Review:
Writing on the Edge, Special Hypertext Section. Volume 2, number 2 (Spring 1991), 78-159 + disk. Stuart Moulthrop, guest editor.

We live in the 1990s. Most of us consider ourselves logical, linearly thinking academics. Too logical, too linear, perhaps? If by chance or intent, we suddenly find ourselves immersed in an intriguing, interactive, constructive hypertext, our habits of linearity might tempt us to ask directions of the text's editor lounging nearby. Yet whenever we ask which way we ought to go, the editor, looking suddenly like a grinning Cheshire Cat, only says, "That depends a good deal on where you want to get to."

"We don't much care where--" we say.

"Then it doesn't matter which way you walk," says the editor/Cat.

"--so long as we get somewhere," we add.

"Oh, you're sure to do that," says the Cat, "if you only walk long enough!" Then slowly he fades from tail to head, leaving just a shadowy trace of his widely grinning grin. [1]

The spring 1991 issue of Writing on the Edge--acronym WOE--includes a special hypertext section guest edited by Stuart Moulthrop. At the least, this section is novel. At the most, it is a daring move, and one ahead of its time. The move arises from a desire of John Boe and Brian Connery, WOE's editor and founding editor respectively, to publish work about writers, the teaching of writing, and writing theory that edges away from predictable, comfortable, traditional journal articles about writing. This section of WOE also encompasses the hope that WOE readers, on the edge themselves just as much as the journal, will engage themselves with the whole of the special section, a whole that includes the Macintosh disk packaged with the issue, and begin to experience some of the excitement that certain of their colleagues feel about hypertext.

In a journal not solely devoted to computers and writing, this move is daring because it could possibly alienate certain groups of readers: those who have no patience with the feeling that they have fallen down a disorienting rabbit hole, those who feel computers have no place in the teaching of writing, and those who own computers other than Macintoshes. The section also risks being ignored even by computer enthusiasts because of the challenge involved in reading a constructive, as opposed to an exploratory, hypertext. Just as reading a hypertext differs from reading a traditional, linear text, so too does reading this special constructive hypertext section of WOE. Constructive hypertexts, first defined by Michael Joyce in a 1988 article in Academic Computing, are further explained by Stuart Moulthrop in this issue. Constructive hypertexts, says Moulthrop, are not "unitary documents," but instead, "multiple discursive sequences," a conception of writing

based not on circuitries of arrival/departure or 'exploration' but instead on pure extension or 'the lines of flight'. . . . Constructive hypertexts extend not just in material volume but also into the deterritorialized space of electronic writing. They are not closed books but open ranges, discursive improvisations that grant no last word. (p. 154)

Thus readers of WOE's hypertext section, in addition to reading its standard hard copy articles, must navigate through three nonstandard texts, two separate short stories on disk--WOE by Michael Joyce and Izme Pass by Carolyn Guyer and Martha Petry--plus the "decoupled" hard copy linear narrative "The Planes," by John McDaid, then weave back and forth between patches of all three fictions in order to construct a larger sequence encompassing all three. Readers will notice echoes in all three improvisations, which will gradually help them to understand the other fictions, while each still remains a separate "line of flight." Readers will also construct their own larger sequences in different orders and will attempt to weave together the three fictions' echoes using different reading strategies. But because there are gaps and indeterminacies in each person's construction of a larger "text," no one reader will have the "last word."

I have begun this review by raising cautions because I believe that this issue may be difficult, as I said above, even for some computer enthusiasts. Many of us, myself included, will have had only a limited experience with hypertext, and most of that experience will have been with hypertexts of the exploratory type, like HYPERCARD. The distinction lies in the fact that exploratory hypertexts, like books, are self-contained, autonomous texts. Stuart Moulthrop (1991) explains that

despite the appearance of 'interaction,' [exploratory hypertexts] are just as 'autonomously meaningful' as books. Or to be more accurate, like books they also preserve the discursive autonomy of their authors. The user may 'transform' the textual body by following alternative paths or linkways, but that body is protean: it retains its fundamental identity under all transformations. . . . The maze may have many permutations, the circuit many switchings, but in all of them the user still circulates through the same mechanized volume, working toward some predefined end. (p. 153)

Readers need to know that WOE's complete constructive hypertext section, which includes the disk, demands different reading strategies, so that they are forewarned. They need to know to expect the unexpected, to know that the three fictions at the heart of the section are not only autonomous texts isolatable from each other, but also texts meant to be attached to each other. Once immersed in all fictions and the journal's articles, however, most readers will feel the thrill of a strenuous, enjoyable, intellectual workout. The texts give nothing freely, but they return much to those who work hard at navigating their various cells.

For those with no prior experience at reading hypertext, the first pass through these stories is likely to be a dizzying experience. It was for me, even though I knew what to expect. Something about not being able to read linearly frustrates those of us who have spent so many decades of our lives reading from page one to the end and teaching more or less the same way. However, several of the individual hard copy articles provide background. A starting place for readers who want to work through these articles is Stuart Moulthrop's technical introduction for reading the stories; he explains how to click on the tool that enables readers to move around in the text from cell to cell, and how to switch from a cell's text to a schematic representation of the hypertext's structure at that point. Both fictions also contain explanations of the way to navigate through their texts--that is, if the reader can find them embedded in the map portions of the stories and then click on the proper tool that will open the text.

Besides Moulthrop's practical introduction, J. Yellowlees Douglas's (1991) essay "Understanding the Act of Reading: The WOE Beginner's Guide to Dissection" provides a clear, highly readable, entertaining explanation of the way one goes about reading hypertexts like WOE and Izme Pass. The act of reading, Yellowlees Douglas points out, becomes quite conscious when one reads an interactive hypertext fiction like the two here or Michael Joyce's earlier hypertext fiction Afternoon. Yellowlees Douglas (1991) presents a short background on theories about the act of reading, including Sartre, Iser, and Ingarden, discusses reading's cinematic parallels--the "virtually limitless collection of cuts and edits that remain nearly transparent to us" (p. 115)--and then establishes the connection to reading hypertext. Yellowlees Douglas (1991) does point out that unlike ancestors of interactive fiction, this newest type of text has no conventions such as chapters, no "seamless continuity between paragraphs within each chapter," and thus contains new difficulties, for instance, no traceable antecedents for pronouns (p. 115). We hold the texts in virtual electronic space and in the spaces of our memories:

Unlike Ford's paragraphs and chapters [in The Good soldier] , however, WOE contains narrative segments arranged in hierarchies, chains and configurations represented through spatial relations laid out in virtual electronic space. Like a Greek rhetorician who strolls through the imaginary spaces of his elaborate memory-palace rehearsing and rearranging his text, I experience STORYSPACE narratives by navigating through virtual space, by traversing places and the paths that links them. (p. 121)

The kinds of distinctions and parallels that Yellowlees Douglas makes, to The Good Soldier above, or to Lawrence Durrell's web of narratives, The Alexandria Quartet, help us construct a somewhat parallel context for reading WOE and Izme Pass and thus make the immersion into hypertext less disorienting for those of us who are novice readers. Yellowlees Douglas (1991) explains that reading a hypertext demands more of the reader, demands an engagement with the text, a more active involvement, than traditional text:

Without an investment in the text, I become lost. The sheer amount of indeterminacy lurking in many places as well as in the spaces between them demands that I fill in enormous chunks of detail as if I were wrenching a three-dimensional image from something supplied to me in a line drawing. In order to make any real inroads into WOE I have to acknowledge that I'm looking for something, and I need to have a reasonable idea what it is--otherwise, I have no real reason for moving from place to place and no means of gauging whether I should continue reading or give things a rest. (p. 121)

One piece that helps set hypertext in context, and one that raises, in addition, many critical questions in all the senses of the word "critical" is Johndan Johnson-Eilola's " 'Trying to See the Garden': Interdisciplinary Perspectives on Hypertext Use in Composition Instruction." Stuart Moulthrop notes in his introduction that this piece was not part of the original WOE project. Instead, it was solicited separately by the editors of Writing on the Edge. Johnson-Eilola's piece is a thoughtful, intelligent argument about the need for a critical examination of hypertext's relevance in composition instruction. He evaluates hypertext from two perspectives that are "normally mutually exclusive"--technology criticism and cognitive psychology (Johnson-Eilola, 1991, p. 93).

In the technology criticism section, Johnson-Eilola points out issue of politics and power in the use of a technology such as hypertext. Hypertext proponents often claim hypertext liberates students from control, linearity, and artificial academic divisions--thus, by extension, it emancipates and empowers oppressed groups in society. Such current writing about hypertext suggests that qualities inherent in the medium will bring about social change. Yet Johnson-Eilola (1991) cautions that hypertext itself causes nothing; rather, the teachers' successes will level social inequalities:

To view hypertext as the sole cause of the questioning of authority . . . is to ignore the social conditions that created many social inequalities in the first place. . . . Much current writing about technology and empowerment, especially in the realm of hypertext, appears to oversimplify the tangled nature of the connections between technology and society. (p. 96)

Johnson-Eilola then explains that hypertext was not originally conceived of as a tool for revolution but as a tool for productivity. Market-driven hypertext development, Johnson-Eilola contends, will lead to hypertext programs that cause readers not to question a text's authority, but rather to accept it.

Johnson-Eilola also points out dangers involved in large hypertext systems where readers can move freely from text to text, comparing, cross-referencing, posting references, and overcoming the limitations of a single text. Appearing seamless, such large bodies of information discourage readers to look beyond computer simulation for other information, causing us to mistake the quantity of information for the ability to gain knowledge, indeed to miss gaps and inaccuracies in information.

Cognitive psychology, Johnson-Eilola (1991) claims, "allows us to look quantitatively at the mental processes involved in writing and reading hypertexts" (p. 93) and to raise several valid concerns. Johnson-Eilola (1991) examines the two basic levels in the act of reading a text--the "microlevel" and the "macrolevel"--and shows how readers of linear text build macrostructures (p. 104). Hypertext could be used to aid readers who lack knowledge and navigational strategies regarding unfamiliar bodies of information; Johnson-Eilola (1991) says some "strictly textual hypertexts (those lacking graphical maps of the text structure)" cause readers confusion about their location in the text and "this uncertainty may interfere with their ability to assign importance to information they are attempting to integrate into a macrostructure" (p. 106).

Readers often initially have difficulty navigating hypertexts because the hypertext appears to make a "virtue out of lack of organization" (Johnson-Eilola, 1991, p. 106 as cited in Norman, 1988, p. 212)--that is, ideas and thoughts are juxtaposed at will, and no indication of any relationship between text segments seems to exist. Johnson-Eilola ends both the technology criticism section and the cognitive psychology section with a list of questions we must ask as we continue research in hypertext.

The heart of this hypertext issue, the computer disk, is easy to overlook. Readers must be willing to invest time reading the computer disk enclosed in the issue. The disk is crucial because it contains the two interactive hypertext fictions mentioned earlier, both created on the STORYSPACE hyptertext program. One of the fictions, aptly named WOE was written by Michael Joyce, and the other named Izme Pass was cowritten by Carolyn Guyer and Martha Petry. WOE and Izme Pass are stories in themselves. They are also synchronous stories, occurring simultaneously--referential, yet entirely separate. Each hypertext fiction contains both text and a map scheme--the reader can move back and forth between these elements. After first reading as linearly as possible, I finally discovered how to call up the story's map. By doing so, I discovered that my linear reading had caused me to miss out on whole portions of the fiction. So I then tried to settle on one section of the map at a time, clicking on sections embedded in that portion of the map and choosing one embedded section--which sometimes contained even more embedded sections--and then switching from map to text in order to discover whether I could make sense of the cells the authors had chosen to link.

Most of us will spend time with one of the fictions before moving on to the next. When we do move, a few similarities between fictions could trick us into forgetting that the second fiction contains other voices, other points of view. Some of the same characters occur in both stories, an unnamed man and woman, a she with "thighs brown as butter," and a he "with a buddha poof of a belly." Justin, Liam, and Philomena appear in both stories, as do similar scenes, phrases (stones, harem pants, transatlantic flights), and titles of fictional sections: "a," "mandala," "we." After spending time first with WOE and then turning to Izme Pass, I was expecting to hear from the "she" I had encountered, but I was surprised and pleased to find some passages from the perspective of the children. Thus, reading similarly titled sections from both stories opens our understanding of the characters in each story.

Sometimes, too, the understanding occurs not so much from what is written but from what is changed, added, or omitted. For me, the experience of reading constructive hypertexts is a little like reading poetry where a lot happens in the spaces between stanzas, in what the poet chooses not to say, and where we can notice a shifted point of view because of different image clusters. Likewise, what happens in the spaces between the cells of each hypertext fiction, and then in the spaces between the two fictions entwined with the decoupled narrative, determines how each of us will choose to read the stories. The section "We" in Izme Pass, for instance, contains a we that is more indeterminate than the group in the "We" section of WOE. Justin, Mommy, Dad, Steve, Filly, and me are not named, there is no holding of the children, no giving five, thus a less comforting resolution. Other readers may not sense what I do from the elimination of the names of these people.

WOE's special hypertext section contains more than I can cover here. I have said nothing about Carolyn Guyer's and Martha Petry's electronic mail communication discussing their work's creation, or the articles by Jay David Bolter and by Terence Harpold. I have not mentioned, except in passing, John McDaid's decoupled narrative. Suffice it to say, all gets more and more curious, more and more interesting. Stuart Moulthrop has put together a fine range of articles and stories that not only explains but also illustrates this new, fascinating area of electronic text. So for those readers eager to try their hands at reading an entirely different kind of text, I heartily recommend a plunge down this intriguing hypertextual rabbit hole.

Carolyn Handa teaches at American River College in Sacramento, California.

1. Adapted from Alices's Adventures in Wonderland by Lewis Carroll.

Joyce, M. (1988, November). Siren shapes: Exploratory and constructive hypertexts. Academic Computing, 3(4), 10+.

Johnson-Eilola, J. (1991). "Trying to see the garden": Interdisciplinary perspectives on hypertext use in composition instruction. In S. Moulthrop (Guest Ed.) Writing on the edge. Special hypertext section. 2(2), 92-111.

Moulthrop, S. (1991). Polymers, paranoia, and the rhetoric of hypertext. In S. Moulthrop (Guest Ed.) Writing on the edge. Special hypertext section. 2(2), 79-81 and 150-159.

Norman, D. A. (1988). The psychology of everday things. New York: Basic Books.

Yellowlees Douglas, J. (1991). Understanding the act of reading: The WOE beginner's guide to dissection. In S. Moulthrop (Guest Ed.) Writing on the edge. Special hypertext section. 2(2), 112-125.