Ghosts haunt the fiction of David Foster Wallace, disrupting both the world of the story and the plane of the narration, provoking an enquiry into the organising voice of the text. They address this enquiry through the process of their manifestations or utterances, existing in an ambiguous liminal narrative space, calling attention to the authority and position of both character and implied author. I want to suggest that through the abstruse status of these ghost figures Wallace addresses what has been a matter of urgency to the writers of his generation: the presence of the author in the text, and their dialogic relationship with the reader.

American writers of Wallace’s generation, who wrote their first major works in the 1980s, were faced with the ramifications of theoretical discourse upon the form of the novel itself. In particular, the question of “death” haunted the academic environments within which many of these writers including Wallace thrived, as the integration of poststructuralist theory into academy syllabi, notably Roland Barthes’ seminal “The Death of the Author” (1968), interrogated the ontology of the author’s presence in the text. In his 1988 essay “Fictional Futures and the Conspicuously Young,” Wallace situates himself firmly within this generation, claiming that “the climate for the ‘next’ generation of American writers […] is aswirl with what seems like long-overdue appreciation for the weird achievements of such aliens as […] Barthes,” arguing that “the idea that literary language is any kind of neutral medium for the transfer […] from artist to audience […] has been cast into rich and serious question” (Both 63–4). Four years later, Wallace makes a more vexed pronouncement on the post-Barthesian author figure in a review of H.L. Hix’s Morte D’Author (1990), glossing the poststructuralist effacement of the author as an attack on “a post-Platonic prejudice in favour of presence over absence and speech over writing” (Supposedly 140) before offering a surprisingly brief rebuttal in the final lines of the review on behalf of “those of us civilians who know in our gut that writing is an act of communication between one human being and another” (144). In this implicit alignment with “civilians” Wallace seeks, perhaps a little disingenuously, to position himself outside the “ghastly jargon” (Ibid.) of the academy. In this sense he can be described as belonging to the generation identified by Judith Ryan as “troubled by the implications of theory for contemporary life” (Ryan 208), but on the evidence of his fiction Wallace is far too interested in how those implications come to bear upon the positions of author and reader to disengage completely with the matter of authorial effacement.

This oscillatory position is dramatised within Wallace’s fiction through a fixation on the author’s dialogic relationship with the reader, which is enacted through instances of “possession” and ghostliness that implicitly refer to the absence or presence of the dead, among whom can be found the spectral figure of the author. In staging this dramatisation, Wallace practices what Benjamin Widiss describes as “a continual rehearsal of Barthes’ claims, but never an affirmation of them” (5). Widiss suggests that the critical assumption of the steady effacement of authorial presence that developed as a response to modernism, an assumption that finds its apotheosis in Barthes’ essay, is inaccurate, suggesting that “only the most radically chance-driven works […] prove so eager to shed all authorial design” (6). Widiss reads Barthes’ author/scriptor binary as a false dichotomy, desiring instead to more subtly “read the troping of a pervasive textual praxis of solicitation when it is not represented as explicit importuning” (17). Wallace’s fiction, I would suggest, practices this implicit “solicitation” but diverges from Widiss’ rejection of poststructural authorial effacement in its recognition of Barthes’ essay as a necessary moment of importance in literary history. In Wallace’s fiction, authorial presence is implicitly amplified as a way of commenting upon its removal.

This is staged by Wallace through a reification of Barthes’ question “Who is speaking thus?” (142), with the fiction accordingly populated by multiple competing indiscernible voices which originate from powerful, absent, and often ghostly figures. As his career progresses, Wallace imbricates these presences with increasingly visible iterations of the revenant author figure, and an attendant focus on the importance of dialogue with both character and reader. I read this as a process of developing “materiality”, with Wallace performing a vexed dramatisation of Barthes’ claims before obtaining a situation in the later fiction whereby the “revenant” author, who has undergone his theoretical “death,” returns as a modified presence in the text. I do not read the revenant author as a direct revival of the pre-Barthesian author figure, but rather a ghostly return of the dead author, one aware of his existential contingency upon readerly presence and interpretation and committed to a dialogic engagement with those readers.1

As part of this process, I read Wallace as entering into a connected dialogue with two further models of authorial anxiety: Harold Bloom’s ghostly-inflected model of apophrades and Mikhail Bakhtin’s concept of dialogism and polyphony. Apophrades, whereby “the mighty dead return, but they return in our colours, and speaking in our voices, at least in part, at least in moments, moments that testify to our persistence, and not to their own” (Bloom 141), is a form of misprision whereby “the very strongest” poets achieve “a style that captures and oddly retains priority over their precursors, so that the tyranny of time almost is overturned, and one can believe, for startled moments, that they are being imitated by their ancestors” (Ibid., emphasis original). However, as Charles Harris has acutely argued, “the strong precursors Wallace was driven to overtake [also] include himself” (120), and I believe this is borne out by the steadily increased presence of an implied author figure across Wallace’s fiction.2

This presence is also implemented in accordance with an understanding of dialogism as described by Mikhail Bakhtin, who is namechecked by Wallace in “Fictional Futures” in the same list in which he includes Barthes. Marshall Boswell reads Wallace’s approach to the death of the author directly via the Morte D’Author review, in which Wallace praises the way that Hix “amends Derrida by way of Wittgenstein” (Boswell 171), and while I agree that this is a useful approach to the early fiction, I believe that reading Wallace in relation to Bakhtin provides a more sustained career-length model by which to map the problems of authorial monologism staged by the motifs of possession and ghostliness in the fiction. In his famous discussion of Dostoevsky’s polyphonic method, Bakhtin analyses the manner in which the “monologic plane of the novel” is destroyed by the character as “fully autonomous carrier of his own individual world” (Bakhtin 5). Bakhtin praises Dostoevsky’s narratives as “a whole formed by the interaction of several consciousnesses, none of which entirely becomes an object for the other” (18), resulting in “free people, capable of standing alongside their creator” (6, emphasis original). In concluding this essay, I argue that Bakhtin’s concept of polyphony and dialogism offers an important model through which to read Wallace’s spectral response to the effacement, and possible return, of the author figure.

Tracing Possession, Ghosts and Materiality in the Fiction

The narrative of Wallace’s debut The Broom of the System (1987) does not make explicit recourse to the supernatural, but the novel retains a preoccupation with uncanny occurrences and the prescriptive authority of unseen forces. Broom’s plot is driven by the disappearance of a linchpin family figure: Lenore Beadsman’s great-grandmother, also named Lenore, whose whereabouts are not conclusively addressed, but who exerts control over events from her unknown hiding place. The linguistic influence of Gramma, a student of Wittgenstein, on Lenore manifests as extreme ontological uncertainty, as evidenced by her conversation with her psychiatrist Dr Jay:

Suppose Gramma tells me really convincingly that all that really exists of my life is what can be said about it? […] that there’s nothing going on with me that isn’t either told or tellable, and if so, what’s the difference, why live at all? (Broom 119)

Lenore’s appeal is undermined by the fact that Dr Jay himself is under the control of Gramma, with his apparently disinterested responses informed by the same authority that initiated Lenore’s existential crisis. The novel is littered with numerous minor examples of narrative or linguistic possession connected to Gramma’s ghostly influence. Lenore’s controlling lover, Rick Vigorous, is open about his desire to possess her, equating possession with control and explicitly stating: “I am possessive. I want to own her, sometimes” (72). However, Rick’s possessiveness is inflected by Gramma’s disappearance: he knows that his controlling nature “does not sit well with a girl thoroughly frightened of the possibility that she does not own herself” (Ibid.). Subsequently, Rick’s desire to control Lenore manifests itself covertly and metafictionally in his own pseudonymous short stories (191). Furthermore, the sudden “parroting” vocal articulacy of Lenore’s cockatiel Vlad the Impaler, which is interpreted by Reverend Sykes as “the voice of the Lord” (275), is actually due to his ingestion of the pineal supplement partially masterminded by Gramma (148–9).

Gramma, the earliest iteration of a figure I will term the “absent possessor,” is also a site of generational and metafictional anxiety, as it is possible to locate behind Gramma the presence of another absent possessor: Wittgenstein himself, whose influence on Wallace’s writing was profound.3 However Gramma’s intrusion is limited, as despite her substantial level of narrative control, her own voice does not directly appear in Broom.

A number of stories in Wallace’s first short story collection Girl with Curious Hair (1989) develop the motifs of possession from his first novel. This process involves a deliberate transposition of the “parroting” found in the plot of Broom to the narrative registers of the stories themselves. Girl with Curious Hair can usefully, if a little reductively, be read as Wallace’s parroting of the register of several preceding and contemporary writers. Wallace imitates and parodies the narratives of Bret Easton Ellis (“Girl with Curious Hair”), Robert Coover (“Lyndon”), William Gass (“John Billy”), Philip Roth (“Say Never”) and finally John Barth (“Westward”).4 While this technique can come across as more than a little obnoxious (Wallace himself later referred to it as “formal stunt-pilotry” [“Interview with Larry McCaffery” 25]), the sequence of stories presents a slow convergence whereby the implicit mimicry of existing authorial styles builds toward a “breaking out” of the implied author in the death-of-metafiction fable “Westward the Course of Empire Takes its Way.” This concluding story, operating as an “Armageddon explosion” for the form, engenders a new approach to metafiction in Wallace’s writing (Ibid. 41). In this way, it is possible to see Girl with Curious Hair as an iteration of Bloom’s apophrades, whereby Wallace deliberately tries to force his predecessors (as well as his contemporaries) to speak in his own voice: a methodology more aggressive than Bloom’s “holding open” of the work to the predecessor, and nakedly, deliberately artificial in its mimicry.

The dramatisation of the surrender of one’s personality and voice to another is substantially developed in “Westward the Course of Empire takes its Way,” and is illustrated in the following passage where the narrator relays protagonist Mark Nechtr’s anxieties about his emotions:

[…] it’s like he’s denied access to them. He doesn’t ever feel in possession of his emotions […] Except when he shoots, he very rarely feels anything at all (Girl 303).

Mark’s anxieties involve a loss of possession of the modes of expression resolvable through the performance of an action. The rough equivocation of “writing out” or “shooting” (the latter explicitly equated with writing in “Westward”) locates the regaining of one’s own expression in an explicitly metafictional gesture – the writing of the self. However, Mark’s narrative is apparently “written out” on the page by another, heterodiegetic narrator, an unnamed member of his creative writing workshop. If Mark’s mode of expression seems, then, to actually be directed by an “absent possessor,” the climactic retelling of Mark’s own short story further contests the site of narration and ownership of the narrative, making the positions of both character and narrator unclear (264). This is achieved via some deliberately ambiguous metaleptic shifts in narration: when toward the end of “Westward” the mysterious narrator begins to relate Mark’s own metafictional story, a story Mark feels is “not his own” (355), the register becomes extremely uneven, moving between a third-person retelling of the content (356) and context of Mark’s story, and an apparently unmediated telling of the story itself (363), which relates a power struggle between a jailed “archer, named Dave” (356) and his violent and apparently omniscient counterfeiter cellmate “whose name is Mark” (361).

This convoluted amalgamation of character, implied author and narrator is substantially more complex than the possessor-possessee relationship in Broom, confounding the more simplistic directional flow of preceding depictions of possession or lack of autonomy. For the first time, it also makes the process suggestive of the presence of Wallace himself (“Dave”) in the text. “Westward” itself is, of course, Wallace’s explicit rewriting of another metafictional story: John Barth’s “Lost in the Funhouse” (1968). This subversive rewriting of a story by Wallace’s “patriarch for my patricide” (“Interview with Larry McCaffery” 48) explicitly associates the dramatisation of loss of voice, or possession by an unseen force, with the anxiety of influence. Here the process of anxiety is transfigured, via the story’s funhouse-like hall of mirrors, into a multivalent process where possessor figures (Barth, Mark the counterfeiter, Wallace’s implied presence) form an amalgamated chain of possession and counter-possession whereby each appears to infect or influence the other’s narrative position. While Wallace’s name is invoked, and his authorial presence thus implied, it is deferred within a system whereby the story’s narrative is not pervasively, palpably controlled by any single agent. While Boswell reads “Westward” as Wallace enacting a clinamen of Barth’s story (103), I believe that the amalgamated chain of influence, counter-influence and metafictional mimicry described above ultimately brings the climax of the story closer to a multidirectional parodic form of apophrades.5

Infinite Jest (1996) takes the motifs of possession, metafiction and authorial presence and marries them to a diegetic environment that is explicitly supernatural, marking the beginning of a sustained interaction with the motif of the undead narrator in Wallace’s work. This embrace of the supernatural is a systematic development of the earlier models of possession, autonomy-anxiety and metaleptic narrative fluidity, and is continually linked, ever more explicitly, to the figure of the author.

The return of the dead in Infinite Jest, specifically the return of a dead father figure who is also a cinematic auteur, stages even more explicitly than “Westward” a dramatisation of the anxiety of influence.6 In doing so, it also creates a site of anxiety over narrative authority, as Wallace builds one of the novel’s key sequences around a ghostly figure commensurate with what Brian Richardson terms the permeable narrator: “the uncanny and inextricable intrusion of the voice of another within the narrator’s consciousness” (95). Richardson argues that the presence of the permeable narrator provokes a series of questions about vocality and narration:

In its starkest form, the basic question remains: is the voice internal or external? If external, what is its possible source [?] If internal, is it a projection or delusion, or is it the voice of another character speaking through him due to his internalization of the other’s speech? Then again, is it somehow both internal and external at the same time, say a preternatural voice sounding within him, daimonlike, or is it the voice of an author creating and directing the thoughts of the narrator? Or is one speaker simply making all the voices up? All of these hypotheses are plausible, yet each is contradicted at many points by some aspect of the other possible answers (98).

In Infinite Jest, Wallace dramatises this confusion over the originating “site” of the voice through the appearance of the “wraith” of Dr James Incandenza to Don Gately, which takes place toward the end of the novel. The wraith’s presence is, however, foreshadowed in a number of earlier scenes. James Incandenza’s father makes the materialist suggestion that in order to achieve success at tennis his son must become “a machine in the ghost” (Jest 160), an unwitting inversion of Gilbert Ryle’s critique of Cartesian dualism.7 Stephen Burn, in an analysis of Infinite Jest and neuroscience, argues that “materialism is a monistic thesis that does away with appeals to soul or spirit in its insistence that mind is simply an emergent phenomenon of the biological matter of the brain” (Burn, Guide 50, emphasis original).8 This philosophical position is ironised when James later appears in the narrative as a spirit, the inversion of Ryle’s famous quote reified through the implantation into James’ cranium of a “gyroscopic balance sensor and mise-en-scene appropriation card and priapistic entertainment cartridge” (31); there is then, we can presume, a machine in Incandenza’s ghost. The doctrine of materialism, which, following James Incandenza’s father’s pedagogical approach, is taught at Enfield Tennis Academy and ultimately leads in part to Hal Incandenza’s disintegration (“I’m not a machine,” he protests during his mental and physical breakdown [12]) results in a form of self-taught social isolation. If the narrator of Girl with Curious Hair’s “Everything is Green” feels despair when he looks at his emotionally unresponsive lover and feels that “there is all of me going into you and nothing of you is coming back any more” (230), Infinite Jest performs a disturbing escalation of this monadic “sealing up” of the self, a process that becomes explicitly associated with questions of supernatural, generational and authorial anxiety.

In a conversation between Hal and Orin Incandenza, a discussion of superstition leads to the implication that “primitive” beliefs lurk beneath the surface of the materialist mindset. Hal subsequently invokes his anthropological knowledge of Canadian tribes people:

The Ahts of Vancouver used to cut virgin’s throats and pour the blood very carefully into the orifices of the embalmed bodies of their ancestors […] apparently the Ahts tried to fill up the ancestors’ bodies completely with virgin-blood to preserve the privacy of their own mental states. The apposite Aht dictum here being quote “The Sated ghost cannot see secret things.” The Discursive OED postulates that this is one of the earlier on-record prophylactics against schizophrenia (243–44).

This passage is significant in its development of the “ghostly” influence wielded by Gramma in Broom and the fears of loss of psychological authority explored in Girl with Curious Hair.9 Firstly, the absent possessor is configured as a literal ghost – a revenant spirit from beyond the grave. Secondly, the revenant is associated here with mental intrusion, which according to the Ahts is something to be feared, hence the contemporary association with mental illness. Finally, the spirit is associated with the figure of the ancestor. This composite figure, a ghost of the ancestor who permeates or possesses your own mind, is a significant advancement of the ideas about loss of authority explored in Wallace’s earlier fiction while remaining tied to the anxiety of influence; of the ancestor controlling your brain. This is a literalising and reversal of Bloom’s return of the dead: in apophrades the voice of the dead appears, almost inexplicably, to have been generated through the work of the living, who have superseded their place in the canon. In Infinite Jest the dead literally return, with the express intention of possessing the brain of their descendants.

The appearance of James Incandenza’s wraith to the hospitalised Don Gately is a concretisation of the tropes of possession, inheritance and authorship that have obliquely haunted Wallace’s fiction since his first novel. Tom LeClair initially associates the wraith with the author himself (“Wallace enters his narrative as a tall, lexically gifted, and etymology-conscious ‘wraith’” [LeClair, “Prodigious” 32]) before conceding that “the ‘wraith’ sounds like a combination of Hal and his father.” However, he concludes that the wraith sounds like this because Wallace is himself a “prodigious collaboration” of those characters (33). In fact, the wraith implicitly identifies itself to Gately as a manifestation of James Incandenza, but LeClair’s article remains valuable for its foregrounding of the conundrum of authorial presence posed by the wraith’s manifestation.10 Elsewhere, LeClair associates the living James Incandenza with Thomas Pynchon (33), another connection of the revenant figure with the literary ancestor.11

The wraith is a site of authorial confusion, and the manner in which it communicates with the supine and effectively dumb Gately stages an advancement of the “absent possessor” figure from Wallace’s earlier fiction. The climax of “Westward” suggests an end to the more comprehensible directional flow of possession via the introduction of an amalgamated set of possessing and counter-possessing narrative voices, but the wraith-Gately interface appears to simplify the originary flow of the possession by having both possessor and possessee present in the same room. However, the manner of the possession is complex. The wraith converses with Gately through his own “brain-voice” (831), implanting thoughts into Gately’s own consciousness including words that are not in his vocabulary but which nevertheless appear as part of his own thought process. This results in a narrative register that is virtually impossible to disentangle, as it is unclear to what extent the wraith is inflecting Gately’s “brain-voice.” For the first time in Wallace’s fiction, we see the kind of “permeable narrator” described by Richardson, whereby the dead influence the linguistic choices of the living, but here the possession is framed, however problematically, in terms of dialogue.12

The confounding status of the wraith’s voice problematises the possibility of an undisrupted narrative plane. The wraith itself takes a pragmatic approach, telling Gately that he should stop concerning himself about whether he is dreaming or not and “just capitalize on its presence” (Jest 830). This sentiment, with its pointed use of the world “capitalize”, is delivered by a spirit who was in life a failed communicator, evokes Wallace’s concurrently articulated dissatisfaction in ‘E Unibus Pluram: Television and US Fiction’ with the capitalistically-inflected ironising of postmodern art of which much of James Incandenza’s work is taken to be representative. Therefore, this complex iteration of apophrades ultimately involves Wallace bringing back the dead and having them parrot their (flawed) sentiments in his own narrative register, while in a radical ironising of the process the sentiments of the ghost could also be attributed to some of Wallace’s own contemporaries (the return of the living who embody dead sentiments) criticised in “E Unibus Pluram” for being “reverently ironic” (Supposedly 76, emphasis original).13

While one might be wary of LeClair’s direct association of the wraith with Wallace, it is important to note that while he associates the living James Incandenza with Pynchon, he does not extend that association to his ghostly presence. The wraith, as with the amalgamated narrative voice at the end of “Westward,” seems then to represent an amorphous, “flickering” accretion of sentiments, some of which appear to have specifically been modified through the process of death itself. As well as occupying the brains of the living, wraiths move at “the speed of quanta” (Jest 831), which has infused Incandenza’s wraith with a new empathetic facility. Wallace suggests to McCaffery that for living humans “true empathy’s impossible” (22), but the wraith permeates the barriers of consciousness, appearing at one point to literally feel Gately’s pain (839). The “dialogue” between Gately and the wraith, with its associated amalgamation of narrative contradictions and irresolution, presents a significant empathetic change in the mode of possession in Wallace’s fiction and a move, albeit a problematic one, toward a dialogic model.

Companion Ghosts

In Infinite Jest the “absent possessor”, who has until now occupied an ambivalent and occasionally malignant role, becomes present in insubstantial ghostly form, and engages in direct conversation with the mind it possesses. The “apparition” of this figure and the newly dialogic nature of the relationship results in the establishment of what I term the “companion ghost” figure in Wallace’s fiction. The companion ghost carries some traits of the absent possessor (thus Gately’s initial confusion over whether the wraith represents God or his disease [833]), but through its manifestation to the possessee and its tendency to directly entwine its consciousness with theirs, it engenders something akin to empathetic conversation. This figure still bears the traces of ancestor and author, but makes plain its desire for interaction, rather than a kind of remote orchestration. In Infinite Jest, this interaction is marked by a degree of uncertainty over the authority and agency of the wraith, who intrudes uninvited into Gately’s consciousness in an act of “lexical rape” (832). However, the mode of interaction here is more benevolent than the consciousness-controlling possessor figures of the earlier fiction. The wraith’s lengthy speech about the importance of the voices of “figurants” in his methodology of “radical realism” transfigures the possessor from a monologic figure to one at least focused on a polyphonic approach to narrative in which all character voices can be heard.

The figure of the companion ghost is developed radically by Wallace in Oblivion (2004), but in the intervening collection, Brief Interviews with Hideous Men (1999), Wallace performs a form of entr’acte which focuses on the associated granular development of metafictional tropes from “Westward,” but is nevertheless essential to the maturation of the possessor figure. The collection develops the conversational possessor-possessee relationship from Infinite Jest via the juxtaposition of the brief interviews with the explicitly metafictional story “Octet” which, it is implied, should be read as an inversion of the format of the brief interviews themselves (Brief 123). The notorious structure of the brief interviews involves the deleting of the female interviewer’s voice; the opinions espoused by the hideous men are so inherently monologic that a reciprocal dialogue cannot be obtained. Conversely, in “Octet”, which structurally resembles the interviews in its Q&A format, the narrator, a “fiction writer,” ultimately throws himself on the mercy of the reader. Unlike the occluded amalgamation of fictional and implied authorial voices in “Westward,” “Octet” offers in its appeal to the reader something close to a reciprocal form of metafiction.14 While the conversation between Gately and the wraith in Infinite Jest occurs between intradiegetic characters, here the ghostly presence is inverted and becomes extradiegetic, mapped on to the reader, who is given possession over the narrative’s success. The evident importance of this appeal is reflected in Wallace’s changing of the ending of “Octet” during the writing process, where the original final line of the main text (“Q: Self Evident”) was ultimately replaced by “So decide” (Brief Box 1.7).15

Wallace’s 2004 collection Oblivion marks an explicit return to ghostly matters. “Good Old Neon” appears at first to be a confessional monologue spoken to an unidentified partner, before a casual reference to the narrator Neal’s suicide (“it gets a lot more interesting when I get to the part where I kill myself” [143]) reveals that the speaker is a ghost. It also appears that Neal’s ghost can take physical form when he reveals that he’s “sitting here in this car” (152). “Good Old Neon” presents an advancement of both the wraith-Gately interface in Jest and the figure of the “companion ghost,” itself an inflection of and progression from the earlier “absent possessor.” There are several striking similarities between Neal and the wraith: both can move outside of linear time, with Neal’s explanation that dying “takes forever” (180) mirroring the wraith’s observation that death involves “everything outside you getting really slow” (883); both have also experienced epiphanies regarding their own sense of solipsism when watching the show Cheers (Jest 835; Oblivion 168–9) The embracing of communication in the post-corporeal iterations of Neal and the wraith are also drawn in stark contrast to the careers they pursued before their death; Neal worked in advertising, an industry repeatedly linked by Wallace to metatextually “heaping scorn on pretensions to […] virtues of authority and sincerity” (Supposedly 61), while Incandenza’s avant-garde film-making career became blighted by “metacinematic-parody” (Jest 703). In this respect, their newfound incorporeal ability to be able to converse with or enter the consciousness of another mirrors both that metafictional “Armageddon-explosion” that occurs at the climax of “Westward”, where the metafictional amalgamation of voices turn outward towards “you” (Girl 373), and the transference of power to the reader by the fiction writer at the end of “Octet”.

While the base similarities between Neal and the Wraith are striking, the degrees of “possession” in “Good Old Neon” are substantially more convoluted and complexly metafictional than in previous scenarios. Neal’s assertion that ghosts can “be in anyone’s room” (178), as well as Wallace’s note on his drafts of “Good Old Neon” that “[ghosts talk] to us all the time, but we think their voices are our own thoughts” (Oblivion Box 24.2), indicate a multiplication of the potential sites of possession, moving away from earlier scenarios where possession takes place between two individuals and suggesting, finally, that a conversation between the dead and living can occur in multiple consciousnesses simultaneously. This model is implemented in “Good Old Neon” twofold: firstly, the reader is manoeuvred into the listening position occupied by Don Gately, so that rather than being a witness to the dead-living conversation, they are implanted within it through the process of reading, resulting, as in “Octet,” in a radical and uncapped extradiegetic proliferation of listeners (a number that will continue to expand every time someone reads “Good Old Neon”). Secondly, this proliferation is mirrored intradiegetically by the sudden outward shift in narrative focus at the climax of the story, where Neal frames the instantaneity of “this whole seemingly endless back-and-forth between us” within the miniscule details of the lives of five supporting characters before entering, wraith-like, into the empathetic consciousness of “David Wallace,” who has a “totally unorganizable set of inner thoughts, feelings, memories and impressions” (180). This sudden narrative shift, with all its metaleptic and metafictional possibilities, represents a significant broadening out of the modes of character-to-character possession that have been seen thus far, and deploys a character with the same name as the author. While Neal’s narrative is a monologue, unlike the possessor-possessee dialogue in Infinite Jest, the marriage of that earlier interdimensional dialogue with the form of “Octet”s empathetic appeal to the reader results in a mode that, while retaining the motifs of possession in Wallace’s fiction, ultimately gestures towards a metafictionally-inflected, extradiegetically-directed relationship between narrator and reader that is not based in a monologically-motivated power relationship.

Neal is endowed with a number of the characteristics that separate the “companion ghost” from the “absent possessor”; direct manifestation to the listener, a narrative register concerned with interaction rather than remote orchestration, and a less monologically invasive, more empathetic position. The explicit positioning of Neal as a contemporary of “David Wallace” is also a development in the process of apophrades that shadows Wallace’s possessive or ghostly figures. While Jest sees Wallace performing a conflation of the return of the dead with the voices of both his literary ancestors and his irony-bound contemporaries, “Good Old Neon” places the revenant in direct proximity to an iteration of Wallace himself, an alignment compounded by “David Wallace’s” sense of himself as a “pathetically self-conscious outline or ghost of a person” (181). It appears, then, in accordance with Harris’ belief that Wallace wished to “overtake […] himself” (120) that the anxiety of influence in Wallace’s work has caught up with him, the fiction now less concerned with the process of overcoming external influence and more with internal refinement. What begins to take place in the metafictionally inflected “Octet,” before finding a more established form in “Good Old Neon,” can be described as Wallace performing a mediation of his own prior style, an interrogation of and escape from the voice which, until this point, he has possessed.

The Dialogism (or Otherwise) of Possession

In “Joseph Frank’s Dostoevsky” in Consider the Lobster (2005) Wallace argues that Mikhail Bakhtin’s praise for Dostoevsky’s polyphony, a method that “supposedly allowed him to refrain from injecting his own values into his novels” should be framed as “the natural result of a Soviet critic’s trying to discuss an author whose ‘reactionary’ views the State wanted forgotten” (Lobster 269). This footnoted observation, putatively a critique of the methodology of Dostoevsky’s biographer Joseph Frank, is important to an understanding of Wallace’s own approach to monologism and polyphony, his use of “absent possessors” and “companion ghosts,” and ultimately his own authorial presence in the text.

The situation of one consciousness as object for another, aligned with monologism in Bakhtin’s formulation, can also act as description of the process of possession dramatised within Wallace’s earlier works. The possessor-possessee relationship between Lenore and Gramma works on this principle, with Lenore’s mode of expression, and even her name, appropriated and directed by her absent great-grandmother. However, while Gramma is the controlling consciousness of Broom, the novel therefore taking the process of monologism as subject, Adam Kelly suggests that the narrative of Broom itself is ultimately monologic, as it merely performs “a gesture toward an open system and a readerly dialogue” that is ultimately subjugated to “the desire to control meaning and the reader’s agency” (Kelly 273, emphasis mine). A specific example used by Kelly is the fallacious ambiguity apparent in the final line of the novel (“I’m a man of my” [Broom 467]), whereby a sentence that is apparently left open for the reader to complete has in fact only one plausible outcome; the reader is essentially controlled into providing the only word (“word”) that will work syntactically (Ibid.) Kelly reads Infinite Jest as a move toward a less monologic approach to dialogue, using the example of the lengthy Marathe-Steeply interface to illustrate how Wallace has progressed to “a dialogic context in which both sides of the argument can be offered to the reader, without a clear authorial conclusion drawn” (275).

While I concur with the argument that “the desire to control meaning and the reader’s agency” is a quality that is gradually refined out of Wallace’s work, I believe that this development from monologism to dialogism is achieved not only through the development of the kind of peer-to-peer dialogue that Kelly identifies, and indeed that Bakhtin so praises in Dostoevsky, but also through a gradual revelatory process based around the changing modes of possession (both physical and metaphysical) in Wallace’s fiction. In the transition from the “absent possessor” to the “companion ghost,” Wallace addresses the question of his own monologic tendencies before finding a solution that allows him to dramatise and separate out those same tendencies. Essential to this process is Wallace’s acknowledgement in “Joseph Frank’s Dostoevsky” of the nature of Bakhtin’s analysis: that the advocacy of polyphony is also inherently bound up with the author-critic’s own position in relation to their work. This process can be observed most directly in the transition between the modes of possession employed in Infinite Jest, “Octet” and “Good Old Neon,” as these texts all feature the materialisation of the “companion ghost”: a visual apparition of the dead.

I have argued that the dialogic process between the wraith and Don Gately is not founded upon an equilateral power relationship. While the wraith professes to be able to empathise totally, the terms of their conversation are founded on the wraith’s decision to invasively enter Gately’s “brain voice” and, as in Warren’s “free indirect wraith” formulation, it is unclear what degree of control the wraith has over Gately’s consciousness. In an extrapolation of this position, Timothy Jacobs argues that the wraith’s lexical control can be extended to incorporate the entire narrative. Jacobs suggests that the wraith is “the text’s mediator, the centering and orienting presence that organises the entire narrative structure,” and that its presence can implicitly be felt from the first page of the novel, with the pronouncement “I am in here” attributable to the wraith’s possession of the entire narrative (56–59). The wraith, Jacobs argues, is the master mediator:

All is mediated, the polyphonic voices collated, by the wraith […] the narrative is dialogic, yet also complexly monologic in the sense that the wraith assembles the many voices through his own voice (75).

Jacobs’ suggestion that “the wraith […] serves as a transmission of the author’s embedded consciousness” (62) positions Wallace in a simultaneous “flickering” enactment and defiance of the “dead” author; the wraith remains as a trace of authorial presence.16 Whether or not one agrees that the wraith mediates the entire novel, Jacobs’ conflation of the wraith’s “centering and orienting presence” with Wallace himself engenders an important question, one fundamental to Wallace’s representation of possession: what exactly are the ramifications of this “dialogic, yet also complexly monologic” model in terms of the novel itself? In interviews while writing and publicising Jest Wallace oscillates between guardedly suggesting to McCaffery that “fiction’s duty [is not to] edify or teach” (26) and more prescriptively asserting that “principles and values in this country […] [seem] to me like something our generation needs to feel” (60, emphasis mine). Is there a risk that, through the ideological “centering and orienting” of Jest, Wallace’s “possession” of the narrative comes at the cost of a “plurality of consciousnesses not linked to a single ideological common denominator” (Bakhtin 17)? Jacobs himself recognises the monologic ramifications of the “centering and controlling” wraith, stating (in a comparison between the wraith and Eliot’s Tiresias in “The Waste Land”) that “to posit a governing consciousness risks the attenuation of the [work’s] many individual voices” (99).

While it is difficult to reconcile these positions, I read the presence of the wraith in Infinite Jest and the attendant development of the “companion ghost” as representing a moving toward an engagement with dialogism in Wallace’s work through awareness of his monologic tendency. LeClair argues that there is a direct correlation between the wraith’s version of “radical realism” and Wallace’s own fiction (32–3), but the wraith’s iteration of radical realism is not identical to Wallace’s approach as it is framed within the narrative of Jest as the work of a failed artist. A more useful correlation between Wallace and the wraith’s methodologies can be found in LeClair’s later explanation that, while “Infinite Jest” (the film) is “single-voiced”, Infinite Jest the novel is “multifarious and multivocal” (34).17 This is a useful approach to thinking about how Wallace begins to address the “single-voiced” monologic tendency in his own work. By embedding into his own novel the monologic work of a failed artist (an artist whose filmography bears some significant similarity to Wallace’s own previous work), and framing that artist’s gestures toward polyphony within a dialogue with a character (Gately) who chastises the artist for his self-centred approach (835), a monologic artistic tendency is critiqued within a dialogic framework. While this may not entirely rescue Infinite Jest from monologic tendencies, the objective of the episode is a movement toward its dissolution, a dramatisation and transfer of monologic power that is enacted in the uneven possessor-possessee relationship between Gately and the wraith. This technique not only operates as a simultaneous dramatisation and critique of monologism within a polyphonic framework, but also draws attention implicitly to the conflation between the wraith’s power of possession and extradiegetic control of the narrative by an author figure.

This approach is complemented by a change in the materiality of the “companion ghost”. The “absent possessors” of Wallace’s earlier work have a fundamentally hidden quality, whether it involves not being physically present or hiding figuratively behind the mask of irony. Conversely, the “companion ghost” is always in some sense present. This presence has a gradual adverse effect on the degree of possession afforded, before leading (in “Octet” and “Good Old Neon”) to a more equilateral, co-creative relationship between speaker and listener. The wraith and Neal also have a “quantum” quality that allows them to move almost imperceptibly between places and consciousnesses, an ability identified with polyphony by Bakhtin when he describes Dostoevsky’s method in the following terms:

The fundamental category in Dostoevsky’s mode of artistic visualising was […] coexistence and interaction. He saw and conceived his world primarily in terms of space, not time […] in every manifestation of the present he strives to glimpse a trace of the past, a peak of the present-day, or a tendency of the future; and as a consequence, nothing for him is arranged along a single extensive plane (Bakhtin 28, emphasis original).

If this passage recalls the wraith’s ability to move at the speed of “quanta,” it is even more redolent of Neal’s experiences of simultaneity of time and space after death.18 While the wraith’s mode of possession remains yoked to elements of monologism, the simultaneous availability of “everyone’s room” after death in “Good Old Neon” dramatises a systematic totality of empathy whereby possession no longer operates in a hierarchical manner, and the matter of “telling” that so constricts Wallace’s earlier protagonists has been dissolved. This dissolution of possession has, of course, also been informed in part through the aforementioned “opening out” of control of the narrative to the reader in “Octet.”

The Revenant Author

These “ghostly” approaches toward dialogism and metafiction lead inexorably toward the question of Wallace’s own authorial presence in his texts. As I have illustrated, Wallace’s own nominal appearances in his fiction, like those “absent possessors,” are initially hidden before emerging, by gradations, in an apparition commensurate with the introduction of the “companion ghost” in the later work. I believe that this slow accretion of presence, rather than a simple, brute reassertion of authorial control, operates in tandem with Wallace’s approach to apophrades to create a contingent, revenant author figure, “killed” by Barthes’ essay, who has reasserted their presence on the basis of material contingency and through a process of dialogism. Wallace’s revenant author accepts “the birth of the reader,” but refuses to submit to its own effacement, instead proposing an author-reader relationship that is explicitly dialogic.

Wallace’s authorial presence in the diegetic world of his fiction materialises under more than one name (Dave, David Wallace) and none of these personae should, of course, be taken as exactly commensurate with the author himself (the name “Foster” is itself an amendment to David Wallace, the name by which he went to friends and colleagues).19 Moreover, in the earlier fiction Wallace disguises or effaces some more explicit authorial references. In the original draft of Broom Lenore was born in 1962, the same year as Wallace (Broom Box 3.7), but in the published text the date is 1966. Bonnie Nadell later suggests moving the date of the action to 1986, to which Wallace replies that he “needed an Olympic year”, which doesn’t explain how the novel ends up being set in 1990 (“Letter to Bonnie Nadell”). This refining out of explicitly biographical data from a novel that Wallace later described to McCaffery as “a coded autobio” (41) is more dramatically enacted in a couple of changes made to the published version of “Westward”. Firstly, the name of the abusive counterfeiter with whom “Dave” shares a cell is not Mark, but Barth (Girl Box 14.6). Secondly, the mysterious “L____”, for whose death Dave is convicted, is called Gale (Girl Box 15.2), which was the name of an ex-girlfriend of Wallace’s (Max 58). In a letter to Gerald Howard in April 1988 about the legal fall-out caused by the use of real names in the manuscript of Girl, Wallace explains that “the first draft’s use of a real person’s name is testimony to my stupidity in terms of balancing literary and real-world considerations […] I need the circularity of “Westward”s character contriving a fiction in which I as real person am character.” Later in the same letter, Wallace indicates that L “stands for Lenore, if anything” (“Letter to Gerald Howard”). While Wallace is prepared here to use his real name for the first time, it still occurs within a complex amalgamation of registers. It does, however, represent a degree of progression from the coded and redacted biographical data from the drafts of Broom. Two Davids appear in the drafts of Infinite Jest, and both are subsequently erased. In the first iteration of the “professional conversationalist” scene, which appears to date from 1986 and has the title “What are you exactly – unadorned autobio – automabiography,” Hal is called David. The scene is set in 1974 and “David” is 13, while Wallace would have been 12 in 1974 (Jest Box 15.6). Moreover, in a later draft the character later named as Marlon Bain is called David Foster Wallace (Jest Box 16.5).

From Broom to Infinite Jest, this process of “hide-and-seek” appears to bear out Wallace’s response to McCaffery that “I’m an exhibitionist who wants to hide, but is unsuccessful at hiding” (43). However, the encroachment of Wallace’s presence in his own fiction is also concurrent with the transition of the “absent possessor” to the “companion ghost,” and their attendant associations with monologism and dialogism. The change in the mode of Wallace’s presence in “Octet” and “Good Old Neon,” a development from the earlier manifestations in Broom, Girl and Infinite Jest, mirrors the difference between the absent possessor’s remote orchestration of narrative action (Gramma in Broom, the “hidden” Wallace) and the companion ghost’s presence in and sacrifice of control of the narrative to their dialogic partner (Gately, the reader in “Octet”). While the presence of an author in a text might initially be taken as a reinforcement of monologism, reminding the reader of exactly who is controlling the narrative, Wallace’s appearance in “Good Old Neon” actually has the reverse effect, an effect that is comparable to Bakhtin’s assertion that the dialogic author creates “free people, capable of standing alongside their creator.” At the climax of “Good Old Neon,” where Neal’s ghost reveals the dialogic simultaneity of “this whole seemingly endless back-and-forth between us” with the actions of “David Wallace,” a repositioning of Wallace’s implied presence from monologic remote orchestrator to dialogic companion has been effected. “David Wallace” has been separated out from both absent possessor and companion ghost to stand, in a literalising of Bakhtin’s formulation, “alongside” his own characters in a moment of diegetic temporal simultaneity. Bakhtin explains that a dialogic character’s word does not “serve as a mouthpiece for the author’s voice […] it sounds, as it were, alongside the author’s word and in a special way combines […] with it” (Bakhtin 7, emphasis mine). At the climax of “Good Old Neon,” Neal’s voice does exactly this, simultaneously combining with Wallace’s experience of reading about Neal’s suicide and allowing “David Wallace” himself the final word of the story. Like Broom, this final word is itself “word,” but rather than provide a fallacious illusion of “openness,” this final word affords the revenant author the recognition of the contingent survival of his own voice.

Ghost Writing: Future Directions

The Pale King is ultimately an assembly, rather than a finished novel. It is therefore, despite the presence of material written after the publication of Oblivion in 2004, more difficult to categorically pronounce on Wallace’s developmental employment of ghosts and possession in The Pale King. However, it is possible to make some informed observations about his use of these motifs throughout the drafting of the novel, and Wallace’s own presence in the work, to speculate on the ongoing modulation of these forms and commitment to polyphony and to consider possible future directions in the development of the revenant, dialogic author.

In the early stages of the composition of The Pale King, when the novel was titled Sir John Feelgood, Wallace planned to have a ghost narrate the entire work (Pale Box 37.3). While it is my contention that this ambitious narrative model eventually became the basis of “Good Old Neon,” as the ghost narrator does not survive in any drafts of the novel after the publication of that story, The Pale King contains several significant references to ghosts, the most explicit of which were written during and after 2005, and which retain the associated developmental preoccupation with dialogism that has characterised Wallace’s use of the supernatural. The dialogues between Meredith and Ed Rand, who is described as “white as a ghost” (490) and “a corpse” (500) first emerge in November 2005, while the material on Garrity and Blumquist, the two “actual, non-hallucinatory ghosts haunting Post 047’s wiggle room” (Pale 315) appears to have been written in April 2007 (Pale Box 38.1). Meredith and Ed’s dialogues, which are of course framed within another dialogue between Meredith and Drinion (a character who also has supernatural, if not exactly ghostly, powers) operate as a problematised version of the ghostly dialogism in Wallace’s fiction, as both dialogues are largely fallacious, tending toward the monologic: Ed and Meredith dominate the respective conversations, using the dialogic partner as a way to talk about themselves. Central to the frustrated attempts at communication in this sequence is the fact that, while Drinion and Ed are given supernatural attributes, they aren’t actually ghosts. Ed is described with recourse to the incorporeal, but he never actually dies (Pale 509), and while Drinion has an inhuman ability to concentrate, he cannot hold a proper empathetic conversation.

The ghosts Garrity and Blumquist, who are both primarily identifiable through their vocality, correspond strikingly to the respective figures of the absent possessor and companion ghost in the earlier work, even occupying a historical trajectory that mirrors the development of the trope in Wallace’s fiction. Garrity dates from “an earlier historical period” and his verbosity is often confused with the sound of the IRS workers’ own “yammering mind-monkey,” thus giving him characteristics of the “absent possessor,” while Blumquist “basically sits with you […] the wigglers find him companionable” (316, emphasis mine). Garrity’s monologic logorrhoea appears to be connected to his previous job as a mirror inspector, which visually stages an unending, tortuous meeting of the self. Conversely, Blumquist’s outwardly focused dedication to his public sector job (he works, notably, on “partnerships” [28]), which results in his being found dead at his desk (27), corresponds to a lack of vocality (he “didn’t say anything” [28] and even as a ghost “no one ever speaks of him” [316]) implicitly similar to the communicative limitations of language expressed by Neal’s ghost in “Good Old Neon” (Oblivion 178). Blumquist’s silence represents perhaps the most extreme iteration of the companion ghost in Wallace’s fiction, a mode of dialogism represented, in a partial amendment of Bakhtin’s terms, by coexistence and interaction that is not actually vocalised.

While the presence of the author-narrator “David Wallace” points toward a further “standing alongside” of author and character as established in “Good Old Neon,” this figure states that in order to conform to the specifications of a legal disclaimer, he has “polyphonized” a memoir of his experiences in the IRS (72). The situation of “David Wallace” as a writer-in-the-world can be read as a specific turn by Wallace toward an interrogation of his own public literary persona (notably the now well established “David Foster Wallace” narrative register of the nonfiction, as well as a reflection of his extensive legal troubles while publishing Girl with Curious Hair), but the chapter in the published novel is just one of a series of possible scenarios explored in Wallace’s drafts for the novel which address the legal status of the narrative and which make explicit reference to the figure of the ghostwriter.20 Another version of this chapter involves an earlier iteration of the character, a journalist named Frank Brown, assisted by a “friendly ghost” who is helping him to fictionalise a specific percentage of his memoir (Pale Box 39.2). Yet another draft suggests that the novel will feature a “Co-Author’s Foreword” (Pale Box 39.7) in addition to the authorial foreword from the published text.

The radically different nature of these scenarios are best approached through a process of genetic criticism which maps a broad topographical model of the tropes of the novel’s drafts, as this is the nearest we are likely to get to understanding The Pale King as a totalised narrative system. In this instance, it yields an illuminating insight into where Wallace’s use of the tropes of ghostliness and possession may have led had the novel been finished, with a narrative scenario where authorial presence and “friendly ghost” speak as one, and perhaps as one another, in a co-operative, co-creative partnership (a scenario compounded by Wallace’s use of the term “ghost conflation” to denote the synthesis of two individuals in the IRS’s faulty computer system [414]). This suggests, however tentatively, that Wallace was continuing to develop a narrative mode that sought to build upon and potentially supplant the ghostly questions of anxiety and monologism that attend the death of the author.