[The following essay is a slightly edited version of an article posted to the USENET newsgroup alt.hypertext as message <506@amanue.UUCP> on 1 Dec 91, in response to a query by Shawn Fitzgerald asking if there are any hypertext writers "out there".]
I am a poet who has been working in various non-linear poetic forms since 1966. My recent work is a series in progress called Intergrams, realized as HyperCard stacks. It is not simply hypertext but uses the concept of links along with other techniques as a continuation of work I've been doing for a long time. My early non-linear work might be described as polylinear; these pieces were called word nets and consisted of interlacing skeins of writing that wandered over a two-dimensional surface. Later I became interested in how to achieve word clusters, a direct analogue of the concept in music called tone clusters that I heard my friends in the music crowd talking about. Immediately this presented a problem: how to integrate word clusters into a larger syntactical structure. By juxtaposing words together with no structural relation at all, the contribution of the word itself (as opposed to context) to defining its syntactical place was completely disrupted. This led naturally to the idea of creating a separate channel to carry syntax. Thus began in 1968 a long series of works exploring a diagrammatic syntax. As time went on I became more and more interested in the diagram concept and less interested in word clusters. Several years ago I realized that with bitmapped graphics and a mouse, I could actually implement word clusters easily: with bitmapped graphics words could be physically overlaid -- juxtaposed in the same physical and logical space -- and as the mouse is moved to various hot spots one phrase at a time could be brought forward so as to become readable. Intergrams combines this idea with the diagrammatic syntax. The hypertext link concept is used to navigate layers in the syntax. The concept of a diagrammatic syntax may seem unrelated to hypertext, but I will have more to say about that below.
Intergrams is forthcoming in the fall of '92 from Eastgate Systems.
Let me move on now to a broader discussion of why I think poets in particular have been slow to make use of hypertext. (I don't feel able to speak about fiction from personal experience.)
Some may consider this a prejudice, but I consider it self evident that the different arts respond at differing rates to technological change. The first to grab hold of new technologies always seem to be the composers. This is not surprising; all music -- with the exception of a cappella singing -- is inherently technological; a musical instrument is a technological device. Composers have been quick to take advantage of high technology for centuries. As Andrew Culver pointed out to me, the organ was very high technology indeed when it was new. More recently, Cage enunciated the acoustical basis of music back in the 30's before electronic music was physically possible (that has always amazed me!); at the moment electronic instruments became possible, composers were using them even though they were far beyond what an individual person could afford. (My recollection is that the Columbia-Princeton electronic music studio cost something like $80,000 back in the early fifties.) By contrast, visual artists tend to wait until a technology is poised just on the verge of being a consumer technology before jumping in. E.g. for years the only video artist in the world was Nam Jun Paik, out there all by himself; only when video was just barely too expensive to be a consumer technology did we have an explosion of video art.
Poets, alas, are the worst. We tend to wait until technology has been around for a while before jumping on it. Although poets tend not to be very conservative politically, we do tend to be conservative æ sthetically, turning a blind eye to techniques that have been in use in the other arts for decades when we might make perfectly good use of them. Look at what happened in the early part of the 20th century. At almost exactly the same time Cubism turned the art world upside down and Serialism turned music upside down. The comparable movement for poetry would have been asyntactic poetry, but even today poets who have gone beyond syntax are in a tiny minority. No one considers it strange if a visual artist practices abstraction, but this is still controversial in poetry. There was of course Gertrude Stein. Robert Duncan acknowledged her influence, but still Gertrude Stein's influence on writing was nothing like that of the Cubists or the 12-tone composers. Gertrude Stein is like a stream that goes underground. (It comes up as Cage's music! I once asked Cage if he had ever written anything about Gertrude Stein, and he answered no, that she was "too close". Pretty revealing, considering that he has written about Duchamp.) No, the writer from that period who emerges as the influential figure is Joyce, and again the impact of Joyce on writing is not comparable to the Cubists or Serialists.
But give us time, things are changing. I've been amazed that most of the poets I know now have some kind of microcomputer or other to use for word processing, even though the stereotype of poets as having meager incomes is largely true. (However! The problem of poets not being able to afford expensive machines is still having an impact, since many poets who have computers may not have high-resolution graphics.) It may take a decade or two for serious amounts of hypertext poetry to emerge, but be patient, it will come.
Just because technology and commerce have set the table with toys does not mean that artists of whatever stripe will flock to play with them; to understand why any technology is or is not used in a particular art you must understand what is going on in that art. The post World War II era has seen a ferment in American poetics; but until recently that ferment has led in various directions that would not be conducive to hypertext. Let us look at some of the "schools" that emerged from the 50's and 60's. The three most influential groups of poets to emerge were (not in any particular order) 1) the Black Mountain poets; 2) the Beat poets; 3) the New York School. (A great number of other poetries have also flourished, but were less well organized into movements.) All three of these have had aspects that, if anything, lead away from hypertext. Projective Verse is hardly something that can be explained in a couple of sentences, but for me two of the key ideas Olson was talking about were Composition by Field (this is actually Duncan's term) and a kind of real-time image concept. Composition by Field means roughly that if a continuation occurs to the poet that works locally ("within the field") it should be applied without reference to any global constraint. I suppose this could be interpreted as consistent with hypertext links, but my feeling is that in practice this tends to reinforce linearity, not encourage departure from it. The real-time-image concept is a bit more obscure; the idea is that the time sequence of the poem should be a kind of image of the real-time act of composing the poem. Or in Olson's case, the real-time sequence of breaths should be the same time sequence as the natural prosody of the poem. Allen Ginsberg went even further in articulating a real-time concept of poetics, speaking about composing directly into a tape recorder. (He wrote a fascinating book called Improvised Poetics -- probably out of print -- which was the most interesting statement of poetics to appear following Olson's famous "Projective Verse" essay.) On the other hand, the "personism" of Frank O'Hara tended to deemphasize technical means altogether.
An immense cautionary tale, which looms large in the minds of many poets when alternate media, poetic forms, etc. are discussed, is the unhappy experience the poetry community at large had with Concrete Poetry. The concrete poets became prominent at roughly the same time as the movements I discussed above. Concrete poetry is difficult to describe in a sentence, but its practitioners use language elements visually, typically with the "language content" -- such as syntax and semantics -- completely stripped away. Many of the concrete poets were extremely polemical in pronouncing "straight" poetry (their term!) obsolete. Needless to say, this did not endear them to the rest of the poetry community. It was a tremendous tragedy. Instead of an amazing dialogue, that would have happened at roughly the same time as such ferments as Cage's Music and a whole succession of movements in art, what resulted was a total fracturing in which the concrete poets ended up talking only to themselves. To site a specific example: there is a point in "Projective Verse" where Olson discusses the typewriter as providing a precise grid for measuring the poem. The concrete poets could have stepped into this and challenged that there is a continuum available, why be limited by a grid? A lively discussion could have resulted. It didn't. Please think about this whenever you are tempted to make bald pronouncements to writers that hypertext is the future of language. That may indeed be the case, but it is a difficult challenge to educate writers to the wealth of new techniques that technology affords without disparaging the work they are now doing.
In recent years the emergence of the so-called Language Poets has changed the situation in poetics considerably. (My feeling that the term "Language Poets" must be prefaced by a distancing qualification is painful; I know many of these people, and it is arguably "my scene". But I cannot in good conscience condone the hijacking of the word `language' by a small group of poets, nor can I fathom the seeming acquiescence in this of the poetry world at large. The term has long since become an accomplished fact; still one must protest!) This movement has grown from the work of early pioneers Clark Coolidge and Jackson Mac Low. I will not propose to give a definition of it, but there are many aspects in the poetics of these poets that make it much more likely for hypertext to be accepted.
I do have a prejudice that the best art will come from a situation where the artist is preadapted to such concepts as hypertext by an ongoing æ sthetic. It is not so simple for someone used to writing one-word-followed-by-the-next designed to be read the same way to be suddenly handed hypertext and to make good use of it. On the other hand, writers who have been using various other kinds of experimental techniques that may seem unrelated to hypertext are much better adapted to the non-linearity of hypertext. If you would seek out hypertext writing, seek out experimental writers in general; if they are not currently working with hypertext, offer it to them.
I would like to now go into some detail about specific elements of poetics that I believe would predispose a poet to being receptive to hypertext.
By precomposition I mean composing a work in layers, where work in one layer may affect the entirety of the succeeding layer. Precomposition is widely practiced in music; this is one area where visual artists have taken the lead and have been doing precomposition for centuries. At one time making a painting meant scouring the woods for materials to grind into pigments, then building a support, applying the priming, applying various undercoats -- all before a single square millimeter of the final surface of paint was applied. Despite the great heterogeneity of American poetry, there has been almost a conspiracy of romanticism about the act of composing itself that has led poets to look unfavorably on precomposition. Many of the language poets do practice precomposition, and consider language as material in much the same way that a visual artist might consider paint as a material. An important part of the hypertext act is the linking of pre-existing materials. This does not sit well with a poetics that emphasizes real-time poetics or the spontaneity of the act of composition; on the other had it agrees very well indeed with an established practice of precomposition.
A hypertext is a fundamentally indeterminate work: the user is totally free to choose whether or not to follow a link, and the node sequence experience of each "reader" (we really do need a new verb here) will be different. Cage has opened music very wide indeed to indeterminacy; many dozens of composers have used various forms of aleatory methods, even though they may not agree with other parts of Cage's æ sthetic. One might even argue that the improvisational bent of classical jazz is a preadapting form of indeterminacy. While Jackson Mac Low stands out as having used various forms of indeterminacy for decades (he is a long-time friend and associate of Cage, and began using chance at about the same time as Cage did), the use of indeterminacy in poetry is not common. Indeterminacy in music tends to happen in the degree of control or choice in the instructions given to other people. (The musicians.) Poets are not used to giving instructions. Composition is a solo act, performance is a solo act. (Even in group readings, which are serially solo.) There have existed such traditions as the Japanese Renga, in which several poets would each in turn produce a line, but poems for multiple voices are not common. This is one area where there may be no obvious preadapting poetics; exposure of poets to indeterminacy is as likely to come from confrontation with hypertext itself as from anything I know of that is happening within poetry.
Perhaps this one is so obvious that it hardly needs mentioning. This is an area where the devastation caused by concrete poetry is particularly painful. There is no natural temptation for poets to experiment with diagrams or other forms of non-linear structure; even many of the language poets are amazingly conservative in the cosmetic "look and feel" of the poem. The difficulty of giving up linearity is something about which I can speak personally, having been doing non-linear work for 25 years. There is no doubt about it: doing non-linearly structured writing is a leap, an immense leap. That writers are not jumping in en masse is hardly surprising. I fear it will take an entire generation before it is as common as the advocates of hypertext would like.
"Traditional" hypertext entails imposing a non-linear structure on a locally linear substrate. Face it: this is an inherent asymmetry. The act of continuing to read on is simply nothing like the act of following a link. This may be of little concern to medical students making use of a hypertext medical information base, but to writers, who have an æ sthetic, it is a major concern indeed. Batting back and forth between ordinary linear reading and following links is a kind of bending back and forth that can produce the æ sthetic equivalent of metal fatigue.
There is one obvious cure for this problem: the concept of hypertext must be extended all the way into the fine structure of language itself, at least as far as syntax. This is an extreme position, a radical position. This is the place where my diagram poems began, more than 20 years ago; the atmosphere up there is so rare, the concept is so radical, that I have had to climb down from there to produce works that other people can understand. Still: if you truly believe in hypertext then you must be prepared for those who believe that we have not arrived at hypertext until we begin forming individual thoughts as hypertexts, thoughts that cannot be expressed in any other way. The hypertext link-following mechanism can be used to carry the actual syntactical relationships of thought. A diagram syntax is the first step in this process. There is no fundamental difference between diagrams and hypertext, except that a hypertext is a diagram where the diagrammatic links go off the edge of the page. Or to put it slightly differently, a hypertext is a virtual diagram with an unbounded virtual page.
One is somewhat daunted in making such proposals by the mass of evidence concerning the degree to which some parts of our language capability may be biologically based. Like a lot of poets, I am sure, I initially found the idea that the brain is biologically wired for syntax to be distasteful. Bearing in mind that any mixing of biology and the social behavior of human beings is potential social dynamite, still, the amount of evidence for some degree of biological basis for language is now nearly overwhelming. If the brain is indeed wired for syntax as we know it, how are we to achieve this fine-grained hypertext, this hypertext syntax? It will be difficult. Still, no one would deny that the conformation of our bone structure is largely biological, but that has not abolished the dance. To the contrary: we appreciate all the more keenly the work of a dancer who expands our ideas of what we thought the human body could do. If syntax is biological, it makes all the more valuable artistic work that will take us beyond this. That it is so difficult is part of what makes it so alluring.
To be candid, I expect few takers along this road, at least for a while. There is another method of dealing with the awful asymmetry of hypertext which I suspect that poets in particular will find much more congenial. We are accustomed to taking our word journeys on the sentence rail. By composing word groups that are static, that "don't journey", the way is open to using other forms of travel to navigate from one word group to another. The proposal above entails abandoning traditional syntax and using hypertext itself as the syntactical channel; this proposal would keep syntax locally but abandon the sentence as the vehicle for word travel. All travel would be by hypertext link. A poet would obviously be preadapted for this idea by a more centered kind of writing, possibly even a minimalist kind. Again, we are talking here about a specific æ sthetic bias, one which it is unreasonable to expect of any but a minority of poets, at least at first. (My current work uses both the more radical kind of syntax and this second idea.)
The culture of contemporary American poetry is an oral one. Those who care
about their poetry expect to be able to hear the poet recite it. Believe me: to
have work which cannot be recited -- because of its intrinsic structure
-- is a severe handicap. It means virtually renouncing "the scene".
Asking a poet to do this is asking a lot. I believe very fervently that McLuhan
had it exactly backwards, that speech is inherently linear, by virtue of the
real-time synchronization constraint between speaker and listener, whereas
writing, by allowing for diagrams, has the inherent capacity to be non-linear.
Perhaps the single most exasperating thing I have faced in my career as a poet
is the reaction to inherently non-linear work: "Well, I can't believe there's
not some way to perform this!" I get this all the time, from people who
should know better. The urge is just too strong, the sense that the community
comes together at readings is too overpowering for most poets to face
the idea of work that is unrecitable. I believe that this is one area where
only the technology itself will improve the situation. Yes, this is an
area where the lack of venue for non-linear writing hurts. It will take the
emergence of a network community of writers to overcome this problem.
And when it does come about, there will be an acute problem of making
sure there aren't two communities, separated from one another, the networkers
ghettoized from the oral poets. Again, it is CRUCIAL that those of us
who are experimenting with new technologies in writing do our utmost to avoid
anything that even smacks of hinting that there is something somehow "wrong" or
"archaic" about those pursuing traditional media. We experimentalists
need the traditionalists -- without the tradition, how can there be an
experiment? But the converse is not true. The traditionalists can go on quite
happily without us. The risk that hypertext writers will become ghettoized is
very great. It will take great skill to keep the lines of communication open.
Jim Rosenberg's home page o Poetics