The Hypertext Thesis: A First Experiment

Barry J. Fishman '89

10 December 1988

The files credited to BJF on this system comprise the first hypertext thesis. The subject of this thesis is the writing of Graham Swift, a young British novelist whose book, Waterland, is taught as part of English 32 at Brown University. The initial thesis was to show progression in the works of this one author. By looking at Graham Swift's collection of short stories and his four published novels I hoped to see a pattern of evolution and maturation over time. I proved my thesis in a number of ways, but the end product does not necessarily demonstrate this proof. Instead, what you will find on Intermedia is a large body of materials that deal with Swift's writing. These materials may be approached by the user in any one of a number of ways, only some of which definitively prove my thesis to be true. The reason for this lack of a clear path through my work is connected to my implementation of hypertext as a learning tool, and not a teaching tool.

The traditional English undergraduate thesis begins by making an assertion, usually about a narrow or limiting aspect of an author's work, which is then defended or proven over the course of about fifty pages. As I worked on my hypertext thesis, people would often ask me, "How many pages do you have?" I found this to be a confounding question, as I had no clear answer. I knew that I had written X number of pages, but the actual page length of my work in finished form is impossible to determine. The reason for this lies in the definition of hypertext itself. At its simplest, this was stated by Theodor Nelson in the early 1960's as "non-sequential writing." Along these same lines, Intermedia involves non-sequential reading. As a result there is no beginning or ending point for my thesis. My work is broken down into roughly sixty discrete units that can be read in any order the user desires.

The organizing structure of my work is very simple. There are two main entry points. The first is called G. Swift OV and provides direct access to overviews of Swift's major works, an overview of book reviews about Swift, some general material about literary techniques (created by other authors), and access to a number of papers written by me that have no direct connection to any one particular work of Swift's. The other entry point, WATERLAND OV is at a lower level than the first. It was created by Professor Landow, and then modified by me. This is the entry point best suited to students in English 32, who study only Waterland . There are a number of short essays written by me that are linked to this overview, and there are also a number of links made back to the top-level G.Swift OV as well as to overviews of Swift's other books.

I began my work in September, 1988 with this structure in mind, using it as a guide for my work. I realized quite early that no erudite criticism of Swift existed, only numerous book reviews. While these book reviews are at times quoted in my work, I in no way treated them as literary criticism. The wide swings of opinion from one magazine reviewer to the next formed an interesting parallel to an aspect of Intermedia that I found much more useful to an understanding of how hypertext should function. When listening to a professor in class or reading a great work of criticism in the library, it is very easy to accept the arguments made as fact, precluding a broader view of the topic. Intermedia circumvents this by providing the student with summaries of critical arguments that can be directly compared to competing philosophies. This has the effect of removing the authoritative voice from the material, placing the emphasis instead on the student's ability to think critically and draw conclusions from opposing theories. Although the book reviews I provide are not all-encompassing, they do provide a number of different opinions on the relative successes and failures of each work. The effect is that no work is "deified," that is to say that no work is placed in an unassailable position of greatness. The student is encouraged to read the reviews, easily accessed since they reside within the system, and then to draw his or her own conclusions about the success of the work.

After reading and thinking about Swift's novels and short stories, I broke my reading notes down into a list of subjects that I felt I could effectively deal with in under four pages -- the recommended length of the average hypertext document (The Rhetoric of Hypermedia, IRIS, 1988). I then began to create the documents that are subsumed under each major OV. As I worked I made notes, both written and mental, for "links" that needed to be made between documents. A link is, in short, a marker inside of a document that lets the reader access another document that contains relevant or useful information. After completing all of the book-specific documents, I proceeded to create several longer essays that covered topics I felt could not be assigned to just one work. The end result of this format is that there is a great deal of repetition from one document to the next. My intention was that a student who spends five minutes flipping through my work will get the same basic information as somebody who spends several hours. My own experience with Intermedia is that the longer I use the system, the less I actually read. Instead I learned to scan documents in search of useful information. More frequently, the links and "block extents," selections of text that are highlighted at a link destination, directed my attention to a very brief passage of a longer essay. The result was that I rarely read introductions or conclusions to essays, instead just looking for the information at the heart of the matter.

With this technique of using Intermedia in mind, the work that I had to do after all of the writing was complete became of utmost importance. Once I had transferred all of my files into the Intermedia database, I began to read through them all in order to make links. The linking proved to be a very difficult and frustrating task. The end goal is to provide links which satisfy the reader upon completion. If one sees a link marker in a document there is a whole series of expectations placed upon the information at the other end of that link. If a reader follows the link and finds information at the other end that is either of no use or is not what was expected, he or she will be let down. It was very difficult to make these links so that they would be useful to the reader because in many cases my own knowledge of the material precluded an open-minded approach.

Another troublesome point is that all links are bi-directional. This means that if it makes perfect sense to link between an essay on Shuttlecock and a document about detective fiction the relationship must also be clear to the reader upon linking in the other direction. The difficulty of this task is increased when the link takes you into the middle of a document, causing some amount of disorientation for the user. I was able to avoid some of this by employing descriptive "block explainers," the messages that appear after the document name when making a link, but this did not solve the problem completely. In the final analysis, the problem is not as severe as it sounds, but there is a potential for some difficulty, especially for novice users, in navigating through the system.

For the final stage of my task I familiarized myself with the contents of Intermedia that are already in place. Referring back to my own work, I found many points of common interest where a link could prove useful. I made many links between my own work and topics of general interest such as Literary Techniques, World War I, History, and London. What these links demonstrate is in fact the key to Intermedia's real power. No work stands alone in English Literature, it forms a part of a greater body of work. Swift's work in particular gains in significance if the reader is familiar with contemporary historical events and the literature that precedes it. The Intermedia user is made aware of these events through links to time lines, historical overviews, graphic documents, and essays on related subjects. The result is that the student learns to think about literature as an expansive subject that must be understood in relation to an infinite number of parallel events and trends.

The nature of this expansive mode of thinking is that my thesis could very easily be judged as unfocused. To make such a judgement would be to miss the point of hypertext and Intermedia. The hypertext thesis will certainly continue to evolve, taking many different forms along its course. My first interpretation of the form, an experiment in every sense of the word, demonstrates the importance of preventing a single intellectual argument from dominating a body of work intended from its creation to encourage critical thinking and understanding among its users. The job of the author in this type of environment is to first make his or her writing understandable and then to make it easily accessible by the various means provided in Intermedia.

In the end, the success of my work will be judged by the students who use it in the years to come. The existence of these students is what caused my thesis to evolve away from a very directed study of one aspect in Swift's writing into a more generalized approach. If a student begins in the G.Swift OV and reads first the document entitled "Progression in One Author's Works," then my thesis would be clearly stated and what follows (through links) becomes the support. If almost any other path is taken, the student is on his or her own in terms of drawing conclusions about Swift's work. This is not a shortcoming but instead a strength of the form. Although I have my own particular axe to grind, I can avoid imposing that personal vantage point upon the student's learning. Intermedia is in this way not so much a "teaching tool" as it is a "learning tool."

In conclusion, Intermedia's status as a learning tool was this project's initial attraction. Instead of writing a thesis that would be filed away at the John Hay Library never to be seen again, I have been able to create a body of work that will be actively used by students for learning purposes. My work is also a "living" thesis. Whereas most students finish their work and then put it away never to be worked on again, a hypertext thesis is always open to revision and addition. Just as I have built upon a basic set of materials provided by Professor Landow, so those who follow me will continue to expand upon my work. I feel very strongly that undergraduate learning should be conducted in an open and dynamic environment. Hypertext and Intermedia challenge by their very design a static approach to literature, encouraging change and new ways of thinking. It is this difference that becomes the greatest challenge to the author of a hypertext thesis.

United Kingdom