INTERVIEW WITH MARK BERSTEIN,
EASTGATE´S CHIEF SCIENTIST

Between art and science





Mark Bernstein at Eastgate´s
stand at the Conference


- Where are the Hypertexts? (Here is a tricky question)

First, it's important to remember how far we've come. We do have many fine hypertexts. The Web is wonderful, it's flourishing, and -- though much of the Web is only superficially hypertextual -- it's making great strides. Still, our virtual shelves are far too bare.

Where are the hypertexts? Some of them are hiding from silliness, from trivial arguments propped up as pronouncements and cultural edicts. Worries like "twitchy screens", being too much like TV, or the Bolter test (reading in the bathtub); these are hard to take seriously, but journals and journalists often appear to take them seriously.

Some of our hypertexts are lost in the chasm between the two cultures. We think that art and science should be partners, but we also behave as if we're shocked and upset if they come too close. To entwine technology and art seems strange, unnatural, and queer. Poets think nothing of expecting people to learn medieval Italian for the sake of Dante, but learning a smidgeon of programming is often regarded as an apalling imposition. We have to change this.

Some hypertexts are frightened of literary theory, or delayed because writers don't grasp immediately that writing is not trivially linear, even on paper. Narrative always loops and twists, we always write with tangles and mirror-worlds, with montage and feints. Hypertext gives us new freedom to do this, and at times the freedom to do well what we have always done with difficulty terrifies us.

- Why are most hypertexts about fragmented identity or about writing?

Yes, these are two common themes. But it's probably an exaggeration to say that "most hypertexts" are about them.

Still, these themes *are* quite common, and they also give rise to a strange coincidences. For a while, car crashes were ubiquitous (afternoon, Victory Garden, I Have Said Nothing, Uncle Buddy's Phantom Funhouse, Ambulance). Dismemberment is a common image too (In Small & Large Pieces and Patchwork Girl -- two works that are otherwise completely unlike;also Cyborg: Engineering The Body Electric).

Some of this is coincidence, of course. Sometimes it's just allusion. Of course, some aspects of hypertext -- its fragmentation, its propensity for loops, its ties to the machine -- also suggests these themes. And writing about writing is very much in the air in the late twentieth century: just as mid-20th century painting can be thought of as being "about paint", lots of late 20th-century writing tries to understand the quality of the medium.



Mark prepares his Keynote speech


- Why are our hypertexts so avant-garde and there are no more genres?

In part, by definition: today, unfamiliar literature is consigned to the avant-garde by definition, and tends to be rewarded in proportion to its novelty -- what Michael Joyce calls "inexorable nextness". We don't have hypertext genres, yet, because we don't yet have enough hypertexts. You can see hints of the formation of genre here and there --there's an interesting relationship between Arnold's "Lust" and Larsen's "Samplers", for example -- but genres form by accretion.

If by "genre" you mean the pejorative sense -- as in "genre fiction" -- the problem is technical: many currently-popular genres operate as reenactments of a ritual sequence, leading is afresh through familiar places to a resolution that we can anticipate but that, through the writer's craft, still delights and surprises us. The mystery, in particular, is a ritual: the world is damaged, and through these mean streets the protagonist must walk to repair the rent and to restore the world to a tolerable state. It's hard to square liturgical sequence with hypertext. This might well prove possible, but people attempting to do this have often fixated on mysteries as puzzles. They've been working on the wrong part of the problem.

- What do you miss in hypertext literature?

Mostly, I wish there were more of it! Where *are* the hypertexts? Yes, our virtual shelves are crowded -- far more crowded now than they were back in '87! But so much more could be done.

We have almost no hypertext writing in the sciences, in engineering or mathematics, even in computer science.

Our hypertexts tend to be serious, and they're often sad. That seems wrong to me. Where is joy? Where love? Where are the schlemiel and the schlemazel hiding? Don't get me wrong: seriousness is good and important. I don't especially want to go have a beer with the boys from _Godot_ or to go birding with anyone from _Macbeth_. But range is important. (I'd feel more comfortable, too, if I absolutely *knew* we could do this; pianists may not take great aesthetic satisfaction in exercises and scales, but it's good to know what you can command when you need to)



- What role does Eastgate play in this Hypertext world?

We stand between the candle and the stars.

On the one hand, Eastgate is a technology-driven company. We create hypertext tools, software that tries to let people write in ways that they couldn't write before. That means we're closely attuned to the computer science and software engineering on which progress depends.

On the other hand, Eastgate is a hypertext publisher. We look for fine hypertext, hypertext that can change people and that can transform the medium. We aren't deterred by challenging work.

So Eastgate stands in an interesting place, switching constantly from between tangible and intangible, between art and science, between pressing deadlines and decade-long horizons. It's not uncommon to switch from studying cardboard manufacturing one moment to designing new link algorithms the next.

As it happens, one of our hypertext tools, Storyspace, has come to play an interesting role in hypertext research; it's become for many people at this conference the baseline system, the general-purpose hypertext tool for which people reach first. You see Storyspace maps peeking out from all sorts of places here in Darmstadt -- in sociological textbooks and one-the-fly panel notes to hypertext fiction workshops.



Mark´s Keynote Speech Characters


- Eastgate seems like the perfect example of "bridging the gaps" between Science and Art. Why have you achieved a balance that the rest don´t seem to understand?

Keeping your balance is always hard!

Connecting Science and Art is what we do, it's the core idea of Eastgate. If we've succeeded, it's because of an exceptionally talented group of people, and an exceptionally patient group of investors and Directors.

- Why is a Conference like this useful?

Research conferences are still the best way for sharing new results and new ideas. Journal publication and book publishing simply take too long.

Another important part of The Hypertext Conferences is the chance to discuss new ideas with some of the brightest people in the field. Last year, for example, I wanted to adapt parts of Marc and Jocelyn Nanard's seminal MacWeb knowledge representation system to be the core of the new Storyspace/HTML export tools. I collared Jocelyn Nanard at the reception to ask for implementation advice, the kinds of details that aren't always included in papers but that can be really important in building systems people use every day. Instead, she told me I was approaching the problem the wrong way, and we wound up with an ad hoc conference among me, Marc and Jocelyn Nanard, and Daniel Schwabe. We laid out the design for the new HTML export facility in about an hour, a design that's turned out be be one of the nicest parts of the new Storyspace.

- Did you find any interesting new idea this year to work about? (or is it top-secret?)

It usually takes me a month or so to sort out the key idea or two. Peter Nürnberg's idea of "structural computing" is intriguing, and the demonstration of a collaborative reimplementation of VIKI built on top of a structure server was very striking. This suggests a number of implementation schemes for our new system project. The Writers Workshop created a huge bag of "things writers want"; I expect to spend many hours sorting through the list.

It's interesting, too, to see the way my 1998 "Patterns Of Hypertext" paper has been read, and the way this vocabulary (and the idea of patterns) is seen so differently by different people. In particular, there was a really interesting group of critical papers (Walker, Tosca, Calvi, and Rau all come to mind) and some very interesting workshop papers on pattern languages (Schwabe, the Nanards, Bieber). The two bodies of work have very different attitudes and languages; reconciling these approaches will require some careful thought .

- You have a long Hypertext-Conferences experience... how has the field changed over the last few years?

In the very early years, people were most worried about the navigation problem and about expedients (like standards and models) to deal with the enormous cost of creating the early hypertext systems. Both of these turned out to be partly chimerical: the navigation problem was mostly an illusion. Gypertext systems are no longer terribly expensive to make, so it's not vital to codify everything to some lowest common denominator.

What we all missed in the early years was that hypertext writing really would prove challenging, that it introduced interesting new questions of rhetoric and technique. For a time, this led to lots of tension between writers and engineers. This tension has finally dissipated, leading to the wonderfully cooperative spirit of this year's conference. Deena Larsen's workshop for writers, for example, was overflowing not only with writers but also with systems builders. That's a new (and wonderful) phenomenon.



Mark Berstein and Marc Nanard


- How do you see the future of this Conference?

It's an outstanding conference, probably the most consistently strong conference I know. People worried for a while that hypertext would disappear because it would be too easy and too obvious; that hasn't happened yet. We still have lots of interesting challenges.

I'm hoping to see more systems work soon, work that builds on new software technologies to explore some strange new ideas.

- Is the WWW´s importance shaping new approaches to hypertext? or do we live with our backs to the real world?

Very much so. The Web provides a wonderful array of documents and an incredible testbed for ideas. Mozilla may be very important to research in the next few years; it should soon be quite feasible for a researcher to create a very sophisticated browser (or other Web client) to explore off-the-wall ideas.

We need to remember what the Web is and what it isn't. The Web is not a user interface; we can (and must) get beyond blue links.

- Many thanks for your time and for having answered all our questions.


© Susana Pajares Tosca, 1999 for the text.
© Mark Bernstein, Jocelyne Nanard, Daniel Tietze, 1999 for the pictures.