Go to Semiotic Review of Books Home Page
Go to SRB Highlights
Go to SRB Archives

SRB Archives

This review appeared in Volume 6 (3) of the Semiotic Review of Books.

Texts Unbound

Ken Paradis

HYPER/TEXT/THEORY George Landow (ed.), Baltimore: Johns Hopkins University Press, 1994. ISBN 0-8018-4837-7

In Hypertext: The Convergence of Contemporary Critical Theory and Technology (1993) George Landow identifies the need to assess the communicational nature of new electronic writing technologies. In critical theory he finds tropes and discourses which can be used to begin to articulate the textual forms and associated cultural practices of these communication forms. For Landow, "hypertext", a catch-all term designating the range of electronic writing practices, "has the potential to serve as a laboratory for theory while theory illuminates the design, use and cultural effects of the new electronic technologies" (1994: 2). This contention is extensively explored and evaluated by the contributors to Landow's HYPER/TEXT/THEORY (1994).

The types of writing most familiarly associated with hypertext are those made possible by software which allows sets of documents to be more or less freely linked to each other, in which the linked document can be brought onto the same screen and read together with the "original" document. Axial hypertext systems allow "notes" to be brought alongside and read with the primary text, the relation often figured as a tree trunk with branches. Links in these hypertextual forms function analogously to the asterisks, endnote numeration, or indicators of notes written in the margin of printed text, though in hypertext the physical marginality of these paratexts need not be maintained. Hypertextual networks are more complex forms often figured as webs, stacks of cards, maps or rhizomes. In these forms the notion of an originary or central text disappears, replaced by the operating concept of a potentially infinite set of textual chunks (frequently called, after Roland Barthes, lexia) , which can be linked, at potentially any point, to any or all of the other bits and chunks of text. The "navigator" of this field has the ability to move through the lexia at almost any speed, in almost any pattern or configuration (depending on the particular hypertext environment and software), frequently having the ability to create new links and add new lexia.

Technological limitations restrict the movement of the navigator in all extant hypertextual environments, which are limited in the size of their hypertextual field, the "granularity" (decomposibility) of their lexia, the number and nature of possible linkages and the power of their search mechanisms. The non-existent "pure" network hypertext environment envisioned by Ted Nelson as the "docuverse" and by Landow as the "metatext" (1993: 8) is the repository of human information arrayed in a field accessible to an incredibly powerful search mechanism, allowing the user to freely forge associational links between any of the data units. It is an infinitely dense web of infinitely "granular" lexia, allowing (or condemning) the reader to completely organize her own informational existence. This vision, in which the individual is seen to be completely controlling or completely controlled by this ocean of information, stands as the dream and nightmare of the techno-optimists and techno-pessimists who speculate on the cultural impact of the coming electronic society.

The production of writing in Nelson's docuverse is analogous to the production of meaning in models which have been developed to describe the operations of signs. Eco's Global Semantic System (under the functional logic of the "model Q"), for example, designates a loose arrangement of signs and sign clusters analogous to the largest conceptions of intertextuality, implicating language and all its possible textual usages (1979: 68). "Writing" under these models takes the form of the inscription of connections between nodes of the informational set, these inscribed relations then subject to intersection (interpretation) through the establishment of relations to other nodes. Inasmuch as the "reader" moving through a network hypertext can choose which links to follow and which to ignore, this person has the ability to recombine the lexia in accordance with her associational impulses, in effect "writing" her own narrative ofprogression through the field. Thus "reader" and "writer" seem clumsy when used to describe hypertextual operations, replaced by most theorists by a term such as "wreader", "riter", "screener" or "navigator".

The essays of Espen J. Aarseth and Gunnar Liestol consider the validity of the notion of hypertextual "non-linearity". Aarseth makes the distinction between the manifest text on the page or screen, which he designates the informative, and the text-as-read, designated the interpretable. For Aarseth, the non-linearity of hypertext is valid only in its most basic, mathematical sense: hypertext inscribes no fixed linear sequence of "informative" text like the unidirectional line of print running through a book. But hypertext shares a certain linear phenomenality with printed text in that the act of reading, any act of reading, is a temporal process of sequential encounter with text which can be spatially figured as a line; on the "interpretable" level hypertext is as "linear" as print, a consideration which severely problematizes offhand attempts to label hypertext "nonlinear".

That established, Aarseth's project involves considering the ways in which hypertext can be structured to affect the temporal line of reading. For Aarseth, hypertext (or any text) can be described in terms of dynamism (the fluidity of the size of the field), determinability (the fluidity of the sequence and internal arrangement of the field), transience (the ability of the screener to control the temporal speed and direction of her movement through the field), maneuverability (the screener's freedom of "spatial" movement within the field), and user function (the "genre" of the hypertext which determines its use: game, encyclopedia, fiction etc.). Aarseth's categories are pan-textual, clearly establishing the continuity between print based and electronic forms. He notes, however, that the hypertextual link, while a form of intertextuality, cannot precisely be claimed (as have other theorists) to literalize intertextuality itself. Where the hypertextual link makes the intertext virtually manifest to the reader, presenting a complete and present companion text, print based forms like allusion or quotation necessitate the shadowy half-presence of the alluded-to intertext and effect quite a different reading experience. The hypertextual link-facilitated reading experience, Aarseth feels, is the phenomenon which is of the greatest potential import to literary theory in its investigation of "how the sign system is used to construct and explore the possibility of a text-based representation of identity" (83).

Liestol directs his attention to what he dubs "the reader's narrative in hypertext" (88). He uses the example of Wittgenstein's aphoristic book Philosophical Investigations (a set of densely inter-allusive but discrete fragments physically arranged in an invariable linear order) to demonstrate, (complementing Aarseth's contention) that reading, any form of reading, is simultaneously both figuratively linear and non-linear. Though the act of reading is temporally sequential, and though a narrative may be linear, the processes of text-actualization can best be figured spatially as the formation of complex, multi-dimensional structures such as dense webs or nets. As the reader establishes the text's complex inter- and intratextual orientation it seems to both reach beyond and double back upon itself.

For Liestol, the web of associations invoked by the print-based text is assumed by the reader to be an imperfect realization of the text's original, intended associational web. But because the author in hypertext can only firmly fix certain links and establish the boundaries of the individual lexiae it is the reader who assumes the role of the final organizer; it is she who decides the arrangement of the final web of relations. So any hypertextual reading, Liestol claims, is recognized as merely one of the possibilities in the non-hierarchical field of reading possibilities tolerated by the macro-structure of the hypertext web. In addition to establishing the linear sequence (the argument or narrative line) and attempting to reconstruct the net of linguistic relations, the reader of hypertext has also to conceptualize her own reading as a member of a much larger set of possible readings. Thus, though "linear" is a misleading figure for any reading, "non-linear", as simply its negation, describes hypertext less well than Liestol's proposed replacement (admittedly limited in its figural commitment to geometry): "multi-linear".

As long as hypertext is given a primarily encyclopedic function (which claims no semantically significant organization above the level of the lexia), this toleration of a multitude of reading sequences poses little theoretical problem. But when whole hypertexts are harnessed to the task of communication the situation becomes far less clear. David Kolb picks up this problem, investigating the status of terms like "intention" and "understanding" for the readers and writers of hypertext for whom the presumption of a primary and priviledged textual macro-organization (authorial intent) cannot be maintained with regard to the nonhierarchical field of linear reading possibilities presented by hypertext.

Kolb senses that philosophical discourse, with its Socratic impulse towards "truth and final grounds" (325), seemingly requires an efficient (preferably transparent) communicative medium if its constitutive notions of rigour and consistency are to be communally evaluated. After all, Kolb asks, "What would thinking mean if it were not providing form and focus, definite claims, critical judgment, beginnings, middles and ends, and so preventing an indefinite accumulation of words and images?". "This question", he states, allowing himself to consider possibilities of philosophical discourse in hypertext, "remains open" (329).

Hypertext, he indicates, could have several philosophical uses. For example, it already provides powerful research tool, allowing information to be processed more quickly and easily. As well, it could provide a medium in which philosophical texts are more easily grafted onto and into one another, accelerating the present philosophical multilogue. More radically, he notes, it could provide a medium for articulating the fundamental but often neglected non-linear aspects of philosophical discourse. The linear argument, Kolb explains, which is seen as the means of the work of philosophy, is always surrounded by and grounded in what Kolb calls "fluid discourse". This is the commentary, digression, extrapolation and rationalization which support the argument, in which there are "no primaries or fixed meta-levels because it is here that such things get established" (331). This discourse could flourish in hypertext, Kolb notes, though he remains rather vague about its status. He acknowledges that while it should not be made to replace the line, its own function could be better appreciated and its possibilities explored. For example, Kolb notices that Hegel's writing, in its attempt "to show the mutual constitution and interdependence among the categories that structure thought" (332) is necessarily and integrally less like a line moving through a field than like a set of progressively larger nested spheres, shaped largely from this fluid discourse.

But this does not mean that Hegel's Phenomenology would be unchanged if it were atomized and rearticulated in an open hytertextual web. Like linear argument, this intricate globalized form of argument still presumes the conditions of intent and understanding and still requires the authorial organization of the material which can be recognized by the reader. By largely removing the ability of the author to determine the relations (and the sequence) of the parts of the text within a bounded whole, hypertextual writing could not marshal the coherence necessary to globalize the mass of fluid discourse which Kolb identifies in Hegel's systems. So Kolb calls for hypertext systems with 'smarter', more overdeterminable links which operate to allow some intermediate structures or subtexts to be built, thereby allowing the author some power to communicate macrolexial argument patterns to the reader.

But while Kolb indicates ways in which hypertext could be tamed and made more communicatively useful, he also gestures vaguely to a philosophical paradigm beyond the communicative which might necessitate precisely the fluidity, dynamism, transitority and "noise" of hypertext. Though unsure if this writing would be considered philosophy under contemporary criteria, Kolb indicates that the line or the globalizing system could be replaced by the structured aggregate, "with parts that remained outside of other parts, in an assemblage that did not demand balanced unity, wherein changes in one place did not necessarily demand changes elsewhere" (339). Whether expressed in the DeleuzoGuattarian language of a rhizomaic assemblage operating through relations of immanence and contingency, or in terms of Lyotard's notion of the paralogy, (the negotiation of effect through language games in which meaning is a non-universalizable, local game-function), Kolb seems to lean toward what is frequently identified as the communicative reality of network hypertext, what Stuart Moulthrop has chosen to call the condition of "global variability in tension against local coherence" (388).

Narrative shares with philosophical discourse the presumption of a global (rather than local) logic (reposing in the figure of the narrator) which organizes the field of text from beginning through middle to end, a shared presumption which seems to invalidate its exploitation of the most radical communicative attributes of hypertext. J. Yellowlees Douglas, in her reading of Michael Joyce's hypertext fictions White Afternoon and WOE, evaluates the possibility of hypertext to engender closure, "the sense of an ending" so necessary to narrative. For Douglas the narrative sense of closure emerges when the work becomes coherent through the reader's apperception of global patterns of organization; it is posited (and thus created by the reader) rather than necessarily provided by the text. Douglas seeks to understand how global patterns can be apperceived under conditions which preclude the possibility of exhausting either the textual set or the ways in which it can be organized to be read.

Douglas finds that some of her readings simply do not account for enough of the data encountered or do not satisfy the demands of narrative logic in terms of character resolution, allusions, informational gaps, etc. But eventually she undertakes a reading, which though not exhaustive, she decides does indeed provide a satisfactory sense of closure. For Douglas this reading provides what J. David Bolter calls the recognition of a "structure of possible structures" (164), the point at which a meta-structure can be hypothesized which could integrate even unencountered material. This "final concluding metaphor which organizes patterns in the text into a coherent, tangible whole" (166) is, for Douglas related to the discovery and identification of a "central junction" (171) in the text of White Afternoon which allows a limited field of possible structures to emerge (and implicitly, by noting the centrality of this junction, allows other structures to be deemed marginal or in some way invalid). In reading Joyce's hypertext fictions Douglas constructs heuristic meta-stories, hypotheses of possible structures which make comprehensible the narrative data she encounters. Closure, even in this looser form, remains necessary to hypertext narrative. On this basis Douglas decides that hypertext narratives can be seen as varieties of Eco's "open works" (1979), works which tolerate and demand more imaginative investment of the reader in exchange for less communicative efficiency.

Where Douglas finds reading patterns which can (at least partially) accommodate the communicative differences of hypertext, Mireille Rosello moves in another direction, attempting to theorize the ways in which reading itself (and with it, thought) could change to more fully exploit the textual possibilities made available by electronic communication technology.

Rosello condemns the movement to figure hypertextual interpretation in the rapidly-solidifying terms derived from print based reading. She feels that new, less constraining figurations must be developed which better articulate the radical differences of hypertext. For example, reading hypertext is commonly expressed in geometrical, if not geographical, terms such as "moving through", "lines", "webs", "nets", "stacks", "fields" etc. These, however, do not convey hypertext's virtual possibilities of absolute contingency, its inherent independence from notions of proximity. Lexia are not physically nearer or further from each other, so their figuration on web map of lines, centers and margins is conceptually misleading. Though she concedes that spatial tropes are inevitable she feels they must be used as self-consciously as possible to figure hypertext's possibilities. Escaping these tropes may be necessary to conceptualize, articulate and develop hypertextual social forms. "Perhaps", she writes, "a new geometry of space is needed in order to invent communities that will have little to do with proximity and context" (132). Unlike Kolb and Douglas who feel that some form of structure must be imposed on or retained in the hypertextual field to facilitate communication, Rosello focuses on the possibilities of coherence at the most granular, local, level. She seeks to explore the possibilities of text at the threshold of communication, investigating the status of meaning in a field of random signification.

She does this by displacing the end-oriented thrust of metaphors such as "traveling" or "exploring" (getting somewhere, getting something done/mapped out) which are used to figure communication, with a geographically non-teleological figure: that of the wanderer. Meaning, for the wanderer, is entirely the product of immediate movement rather than the end (and recovered beginning) of the journey. Movement results in encounters with various points (lexia) between which relations can be developed, and in the relation meaning emerges as the points gain the ability to be expressed in terms of one another. Multiple encounters organized through a single perspectival nexus links these meanings into narrative, a non-teleological narrative of movement itself. As Rosello writes, "If we let randomness function like a paradoxical organizing principle, we may be capable of hearing new and intriguing songs" (147).

But Rosello's liberatory move to randomness is countered by the centripetal exigencies of reading. She characterizes the wanderer as one without fear of disorientation, but she does not explicitly acknowledge that interpretation is precisely the activity of orienting points to one another and of perspectival node to field. As soon as the wanderer begins endowing the relations between events and encounters as anything other than mere and absolute contingencies the presumption of randomness necessarily disappears. It is replaced by a narrative logic, an essential attribute of which is extension; narrative grants the encounter a mythos (a past and future) and a cosmos (a world order in which it could occur). The random contingency has been mapped onto a narrative plane which engenders, if not the determination of causality, at least the possibility of global comprehension. Rosello notes that in one of her example texts, the interpretation of random movement produces in the reader an "oscillation between the feeling of being lost and the intuition that new combinations are needed" (150). In other words, random reading could only be a reading toward the end of randomness, toward an order which would engender comprehension. The principle of randomness so titilatingly offered by hypertext writing is again shown to be incompatible with demands of reading.

Unlike Rosello who seeks to identify a new way of reading more suited to hypermedia, theorist Gregory Ulmer describes a new cognitive mode which, he claims, inevitably arises from this form. Ulmer decides that hypertextual writing is fundamentally distinct, governed by what he calls "hyperrhetoric", an associative rather than communicative writing form governed primarily by the inferential process of "conduction". Ulmer feels that hyperrhetoric conforms to Gilles Deleuze's description of the situation in which two non-parallel series are brought into relation with one another through their participation in "a paradoxical case shared by both series without being reducible to either one" (352). Upon encounter with the two series the interpreter casts associatively through his mental encyclopedia for these paradoxical cases which can function as hyperrhetorical "conductors", a mutually peripheral sign which does not function as a term through which the first can express its relation to the second, or vice versa. The inscription of the movement first from one term to the conductor, then from the conductor to the second term (which produces a ground upon which a different relation of the first to the conductor can be articulated, and so on) produces, Ulmer feels, a serial form of composition based on contingent relations rather than on the adherence to a transcendent structure of composition.

Unlike the supposedly rational "concept" engendered by other logical processes, Ulmer claims that conduction is essentially intuitive, finding "word-things evoking feelings. Mood" (371). Ulmer's essay records the "writ[ing of] an intuition" (353) by finding and articulating the conductor of two different signifying complexes.

Why hypertext necessitates, generates or is characterized by hyperrhetoric is never explained; the connection between the two phenomena is frequently and emphatically stated but never justified. His own essay demonstrates that "conduction" is easily undertaken outside of a hypertext environment, that is, if his essay actually records a conduction. This confusion exists because his own recorded associative processes, though explained in terms of Deleuzian and Wittgensteinian models labelled "conductive", have as little to do with those models as with models of any other logical process of inference. As well, the differences between "the print age" and "the new electronic age", nods to Walter Ong notwithstanding, are never questioned but presupposed, as are the viability of his never explained but frequently used neologisms such as "the popcycle".

In his concept of serial composition Ulmer does at least tend toward a form of writing that provides local coherence in a condition of global variability. But Terence Harpold, like Rosello and Douglas, reasserts the primacy of the reader's cognitive need to presume some form of global consistency as the precondition of his ability to form conclusions.

Like Douglas, Harpold notes that any conclusion reached in a hypertextual encounter is contingent, produced not after the text has been fully encountered but after the reader (out of exhaustion, boredom or the sense of an ending) simply chooses to stop. He recognizes that reading is guided by a "will to make sense" (193), a will which makes any assertion of mere chance (or randomness) in reading duplicitous, because reading renders even chance potentially meaningful. Reading then involves an implicit faith in understanding, a condition which, Harpold insists, involves the invocation of a figure of a Communicator. This "absent interlocutor" (194) functions as the repository of meaning which the reader seeks and in whom lies the end of the textual dialogue. "The [reader's] inquiry", he writes, "is, even before it is made presumed to be answerable by a voice within that scene, one that will fulfill the dream of a convergence of text's substance and signification" (194).

But in following hypertextual links through a field the reader knows that it is she herself who is organizing the material sequentially. To make or follow a link is irreversibly to enact one's own will in the absence of an authorial indication of a correct reading pattern, the recognition of which forces either the abandonment of the absent interlocutor figure or an obsessional faith in the saturation of the link (and thus the entire textual field) with that figure's semantic intent. The hypertext is either communicationally meaningless, or is impossibly meaning-full. For Harpold there is either no reason to read, or always too much to read. Unlike Rosello's reader, content to take what comes from a random wander through a textual field, Harpold's reader is a much darker figure, doomed to become either quickly disengaged after recognizing that there are no conclusions to be had, or to become obsessional in her perception of a meaning-saturated environment which demands endless reading.

The social aspect of the problem of coherence with regard to hypertext is explored by Charles Ess. He notes that hypertext has been frequently claimed both as a tool of emancipation toward a fuller democracy, and as the instrument of ever-greater institutional or corporate oppression. His project, using the political philosophy of Jüautrgen Habermas, is to both describe and prescribe the ways in which hypertext can and does function to further the ends of democracy. Ess recognizes that because the more anarchic potentials of hypertext are either not useful or dangerous to institutions with an interest in public compliance, there exists the potential for electronic communication to be yoked to the ends of institutional informational hegemony. This possibility impels Ess to invoke Habermas' doctrine of communicative reason as a sort of charter for electronic communication.

Communicative reason finds truth in consensus rather than truth in a mimetic relationship with Being. Therefore "truth" can be negotiated to fit the immediate conditions it need govern without having to be potentially universalizable. It can be a product of local coherence which is not invalidated by its enactment in a situation of global variability. This truth is negotiated rationally by all interested parties, in which every party is allowed to express him or herself, question any assertion and introduce any assertion, but, significantly, cannot coerce the ability of other parties to express themselves.

Electronic communication, Ess notes, provides an ideal medium for the development of this communicative reason. He insists that a "discourse ethic" is incipient in present forms of electronic communication and must be adopted if hypertext is to contribute to "grass-roots democracy" (245), thus avoiding hypertext's totalitarian possibilities. Because "As Habermas shows, the preference for a democratic polity... lies implicit in the structure of everyday discourse as such" (246) Ess feels that electronic communication can operate as a site in which the democratic preference can be played out.

But the notion of self-expression presumes a Cartesian a priori autonomy of the ego and a Romantic faith in organic essentiality, as well as the possibility of a non-coercive discourse governed entirely by rationality enacted in a perfectly free site of dialogue. Ess dismisses all critique of these conditions as the work of "poststructuralists and postmodernists" (249). These theorists, he notes through Habermas, "undermine the possibility of opposing totalitarianism by rejecting Enlightenment notions of freedom, individual autonomy, communal solidarity and democratic self-determination" (248). This may be a fair claim, but it neither refutes nor invalidate the poststructuralist critiques of the ideas so central to Habermas' discourse ethic.

Like Ess, Stuart Moulthrop considers claims of the emancipatory potential of electronic communication but where Ess sees the possibility of appropriating hypertext technology to the task of working toward an ideal state (a version of Enlightenment Democracy), Moulthrop seeks to explore the ways in which hierarchy and power structure itself will modulate under the influence of conditions of new communicative media.

Like Ulmer and Rosello, Moulthrop envisions the possibility of hypertext to provoke an epistemic change in cognitive processes, from those organized around ideas of logos or "the law of substances" (301) to those organized around nomos, "the designation of places or occasions" (301); around becoming rather than being. Moulthrop's model for this new cognitive mode is Deleuze and Guattari's "nomadic thought", a practice which "attempts to displace a language founded on logocentric, hierarchically grounded truth and replace it with an unfounded play of anarchic, contingent paralogies" (301).

Textually Moulthrop finds this distinction echoed by Barthes, who distinguishes between "the work" and "the text" in which the former is communicative, while the latter is a site of signification. The problem, for Moulthrop, occurs with the recognition that culture is manifest communicationally, and that communication is seemingly inextricable from logos. Though it may be possible to arrive at a point of becoming (a singularity) through a recombination of hypertextual lexia, in order to make that point culturally significant it must be communicated. And because communication must express this singularity in terms of other, culturally known things, the singularity of the point of becoming is eradicated and it becomes merely a variation of a substance. Therefore, while hypertext will ultimately be useful (in making a reproducible, communicable product) only in a highly territorialized form, its nature is latently highly deterritorializing, making it remain a potent site for the operation of nomad thought in a logos ruled world.

The ambivalence of Moulthrop's judicious evaluation of hypertext's political implications arises from the inability of language to articulate nomos. One of the tropical fields of logos which most fundamentally curtails the conception of hypertextual potential is the language of spatiality. This is the concern of Martin E. Rosenberg's extraordinary contribution to the volume.

Physics, for Rosenberg, assumes that "time is as symmetrical as space and geometry can represent, with certainty, the physical world" (278). It describes a set of natural laws (the laws of Being and nature) which are transcendent with regard to time, reducing time to a function of determinable causality. Geometry offers a transcendent perspective, from which (the presumably stable field of) space can be organized and movement rendered reversible with regard to time. Geometry lets you get back to where you were, by letting you know how you got there and where you went.

Rosenberg notes that the emancipatory claims forwarded for hypertext can be roughly placed in a tradition of avant garde aesthetic theory which posits epistemological (and frequently social) liberation through the restructuration of practices of signification. These changes are commonly figured in the tropes of physics, in which the laws (logos) of physics are used to represent domination or restriction, while the departure or escape from these laws are used to represent the condition of freedom. The avant garde, Rosenberg argues, has figured liberation since the nineteenth century in terms of contingency rather than causality, becoming rather than being. To implement change, the avant gardistes appealed to a reality of universal instability. Everything always changes, and though those changes can only be measured statistically (usually), the ability to truly return is an illusion. Movement, because of its temporality (bracketed off by geometry), is irreversible. In this universe of micro-instability causality becomes a necessary illusion rather than a law, and contingency becomes the condition of reality. Of course cultural law and domination (as we know it) cannot exist in recognition of contingency. Thus its appeal to the avant garde.

But, Rosenberg notes, there are no meaningful tropes to articulate this condition except those which negate figures inscribed through physics and geometry. Worse, hypertext is made more useful through aids like web-maps, links and lines, which, though allowing hypertext to be conceptualized, reproduce the illusion of a transcendent, atemporal perspective from which the hypertextual field can be manipulated.

Thus Rosenberg arrives at the dilemma which plagues virtually every theorist in this volume. For hypertext to be made intelligible (the project especially of Aarseth, Liestol and Douglas in this volume) it must be articulated in tropes accessible to extant meaning fields and practices, practices which are perhaps epistemologically inherent (Harpold). Yet the exploitation of the possibilities of difference offered by hypertext requires either its yoking to the service of some anticipated condition (Ess), or to some radical break with the very tropes with which we think in and about it. Kolb, Ulmer and Rosello speculate with regard to the form of such breaks, and Moulthrop evaluates their possibility and practicability.

Rosenberg's evaluation, echoing Moulthrop's, seems particularly apt. He notes that possibility of difference arises from the ability of the hypertextual link, as a temporal act, to be fundamentally dislocating. Moving across a link can change everything, can be an irreproducible and irreversible encounter. This recognition of radical difference is fundamentally disorienting to the hypertext reader, making the question "Where am I?" invalid with the recognition of the irreproducibility and incomprehensibility of the present nexus. Of course, for the hypertext reader sequence is often relatively unimportant, and potential confusion is negated by orientational aids such as web maps which reinstate the comforting illusion of temporal reversibility. But the link, what Harpold (citing Zizek) calls "pure act" can be a moment of bifurcation, a moment of real newness and changed only if its power of dislocation is enough to force the reader's reconfiguration of the environment to the disjunction rather than having the disjunction made intelligible within some prexisting conception of its environment.

Rosenberg quotes physicist Ilya Prigogine regarding the nature and environment of bifurcations in self-organizing systems: "We expect that near a bifurcation, fluctuations or random elements would play an important role, while between bifurcations the deterministic aspects would become dominant" (289). The possibility for difference then, for change and, to adopt the language of the avant garde, for emancipation or liberation offered by hypertext lies in its potential to incite bifurcation. The barely suppressed temporality of hypertextual reading engenders a high potential for the disorienting recognition of the contingency of the system, a potential more easily realized if randomness, as Rosello advocates, is a condition of the encounter.

Landow's HYPER/TEXT/THEORY articulates both the limitations and possibilities of hypertext, articulated in part here by Rosenberg. The avant garde emancipatory potential of electronic communication is curtailed by its role as a communicative cultural tool, its potential for critical knowledge in competition with its need for use as a medium of culturally functional knowledge. As Rosenberg notes, "no matter how self-conscious I may be in critiquing the social costs of logocentric thinking..., as a teacher I recognize that logocentric thought is precisely what my students need to master as a discourse which empowers them in the world" (293). But as the recognition of contingency has been valorized by the avant garde as "the initial condition for all creativity in intellectual systems and for life in physical systems" (294), the disorienting, displacing potential of hypertext may both allow and condemn it to function as an organ of epistemological, and ultimately cultural, change.

Bolter, J. David. 1991. Writing Space: the Computer, Hypertext and the History of Writing. Hillsdale NJ: Erlbaum and Associates, 1991.
Eco, Umberto. 1976. A Theory of Semiotics. Bloomington, IN: Indiana University Press.
--- 1979. The Role of the Reader. Bloomington, IN: Indiana University Press.
Landow, George (ed.). 1993. Hypertext: The Convergence of Contemporary Critical Theory and Technology. Baltimore, MD: Johns Hopkins University Press.

Ken Paradis is a Ph.D. student in the English department of McMaster University in Hamilton, Ontario. His dissertation, entitled Paranoic Hermeneutic Models in Pychon, Auster and Hypertext Fiction examines paranoic cognitive patterns of textually elicited interpretive response.

Go to Semiotic Review of Books Home Page
Go to SRB Highlights
Go to SRB Archives