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

Chapter 1

The Effects of Narrative-"You"

1. Naming Narrative-"You"

"Second-person" narrative is a mode that may be defined, provisionally, as narrative in which the "second-person," personal pronoun "you" is used to identify and directly or indirectly address a protagonist. Or as Fludernik remarks, it is "narrative whose (main) protagonist is referred to by means of an address pronoun (usually you) and [which] frequently also [has] an explicit communicative level on which a narrator (speaker) tells the story of the 'you' to (sometimes) the 'you' protagonist's present-day absent or dead, wiser, self" (Fludernik, 1994b: 288). If both definitions appear vague and negotiable, they are necessarily so because of the often equivocal nature of the "second-person" pronoun within narrative discourse. Its fluidity and non-conventionality have ensured that more exclusive, categorical definitions are difficult to formulate. Darlene Hantzis, for instance, asserts that the "you" in "second-person" point of view proper will not only dominate the "third-person" pronouns that occur in any text, but will be deployed to the complete exclusion of the "first-person" narrative pronoun. Mary Francis Hopkins and Leon Perkins (1981), on the other hand, while working within the same tradition of point of view criticism as Hantzis, propose that an explicit "I" narrator does not necessarily disqualify a text from being categorised as a "pure" "second-person" point of view. Similarly, Gerald Prince's definition assumes the "you"-protagonist must always be the narratee (Prince, 1987: 84), yet Richardson more circumspectly proposes that the "you" is only "generally the work's narratee" (Richardson, 1991: 314). In fact, appreciative of the difficulties of adequately defining the "second person" as a distinct field, Richardson proceeds on the premise that the "very essence [of the "second person"] is to eschew a fixed essence" (Richardson, 1991: 311). His approach, consequently, is to enumerate tendencies that any instance of narrative-"you" may present to the reader rather than to stipulate invariant conditions, though like Fludernik, Hantzis and Prince, he too proposes that at the very least it is "narration that designates its protagonist by a second-person pronoun" (Richardson, 1991: 311). These tendencies, he writes, include present-tense narration and allegorical undertones, but most typically - as a matter that will be taken up at length below - the reader will be faced by an "irreducible oscillation" between the intimate voice of a "first-person" narration limited in the breadth of knowledge it can have access to, and the distant, omniscient voice of a "third-person" narration, simultaneously inviting and precluding identification with other pronominal voices (Richardson, 1991: 313). It is this characteristic oscillation that differentiates the "second-person" from types of displaced "first-" or "third-person" utterances found in authorial colloquies to a "gentle reader" (in Brontè, but Thackeray and Nabokov, too) or internalised debates and subvocal dialogues (of Beckett and Sarraute) (Richardson, 1991: 310), each of which, he writes, employs "you" in ways "easily comprehended by traditional dyadic theories of point of view" (Richardson, 1991: 311). Uri Margolin approaches the problem of defining the "second person" in a similar way listing a number of typical characteristics that he describes as necessary in defining "second-person narrative as a distinct variety of narrative discourse" (Margolin, 1997/1991: 6).

• The presence of a single global narrator on the highest level of textual embedding, such that the whole (fictional) discourse originates with him or her;

• The presence of numerous instances of "you" in his or her discourse, oral or written;

• The majority of these "you" instances refer to a narrated rather than communicative "you;"

• The speech acts of the narrator concerning the "you" thus go beyond apostrophes, questions, orders, etc. (Bühler's Appellfunktion) and tend toward the constative or representative, that is, reporting (Bühler's Darstellungsfunktion);

• The narrated you is a central agent in the sequence of events being recounted;

• The events/actions/states involving this "you" are specific and individual as regards their time and space, as opposed to the purely typical or recurrent (generic you, "you" as equivalent to "one" or "everyone"). (Margolin, 1997/1991: 6)

Not all six features are incontestable, of course, particularly given the case that instances of "second-person" narrative may yet be produced within narratives that do not unequivocally identify a single global narrator, as shown by Pynchon's Gravity's Rainbow (1975) and Daniel Gunn's Almost You (1994) (see Chapter 6). Yet Margolin's list of features does offer a practicable way of characterising the "second person" as generally distinct from the "first" and "third," and it succinctly provides, for those who require it, a means of provisionally establishing what Fludernik has called "the golden proportion between 'real' second-person texts and other fiction using the second-person pronoun in interesting and potentially significant ways" (Fludernik, 1994b: 284).

Bruce Morrissette's (1965) conclusion in this respect is that, in order for "second person" narrative to achieve a properly "literary" status in narrative (as opposed to being a flawed narrative's stylistic "trick" or affectation), it must fulfill certain classical definitions of narrative, such as dual time. 17 Traditional definitions of literary narrative insist that narrative necessarily involves this duplicity of time, which serves to establish a story-telling situation in the mind of the reader. The reader recognises that a narrative is being delivered. Within this story-telling situation, there is what can be called the "time of the narrating," the moment or period in which the narrator speaks or tells the events of the tale to whoever might be listening or reading. Then there is the time of the events of the story, the time in the near or distant past when the events occurred. The element of the "second person" must also fulfill the requirement of having a specifiable sense within the narrative logic or thematic structure of its particular text. For instance, Morrissette describes the "you" address of La Modification as functioning in the context of a crise de conscience or moral struggle, and that of Stout's How Like a God (1929) as functioning in the context of a retrospective self-judgment. And likewise, he describes the "you" address of Ashmead's The Mountain and the Feather (1961) as ineffectual for its lack (as he sees it) of any such logical context for the "you" (Morrissette, 1965: 137-38). In defining what a "second-person" narrative must involve for it to be judged a work of literature, it would appear that Morrissette is hoping to reappropriate the "second person" from its use in marginalised, popular, and mass forms of discourse - forms that tend to use conversational, vernacular or "oral" styles- on the assumption that these present a damaging or corruptive threat to the literary. But, as Kacandes argues, quite the opposite might be the case. The "oral" or conversational character of some forms of the "second person," in its foregrounding of a communicative circuit and courting of the reader's involvement, may in fact function as a bid to close (or at least to renegotiate) that alienating gap between speakers and hearers, between story-tellers and audiences, opened up by literacy (and exacerbated by the affections and assumptions of a "high" literature and criticism that sets itself at odds against a "low") (Kacandes, 1990: 224).

As Margolin, Richardson, Fludernik and others attest, any theory of narrative-"you" must also fully acknowledge the complexity of gendered, formal and familiar, singular and plural relations and functions of pronouns and pronominal verbs available (or unavailable) to speakers of other languages than English. And it must acknowledge that the analysis of each particular work of "second-person" narrative will need to be attuned to how these elements might function in each specific text relative to the particular language. English has a sole conventional "second-person" pronoun in current usage ("you") and a lone corresponding conjugation of the verb "to be" ("are"). In comparison, Spanish provides for its speakers five pronouns whose grammatical reference is a "second person" and eight corresponding forms of the verb "to be." Each of the five pronouns is heavily circumscribed by relations of power through conventions of polite and impolite usage. For instance, because of the difficulty of translating the systematic use of both plural/formal and singular/informal "second-person" pronouns in one work, Julio Cortázar's story "Usted se tendió a tu lado" (1977) remains unpublished in English translation, omitted from Gregory Rabassa's translation of Cortázar's Change of Light (1980) [Alguien que anda por ahi]. An English translation of "Usted se tendió a tu lado" might only become graceless and prolix, or otherwise condense the multiple modalities of the Spanish pronouns down to the portmanteau of the English "you." Likewise, in Italian, as Irene Kacandes demonstrates in her reading of the masculine "second-person" pronoun of Italo Calvino's If on a Winter's Night a Traveller (1982), the ways in which male and female readers are likely to engage with the "second-person" text are quite different. To identify with the you of the novel's opening passages, Kacandes asserts, the female reader "will have to read as a man," so that the novel in Italian "marginalises women" in a way that it need not marginalise its male readers, and in a way that it does not so immediately marginalise its female readers in English (Kacandes, 1993: 181).

The English language's relative paucity of "second-person-ality" can also obscure a second issue central to any adequate definition of "second-person" narrative, as Monika Fludernik points out in "Second Person Fiction" (1993). Some languages employ pronouns that, although not strictly "second person," do possess that pronoun's function of address: the German polite "third person" plural "Sie" and the Italian "third person" singular "Lei" being two cases. Because of this, and arguing that "the central irreplaceable constituent" of "second-person" narrative is not grammatical person but its allocutionary function, its function of address, Fludernik suggests that the generic designation "second-person" narrative is "a misnomer of major proportions" (Fludernik, 1993: 219). It should rather be called "pronoun-of-address fiction." Although I will remain with the convention of calling these "second-person" fictions, Fludernik's point that the address function is one of its principal constituents cannot be emphasised enough. As I will argue, this function is fundamental to the most penetrating - and often the most unsettling - effects of "second-person" textuality on its readers.

2. The Addressee Identity Model of "Second Person-ality"

It will also be useful at this early point to sketch the narrator-character-narratee-reader complex implicit as a powerful hermeneutic frame in many discussions of "second-person" narrative. In so far as the model that follows (figures 1. to 4.) does address categories of "second-person-ality," it should also be thought of as a meta-model whose object is not so much "second-person" textuality itself, but what is common or typical within its criticism and the various approaches proposed for its analysis. To that extent, I do not mean it to be taken in any straight-forward way as an analytical model of "second-person" modality, although it can clearly stand as such. Rather, it should be taken as representing precisely that mode of analysis to be interrogated throughout this thesis.

The fact that it is often difficult to fix the identity of the "you" to any one point within a narrator-character-narratee-reader nexus is generally (if frequently implicitly) accepted by analysts of "second-person" narrative - though not all, Holden and Bonheim amongst them, take this "referential slither" or "sloppy identity" as a positive feature. It is proposed that the "Protean shape-shifter," as Bonheim describes it, eludes narrative closure by refusing a final fixing of its referent. The "you" may point back to the narrating voice as self-address, it may point to a particular and particularised character dramatised within the narrative, or it may point to the reader/narratee - or it might point to no "person" at all. These relations will be greatly complicated by matters of "distance," involvement in the story, and so on, but they can be sketched as follows. 18

The first relationship of the addressee identity model is that in which a narrator addresses him or herself as "you" and describes to him or herself from a greater or lesser distance his/her own acts and thoughts, as occurs in Butlin's The Sound of My Voice (1987) and Frank Morehouse's "Walking Out" (1964).

figure 1

How, in bed early Thursday morning, do you explain to your father and mother whom you have lived with for 22 years, that you do not want to go to work and that you do not want to see your friends? How do you explain that you'd rather not see them too? How do you explain that the idea of working and the idea of seeing your friends makes you feel sick in the head? How do you explain that you think that life stinks and that you want to lie down somewhere on your own?

So I lay on my back in bed early Thursday morning knowing that the clock was ticking toward seven o'clock. . . . (Morehouse, 1964: 24)

As made explicit in the shift from "second-" to "first-person" here, the first relationship is one in which the implicit "I" is transformed into or disguised behind a "you." As Margolin proposes, in all such "I"-"you" transformations, the "you" in his or her "receiver role is always present in the speaker's immediate communicative field," but as topic might be temporally separated, situated either in the past, present or future in respect to the speaker's temporal position (Margolin, 1986-87: 196). If the transfer is in the present tense, writes Margolin, the reader should expect to find either of two subclasses.

In the first, an internal dialogue is taking place in which the "you" can also assume the sender role and "talk back" to the speaker. The more frequent case, however, includes an implicit, non-self-indicating I-speaker addressing an aspect or layer of his [or her] self in the explicit "you" form and not receiving any response from his split-off aspect. (Margolin, 1986-87: 196)

The speaker observes and addresses an aspect of him-/herself that is "regarded by him for the time being as an externalised object. The initial, unified I is split in two, and a hierarchical, non-reversible relation is established between the implicit I-speaker (superior half) and the aspects being addressed (subordinate half)" (Margolin, 1986-87: 196). Margolin writes that this one-way communicative situation is ideal for the thematics of self-confrontation, in which the superior half typically "questions, orders, assesses, or judges the subordinate one, articulating what this 'you' cannot or will not articulate" (Margolin, 1986-87: 196). Conversely, the speaker might "make explicit those things of which the 'you' splinter is unaware, which it has forgotten, suppressed or failed to comprehend," taking the role of some form of super-ego, self-consciousness or conscience (Margolin, 1986-87: 196). If the narration presents the "you" in the past tense, on the other hand, writes Margolin, "we are dealing with . . . a case of what Bühler has termed 'deixis am Phantasma'" (Margolin, 1986-87: 196). "In it, the speaker conjures up a past version of himself and talks to and about it as if it were present in his immediate deictic field, while distinguishing it from his present identity through difference in tense and person, e.g., 'How miserable you were then, my poor child'" (Margolin, 1986-87: 169). Each of these forms of dissembling-"I," and those to be discussed next, might be kept firmly in mind when I turn to Daniel Gunn's novel Almost You (1994) in Chapter 7. The deep ambiguities of Almost You are such that the reader might take any one of these cases as dominating Gunn's novel.

If figure 1 describes the relationship in which a narrator addresses him or herself, figure 2 concerns the relationship in which the referent of the "you" is a character within the story other than the narrator.

figure 2

In this second form, the narrator - whether dramatised or effaced: whether the narrator of Fuentes's A Change of Skin (1968) or of Calvino's If On a Winter's Night a Traveller (1982)might directly address a "you" who is more or less conscious of the address, or figure the character within some form of apostrophe.

In the shop window you have promptly identified the cover with the title you were looking for. Following this visual trail, you have forced your way through the shop past the thick barricade of Books You Haven't Read, which were frowning at you from the tables and shelves, trying to cow you. (Calvino, 1982: 10)

Although the first figure tends to involve what Genette calls autodiegesis, it is not possible to superimpose his classifications of autodiegesis, homodiegesis and heterodiegesis in any easy way over figures 1. and 2. For Genette, the heterodiegetic narrator never appears as a character, telling stories from which he or she is absent, whereas the homodiegetic narrator, subject to degrees of presence, will be present in the diegetic world of the text, participating in the story's unfolding events in some capacity or another (Genette, 1980: 244-45). In "second-person" texts, these degrees of presence can often involve degrees of self-presence or self-consciousness, in such a way that one element of the narrator's self or self-identity is telling a story about another self or identity--e.g., Christa Wolf's Patterns of Childhood (1990). The closest degree of presence produces the autodiegetic narrator, a variety of homodiegesis in which the narrator self-consciously tells his or her own story as the principal character (Genette, 1980: 245). Whereas figure 1. most frequently involves the "you"-narrator presenting him or herself as the principal protagonist, however, it is not inconceivable that a narrative may present a self-referential "you"-narrator who does not make him or herself the principal protagonist, reserving that status for another. So although the second figure would always seem to describe heterodiegetic narrators, the first figure may describe autodiegetic or merely homodiegetic narrators.

The third relationship differs from the first and second in that, where the "you" addressed in those is a characterised participant of narrative events, the "you" addressed in the third is not, although it may yet be figured within the diegetic frame of the text.

figure 3

It is the mode in which the narrator directly addresses the (implied) reader or a more or less specified narratee. The narratees of Ring Lardner's "Hair Cut" (1981) and Margaret Atwood's "Rape Fantasies" (1978) for instance, are positioned within the story as silent but powerfully present addressees of each monologue. The narratees of Barnes's Talking It Over (1991) and Sterne's The Life and Times of Tristram Shandy (1940) on the other hand, are situated outside the story and coincide more or less closely with each story's respective (implied) reader. The "you" readers harangued and cajoled by Barth in Lost in the Funhouse (1969) and Federman in Take It Or Leave It (1976), too, belong to this mode.

I got another barber that comes over from Carterville and helps me out on Saturdays, but the rest of the time I can get along all right alone. You can see for yourself that this ain't no New York City. [. . .]

You're a newcomer, aren't you? I though I hadn't seen you round before. I hope you like it good enough to stay. (Lardner, 1981: 786-87)

I have dropp'd the curtain over this scene for a minute,—to remind you of one thing,—and to inform you of another. (Sterne, 1940: 144)

Dammit! If you guys keep talking all the time / and at the same time / we'll never get it straight! We'll never get there! Do you think it's easy to tell a story? Any story? HEY! Particularly when it's not YOUR story—a second-hand story. Anywhere? To retell a story which was already told from the start in a rather dubious manner? Do you think it's easy to set it up so that it looks coherent? Or even readable? Not to mention credible? I tell you it's not easy. A life story (or even parts of it)! (Federman, 1976: n.pag.)19

In practice, however, there is no clear boundary between the third and the second form of "second-person" modality. This partial dissolution of boundaries is a necessary consequence of the varying degrees to which a "second-person" addressee can be involved or implicated in a story. The forms of "second-person" narrative might be thought of as points of a continuum. At one end, the figure addressed is a fully dramatised and psychologised protagonist; further along, the identity addressed as "you" is an undramatised but story-embedded narratee; and further still is the (putative) reader him or herself.

A fourth form, which is rhetorical in character and points to no particular "person" at all, is deployed to designate a generalised object..

The boy rose early, and tried to part his hair. [. . .] If you don't get your part right you get all cockies up the back. (Gasmire, 1990: 179)

figure 4

Conveying an extensive range of potential functions or rhetorical forces (such as didacticism, judgment and so on), these modalities become rendered in criticism (perhaps too loosely) by neologisms such as interrogative-"you" (named for rhetorical effect), and advertising-"you" (named for genre). It is also the form that in Chapter 3, where it is discussed at length, I designate as the "second person's" sweeping figural mode.

Models of classification, including the paradigmatic four-part system described here, however, have at least two important features of which criticism must remain mindful. Firstly, such models tend to define examples of the "second person" in terms of clearly differentiated "types." Mary Francis Hopkins and Leon Perkins perhaps best typify this approach. Working within the tradition of point of view criticism, they set out to demonstrate that "narrative-you" proper (named after Morrissette's example) can be as multiple and varied as "first-" and "third-person" points of view. They illustrate this through coining such categories as "second-person-limited omniscience," "second-person-personal," and "second-person-impersonal centre of consciousness." In their introductory remarks, they readily accept "the hopeless generality of long-accepted terms such as 'point of view,'" yet insist that "these terms remain useful, even retain vestiges of their original sense, not only because they dominate our heritage of narrative theory, but also because, recognised as areas of concern rather than single items, they name what must still be our basic considerations" (Hopkins and Perkins, 1981: 119). Rather than hoping to renew critical language, Hopkins and Perkins "simply introduce additional terms and descriptions to clarify the distinctions. . . . [It is the case that] categories of person . . . pervade our discussions of point of view; therefore, 'second-person point of view' can name a class of narratives or stretches within longer narratives distinct from first-person and third-person narratives" (Hopkins and Perkins, 1981: 119). Certainly, they are right in suggesting that metaphors of point of view are powerfully entrenched in both literary and social discourse. But as Richardson warns, the Protean fluidity of "second-person" fiction does not lend itself to such approaches and attempts to categorise it in terms of manifest features is bound to fail (Richardson, 1991: 311). His own solution is to describe the limits of his three forms of "second-person" modality, the "standard," "autotelic" and "subjunctive," as clusters of typical characteristics or tendencies rather than as sets of constitutive features. His principal argument is that "second-person" narrative plays on the boundaries of other narrative voices, and that it is always conscious of this status and often playfully or transgressively disguises in its movement of the you-referent and narrative voice across the terms of the nexus illustrated above (Richardson, 1991: 314). The "you" of the "standard" form by and large simultaneously designates the narrator and narratee--in the way of self-address, as Jay McInerney's Bright Lights, Big City (1986) and Fuentes's Aura (1975)--but nevertheless remains inherently unstable and subject to Bonheim's "referential slither," constantly threatening to merge "with another character, with the reader, or even with another grammatical person" (312). That is, the "you" may at any time turn away from designating its protagonist-focaliser to merge with or address a reader outside the fictional world (whether actual or implied) or another character. It can even, as in Christa Wolf's Patterns of Childhood (1990), Lolo Houbein's Walk a Barefoot Road (1988), or Ding Xiaoqi's Maidenhome (1993), and even Faulkner's Absalom! Absalom! (1971), merge with another grammatical person, complexly blending "second-," "first-" and "third-person" narrative pronouns.

In the "autotelic" form, the focus becomes narrower: although the "you" may still intermittently refer back to the narrator, it will no longer do so systematically as self-address, as the "standard" form has the tendency to do. But having risen above diegetic level, the narrative voice swings its address across the character, narratee, and implied and actual reader, oscillating between "first-person"/internal and "third-person"/external points of view. Its defining characteristic is a "direct address to a 'you' that is at times the actual reader of the text and whose story is juxtaposed to, and can merge with, the characters of the fiction" (Richardson, 1991: 320). For instance, in McInerney's Bright Lights, Big City (1986), Richardson argues,

the reader knows that he or she is extradiegetic, outside the narrative, and only assumes identity with the main character as part of the act of play in which reading consists. Calvino's sophisticated strategy [in If On a Winter's Night a Traveller, alternatively,] is to catch you, the extradiegetic reader, off guard, and make you the subject of diegesis, thereby spiriting or abducting you into the narrative. (Kirby, cited in Richardson, 1991: 321)

In Richardson's "subjunctive" form, too, the narrator and narratee remain typically distinct, more firmly even than in the "autotelic," but this partial closing of the slide of the "you" is powerfully compensated for. Its tendencies towards imperative verb forms and the future tense, its equivocal specification of character, setting and chronology, its sense of generality as "a kind of enumeration and interrogation of the typical," and its enhanced allegorical quality (Richardson, 1991: 320), maintain the deep ambiguities of reference and sense enabled not only by the "second person's" irreducible oscillation of internal/external points of view, but also its indefatigable movement across the terms of that complex.

Another solution to the problem of categorising "second-person" modalities, already suggested by my description of the way the boundaries between the second and third forms in the addressee identity model readily collapse, is to think of the forms as resting along a linear continuum. This is the strategy Fludernik adopts in "Second Person Fiction" (1993) and "A Test Case for Narratology" (1994c) and which Irene Kacandes uses in "Narrative Apostrophe" (1993). Both approaches classify instances of pronoun-of-address fiction comparatively against other nodes of a spectrum rather than in terms of characteristics they might be thought of as having in and of themselves. Fludernik puts the argument that, in some measure or another, texts either do or don't figure a communicative exchange between the narrator figure and actants, so that, on the one hand, we have narratives that may be called communicative texts, and on the other, texts that may be called noncommunicative texts. The term "communicative" in noncommunicative and homo/heterocommunicative narrative, she writes, is intended to refer to the "communicative circuit between a narrator (or teller-figure in Stanzel's typology) and the immediate addressee or narratee who is at the receiving or interactive end of that communicational frame" (Fludernik, 1994c: 446). Her continuum insightfully marries elements of Stanzel's typological circle 20 and the hetero-/homodiegetic divisions of Genette's typology, producing a descriptive model that examines not so much the type of narrator as the relationship between narrative agents, a relationship that is necessarily anthropocentric in so far as it is constituted as an act of communication. The Genettian model, however, in determining the homo- or heterodiegetic status of a text, treats only the narrator. It cannot distinguish whether the "you" is or is not a character within the story. Fludernik argues that:

Homo- and heterodiegesis simply "tick off" actantial roles and their recurrence or non-recurrence on the narrational plane. Stanzel's schema, on the other hand, talks about a continuity in the realms of existence, stressing the first person narrator's existential involvement in the fictional world versus the third person narrator's aloofness from it. (Fludernik, 1993: 220)

She aligns her communicative and noncommunicative categories respectively with Stanzel's teller and reflector modes. That is, very generally, in the first set of texts some form of narrator or narrative voice will be present: teller modes figure somebody talking to somebody else. In reflector texts, the story is presented through the point of view of the characters: reflector modes are constituted by "flow of consciousness" narrative discourse. In her model, as in Stanzel's, "the communicative level . . . is logically constituted by the reader's construction of teller-narratee interaction on the basis of a series of data triggering a communicational 'frame' (somebody is talking to somebody else) . . . " (Fludernik, 1993: 447). Her model thus places noncommunicative texts at one pole and communicative texts at the other, dividing the communicative texts into heterocommunicative and homocommunicative narratives, and the homocommunicative again into homodiegetic and what she names homoconative narrative.

Homocommunicative texts share realms of identities between the personae on the communicative level and the fictional personae: that is to say, either the narrator or narratee or both are also characters in the fiction. Heterocommunicative texts, on the other hand, completely separate the realms of plot agents (characters) and interactants on the communicative level (narrators and narratees). (Fludernik, 1993: 447)

In the heterocommunicative "second-person" text, the narrator plays no actantial role in the narrative, and the role of the "you" is solely that of protagonist. In the homoconative narrative, the narrator again plays no actantial role, but the "you" is now both protagonist and narratee. That is, in Stanzel's terms, the narrative is one in which the narrator, even while telling the story to the characters/narratees, does not share realms of existence with them. It is a non-realistic situation that Genette's terminology is unequipped to describe: to class such instances of narrative-"you" solely as heterodiegetic is entirely to miss their point (Fludernik, 1994c: 446). 

Kacandes's approach again asks the questions implied by the addresser-addressee relations foregrounded by the addressee identity model, but, insightfully, she also draws on the "second person's" grammatical function as shifter. Roman Jakobson describes shifters as a particular set of words whose general meaning "cannot be defined without a reference to the message" (Jakobson, 1971: 131). The notion of the shifter as Jakobson and Benveniste develop it, drawing on Jesperson's original coinage, has its roots in the linguistic concept of deixis.

Deictic expressions are terms such as "here" and "over there" which point to features of the surrounding context. In that deictic terms act as pointers they are sometimes called "indexical expressions." The referents of deictic expressions are constantly shifting as the relationship between utterance and context changes. For example, within a conversation the person identified as "I" changes as speakers change (and in much more complex ways when a speaker quotes the talk of another . . . ). The term "shifter" is thus also used to refer to deictic expressions. (Duranti and Goodwin, 1992: 43)

Quoting directly from J. Lyons's Semantics (1977) and "Deixis and Subjectivity" (1982), Peter Jones gives this standard account of deixis as being based on the two suppositions that:

1. the basic function of deixis is to relate the entities and situations to which reference is made in language to the spatio-temporal zero-point - the here-and-now of the context of utterance;21

2. the zero point (the here-and-now) is "egocentric," in the sense that "the speaker, by virtue of being the speaker, casts himself in the role of ego and relates everything to his viewpoint. (Lyons, cited in Jones, 1995: 27-28)22

Jones traces the standard account back to Bühler's account, in which "the word now, like the words I and here, is positioned at the origo or some source of the coordinate system of subjective orientation and is therefore assumed to be egocentric" (Jones, 1995: 38). Jones argues that the standard account has absorbed this assumption entirely, illustrating this again through Lyons.

Egocentricity is temporal as well as spatial, since the role of speaker is being transferred from one participant to the other as the conversation proceeds, and the participants may move around as they are conversing: the spatiotemporal zero-point (the here-and-now) is determined by the place of the speaker at the moment of utterance. (Lyons, cited in Jones, 1995: 38).23

In taking up the notion of the shifter as a central term, Kacandes's method asks not only "who speaks" and "who/what is addressed," but crucially: "Can the 'you' respond?" (Kacandes, 1993: 69). That is, can the "I" and the "you" implicit in any "second-person" utterance exchange places? Kacandes suggests that what these questions produce is not a range of discrete units, but a spectrum of reversibility spanning from statements like, "You can lead a horse to water but you can't make it drink," to "You can take that horse to the water trough, Fred." These mark, respectively, an apostrophic pole, in which (any linguistic markers of dialogue notwithstanding) communication can flow only in one direction, and a dialogic pole, in which total reversibility is possible, although not necessarily realised (69).

If we select as a main criterion the ability of an addresser and an addressee to actually exchange positions, in other words, to have a conversation, we could create a spectrum where direct address such as we know it from quoted dialogue [e.g., Nicholson Barker's Vox (1992)] would be on one end and what I suggest we call radical narrative apostrophe . . . would be on the other. (Kacandes, 1993: 31)

In radical narrative apostrophe, Kacandes says, the identity of the "you" is never stable and the source of the narrative discourse is always inexplicable: who is it that speaks the "you," and who to? This is one of the reasons she describes stories like Barthelme's "Moon Deluxe" (1983) as the most unsettling forms of the "second person:" the relationships are never, finally, clear. Beyond this is the use of "you" as a generalised "everyone" or abstract "someone," as in Hawkes's The Lime Twig (1961), and beyond that again is the use of "you" for "one," where "no dialogue is possible because the address is to an abstract or generalised construct, not a specific person," in which case no human addressee is constituted as a respondent (Kacandes, 1993: 69-70). Her argument for the strong effect of address in radical narrative apostrophe, however, implies that the effect does not derive entirely from its position on this continuum. Standing adjacent to generic-"you," its sense of address should be weak, and, indeed, argues Kacandes, its dialogism is even further arrested by the "bizarre" situation of the speaker's effacement (Kacandes, 1993: n1 68). Not only can the addressee not respond, but the interlocutory and ontological status of the addresser is also put to question: who, where and even what is "the speaker"? The effect of address experienced by readers of radical narrative apostrophe, argues Kacandes, derives not only from the possibility of address coded by the "second-person" pronoun, but from the way in which apostrophe itself necessarily assumes a "third term"the onlooker, eavesdropper, witness, theatre patron, reader, who is assumed by the utterance. Incapable of responding (or if doing so, responding inappropriately), the "third term's" absence would make the classical apostrophic gesture of directly addressing an inanimate object or absent person pointless: as a speaking for no-one (Culler, 1981: 141). Kacandes writes that:

The apostrophic circuit can be described as address to a specific other which is witnessed by an audience, with the effect that the later, although not directly addressed, is moved in some profound way by the verbal gesture. Apostrophe provides me with a model for the double reaction that many uses of the second person in fiction trigger in actual readers: on the one hand, a sense of being moved and on the other, the recognition that the address is to another. In apostrophe there is always a gap in the communicative circuit; one is never addressed directly. (Kacandes, 1993: 67-68)

Kacandes's need to go outside the logic of her continuum to explain radical narrative apostrophe's anomalous effect of address - however cogently she accounts for it - is significant. These effects are not to be contained by a linear model. As immediately appealing and useful as her notion of a continuum of reciprocity is as a way of characterising "second-person" utterances, I will suggest in Chapter 3 that taking reciprocity as a constitutive term entrenches analysis in a dyadic communication model whose points of reference are ultimately limiting.

3. Privatised Selfhood and the Anthropocentric Imperative

The critic must remain mindful that each of the models described above depends upon metaphors of "person." They do not so much offer an analytical perspective as represent the ways in which conventional practice administers the hegemony of what I will call Cartesianism by inviting (even necessitating) us to read such narratives in terms of who speaks/sees/overhears. Wojciech Kalaga, in discussing his concern over poststructuralist criticism's response to the notion of the implied reader - an central figure in many discussions of pronoun-of-address fiction - usefully describes Cartesianism as a principle which

although discarded now by poststructuralism, has nevertheless saturated the general way of thinking and has thus not only determined the popular view of the mind and its relation to reality, but has also become a comfortable and comforting axiom: the axiom of the primacy and autonomy of the human mind with respect to the external world. When I say "Cartesian," I use the adjective as an emblem covering all sorts of conceptions according to which the mind is seen as an isolated monad which can reflect the external world, but which could exist independently of that world. (Kalaga, 1990: 145)

This is the strict sense in which I mean Cartesianism to be taken. Rather than bound by the narrower historical, philosophical context of Descartes's Meditations, it is to be understood as epistemic, that is, as a concept that transcends the historical moment of its namesake. Descartes's meditations on the self, condensed into the enunciation cogito ergo sum, are emblematic - or symptomatic - of the general notion of the privatised self being addressed here, and it is this crucial enunciation that I will examine in depth in the following chapter. Other fundamental elements of Descartes's philosophy, however, such as the division of the mind and body, are to that extent beside the point. Husserlian phenomenology, for instance, denies that the mind and body are split as Descartes argues, but it too, I will argue later, acquiesces to the hegemony of Cartesianism. 24 What the notion of Cartesianism directly addresses is a framework, a superstructure, that conceives of cognition as

an unmediated, direct, and frequently an intuitive process. A consciousness, a self is confronted with a universe of objects, of texts . . . , which it cognises, but with respect to which it is antecedent and primary.

The Cartesian concept of the subject is a comforting concept because it gives us self-confidence, it draws clear-cut contours of the external world and gives us a sense of power over that world. (Kalaga, 1990: 145)

Each of the models described so far depends upon the hegemony of this Cartesianism. Each, that is, constitutes an invitation to identify (with) clearly delineated roles of anthropomorphic speakers and listeners, powerfully working to contain meaning by prefiguring generic (that is, highly conventionalised) relations between story-tellers and audiences. They are models that originate in "a certain prejudice," as Kalaga remarks, borrowing the notion of "prejudice" and prejudicial thinking from Descartes's own meditations. This prejudice, writes Kalaga, is the dominant Western notion of the "self." Discussing Charles S. Peirce's criticism of Descartes, Walter Benn Michaels phrases it this way: "For Descartes the self is neutral"unprejudiced," it begins by believing in nothing except its own existence; for Peirce, the self is always committed - it cannot begin by calling into question all its beliefs, for these beliefs, these "prejudices," are "things that it does not occur to us can be questioned" (Michaels, 1980: 194).

Cartesianism, I will argue, runs to the very heart of prevailing practices of reading and of the theories of narrative that apply themselves to identifying and explicating these practices. It is realised most strongly in literary texts within anthropocentric modes of reading. As Uri Margolin points out in respect to narratological analysis, there is a widely held, underlying presupposition "that individual subjects, as narrators and as participants in the recounted states or events, can be recovered (recuperated) from the linguistic signs of the text" (Margolin, 1986-87: 183). In the context of critical analysis, this need not be an inappropriate supposition. As Fludernik proposes, since a majority of literary narrative texts operate on apparently realistic grounds - grounds that doggedly propagate anthropocentric notions of character, narrator and narratee - it is important to read and analyse them "within the framework of the mimetic presuppositions that they invoke" (Fludernik, 1994c: 457). But it is also a presupposition that for the most part is entirely naturalised and transparent. The issue is not that subjects can be recuperated from textual inscription, but the particular nature of the subject recuperated - expressly, Cartesianism's subject.

It needs to be observed immediately, of course, that very many critics would freely own to the Cartesianism of their arguments and to holding the doctrine of the unified self. The influential transactional theorist Louise Rosenblatt is a case in point - a case all the more significant for the inroads being made by transactional theory into literary pedagogy. 25 Rosenblatt makes her position explicit when, towards the end of The Reader, the Text, the Poem (1978), she describes structuralist (and presumably poststructuralist) approaches as articulating a refusal of the older categories of human nature. What is to be rejected in their different conception of the human being is their anti-humanism. That is, what is to be rejected is their assumption "that the individual consciousness is somehow a kind of construction, something to be seen as merely a collection or intersection of patterned forces, social and natural" (Rosenblatt, 1978: 172). The Transactional view

reveals the individual consciousness as a continuing self-ordering, self-creating process, shaped by and shaping a network of interrelationships with its environing social and natural matrix. Out of such transactions flowers the author's text, an utterance awaiting the readers whose participation will consummate the speech act. By means of texts, we say, the individual may share in the funded knowledge and wisdom of our culture. (Rosenblatt, 1978: 172-73)

Rosenblatt argues that the literary work of art constitutes "an important kind of transaction with the environment," with the world we experience, "precisely because it permits self-aware acts of consciousness" (Rosenblatt, 1978: 173). Terry Eagleton argues, however, that as a consequence, one unavoidable implication of transactional criticism, and of reader-response criticism generally, is that "what you get out of the work will depend in large measure on what you put into it in the first place, and there is little room here for any deep-seated 'challenge' to the reader" (Eagleton, 1983: 81). Indeed, as he writes:

Everything about the reading subject is up for question in the act of reading, except what kind of (liberal) subject it is: these ideological limits can be in no way criticised, for then the whole model would collapse. In this sense, the plurality and open-endedness of the process of reading are permissible because they presuppose a certain kind of closed unity which always remains in place: the unity of the reading subject, which is violated and transgressed only to be returned more fully to itself. . . . [W]e can foray out into foreign territory because we are always secretly at home. (Eagleton, 1983: 79-80)

Eagleton argues that what surreptitiously underlies the apparent open-endedness of response theories are doctrines of the unified self and the closed text (Eagleton, 1983: 80). Indeed, the will towards the closure of a reading of a text (if not also of the text itself) is intimately bound into a notion of the self that I have been characterising as Cartesianism. Rosenblatt is unequivocal in her declarations concerning the polysemous character of texts: "there is no one absolutely 'correct' meaning of a text" (Rosenblatt, 1988: 13). She resiles, however, from the purported extreme relativism of deconstructionism, and even of other reader-response critics such as Stanley Fish who are accused of proposing that every single reading of a text may be different and will be equally valid. Rosenblatt writes that:

The structuralists, the formalists, and others are entitled to choose whatever type of descriptive system they wish to use, and to focus on any level of analysis, but the essence of a work of art is precisely that a consciousness is living through a synthesising evocation from a text which involves many - one is tempted to say all - levels of the organism. (Rosenblatt, 1978: 173)

In order to explain the processes by which texts motivate meaning, Rosenblatt discusses the concept of selective attention, in which "what is brought into awareness, [and] what is pushed into the background or suppressed, depends on where the attention is focussed" (4).

When we see a set of marks on a page that we believe can be made into verbal signs (ie., can be seen as a text), we assume that it should give rise to some kind of more or less coherent meaning. We bring our funded experience to bear. Multiple inner alternatives resonate to the words as they fall into phrases and sentences. . . . If the marks on the page evoke elements that cannot be assimilated into the emerging synthesis, the guiding principle or framework is revised: if necessary, it is discarded and a complete rereading occurs. . . . Finally, a synthesis or organisation, more or less coherent and complete, emerges, the result of a to-and-fro interplay between reader and text. (Rosenblatt, 1988: 11)

Transactional reading - and this is true of many strands of response theory - is a continual process of closing down meaning as much as it is a process of opening it up. To the extent that I agree that readers seek "some kind of more or less coherent meaning" in literary texts, I would insist that this behaviour - described by Rosenblatt herself as normative - be understood in the context of the arguments taken up in Chapter 3 regarding the conventional practices of constraining meaning-making. Most crucially, in the process of unifying texts, readers find images of their own (putatively) unified worlds and selves. This is the point Eagleton makes forcefully when observing that what lies behind Rosenblatt's process of seeking a "more or less coherent meaning," is in fact, "the influence of Gestalt psychology, with its concern to integrate discrete perceptions into an intelligible whole. It is true that this prejudice runs so deep in modern critics that it is difficult to see it as just that - a doctrinal predilection, which is no less arguable and contentious than any other" (Eagleton, 1983: 81). The return of the doctrine of the closed text throws us back into the doctrine of the unified subject. Eagleton is speaking here, in fact, about Wolfgang Iser's formulations of response theory, in which "different readers are free to actualise the work in different ways, and there is no single correct interpretation which will exhaust its semantic potential. But this generosity is qualified by one rigorous instruction: the reader must construct the text so as to render it internally consistent" (Eagleton, 1983: 81). The similarities of this to Rosenblatt's formulation are clear, even though she allows that these internal coherences need only be "more or less" complete. Where Iser speaks openly of reducing a text's polysemantic potential to some kind of order, Rosenblatt speaks about arriving at a coherent synthesis or organisation of a text's potentials. Eagleton's point, as true of Rosenblatt as it is of Iser, is that: "Unless this is done, the unified reading subject will be jeopardised, rendered incapable of returning to itself as a well-balanced entity in the 'self-correcting' therapy of reading" (Eagleton, 1983: 82). And it is this, very clearly, as much as the anarchic threat of endless readings, that Rosenblatt amongst others will not tolerate.

Many critics, on the other hand, do recognise the desirability of challenging Cartesianism's modes of reading for the insights into literary textuality that may emerge. Julie Solomon's discussion in "Fictional Questions: Illocutionary Force in Literary Communication" (1990), for instance, takes the notion of the narratee as it is originally discussed by Gerald Prince (1971) and developed by Mary Anne Piwowarczyk (1976) 26 with the intention of recasting it in less anthropocentric terms. Her hope is that this might broaden its usefulness to criticism, particularly speech-act criticism, by turning the narratee into a much more neutrally formalist category (Solomon, 1990: 89). Mary Louise Pratt's description of the narratee states its anthropocentrism succinctly:

In contrast with [criticism's] usual treatment of the encoded reader as a highly abstract, loosely defined function or position or cluster of suppositions, one is struck in Prince's approach by an insistence on construing the narratee as a full-fledged fictional person, with a genuine discernible personality whether or not "he" is actually named or referred to in the text. [ . . . ] Repeatedly we are referred to the task of constructing a "portrait" of the narratee. . . . These turns of phrase are not in themselves intended to be taken very seriously, but they nevertheless invoke a whole set of values of personality, individuality and privatised selfhood that informs all of Prince's analysis. . . . (Pratt, 1982: 211)27

Solomon's proposition is that analysis of the narratee might be extended beyond its use as an element by which the narrator might be more fully revealed, to an analysis of the narrative discourse itself as rhetoric. Arguing that the narratee might be approached in terms of literary speech-act theory, she proposes that: "Who is speaking to whom becomes subordinated to the question of how they are speaking to each other and why. . . . " (Solomon, 1990: 89). I would argue, however, that this does not remove the model of the narratee very far beyond Prince's original anthropomorphic definition. In Solomon's own words: "Looking at illocutionary force doesn't just reveal what sort of person the narrator or narratee is, but it also shows what they are doing, or think they are doing or trying to do in speaking: the communicative project(s) of the text" (Solomon, 1990: 89). Designated by pronouns, the narratee unapologetically remains a "person." Even if the qualities of "person-ality" were to be systematically withheld in any critical description, the application of notions of illocutionary force ensures that the narratee is still seen to do, to think, to speak, to listen, that is, is seen as Cartesianism's privatised self. Solomon has not, then, finally escaped the centripetal pressure to return to Cartesianism's centre.

An alternative approach is to recognise the pressures placed on the interpretation of narrative texts by anthropocentrism as fundamental rather than incidental, and incorporate this perception within the very definition of narrative. In Towards a "Natural" Narratology (1996), for example, Fludernik argues that what lies at the base of narration is not "plot" or "story" per se, 28 but more fundamentally, "experientiality of an anthropomorphic nature" (Fludernik, 1996: 26). Fludernik forces anthropomorphism to the foreground by "disqualify[ing] the criteria of mere sequentiality and logical connectedness from playing the central role" in defining narrative in favour of a notion that holds that "action belongs to narrative as a consequence of the fact that experience is imagined as typically human and therefore involves the presence of existents who act" (Fludernik, 1996: 26-17). In a sense, Fludernik is doing nothing more here than observing what many critics, including Forster himself, have asserted before. As Forster declares, narrative is always, at some level, about people.

Since the novelist is himself a human being, there is an affinity between his subject matter which is absent in many other forms of art. . . . The painter and the sculptor need not be linked: that is to say, they need not represent human beings unless they wish, no more need the poet, while the musician cannot represent them even if he wishes, without the help of a programme. The novelist, unlike many of his colleagues, makes up a number of word-masses roughly describing himself. . . , gives them names and sex, assigns them plausible gestures, and causes them to speak by use of inverted commas, and perhaps to behave consistently. These word-masses are his characters. (Forster, 1976: 54-55)

By and large, Forster's frame of reference for the novel is unapologetically realist. Fludernik's argument, on the other hand, while maintaining anthropomorphic character as a principal constituent of narrative, tries to disengage the conventional, uninterrogated correspondence between anthropomorphism and realism to position the realist text within larger systems. In "A Test Case for Narratology," she makes the observation that "[b]roadly speaking, narrative theories . . . rely on narrative parameters that reflect typically realist assumptions about narrative" (Fludernik, 1994c: 456-57). In this she singles out the work of Genette, Stanzel, Chatman, Bal, Prince, Rimmon-Kenan, Bonheim, and Lanser. But as she asserts:

This is not to say that these theories cannot deal with departures from the realist mode; on the contrary, fantastic and postmodernist kinds of writing are easily categorised in terms of infractions of the narratological rules of selection and combination. The realist presuppositions specifically touch on prototypical ways of story telling: few and far between are the theories that allow for narratives without a narrator; rare the theory that accommodates noncanonical types of story telling (in the present tense, simultaneous narration, second-person narrative) on a par with canonical first- or third-person past-tense texts. (Fludernik, 1994c: 457)

To identify a critic as operating within a realistic framework, she insists however, is not necessarily a point of criticism (Fludernik, 1994c: 457). She proposes that:

The solution, theoretically at least, does not lie in the complete abandoning of realist parameters and categories but in the integration of realist parameters within an encompassing theoretical model that can treat the realist case as the special instance of prototypical story telling without, at the same time, promoting the realist mode to a position of theoretical centrality. (Fludernik, 1994c: 457)29

Fludernik's own model, described earlier, is unabashedly, though also very mindfully, anthropocentric. She makes anthropocentrism a conspicuous constituent of the literary object itself, robbing the hegemony of Cartesianism of its transparency and privilege. The value of her model, I suggest, resides in its endeavour to disengage assumptions about realism from anthropomorphism, and so resides, too, in its implicit identification of an anthropomorphic imperative as being even more fundamental to the way we read narrative than realism itself. As I will explore in Chapter 2, however, certain consequences and problems for literary criticism and theory follow directly from this: what is the critic-reader to presume about the nature of the anthropocentric self?

4. An Experimental Study of Apparent Behaviour

Whatever is to be assumed about the nature of the social subject, the powerful inclination to see, to imagine, the forms of agents within narrative textuality and to assign these forms anthropomorphic and even psychologistic qualities seems deeply ingrained. Indeed, the tendency may be fundamental to the cognitive activity of making sense of our experience of texts - fundamental, at least, to that "certain species of life" that is Cartesianism's constant self. The vigour of this tendency to take up anthropocentric modes of interpretation has been strikingly illustrated.

In 1944, Fritz Heider and Marianne Simmel published an analysis of responses to a rudimentary, two-and-a-half-minute animation in which a large triangle, a small triangle and a circle are shown moving in various directions and speeds around a field. The only other figure in the field, as illustrated below, is "a rectangle, a section of which could be opened and closed as a door is" (Heider and Simmel 244).

figures 5 and 6

Figure 5 illustrates Heider and Simmel's description of the animation's opening frame.
Figure 6 shows Heider and Simmel's illustration as it appears in their article.

To open a reconstruction of Heider and Simmel's animation, open this link: PLAY ANIMATION    

Heider and Simmel's objective in "An Experimental Study of Apparent Behaviour" (1944) is to demonstrate the usefulness of such an endlessly variable animation as a tool in the investigation of "the perception of the behaviour of persons" (Heider and Simmel, 1944: 251). They also propose, more tentatively, that the method may help explore our habit of attributing causality to behaviour and of arranging behaviour into connected sequences - specifically into coherent, causally motivated stories. Given that their object is "the behaviour of persons," Heider and Simmel allow themselves the assumption that their subjects will readily interpret the changes in the field - realised in the mind of the viewers as the movement of shapes"in terms of actions of animated beings" (Heider and Simmel, 1944: 259). For instance, virtually all their subjects regard the rectangle's moving segment as a door, and do so, Heider and Simmel argue, because it moves only when T, t or c are in contact with it. They also assume that these actions will be interpreted in terms of logically connected sequences of behaviour that are likely to be unified into a coherent story. Indeed, they explicitly direct two of their groups to do exactly this - to interpret the three shapes as persons and to tell the story of what they see - having first tested their assumption against a "blind" control group. Only three of their subjects, one in Group I and two in Group III, describe the figures "almost entirely in geometric terms" (Heider and Simmel, 1944: 246). Of these, Heider and Simmel deal solely with the lone heretic from Group I, summarily dismissing the two from Group III as being unable to follow instructions and, by inference, as providing invalid responses. Members of Group II - which Heider and Simmel designate their main sample - were provided with considerable direction. They were "instructed to interpret the movements of the figures as actions of persons" and were provided with a questionnaire of nine leading questions to be opened after the animation was viewed. The members of the third sample were instructed likewise, though were given an abridged questionnaire and were shown the animation in reverse. 30

Heider and Simmel themselves describe the animation in highly anthropocentric and storied terms, though they readily acknowledge this and excuse themselves by arguing that to describe the animation "in purely geometrical terms would be too complicated and too difficult to understand" (Heider and Simmel, 1944: 245). They find it much more concise, for instance, to say that the large triangle "moves towards the house, opens door, moves into the house and closes door" than to attempt a necessarily longer, more "objective" description that details trajectories of movement and contact between shapes and the pivoting out of a segment of the rectangle. But Heider and Simmel's concision is achieved at a cost. Words such as "house" and "door," and other words that appear in their description such as "fight," "wins" and "chases," all derive their sense from human-centred frames of reference, and so in turn bind their scenario into an inescapable anthropocentrism. The apparent clarity of the words, "the big triangle opens the door, moves into the house and closes the door," comes not from the precision of their reference, but paradoxically through the activity of heavily overdetermined, polysemic signs rapidly conveying meaning, including multiple, prototypal scenarios (e.g., what opening a door looks like) and models of interpersonal relation (e.g., what it means to be larger than another entity, and vice-versa). It might also be observed at this point that it would be inexact to attribute Heider and Simmel with authorship - in any but the most remote and inadequate sense - of the stories recorded by the viewers of the animation. Although Heider and Simmel as "producers" of the text are clearly mindful of the cinematic effect they wish to produce and have therefore made considered choices regarding the placement of the geometric shapes from frame to frame, they have not thereby made the animation coherent. They can do no more in this respect than lend it the mannerisms and semblances of structure, doing so by drawing upon the same semiotic practices as their audiences are likely to engage when viewing it. The final organisation of the structure in each reading/viewing, and more crucially the meanings drawn out of the text, are demonstrably not Heider and Simmel's to control.

It is Heider and Simmel's control group that is of most interest to this discussion. To establish the validity of their assumption that the animation's viewers would indeed interpret the shapes as "persons," Group I were told simply to "write down what happened in the picture" (Heider and Simmel, 1944: 245). This simple instruction, it should be noted, is itself far from neutral. It does guide the viewer on how to respond to the images on the projection screen by inviting them to deem that "something happens." That is, it instructs them to infer some kind of agency, and to construe that agency in terms of an event, obliquely but inevitably invoking "storiedness" as an interpretive frame. The overwhelming uniformity of response in this respect, however, seems conclusively to support Heider and Simmel's assumption: as observed, all but one of the control group's thirty-four subjects respond in kind, regarding the movements in the field as "the actions of animate beings, in most cases of persons; in two cases of birds" (Heider and Simmel, 1944: 246). Moreover, a little over half of the sample "reported a connected story" (Heider and Simmel, 1944: 246). But what is conclusive in the responses Heider and Simmel report from all three sample groups is, firstly, the readiness of the interpreters to attribute anthropomorphic qualities to textual inscriptions - in this instance geometric shapes - and secondly, the tendency to organise around these inscriptions the most elaborate narratives and motivations. T is aggressive, warlike, belligerent; or strong and resourceful; or a good business man; or stupid and confused; even unattractive and defensive. t is valiant, independent, resentful of being bullied; is timid; is sly, wary; evil; intellectual; devoted; attractive; even a coward. c is fearful, cowardly, meek; helpless and dependent; shrewd; courageous and resistant; teasing, very refined; even opportunistic. T, t and c may even become allegorical, not leaving anthropocentric interpretation behind, but carrying it to a new level, so to speak. For instance, screened in reverse, the animation may tell a story in which:

Man (T) finds himself in chaos, which finally resolves into a sort of cell representing Fate. He is able to free himself (but only temporarily), when Woman (c) accompanied by Evil (t) comes upon him, and disrupts his momentary peace. He feels called upon to rescue her, but Evil imprisons them both by Fate, from which Man escapes, leaving the woman there for safe-keeping. He at first seems to vanquish Evil, but Woman comes into the picture again and again disrupts Man. She goes off with Evil, as he seems the winner of the struggle, and Man, not understanding her, himself, or anything, resigns himself to Fate. (Heider and Simmel, 1944: 252)

As an illustration of Margolin's remark on the supposition that individual subjects as narrators and participants in a narrative's recounted events can be recovered from the linguistic signs of the text (Margolin, 1986-87: 183), Heider and Simmel's "An Experimental Study of Apparent Behaviour" does run up against a limit - but a limit of a very particular and telling kind. No "narrator" as such might be traced in Heider and Simmel's animation. Many film theorists argue for the general inapplicability of the notion of "the narrator" to film texts. Put crudely, the argument is that the film text involves acts of Aristotelian mimesis rather Aristotelian diegesis: film shows rather than tells. 31 And manifestly, whereas in respect to literary texts one can point to discourse that seems uttered by whatever voice "speaks" the text, since the reader will have before him or her the discourse's material form - the permanent record on the page of the narrator's very words - one generally cannot point to such discourse within film texts, and so it becomes harder for film theorists to posit a voice that "speaks" the text. This specificity of the medium aside, the animation is taken by most of its viewers as involving "individual subjects," T, t and c. And crucially in terms of defining the animation as a story, as narrative, these subjects are taken as participants in events that are recounted - if not recounted by a narrator, then certainly, at least, within a narration. Taken by more than half of the participants in the minimally directed Group I and virtually all of the participants in Groups II and III as "telling the story" of what three people did when they met on that occasion, Heider and Simmel's animation deftly illustrates just how willing we are to see in the world around us what we take to be our own forms and behaviours - but forms of subjectivity and modes of behaviour, as yet, defined by Cartesianism.

5. The "Second Person" as Rhetoric and as Address

Schemes of classification as Richardson's, Kacandes's and Fludernik's, no matter how well conceived, can only do so much for analysis and criticism: there remains the much broader question of what pronoun of address modes of narrative can do. The effects of "second-person" narrative on the reader, particularly of the mode that I will call Protean-"you," can be very striking - at once engaging and alienating. The very "shiftiness" of the "second-person" pronoun, it seems, must play no small role in this. In "A Nowhere for Us: The Promising Pronouns in Cortázar's 'Utopian' Stories" (1986), Doris Sommer recounts how "[i]t used to make Cortázar rather nervous to think about pronouns, those 'shifty' terms that, linguists tell us, are the least stable of signifiers" (Sommer, 1986: 231), a nervousness produced, she writes, for "the breakdown of Platonic assumptions about meaning and presence lying beyond words" (Sommer, 1986: 250).

Morrissette cogently argues that the affective core of "second person" modality is constituted by its rhetorical strategies. Its function of address, which it rarely relinquishes, no matter how dispersed or generalised, works variously to generalise, universalise, typify, moralise, axiomise, coax, goad, question, interrogate, evaluate, judge, reproach, challenge, exhort, mock, accuse, admonish, advise, instruct, lecture and so on. Gainfully turning the methodological tools of speech-act theory to the analysis of "second-person" narrative, Kacandes adds to these an array of specifically linguistic rhetorical/grammatical functions: the conative, referential, imperative, vocative, appellative, phatic, performative and so on. As she suggests, however, even while narrative "you" certainly exhibits tendencies towards particular functions, such lists might only ever be incomplete because it is likely that any particular function of language can come to dominate any "second-person" passage, a situation that immediately accounts for the vast array and fluidity of tonalities exhibited by "second-person" texts.

The reason the "you" rarely relinquishes its sense of address can be explained in part by observations made by Ann Banfield in Unspeakable Sentences (1982)--and glossed by McHale in his article "You Used to Know What These Words Mean" (1985). Banfield suggests that the prime marker of a communicative act between an addresser and an addressee is not the "first-person" pronoun, but the "second," because whereas "I" does not necessarily imply "you," "you" does imply "I." McHale, in turn, points out that wherever a "you" does appear in an utterance, readers are justified in trying to isolate an addresser and an addressee, and in trying to identify the nature of the communicative circuit that joins them, reading the utterance in terms of the particular textual entities and relations of power, status, and psychology subsequently assumed to be involved (McHale, 1985: 95). "You," he says, "is a sign of dialogue, conveying some vocative appeal, some sense of address, even in its most 'innocent,' impersonal instances, and we cannot help but respond dialogically to it in some measure" (McHale, 1985: 112). It is to the "second person's" dialogic quality, incidentally, that Morrissette, Kacandes and others attribute one of its most striking rhetorical effects, its didacticism. Dialogue and its direct, inescapable, dialectical exchange, it appears, possesses a pedagogical/didactic function par excellence. Plato's Socratic dialogues attest to the "second person" pronoun's pedagogical value. Longinus, too, writing in the first-century AD, not only notes its use by the didactic poet Aratus (Longinus, 1965: 135), but deploys it extensively himself. The modern student of literature reads Longinus's On the Sublime (1965) over the addressee's shoulder, as it were, so that while Postumius Terentianus is both named and directly addressed with the "second-person" pronoun throughout, other recipients do not feel excluded - addressed, if not by Longinus, then certainly by the lesson itself. The reader of the much more generalised "you" of "The Apology"a text noteworthy for its use of a courtroom-"you"is similarly involved in the world of the text, not so much to read over the shoulders of the (unnamed, uncharacterised) addressees, but to look over their shoulders, involving him or herself as a third, silent but attentive and welcome participant in a (dramatic) dialogue.

Like McHale, Kacandes points to linguistic functions in an attempt to explain why the "you" clings to its sense of address and dialogue, though for her the work is done by the appellative rather than the vocative. Indeed, it may be that the vocative of itself is insufficient to carry the weight of address in particular forms of "second-person" narrative. These are the radical apostrophic narratives at the extreme of her continuum in which the identity of the "you" is least stable: in which the vocative falters because the "you" pronoun can no longer be relied upon to single out the addressee. The appellative, however, works not to designate addressee, but to attract attention - which it does much in the manner of the hailing gesture of a man-in-the-street (or of Louis Althusser's policeman calling, "hey, you there").

Michel Leiris puts it more eloquently in his introduction to the 18/10 edition of La Modification, a critical piece that uncannily presages the opening of Calvino's novel. Leiris writes:

You have between your hands a brand new copy of La Modification, a novel by Michel Butor. You leaf through this book, and in the course of this slight effort, you read some paragraphs at random. What is it that strikes you from the outset? [ . . . ] [It] is yourself, reader, that the writer seems to politely challenge and it only takes a few brief glances over the printed lines while you handle the paper-knife in order to sense that you are in the presence of an invitation, or else a summons/warning [sommation]. (Leiris, 1975: 287). 32

As can be argued of the narratee generally, one of the "second-person's" roles is the positioning of the reader in relation to the story being told, the text implicitly and sometimes explicitly constitution of the reader as a particular "type" of reader, that is, instructing him or her on what attitudes (thorough-going unanimity, sympathy; credulity; skepticism; judgment; affront; and so on) should be taken to the characters, events, and narration. The notion of the "second person" as sommation, however, as carrying appellative force - indeed, the act of "second-person" narration itself - would also seem to beg discussion of the constitution of the reader explicitly in terms of the Althusserian theory of interpellation. Yet this is an area that theorists and critics have skirted, addressing it by and large only in the vaguest terms. William Waters, for instance, in talking about the openness to the reader of the variety of "you" whose referent is not conclusively fixed, writes that:

"you" tends to hail; it calls everyone and everything by their inmost name. The second person pronoun is address itself. One can read unidentified "I" or "she" with comparatively small concern, but the summons of unidentified "you" restlessly tugs at us, begging identification. Thus a centrally important question . . . is our "tilt" of reading when confronted by such a call. How will we stand? (Waters, 1996: 130)

The beguilingly candid nature of the "second person's" hailing gesture, of course, is deceptive: highly complex processes of ideological interpellation work at far deeper levels than that of explicit address. Althusser's own illustration of the hailing policeman is intended as metaphor. Indeed, by and large, and in spite of the "second person's" promise of unconventionality, the ideological work done by a "second-person" fiction text (of constituting concrete, ideologically inscribed subjects) will be no more or less hegemonic or resistant than the work done by any other fiction text. Nevertheless, "second-person" fiction does appear to involve an initial, immediate appeal or call to the reader above and beyond that manifested in the discourse of conventional fiction. This immediate or surface interpellative gesture is most clearly evident at moments in which the reader is addressed - or more properly, moments at which the reader recognises him or herself in, identifies with, the (implied) reader constituted by that address. But, as I have made clear, in "second-person" narrative an utterance need not be explicitly addressed to the reader for him or her to experience the utterance as address. 33 Some texts, in fact, make effective use of an initial equivocation in this respect - as in the opening of If on a Winter's Night a Traveller, in which an initially ambiguous "you" resolves into a "you"-protagonist. As a gesture of interpellation, this moment has much in common with Kaja Silverman's description of the "particularly lucid example of textual hailing" that opens Frank Capra's film It's a Wonderful Life (Silverman, 1983: 49). Whether a text opens with the image of a town-limits signpost that announces, "You are in Bedford Falls," or a first line that states, "You are about to begin reading Italo Calvino's new novel," in the absence of a clear referent for the pronoun, that pronoun might easily be taken up by the viewer/reader. Indeed, before the fictional character arrives on the scene, the only way for the viewer/reader to make sense of the statement is to situate him or herself as the subject of speech, the "you" of the utterance, as the only subject available to fill it, so that "the rest of the sentence . . . organises itself around the viewer [or reader], locating him or her in the narrative space" in the moments before the protagonist has come to claim it (Silverman, 1983: 49). And having been called into that space, having been interpellated as that subject, Silverman argues, the reader/viewer permits his or her subjectivity to be carried forward by the figure of the protagonist (Silverman, 1983: 49-50)albeit that in the Calvino there will be more subsequent resistance to being carried in such a way than there will be in a work of classical Hollywood cinema.

The implications of Morrissette's formulation of the guide book mode bear attention in this context, too. Morrissette suggests that the generalising function inherent in this mode results in the writer of the guide book including him or herself in that same "you" address, so that the speaker is simultaneously the listener: and vice-versa. He suggests that, if this is the case, the engaged reader, identifying him or herself in the "you," adopts not so much the position of a tourist/protagonist, but that of the speaker, the narrator (Morrissette, 1965: 6). This, indeed, is the quintessential moment of interpellation: Althusser's theory "suggests that we are hailed or summoned by the ideologies which recruit us as their 'authors,' their essential subject" (Hall, 1985: 102). In La Modification, for instance, the voice that says "vous," writes Morrissette, "is less that of the character than of the author [read narrator] or, better still, that of a persona, invisible but powerfully present, who serves as the centre of consciousness in the novel" (Morrissette, 1965: 130). If the reader's identification falls to this effaced persona, to the "source" of the narrative discourse, rather than to the "second-person" pronoun's referent, Léon, then the reader is positioned as an interrogator and judge. It is as if the subject positions of Longinus's treatise, and of Plato's Socratic dialogues, too, have been inverted, so that - having alighted on the voyeuristic, moral-ethical high ground of the judgmental spectator - the engaged reader no longer stands behind the addressee, but at the shoulder of the addresser. Mariolina Salvatori (1986), however, reminds us of the ethical danger of such a position. For Salvatori, the "second-person" narrative pronoun in Calvino's If on a Winter's Night a Traveller (1982) is one element of a dynamic in which the novel's explicit inscription of the writer's authority and the reader's seeming autonomy functions as a form of flattering entrapment of the reader and therefore as a closure. The dynamics of "second-person" narrative can constitute a trap, not so much because we as readers "run the risk of slipping into the pronominal space that the you-Lettore occupies, but because, having been made to take an ironic, omniscient stance toward him, we might pass judgment on him without realising that we are passing judgment on ourselves" (Salvatori, 1986: 196). Indeed, what Salvatori warns against is that moment of interpellation in which we have become the "author" of the statement itself without realising its fullest implications.

6. Inscribing "You," the Reader

Not all critics, of course, accept that the essentially interpellative address function of the "second-person" pronoun remains equally active when the "you" becomes a protagonist (rather than a direct addressee), arguing that the "you" utterance must remain unparticularised and general for the reader to feel it as calling upon him or herself. It would be hasty, of course, to assume that the "you" address perforce leads the reader into identification, whether with the position of a "you" protagonist, the narrator, or the narratee/implied reader - just as it would be hasty to take its confronting address as always and inevitably alienating. And this is the ground that critics have staked out for the early rounds of the debate, that of whether or not, and why, any particular "you" address (experienced as singular and direct or as generalised and non-specific) draws actual readers into a narrative, facilitating or hindering identification. Bonnie Costello, for instance, contends that the direct address simply brings the reader into the text not so much as a witness overhearing a confession, but rather as a self-conscious "internal audience"much like the experience of the French nobleman at the theatre whose chair is set upon the stage itself and is thereby brought into the midst of the action. This inscription, however, Costello says, is always in some sense a closure. Inscribing an implied reader in this way "makes him as a fiction, a puppet to be acted on symbolically, with the hopes that this magic will penetrate to the heart of the actual reader" (Costello, 1982: 499). Given that "you" utterances remain in some degree reader address, as was argued earlier, this "magic" might be thought of as consisting of elements of the "you" identity dividing, one figure sitting attentive in front of the book, the other splitting off as a "you-other-than-ourselves," a re-imagined self that in many "you" narratives is simultaneously addressed along side us at any instant the "you" is uttered (Salvatori 169).

James Phelan makes a similar observation about the frequently equivocal overlap between character address and reader address, proposing that "what happens as we read 'You are unsure of how to react' is frequently an important dimension of reading second-person narration" (Phelan, 1994: 351). Readers of such narratives, Phelan argues, might very well find themselves occupying at one and the same time the position of observer - as the position familiar in reading "from which we watch characters think, move, talk, act" (Phelan, 1994: 351)and the position of addressee. He goes on to conclude that "the fuller the characterisation of the "you," the more aware actual readers will be of their differences from that "you," and thus, the more fully they will move into the observer role, and the less likely this role will overlap with the addressee position" (Phelan, 1994: 351). Katherine Passias has a similar argument. Any reader faced by a highly particularised "you" referent, she says, necessarily repudiates the utterance's illocutionary function in the self-knowledge that the particularities of the "you" (its personality, its acts, its physical context, and so on) do not describe "me" at all. At the point at which "a very personal experience" is described in such a way that the "second-person" pronoun becomes the referent of a character or actant within a narrative it loses all sense of its linguistic function of address and becomes a mere substitute for "I" (Passias, 1976: 199). Phelan is more circumspect, and appears to allow that an address function may yet leave its trace, observing it is sometimes "not easy to say who you are" (Phelan, 1994: 351). Nevertheless, the fact of greater characterisation of the "you" figure allows the reader to fall more easily into familiar and conventional reading practices - practices in which narrative pronouns are not ordinarily experienced as pronouns of address. "In other words, the greater the characterisation of the 'you,' the more like a standard protagonist the "you" becomes, and, consequently, the more actual readers can employ their standard strategies for reading narrative" (Phelan, 1994: 351).

Kacandes's position on the inscription of the reader is that it is not solely vocatives and appellatives that bring the reader into the text. Powerfully, this is also the function of performatives, of utterances "in which to read something is to do what one reads" (Kacandes, 1993: 65-66). Within "second-person" narrative this becomes "an action that the reader fulfills by reading a second-person statement" (Kacandes, 1993: 169n). The frequently-given example is the statement, "you are reading this," which, if read, is simultaneously performed. The statement, "you are dead," on the other hand, if read, is not performed (and if "performed," can hardly have been read). As the performative, Kacandes argues, the "you" utterance can be construed by a reader as being addressed not only, firstly, to the work's narratee, and secondly, to whatever encoded/implied reader might be posited, but also to the actual reader him/herself (Kacandes, 1993: 66). The statement, "you are reading this," is a felicitous statement, because indeed the reader of this article has performed the "you" utterance in the act of reading it. Kacandes suggests that part of the pleasure of reading the opening lines of If On a Winter's Night a Traveller arises precisely because of the tensions generated between the felicity and infelicity of performatives at the level of reading. In the first instance, as actual readers, "we felt the surprise of an unwilled performative," but if we consider the line a moment, we soon realise that the statement is not entirely felicitous (Kacandes, 1993: 169). The case is not that "[y]ou are about to begin reading Italo Calvino's new novel, If On a Winter's Night a Traveller"a beginning, incidentally, that the narrative substantially defers - because the actual reader has already begun Calvino's new book. "For the attentive reader, then, what this opening line accomplishes is both a seduction to feel addressed and a realisation that the call is not quite accurate . . . " (Kacandes, 1993: 170). But what also fails or becomes radically unstable in this infelicitous moment, as I have said, is the vocative function that inheres in any "second-person" utterance as an element of the allocutionary function, in that the reader is immediately unable conclusively to determine whether the "you" utterance refers to him/herself, to a character, or to some intermediary form of narratee. This does not, however, rob the performative of functionality. Rather, the very inability to fix the vocative's referent maintains a channel through which the reader, still admitting to the possibility of being the designated object, can felicitously perform the act referred to in the utterance: and having performed the act, as addressee, the reader is doubly inscribed.

7. A Point of View

Darlene Hantzis, too, offers a persuasive argument for the process through which readers are drawn into "second-person" texts. She provides one of the most thoroughgoing explorations of the ambiguous reader-character-narrator complex, albeit one that remains committed to the terms of narrative point of view - and, I will argue, to a deep seated if covert Cartesianism. Having installed the "second person" in a triumvirate of discrete modalities, however, she immediately stipulates that the "second person" point of view functions quite differently from the "first" and the "third." She retheorises Lois Oppenheim's (1980) suggestion that the "you" voice oscillates between self-address and omniscient commentary, ceaselessly moving the engaged reader between an internal and an external subject position. Oppenheim and Hantzis regard this oscillation as constituting an "intersubjectivity" that is always in process, always in the act of becoming. "Second-person" point of view, Hantzis claims, is a point of view that rejects traditional concepts of narrative subjectivity and authority, predicated as those are on the notion of a relatively stable and continuous, self-sufficient subject in control of its own discourse. A "second-person" narrator and point of view can produce no such privileged subjectivity, nor can it guarantee its own authority. The point of view of a "second-person" narrator, she says, "is ephemeral and un-author-ised outside the voicing of the text; it [only] persists so long as the narration persists" (Hantzis, 1992: 75). This statement, I suggest, raises some questions about assumptions regarding the status of "first-" and "third-person" points of view, and I will turn to these in my final chapters: in what sense can they be said to be any less ephemeral?

Hantzis proposes that "second-person" point of view comes into being only "when the second person pronoun simultaneously houses the narrator, actant, and narratee(s) in a narrative text"and she subsequently adds the reader and the author to this house (Hantzis, 1992: 44, 77). Its process is one in which the reader, against all conventional tendencies towards stability and exclusivity, "continually places her/himself in and continually displaces her/himself from the 'you' while simultaneously placing and displacing others in and from the 'you'" (Hantzis, 1992: 69). Like Richardson, Hantzis argues that it is this characteristic oscillation and ambiguity that differentiates the "second-person" point of view from types of displaced or disguised "first-" and "third-person" utterance that employ "you" in ways "easily comprehended by traditional dyadic theories of point of view" (Richardson, 1994: 311).

It has become a commonplace in literary studies to assume that character identity or "subjectivity" is conferred by narrative and textual practice - that it is manufactured in tissues of narrative statements. Terms such as actant, agent, and so forth have been coined to clarify our understanding of this narrative function. Ross Chambers further counsels us to keep in mind that the subjectivity of the narrator and narratee, likewise, is a textually produced chimera - and produced (as I will further discuss in my final chapter) by the very discourse they seemingly utter/respond to (Chambers 1989b: 37). To borrow from observations Roland Barthes makes about narrative "person" in The Rustle of Language (1986), a narrative's "person" will more helpfully be thought of as the support parsimoniously granted to a very mobile and ill-attached language, a support (or "slot") constantly posited and withdrawn that shifts from object to object as the mark of a (fictional) continuity/identity (Barthes, 1986: 181). For structuralist poetics - one of the contexts Hantzis establishes for her argument - a narrator is constituted as a subject by the positing of an Other that guarantees the narrator's own subject status, the narrator straightforwardly enough appearing to hail this Other through an address that presupposes that Other's identity (for instance, the identity of the ideal reader, assumed as a self-sufficient subject outside the text). This in fact is the tactic Pierre Gault (1978) ascribes to the narrator of The Lime Twig (1961) in Hawkes's deployment of narrative "you:" indeed, it describes, too, the interpellative action of the narrator-narratee exchange (between an absent-"I" and a projected "you") discussed earlier in relation to the opening moments of Calvino's If on a Winter's Night a Traveller and Capra's It's a Wonderful Life (in so far as the film can be said to have a "narrator"). The narrator or narrating instance, that is, is constituted as one "knowing" subject addressing another, similar "knowing" subject: but both, of course, are produced within the narrative discourse, one as the origin and anchor for this knowledge, the other as the appropriate (curious, desiring, needful) recipient and guarantor of such knowledge (Chambers, 1989b: 36). In the case of Hantzis's "second person" point of view, however, the experience of subjectivity is achieved not through one subject's hailing gesture toward another, but through the articulation of the "you" address itself, in such a way as to produce the narrator-character-narratee-reader complex as the site of a self-supporting intersubjectivity. The act of speaking "you," she states, mutually and simultaneously constitutes the subject status of the narrative entities. The "second-person" point of view's new vision of the world and one's place in it, then, depending as it does on "old habits," is what a social subject (whatever might constitute it) comes to and experiences through the "second-person" point of view - is what a reader, a prior identity, comes to and experiences as an epiphany.

Critical and analytical literature about the "second person" attests that the "second person" can at times effect the reader to produce a deeply unsettling experience of reading, a feeling of being "moved in some profound way by the verbal gesture" (Kacandes, 1993: 67). But it may not be the case at all that this experience is that of a radicalising intersubjectivity. What Hantzis does not acknowledge - a point I will take up in Chapter 7is that while this most fragile of points of view explicitly challenges the assumption of the unitary textual subject that underwrites "first-" and "third-person" modes (Hantzis, 1992: 93), it simultaneously underwrites that same self-identical subject by producing it as a constitutive precondition, embracing rather than critiquing what has been called the problematic subject of knowledge. Predicated on the experiential fact of the reader's own subjectivity, the subject of knowledge is left intact outside the text. Far from radically destabilising normative subject positions, "second-person" point of view proper appears to reinstitute the authority of an autonomous, originating subject, firmly ensconced as "second-person" proper's enabling condition.

I would like to suggest, then, that the task for critical literary theory in exploring the striking effects of "second-person" narrative might not be an analysis of it as a collusion between subjects (as described by Hantzis), but as a failure of unconventional, unauthorised narrative discourse to manufacture the necessary "token of the type 'human subject.'" Nor would the "second person's" power be understood to arise in the production of an ephemeral and revisionary point of view. Indeed, as I will argue in the following chapters, and somewhat contrary to Hopkins and Perkins's thesis, continuing to construe the issues of "second-person" narrative textuality principally in the critical and analytical terms of point of view may lead us to misconstrue the nature of narrative-"you's" most penetrating effects. Its power, I will argue, does not arise in the production of an intersubjectivity predicated on mutually authorising, always fading subjects, but rather through its confronting and perilous articulation of what much post-Saussurian cultural theory assumes to be the actual condition of subjectivity: subjectivity, that is, conceived of as constituted in language as an apparent unity - as ruptured, disparate, and produced.

the skinny

top

Index

1

2

3

4

5

6

7

Notes

Bibliography

Links

Slackwire