militantly asserted his individualism, joined Peirce in dismantling the transcendental singular subject by identifying the individual thinker, the "I think," with the stream of thoughts and of consciousness. After Peirce and James, Bertrand Russell completed the empiricist deconstruction of transcendentalism by analyzing all personal and demonstrative pronouns as "egocentric particulars." The conceptions of the self, individual and social, then become constructions with the help of a logic of relational structures and processes (Russell 1948: 300-303; Singer 1984b). If uttering "I think" can no longer give me indubitable assurance of my existence and identity, at least uttering "we think"

282 Milton Singer and "we talk," and tilling in other pronomial iconic diagrams, will provide us with as much assurance of our existence and identity as we can reasonably expect to find. The English personal pronouns have had unusual stability over long periods of time. A Linguistic Introduction to the History of English asserts that "pronouns as a paradigmatic set have undergone less change since OE times than any other part of speech" and compares the modern forms with the OE forms to support the assertion: Corresponding to ic, me, mln, we still have I, me, mine; for we, us, Orer, we have we, us, our, for pü, pe, pin, we have thou, thee, thine, for ge, eow, SOwer we have or had ye, you, your, for he, his, him, we have he, his, him, for hit we have it, for pis, past, hwa, hwset, we have this, that, who, what (Bloomfield and Newmark 1963:163). This stability is not inconsistent with the reduction in the set of modern pronominal forms as compared with the old forms, a reduction that is in part the result of an unconscious lingmstic drift, according to Sapir: at least part of the case feeling in he and him is to be credited to their position before or after the verb. May it not be, then, that he and him, we and us, are not so much subjective and objective forms as pre-verbal and post-verbal forms, very much as my and mine are now pre-nominal and post-nominal forms of the possessive (my father but father mine·, it is my book but the book is mine)? That this interpretation corresponds to the actual drift of the English language is again indicated by the language of the folk. The folk says it is me, not it is I, which is "correct" but just as falsely so as the whom did you see? that we have analyzed. I'm the one, it's me; we're the ones, it's us that will win out — such are the live parallelisms in English today. There is little doubt that it is I will one day be as impossible in English as c'est je, for c'est moi, is now in French [or for c'est nous, if Sapir were writing today — M.S.] (Sapir 1949[1921]: 166-67). That these unconscious changes in linguistic usages may themselves be closely connected with changes in social interactions and conscious forms of communication was a major thesis of the symbolic interactionists influenced by

Pronouns, Persons and the Semiotic Self 283 Peirce. A lucid and incisive formulation of that thesis was stated by Dewey in his book on politics The Public and its Problems: Associated or joint activity is a condition of a community. But association itself is physical and organic, while communal life is moral, that is emotionally, intellectually, consciously sustained.... Associated activity needs no explanation; things are made that way. But no amount of aggregated collective action of itself constitutes a community. For beings who observe and think, and whose ideas are absorbed by impulses and become sentiments and interests, "we" is as inevitable as "I." But "we" and "our" exist only when the consequences of combined action are perceived and become an object of desire and effort, just as "I" and "mine" appear on the scene only when a distinctive share in mutual action is consciously asserted or claimed.... Interactions, transactions, occur de facto and the results of interdependence follow. But participation in activities and sharing in results are additive concerns. They demand communication as a prerequisite (Dewey 1954[1927]: 151-152). And communication, for Dewey, demands a new medium — of signs and symbols, and their meanings — a medium which makes possible new levels of memory and anticipation, action and planning: Only when there exist signs or symbols of activities and of their outcome can the flux be viewed as from without, be arrested for consideration and esteem, and be regulated. Lightning strikes and rives a tree or rock, and the resulting fragments take up and continue the process of interaction, and so on and on. But when phases of the process are represented by signs, a new medium is interposed. As symbols are related to one another, the important relations of a course of events are recorded and are preserved as meanings. Recollection and foresight are possible; the new medium facilitates calculation, planning, and a new kind of action which intervenes in what happens to direct its course in the interest of what is foreseen and desired (Dewey 1954[1927]: 152-153). The new medium of signs and symbols not only makes possible a new kind of significant communication in social interactions and associations, but thereby

284 Milton Singer transforms the physical interactions and associations into a moral and intellectual community of thinking and feeling human selves: Symbols in turn depend upon and promote communication. The results of conjoint experience are considered and transmitted. Events cannot be passed from one to another, but meanings may be shared by means of signs. Wants and impulses are then attached to common meanings. They are thereby transformed into desires and purposes, which, since they implicate a common or mutually understood meaning, present new ties, converting a conjoint activity into a community of interest and endeavor. Thus there is generated what, metaphorically, may be termed a general will and social consciousness: desire and choice on the part of individuals in behalf of activities that, by means of symbols, are communicable and shared by all concerned. A community thus presents an order of energies transmuted into one of meanings which are appreciated and mutually referred by each to every other on the part of those engaged in combined action. "Force" is not eliminated but is transformed in use and direction by ideas and sentiments made possible by means of symbols (Dewey 1954[1927]: 153). It should not be necessary at this point to remind the reader that the search for and discovery of personal and social identities is not a hermetically sealed process of intuitive introspection inside individualized thinking substances, separate and distinct from bodily substances and isolated from one another. That the process depends on social interactions and associations with other minds and bodies, mediated by all available modalities of communication, verbal and nonverbal, is by now recognized as an implication of Peirce's conception of a semiotic self, and of the elaboration of that conception by the symbolic interactionists. The use of the first, second and third person pronouns, and the corresponding forms of the verb, are implicated in this conception of self because they are implicated in the unified structures of human interactions, associations, and communication. The use of these pronouns is a condition for the formation of selves and persons because it is also a condition for the formation of human communities and human society.

Pronouns, Persons and the Semiotic Self 285 Notes 1.

I am deeply grateful to Benjamin Lee and Greg Urban, to the Center for Psychosocial Studies, and to the other participants in the 1986 American Anthropological Association symposium "Semiotics, Self and Society" for the honor they have paid my book Man's Glassy Essence: Explorations in Semiotic Anthropology without embalming it. The title "Semiotics, Self and Society" is almost identical with the title of George Herbert Mead's lectures Mind, Self and Society posthumously published in 1934 and edited by his student and successor at the University of Chicago, Charles W. Morris. It would not be an exaggeration to say that the implicit agenda of the symposium, as well as of the present book, is to consider whether the substitution of "semiotics" for "mind" in the title provides an accurate marker of the changes that have taken place in our conceptions of the relations of self and society during the last half century. If there is one person chiefly responsible for replacing "mind" by "semiotics," it is Charles S. Peirce and his theory of signs. Although scholars from philosophy, linguistics, psychology, anthropology, and comparative literature have contributed to the recent development of a semiotic analysis of the relation of self and society, we must, I believe, acknowledge Peirce as the granddaddy of them all. My own awareness of Peirce's contribution goes back to the 1930s when I was a student in psychology and philosophy at the University of Texas, Austin and then at the University of Chicago. Dissatisfaction with Watsonian behaviorism led me first to embrace Mead's "social behaviorism" and Peirce's anti-Cartesianism (Singer 1936). In Austin this move was encouraged by three former Mead students — Clarence Ayres in Economics and George Gentry and David Miller in Philosophy. In Chicago, the encounters with Charles Morris, Bertrand Russell, and Rudolf Carnap introduced me to the method and history of symbolic analysis in pragmatics, semantics, and syntactics (Singer 1940). The later associations in the 1940s, 1950s, and 1960s with Robert Redfield, Fred Eggan, David Mandelbaum, M.N. Srinivas, Surajit Sinha, and with other anthropologists at Chicago in those years (Stocking 1979), as well as with South Asianists such as W. Norman Brown, Stella Kramrisch, V. Raghavan, Nilakanta S has tri, T.K. Venkateswaran, Hans Van Buitenen, Edward C. Dimock, A.K. Ramanujan, among others, and three trips to India, gave me an opportunity to develop more contextual and comparative cultural methods

286 Milton Singer

2.

3.

4.

of symbolic analysis (Singer 1980[1972]). Conversations with my colleague S. Chandrasekhar have greatly helped my understanding of modem physical theories and their relations to physical reality (Chandrasekhar 1987). In the 1960s, Claude L^vi-Strauss's structural interpretation of RadcliffeBrown's second lecture on totemism challenged me and many other anthropologists to reconsider the history of structural analysis in social and cultural anthropology and its relations to theories of symbolism and conceptions of the self (Singer 1984b). The result of this reconsideration was a return to Peirce's theory of signs ("semiotics") as a source for a "semiotic anthropology." The public christening took place in 1976 when Tom Sebeok invited me to lecture at Bloomington (Jean Umiker-Sebeok 1977, Thomas A. Sebeok 1978). This lecture, the 1978 American Anthropological Association lecture on "Signs of the Self," and the lectures and articles which soon followed constituted the substance of Man's Glassy Essence: Explorations in Semiotic Anthropology (Singer 1984a). The emphasis on personal pronouns in my paper, especially I, it and thou, was suggested to me when Max Fisch sent me in 1979 some comments on the "Signs of the Self lecture together with photocopies of Peirce's 1859 and 1861 pages on I, it, and thou. Fisch's keen and persistent interest in my work and his own illuminating historical studies of Peirce's many-sided contributions to semiotics (Fisch 1986) have been a constant source of inspiration and information. As a small token of my appreciation of his important contributions to the revival of Peirce studies, I am glad to dedicate my paper to him. Peirce's relation to these more recent developments has been noted by Lewis 1918: 79-106, 354-62; Singer 1984a: 22-26; Fisch 1986: 34-78, 171-183, 248-260; Zellweger 1982; Hawkins 1976. Bettelheim's article first appeared in The New Yorker. It was later expanded into a small book, Freud and Man's Soul. See also Ernest Wolfs paper in this volume. See Singer 1984a: 89-96; Benveniste 1971: 203; Seidel 1975. Dumont's Essays on Individualism (1986) deals historically with a broad anthropological concept of "ideology" but needs to be supplemented with a linguistic history. Also see Urban (1986) and (1988), and Wertsch (1987).

Pronouns, Persons and the Semiotic Self 287 References Anderson, Q. (1971). The Imperial Self. New York: Alfred A. Knopf. Benveniste, E. (1971). Problems in General Linguistics. Coral Gables: University of Miami Press. Bettelheim, B. (1983). Freud and Man's Soul. New York: Alfred A. Knopf. Bloomfield, M.W. and Newmark, Z. (1963). A Linguistic Introduction to the History of English. New York: Alfred A. Knopf. Boas, F. (1911). Introduction to Handbook of American Indian Languages. Bureau of American Ethnology Bulletin 40. Washington, D.C.: U.S. Government Printing Office. Boon, J.A. (1982). Other Tribes, Other Scribes: Symbolic Anthropology in the Comparative Study of Cultures, Histories, Religions, and Texts. Cambridge: Cambridge University Press. Boon, J.A. and Schneider, D.M. (1974). Kinship vis-d-vis myth: contrasts in L6vi-Strauss' approach to cross-cultural comparison. American Anthropologist 76: 799-817. Bourdieu, P. (1984[1979]>. Distinction: A Social Critique of the Judgement of Taste. Cambridge: Harvard University Press. Brunot, F. (1899). Pricis de Grammaire Langue Historique de la Langue Frangaise. Quatrifcme Edition. Paris: MassonetCie. Brunschwig, J. (1983). Aristote et l'effet Perrichon. In Hommage ä Alquiö: La Passion de la Raison, pp. 361-377. Paris: Universitäres de France. Buber, M. (1970). I and Thou. New York: Charles Scribner's Sons. Bucher, B. (1985). An interview with Claude L6vi-Strauss, 30 June, 1982. American Ethnologist 40: 360-68. Campbell, C. and Murphy, A. (1980). Things We Said Today: The Complete Lyrics and a Concordance to the Beatles' Songs, 1962-1970. Ann Arbor: The Pierian Press. Chandrasekhar, S. (1987). Truth and Beauty: Aesthetics and Motivations in Science. Chicago: University of Chicago Press. Cherry, C. (1966). On Human Communication. Cambridge: M.I.T. Press. Collins, S. (1985). Categories, concepts or predicaments? Remarks on Mauss's use of philosophical terminology. In The Category of the Person, M. Carrithers; S. Collins; and S. Lukes (eds.), 46-82. Cambridge: Cambridge University Press.

288 Milton Singer Daniel, E.Valentine (1984). Fluid Signs: Being a Person the Tamil Way. Berkeley: University of California Press. (1987). Therapeutic history. Contributions to Indian Sociology 21: 77-83. Dewey, John (1954 [1927]). The Public and its Problems. Chicago: The Swallow Press. Doi, T. (1986). The Anatomy of the Self: The Individual vs. Society. Tokyo: Kodansha. Dumont, L. (1986). Essays on Individualism . Chicago: University of Chicago Press. (1985). A modified view of our origins: the Christian beginnings of modern individualism. In The Category of the Person, M. Carrithers; S. Collins; and S. Lukes (eds.), 93-122. Cambridge: Cambridge University Press. Einstein, A. (1982[1949]). Reply to criticisms. In Albert Einstein: PhilosopherScientist, vol. 2, P.A. Schilpp (ed.). LaSalle, Illinois: Open Court. Einstein, A. and Freud, S. (1933). Why War? Paris: League of Nations. Erikson, E. (1958). Young Man Luther. New York: W.W. Norton & Co. (1969). Gandhi's Truth. New York: W.W. Norton & Co. (1974). Dimensions of a New Identity. New York: W.W. Norton & Co. (1981). The Galilean sayings and the sense of "I." Yale Review 70: 32162. Faur, J. (1986). Golden Doves with Silver Dots: Semiotics and Textuality in Rabbinic Tradition. Bloomington: Indiana University Press. Fernandez, J.W. (1985). Exploded worlds — text as a metaphor for ethnography. Dialectical Anthropology 10: 15-26. (1986). The argument of images and the experience of returning to the whole. In The Anthropology of Experience, V.W. Turner and E.M. Bruner (eds.). Urbana: University of Illinois Press. (1986). Review of M. Singer, Man's Glassy Essence. American Anthropologist 88: 768-69. Fisch, M.H. (1982). Introduction to Writings of Charles S. Peirce. A Chronological Edition, vol. 1, M.H. Fisch, et al. (eds.). Bloomington: Indiana University Press. (1986). Peirce, Semeiotic, and Pragmatism. K.L. Ketner and C.J.W. Kloesel (eds.). Bloomington: Indiana University Press. Freud, S. (1953[1930]). Civilization and Its Discontents. London: The Hogarth Press.

Pronouns, Persons and the Semiotic Self 289 Friedrich, P. (1978). Sappho's subjectivity. In The Meaning of Aphrodite. Chicago: The University of Chicago Press. (1979). Speech as a personality symbol. In Language, Context and the Imagination. Stanford: Stanford University Press. Geertz, C. (1973). The Interpretation of Culture. New York: Basic Books. (1983). Local Knowledge. New York: Basic Books. Gerow, E. (1982). What is Karma (Kim Karmeti)? an exercise in philosophical semantics. Indologica Taurinensia 10: 87-116. (1984). Language and symbol in Indian semiotics. Philosophy East and West 34: 245-60. Grant, U.S. (1962[1885-1886]). Personal Memoirs, E.B. Long (ed.). New York: Grosset and Dunlap. Greenberg, J.H., ed. (1963). Universals in Language. Cambridge: MIT Press. Groddeck, G. (1961[1923]). The Book of the It. New York: The New American Library of World Literature. Habermas, J. (1987). Theory of Communicative Action, vol. 2. Boston: Beacon Press. Hawkins, B.S. (1976). Peirce's and Frege's systems of notation. Graduate Studies Texas Tech University, K.L. Ketner, et al. (eds.), 381-89. Lubbock, TX: Texas Tech Press. Heimann, B. (1964). Some Aspects of Indian Thought. London: Allen & Unwin, Ltd. (1937). Indian and Western Philosophy: A Study in Contrasts. London: Allen & Unwin, Ltd. Hintikka, J. (1962). Cogito ergo sum: inference or performance. Philosophical Review 71: 3-32. Hobart, M. (1986). Review of Man's Glassy Essence. Man 21: 388-89. Holloway.E. (1949). Walt Whitman. Encyclopedia Brittanica,\ol.23. Ingram, D. (1978). Typology and universals of personal pronouns. In Universals of Human Language, vol 3: Word Structure, J. Greenberg (ed.), 213-247. Stanford: Stanford University Press. Jakobson, R. (1972). Verbal communication. Scientific American 227: 73-80. Kahn, C. (1979 [1966]). Sensation and consciousness in Aristotle's psychology. In Articles on Aristotle, vol 4, J. Barnes et al. (eds.), 1-31. London: Duckworth. Kakar, S. (1978). The Inner World: A Psychoanalytic Study of Childhood and Society in India. Delhi and New York: Oxford University Press.

290 Milton Singer (1982). Shamans, Mystics and Doctors. New York: Alfred A. Knopf. Kant, I. (1965 [1781]). Critique of Pure Reason, N.K. Smith (trans.). New York: St. Martin's Press. (1974 [1798]). Anthropology from a Pragmatic Point of View, M.J. Gregor (trans.). The Hague: Martinus Nijhoff. Kaplan, J. (1980). Walt Whitman, A Life. New York: Simon and Schuster. Kipling, R. (1926). We and They. In Rudyard Kipling's Verse. New York: Doubleday. Leach, E. (1985). Review of M. Singer, Man's Glassy Essence. American Ethnologist 12: 154-56. Leaf, M. (in press). Milton Singer's pragmatism. In Innovation and Tradition in South Asia, M. Ames and M. Leaf (eds.). Toronto: University of Toronto Press. Lee, B. (1985). Peirce, Frege, Saussure and Whorf: the semiotic mediation of ontology. In Semiotic Mediation, E. Mertz and R.J. Parmentier (eds.), 99-128. New York: Academic Press, Inc. (1986). The Objectification of Subjectivity. Ph.D. Thesis. Department of Anthropology, The University of Chicago. L6vi-Strauss, C. (1963). Totemism. Boston: Beacon Books. (1964). Le Cru et le cuit. Paris: Plön. (1966). The Savage Mind. London: Weidenfeld and Nicholson. (1970). The Raw and the Cooked. New York: Harper and Row. (1971). L'Homme nu. Paris: Plön. (1974). Tristes Tropiques. New York: Atheneum. (1976[1960]). The scope of structural anthropology. In Structural Anthropology, vol.2. New York: Basic Books. (1979). Myth and Meaning. New York: Schocken Books. (1981). The Naked Man. New York: Harper and Row. (1984). The View from Afar. New York: Basic Books. Levinson, S.C. (1983). Pragmatics. Cambridge University Press. Lewis, C.I. (1918). A Survey of Symbolic Logic. Berkeley: University of California Press. Marriott, McK. (1976). Hindu transactions: diversity without dualism. In Transactions and Meaning, B. Kapferer (ed.), 109-142. Philadelphia: ISHI. (1988). The open Hindu person and interpersonal fluidity. Unpublished manuscript.

Pronouns, Persons and the Semiotic Self 291 Martin, A.B. (1930). A Vocabulary Study of 'The Gilded Age." Western Groves, MO.: The Mark Twain Society. Masson, J.L.M. (1980). The Oceanic Feeling. Dordrecht: Reidel. Mauss, M. (1985[1938]>. A category of the human mind: the notion of person; the notion of self. In The Category of the Person, M. Canithers; S. Collins; and S. Lukes (eds.), 1-25. Cambridge: Cambridge University Press. Mead, G.H. (1922). A behavioristic account of the significant symbol. Journal of Philosophy 19: 157-63. (1930a). Cooley's contribution to American social thought. American Journal of Sociology 35: 693-705. (1930b). The philosophy of Royce, James and Dewey in their American setting. International Journal of Ethics 40: 211-231. (1934). Mind, Self and Society from the Standpoint of a Social Behaviourist, C.W. Morris (ed.). Chicago: The University of Chicago Press. Miller, E.H. (1969). A Century of Whitman Criticism. Bloomington: Indiana University Press. Mirror of Gesture (1970), A. Coomaraswamy and G.K. Duggirala (trans.). New Delhi: Munisharam Manoharlal. Momigliano, A. (1985). Marcel Mauss and the quest for the person in Greek biography and autobiography. In The Category of the Person, M. Carrithers; S. Collins; and S. Lukes (eds.), 83-92. Cambridge: Cambridge University Press. Morris, C.W. (1938). Foundations of the Theory of Signs. Chicago: University of Chicago Press. (1970). The Pragmatic Movement in American Philosophy. New York: Brazille. (1971). Writings on the General Theory of Signs. The Hague: Mouton. Mosko, M. (1984). Anthropology Peirce-eived. Reviews in Anthropology 2: 226-35. Obeyesekere, G. (1987). Re-reflections on Pattini and Medusa. Contributions to Indian Sociology.21: 99-109. Peirce, C.S. (1931-58). Collected Papers of Charles Sanders Peirce, 8 vols., C. Hartshorne; P. Weiss; and A.W. Burks (eds.). Cambridge: Harvard University Press.

292 Milton Singer (1955[1940]>. Philosophical Writings of Peirce, J. Buchler (ed.). New York: Dover Publications. (1977a). Semiotic and Signifies, C.S. Hardwick (ed.). Bloomington: Indiana University Press. (1977b). The New Elements of Mathematics, III, C. Eisele (ed.). The Hague: Mouton. (1978). Ideas, stray or stolen, about scientific writing. Philosophy and Rhetoric 2: 147-55. Includes a bibliographic note by J.M. Krois. (1982). Writings of Charles S. Peirce. A Chronological Edition, vol. 1, M.H. Fisch, et al. (eds.) Bloomington: Indiana University Press. (1984). Writings of Charles S. Peirce. A Chronological Edition, vol. 2, M.H. Fisch, et al. (eds.) Bloomington: Indiana University Press. (1986). Writings of Charles S. Peirce. A Chronological Edition, vol. 3, M.H. Fisch, et al. (eds.). Bloomington: Indiana University Press. Quine, W.V. (1981). Five milestones of empiricism. In Theories and Things. Cambridge: Harvard University Press. Perinbanayagam, R.S. (1985). Signifying Acts: Structure and Meaning in Everyday Life. Carbondale: Southern Illinois University Press. Radcliffe-Brown, A.R. (1952). Structure and Function in Primitive Society. London: Cohen and West. (1957). A Natural Science of Society. Glencoe, IL.: Free Press. (1958). Method in Social Anthropology, M.N. Srinivas (ed.). Chicago: The University of Chicago Press. Ramanujan, A.K., trans. (1981). Hymns for the Drowning: Poems for Visnu by Nammalvar. Princeton: Princeton University Press. Redfield, J.M. (1975). Nature and Culture in the Iliad: The Tragedy of Hector. Chicago: The University of Chicago Press. Redfield, R. (1953). Relations of anthropology to the social sciences and the humanities. In Anthropology Today, A.L. Kroeber (ed.), 728-38. Chicago: The University of Chicago Press. (1962). Art and icon. In Human Nature and the Study of Society, M.P. Redfield (ed.), 468-489. Chicago: The University of Chicago Press. Roland, A. (1982). Toward a psychoanalytical psychology of hierarchical relationships in Hindu India. Ethos 10: 232-253. (1983). Psychoanalysis without interpretation: psychoanalytic therapy in Japan. Contemporary Psychoanalysis 19:499-505.

Pronouns, Persons and the Semiotic Self 293 Rosen, L. (1984). Bargaining for Reality: The Construction of Social Relations in a Muslim Community. Chicago: The University of Chicago Press. Russell, Bertrand (1948). Human Knowledge: Its Scope and Limits. New York: Simon and Schuster. Saberwal, S. (1985). Review of M. Singer, Man's Glassy Essence. Contributions to Indian Sociology 19: 211-212. Sahlins, M. (1985). Islands of History. Chicago: The University of Chicago Press. Sapir, Edward (1949[1921]). Language: An Introduction to the Study of Speech. New York: Harcourt Brace Jovanovich. Schell, J. (1982). The Fate of the Earth. New York: Knopf. Schiller, F. (1967 [1845]). On the Esthetic Education of Man. Oxford: Clarendon Press. Schutz, A. (1962). The structure of the social world. In The Phenomenology of the Social World, 139-214. Evanston: Northwestern University Press. Sebeok, T.A, ed. (1978). Sight, Sound and Sense. Bloomington: Indiana University Press. (1979). The semiotic self. In The Sign and Its Masters. Austin: The University of Texas Press. (1981). The Play of Musement. Bloomington: Indiana University Press. (1983). One, two, three spells UBERTY. In The Sign of Three: Dupin, Holmes, Peirce, U. Eco and T. Sebeok (eds.). Bloomington: Indiana University Press. (1986[1985]). President Grant's final interpretant. In I Think I Am a Verb. New York: Plenum Press. Seidel, G. (1975). Ambiguity in political discourse. In Political Oratory and Language in Traditional Societies, M. Bloch (ed.). New York: Academic Press. Sewell, D.R. (1987). Mark Twain's Languages. Berkeley: University of California Press. Silverstein, M. (1976). Shifters, linguistic categories and cultural description. In Meaning in Anthropology, K. Basso and H. Selby (eds.), 11-55. Albuquerque: University of New Mexico Press. (1979). Language structure and linguistic ideology. In The Elements: A Parasession on Linguistic Units and Levels, P.R. Clyne; W. Hanks; and C. Hofbauer (eds.), 193-247. Chicago: Chicago Linguistic Society.

294 Milton Singer (1985). Language and the culture of gender: at the intersection of structure, usage and ideology. In Semiotic Mediation, E. Mertz and R. Parmentier (eds.), 219-259. New York: Academic Press. Simon, B. and Weiner, H. (1966). Models of mind and mental illness in ancient Greece. Journal of the History of the Behavioral Sciences 2:303-314. Singer, M. (1936). George Herbert Mead's Social Behavioristic Theory of Mind. M.A. thesis. Department of Philosophy, University of Texas at Austin. (1940). On Formal Method in Mathematical Logic. Ph.D. thesis. Department of Philosophy, University of Chicago. (1961). A survey of personality and culture theory and research. In Studying Personality Cross Culturally, B. Kaplan (ed.), 9-90. Evanston, 111.: Row, Peterson. (1976). Robert Redfield's development of a social anthropology of civilizations. In American Anthropology: The Early Years, J. Murra (ed.), 187-260. St. Paul, Minn.: West Publishing Co. (1980 [1972]). When a Great Tradition Modernizes: An Anthropological Approach to Indian Civilization. Chicago: The University of Chicago Press. (1984a). Man's Glassy Essence: Explorations in Semiotic Anthropology. Bloomington: Indiana University Press. (1984b [1973]). A neglected source of structuralism: Radcliffe-Brown, Russell and Whitehead. Semiotica 48 (1/2): 11-96. (1985). Comments on semiotic anthropology. American Ethnologist 12: 549-53. (1986). "We talk, therefore we are" — some remarks on the semiotics of identity. Journal of Social and Biological Structures 9: 353-64. Spicer, E. (1980). The Yaquis: A Cultural History. Tucson: University of Arizona Press. Spiro, Μ. (1979). Whatever happened to the id? American Anthropologist 81:513. (1982). Oedipus in the Trobriands. Chicago: The University of Chicago Press. Staal, F. (1966). Indian semantics, I. Journal of the American Oriental Society 86: 304. Stocking, G. (1979). Anthropology at Chicago. Chicago: The Joseph Regenstein Library, The University of Chicago.

Pronouns, Persons and the Semiotic Self 295 Swinton, W. (1872[1859]). Rambles among Words: Their Poetry, History and Wisdom. New York: Ivison, Blakeman, Taylor and Co. Tambiah, S.J. (1986). Sri Lanka: Ethnic Fratricide and the Dismantling of Democracy. Chicago: The University of Chicago Press. (1987). At the confluence of anthropology, history and Indology. Contributions to Indian Sociology 21: 187-216. Thompson, J.B. (1984). Symbolic violence: language and power in the writings of Pierre Bourdieu. Studies in the Theory of Ideology, J.B. Thompson (ed.), 42-72. Berkeley: University of California Press. Thompson, M. (1963 [1953]). The Pragmatic Philosophy of C.S. Peirce. Chicago: The University of Chicago Press. Thoreau, Henry D. (1937[1854]). Waiden and Other Writings. New York: Random House. Turner, V.W. (1983). On the Edge of the Bush: Anthropology as Experience, E.L.B. Turner (ed.). Tucson: The University of Arizona Press. (1986). Dewey, Dilthey and Drama: an essay in the anthropology of experience. In The Anthropology of Experience, V.W. Turner and E.M. Bruner (eds.), 33-44. Urbana: The University of Illinois Press. Twain, M. (1889). A Connecticut Yankee in King Arthur's Court. New York: C.L. Webster and Co. (1959). The Autobiography of Mark Twain. New York: Harper. Umiker-Sebeok, D.J. (1977). Semiotics of culture: Great Britain and North America. Annual Review of Anthropology 6: 121-135. Urban, G. (1986). Rhetoric of a war chief. Working Papers and Proceedings of the Center for Psychosocial Studies, No. 5. Chicago: Center for Psychosocial Studies. (1987). The "I" of discourse. Working Papers and Proceedings of the Center for Psychosocial Studies, No. 10. Chicago: Center for Psychosocial Studies. (1988). The pronominal pragmatics of nuclear war discourse. Multilingua 7(1-2): 67-95. Varenne, H. (1983). American School Language. New York: Irvington Publishers. Vendler, Z. (1972). Res Cogitans: An Essay in Rational Psychology. Ithaca: Cornell University Press. Vernant, J.P. (1986). The private man within the city-state. Lurcy Lecture, The University of Chicago.

296 Milton Singer Watt, W.C. (1986). The self itself: review of M. Singer, Man's Glassy Essence. Semiotica 58: 379-83. Weiner, M. (1978). Sons of the Soil: Migration and Ethnic Conflict in India. Princeton: Princeton University Press. Wertsch, J. (1987). Modes of discourse in the nuclear arms debate. Working Papers and Proceedings of the Center for Psychosocial Studies, No. 8. Chicago: Center for Psychosocial Studies. Wheeler, J.A. (1981). "Bohr, Einstein, and the strange lesson of the quantum." In Mind and Nature, R.Q. Elvee (ed.). San Francisco: Harper and Row. Whitman, W. (1977[1855]). Leaves of Grass. M.Cowley (ed.). New York: Penguin Books. Yngve, V.H. (1986). Linguistics as a Science. Bloomington: Indiana University Press. Zellweger, S. (1982). Sign-creation and man-sign engineering. Semiotica 38: 1754.

Afterthoughts Benjamin Lee and Greg Urban

An edited volume is like a self unfolding before the reader's eyes. Its introduction presents what in reality is a retrospective construction as if there were an priori unity among all its diverse components; the afterword struggles to create a stable vision of future possibilities into which the editors hope the work can insert itself. This volume even has an emotional tenor — the respect paid by colleagues and former students to Professor Milton Singer — which acts as the underlying metatextual principle unifying its diverse explorations. Unity among dialogical diversity thus becomes both a discovery that these texts have made about how person, self, and identity are created, and the linguistic metaphor that has animated them. Each paper has analyzed some particular aspect of how lexical items, pronouns, grammatical patterns, or genres contributes to processes of self and other articulation. Despite the specificities of each analysis, their co-presence and inter-textual cross references suggest that future work might try to discover some framework they all share; the trope of dialogue may yet turn out to be more than metaphorical for the study of person, self, and identity. In lieu of such a framework, looking at the development of a specific genre or style in which issues of metalanguage, pronouns, and the representation of subjectivity intertwine will serve here to illuminate some of the dimensions that future work might consider. Bakhtin (1986) has suggested that different genres contain different world views based upon their particular thematic content, style, and compositional structure, a view which adds a functional moment to Whorf s more grammatically focused speculations. Of critical importance for him is the distinction between primary and secondary speech genres. Primary speech genres are parts of everyday speech —jokes, verbal dialogues, everyday narration, etc. — while secondary speech genres such as novels, dramas, commentaries, scientific and philosophical texts, etc. incorporate and represent the primary speech genres. This incorporation and representation is both a decontextualization of the primary genre from its normal situation of occurrence and its recontextualization within the secondary genre. Secondary genres thus stand in a meta-linguistic relationship to primary genres and their existence depends upon the meta-linguistic forms and structures available in their particular languages of expression.

298 Benjamin Lee and Greg Urban Several of the papers in this volume have shown that systematic relations between metalanguage, pronouns, subjectivity, and genres can be found in the ways for reporting speech and thought present in everyday speech and many secondary genres. The primary and secondary genres of what Whorf called Standard Average European (mainly the modern Germanic and Romance languages, including English, French, and German) use both direct and indirect quotation to depict the "inner" mental activities of people. In such secondary genres as novels, these forms become the basis for various types of narration. For example, omniscient narration uses the distinction between reporter and speaker/source as the basis for the contrast between the omniscient knowledge of the narrator and the far more limited consciousness of the characters which can be represented as inner speech or thought using either direct or indirect quotation. Around 1600 there develops first in France and then rapidly spreading across Europe a style for representing consciousness which combines the third person objectivity of indirect quotation with the subjective expressivity of direct quotation. Variously known as "style indirect libre," "free indirect style," or "erlebte Rede," it is championed by Flaubert, and becomes widely used by such French novelists as Zola and Maupassant; its English practitioners include Jane Austen and Henry James, and it also becomes one of the styles characteristic of the genre of stream of consciousness novels. In this style, the time of the verbs (is characteristically past) and the person of the pronouns is typically third person. These forms are combined with subjective expressions usually associated with direct speech and the spatial and temporal deictics normally (»referential with the present tense become coreferential with the past tense. Unlike inner speech derived from direct quotation, all forms dependent on the second person, such as imperatives, are eliminated; it is as if there were an objective, third person view of a subject whose consciousness unfolds before the reader's eyes and whose narrated monologue loses all traces of communication to some second person addressee: Tomorrow was Monday, Monday, the beginning of another school week (Lawrence). La maison 6tait bien triste, maintenant (Balzac). Aber am Vormittag hatte sie den Baum zu putzen (Berend).

Afterthoughts 299 The effect of the style is, as Flaubert put it, to "transport oneself into the characters, not draw them to oneself by creating an objective report of the character's mental activities which reduces the distance between narrator's and character's consciousness. In some cases there can be an almost complete identification where it is difficult to determine whose voice is speaking, that of the narrator or that of the character, as in the following passage from Jane Austen's Emma: Never had she felt so agitated, mortified, grieved at any circumstance in her life. She was most forcibly struck. The truth of his representation there was no denying. She felt it at her heart. How could she have been so brutal, so cruel to Miss Bates! (Austen 1966: 376). In the development of this style into what has been called stream of consciousness technique, the balance between the indirect style and omniscient narration that is present, for example, in Austen's novels, begins to shift in favor of the dominance of the indirect style of narration; this is accomplished by such techniques as the reduction of linear plot sequence and a subtle playing with the different frames which can be used to represent speech and thought, resulting in the creation of an image of a continuously unfolding internal mental world, as the following passage from Virginia Woolfs Mrs. Dalloway indicates: What a lark! What a plunge! For so it had always seemed to her, when, with a little squeak of the hinges, which she could hear now, she had burst open the French windows and plunged at Bourton into the open air. How fresh, how calm, stiller than this of course, the air was in the early morning.... (Woolf 1976: 3). In still later forms, free association is introduced which breaks up the unity of this inner "stream" into a multi-stranded flow. Just as Flaubert thought that "style indirect libre" was a way of exploring consciousness which objectively revealed what had previously been obscured, so too did champions of stream of consciousness believe that they had uncovered a new level and kind of mental activity. However, several questions of a Bakhtinian and Whorfian nature immediately arise. Did these genres discover something new

300 Benjamin Lee and Greg Urban about consciousness which other genres had failed to reveal, or had they discovered forms specific to their socio-historical periods? Or, pushing still further, perhaps the views of subjectivity were creations of the genres and reflected nothing at all about consciousness. But if every view of consciousness is presented through some secondary genre, how do we know what is the product of the genre and what preexists it? Is there a consciousness independent of our narrative frames and how do we get at it? A Whorfian might reply that the only way to discover what such a consciousness might look like would be to systematically compare different views and discover what was common to all of them or at least what systematically covaried. For example, in the Chinese novel there is also a form which blends properties of direct and indirect quotation, possesses no second person forms, and is used to represent psychological activity; when translated into English, the free indirect style is usually employed, as the following passage from the eighteenth century Chinese classic The Story of The Stone shows: Lin Dai-Yu dead! A world from which that delicate, flowerlike countenance had irrevocably departed! It was unutterable anguish to think of it. Yet his sensitized imagination did now consider it — went on, indeed, to consider a world from which the others, too — Bao-chai, Caltrop, Aroma and the rest — had also irrevocably departed. Where would he be then? What would become of him? (Xueqin (1986 [1977]: 41-42). Among modern writers, this style develops into a form which in translation seems similar to the passage from Mrs. Dalloway quoted earlier: He had not been back to his old home for more than twenty years. Whose fault was it that he had been born from the wrong womb? Landlord, landlord! He had gone back once in 1956, once was enough for loneliness — he only stayed at home four days, but his self-criticism lasted twenty-two years! What troubled him was that the purpose of life could surely not be self-criticism. Luckily, all of that was in the past (Wang Meng 1983: 140).

Afterthoughts 301 If other comparisons yielded similar results, could we not postulate something like a Cartesian-like representation of internal consciousness? Or could we not at least attribute such a model to the Chinese? Yet the Chinese case turns out to be more complicated than at first glance despite the similarities. In Standard Average European, verbal tense is obligatory and free indirect style can be easily identified by such syntactic phenomena as the coreferentiallity of normally present time deictics with the past tense (as signalled in the verb). The relatively clear boundaries between this style and direct and indirect quotation allow the author the possibility of a continuous and unambiguous use of the form to emphasize the continuity of internal consciousness. In Chinese, the lack of verbal tense, the aspectual system of Chinese verbs, and the wide latitude in anaphoric deletion makes it often impossible on a purely syntactic level to distinguish the three forms; the surrounding context may be the only key to disambiguating free indirect style from direct or indirect quotation. Does this difference mean that the Chinese have no notion of a Cartesian-like subject? In the Chinese case, the narrative creation of such a subject would require a consistent contextual disambiguation of the form. Since the surrounding contexts are subject to constraints of the given genre and genres themselves are supra-individual products of cultural-historical forces, the analysis of subjectivity would therefore immediately lead to an anthropology of discursive practices. At the same time that comparison brings into question the idea of a consciousness independent of our genres for representing it, it also provides an insight into why we might think that there is such a consciousness. It is precisely the regularities of the metalinguistic relations between secondary and primary speech genres encoded in forms such as pronouns that are the ground for our representations of self and subjectivity as genre independent. The free indirect style, by blurring distinctions present in direct and indirect quotation, seems to provide a transparent window into a consciousness whose expression of its subjectivity seems to be independent of any act of communication directed towards another, as the lack of any syntactic forms based upon the second person seems to indicate. Since this form can be unambiguously differentiated from the other ways of representing speech and thought, we can create the representation of an autonomous subject by separating this style from its genric contexts and then "objectifying" the form of consciousness implicit in it; the Chinese example provides the contrasting case where the style that contains the representation of such a subject cannot be systematically grammatically disambiguated from the other modes of representing speech and thought. Our model of the Cartesian

302 Benjamin Lee and Greg Urban subject which is independent of any form of communication may itself be dependent upon the metalinguistic relations between genres which allow for its possibility. However, if our representations of consciousness are genre dependent, then all the complications raised by the Chinese example re-emerge. The problem of the relation of language to self and identity formation forces us back to the dialectic between language and culture which helps to constitute iL In both the Western and Chinese cases we find a similar use of direct and indirect quotation to form a new, hybrid style which is increasingly used to depict psychological activity. There are syntactic similarities as well, including a Cartesian-like moment in which all references to a second person addressee are removed, producing not an interior monologue based upon direct or indirect quotation, but a "narrated monologue" in which the subject seems to be expressing himself to no one. As soon as we turn to the contexts in which these forms occur, however, the contrasts emerge. The difficulty of maintaining an unambiguous usage of the style in the Chinese case produces a model of self in which there is a constant intermingling of different narrative voices, with no core of Cartesian autonomy. Chinese writers constantly exploit this "ambiguity" for creative effect, producing various indigenous genres of "self-narration" which depend not only upon the particular metalinguistic resources available for representing consciousness but also the metalinguistic relations between genres. If a genre is seen as the intersection of cultural-historical and linguistic processes which mutually influence one another, then a view of an autonomous self can only be created by holding both language and culture constant. These issues all come to the fore in a letter written by Wang Meng, who later became the cultural minister of China, to some students about his use of stream of consciousness technique. He had been the major proponent of this style in China, leading some conservative thinkers to criticize him as a modernist. Since then, modernism has become a term of approbation, reflecting the political struggles in which literature is involved in China. The style that Wang Meng uses is representative not only of a particular way of depicting consciousness, but of the intellectual movement of opening up to the West, a microcosm of the issues facing modern China and also facing us as analysts: I have to admit that not long ago I read some foreign "stream-ofconsciousness" novels and stories. After finishing some of them I felt just as you did — they made me feel confused. Naturally, I can neither accept nor indiscriminately imitate those kinds of

Afterthoughts 303 morbid, abnormal, mystical, or solitary mental states. They did, however, inspire me to do something: write about people's feelings.... Of course, we must be fully aware of the dangers of assimilating and drawing lessons from this technique. We are in absolute disagreement with mysticism and antirationalism. We believe that it is wrong to place feelings, impressions, and thought associations in opposition to reflection, conceptualization, and judgement We believe that the former are the foundations for the latter and that the latter are the results of the former.... We use a little "stream-of-consciousness" not in order to go insane, or to give vent to some tragic mal de siecle; but rather in order to portray a more profound, more beautiful, more ample and more civilized soul.... For all these reasons, our "stream-of-consciousness" is not a stream-of-consciousness that urges people to escape reality by an inward flight; it is rather a healthy and substantial self-feeling that urges people to face both the objective and the subjective worlds, to love life and to love the human heart (Wang Meng 1984: 25-28).

References Austen, J. (1986). The Novels of Jane Austen, Vol. IV, Emma, R. W. Chapman (ed.). London: Oxford University Press. Bakhtin, M.M. (1986). Speech Genres and Other Late Essays, V.W. McGee (trans.). (=University of Texas Slavic Series 8). Austin: University of Texas Press. Meng, W. (1983). The Butterfly and Other Stories. Beijing: Panda Books. (1984). An open letter on stream of consciousness. In Modern Chinese Literature 1(1), 25-28. Woolf, V. (1976). Mrs. Dalloway. London: Hogarth Press. Xueqin, C. (1986[1977]). The Story of the Stone, Vol. II, David Hawkes (trans). New York: Penguin Books.

Index

"aloneness", culturally defined, 6-7, 74-78, 80-83 "aloneness-disorder," 3, 6, 69,73-78, 80 fig., 83-84, 83 fig., 85, 91-95 (see sociality; suicide) amae, 3 American individualism, 247,251, 279, 281 anaphora coreference, 31-34,35,45 device, linguistic, 27, 30-31, 3435,41 ' Τ , 29-38, 41-42,46,49-50 indexicality, 31, 41 (see quotation) anthropology critique of discourse, 124-27,13336,146 n. 8,153-69 cross-cultural therapy, 154-58 as discursive practice, 121-29, 139-44,153-54,156 of Ireland, 9,127-39,142 and the other, vi, 125-27,130-31, 145 η. 3,160 "self-help", 10,153-56,167-68 semiotic, 1, 96, 123-27, 139-44, 156, 161, 171-72, 174-75, 301 Appadurai, Arjun, 73 "ecology of effect", 73,96 Augustine, Saint, 256, 260 Austin, John L., 209-11, 231 Bakhtin, Mikhail, 12 n. 3, 29. 53, 142, 297-98, 301-2 Bastian, Adolf, 174 Beatles, the, 159, 241-42, 278 Benedict, Ruth, 132, 261

Benveniste, Emile, 27-31, 37, 48, 209, 251, 252, 258, 264 Bertillon, Alphonse, viii Boas, Franz, 174,186 body, viii-ix, 2-4, 6, 8-9, 10-11, 16, 221,224-26 (see mind, body dualism) (see mind, body dualism) Boon, James Α., 9,10 Bourdieu, Pierre, 252, 254 Buber, Martin, 173 Coleridge, Samuel Taylor, v, 240, 244 communication, patterns in, v, 1-2, 8, 27-29, 56, 172, 174-75, 181, 184-85, 187, 232-42, 264-66, 272-84 consciousness, vi, 15, 56, 90,143, 18184, 185, 208, 223, 226, 23032, 240, 249, 272, 299-303 alteisense, 89,92 culturally specific, 4,298-303 in literature, 243,297-303 self-, 15, 177, 185, 193, 223, 227,232, 249, 255-59 stream of, vii, 123, 176, 281, 299, 303 (»reference, 30-34,35,207 linguistic universal, 33-34 cultural code in discourse, 8, 29, 30, 35-41, 298-303 nosology, 69, 74-78, 80-83, 8688, 91-95 referential, of self, 1-4, 63-67 culture and anaphora, 35-38,41,45-46, 49

306 Index Arabic, social relations in, 8-9, 101-6

Cheyenne, Northern, 5-6, 53-67 and Islamic law, 106-116 Peircean third, 91 performances/ritual, 36-41,44-48, 48-50,174-75, 186 personality studies, viii, 261 and responsibility, Arabic, 101, 102-3, 105,111-13 and self, 1-6, 9,29,46-50 as semiotic process, 35-36,46-48, 86, 91-96,175, 239 Tamil, categories of, 6-7, 72-73, 74-78 (see Singer, Milton; tuism) Daniel, E. Valentine, 3, 6-7, 274 deixis, 4, 201, 207 pronominal, 3, 121, 243 De Valera, Eamon, 133, 135, 138, 141, 144, 147 n. 14, 148 n. 15 Descartes, Rene, 2, 96, 183, 188n. 1, 190, 211-27, 260, 281 cogito ("I"), 4, 10, 12 n. 3,193, 209, 217-26, 231-32, 247, 25557. 261, 272, 301-2 dualism, 195, 209, 212, 217-26, 231,257 language and thought, 4, 221-27 method, 218-25 mind, 199-200, 211-12, 224-26 self, 173 Dewey, John, v, 117 n. 4, 132, 176, 188 n. 2-190, 261-62, 266, 28384 discourse^) anthropological, and representation, 121-44,153-69 competing, and sense of self, 811, 128, 136, 141-44 (»reference in, 30-34,35 direct/indirect, 31-34,199-209, 226 legal, 7-9, 11 pronominal reference in, 29-50, 174 therapeutic, 7-9,11, 153, 154-58, 167 Doi, Takeo, 3 Dumont, Louis, 255-56

Dürkheim Emile, 73, 96, 177, 249, 250, 254, 256, 258-62 East/West, 159-61,167,270-75 ego, 92, 184, 187, 218, 255-57, 260, 269, 272-73 and self, 15-17,19-20, 53, 65, 270-75 id and, 15-16, 20. 23, 50, 266-67, 269, 273-74, 275 in Kant, 230-31 pronoun "Ich", 15 psychoanalysis, 18-20 super-ego, 15-16, 20, 23, 36, 50, 273 "system", tripartite, 15-16, 26567,273 (See Descartes, cogito; "I") ego psychology, 18-19 egotistical era, 240, 249, 251, 273 Emerson, Ralph Waldo, 240, 243, 279 epistemology, 172-86,188 n. 2-190, 219-24, 261-66 Erikson, Eric, 267-70, 273-75 ethnography, viii authorial 1,121-26, 137, 141, 153-58,162-67, 247-49, 274, 275-79 Ireland, 127-30,136-39, 142,146 n. 8 of the "other", 125-26, 130-33, 157, 159-60, 168 semiotic nature of, 82, 139-44, 168-69,174 firstness, 78, 88-90, 92-93, 234, 265, 272 Frege, Gottlob, 195, 199-200, 239 Freud, Sigmund, 15-18, 23, 24 nn. 2-3, 261, 265, 269 Freud-Groddeck theory, 266, 273 functionalism, 186 Geertz, Clifford, 250,261 gender. 28. 54-57,58 Goffman, Erving, xii Hegel, G.W.F., 177 Hose, Charles, 153,162-68 Hume, David, 10, 178-81, 183, 229, 231, 281 Hymes, Dell, 171

Index 307

"Ι", ν, 4 "anaphoric" 29-38,41-42,46,4950 authorial, 243-249,279-281 Cartesian 220-22, 227,231 "de-quotative" 27, 36-38,42-45, 47-48, 49 in discourse 27-50, 43 fig. index referential 27-31, 37-38, 4142, 47-48, 49-50 in Kant 230-31, 259 "metaphorical", 27, 36 in Peirce, 264-267 "projective" 37, 39, 42-46, 47, 49 "theatrical" 36^1, 42-44,47-48, 49 token-reflexive, def of, 28 (see Descartes, cogito; ego; Kant; quotation) icon (sign), 46-48, iconic diagram, x, 237-39, 242, 282 idistical era, 240, 241, 247, 249, 275,272-73 index (sign) to self, 1-2, 5 and anaphora, 27, 30-31, 35-37, 41 referential, 27-29, 30-38, 81, 196, 200, 216, 264, 270-72 India, 153, 159, 161, 174, 229 interpretant, 144 defined, 78-80, 84 in Peircean triad, xii, 69, 78-82, 79 fig., 83-86, 88-95, 230, 232, 263-64 Jakobson, Roman, 233, 258 James, William. 173, 176, 229, 23435, 243-44, 252, 261-62, 266, 278, 281 Joyce, James 131, 136, 138, 140, 142,147 n. 10, 159 Jung, C.G. 261 Kant, Immanuel 10, 229, 259, 281 pragmatics, 172-74,178-86 social psychology, 181-84, 187 self in, 178, 181-83, 226, 230-31, 279 and semiotics, 171-87 subject-object dialectic v, 174, 180, 182-84, 230

transcendental in, 173-74,181-84, 230, 265 Kohut, Heinz, 2, 18-20, 36, 265 language Algonquian, 54 analogy, 194-200, 225 Cheyenne, Northern, 53-67 Chinese, 300-3 construction of self, 2, 27-29, 59, 258-60 and culture, 4-6,193-209 English, 30-33, 54, 130-31, 135, 160, 197, 199-209, 214-15, 234, 275-79, 282 French, 199-209, 215, 259-60, 275-79, 282 Gaelic, 130-32 Hopi, 197 Irish, 130-31 Latin, 199-200, 202-4, 209, 21415,234 mental process, 194-95 and national identity, 130-36, 14243, 147 n. 12 reference, 215 Sanskrit, 159, 260, 272-274 Sapir and Whorf, 186, 216 Shokleng, 30, 32-34, 38-41, 42, 44-46,247 Standard Average European (SAE), 195-209, 227, 298 Tamil, 3, 6-7, 69, 73-75, 270-72 (see culture) law. Islamic, 108-112 classical, foreign influences on, 114-16 cultural dimensions of, 5, 8, 101, 106-111, 116 hadd/ta'zir, 108-9 justice, 112-13 Leach, Sir Edmund, 156,167 Leaf, Murray J., 7, 229-232 Lee, Benjamin, 2,4-5, 10, 231-32, 247 Le'vi-Strauss, Claude, 121, 156, 233, 247-49,275-79 structuralism, 239, 247, 254, 274, 275-79, 286 η. 1 Malinowski, Bronislaw, 174, 186 "man's glassy essence," vii (see Singer, Milton)

308 Index Mauss, Marcel, 1-4, 7-8, 10, 274 category of person, 1,259-60 Mead, George Herbert, 7,132,173, 175, 185, 249, 252, 258, 26162, 270, 273, 285 η. 1 Mead Margaret, 155,157,261 memory, 283-84 individual and, vi-vii, xii and narrative, 39,44,49 Mehta, Gita, 153-55, 158-62 metaphor indexical references of, 34-36 mind, vii, 173, 184, 185, 188 n. 1, 194, 199-200, 215, 264, 285 η. 1 body dualism, 195, 212, 216-17, 224, 257 (see Descartes) Morris, Charles, v-vi, xii, 236, 26162, 285 η. 1 social phenomenology, vi, myth, 28, 32, 38, 39-41, 44-46, 47, 128, 130, 248, 276 Naipaul, V.S., 162 narrative discourse indexical cues to multiple selves within, 37-48,49-50 nationalism Gaelic/Irish, 130-33,135,142-43, 146 n. 9 nouns, 27-28, 33, 55, 196-97 oath, 104, 106-7 other and the unconscious, viii, and the ethnographer, vi, 125-26, 157, 159-60, 168 (see self) panpsychism, ix Peirce, Charles Sanders, v-vii, ix, xi, 1, 29, 53, 218 anti-Cartesianism, 193, 229-30, 232, 249, 255-58, 264, 272-75, 279-82 communication and signiGcation, 123, 233-39, 264-66, 272 and consciousness 177, 240, 263, 272 dialogical self, 123, 124,143-44, 177, 232-33 and Freud, 266-70

historical eras, triad of, 240-43, 249,279 and Kant, 171-74,182-84,188-90 nn. 1-2, 229-31 phenomenology 78, 81, 85, 88-95, 96, 234, 257, 262-66, 273 phenomenology of self, 261-66 pronominal triad, xi, 172-77,18486, 190 n. 4, 229, 233-39, 24042, 255-56, 264-69, 272-75, 279, 280 fig., 284 psychology, 92 semiotic formulation of, 69, 78-81, 84-86, 88, 92. 232-39, 264-66, 272 sign triad, 69, 78-80, 79 fig., 8892, 230, 232-33, 235-39, 241, 253-64, 272 (see symbolic interactionists) person in Arabic culture, 101-7, 111 category of, vii-xi, 1, 7-9,16, 5357, 111, 184-85,186, 255-61 hierarchy in Cheyenne, 59-64 in South Asia, 90, 98 n. 10 pronouns, personal, v, 2-4, 11, 2729.260 authorial use of. 240-55,275-81 in Cheyenne (Algonquian), 53-55, 57-63, 60 fig «»reference, 30-32 cultural dialogicality, 4-5, 54, 58, 239-44, 246-47,249-55, 266-67 dicourse and, reference in, 4,27-50, 43 fig., 54, 58 in a dyad, 3-5, 8, 28, 29,45, 157, 185,234, 242-43, 247-58, 26769 first person singular, 3-5, 15, 2733, 35-37, 39-41, 44, 49, 57-61, 63-64, 121, 214, 240, 243-49, 251-52,256, 275-79 fourth person, 61-63 and gender, 28,54-58 hierarchal ordering of, Cheyenne, 59-64 Kin terms, 65-67 linguistic universal, 29-30, 234, 266-70 in literary genres, 240-51 and the other, 5

Index 309 and reference, 2, 3,27-31,33-34, 38, 54, 59-64, 201, 214, 237-55 refering to self, 3-5, 63 second person, 5, 27-29, 32-33, 35, 38, 53, 55, 57-61, 63, 67, 240, 246-47, 298 semantic reference, 27-28,31 semiotic function of, 27-29, 36-42, 4446 third person, 4, 5, 11, 28, 30, 33, 39-41, 44, 57-63, 145 n. 3, 214, 240, 264, 269, 298 "we", v, 247-55, 275-82 (See anaphora; "I"; Peirce, C.S.) performative statements, 209-213, 227 phenomenology, v, 81, 96 Plato, 178, 245, 256, 258 political discourse and self identity, 7-9 pragmatism, 174, 176, 179-88 proper names, 236-37 psychology, 2-3, 15, 18-19, 20-23, 185-86, 265 psychoanalysis, 2, 6, 17-23, 29, 36, 50, 265, 267-70, 274 empathy (einfuhlung), 17-20, 23, 23 η. 1 quotation, 4, 27-29, 36, 200-203, 205-9 as metaphor, 34-36 direct/indirect, 4, 30-34,37-38,4144, 47, 48, 202, 213-14, 298, 300-301 iconic reference, 34 (see coreference) Quran, 101, 1024, 108 Radcliffe-Brown, A.R., 253-54,261, 286 η. 1 Radin, Paul, 261 Redfield, Robert, 176, 184, 229, 249, 261 responsibility in Islamic Tradition, 101, 105, 111-13 Ricoeur, Paul, 27, 29, 276 Rorty, Richard, 172, 182, 188-90 nn. 1-2 Rosen, Laurence, 5, 7-8, 274 Russell, G.W. (AE), 132

Sapir, Edward, 171, 186, 216, 261, 282 Saussure, Ferdinand de, 29,258 Schiller, Frederick, 267, 273, 279 Schutz, Alfred, v-vi, 250, 261, 270 secondness, 78, 88-93, 96, 234, 265, 272 self Arabic narrative, 103-4 biological individual, vi, viii-xi, 1-4, 810, 16 and the body, viii-ix, 1-10, 16 Buddhist idea of, xii Cartesian, 173 classical Greece, 255-60,275 and consciousness, 143,181-84 cross-cultural comparisons, 3-10, 270-275 cultural constructions) of, 1-9, 27, 29, 50, 54-56, 63-67 cultural performance and, 35-36, 41-48, 186 cultural universal, 1-4,243-44, 259-60 dialogic nature of, 5,12 n. 3,29, 53, 58-67, 105, 172, 186, 193 embedded in discourse, 10,27-29, 35-36,48-50 experience of, 3, 6-9, 16 god, 173, 183, 190 n. 3, 258-59 and individual. Western philosophy of, 1-11 and individuality, vi, xii, 2-4,11, 20 in narrative, 28, 37-46,49 Indian idea of, 175,184-185,260, 270-275 language and culture, 27,29,4850, 59, 239-60, 275, 297-303 multiple identities of, 46-50,171, 177-78,184-85, 229-33 not-self/other dialectic, vi, 1,5, 6367, 89 other, vi, xii, 5-7, 11, 29, 53-54, 143, 157, 168, 186, 267-69 otherness, iconicA'eferential, 29, 46-48,48 fig personal sense of, vii, 2-4 psychoanalytic approach to, 2,1523,272-75 reflexive awareness of, 1-2 and semiotics, v-xii, 1-11, 123 and signs, viii, 229-84, 285 η. 1

310 Index and unity, 2,11 violence, 70, 95 (see ego; "I"; person, pronouns, personal) self psychology, 2, 3, 15, 19, 20-23, 36,265 selfobject(s), 2-3, 6,11, 21-23 semantic(s), viii, 74, 174 pronominal reference, 27-28, 3031,34-35, 41,237-39 semiotics (semeiosis), 82-83, 96, 236 anthropology, 1, 96, 139-44, 156, 161 biology, vii-xii, 1-3 and Cartesian empiricism, 96, 22930 and community, 95-96 communication, viii-ix, 34, 172, 174-75, 187, 188 η. 1 definition of, 230, 233 de-quotation, 36-41,47 epistemology, 5, 171, 175-79, 229-33, 275 intellectual tradition of, v-vii, 1-11, 27, 171-87, 191 n. 5, 249, 25962, 266-67, 270, 281, 285 η. 1 and language, 27-29, 37-50, 74, 172, 188 η. 1, 193-95, 233 Peircean, 69, 78-80, 84-86, 88-92, 168, 194 and phenomenology, ν-vi psychoanalysis, 265, 266-70, 27375 self-awareness, 1-2, 5-6 symbolic process, 173, 174, 187 (see communication; firstness; Peirce, C.S.; secondness; symbolic interactionists; thirdness) semiotic self, v-xii, 123, 169, 172, 187, 233-35, 262, 281, 284 pronouns and, 3-5, 8, 172, 242, 270-75 sign(s/representamen), 230, 232 anaphoric substitute, 31, 41 dyadic relation of, 90 iconic, x, xi, 34, 44, 46-47, 81-82, 90, 95, 235-39, 265-66, 272 ideas as, 173 indexical, 1, 5, 27, 28, 30, 34-35, 37, 42, 46, 81, 90, 196, 200, 216, 235-39, 265-66, 271-72

interpretation of, xii, 81-82, 84-85, 88-94, 144, 195-209, 214-16, 230-31,235-39,271-272 Peircean definition of, 78-80 systems of, vi-viii, 1-2 symbol, 81, 90-91, 128,136, 15758, 235-39, 265-66, 283-84 as thought, 174, 176-77, 178-84, 194-95,199-200, 211-17, 22027, 231, 236 (see interpretant) Signification patterns of, v, 2-29, 30-34, 41-48, 195-200 Silverstein, Michael, 27, 232, 247 Singer, Milton, v, vii, xi-xiii, 1, 11, 27, 29, 49, 53, 69, 90, 193 "cultural performances", 174-75, 185, 186 epistemology, 178-79,186 Man's Glassy Essence, 124, 153, 175, 229-33, 263 and Peirce, 171, 173-75, 176, 177, 184, 187 and semiotics, 1, 173, 175-78, 187, 194 semiotic comparison, 168-69, 25355,270-72 semiotic self, 172-73, 175, 187 sociality, 73, 87, 91-92, 102, 109 "we-reality," v, 250 society, individual and, xi, 29, 35-36, 46, 53-57, 63-66, 87, 91-92, 101-4, 113, 177-78, 181, 18385, 247-55, 259-62, 281 (see "aloneness"; culture; self semiotics;) soul, 173, 182-84, 221-24 Straus, Terry, 5-6, 243, 274 structural analysis, 29, 135, 142, 143, 247, 253-54 suicide, 69, 70-73, 87-88, 95 symbol (see sign) symbolic Interactionists, tradition of, 173, 177, 229-30, 232, 261-62, 266-67, 270, 282-84 symbolic process, 173, 174, 177 Synge, John Millington, 131, 142, 147 n. 12 thirdness, 78, 88-94, 234, 265, 272

Index 311 Thoreau, Henry David, 243, 244, 279, 281 thought, xii, 17, 176, 194, 212-13,

221

token, linguistic, 27, 28, 31, 47, 81 trance/possession, 37, 39, 41, 42, 45, 77-78 translation, 214-15, of Livi-Strauss, 248-49,275-78 of literature, 298-303 of pronouns, 32-33, 58 Strachey, of Freud, 15-16, 23 η. 1, 24 nn. 2-3, 267 tuism cultural, Cheyenne, 6, 53, 58-63, 66-67 historical, 234, 240-41, 243, 249, 251,272-73 Tumbull, Colin, 153-59 161-62, 167-68 Turner, Victor, 250, 261 Twain, Mark, 24647, 278 Type, linguistic, 28 Urban, Greg, 2,4-5, 247, 266 Varenne, Herv6, 9,121 Vendler, Zeno, 209-17, 222, 225-27, 231-32 Whitman, Walt, 243-47, 251, 270, 274-75, 278, 279, 281 Whorf, Benjamin Lee, 186,195200, 209, 211, 216-17, 227, 232, 298-300 Wittgenstein, Ludwig, 193-95, 199, 218, 232, 261 grammatical analogy, 194-95, 199200 Wolf, Ernest S„ 2-3, 265, 274 Wundt, Wilhelm Max, 174, 185-86, 189 Yeats, William Butler, 130-31, 133, 135, 138, 141, 143, 145 n. 4, 159

m

m m m m m m

m m

m m m

mi

Bruce B. Wavell

Language and Reason 1986.14.8 cm χ 22.8 cm. ΧΧΠ, 353 pages. Cloth. ISBN 3110106841 (Approaches to Semiotics 74] A book of philosophy, Language and Reason describes the role of common sense as the origin of both scientific reasoning and the human decision-making process. The book explicates the subtlety and sophistication of the deliberative procedure of common sense or "natural reason". The presentation of a "natural logic" derives its elements from an analysis of the structure of language, offering a comprehensive semiotic for linguistic analysis and an analysis of common sense deliberation, the product of centuries of evolution. Scientific reason, which has fallen into neglect in our cultural preoccupations with science and technology, now urgently needs rediscovery and refinement.

m m

m

mouton de gruyter Berlin · New York

mouton de gruyter Berlin · New York

MULTILINGUA Journal of cross-cultural and interlanguage communication Editor-in-Chief: Richard J. Watts, University of Berne, Switzerland MULTILINGUA was first published in 1981. As of 1987, Volume 6, the journal has been reconstituted under a new editorial board redirecting its scope to cross-cultural and interlanguage communication. MULTILINGUA is an international, interdisciplinary journal aimed at the enhancement of cross-cultural understanding through the study of interlanguage communication. To this end it publishes articles and short notes in fields as diverse as cross-cultural differences in linguistic politeness phenomena, variety in what is traditionally regarded as one culture, conversational styles and the linguistic description of nonstandard, oral varieties of language, strategies for the organization of verbal interaction, intracultural linguistic variety, communication breakdown, translation, information technology, and modern methods for managing and using multilingual tools. MULTILINGUA considers contributions in the form of empirical observational studies, theoretical studies, theoretical discussions, presentations of research, short notes, reactions to recent articles, book reviews, and letters to the editor. MULTILINGUA is published in four issues per volume (approx. 448 pages). Subscriptions and single or back issues can be ordered from your local bookseller or any subscription agent or directly from MOUTON DE GRUYTER (a Division of Walter de Gruyter) at either of the following addresses: For North America Walter de Gruyter, Inc. 200 Saw Mill River Road Hawthorne, NY 10532 USA

For all other countries Walter de Gruyter & Co. Postfach 1102 40 D-1000 Berlin 11 Federal Republic of Germany