Truth, Beauty, and Hypertext

by Robert Kendall

(also in SIGLINK Newsletter, Vol. 6, No. 1)
[cover date: Feb. 1997; publication date: June 1998]

Hypertext literature is a fascinating meeting place of art and science. The engineers and computer scientists have brought their precisely adjusted information tools to the table, the writers have brought their aggressively subjective imaginations, and a new genus of writing has emerged from the transaction. The technologists are happy to build up their user base and the writers are grateful for an offbeat entryway into new artistic territory, but one can often detect a certain uneasiness in the attitude of each group toward the other. The scientist, however much a fan of a good novel, may instinctively feel that the needs of artistic creativity aren't terribly relevant to the pursuits of technical research. The writer, meanwhile, even though engrossed at the computer screen, may be suspicious of letting the calculated, objective necessities of technology impinge upon the imagination. "Experimental" writing takes place in a laboratory unfamiliar to the sciences.

Nonetheless, this transaction between art and science should transpire with more than just a grudging acknowledgment of the deal's practical benefits for both sides. Both parties share more deeply rooted concerns than they may realize. Let's consider the scientist's perspective first.

In his book Machine Beauty: Elegance and the Heart of Technology, David Gelernter argues eloquently that the scientific endeavor is inextricably linked to aesthetics as it pursues simple yet powerful and elegant (i.e., beautiful) principles to explain things or make things work. Even the physicist must rely on an eye for conceptual beauty to help determine whether an untested postulate is likely to prove true. Gelernter declares that computer science too often looses sight of aesthetics, "banging its head against the wall . . . in an effort to put programming on a mathematical basis," an approach that has led to clumsy, overly complex, hard-to-use software. He insists that "in the computer world, beauty is the most important thing there is." Computer scientists must cultivate above all an intuition for the beautiful--an aesthetic sense similar to the one that guides artists--if they are to arrive at elegant and therefore truly useful products.

What does hypertext literature do for the engineer? It pushes software into the domain of art, where the demand for beauty (not to be confused with surface prettiness) is absolute and unforgiving. Hypertext literature becomes a uniquely rigorous test bed for interface and systems design because it heightens the aesthetic expectations of the user. A satisfying reading of a hypertext story or poem depends upon how successfully all the details of the reading experience fit together into an effective whole. Navigation becomes not simply a means toward an end but an end in itself.

The hypertext designer typically tries to meet the traditional needs and expectations of readers of nonfiction information sources and reference materials. Such readers are likely to want to search for specific information or browse through parts of the work and may have no desire to digest the entire corpus. They may expect to skim or skip over much text that is of little immediate interest. Hypertext facilitates browsing and information retrieval by minimizing readers' contact with text that's irrelevant to their current needs.

Literary readers, on the other hand, come to a text expecting (or at least hoping for) all of the reading experience to be of equally high quality. There's little tolerance for having to push aside writing that doesn't contribute effectively to the current state of a story's unfolding. Distracting navigational complications can tarnish the reader's memory of a poem, leaving a stain that isn't erased by the eventual arrival at some rewarding textual destination. The reading itself--not just the text it is based on--must be a work of art. Every navigational choice has to count.

Accommodating the desire of the literary reader for a perfect reading may seem a tall order for hypertext--perhaps even an attempt to bend the medium to purposes for which it is ill suited. Yet applying the technology to literature doesn't alter one of the primary goals of hypertext, which is to put at the reader's fingertips the text that will be of most interest at any given moment. On the contrary, the demands of literary hypertext compel us to keep this goal more clearly in view.

In nonfiction hypertexts, the integrity of the overall reading may not be paramount, as it is in hypertext literature. The specific nuggets of information that the reader comes away with may be more important than the details of how they where obtained. Yet an artfully shaped reading is always likely to be more rewarding, more efficient, and more memorable than a clumsily structured one, even if both encompass most of the same text. Ensuring an artfully shaped reading requires putting an artist's sensibility--an understanding of the aesthetics of writing--behind the system that will guide the reader through the text. Just as a novelist's sense of pacing or a poet's skill with metaphor can often benefit nonfiction prose, so too can a literary aesthetic often benefit the design of a general-purpose hypertext delivery system.

The value of the literary viewpoint to hypertext research runs deep because hypertext belongs to the domain of conceptual rather than hard science. For the same reason, the benefits flow in the other direction as well. In my next column I'll look at what computer science can do for the poet or fiction writer.

Reference

Gelernter, David. Machine Beauty: Elegance and the Heart of Technology (New York: BasicBooks, 1998).