1. On June 21, 1992, the front page of the New York Times Book Review proclaimed "The End of Books." This headline was the title of an article by Robert Coover on hypertext, and while the article presented a highly optimistic evaluation of this new art form, Coover's title was clearly playing on the technophobic fears of his book-loving audience. Coover argues that hypertexts stress "fluidity, contingency, indeterminacy, plurality, [and] discontinuity" instead of the traditional, linear narrative (25). In place of the traditional narrative structure with beginning, middle, and end, hypertexts provide endlessly "branching options, menus, link markers and mapped networks" (23); Coover adds that even character and plot are "two traditional narrative elements that are decidedly in jeopardy" (24).



2. Of course, Coover is aware that all of these features have appeared in print texts; he even begins his article with the following disclaimer: "[T]hrough print's long history, there have been countless strategies to counter the line's power, from marginalia and footnotes to the creative innovations of novelists like Laurence Sterne, James Joyce, Raymond Queneau, Julio Cortázar, Italo Calvino and Milorad Pavic, not to exclude the form's father, Cervantes himself" (1, 23). Coover's own fiction could also be added to his list; his non-linear short story, "The Babysitter," for example, could easily be read as a proto-hypertext narrative. And his prophecy of the end of character and plot seems to echo Alain Robbe-Grillet's 1957 essay "On Several Obsolete Notions," in which Robbe-Grillet argues that the concepts of character, story, form, and content are no longer relevant to contemporary literature (25-47). However, Coover quickly adds that "true freedom from the tyranny of the line is perceived as only really possible now at last with the advent of hypertext" (23), and he concludes by stressing that hypertext is "truly a new and unique environment" and its principles will displace those of traditional narrative "in the same way that relativity not so long ago displaced the falling apple" (25).



3. This article seems to encapsulate all of the significant debates surrounding hypertext over the past decade. These debates have been largely concerned with questions of authorship, of narrative, and of what constitutes a text, as well as questions about the role of technology in the act of writing and reading. While many critics have attempted to explain the differences between print and hypertext by focusing on formal properties of the texts themselves, there does not seem to be any clear consensus as to how print narratives actually differ from hypertext narratives, and in this paper I would like to suggest that it is ultimately impossible to account for the differences between these two technologies solely by examining narrative structure. Through a close reading of two of Michael Joyce's now classic hypertexts, Afternoon and Twelve Blue, and an analysis of the ways in which the characteristics of these hypertexts are compatible and consistent with the goals and techniques of print texts, this paper will attempt to illustrate the difficulties of determining a precise difference between print and hypertext narratives.



4. Perhaps the most frequently debated issue involves the question of authorship: specifically, who is authoring these narratives, or whether hypertext eliminates the notion of authorship altogether. Coover's article, which was written at a time when Brown University's Hypertext Workshop had only been in existence for two semesters, was one of the major catalysts for discussions of hypertext as a new art form which demanded new skills and new narrative techniques. In an article written by Coover in 1993, he even added that there would be a "transitional time" before writers fully adjusted to working in the new medium ("Hyperfiction" 10). Such a claim might suggest that hypertext writers are today's avant-garde artists; however, hypertext is more frequently championed by critics, such as George Landow, as eliminating the author altogether and thus serving as a perfect illustration of Post-Structural theory. Obviously Landow does not hold that this theory only applies to hypertext, but rather he sees hypertext as a way of testing Post-Structuralism (Hypertext 2.0 2).



5. This claim, that hypertext decenters the author and empowers the reader, has been contested in several different ways. Some critics, such as Espen Aarseth, argue that hypertexts are not interactive enough (178), and rather than privileging "hypertext" as a new form of textuality, Aarseth argues for the term "ergodic literature," which could be applied to either print- or computer-based texts in which the reader can truly interact; Aarseth even goes so far as to question whether there is a difference between hypertexts and print texts, and in his book Cybertext he states that "no such essential difference is presumed" (17). Other critics, such as Tony Tremblay, argue that print texts are actually more interactive than hypertexts because they allow the reader to write marginal comments, and he adds that readers of print texts are equally as active in that they "constantly jump forward and backward, left and right, reading as much between the lines as between the margins" (127). And there is yet a third group of critics, such as Silvio Gaggi, who argue that "although the reader's ability to make choices seems to indicate control and empowerment, that empowerment may be specious" (105). The reader is free to forge his/her own path, Gaggi argues, but "[t]he complexity of the Web and the possibility of having to make decisions without sufficient information regarding where any choice may lead can result in a disorientation that precludes meaningful freedom" (105). Gaggi suggests that, despite claims of empowerment, the technology is itself inherently tied to corporate and military interests (116-7), and Tremblay similarly points out that the interactivity of hypertext merely obeys the logic of corporate advertising: "[H]ypertext is the dreamwork of advertising, lending itself to a seemingly naturalized peer pressure that is an integral part of the corporate experience" (126).



6. The other characteristics of hypertext which Coover champions are its non-linearity, its multiplicity of narrative threads, and the fact that hypertext narratives are infinite, unbounded, and without closure, but the claim that these aspects are traits of hypertext and not of the traditional narrative has similarly been met with resistance. While some critics, such as Jay David Bolter, argue that non-linearity marks a crucial distinction between print and hypertext because print narratives are inherently linear ("Rhetoric" 287), others, such as Ilana Snyder, argue that there is a tradition of non-linearity in print texts which stretches back as far as Tennyson (98). Although Bolter acknowledges that there are print narratives which are non-linear, a sub-genre he refers to as "topographic writing," he argues that these print narratives work against the medium in which they were made and that they merely reveal the limits of print technology itself (Writing Space 143). Frank Kermode's theories of narrative, however, would seem to suggest that non-linearity is not simply a feature of experimental writing, but rather it is an inherent feature of narrative itself. Like the hypertext link, Kermode's notion of "peripeteia," the reversal or turn of events in a narrative, depends on the falsification of the reader's expectations, and "[t]he more daring the peripeteia, the more we may feel that the work respects our sense of reality; and the more certainly we shall feel that the fiction under consideration is one of those which, by upsetting the ordinary balance of our naïve expectations, is finding something out for us, something real" (18). The hypertext link would seem to be the most daring peripeteia of all because of the way it shocks the reader and defies any narrative expectation whatsoever.



7. This conception of narrative as a form which is inherently filled with shocks and disjunctions would leave only one feature which could distinguish print narratives from hypertext narratives: the fact that print narratives are bound by a beginning and an end while hypertext narratives are unbounded and endless. According to Kermode, narrative relies on beginnings and endings because it is the sense of an ending which gives narrative its meaning and helps the reader to become aware of his/her own sense of mortality. Molly Abel Travis similarly argues that a lack of resolution deprives the hypertext reader of any pleasure (119). However, most critics seem to agree that hypertext does provide a sense of closure. According to J. Yellowlees Douglas, for example, the only difference between the ending of a print narrative and a hypertext narrative is that the ending of a hypertext occurs at a moment chosen by the reader once he/she believes a mystery has been solved (8). Even Landow admits that hypertexts still provide a sense of an ending (Hypertext 2.0 192). Hypertext writer Michael Joyce also argues that closure simply occurs when the reader becomes tired.



8. Like Coover, who sees similar features in print narratives going back to Cervantes, the critics engaged in these debates seem acutely aware of the print antecedents to hypertexts, and while they struggle to define the precise differences between print and hypertext narratives, none of their arguments are ultimately persuasive. Many critics, such as Susan Lang, have even begun the process of incorporating hypertext into literary studies by applying hypertext technology to printed texts (305), which would suggest that print narratives are not radically different from hypertext narratives. While this is certainly an important and useful approach, particularly in its application to pedagogy, it might also be useful to look at the ways in which hypertexts are themselves compatible with traditional literary studies. A close reading of Joyce's Afternoon and Twelve Blue, and an analysis of the ways in which the characteristics of these hypertexts--such as reader interactivity, non-linearity, multiplicity, and unboundedness--are compatible with the goals and techniques of print texts, will illustrate the difficulties of determining a precise difference or specific differences between print and hypertext narratives.



9. Afternoon begins with a highly traditional and compelling narrative hook: "I may have seen my son die this morning." Like the most effective openings of print texts, this line raises many questions: Is the narrator's son dead or not? Why isn't the narrator sure? And what happened? The reader is presented with several options, in terms of words. The reader can click on any word, and the word selected will link to a new page, but no matter which word is chosen, the reader always eventually returns to the primary path of the story. As the story unfolds, it is revealed that the narrator believes he saw his ex-wife's car in an accident; and there were two bodies on the ground which may have been hers and their son Andy's. The narrator begins to play detective as he attempts to track down clues, but of course, there are many obstacles in his way, including his own status as an exile from the family, the lack of cooperation he receives from his ex-wife's lover, Desmond, the "half-blind musician," and the recurring failure of communications technologies--the latter a tangible example of narrative breakdown. The narrator attempts to return to the scene of the accident, but all he finds are impressions in the ground where the bodies were lying and a paper floating in the wind which turns out to be a report his son wrote for school.



10. The linear part of the story ends at the point when the narrator decides to call Lolly. And then the reader is presented with more options. Clicking on the word "Lolly" reveals an enigmatic phrase, "Love or death?" When this path is pursued, some obscure details about the narrator's relationship to Lolly are revealed, or quotations from literary texts such as Cortázar's Hopscotch and "Blow Up," or Sterne's Tristram Shandy, but all these links eventually lead back to the sentence, "I may have seen my son die this morning," and the story begins again--or is it continuing? Joyce also makes several references to the films of Michelangelo Antonioni, such as Red Desert and Blow Up, the latter which is based on the Cortázar story. These quotations, reflecting the labyrinthine structure of the hypertext itself, include frequent references to Frank R. Stockton's classic labyrinth story "The Lady, or the Tiger?" Joyce also quotes Tolstoy: "[G]enuine drama occurs on the upward or downward slopes, never at the apex." This quote seems particularly appropriate for a hypertext in which a reader can move from slope to slope and avoid the apex entirely.



11. Although Douglas and Travis both argue that Afternoon has a definite ending (Douglas 8; Travis 121), my readings never achieved any particular climax. For me, Afternoon never seems to answer the questions raised by its narrative hook, and I was never quite sure whether the answers existed within the text or not. But as Nancy Kaplan points out, hypertexts do not culminate with the unraveling of any plots or the revelation of any secrets: "How will we know when we're done with the reading? Either the reader exhausts the text or the text exhausts the reader . . . hypertext [reading] is never done." Joyce describes this feature of hypertext in a slightly different way: "Hypertext is the confirmation of the visual kinetic of rereading" ("Nonce" 580). In other words, Joyce claims that the questions posed by a hypertext are never answered and that the reader is never finished reading; rather, the reader is always engaged in the act of rereading, which is an accurate description of my own experience reading this hypertext.



12. Afternoon is a somewhat anomalous hypertext because, although it lacks a clear-cut ending, it still contains a clearly defined beginning which leads into a seemingly linear storyline. One of Joyce's later hypertexts, Twelve Blue, though, has no beginning whatsoever. Instead, it presents the reader with a graph of twelve parallel lines which each pass through eight points on a grid. These lines are jagged and frequently intersect or overlap one another. The grid itself provides 96 different nodes which each link to pages of text. This hypertext also differs from Afternoon in its absence of any primary storyline. In place of this linear foundation is the grid itself, sections of which appear to the left of each page and provide links to other pages. Unlike my readings of Afternoon, which all seemed roughly similar because of the presence of a single storyline, my readings of Twelve Blue were much more varied and unpredictable. There seemed to be several possible main characters and several different storylines which all twisted around each other.



13. The first time I read Twelve Blue, it seemed that the mysterious drowning of a deaf boy, Samantha's boyfriend, was the central focus of the narrative, but in my second reading, it seemed that Javier's discovery of a photograph of his great-grandmother in Ed Stanko's hotel was a pivotal scene. Although I often read the same page several times during the course of reading, or rereading, I was struck by how the meaning of individual pages changed each time they were placed within a different context. I read the same page about Eleanor pleading insanity at least six times before I learned that Eleanor had murdered Ed Stanko--and that familiar page suddenly took on a more profound meaning. Through the act of rereading, therefore, the expected became the unexpected, and the familiar became strange.



14. Unlike Afternoon, Twelve Blue does not contain any literary quotations; rather, Joyce plays other, more subtle, textual games. Every page, for example, repeats the word "blue" at least once, such as "blue cars on the whirly ride" or "Javier and Beth travel through the Blue Ridge Mountains," and the story is full of water imagery. For example, Lisle is sewing a quilt of the river which is twelve feet by eight feet, an obvious replica of the hypertext's grid. There is also a minor character, the wife of a scientist, who drowns in a diving accident by getting tangled up in seaweed. The narrator muses, "Life is a river that flows both ways, it doesn't do to get caught up in the threads the water weaves." Water therefore serves as a metaphor for the transience of the characters' lives: Their relationships are always changing, their families quickly form and then collapse, and their roles are constantly in flux. The only character who seems resistant to change is Ed Stanko, a bitter old man, and in my reading Stanko was the only character who died; the characters who survive all manage to adapt to the fluidity of their lives.



15. This fluidity is also conveyed through the often confusing use of names. Many of the characters have more than one name, or have names with several possible meanings. Beth's name is actually Tevet, meaning "April," but the name can also be pronounced "Tebet" or "Tebeth." Javier's new girlfriend Lisle is nicknamed Lee, the same nickname as Javier's lesbian ex-wife, Aurelie. And characters often misremember each other's names, such as Samantha, Lisle's daughter, who always confuses Aurelie's name with Lorelei (another aquatic reference). Everything seems fluid in the story, not only the plot but also the characters themselves and their names.



16. The characters also seem to be aware that the story they inhabit does not focus solely on them, but rather it merely passes through their lives from time to time. For example, the narrator discusses Lisle's philosophy of life: "she had taught herself abandon, taught herself that they were not minor characters, she and her daughter [Samantha], but at the center of something flowing through them." The plot of the fiction does not follow these characters, it is not organized around them, but rather it flows through them. The narrative thread the reader chooses might follow one character for a few pages and then a link will send them off to another character. These links seem to be a perfect illustration of Kermode's notion of "peripeteia," the reversal or turn of events in a narrative. Just as the peripeteia in traditional narrative is designed to falsify the reader's expectations and take him/her in unforeseen directions, so do these links falsify expectations by constantly surprising the reader and keeping him/her guessing how the narrative will unfold.



17. Kermode also discusses how certain experimental works, such as Jean-Paul Sartre's La Nausée, had a higher degree of peripeteia. Written in 1938, La Nausée is constructed around contingent or accidental relationships between events rather than cause-and-effect relationships. Sartre wrote this novel as an episodic work in which each narrative move had no relation to the last; in this way, he was attempting to write a narrative which reflected everyday life and human consciousness and therefore had a heightened "sense of reality." When peripeteia is taken to this degree, its similarity to the hypertext link becomes even more clear: Peripeteia, in its purest form, is a structural chaos which seems to express the actualities of lived experience rather than the confined limitations of a fixed narrative. Sartre's use of peripeteia to reflect the workings of consciousness also supports the claims of critics such as Landow and Joyce, who argue that the chaos of hypertext links replicates mental processes (Landow and Delany 4) and/or that they can be used to discover our own "distinctive structures of thought" (Joyce, "Siren Shapes" 13).



18. Kermode also claims that the most essential components of narrative are beginnings and endings because they are connected to the process of making human life meaningful, of giving value to the interval between birth and death: "[I]n 'making sense' of the world we still feel a need, harder than ever to satisfy because of an accumulated scepticism, to experience that concordance of beginning, middle, and end which is the essence of our explanatory fictions" (35-36). While Afternoon does have a definite beginning, its ending is completely elusive, giving the reader the impression that he/she is merely abandoning the text rather than finishing it. And Twelve Blue is even more radically disrupted: There is no beginning nor end, only a seemingly endless series of "pages." According to Kermode's definition of narrative, these hypertexts would fail to help us make sense of the world and our own mortality because they fail to establish any concordance between beginning, middle, and end.



19. Joyce acknowledges that there is a connection between narratives and mortality, although for him this connection does not rely solely on the presence of a clear beginning and ending: "Either our lives seem a line in which our reading has ever circled, or our lives seem to circle on themselves and our reading sustains us in its directness and comforts us in its linearity" ("Nonce" 580). Rereading, according to Joyce, is not a unique requirement of hypertexts but rather the standard practice readers engage in every time they read:

    Our choices change the nature of what we read. Rereading in any medium is a conscious set of such choices, a sloughing off of one nature for another. . . . [W]e linger or shift back intentionally upon a text, making each reoccurrence or traversal its own new or renewed text, the exploration of a dark seam of meaning that mere choice seems to illuminate and (we hesitate to suggest) create for us. ("Nonce" 581)

    Narratives, therefore, not only create meaning for us, but our reading choices contribute substantially to that process of creation. Joyce also extends this theory of rereading to the question of non-linearity and mortality. By drawing on the work of Hélène Cixous, Joyce argues that hypertext narrative is still capable of invoking the reader's sense of mortality, not through the sense of an ending but rather through the movement between reading and writing: "Cixous seamlessly moves from reading to writing, seeing in the exchange between them a recognition of mortality, which is to say the body" ("Nonce" 585-586).

20. However, Kermode's claim that beginnings and endings are the most essential components of narrative has also been questioned by literary critics, such as Peter Brooks, who have attempted to theorize the middle of narrative. Brooks clearly agrees with Kermode that narratives are related to mortality, but he points out that "between beginning and end stands a middle that we feel to be necessary . . . but whose processes, of transformation and working-through, remain obscure" (96). While repetition has frequently been claimed as a characteristic feature of hypertext, Brooks argues that narratives are always "in a state of repetition, as a going over again of a ground already covered" (97). To explain the function of this repetition he employs Freud's fort/da game, in which the child reenacts the disappearance of the mother in order to gain a sense of mastery over an uncontrollable situation (97). Brooks argues that, like the fort/da game, narrative is an attempt to gain mastery and assert "control over what man must in fact submit to--choice, we might say, of an imposed end" (98). Brooks goes on to suggest that "[w]hat operates in the text through repetition is the death instinct, the drive toward the end" (102); however, he also adds that repetition "retards the pleasure principle's search for the gratification of discharge" (102). In other words, narrative repetition drives the reader toward the inevitable ending, but it also functions as a constant postponement or delay of this ending, thus giving the reader a sense of mastery or control over it. Brooks eventually claims that by constantly postponing the ending, repetition "ultimately subverts the very notion of beginning and end" and suggests that "the interminable never can be finally bound in a plot" (109). Brooks also compares narrative to the process of analysis. Like analysis, Brooks argues, narratives "impose an end which yet suggests a return, a new beginning: a rereading." He also claims narrative "wants at its end to refer us back to its middle, to the web of the text: to recapture us in its doomed energies" [emphasis added] (109-110). Although Brooks does not discuss hypertext, his description of narrative as a "web" of endless repetitions, which is constantly being reread, without beginning or end, would seem to suggest that hypertext narratives with the same features merely reflect the fundamental structure of narrative itself.



21. The only way in which Twelve Blue might still remain distinct from print narratives is through its use of images. It not only uses an opening graphic, which I have already discussed, but it also incorporates illustrations of a riverbank. Landow and Delany employ the term "hypermedia" to describe hypertexts which incorporate images and sound; the promise of hypermedia, they claim, is that it "takes us even closer to the complex interrelatedness of everyday consciousness" (7). Such a practice of hypermedia seems to directly address Vladimir Nabokov's criticism of James Joyce's "stream of consciousness" technique: "[S]tream of consciousness is a stylistic convention because obviously we do not think continuously in words--we think also in images; but the switch from words to images can be recorded in direct words only if description is eliminated" (363). By combining images, text, and sound, therefore, hypermedia has the potential to replicate consciousness more precisely than print texts. Nabokov also points out a second difference between stream of consciousness and actual consciousness: "[S]ome of our reflections come and go, others stay; they stop as it were, amorphous and sluggish, and it takes some time for the flowing thoughts and thoughtlets to run around those rocks of thought. The drawback of simulating a recording of thought is the blurring of the time element and too great a reliance on typography" (363). Hypertext also seems to resolve this problem by not relying on a strict typography and by blurring the time element of reading through frequent and irregular stops and starts. Hypertext's ability to duplicate "everyday consciousness" can therefore be seen as a natural extension of the modernist project.



22. The characteristics of hypertext which were praised by Coover nearly a decade ago, such as reader interactivity, non-linearity, multiplicity, and unboundedness, can clearly be seen as natural extensions of techniques and goals which are fundamental to print narratives as well. However, there remains an understandable urge to define precisely what is new about the medium of hypertext, or, more generally, to define the relationship between individual narratives and the technologies with which they are written and read. For example, Joyce argues that the question hypertheorists should be asking is not whether print texts anticipated the innovations of electronic textuality but rather whether differences in reading characterize them ("Nonce" 587). The key difference, Joyce adds, is that the hypertext reader can never go back to the same text again, and thus rereading is always an "unreading" which undermines or effaces a previous reading ("Nonce" 589). Tremblay similarly focuses on how hypertext changes the act of reading rather than its supposedly transgressive narrative form. "It seems only so much common sense that a message written with a typewriter . . . differs from the same message written with a computer" (123), Tremblay argues, but the only real distinction he sees between hypertext and print text is that "with the advent of the computer, a populist medium is for the first time in history both [a] writing and reading machine" (124). In other words, what is truly unique about hypertext is that it can only be read with a computer, and the computer, Tremblay argues, "foregrounds the visual and auditory to the detriment of the tactile and haptic" (128). This leads Tremblay to address what he calls the "physiological dimension" of reading and the tactile qualities of the book, such as the size, shape, sound, and weight of the page; Tremblay even goes so far as to mourn the loss of the book's smell (140).



23. Tremblay's argument is certainly not new to hypertheorists, and most of them respond to these claims as simply conservative and reactionary; Coover even anticipated such a response when he imagined a hypothetical reader resisting hypertext by saying, "what do you mean, you can't take it to bed with you?" ("Hyperfiction" 10). However, it seems that a consideration of the physiological dimension of reading does not simply mark a resistance to hypertext but rather a valuable insight into the essential difference between hypertext and print text. Instead of explaining this difference through the formal features of hypertext narratives, such as reader interactivity, non-linearity, multiplicity, and unboundedness--features which fail to adequately distinguish the experience of reading hypertext from the experience of reading print text--recent critics, such as N. Katherine Hayles, are calling for a new form of "media-specific analysis" which "moves from the language of 'text' to a more precise vocabulary of screen and page, digital program and analogue interface, code and ink, mutable image and durably inscribed mark, texton and scripton, computer and book." In other words, rather than making claims about the ways in which the formal features of hypertext narratives affect the process of reading--claims which can be just as easily applied to print narratives--media-specific analysis focuses instead on the technology of hypertext itself and the ways in which this new form of media affects the physiological dimension of reading. By focusing on the material differences between computers and books, for example, media-specific analysis promises to provide a much more accurate description of the ways in which the technology of hypertext affects the reading experience and the ways in which the act of interpretation is mediated by the relationship between the writing and the reader. Such an approach thus offers a much more precise explanation of the essential differences between hypertext and print text, as well as the ways in which the computer collaborates with the reader in the creation of meaning.

Works Cited