I chose to write about hypertext in hypertext partly to see what the potential of the medium for argumentative rhetoric really is. If it could do some things as well as or better than linear prose, I reasoned, my questions would be at least partly answered.
In writing this text I found that I was constantly walking a very fine line between suggesting a default path for the reader (which seems to me to destroy much of the purpose of hypertext) and leaving the reader to bumble through a cloud of random associations.
Unlike print forms, there aren't very many models to work with in this form. I found myself most heavily influenced by Kolb's Socrates in the Labyrinth, which uses a web format in which usually more than one link runs to and from each node. I have suggested which nodes I think might be productively associated without leading the reader too obviously. Different readers with different interests will build the text differently, though I suspect that long habit will lead most readers to try to "exhaust" the text by tracking down every node. (This will be difficult when some of the links lead out to other parts of the WWWeb, each with its own links to yet more texts.)
The text took shape around several centres of gravity, sites of attraction if you will, that tended to pull certain nodes into closer relationship than others. There is a loose cluster of nodes about hypertext, argument, and form. There is another on the effects of hypertext on both readers and writers. There is a third on the implications for teaching. In addition, there are some that seem to be at a higher level of generality which speak about hypertext in general. But most of the nodes in each cluster seem to connect across to nodes in other clusters as well.
One choice that I found I had to make was how much redundancy to tolerate. Key phrases and ideas turn up in a number of nodes. In linear text, the game is to repeat an idea in such a way that it is obvious that you know you are repeating it for a sound rhetorical reason. When you can't be sure whether the reader is encountering an idea for the first time or the seventh, this strategy is denied you. I settled on allowing a small amount of redundancy in the nodes themselves and counting on readers to follow links back and forth to generate their own redundancy. I am hoping that readers will traverse some of the more key nodes a number of times, finding more in them each time they come at them from a new direction, rather than saying "Dammit, I've already read that one."
What, I now ask myself, is the point? Surely there must be more to this than the game (rather inconsequential it seems to me) of illustrating postmodernist assumptions about the instability of texts.
I can't say for certain what the point is for readers, but I have found certain interesting effects as a writer. I have found that writing in this form makes one resist closure. Every node is somehow questioned, extended, and deconstructed by some other node. The relentless drive toward a conclusion, even a tentative one, that print texts seem to demand is undercut by the demands of this new form of text. Whenever a series of nodes seemed to be working their way toward a final-ish sort of claim, I found myself deliberately looking for competing options, finding opposing viewpoints, or writing metatext that would question the text I was writing.
This seems to be an interesting twist on Richard Coe's assertion that form is heuristic--that certain forms focus the writer on certain modes of thought. The five-paragraph theme, Coe notes, has spawned generations of students who think that there are exactly three reasons for everything. The hypertext, I find, spawns a mindset that questions everything, sets everything in opposition to everything else. It spawns questions, resists answers.
This might be one answer, then, to my question of what role hypertext might play in discursive rhetoric. When one has a specific claim to make, hypertext may not provide much advantage over linear text except for an ability to embed longer quotations and handier references. But when one wants to explore and to question, the more radical forms of hypertext help one think (not merely write) in an exploratory mindset.