The Second Person: A Point of View?
The Function of the Second-Person Pronoun in Narrative Prose Fiction.

Part 1. The Prejudice of Person

1. Dreams of Geometry and Voice

In proposing in the previous chapter that Cartesianism runs to the heart of prevailing practices of reading and theories of narrative, I singled out two particular habits of reading that one way or another will remain an implicit focus throughout this thesis. To reiterate remarks by Uri Margolin, prevailing pre-critical and critical reading practices have long taken it for granted, firstly, that individual subjects, as narrators and as participants, can be recovered from the linguistic signs of the text, and secondly, that these individuals can be described in terms of identity and location in space and time in the text-world (Margolin, 1986-87: 183). This is not merely to say what critics such as Forster and Fludernik have said all along that narrative, especially narrative prose fiction, is about people. The point is more crucial: that we construe the forms of the people we find in the linguistic signs of prose narrative texts in terms of particular modes of subjectivity.

In this chapter, I want to deal quite specifically with this matter and provide a necessary background, as a theoretical and philosophical context, for the thesis as a whole. I will develop my critique of Cartesianism and its conspicuously self-ish notion of the human subject further, drawing in particular from discussions by Andrew Gibson and Uri Margolin. I will then outline the quite different notion of the human subject that informs this thesis, and that constitutes a principal assumption within my later arguments about the striking effects of Protean narrative-"you." I would suggest that the truth of Margolin's remarks regarding criticism's faith in the recoverability of subjects from literary textuality is nowhere more evident than in our investment, on the one hand, in notions of voice, and on the other, notions of Euclidean space. We might be said to experience both space and voice at the most fundamental levels of day-to-day life. It might also be said that our experience of both is fundamentally illusory, certainly in so far as Cartesianism affects the sense we make of such experience.

Andrew Gibson takes up this issue in Towards a Postmodern Theory of Narrative (1996). Quoting Nietzsche directly, he asserts that,

"There exists neither 'spirit,' nor reason, nor thinking, nor consciousness, nor soul, nor will, nor truth: all are fictions." In [Nietzsche's] The Will to Power, substance, attribute, unity, knowledge, causality, object, subject, space, time are all laid open to the same sceptical interrogation as aspects of "our metaphysical dogmatism." [ . . . ] They are the product of "the inventive force" that has "laboured in the service of our needs," and their "value for life is decisive." (Gibson, 1996: 34) 34

Like Cartesianism's notion of the self, Euclidean space, too, "is a conditional 'truth' . . . a mere idiosyncrasy of a certain kind of animal" (cited by Gibson, 1996: 39). 35 The problem of "the space of representation," it seems, has long been a topic for thought about narrative. Understood in terms of the linguistic reproduction of "the space of the real, the homogeneous space of the world" through which the characters move, the argument concerns issues of representation and mimesis, and involves the reader in imagining a particular kind of space, a spatiality that is homogeneous, linear, constant, usually three-dimensional, and so on. But thought about narrative, particularly within critical theory, has also traditionally concerned itself with another kind of space. Gibson describes the connection between the two kinds of space as profoundly ideological. Besides the space of the real, there is the space of the model or describable form (Gibson, 1996: 3). "In this second dimension," he writes, "the narratological imaginary has been haunted by something like the reverse of poetic intuition, by dreams of the geometric" (Gibson, 1996: 3). Throughout narratology, from Propp's notion of "spheres of action," Iser's "Gestalten," Greimas's "semiotic square," Eco's "intertextual frames" to Chatman, Stanzel and even Barthes, writes Gibson, "[t]he geometry of the text and its intelligibility become inseparable" (Gibson, 1996: 4). One need only look back over some of the models illustrated in the previous chapter and forward at the following chapter to see manifestations of this dream of the geometric. The model that Gibson singles out is Chatman's now classical model of the implied reader (see Chatman's Story and Discourse, 1978: 151. Chatman's "Diagram of Narrative Structure" on page 267 of Story and Discourse provides another, much more elaborate, brachiate illustration of Gibson's point).

Fig. 7. Chatman's diagram of the "narrative-communication situation"

It is a model which, as Chambers remarks, can "leave one wondering exactly what is the point of such distinctions between subjects that are in effect versions of one another. . . . It is not really surprising that Chatman immediately begins telescoping them in his category of the 'nonnarrated story,' which is one that only has an 'implied author'" (Chambers, 1989a: 28). One need only consider how the model reproduces itself around the central arrow in a receding nest of story levels to glimpse its schematic appeal to structuralist and formalist literary criticism. John Barth's parodic "orgy of punctuation" in the story "Menelaid" illustrates this precisely. Menelaus, as an extra-homodiegetic, "disembodied" narrator (referring to himself in the third person), recounts a conversation between himself and Helen of Troy in which he had asked why she would choose him from amongst all her suitors to be her husband. But his question and Helen's answer has been recounted several times already, each instance bound into a chain of acts and consequences.

" ' " ' " ' "Speak," Menelaus cried to Helen on the bridal bed,' I reminded Helen in her Trojan bedroom," I confessed to Eidothea on the beach,' I declared to Proteus in the cave mouth," I vouchsafed to Helen on the ship,' I told Peisistratus at least in my Spartan hall," I say to whoever and where I am. And Helen answered:

" ' " ' " ' "Love!" ' " ' " ' " (Barth, cited in Chatman, 1978: 256)

Ironically enough, perhaps, the passage is one of Chatman's own principal illustrations of the nesting of narrative levels, Chatman assiduously identifying each level as it passes, tracing each instance of the utterance from its initiation out to the extra-homodiegetic narrator and the final recounting. The critical space of the text is made to seem utterly commonsensical and logical, admitting to a "purity, clarity, uniformity and universality of narrative space" (Gibson, 1996: 8). The virtue of geometrical thought, writes Gibson, "is that it releases us from merely linear models. But that virtue is compromised, because geometrical segmentations appear according to an irreversible and, above all, a univocal order" (Gibson, 1996: 20). It is Menelaus's voice that dominates here: it is he alone who speaks, who quotes earlier speech, who alone addresses the narratee.

Gibson argues, moreover, that "[i]f narratology has never escaped a 'geometrics' of structure, it has surely never escaped an accompanying insistence of the full presence of a 'centre,' of 'the original or transcendental signified,'" has never, that is, escaped from the metaphysics of the Enlightenment (Gibson, 1996: 22). Gibson contends that within the various fields of narratology "there is the compulsion to practice a kind of boxing of narrative. . . , confin[ing] it within a fixed spatial arrangement" that is subject not only to structures of classification as elaborate and hierarchical as The Great Chain of Being, but also to "a structure of supervision," the layers and boxes typically "arranged in a downward-looking perspective" (Gibson, 1996: 215). It is a perspective, indeed, that reminds one of nothing so much as Foucault's panopticon. This "geometry of levels," Gibson writes, has a comforting simplicity and clarity. "With narrative levels, you know where you are" (Gibson, 1996: 216). Against these "dreams of the geometric" and the fixed centre from which we look out across Euclidean space, Gibson offers arguments by Deleuze and Guattari, who, he contends,

have waged war on the notion of strata, whether designed as the object of literary-critical, philosophical or scientific attention. For strata are always constructions, produced on the basis of a specific choice about how to see matter. Hence Deleuze and Guattari's deployment of metaphors - the rhizome, the "plane of consistency," transversals, the plateau - that offer images of a very different spatial organisation to that of the geometrical "box." These images flatten space, or multiply it, or cut across strata, redistributing their elements. (Gibson, 1996: 220)

Following Cornis-Pope, Gibson argues that "narratology must be persuaded to give up its 'fantasy' of taxonomic ordering and scientificity, 'its schematising rationality devoted to intellectual mastery'" (cited Gibson, 1996: 14-15). 36 One might wonder, though, if in fact this would mark the end of narratology itself. Perhaps a better prospect would be to allow narratology, resilient as it has proved to be, to reflect on these issues through "the successive application to the same texts of the narratological and the post-structuralist models [to] enable a principled comparison of the two, their logical relations and their relative modelling capabilities" (Margolin, 1986-87: 209). The object would not be the manufacture of a synthesis of two models, but the facilitation of more penetrating insights into the nature of narrativity and the nature of narrative (and its elements of "person," story, causal relations and so on) as socially, not merely textually, constitutive.

Narratology's treatment of voice is similar to its treatment of space. Echoing Margolin, Gibson writes that anthropocentricism and its investment in the notion of "voice" in narrative is essentially "predicated on the assumption that a narrator is a character present in a mystifying singleness throughout her or his discourse" (Gibson, 1996: 164). The seeming presence of a distinct voice within a text is taken by us as evidence of the presence of narrating subjects, just as the "self-presence" of our own "living speech" is taken as evidence of our own status as subjects. For Derrida, writes Gibson,

the privilege historically awarded to speech is inseparable from a particular valuation of reason, the conscious will, the self-identity of consciousness itself. On the one hand, voice guarantees the certitude and unity of "inner existence." It guarantees our self-presence, too, for the uttering self is deemed to be immediately present to itself as living consciousness. 37 Speech at once expresses a whole meaning that comes from within and is intimately tied to thought. Equally, speech has classically been understood as "mastery of objective being,"38 as guaranteeing the presence of the object "aimed at" by the speaker. (Gibson, 1996: 167)

As Derrida puts it, the classical conception of the voice and of voiced words provides them "a strange privilege" for the apparent absolute proximity of voice and being, of voice and the meaning of being, of voice and the ideality of meaning (Derrida, 1976: 12). For Aristotle, Derrida writes, the voice is conceived of as having a relationship of essential and immediate proximity with the mind, so that spoken words become seen as the symbols of mental experience and written words become seen as the symbols of spoken words (Derrida, 1976: 11). From Aristotle down through Descartes and beyond, it is by the virtue of hearing oneself speak that the subject affects itself. Within the logocentric notion of being, the apparent immediate presence of the voice to the self joins with and supports the general meaning of being, of being as presence, to form what appears an unassailable, self-evident truth. Indeed, this logocentric "truth" of hearing "oneself-speak" comes to form the basis of Descartes's first truth: cogito, I think. "The concept of voice in writing or as promoted by writing," Gibson concludes, "always involves a certain idealisation of 'vocal power' and risks retreating into a 'complacent daydream'" (Gibson, 1996: 166). 39 Gibson is perhaps stating the obvious when he reminds us that no-one actually speaks in literary narrative, and likewise, that no-one addresses anyone else. There is no material speech of any kind: all we have is text. "But it is precisely this," he contends, "that - locked in our stubborn anthropomorphic prejudices, insisting on the identity of the oral and the literary, of spoken and written narrative - we have been unable to recognise and dare not think" (Gibson, 1996: 148). This, indeed, is one of the arguments underlying Ann Banfield's Unspeakable Sentences. Following Banfield's lead, Gibson points out that we in fact do not actually encounter narrative "voice" in narrative at all.

We encounter an equivalent of voice that we have created for ourselves as a simulacrum and insisted on identifying with voice. Narrative voice is a theoretical construction, used, not only to give narrative a secure and unitary foundation, but to protect us against the disturbing recognition of what Banfield calls the "abstract" nature of the language of narrative as it is "made possible by writing." (Gibson, 1996: 146) 40

Like narrative "person" itself, then, voice too is "inseparable from and cannot exist without a hermeneutic procedure" that is steeped in Cartesianism's anthropocentric imaginary (149). Although taking issue with Fludernik's retention of anthropomorphic schemata (see page 29 above), Gibson does allow that:

What is clear from [Fludernik's] work . . . is the extent to which the concept of voice relies on and is inseparable from a philosophical essentialism. . . . The very assumption of a voice itself constitutes a mode of reading. . . . Yet Fludernik is right, too: it is not as easy to "think beyond" the concept of voice as Barthes would have us believe. The idea of voice needs to be further problematised, its resistance further weakened. But it is not to be written off. (Gibson, 1996: 150-51)

Conceptions of voice and the geometry of structure within traditional narratological analytical approaches to literary narrative, then, are clearly in need of scrutiny for their explicit and implicit assumptions about narrative's hypothetical (and real) worlds. Yet it might also be the case, as Margolin and (more grudgingly) Gibson imply, that the methods of narratology should not be repudiated out of hand. Rather, what we must explore is the role within the fiction-making process (when reading, writing, imagining) of the "truths" about narrative that narratological analysis has so keenly illuminated for us. That is, we must question, and be sceptical towards, the part played by our own prejudices about narrative (fiction) in our acts of making narrative (fiction). The anthropocentric categories and reconceptions of narrative "person" in particular (and the metaphors of "person," voice, point of view, narrative and analytical space more broadly) so thoroughly inform the social significance and signification of texts as to be inseparable from prevailing reading practices. Certainly, they could not be excised from reading practices without simultaneously relinquishing some degree of a text's social or thematic or emotional or, dare I say, ethical meaning.

2. Geist

"Voice as represented by narratology," Gibson writes, "is actually always a metaphor for intention, meaning and totality. Voice in narrative is Geist" (Gibson, 1996: 169). This anthropomorphic Spirit, of course, is none other than Cartesianism's "I" in all its self-identical, disembodied, self-affective density. As was made clear in my earlier discussion of Cartesianism, it is not my intention to enter into a critique of Cartesian thought per se. An exploration of the Cartesian mind-body divide would move us away from - though never contradict - the much broader sense I, like Kalaga, give Cartesianism. It will be useful nonetheless to pursue more narrowly and briefly a discussion of that aspect of Cartesian thought that does flow into Cartesianism, and which might therefore illuminate some of the assumptions underpinning this thesis and the attitude taken toward Cartesianism. To start, one might turn to Descartes's own "first truth:" that he, it, "x" thinks. Discussing Nietzsche's critique of Descartes in Nachgelassene Fragmente, Sarah Kofman writes that Nietzsche suspects that in identifying cogito as his first principle, that is, in claiming, after he has doubted all, that what he can be assured of is the existence of thought itself, Descartes has "confused an idea and a word." Cogito is a "pure word" rather than an "idea" because "it refers to no reality" (Kofman, 1991: 184). Cogitare, to think, she writes, is a pure abstraction (184). From Descartes's first fiction there consequently follows "[a]n entire series of fictions . . . , including that of the thinking subject," which is to say, as Nietzsche puts it, both "the deed and the doer are fictive" (Nietzsche, cited in Kofman, 1991: 184). Indeed, Nietzsche has it that the whole representation by Western ontology of the condition of the "doer" is comprised of "[a] mobile army of metaphors, metonymies, anthropomorphisms . . . truths are illusions of which one has forgotten that they are illusions'" (Nietzsche, cited in Pheby, 1988: 23).

If you remove from the cogito its tissue of simplicity, if you unravel it and cut it up, you will find a certain number of elements that are so many concealed beliefs.

First: a postulate grafted onto the previous fiction of a thinking activity: "it (es) thinks." Then a second postulate: this es is an Ich, I believe that it-is-I-who-thinks.

To this belief, this superstition, it is easy to oppose the following fact, namely, "that a thought comes when 'it' wishes, and not when 'I' wish, so that it is a falsification of the facts of the case to say that the subject 'I' is the condition of the predicate 'think.' It thinks; but that this 'it' is precisely the famous old 'ego' is, to put it mildly, only a supposition, an assertion, and assuredly not an 'immediate certainty.'" (Kofman, 1991: 185) 41

As Nietzsche seems at pains to point out, recourse to "it" as well as or in preference to "I" solves very little indeed, because "it thinks," Nietzsche argues, also contains the prejudices of "a belief." This belief is that thinking itself "is an activity for which one must imagine a subject, if only 'it' . . . " (cited in Kofman, 1991: 178). But having set aside "I (think)" cogito - and the problematic "it (thinks)" cogitat - Nietzsche wonders whether or not instead, "in order to record a state of fact free from articles of faith," we might just say cogitatur "to (think)" (cited in Kofman, 1991: 178). "But this is to delude ourselves once again, for even the passive form contains articles of faith and not only 'states of fact;' when all is said and done, it is precisely the state of fact that does not let itself be laid bare; 'belief' and 'opinion' are found in cogito or cogitat or cogitatur" (cited in Kofman, 1991: 178). It seems, then, that one can only draw from the sequence cogito, cogitat, cogitatur, not some "logical" consequence - the "fact" of the thinking subject - but "a simple tautology: 'Something is believed, from which it follows that something is believed'" (Kofman, 1991: 191-92). The ergo of the ergo sum shouldn't be thought of as announcing the necessity of a logical conclusion, but rather as translating

only the necessity of a very strong belief raised to a veritable article of faith: the belief in the subject and the predicate, that is, the belief in identity, in fixity, in the whole Aristotelian logic founded only on grammar.

"There is thought," therefore "there is a thinking subject:" this is the upshot of all Descartes's argumentation. But this amounts to positing as true, a priori, our belief in the concept of substance; to say that there is thought there must also be "something that thinks" is a formulation of our grammatical habits that presupposes an acting subject behind every act. In short, here already one constructs a logical and metaphysical postulate, one does not simply record it. . . . By the Cartesian path one does not reach absolute certainty but only the fact of a very strong belief. (Kofman, 1991: 186)

But what follows from the realisation that our experience of subjectivity is steeped in "a certain prejudice" need not traumatise us too deeply, because as Nietzsche avows: "The fictive world of the subject, of substance, of 'reason,' etc., is necessary" (cited in Kofman, 1991: 194). It appears, writes Nietzsche, and as Heider and Simmel's animation demonstrates, that "there is in us an organising, simplifying, falsifying power, a maker of artificial distinctions. 'Truth' is the will to master the multiplying sensations, to organise phenomena into definite categories" (cited in Kofman, 1991: 194). The fact that the "will to truth" can only ever produce fictions may become quite beside the point. This is certainly what C.S. Peirce's critique of Cartesian thought argues on this issue. As Michaels puts it, when Peirce attacks Descartes, "it is not for concealing prejudices, but for imagining a category of the unprejudiced . . . " (Michaels, 1980: 195). Rather than condemn Descartes for the prejudice that may be revealed behind his conceit of doubt, as is Derrida's deconstructive move, Peirce embraces it, describing it as revealing "the most essential point of Cartesian philosophy, which is to accept propositions that seem perfectly evident to us is a thing that, whether it be logical or illogical, we cannot help doing" (Peirce, cited in Michaels, 1980: 197). Michaels proposes that when Peirce embraces this moment in Descartes's argument, "he is embracing not the error in logic but the alternate notion of a self which is no longer neutral, no longer context-free" (Michaels, 1980: 197). If, again, this robs us of the fixed, singular centre that is always our safest vantage point, so be it. The point of any inquiry into Cartesian thought and any critique of Cartesianism, as Kofman writes, should be focused on a particular problem.

[E]ven if cogito ergo sum was irrefutable from a logical point of view, even if it did not rest on "illogical" postulates, vicious or tautological circles, and a contradictio in adjecto-one would still wonder why Descartes believed the certain to be worth more than the uncertain, why he believed in the value of logic.

And if to opt for the value of logic is to refuse to let oneself be seduced by this serpent, life (and all vita is femina), to refuse to be charmed by this veil "interwoven with gold, a veil of beautiful possibilities, sparkling with promise, resistance, bashfulness, mockery, pity, and seduction," to refuse to be conquered by the charm of woman, for "ah yes, life, life is a woman" 42 then to opt for logic is to give proof of a lack of virility of the instincts, and it is, definitively, to opt for a truth that is inertia, to opt for death. (Kofman, 1991: 194)

Over the last century or so, then, the power of the centre to hold and remain singular, the power of the philosophical subject to hold sway over our modes of thought, has shown some evidence of decay: witness Peirce, Nietzsche, Foucault, Derrida, Cixous, Kristeva, Irigaray, Deleuze and Guattari, amongst the others. Yet, still supported by the metaphysical tradition and the various forms of the subject nominated by Aristotle, Plato, Descartes, Hegel, Kant, Husserl, and so on, the hegemony of the Cartesian cogito remains powerfully entrenched and influential. Were this not the case, perhaps, if Cartesianism's hegemony were less efficacious and we more open to the forms of subjectivity met in such texts as I will discuss in later chapters, it might be that Protean "second-person" narrative would have scarce the effects that it does.

Part 2. The Pragmatic Snark

3. Foucault's Subject

It has been put that the problem of the reader's subjectivity must be a false problem, because it relies on a "wrong" self, that is, on Cartesianism's self: the autonomous reader confronting an autonomous text (Michaels, 1980: 198). Before going further, then, having argued against Cartesianism's notion of subjectivity and the self, it is necessary to digress to offer at least the outline of an alternative, non-anthropomorphic notion of self and subjectivity. The notion of subjectivity that I will offer in place of the notion circulated by Cartesianism is essentially the notion of the self posited by C.S. Peirce, a conception that is of particular value because it is not only a theory of the subject, but also a theory of meaning, of semiosis. My discussion will draw upon Peirce's conceptions of the self and identity and on his theory of the sign and semiotics, especially as Vincent Colapietro discusses these matters, and on Anthony Kerby's conception of the "semiotic subject." The issue I am dealing with in settling on one particular conception of subjectivity in preference to another, as Colapietro puts it, "primarily concerns whether one is going to adopt a subjectivist approach, in which an isolated and, in effect, disembodied human subject is treated as the ultimate locus of meaning and truth," or an approach "in which some human community functions as the fundamental source of both meaning and truth" (Colapietro, 1989: 27). Colapietro identifies the former directly with Descartes's subjectivism and with Locke's "new way of ideas," and suggests that Peircean semiotics "promises a way out of this subjectivism, for in granting priority to signs over ideas it shifts the focus from what occurs within a finite, individual consciousness to what occurs between social beings within a common framework of experience and action" (Colapietro, 1989: 27).

It needs to be noted, too, that my position on subjectivity is also broadly informed by Foucauldian assumptions on the nature of the human person's relation to language, which Chantal Mouffe very succinctly sums up this way:

The social agent is constructed by a diversity of discourses among which there is no necessary relation but a constant movement of overdetermination and displacement. The "identity" of such a multiple and contradictory subject is therefore always contingent and precarious, temporarily fixed at the intersection of those subject positions and dependent on specific forms of identification. This plurality does not, however, involve the coexistence, one by one, of a plurality of subject positions, but the constant subversion and over determination of one by the others that makes possible the generation of totalising effects within a field characterised by open and determinate frontiers. (Mouffe, 1992: 28)

Foucauldian theory provides me with the necessary conception of subjectivity as discontinuous and contingent and as affected within language, as opposed to Cartesianism's conception of subjectivity as unified and self-sufficient. It proposes, moreover, that discourse is not to be conceived of as "the majestically unfolding manifestation of a thinking, knowing, speaking subject, but, on the contrary, [as] a totality, [as a regularity] in which the dispersion of the subject and his discontinuity with himself may be determined" (Foucault, 1991: 55). Indeed, Foucault argues that one of the functions of discourse itself is to make the radically discontinuous and dispersed human subject appear unitary, and to make it appear, moreover, as the origin and principle of unity of the discourse that speaks it. In "What is an Author" (1979) and (more obliquely) in Archaeology of Knowledge (1991), he suggests that knowing the identity of the subject requires the analyst to see the subject's identity as being always situated, as being produced in no small part within the fields of discourse through which it moves. Toward the end of "What is an Author" he offers the insight that, as an analysis of the circulation, vaporisation, attribution and appropriation of discourses, the methodology brought to bear on the analysis of the "author-function" might be productively extended to an exploration of the category of the subject. "Clearly," he writes, "in undertaking an internal and architectonic analysis of a work (whether it be a literary text, a philosophical system, or a scientific work) and in delineating psychological and biographical references, suspicions arise concerning the absolute nature and creative role of the subject" (Foucault, 1979: 137). The individual subject - expressive, unitary, and self-identical, Platonic, Cartesian, or Kantian, whether self-knowing or not - is put into question. But Foucault immediately adds that the category of the subject should not be entirely abandoned. It should be reconsidered, not to restore to it the theme of an originating subject, but to seize its modes of functioning, its points of insertion/intervention into discourse, its system of dependencies (Foucault, 1979: 137). As he writes: "We should suspend the typical questions: how does a free subject penetrate the density of things and endow them with meaning . . . ? Rather, we should ask: under what conditions and through what forms can an entity like the subject appear in the order of discourse . . . ?" (Foucault, 1979: 137).

It is important to remember, however, that for Foucault the subject is not an entity. Whatever the subject's relation to the entity, it is not a relation of identity. The category of the subject is a function. Crucially, it is a function of enunciation, and as such is constituted in and by the act of making a statement, even (indeed, most frequently) by the act of making a statement to one's self. It is also, however, an "empty" function, a "vacant space" (Foucault, 1979: 95) bounded by, and only identifiable in, a particular and specifiable network of discursive relations - relations that are themselves endlessly variable and multiple. Each shift of the boundaries defines another "vacant place" from which statements can be made. Statements themselves, however, described by Catherine Greenfield as the constitutive parts of discursive formations, in this respect

represent neither the formal principles of the object "language," nor the conditions of a "social real." That is, statements do not have a representative function but are better thought [of] as productive of effects. Statements are not the effects of some pre-discursive entity to which we have recourse through their mediation; they have or produce effects which we can call knowledge effects. (Greenfield, 1984: 48)

In so far as the subject is constituted in and by the act of making a statement - not by the very act itself (as a variety of Cartesian subject, capriciously announcing: "I speak, therefore I am")it is through, and as, a process that situates the individual speaker in a complex of discursive relations in which s/he has (or hasn't) the right or the obligation to participate, and in which appropriate and inappropriate forms of statement are clearly delineated. Series of statements immerse and re-immerse the subject of knowledge in realms in which what is or is not known/validated is constituted not by itself or by the speaker, but within the field. To specify the subject in its discontinuity, Foucault argues, is to specify the rules of the particular discursive formation and its practice.

As suggestive and valuable as Foucauldian theory is to an anti-anthropocentric, non-Cartesian discussion of subjectivity, however, it nonetheless says too little about the daily experience of the continuity of identity that we do by and large encounter as social subjects. Alone, Foucauldian theory provides too general a field of material to adequately prosecute my thesis. Principally, it is a means by which one might describe the regularities of the discursive practices through which the social subject is realised, without necessitating that these practices be taken as the synthesis of a subject, and without assuming them linked by a system of relations intended "by the synthetic activity of a consciousness identical with itself, dumb and anterior to all speech" (Foucault, 1991: 54). To analyse and define the subject in this sense would be to treat with specificity the discursive practice of the statements themselves (Foucault, 1991: 55). It would be to describe, firstly, the systematisation of who has the right to speak, involving (i) a criteria of competence and knowledge, (ii) a system of differentiation and relations with other individuals/groups, and (iii) a definition of the speaker's role in relation to society as a whole (Foucault, 1991: 50). Secondly, it would be to describe the specific institutional sites from which the enunciations are made and from which each enunciation "derives its legitimate source and point of application (its specific objects and instruments of verification)"sites such as The Library as the documentary field of books, dissertations and essays that can be supplied to inform the speaker (Foucault, 1991: 51). And thirdly, it would be to describe the roles (of questioning, listening, observing, resisting and so on) adopted by the speaker that situate him/her in perceptual and informational networks whose specifiable boundaries "delimit the wheat of relevant information," hiving off or invalidating knowledges inappropriate to the authorised performance of the role (Foucault, 1991: 52).

Peirce, on the other hand, addresses the problem of the subject's experience of selfhood and self-identity more closely. Peirce, particularly through Colapietro's reading of him, provides a notion of the self that is not only complementary to the Foucauldian image of subjectivity and the self, but provides the explicit links between the social subject and the textual subject that will enable me to posit a notion of subjectivity that sustains the arguments of my thesis. For Peirce, as for Foucault, the subject is realised within semiosis; the person becomes a social subject only within language - or in Peirce's words, "the subject in its innermost being is itself a form of semiosis" (cited in Colapietro, 1989: 37). In this context, I will also draw upon Anthony Kerby's discussion of the semiotic self. Crucially, what Peirce and Kerby provide is an element of the theory of subjectivity necessary to my propositions about the deeply unsettling effect of Protean "second-person" narrative. Each provides a model that is radically opposed to the entrenched Cartesian orthodoxy that utterances originate definitively with the individual subject. In its stead, each proposes a model of signification that not only dethrones the autonomous subject from the centre of sign production, but posits a practice of semiosis in which no moment is originary.

4. A Speaking Body

Where the completion of the picture needs to begin, perhaps, is with the assumption of the "fact" of the subject itself: "that indeed there is a subject that apparently underlies any possibility of signification or process of communication" (Colapietro, 1989: 28), a sort of "ghostly presence," to paraphrase Umberto Eco, that will finally make an unavoidable appearance wherever there is discourse, as the concrete producer of signs. 43 Time and again this "ghostly presence" has backed literary critical theorists into a corner where they seem left with no alternative but to assume the common-sensical status of the authorial subject as the originator of literary texts. Many critics, of course - like E.M. Forster and Rosenblatt and the entire humanist tradition of literary criticism - rush to that corner freely, espousing Cartesianism's position: of course authors are the origins of texts. Others, like Booth, go to that corner more slowly and circumspectly, finding it necessary to speak about the category of the author as seemingly originary, but also finding the prospect highly problematical. Booth's solution is to institute a category of hypothetical subject he calls the "implied author," proposing that the "historical author" necessarily remains a mystery on the far side of texts, that the only authorial subject the reader can realise from the text is an inscribed subject, a hypothetical postulate posited within the reading of the text as the ultimate origin of its discourse. Many critics have subsequently argued against Booth's category, Genette for example insisting that there is "no apparent instance of an implied author that cannot be simply identified with either the real author or the narrator" (Genette, 1988: 139-40). Bal's more acerbic criticism is that the notion became "very popular because it promised . . . to account for the ideology of the text" in such a way that it "would have made it possible to condemn a text without condemning its author and vice versa - a very attractive proposition to the autonomists of the '60s" (Bal, 1981: 42). Lanser treats the category with similar misgivings, arguing that it dangerously severs aesthetics from ideology and that the actual author should still be held directly responsible for his or her own text (Lanser, 1981: 50). Throughout such criticism, Booth remains sufficiently comfortable with the concept that he can write, nearly two decades after its introduction in The Rhetoric of Fiction (1961), that "[a] respect for implied authors can be the most important single enabling act of criticism" (Booth, 1979: 278).

As over-large as such a claim might be in the context of contemporary criticism, my own concern with the term, as illustrated within William Nelles's concise summary of the various formulations of the category, is that the implied author and criticism of the category remain subject to Cartesianism's gravity. For Nelles, the implied author is "a critical construct, inferred from the text and with no existence outside that text," who readers conceive of as having "consciously created and intended every implication, subtlety, ambiguity, and complexity that can be discovered in the text" (Nelles, 1993: 26-27). Nelles also supposes that the distinction between the historical and implied author

is not entirely nor even primarily one of reality versus fictionality, but rather of the type of evidence admissible in the construction of them by the reader: the historical author may be created by recourse to any data while the implied author may be constructed only on the basis of the literary text being analysed. (Nelles, 1993: 26)

Nelles seems to imply here that the images of both authors are constructed by the reader, yet it is not a problem that he pursues. The notion of the implied author by and large clearly leaves the underlying nature of the social subject, that is, of the historical author that is the implied author's necessary corollary, uncontested. I would suggest that Nelles is right to assert that the difference between the historical and the implied author is not that of reality versus fictionality: so far as the historical author is still construed as an autonomous, originary subject, I would put it, in the Nietzschean terms sketched above, that both categories are fictive. As a category predicated on Cartesianism's notion of the subject, the implied author is a fiction of a fiction.

Anthony Kerby's use of the category of the implied author also explicitly reiterates the prevailing truism that readers will always tend "to attribute, in one way or another, almost all texts to an implied author" (Kerby, 1991: 106). But in Kerby's model, although he does not put it in these terms, the implied author becomes somewhat of a "ghost in the machine." It is a hypothetical entity required by the hegemony of Cartesianism and by our habit of attribution. And beyond this, the category also has a function in terms of (self-)affirmation. In setting this category of implied author in place, Kerby is in part saying that others in the world (i.e., the reader) powerfully (re)affirm our (i.e., the author's) identity by recognising us as speaking and by engaging reciprocally in discourse. The speaking subject is always immanently a "listening" subject (so to speak), and is caught up at length in Benveniste's "I"/"you" dyad. The "I" does not speak all the time, but sometimes becomes a listener, even (and perhaps most importantly) to itself. 44 Mobilising Benveniste's discussion of the role in the constitution of subjectivity of the exchange between "I" and "you," Kerby shows there can be no "I" without the "you," not so much for the want of somebody to speak to and to speak back, but because without "you," "I" cannot have any experience of itself (and so cannot, in any sense useful to us, be thought of as possessing subjectivity). By definition, the addressee is always paradigmatically "you" (and this "you" may be the speaker's self in a dialogue, which is a point Peirce elaborates upon in particular), and the addresser always paradigmatically "I."

Kerby's position is that the self's identity is constituted within a triadic, spiraling division of the subject that is comprised of:

the speaking subject or material agent of discourse, the subject of speech or purely linguistic subject of discourse (designated by personal pronouns and other deictical indications), and the spoken subject or subject produced through or by the discourse as a result of its effect on a reader-listener. (Kerby, 1991: 105) 45

Critical of the way in which, in his reading, discussions of the sujet de l´énoncé and the sujet de l´énonciation by Barthes and others have left the "subject of speech" afloat in space, he argues that it should, as Foucault insists, be more definitely situated. Accordingly, he argues that it is the entity's body, "as both the site of narration (the speaking subject) and site of ascription (the implied author #2)," that grounds it (Kerby, 1991: 107). "It is here," he writes, "that our common-sense notions of ourselves as embodied subjects are satisfied" (Kerby, 1991: 107). The great danger, of course, is that explication of "our common-sense notions of ourselves as embodied subjects" may, without due and critical scepticism, leave itself open to Cartesianism's constructions of the privatised self. Kerby works hard to circumvent any return to Cartesianism's fantasy of the self, in part by insisting that the body's role be fully acknowledged. Far from passive, it is a body of drives, habits and memories, "a living body of gestures and articulations that exist in extensive interaction with other acting bodies and the products of semiosis . . . " (Kerby, 1991: 101). And the implied author, in its turn, as it is construed by the reader, is the conventional, prevailing conception of that body as an intentional consciousness.

Fig. 8. Kerby's model of the Semiotic Subject

Kerby describes his model as involving two primary realms, Signification and Body, a situation that has direct implications for the implied author. The first implication is that the implied author need not be embodied. "If I write a letter to a friend, for example, I become for the reader the implied author #2 of the correspondence. If the letter were anonymous the exact site of ascription for the implied author could be lacking; this would lead to implied author #1" (Kerby, 1991: 106). As made explicit in Kerby's model, the implied author always operates at the level of signification, is always and only a hermeneutic category, even if it can sometimes be equated (adequated) with the lived body of an utterer. My own feeling is that this does not indicate a different category or sub-category of implied author, merely a different use of, or value attributed to, the utterances, dependent on whether its implied author is taken to be a single unique body (e.g., the producer of a personal letter), or an anonymous or corporate agency (e.g., the producer of a mass-produced form letter, or the "author" of a movie, involving producer/s, director/s, cinemaphotographer/s, set and costume designer/s, editor/s, and so on). That this value at its base manifests the ideologies of Cartesiansim is demonstrated by the fact that the individual's utterance is typically valorised over the corporate entity's. One might look, for instance, to the value ascribed to "authorship" in relation to auteur cinema as opposed to classical Hollywood cinema. In the former, authorship is forced into the foreground, so much so as to sometimes obscure the film itself; in the latter, authorship will most frequently evaporate altogether as an applicable concept. A second implication is more far reaching. Placed where it is within the model, being constituted from within the level of signification, the implied author is revealed as having the capacity to authorise nothing. It is itself already spoken. In so far as readers manifestly continue to construe the implied author as authoritative, it is very much a self-fulfilling, teleological authority, the reader's final image of the (implied) author guaranteeing the reader's own sense of the text - a notion that in fact very much agrees with Booth's original conception of the category's role in reading.

There is, then, the problem of the "reader" him- or herself. Kerby proposes that subjectivity is produced in some type of auto-affective practice in which we want to be social subjects, we "desire" to speak to one another and to ourselves, and we desire identity. Whatever subjectivity may be beyond this, however, he asserts it is certainly "blind without mediation through the realm of signification, [and] signification is not a neutral mirroring process. Subjectivity, as a form of vouloir dire (a wanting to say, to be, to do), is manifest as the speaking embodied subject that seeks to carry over into expression the implicit truth of itself . . . " (Kerby, 1991: 107-08). Borrowing from Stephen Crites, he points to the "storied nature of our experience," a nature which "holds the past (memory) and future (anticipation) together in the present, creating the more or less unified sense we have of our ongoing lives, a sense upon which our personal identity so thoroughly depends" (Kerby, 1991: 8). That is, in effect, I tell myself a story about who I am. In the case of self-narration of a past event, "the speaking subject is myself qua language user and 'repository of images,'" an embodied entity in communication with other embodied entities funded by an assemblage of memories and acquired knowledges and so necessarily "conditioned and restricted" by language, tradition, memory of past experience, and so on (Kerby, 1991: 105). The speaking subject, in an act of "expression," utters a statement about "I," but of course being spoken - expressed, represented - the "I" is not identical with the speaker, but has been externalised by language. It has become a subject of speech; it has become text. The spoken subject is constituted when, in an act of "participation," this speaking subject/body identifies itself as being "thematically represented" by the subject of speech, in the utterance's "I." It recognises itself (or not) in its own statements (expressions), and so identifying (participating) with the linguistic subject of speech, it acquires a meaning to itself, acquires subjectivity; then it speaks again, and recognises itself/participates again; and again.

The speaking body is therefore in some sense a "pre-subjective" entity: it is a body whose status as a subject is supported by the traces of its prior engagement in this triadic system of semiosis. Valuably, and very neatly, the model shows how momently and momentarily the subject does appear to possess a "prior-ity," a prior self-consciousness, that enters the processes of signification and participation. The point is that this self-consciousness is not a self-sufficient unity, nor is it in any way prior to language. Its "prior-ity" is produced. Its identity is in no way transcendental, because it itself is produced in the preceding moment of participation. Crucially, if one allows that the signification of self-identity spirals on in this way without end, what Kerby's model of the semiotic subject "seeks to emphasise is the division of the subject into different moments with no central and originating core" (Kerby, 1991: 107). Summarising, the two things that I want to take from Kerby into my discussion of Peirce, in particular, are the notion that the subject does possess a "prior-ity" of sorts, and that identity evolves within what becomes a self-guaranteeing dialogue, whose "first words" are irretrievably lost - or at best, are (re)traceable only in what Lacan would call the Real. 46

5. The Skins of an Onion

Like Kerby after him, C.S. Peirce, according to Colapietro, believes it absolutely necessary to acknowledge the "ghostly presence" that is the concrete producer of signs in any conception of the subject (Colapietro, 1989: 28-29). But while Peirce, like Kerby, regards the body as the "site of ascription" of such a subject, since "some sort of embodiment is required for any instance of semiosis," he is more rigorous in his exclusion of anthropomorphic categories (Colapietro, 1989: 84). Peirce offers a conception of human subjectivity much less acquiescent to the hegemony of Cartesianism, refusing to posit for it the ideality of an "author" of any form and avoiding terms such as interpreter or speaking subject that might, however obliquely, "hypostatise a personage and a particular individual mind . . . " (Spinks, 1991: 170). In a sense, it is as if Peirce offers a model of subjectivity in which Occam's razor has been applied to the "speaking subject," because the entity is always already, and only, instituted as a subject within language. For Peirce, it seems, the "ghostly presence" is not the ideality of an imputed "author" that conveniently stands in for the body (and vice-versa), but the manifest, brute presence of the entity's body itself, which is also a body of needs and drives, and as Kerby has it, a body of memory and habits, a body whose materiality may be continuous, but whose identity, whose meaning, can only be contingent. Peirce's person, like Kerby's, is "a living body of gestures and articulations that exist in extensive interaction with other acting bodies and the products of semiosis - speech, texts, art works, and meaningful actions generally" (Kerby, 1991: 101). As Colapietro puts it: "The body is not principally something in which the self is located; rather it is the most immediate medium through which the self expresses" itself to the world (Colapietro, 1989: 39), and less immediately, "the way through which the world expresses itself to the self" (Colapietro, 1989: 58).

So the self is to be conceived of not as a cogito but a sign - a sign, basically, by which the conscious entity knows itself. "[M]y language," writes Peirce, "is the sum total of myself; for the man is the thought" [5.314] (cited in Colapietro, 1989: 29). 47 Nor need this thought be conscious. Peirce writes: "The Action of thought is all the time going on, not merely in that part of the consciousness that thrusts itself on the attention . . . but also in the deeply shaded [or hidden] parts," such that "[a]ll activities of the mind are forms of semiosis" [7.555] (cited in Colapietro, 1989: 40). And like all signs, the self will occur in a context and so will owe much to the specificities of its culture.48 The human subject is not merely divided in the sense of being both a conscious and an unconscious being, it is "also a culturally overdetermined being," to be seen "not as a primordially free source of thought and action but rather as a being so deeply embedded in its time and place as to be largely, though not completely, constrained in its cognition and conduct" (Colapietro, 1989: 40). As Colapietro puts it, paraphrasing George Santayana, "[S]igns are simply the means whereby we acquire an awareness of the fact that we or something else is in the middle of things" (Colapietro, 1989: 23).49 That the self is not completely constrained is a point that Peirce dwells upon. He is as adamant that this self no more falls into the sweeping determinism of many subsequent cultural and social theories any more than it participates in Cartesianism's voluntarism. He argues, writes Colapietro, that three key elements constitute the mind and subtend our behaviour. The first is semiosis, as "the activity of a sign," and the second is habit, as "the disposition to act in a certain way in certain circumstances," a disposition principally composed of what Peirce refers to as "law." Kerby's description of the person as "a being of habit" in this respect agrees closely with Peirce's:

The development of the person will depend on a reflective grasp of, and habitual participation in, [a] network of social communication and praxis. The human subject must thus be situated within the structures that sustain it rather than posited as transcendent to them; it must be implicated in the production of such structures but need not be taken as foundational. (Kerby, 1991: 101)

Indeed, for Kerby, the "unity of habits" stands fully beside "the unity of self-narration" as providing the self with its illusion of unity and continuity, and these two correspond closely with Peirce's two elements of habit and semiosis. Significantly, however, Peirce adds a third: autonomy of self-control, as "the capacity of a person to regulate his or her conduct in light of norms and ultimately ideals" (Colapietro, 1989: 108).

Accordingly, a rational mind is one in which habits grow out of signs as the interpretants of these signs, and in turn, self-control grows out of a hierarchy of habits; however, once this self-control emerges, the possibility arises of having some processes of semiosis and some formations of habit grow out of self-control. (Colapietro, 1989: 108)

Peirce's basic model of semiosis is conventionally rendered thus:

Fig. 9. Peirce's triangle

As Kalaga describes the model, the Peircean sign involves "a mutually dependent coexistence of three necessary correlates" (Kalaga, 1996: 25):

1. the representamen, or the sign-vehicle;

2. the (immediate) object, which is a semiotic projection of the external or represented reality (i.e., of the dynamical object); and

3. the interpretant - the element which belongs to the realm of thought, mediation, cognition or, in other words, to the realm of Thirdness. This third correlate of the sign - the interpretant - is the meaning of the sign. (Kalaga, 1996: 25)

As a category that stands between the object and the interpretant, the representamen seems to draw from Peirce a delight in organic metaphors: it is "simply anything that has roots and bears fruits," and "anything that is grounded and growing" (cited in Colapietro, 1989: 22). Its "fruit," the interpretant, is described by Peirce as "the proper significate effect" of the sign. This "proper significate effect" may be conventional "meaning" as in the sense Saussure gives the signified; or it may be a feeling, which Peirce calls an "emotional interpretant," as a response for instance to non-programmatic music (i.e., music to which no formal narrative is attached); or it may simply be an action, which he calls an "energetic interpretant," as in the "flight or fight" impulse (Colapietro, 1989: 35). Or if the statements are about one's self, such as "I am," the "proper significate effect" of the interpretant would be the identity or Being of the utterer him or herself. The representamen's "roots," its object, which Peirce sometimes also calls the sign's "ground," writes Colapietro, is "that which the representamen presupposes an acquaintance with in order for the sign to convey some further information about it" (Colapietro, 1989: 9)but I will deal with the nature of the object more closely in a moment.

First, I want to turn to the matter of the relation between self and object. Peirce believed that "during any given moment of its life, the self is first and foremost a process in which some species of meaning is evolving" (Colapietro, 1989: 92). And Peirce seems quite adamant as to how this "species of meaning" does evolve into the self. Like Kerby, Peirce insists that the self is constituted within dialogue. Whether intra- or interpersonal, this process is such that the subject is in a condition of "always becoming" between the two sides of a conversation, between the two sides of semiosis. There seems a correlation here with Kerby's notion of the speaking subject as at once also a listening subject. But Peirce in no way implicates himself in Kerby's act of throwing an anthropocentric "sop to Cerberus" for the inclusion of a subjective speaker-function, even that of Kerby's quite rigorously formulated category of the embodied speaking subject. By utterance, Peirce does not mean to imply the presence of an "enunciating mind." Although the "action of signs usually takes place between two parties," the utterer and the interpreter as the representamen and interpretant need not at all be persons, "for a chameleon and many kinds of insects and even plants make their livings by uttering signs, and lying signs, at that" (Peirce, cited in Colapietro, 1989: 22). There is nothing necessarily "mental," writes Colapietro, "about either this quasi-utterer ('a source from which a sign springs') or this quasi-interpreter ('a form into which a sign grows')" (Colapietro, 1989: 19). As in the Foucauldian formulation that I am also deploying in which the subject is construed not as an entity but as a function and product of discourse, so the utterer and the interpreter of Peirce's model are no longer human entities, but likewise, functions within a particular notion of semiosis, functions whose necessary corollaries are human bodies.

What must be noted about Peirce's triadic model is that his notion of the object-sign-interpretant relation corresponds in no way to Saussure's sign-signifier-signified relation. Colapietro argues - somewhat heretically, he claims, at once asserting himself fully justified in doing so - that the relation of sign to the object in Peirce's theory is not that of "standing for" in any simple and unequivocal sense. Certainly, the relation is not specified in the way that Saussure ties down the relation between the signifier and the signified, and "in particular, the object is deliberately not characterised as being necessarily something that is represented or stood for by the sign" (Colapietro, 1989: 7). Peirce's discussion of representation reserves the term "representation" itself for naming "the operation of a sign or its relation to the object for the interpreter of the representation," and reserves the terms "sign " and "representamen" for naming "[t]he concrete subject that represents" (Peirce, cited in Spinks, 1991: 167). Colapietro, for his part, seems to regret Peirce's use of the term at all. He argues for its infelicity and the perspicuity of the term "mediation" instead: "it is better to speak of a sign mediating between its object and its interpretant than to speak of a sign representing an object to an interpretant. . . . In short, 'mediation' is a far more perspicuously general term than 'representation'" (Colapietro, 1989: 18-19). Colapietro concludes that "the function of the sign is simply not captured by [the] term ['representation']" (25)and that "[a] truly general theory of signs," as Peirce intends his theory of semiosis to be, "will show that semiosis is wider than 'representation' . . . " (Colapietro, 1989: 25).

Peirce's triadic model maintains that a "sufficiently complete sign" will involve each of three distinct functions of semiosis: Firstness, Secondness and Thirdness. The nature of Saussure's model, on the other hand, is fundamentally binary, comprising the dyad signifier-signified: "sign" is not a necessary third element beside these, but the structurally unifying name of the system. Saussure metaphorises the sign as a sheet of paper, the front and back of the page logically implying and physically requiring the other: one face standing as the signifier and the other as the signified, the paper in its totality being the sign. A second divergence between the Saussurian and the Peircean conceptions of the sign is marked by Peirce's notion of "unlimited semiosis." As Saussure himself observed, nothing means without difference (Saussure, 1960: 117). At the smallest material level it is the differentiation between phonemes, a differentiation of the signifier, that produces meaning: fish, dish, wish; much, mush, muck. Saussure's insight that in language there are only differences without positive terms, however, seems to conceive of difference as being exhausted in each binary moment: A is not B. A is not C. A is not D. And so on. Peircean "unlimited semiosis" posits quite a different notion of the function of difference in language, a notion that Derrida partly echoes in his concept of différance. For Peirce, there is between the representamen and the interpretant a particular relation of commutability in which interpretants are transformed into representamens, endlessly. In Eco's words, paraphrasing Peirce, the basic condition of semiosis is a weave of "signs sending back to signs, in an infinite regression" (Eco, 1979: 188-89). Anne Freadman has remarked that tracing the original object under these circumstances may therefore be a matter of peeling skins from an onion. 50 Peirce's position is that:

The meaning of a representation can be nothing but a representation. In fact, it is nothing but the representation itself conceived as stripped of irrelevant clothing. But this clothing can never be stripped off; it is only changed for something more diaphanous. So there is an infinite regression here. Finally, the interpretant is nothing more but another representation to which the torch of truth is handed along; and as representation, it has its interpretant again. Lo, another infinite series. (Peirce 1, 339)

One significant implication of "unlimited semiosis" for the self would seem to concern what Derrida addresses decades later in the notion of the "trace," a notion integral to that of différance. Derrida proposed that signifiers need not be given directly by the text, but may also be intrusions by intertexts that can be highly distanced, conscious or unconscious. These signifiers and chains of signifiers, thrown up in the midst of difference like sparks into the air, as it were, are formed and linked by the trace. The trace, or "spur," is a neurological analogy that may be crudely rendered as "the pathway of a thought." More than just a pathway, it is a process which both recalls and erases the last signifier, even as the trace of the next signifier begins to vitiate the present (Speech and Phenomena 13). Likewise, Peirce proposes that within every utterance there are echoes of the utterances of others: the self is never simply a speaker. "There is no intuition or cognition," Peirce writes, "not determined by previous cognitions" (Peirce 2, 284). But unlimited semiosis does not, avows Peirce, lead to an endless free-fall through language: there are rules and laws here. Since language is social, to a large extent this regression and progression is fenced by "community." Our engagement in processes of semiosis always occurs in a social context, making context itself a constituent of the interpretant (Spinks, 1991: 172). That is, the interpretant will always be part of a field that itself will always include myriad "possibilities, actualities, and dispositions (would-bes and generalities)" that profoundly effect interpretance (Colapietro, 1989: 17). In relation to "second-person" narrative, the matter of unlimited semiosis, I suggest, seems powerfully to dramatise the reticulation of signifiers and sliding indeterminacies of Protean "you" as I will discuss it in later chapters.

We need to return now to the problem of the nature of the object, foreshadowed earlier in Kalaga's description of it as involving an "immediate" and a "dynamical" dimension. Just as Peirce does not map over Saussure, Peirce's three terms are in no way easily superimposed term-to-term over Kerby's. Semiosis for Peirce is far more complex than any linear relation (even one disguised as a spiral) can account for. Whereas Kerby draws his model as unidirectional in so far as the subject of speech is logically unable to exert an influence back on the speaking subject, the case is quite different for the relation between the object and the representamen. Peirce does describe his system as being one in which "the object determines the sign and, in turn, the sign determines the interpretant" [1.541] (cited in Colapietro, 1989: 14), but he also stresses the possibility that a sign may become active in relation to the object, in some way determining the object. This is so because the object can in fact take two forms: the immediate object and the dynamical object. The Immediate object "is the object as the sign itself represents the object, and whose Being is thus dependent on the representation of it in the sign. This is how the object can be thought of as being determined by the sign" [4.536] (cited in Colapietro, 1989: 15). The meaning of "granite," for instance, will be determined to some extent by the physical properties of density, texture and so on, as will the meaning of "lounge chair" be determined to some extent by function as well as form, style and comfort. The determinants "density," "texture," "form," "comfort" and so on are themselves subject to the same processes of signification, so that the meanings of "to be dense" and "to be sat upon" themselves call on complex sets of determinants. The Dynamical object, on the other hand, "is the Reality which by some means contrives to determine the sign to its Representation" [4.536] (cited in Colapietro, 1989: 15). Peirce proposes that: "Every proposition which we can be entitled to make about the real world must be an approximate one; we never can have the right to hold any truth to be exact. Approximation must be the fabric out of which our philosophy has to be built" [1.404] (cited in Spinks, 1991: 38). Truth and reality are not posited for Peirce, then, within the object, but rather, are "an affair of Thirdness," are a matter of interpretance (Spinks, 1991: 40). But always instituted in its chains of unlimited semiosis, Thirdness is want to become Secondness, so that even if our sense of reality is a matter of Thirdness, Thirdness itself is never conclusive. The object, then, is never a thing that we can absolutely know, even in its dynamical form, because although we do have access to an objective world, the meanings we make of the world are highly mediated and cannot themselves be thought of as "objective," particularly the meanings we make of our own embodied selves, of ourselves as subjects.

Indeed, according to Colapietro, Peirce seems to have considered the embodied self "a perfect example of . . . a perfect sign" (Colapietro, 1989: 58). Peirce wrote that, like the self, "the perfect sign is perpetually being acted upon by its object, from which it is perpetually receiving the accretions of new signs. . . . In addition, the perfect sign never ceases to undergo change [of a spontaneous sort]" (cited in Colapietro, 1989: 58). The object within this model - identity - is discrete and discontinuous for every cycle of the function of semiosis: in each cycle it must be remade in some sense, in each cycle posited as a new (similar or dissimilar) object-identity. That is, each moment interpretance becomes yet another representamen, the triangulation of the sign implies the object's reconstitution. In terms of a notion of the self, one might construe the dynamical object as the body, that is, as the object in the world, as "a 'being' that is something apart from being represented" (Colapietro, 1989: 15). Although certainly a body of needs and drives, a body in all likelihood inscribed not only by a world of semiosis but also by the writing of genes and of biological and environmental chemistry (a writing that is at best cryptic and vague and for the most part read tendentiously and with prejudice), as object, the body is the only actuality that can be spoken of properly in this context. To repeat Kerby's words, the body is where our commonsense notions of ourselves as embodied subjects is satisfied. The immediate object, on the other hand, is the identity - the sense of self - that is ascribed to the body-as-sign, and consequently, however tautologically, the meaning the self gives to itself.

6. Conclusions

Cartesianism's self-ish logic, revealed as prejudiced - embraced by Peirce as unavoidably so - cannot, I suggest, fully account for the deeply unsettling effects of Protean "second-person" narrative. It will hope to retain the unitary, knowledgeable subject as the touchstone by which the difference of the "you" is measured, so that criticism will speak most frequently about oscillations and multiplicities of subjectivity, seeking the comfort of an authority anterior to such an unnatural "person"the subject that does not oscillate, the subject that is a singular totality. Rather, what this thesis looks towards in exploring narrative textuality, as does Margolin, Michaels and others, is the renunciation of Cartesianism's "error in logic"of the self's faith in its own priority, its mistaken faith in an originary moment beyond the prejudices of semiosis. It seems the case, certainly, that Cartesianism's logic has a role in constituting Protean-"you" as deeply unsettling to readers, but, as I will argue in Chapter 3 in relation to the anthropocentrism of the allocutionary function and more generally in Chapter 7, it is a role that compromises Cartesianism's self rather than supports it. In some measure at least, the Protean-"you's" effects might be described as arising between the reader's faith in an orthodoxy and the uneasy disclosure of its partiality, as arising, that is, between Cartesianism's prodigious hegemony and what philosophical pragmatism and post-structuralism glimpse as the actual nature of the human subject - a nature I assume closer to that characterised by Peirce than by Descartes. Peirce's model provides me with an anti-anthropomorphic notion of semiosis and subjectivity in which there is no centre, no originating moment, and a model in which "prior-ity," always evolving from utterance to utterance, can be isolated only in the moment of the utterance. But importantly, it also provides, as Kalaga asserts, a model that explains our overwhelming sense of the self's continuity. It is a model in which continuity is relational.

[It is] the inevitable semiotic continuity of a sign being interpreted by and within another sign, and that sign being interpreted within the next sign, and so on. Of course, all those traceable relations and semiotic chains are not held in being actually, but only in abstracto (or "potentially") pertain to the "now" of the subject-sign-interpretant. It is this attribute of the subject's becoming that we traditionally call the identity of the self, and it is also this aspect of semiosis that allows us to have the metaphorical expression "I am." (Kalaga, 1996: 32)

Peirce's conclusion is that everybody will tend to admit a personal self, but he writes that they "will admit a personal self in the same sense in which a snark exists; that is, there is a phenomenon to which that name is given. [ . . . ] It is an illusory phenomenon; but still it is a phenomenon" [8.82] (cited in Colapietro, 1989: 63). Peirce's metaphor for the object as the "ground" of the sign, then, becomes somewhat pointed, because it must now be seen as a ground upon which the self may never touch its feet: a discomforting condition indeed for Cartesianism's self to realise itself in.

the skinny

top

Index

1

2

3

4

5

6

7

Notes

Bibliography

Links

Slackwire