[Holquist is an American critic, educator, and translator whose works include Dostoevsky and the Novel: The Wages of Biography (1977) and the biography Mikhail Bakhtin (1985, with Katerina Clark). In the following excerpt from his book Dialogism: Bakhtin and His World, Holquist traces fundamental issues in Bakhtin's theories of language and society.]

Mikhail Bakhtin made important contributions to several different areas of thought, each with its own history, its own language, and its own shared assumptions. As a result, literary scholars have perceived him as doing one sort of thing, linguists another, and anthropologists yet another. We lack a comprehensive term that is able to encompass Bakhtin's activity in all its variety, a shortcoming he himself remarked when as an old man he sought to bring together the various strands of his life's work. At that time he wrote:

our analysis must be called philosophical mainly because of what it is not: it is not a linguistic, philological, literary or any other particular kind of analysis…. On the other hand, a positive feature of our study is this: [it moves] in spheres that are liminal, i. e., on the borders of all the aforementioned disciplines, at their junctures and points of intersection.

But if we accept even so privative a sense of "philosophy" as a way to describe the sort of thing Bakhtin does, the question remains: what kind of philosophy is it?

Stated at the highest level of (quite hair-raising) abstraction, what can only uneasily be called "Bakhtin's philosophy" is a pragmatically oriented theory of knowledge; more particularly, it is one of several modern epistemologies that seek to grasp human behavior through the use humans make of language. Bakhtin's distinctive place among these is specified by the dialogic concept of language he proposes as fundamental. For this reason, the term used in [Dialogism: Bakhtin and His World] to refer to the interconnected set of concerns that dominate Bakhtin's thinking is "dialogism," a term, I hasten to add, never used by Bakhtin himself. There can be no theoretical excuse for spawning yet another "ism," but the history of Bakhtin's reception seems to suggest that if we are to continue to think about his work in a way that is useful, some synthetic means must be found for categorizing the different ways he meditated on dialogue. That is, some way must be found to conceive his varied activity as a unity, without losing sight of the dynamic heterogeneity of his achievement. Before looking at any of Bakhtin's particular works, it will be useful to have some sense of the ideas that permeate them all. This [essay] will seek, then, to lay out in a general way some of the ideas considered by Bakhtin at the beginning of his career, and which—with different shifts of emphasis and new accretions of significance—he never ceased to hold.

Dialogue is an obvious master key to the assumptions that guided Bakhtin's work throughout his whole career: dialogue is present in one way or another throughout the notebooks he kept from his youth to his death at the age of 80. Most of these are lost, some remain in the form of communications so self-directed they are now almost impossible to decipher or understand, while others eventually took on the more public and comprehensible form of published books. But early or late, no matter what the topic of the moment, regardless of the name under which he wrote or the degree of shared communication he presumed, all Bakhtin's writings are animated and controlled by the principle of dialogue. It is becoming increasingly evident that Bakhtin's lifelong meditation on dialogue does not have a place solely in the history of literary theory, capacious as the borders of that subject have recently become. It is now clear that dialogism is also implicated in the history of modern thinking about thinking.

In this it is far from unique: the work of many other recent thinkers, especially in France, combines literary criticism, even literary production, with concerns that are essentially philosophical. But the kind of literature and the kind of philosophy that are woven together in the writings of a Sartre or a Derrida constitute genres significantly different from those that characterize dialogism. Rousseau, Hegel, Nietzsche, and Heidegger, the philosophers recently "discovered" by students of literature, represent, not surprisingly, the literary aspect of philosophy. They are lyrical thinkers, some of whom set out consciously to poeticize metaphysics.

Bakhtin is working out of a very different philosophical tradition, one that is little known, even among many Anglo-American professors of philosophy. The men who constitute a dialogizing background for Bakhtin differ from most thinkers now in fashion in so far as they were, in their own day, very much in the mainstream of academic philosophy. They held chairs in the important German universities and sought to make metaphysics even more systematic than had Hegel (most were, in fact, militantly anti-Hegelian, as was Bakhtin himself). Systematic metaphysics is now out of fashion and the names by which philosophy was defined in the latter half of the nineteenth century are for the most part forgotten. It is difficult for most of us now to conceive the passion excited in their time by such men as Hermann Cohen or Richard Avenarius. And if we take the trouble to look into their books, it becomes even harder, for they are written in the forbidding language of German technical philosophy in one of its more complex phases. And there are very few translations. I mention this tradition (emphatically) not to scare anyone away from a deeper involvement in Bakhtin's philosophical roots, but only to make it clear that such an involvement requires the extra effort always required to go beyond the categories and concepts (and translations) currently in fashion.

Dialogism, let it be clear from the outset, is itself not a systematic philosophy. But the specific way in which it refuses to be systematic can only be gauged against the failure of all nineteenth-century metaphysical systems to cope with new challenges raised by the natural and mathematical sciences. The most spectacular of these failures was the increasingly obvious irrelevance of Hegelianism (right or left) to the new scientific discoveries. As a result, from the 1860s on, more and more attention was paid to Kant: by the 1890s Neo-Kantianism in one form or another had become the dominant school of philosophy in Germany—and Russia.

There are many reasons why the rallying cry "Back to Kant!" proved so successful, but chief among them was a compatibility between Kant's work and developments in the realm of science outside philosophy. Kant himself had taught scientific subjects for many years before he published his first critique and became known as a philosopher. And the first critique was aimed precisely at the kind of pure reason divorced from experience that would bring Hegel's Absolute Spirit into disrepute in the later nineteenth century, an age when empiricism and experiment were yielding such obvious scientific benefits. In the fields of physics, mathematics, and physiology, such men as Ernst Mach and Wilhelm von Helmholtz were explicitly committed to working out the larger implications of Kant's speculative epistemology not in the philosopher's study, but in the scientist's laboratory, as they charted new paths in physics and physiology.

Dialogism's immediate philosophical antecedents are to be found in attempts made by various Neo-Kantians to overcome the gap between "matter" and "spirit." After the death of Hegel, this gap became increasingly apparent in the growing hostility between science and philosophy. Dialogism, then, is part of a major tendency in European thought to reconceptualize epistemology the better to accord with the new versions of mind and the revolutionary models of the world that began to emerge in the natural sciences in the nineteenth century. It is an attempt to frame a theory of knowledge for an age when relativity dominates physics and cosmology and thus when noncoincidence of one kind or another—of sign to its referent, of the subject to itself—raises troubling new questions about the very existence of mind.

Bakhtin begins by accepting Kant's argument that there is an unbridgeable gap between mind and world (but as we shall see, he differs from Kant in assuming that therefore there are things in themselves; there may be things outside mind, but they are nevertheless not in themselves). The non-identity of mind and world is the conceptual rock on which dialogism is founded and the source of all the other levels of non-concurring identity which Bakhtin sees shaping the world and our place in it. Bakhtin's thought is a meditation on how we know, a meditation based on dialogue precisely because, unlike many other theories of knowing, the site of knowledge it posits is never unitary. I use the admittedly cumbersome term "meditation on knowledge" here, because from his very earliest work Bakhtin is highly critical of what he calls "epistemologism," a tendency pervading all nineteenth- and early twentieth-century philosophy. A theory of knowledge devolves into mere epistemologism when there is posited "a unitary and unique consciousness … any determinateness must be derived from itself [thus it] cannot have another consciousness outside itself … any unity is its own unity."

In dialogism, the very capacity to have consciousness is based on otherness. This otherness is not merely a dialectical alienation on its way to a sublation that will endow it with a unifying identity in higher consciousness. On the contrary: in dialogism consciousness is otherness. More accurately, it is the differential relation between a center and all that is not that center. Now, a caution is in order here. Serious questions have recently been raised about the validity of any discourse that invokes the concept of center, as in various versions of what has come to be called "logocentrism." "Center" has often been used as a name for the unreflective assumption of ontological privilege, the sort of mystification sometimes attacked as the "illusion of presence." It is important from the outset, then, that "center" in Bakhtin's thought be understood for what it is: a relative rather than an absolute term, and, as such, one with no claim to absolute privilege, least of all one with transcendent ambitions.

This last point is particularly important, for certain of the terms crucial to Bakhtin's thought, such as "self" and "other," have so often been used as masked claims to privilege. Before we further specify the roles played by these protagonists in Bakhtinian scenarios, the simple yet all-important fact should be stressed again that they always enact a drama containing more than one actor.

Self and other are terms that sound vaguely atavistic in an age remarkable for its celebration of all that is extra- and impersonal. We are frequently told that not only God has died, but so has the subject. And perhaps no subject is quite so moribund as the particular kind that once was honored as author. It has even been argued with self-immolating eloquence that man (or at least Man) himself has died in history. All these deaths are melodramatic ways of formulating an end to the same thing: the old conviction that the individual subject is the seat of certainty, whether the subject so conceived was named God, the soul, the author, or—my self. Bakhtin, too, is suspicious of untrammeled subjectivity's claims; he perhaps least of all is mystified by them. And he attacks such claims at their root, in the self itself, which is why for him "self" can never be a self-sufficient construct.

It cannot be stressed enough that for him "self" is dialogic, a relation. And because it is so fundamental a relation, dialogue can help us understand how other relationships work, even (or especially) those that preoccupy the sometimes stern, sometimes playful new Stoics who most dwell on the death of the subject: relationships such as signifier/signified, text/context, system/history, rhetoric/language, and speaking/writing…. [We must recognize] that for Bakhtin the key to understanding all such artificially isolated dualisms is the dialogue between self and other.

Whatever else it is, self/other is a relation of simultaneity. No matter how conceived, simultaneity deals with ratios of same and different in space and time, which is why Bakhtin was always so concerned with space/time. Bakhtin's thought was greatly influenced by the new concepts of time and space that were being proposed by revolutionary physicists after the collapse of the old Newtonian cosmos. In Newton's mechanics it was possible for physical processes to propagate at infinite velocity through space. This meant that if one and the same action emanates from one body and reaches another body at the same instant, the process is purely spatial for it has occupied zero time. In Newton's universe, the sum of instants occurring simultaneously over all of space add up to a time that is absolute in the sense that it is a flux of simultaneous instants embracing the whole of the universe. It was, in other words, a dream of unity in physics that could serve as the proper setting for a dream of unity in Newton's theology, and which could later underwrite in philosophy the absolute oneness of consciousness in Hegelian dialectic. Dialogue, by contrast, knows no sublation. Bakhtin insists on differences that cannot be overcome: separateness and simultaneity are basic conditions of existence. Thus the physics proper to such a universe are post-Newtonian. Bakhtin grew up amidst battles that raged over the concepts of space and time among such "empiriocritics" as Mach, his Russian followers (primarily Bogdanov) and his Russian opponents (such as Lenin). Of these scientists and philosophers, the most helpful in grasping Bakhtin's thought is Einstein. Although there can be no question of immediate influence, dialogism is a version of relativity.

Einstein invented a number of just-so stories, or "thought experiments," as a way to elide physical limits on experimentation. Although not directly related, these experiments in some ways correspond to Bakhtin's attempts to use the situation of dialogue as a means for getting around traditional limitations of ideas of the subject. Both resort to what might be called a "philosophical optics," a conceptual means for seeing processes invisible to any other lens. More particularly, both resort to experiments with seeing in order to meditate on the necessity of the other. Einstein invented several situations (typically involving people looking at moving objects such as trains) that involve problems in perception raised by the speed of light. For instance, if light travels at a certain velocity in one system and at the same velocity in another system moving without acceleration relative to the first, it is impossible to detect the first system's movement by optical means, no matter how refined: the observer's ability to see motion depends on one body changing its position vis-à-vis other bodies. Motion, we have come to accept, has only a relative meaning. Stated differently, one body's motion has meaning only in relation to another body; or—since it is a relation that is mutual—has meaning only in dialogue with another body.

Dialogism argues that all meaning is relative in the sense that it comes about only as a result of the relation between two bodies occupying simultaneous but different space, where bodies may be thought of as ranging from the immediacy of our physical bodies, to political bodies and to bodies of ideas in general (ideologies). In Bakhtin's thought experiments, as in Einstein's, the position of the observer is fundamental. If motion is to have meaning, not only must there be two different bodies in a relation with each other, but there must as well be someone to grasp the nature of such a relation: the non-centeredness of the bodies themselves requires the center constituted by an observer. But unlike the passive stick figures who are positioned at a point equidistant between two railway trains in the cartoons often used to illustrate Einsteinian motion, Bakhtin's observer is also, simultaneously, an active participant in the relation of simultaneity. Conceiving being dialogically means that reality is always experienced, not just perceived, and further that it is experienced from a particular position. Bakhtin conceives that position in kinetic terms as a situation, an event, the event of being a self.

The self, moreover, is an event with a structure. Perhaps predictably for so attentive a student of Kant and post-Newtonian mechanics as Bakhtin, that structure is organized around the categories of space and time. They articulate what has been called the "law of placement" in dialogism, which says everything is perceived from a unique position in existence; its corollary is that the meaning of whatever is observed is shaped by the place from which it is perceived. Bakhtin explicates this law with a just-so story that uses seeing as a means for grasping what is essentially a non-visual situation. He begins with a simple datum from experience; not an observer looking at trains, but an observer looking at another observer. You can see things behind my back that I cannot see, and I can see things behind your back that are denied to your vision. We are both doing essentially the same thing, but from different places: although we are in the same event, that event is different for each of us. Our places are different not only because our bodies occupy different positions in exterior, physical space, but also because we regard the world and each other from different centers in cognitive time/space.

What is cognitive time/space? It is the arena in which all perception unfolds. Dialogism, like relativity, takes it for granted that nothing can be perceived except against the perspective of something else: dialogism's master assumption is that there is no figure without a ground. The mind is structured so that the world is always perceived according to this contrast. More specifically, what sets a figure off from its dialogizing background is the opposition between a time and a space that one consciousness uses to model its own limits (the I-for-myself) and the quite different temporal and spatial categories employed by the same consciousness to model the limits of other persons and things (the not-I-in-me)—and (this is crucial) vice versa.

At a very basic level, then, dialogism is the name not just for a dualism, but for a necessary multiplicity in human perception. This multiplicity manifests itself as a series of distinctions between categories appropriate to the perceiver on the one hand and categories appropriate to whatever is being perceived on the other. This way of conceiving things is not, as it might first appear to be, one more binarism, for in addition to these poles dialogism enlists the additional factors of situation and relation that make any specific instance of them more than a mere opposition of categories.

For the perceivers, their own time is forever open and unfinished; their own space is always the center of perception, the point around which things arrange themselves as a horizon whose meaning is determined by wherever they have their place in it. By contrast, the time in which we model others is perceived as closed and finished. Moreover, the space in which others are seen is never a significance-charged surrounding, but a neutral environment, i.e. the homogenizing context of the rest of the world. From the perspective of a self, the other is simply in the world, along with everyone and everything else. The contrast between spatial and temporal categories that are appropriate to me and the very different categories I employ to give shape to the other must not be misinterpreted as yet another Romantic claim for primacy of the absolute subject: self for Bakhtin is a cognitive necessity, not a mystified privilege.

We will see this—and the intimate relation dialogism bears to language—if we understand that cognitive time/space is ordered very much as time and space categories are deployed in speech. It has long been recognized that the formal means for expressing subjectivity occupy a unique place in any language. "I" is a word that has no referent in the way "tree," for instance, nominates a class of flora; if "I" is to perform its task as a pro-noun, must not be a noun, i. e. it must not refer to anything as other words do. For its task is to indicate the person uttering the present instance of the discourse containing "I," a person who is always changing and different. "I" must not refer to anything in particular if it is to be able to mean everybody in general. In Jakobson's suggestive phrase, "I" is a "shifter" because it moves the center of discourse from one speaking subject to another: its emptiness is the no man's land in which subjects can exchange the lease they hold on all of language by virtue of saying "I." When a particular person utters that word, he or she fills "I" with meaning by providing the central point needed to calibrate all further time and space discriminations: "I" is the invisible ground of all other indices in language, the benchmark to which all its spatial operations are referred, and the Greenwich mean by which all its time distinctions are calibrated. "I" marks the point between "now" and "then," as well as between "here" and "there." The difference between all these markers is manifested by the relation each of them bears either to the proximity of the speaker's horizon (here and now), or to the distance of the other's environment (there and then). As the linguist Émile Benveniste has remarked [in Problems in General Linguistics, trans. Mary Elizabeth Mary, 1971], "Language itself reveals the profound difference between these two planes." The gate of the "I" is located at the center not only of one's own existence, but of language as well.

This is so because there is an intimate connection between the project of language and the project of selfhood: they both exist in order to mean. The word Bakhtin uses for "project," (zadanie) is another twist on the central distinction between something that is "given" (dan) and something that presents itself in the nature of a task, as something that must be "conceived" (zadan.) The situatedness of the self is a multiple phenomenon: it has been given the task of not being merely given. It must stand out in existence because it is dominated by a "drive to meaning," where meaning is understood as something still in the process of creation, something still bending toward the future as opposed to that which is already completed.

It should be added in passing that brute chronological indicators are no guarantee of whether a thing has meaning in this sense or not, for events initiated in the most distant past, as measured by the clock, may still be fresh and unfinished in cognitive time/space. Dialogism's drive to meaning should not be confused with the Hegelian impulse toward a single state of higher consciousness in the future. In Bakhtin there is no one meaning being striven for: the world is a vast congeries of contesting meanings, a heteroglossia so varied that no single term capable of unifying its diversifying energies is possible.

Since Bakhtin sees the world as activity, it will come as no surprise that he defines existence as an event. But it will perhaps seem contradictory that his term for existence is "the unique and unified event of being" (edinstvennoe i edinoe sobytie bytija), a phrase that recurs with obsessive regularity in Bakhtin's early work, and a formulation so important for understanding Bakhtin that each word requires some glossing.

The activity of the world comes to each of us as a series of events that uniquely occur in the site I, and only I, occupy in the world. If I slash my finger with a knife, an "other" may be intellectually aware that I am in pain, and may even deeply empathize with me. But the pain itself happens to me; it is addressed to where "I" am, not to the other (pre-positions, like pro-nouns, grammatically instance the unique placedness of subjects). One way in which the uniqueness of my place in life may be judged is by the uniqueness of the death that will be mine. However, this uniqueness—in what only appears to be a paradox—is shared. We shall all die, but you cannot die in my place, any more than you can live fromthat site. And of course the reverse is also true: I cannot be in the unique place you occupy in the event of existence.

Nevertheless, the event of existence is "unified"; for although it occurs in sites that are unique, those sites are never complete in themselves. They are never in any sense of the word alone. They need others to provide the stability demanded by the structure of perception if what occurs is to have meaning. In order that the event of existence be more than a random happening, it must have meaning, and to do that it must be perceptible as a stable figureagainst the ground of the flux and indeterminacy of everything else. This unification occurs as the result of an event, the action of me fulfilling my task (zadanie), i. e. by making the slice of existence that is merely given (dan) to me something that is conceived (zadan). I perform this transformation by imposing time/space categories appropriate to the other on what is happening. Remember that those categories differ from self-categories precisely in their ability to consummate, to finish off, what is being perceived, to complete it in time and to assign it a space.

The word "event" as it occurs in the formulation above is particularly complex. The Russian word used, sobytie, is the normal word Russians would use in most contexts to mean what we call in English an "event." But as Bakhtin uses it, certain aspects of the word long-forgotten in its everyday usage, are brought to the fore. The most important of these emerges from the fact that in Bakhtin's philosophical writings the word is almost never used alone, but always in conjunction with the word "being." He insists on being as an event.

The obligatory grouping of these two words in this way is a syntactic doubling that points to the mutuality of their meaning. It points as well to the etymological relations of the two words. In Russian, "event" is a word having both a root and a stem; it is formed from the word for being-bytie-with the addition of the prefix implying sharedness, "so-, co-," (or, as we should say in English, "co-" as in cooperate or co-habit), giving sobytie, event as co-being. "Being" for Bakhtin then is, not just an event, but an event that is shared. Being is a simultaneity; it is always co-being.

[Sergi] Karcevskij, too, meditates on simultaneity [in "The asymmetric dualism of the linguistic Sign," in The Prague School: Selected Writings 1929–1946, ed. Peter Steiner, 1982]: "the simultaneous presence of these two possibilities is indispensable for any act of comprehension." Like Bakhtin—and in marked contrast to the French reading of the asymmetry of the sign that finds its most radical extreme in Derrida's differance—Karcevskij recognized that "opposition pure and simple necessarily leads to chaos and cannot serve as the basis of a system. True differentiation presupposes a simultaneous resemblance and difference." In other words, it presupposes a center and a non-center.

What Karcevskij is saying about language is essentially what Bakhtin is saying about reality as such: the self (the perceiver) and the other (the perceived) exist not as separate entities, but as "relations between two coordinates … each serving to differentiate the other." The coordinates proposed by Bakhtin for modeling this simultaneity are the two sets of time/space categories inherent in each of its poles: self and other (Bakhtin speaks of them as two interacting legal codes). The interaction of the binaries resemblance/difference, and figure/ground, both have at their heart the master distinction of self/other. In cognition, even more than in the physical world, two bodies cannot occupy the same space at the same time. As subject, I must not share the time/space of an object. Using self and other as basic categories does not obliterate the split between subject and object, but it complicates that distinction in ways that make it productive.

The other is in the realm of completedness, whereas I experience time as open and always as yet un-completed, and I am always at the center of space. This condition has certain virtues; in a world filled with the determining energies of impersonal social force, it is a potential source of freedom, the ground of other liberties from constraint of the sort Bakhtin celebrates in carnival…. In common with everything else, however, this openness exists in tension with its dialogic partner, closure. The unfinished nature of self is not mere subjective license: like any border, it is also a limit. The very immediacy which defines my being as a self is the same condition that insures I cannot perceive my self: one way to grasp how far removed the self is from any privilege is to be aware that like anything else, its perception requires temporal categories that are less fluid and spatial categories that are more comprehensive than are provided by the manner in which my "I" is fated to live the event of being. For all their comparative openness, indeed because of it, self-categories cannot do what categories of the other can. Seeing requires a certain outsideness to what is seen, a certain stasis. "In the realm of culture, outsideness is the most powerful factor in understanding," precisely because it permits the finalized quality needed for the whole of a culture to be seen…. But as the primal activity that marks being as an ongoing event, the self "itself" cannot abide even the most minimal degree of fixity.

When I look at you, I see your whole body, and I see it as having a definite place in the total configuration of a whole landscape. I see you as occupying a certain position vis-à-vis other persons and objects in the landscape (you are one other among many others). Moreover, you not only have definite physical characteristics, specific social standing, and so on, but I see you as having a definite character as well. I imagine you as being good or bad at your trade, a good or bad husband, wife, parent, as being more or less close to dying, and a number of other things that sum you up as a (more or less definitely) consummated whole. If we imagine self and other in painterly terms, the former would be non-figurative and the latter extremely hard-edged. And yet I must have some way of forming myself into a subject having something like the particularity of the other. My "I" must have contours that are specific enough to provide a meaningful addressee: for if existence is shared, it will manifest itself as the condition of being addressed (obrashchënnost', or addressivnost'). Existence is not only an event, it is an utterance. The event of existence has the nature of dialogue in this sense; there is no word directed to no one.

It is here we approach the ineluctable association of dialogism and authorship. In order to see this connection, let us go back for a moment to the peculiarity of the first person pronoun. Remember other nouns are signs in so far as their material sound, such as the locution "tree" when we actually pronounce it, evokes the fixed notion of a particular sort of object (some kind of natural growth, let us say). In the signifier "tree" we see a signified tree. Most nouns work something like this, but not the pronoun for the self, for what "I" refers to cannot be seen, at least in the same way that the word "tree" enables us to see a tree.

In order for my specific subjectivity to fill the general slot of the first person pronoun, that word must be empty: "I" is a word that can mean nothing in general, for the reference it names can never be visualized in its consummated wholeness. But this invisibility (which, as we shall see, is akin to the invisibility of the unconscious) is not mysterious. It is a general token of absence that can be filled in any particular utterance. It is invisible only at the level of system. At the level of performance, in the event of an utterance, the meaning of "I" can always be seen. It can be said, then, that the pronoun "I" marks the point of articulation between the pre-existing, repeatable system of language and my unique, unrepeatable existence as a particular person in a specific social and historical situation.

Existence, like language, is a shared event. It is always a border incident on the gradient both joining and separating the immediate reality of my own living particularity (a uniqueness that presents itself as only for me) with the reality of the system that precedes me in existence (that is always-already-there) and which is intertwined with everyone and everything else. Through the medium of the first person pronoun each speaker appropriates a whole language to himself. Much as Peter Pan's shadow is sewn to his body, "I" is the needle that stitches the abstraction of language to the particularity of lived experience. And much the same structure insures that in all aspects of life dialogue can take place between the chaotic and particular centrifugal forces of subjectivity and the rule-driven, generalizing centripetal forces of extra-personal system.

The single world "I" is exploited in language very much as the single eye of the fates is used in Greek mythology. The three old women all pass around the same organ. If they did not share their eye they could not see. In order to have her own vision, each must use the means by which the others see. In dialogism this sharedness is indeed the nature of fate for us all. For in order to see our selves, we must appropriate the vision of others. Restated in its crudest version, the Bakhtinian just-so story of subjectivity is the tale of how I get my self from the other: it is only the other's categories that will let me be an object for my own perception. I see my self as I conceive others might see it. In order to forge a self, I must do so from outside. In other words, I author myself.

Even in this brutalized rendition it will be apparent that things cannot be so simple, and in the event (of being) they are not. First, because the act of creating a self is not free: we must, we all must, create ourselves, for the self is not given (dan) to any one of us. Or, as Bakhtin puts it, "we have no alibi in existence." This lack of choice extends to the materials available for creation, for they are always provided by the other. I cannot choose to model my self as, let us say, a Martian might see me if I have not had experience of Martians. I may, of course, imagine what Martians might be like, and then seek to appropriate their image of me as my own. But even an imaginary Martian will be made up of details provided from previous experience, for in existence that is shared, there can be nothing absolute, including nothing absolutely new.

The self, then, may be conceived as a multiple phenomenon of essentially three elements (it is—at least—a triad, not a duality): a center, a not-center, and the relation between them. Until now we have been discussing the first two elements, the center (or I-for-itself) and the not-center (the-not-I-in-me) in terms of the time/space categories appropriate to each. In taking up the third item, the relation that center and not-center bear to each other, we will have to keep in mind one or two new terms that are crucial to Bakhtin's undertaking. Dialogism is a form of architectonics, the general science of ordering parts into a whole. In other words, architectonics is the science of relations. A relation is something that always entails ratio and proportion. In addition, Bakhtin emphasizes that a relation is never static, but always in the process of being made or unmade.

In so far as a relation involves the construction of ratios, it is aesthetic in much the same way that a statue or a building may be judged in terms of how its parts have been constructed with respect to each other. Relation, it will be helpful to remember, is also a telling, a narrative, an aspect of the word's meaning that Bakhtin will not ignore as he takes the somewhat unusual step of treating the relation of the self to the other as a problem in aesthetics.

By choosing aesthetic categories to discuss questions in epistemology, Bakhtin is drawing attention to the importance in dialogism of authoring. Sharing existence as an event means among other things that we are—we cannot choose not to be—in dialogue, not only with other human beings, but also with the natural and cultural configurations we lump together as "the world." The world addresses us and we are alive and human to the degree that we are answerable, i. e. to the degree that we can respond to addressivity. We are responsible in the sense that we are compelled to respond, we cannot choose but give the world an answer. Each one of us occupies a place in existence that is uniquely ours; but far from being a privilege, far from having what Bakhtin calls an alibi in existence, the uniqueness of the place I occupy in existence is, in the deepest sense of the word, an answerability: in that place only am I addressed by the world, since only I am in it. Moreover, we must keep on forming responses as long as we are alive.

I am always answerable for the response that is generated from the unique place I occupy in existence. My responses begin to have a pattern; the dialogue I have with existence begins to assume the form of a text, a kind of book. A book, moreover, that belongs to a genre. In antiquity, too, the world was often conceived as a book, the text of libri naturae. Bakhtin conceives existence as the kind of book we call a novel, or more accurately as many novels (the radically manifold world proposed by Bakhtin looks much like Borges' Library of Babel), for all of us write our own such text, a text that is then called our life. Bakhtin uses the literary genre of the novel as an allegory for representing existence as the condition of authoring.

The author of a novel may unfold several different plots, but each will be merely one version of a more encompassing story: the narrative of how an author (as a dialogic, non-psychological self) constructs a relation with his heroes (as others). Authors are somehow both inside and outside their work. In literary texts, interaction between author and heroes is what constructs the relation that gives deepest coherence to the other meanings of relation, not least relation understood as a telling.

The particular corner (really an angle of refraction) in apperception where such authoring can take place—the self's workshop, as it were—Bakhtin calls unenakhodimost', or "outsideness" (sometimes rendered into English—from French rather than from Russian—as "exotopy"). The term, as always in dialogism, is not only spatial, but temporal: it is only from a position outside something that it can be perceived in categories that complete it in time and fix it in space. In order to be perceived as a whole, as something finished, a person or object must be shaped in the time/space categories of the other, and that is possible only when the person or object is perceived from the position of outsideness. An event cannot be wholly known, cannot be seen, from inside its own unfolding as an event. As Bergson, an important source of ideas for Bakhtin, puts it: "in so far as my body is the center of action [or what Bakhtin calls a deed], it cannot give birth to a representation." [Henri Bergson, Matters and Memory, trans. Nancy Margaret Paul and W. Scott Palmer, 1911].

In a dialogue that takes place between two different persons (one self/other constellation to another self/other constellation) in physical space, the medium of exchange is, of course, natural language. In such exchanges it is words that fix (if only very fleetingly) meanings. They can do so because syntax, grammar, and the sound laws governing phonology provide a relatively stable armature for marking distinctions in the unstable flux of life outside language. Words can segment experience into meaningful patterns because their essence is so radically differential: they exist only to register differences. As Saussure, summing up his argument at a crucial point [in Course in General Linguistics, trans. Wade Baskin, 1966], says: "Everything that has been said up to this point boils down to this: in language there are only differences." Bakhtin insists that language is also a matter of sameness, but he would certainly agree that "language is only a system of pure values."

And so, argues Bakhtin, is the self. Once again quoting Saussure to gloss Bakhtin, we may say that for the units of existence we call "selves," as for the units of language we call "words," "Their most precise characteristic is in being what the others are not." While the self/other distinction does not operate as a complete algorithm of natural language, it does share with language the three fundamental features of function, means, and purpose. The function of each is to provide a mechanism for differentiating; each uses values to distinguish particular differences, and the purpose of doing so in each case is to give order to (what otherwise would be) the chaos of lived experience.

Let us return, then, to the example of two people regarding each other, each attempting to make sense out of the existence each shares with the other. We may now describe the dialogue of radical self/other distinctions unfolding within their cognitive space in linguistic terms. This dialogue takes place much as dialogues in natural language do: by using particular values to specify otherwise unmarked differences.

In our imagined encounter, the first person will see the second, the "other," through the relation of difference that divides all phenomena either into categories of self or categories of the other, never both. But once the distinction is made that defines the second person as one who must be perceived through the lens of the other, distinctions of a secondary order will follow that fill in the other's general outline with shades of particular differences (the progression from primary to secondary differentiation is, of course, logical, not chronological). These shadings will be made with colors drawn from the palette of specific values that obtain in the event of existence as it manifests itself in a particular time and a particular place. The first person will see the second, then, in terms much too detailed to be frozen into the sort of abstract account that a hapless expositor is condemned to provide as an example. But the terms could reasonably be expected to include such things as how physical appearance is judged (is long hair, or curly hair, or blonde hair, or perhaps no hair a good or bad thing? is being round in the tummy to be "portly" or is it to be "fat"?), also manner of speaking ("common," "stilted," "natural"), politics, relation to the major theory dominating a particular discipline at the moment, and so on. The other is always perceived in terms that are specified socially and historically, and for all the abstraction of our discussion so far, dialogism's primary thrust is always in the direction of historical and social specificity.

The only perspective from which values of such specificity and completeness may be brought to bear on the other is from the position of "outsideness." The first person succeeds in attaining the position needed to perceive the second from outside. But will he or she be able to achieve that extreme degree of outsideness toward the second which Bakhtin calls "transgredience"? Transgredience (transgradientsvo) is reached when the whole existence of others is seen from outside not only their own knowledge that they are being perceived by somebody else, but from beyond their awareness that such an other even exists. It is a cardinal assumption of dialogism that every human subject is not only highly conscious, but that his or her cognitive space is coordinated by the same I/other distinctions that organize my own: there is in fact no way "I" can be completely transgredient to another living subject, nor can he or she be completely transgredient to me.

We touch here on two other important concerns of dialogism: authority as authorship; and authority as power. Transgredience is a topic that bears on the specificity of art within a general aesthetic; and it also bears on the question of power in the state. As we shall see, transgredience, when it is used well, results in art; when used badly, it results in totalitarianism.

Although, then, dialogism is primarily an epistemology, it is not just a theory of knowledge. Rather, it is in its essence a hybrid: dialogism exploits the nature of language as a modeling system for the nature of existence, and thus is deeply involved with linguistics; dialogism sees social and ethical values as the means by which the fundamental I/other split articulates itself in specific situations and is thus a version of axiology; and in so far as the act of perception is understood as the patterning of a relation, it is a general aesthetic, or it is an architectonics, a science of building.

Use of the term architectonics betrays once again Bakhtin's debt to Kant, who used it not only in its technical sense (as a way to refer to any systematization of knowledge), but to emphasize the active, constructive role of mind in perception. By using the same word, Bakhtin also seeks to foreground these aspects; but in addition he wants to draw a line between the kind of authoring we all must do all the time, and the kind of authoring some persons do some of the time, the results of which we then call art. Architectonics involves us all; but the branch of architectonics involving artists is aesthetics proper.

What is the difference between the two? It is the ability of the artist in his or her text to treat other human subjects from the vantage point of transgredience, a privilege denied the rest of us who author only in lived experience (and denied to artists too, when they are not being artists). The author of a novel, for instance, can manipulate the other not only as an other, but as a self. This is, in fact, what the very greatest writers have always done, but the paradigmatic example is provided by Dostoevsky, who so successfully permits his characters to have the status of an "I" standing over against the claims of his own authorial other that Bakhtin felt compelled to coin the special term "polyphony" to describe it. Lesser authors treat their heroes as mere others, a relation that can be crafted in architectonics, and which does not therefore require the aesthetic privilege of art for its achievement: it is what we all do anyway. And then there are those authors who treat their characters not only as others, but as having the otherness of mere things, lacking any subjectivity. They exploit their transgredience of their characters much as scientists exploit theirs toward laboratory rats. This is formulaic pseudo-art, in which all possible initiative within the text is sacrificed to a formula pre-existing the text. If in western movies of a certain kind the "hero" ends up kissing a horse instead of a girl, or if in Stalinist fiction the boy always gets a tractor instead of a girl, we feel no violation, because we understand that neither the cowboy nor the collective farmer has any reserve of subjectivity: they are, themselves, effectively, only horses or tractors anyway.

This formulaic art makes explicit the connection of transgredience to power. For not only is snuffing out the "I" of other subjects bad aesthetics, it is bad politics. Dialogically conceived, authorship is a form of governance, for both are implicated in the architectonics of responsibility, each is a way to adjudicate center/non-center relations between subjects. Totalitarian government always seeks the (utopian) condition of absolute monologue: the Gleichschaltung which was attempted in Germany during the 1930s to "Nazify" trade unions, universities, publishing houses, professional associations, and so on had as its aim the suppression of all otherness in the state so that its creator alone might flourish. Dialogism has rightly been perceived by certain thinkers on the left as a useful correlative to Marxism, for it argues that sharing is not only an ethical or economic mandate, but a condition built into the structure of human perception, and thus a condition inherent in the very fact of being human. But by the same token dialogism differs from the pseudo-Marxism of regimes that use "Communism" as a license for totalitarian government. For as the ultimate critique of any claim to monologue, it is intransigently pluralist.

We have looked at several versions of self/other relations. In so doing certain fundamentals have emerged, not least of which is that dialogism is able to make claims in many different areas because it is basically a theory of knowledge, an architectonics of perception. Dialogism argues that we make sense of existence by defining our specific place in it, an operation performed in cognitive time and space, the basic categories of perception. Important as these categories are, they themselves are shaped by the even more fundamental set of self and other. We perceive the world through the time/space of the self and through the time/space of the other. The difference between the two is a relation of otherness that can be gauged by differing positions of outsideness that are enacted as varying degrees of transgredience. Up until this point we have discussed such relations almost exclusively in terms of the other. We must now address the difficult question of how the self achieves the outsideness it needs to perceive itself.

So as always to be an open site where the event of existence can have its occurrence, the self must never stop in time or be fixed in space. Since, however, being finished in time and being specifically located in space are conditions necessary for being "seen" in perception, the self is by definition invisible to itself. In the wake of a still-potent Romanticism, it is necessary to repeat that there is nothing mysterious about this invisibility, for it is merely structural. The self's non-referentiality can be understood by analogy with the non-referentiality of "I" as the first person pronoun in natural language. If each is to perform its function of indicating a unique place that must be shared by everybody (which is what the self marks in existence, and what the "I" marks in language), then they must both refer to nothing—or at least not refer to anything in the same manner other signs refer.

But the self is like a sign in so far as it has no absolute meaning in itself: it, too (or rather, it most of all), is relative, dependent for its existence on the other. A conventional sign is not a unitary thing, but rather a differential relation between two aspects, a signifier and a signified. In this triad it is the relation that is absolute, not the elements it yokes together, for neither of the two elements exists in itself; neither has any meaning on its own, without the simultaneous presence of the other. Nor is the "self" a unitary thing; rather, it consists in a relation, the relation between self and other. A traditional metaphor representing the unity of the linguistic sign's two elements is the unity shared by the recto and verso sides of the same sheet of paper. But in so far as the self is an activity, such a static means of conceiving it will not do: Bakhtin's metaphor for the unity of the two elements constituting the relation of self and other is dialogue, the simultaneous unity of differences in the event of utterance.

One of Bakhtin's simple illustrations will help us overcome the complexity of this last formulation. If we return for a moment to the situation of two people facing each other, we remember that although they share an external space and time (they are physically simultaneous), inside his or her own head each sees something the other does not. Let us envisage you and me confronting each other. There are certain things we both perceive, such as the table between us. But there are other things in the same encounter we do not both perceive. The simplest way to state the difference between us is to say that you see things about me (such as, at the most elementary level, my forehead) and the world (such as the wall behind my back) which are out of my sight. The fact that I cannot see such things does not mean they do not exist; we are so arranged that I simply cannot see them. But it is equally the case that I see things you are unable to see, such as your forehead, and the wall behind your back. In addition to the things we see jointly, there are aspects of our situation each of us can see only on our own, i. e. only from the unique place each of us occupies in the situation.

The aspect of the situation that you see, but I do not, is what Bakhtin calls your "surplus of seeing"; those things I see but you cannot constitute my "surplus of seeing." You know I have a surplus, and I know you have one as well. By adding the surplus that has been "given" to you to the surplus that has been "given" to me I can build up an image that includes the whole of me and the room, including those things I cannot physically see: in other words, I am able to "conceive" or construct a whole out of the different situations we are in together. I author a unified version of the event of our joint existence from my unique place in it by means of combining the things I see which are different from (in addition to) those you see, and the things you see which are different from (in addition to) that difference.

Such acts of combination are a rudimentary form of "narrativity," or the ability to put myself into scenarios of the kind I see others enacting. I never see others as frozen in the immediacy of the isolated present moment. The present is not a static moment, but a mass of different combinations of past and present relations. To say I perceive them as a whole means that I see them surrounded by their whole lives, within the context of a complete narrative having a beginning that precedes our encounter and an end that follows it. I see others as bathed in the light of their whole biography.

My "I-for-itself" lacks such a consummated biography: because the self's own time is constantly open, it resists such framing limits. Within my own consciousness my "I" has no beginning and no end. The only way I know of my birth is through accounts I have of it from others; and I shall never know my death, because my "self" will be alive only so long as I have consciousness—what is called "my" death will not be known by me, but once again only by others. In order to remain a constantly potential site of being, my self must be able to conduct its work as sheer capability, a flux of sheer becoming. If this energy is to be given specific contours, it must be shaped not only in values, but in story. Stories are the means by which values are made coherent in particular situations. And this narrativity, this possibility of conceiving my beginning and end as a whole life, is always enacted in the time/space of the other: I may see my death, but not in the category of my "I." For my "I," death occurs only for others, even when the death in question is my own.

Since Bakhtin places so much emphasis on otherness, and on otherness defined precisely as other values, community plays an enormous role in his thought. Dialogism is, among other things, an exercise in social theory. Although frequently overlooked by those who tear "carnival" out of its larger Bakhtinian context, extrapersonal social force is accorded so much weight in dialogism that it almost (but not quite) begins to verge on determinism. If my "I" is so ineluctably a product of the particular values dominating my community at the particular point in its history when I coexist with it, the question must arise, "Where is there any space, and what would the time be like, in which I might define myself against an otherness that is other from that which has been 'given' to me?"

In answering this question it will be helpful to remember that dialogue is not, as is sometimes thought, a dyadic, much less a binary, phenomenon. Dialogue is a manifold phenomenon, but for schematic purposes it can be reduced to a minimum of three elements having a structure very much like the triadic construction of the linguistic sign: a dialogue is composed of an utterance, a reply, and a relation between the two. It is the relation that is most important of the three, for without it the other two would have no meaning. They would be isolated, and the most primary of Bakhtinian a prioris is that nothing is anything in itself.

The tripartite nature of dialogue bears within it the seeds of hope: in so far as my "I" is dialogic, it insures that my existence is not a lonely event but part of a larger whole. The thirdness of dialogue frees my existence from the very circumscribed meaning it has in the limited configuration of self/other relations available in the immediate time and particular place of my life. For in later times, and in other places, there will always be other configurations of such relations, and in conjunction with that other, my self will be differently understood. This degree of thirdness outside the present event insures the possibility of whatever transgredience I can achieve toward myself.

At the heart of any dialogue is the conviction that what is exchanged has meaning. Poets who feel misunderstood in their lifetimes, martyrs for lost political causes, quite ordinary people caught in lives of quiet desperation—all have been correct to hope that outside the tyranny of the present there is a possible addressee who will understand them. This version of the significant other, this "super-addressee," is conceived in different ways at different times and by different persons: as God, as the future triumph of my version of the state, as a future reader.

As the need to posit a category such as "super-addressee" outside the present moment makes clear, conditions for creating meaning in the present moment are not always the best. A dialogic world is one in which I can never have my own way completely, and therefore I find myself plunged into constant interaction with others—and with myself. In sum, dialogism is based on the primacy of the social, and the assumption that all meaning is achieved by struggle. It is thus a stern philosophy. This fact should surprise no one, given dialogism's immediate sources in revolution, civil war, the terror of the purges, and exile. But the very otherness that makes it at times a version of Stoicism is also what insures that we are not alone. Dialogism is ultimately an epistemology founded on a loophole, for

there is neither a first word nor a last word. The contexts of dialogue are without limit. They extend into the deepest past and the most distant future. Even meanings born in dialogues of the remotest past will never be finally grasped once and for all, for they will always be renewed in later dialogue. At any present moment of the dialogue there are great masses of forgotten meanings, but these will be recalled again at a given moment in the dialogue's later course when it will be given new life. For nothing is absolutely dead: every meaning will someday have its homecoming festival.

Michael Holquist, in his Dialogism: Bakhtin and His World, Routledge, 1990, 204 p.