Every discussion of electronic technology and hypertext should begin with the qualification that due to the extreme fluidity of this medium, evaluating the state of hypertext and the explosion of on-line scholarship is a bit like evaluating a raging river while one is bobbing around in it. In such a situation the frame of reference is, of necessity, going to be hasty, blurred and instantly outdated.

That said, I would like to offer my own perspective on the relatively recent movements towards doctoral dissertations written using hypertext. My title alludes to the short and amusing essay by Walter Benjamin called "Unpacking My Library", an essay in which he discusses his unregenerate fetish for books. In any discussion of information technologies, especially a discussion in an academic setting, the fetish character of printed books must be kept in mind; all too often electronic media are dismissed as technological embellishments to stable, originary texts by commentators who seem unaware of the technological basis of book production and distribution. Comments about the fetish character of electronic media are themselves couched within a discourse that implicitly regards print technology in naturalized, reverential terms. Sven Birkerts' 1994 Gutenberg Elegies is a particularly vehement example of technology-bashing in which the printed text has curiously transcended the bounds of its mechanical production -- Birkerts refers to the book as "the site of veneration"[1] and "a space sanctified by subjectivity".[2] Accounts such as Birkerts' insist that print is closer, somehow, to the essence of humanity than are the virulent and parasitic forms of electronic media. If we are to take seriously the potential of hypertext as a medium for academic discourse, a reminder that the printed book is already a form of information technology is a necessary step[3] in opening up a space for serious consideration of alternate media possibilities. Benjamin, as an early commentator on the technical aspect of textual production, recognized the tendency to regard books as natural objects while at the same time insisting that written texts, like all media, are constructs of a specific social and technological order.[4]

Allusions to Benjamin's legacy aside, I use the word "unpacking" in my title to indicate the need to examine with care the function of words such as "television", "hypertext", and "books" in our discourse. Such terms have with time assumed a metonymic function so that, for example, "television" refers at once to an individual consumer product, to a much more diffuse concept involving broadcast and network technologies, and to an entire cultural form conventionally disparaged in relation to other cultural products such as music and literature. If Sven Birkerts' response is any indication, text on screens suffers from the stigma of the medium; it becomes axiomatic that electronic texts are frivolous, while serious thought requires the solid linearity of print. Indeed, recently on the electronic discussion group Humanist the subject of reading from the screen was discussed,[5] and several participants, while far from hostile to electronic texts, said that they were so habituated to paper that whenever they encountered interesting material in electronic form, they printed it out and read a hard copy.

What I want to suggest is that a true hypertext can't be read off-screen. It must be experienced as text on a TV monitor. While it is technically feasible to print out the contents of a hypertext, such a print version suffers the same result as a filmed version of a novel: inevitably, changing the medium changes the text. The complex theoretical issues raised by translations from one media to another are well-known, and while media theorists since McLuhan have examined the metaphysical and perceptual distinctions between the material word and its electronic counterpart, such complexities are beyond the scope of the present essay. For the purpose of this paper I want to propose that perceptual issues be bracketed, much as they are bracketed by everyday users of e-mail or the on-line edition of Bowker's Books in Print. I wish to make a lesser claim, that the process of engaging an electronic text is qualitatively different from the experience offered by print technology.

A hypertext, from this pragmatic if undertheorized position, is a special case of an electronic text; it may be confined within a single computer's memory or distributed over a network, but ultimately its perception relies on the palimpsest-like qualities of the TV screen. The media interface of the printed page cannot accomplish the transparent transfer between linked elements in a hypertext. It is primarily this ability to move around freely within a decentred space instead of along a fixed, linear path that constitutes hypertext's most significant divergence from print technology.

What are the consequences of using such a linking system to present the results of doctoral thesis research? I would like to offer a few observations on the effects of a shift in medium, drawing on my own dissertation and a number of other publicly available hypertext projects currently in progress.

My doctoral thesis on the work of two postmodern novelists, Martin Amis and Don DeLillo, explores the manner in which discourses of technology inform their work. My motivation for moving a part of the thesis to hypertext resulted from the realization that an increasing number of useful resources and texts were becoming available on the World Wide Web, and once I began to think about what a hypertext version of my thesis might look like, other possibilities suggested themselves.

In "Re-engineering the Dissertation", a prefatory piece to his own on-line thesis project, Jonathan Yordy suggests that we might think about electronic dissertations in the following ways:

1. An electronic analog of non-electronic processes. This might include putting a dissertation on the Web. Instead of typing the document, it would be marked-up using HTML and posted on the Web.
2. A value-added electronic dissertation. This might be characterized by chapters that are marked up, with hypertext references to source materials and other ideas.
3. The re-engineered dissertation. We'll find out what this means as we re-engineer it (learning by doing)... [6]

And he ends with that ellipsis. While I have no clear idea of what an entirely re-engineered dissertation might look like, I am convinced that a certain amount of HTML tagging and an internet connection adds value to my thesis project. For my purposes, one of the greatest advantages to an on-line hypertext is the capability to link bibliographic references to the entire text referred to. A reader interested in, for example, a point drawn from John McClure's essay on Don DeLillo's novella "Pafko At the Wall" could pause to read it in its entirety, or examine the full text of an on-line interview with Martin Amis to get a better sense of the context in which a quoted passage appears.[7]

Providing an electronic dissertation with a keyword-search function would facilitate the readers' ability to read through the material from a certain thematic perspective, or perhaps enable them to isolate information specific to their own research interests.

Word searches, an on-line glossary, buttons to facilitate movement through the pages of linked information: while such devices explode the text and reduce the contents to a collection of fragments, the prospect of a thesis in fragments does not have to be a source of anxiety. In fact, I contend that this model can facilitate the kind of reading that already occurs in an academic environment. As a researcher, my approach to a work of critical theory or literary interpretation rarely involves following the linear structure of a book. Since it is unlikely that someone else's research is, in its entirety, applicable to my own interests, I check the index and the bibliography, scan the table of contents, and then read those chapters or sections that seem relevant to my research interests. In a book on naturalism in American literature I may only read the chapter on Don DeLillo, thereby treating the text as a collection of fragments despite its putative status as a "linear" text. My feeling is that the kind of reading encouraged by electronic texts which incorporate internal and external links enables the extraction of specific fragments of information and thus parallels the actual practice of humanities research.

The speed and apparent seamlessness of linking strategies assist the writer in expanding parenthetical references that in a conventional dissertation would soon grow cumbersome.

In principle, the hierarchy of primary material/footnote material that exists within a linear, print-bound volume could be entirely eliminated in a hypertext, as the parenthetical elements multiply until they overwhelm the primary material. In such a case, the concept of a thesis is itself undermined. The decentred text is rendered less and less coherent and its argument is dispersed as one aside leads to another. While there is obviously a question as to whether such an outcome is desirable, it appears that in the advanced stages of this process, the (hypertextual) medium has gradually become the (hypertextual) message. If rigorous argument requires a strict, linear, causal logic, then hypertext undoes such rigour, putting in its place a series of related propositions that can be followed by a reader in any order -- or not followed at all.[8] Perhaps at this extreme, where a rhetoric of mastery or "authority" is replaced by a rhetoric of partial knowledge and competence, we glimpse what Jonathan Yordy means when he speaks of "re-engineering the thesis".

My own thesis, presented as a series of chapters on related aspects of new media and literature, participates in the sharing of power with the reader. The table of contents represents one possible ordering of the thesis, but through hypertextual links the reader is invited to approach the material in other ways, moving through the textual space with reference to a single work or a specific theme.

Because both the authors I'm interested in write about the impact of cultural products other than books, hypertext allows me to import examples of other media -- photographs, video stills or clips, and sounds -- and integrate them into the dissertation. In a hypertextual dissertation my references need not be limited to books, just as DeLillo and Amis, the authors I'm working on, respond in their works to many media sources in addition to literature. In my thesis I am trying to argue for the importance of non-print media to a reading of Amis or DeLillo, so moving to a medium where these examples can be displayed is clearly advantageous.

When I began to think seriously about writing a dissertation incorporating the non-linear possibilities of hypertext -- seriously enough to propose a conference paper on the topic -- I believed that there was a certain novelty to the idea. Of course, that was a year ago. There is now a web site, maintained by Matthew Kirschenbaum[9] at the University of Virginia, devoted to the subject of hypertextual theses.

In May 1996 it included a list of about a dozen dissertations, four of them in English literature. In the past twelve months UMI, the company that archives and distributes thesis and research publications in North America, has announced a willingness to publish dissertations on CD-ROM -- which is simultaneously a recognition of the viability of electronic dissertations and UMI's recognition that on-line publications threaten to render obsolete their North American monopoly on thesis storage and reproduction. Universities are scrambling to draft guidelines for submitting electronic theses, though none in Canada has gone as far as the Virginia Institute of Technology, which has announced a policy requiring all graduate students to submit theses and dissertations in electronic format in order to facilitate network storage and access.[10] At the University of British Columbia, the Faculty of Graduate Studies has developed a policy on "alternative thesis formats" which tries to strike a balance between the possibilities of new media and the requirement that "documentation of the research or inquiry process must be provided sufficient to allow others to follow the line of reasoning and evaluate the credibility of the work".[11]

I am tempted to conclude that the genie is out of the bottle, although I do not believe that the Virginia Tech. policy is a particularly useful development. It does not seem that insisting that all theses be submitted in a computer-readable format is any more enlightened that insisting that none be; such a directive is as arbitrary as those that deal with acceptable paper quality or margin width. Nor would I flinch if electronic resources evaporated and it became impossible to complete a hypertext version of my thesis. In a print-bound form it would be a different document, but it would not disappear as a result.

I want to close this paper by mentioning some of the problematic aspects of hypertext. The major area of concern, as I indicated in my opening paragraph, is the rate of technological change in the field. Very little of the software I use for writing or displaying HTML documents existed when I began my Ph.D. four years ago. The phenomenon known as the World Wide Web has only been around for half that time, and new paradigms for cyberspace and their software embodiments are springing, fully-functional, from the heads of programming gods all over the place. A decision to create a thesis in HTML optimized for viewing using Netscape Navigator will probably seem quaint or perhaps willfully perverse in ten years -- or five, or two...

There is also the possibility that even if a hypertext thesis survives in an accessible form, the various external texts to which it is linked may not -- and it may be impossible to relocate them. In this regard the materiality of printed works does have its advantages.

Issues of long-term storage and accessibility are hardly unique to universities. Such problems affect all electronic documents, as archivists grapple with the need to establish standards for the new media. The truly destabilizing aspect of hypertext concerns its implicit challenge to the current view of what constitutes a thesis. The logic of hypertext documents is a logic of contingency, at odds with principles of linearity and necessary sequence, and thus with the current idea of what a thesis should do. What will happen, I think, is that the nature of the dissertation -- which at present is both an intellectual endeavour and an institutional validation -- will alter to better reflect the collaborative and intertextual nature of all research. The Ph.D. is, after all, a mutable form that can be, and has been, altered over time to serve the needs of universities and the wider social order.

I am prepared to accept the risk of participating in a medium characterized by impermanence and constant revolution. On the internet, research material becomes accessible to and distributed among a population of fellow researchers far in excess of what one could expect for most print-bound documents. In time more and more academics will become familiar with the resources available on-line, and realize that, as media theorist Friedrich Kittler says, "all books [including theses] are discourse networks, but not all discourse networks are books".[12] After all, as a recent article in the Chronicle of Higher Education put it, the discourse network of hypertext "isn't about getting away from writing. It's about extending writing and giving the reader a richer, more interactive experience".[13]