Untitled Document

The words of my informants are indicated in italics, whereas other secondary references are enclosed in quotation marks.

Generalizability is neither possible nor desirable in a study of this orientation, which focuses on a human process that is rife with the unpredictable and the variable (Maykut and Morehouse 43-47). However, as Donmoyer suggests, rather than generalizability the value to be found in qualitative studies of this nature lies in the vicarious experience they can offer and possibilities they can suggest (192). However, while avoiding the positivist pitfall common to generalization, I do believe it is both the prerogative and responsibility of the qualitative interpreter to identify commonalities as a basis for theoretical speculation.

I identified, reported, and interpreted a number of contextual concerns at length in the original study (Mason, From Gutenberg's Galaxy). These included focused reflections on the Internet as a writing medium, non-linearity, multimedia, collaboration, "techno-angst," and new metaphors for the new writing space. I will focus here, however, on the particulars of the writing process as it manifested itself among my informants, the questions these findings prompted, and the way in which they suggested to me a new theoretical model.

Writing theory clearly shows that the writing process is so individualistic yet context bound, so full of variables, and so recursive, that it defies rule-governed certainties. This is why writing is so difficult to teach. Yet, interestingly, I found that the hyper-writers I studied, myself included, shared many similar habits or strategies. I also found that many of these tendencies drew on but were yet markedly different from habits commonly reported in a traditional writing environment.

The interpretations offered here might be considered an initial and very tentative response to the kinds of questions George Landow challenges theorists to answer—the very questions that were the starting point for my study. Landow asks,

The perceptions, habits, and strategies that I found most common among a high degree of my informants begin to answer Landow's and other similar questions and, of course, give rise to further questions. I offer my interpretations as a point of departure for theoretic speculation. I have gathered my analyses and interpretations into four principal foci:

The need to visualize both the "look" on the screen and/or the overall design of their hypertextual "docuverse" was arguably the uppermost and often first element that virtually all informants considered when faced with a hyper-writing task. (Docuverse is a term coined by Theodor Nelson to denote a hypertextual "document" in its entirety, including the web created by external links.) Different hyper-writers had different ways of addressing this challenge, but each one relied on some form of visual schematization. Terms such as storyboarding, mind-mapping, sketching, and drawing were commonly used to describe this early process. In a way, Mac spoke for all informants when she described the central importance of this strategy: Well, I think the first thing—and everything plays off this—is that you start essentially with a storyboard, much more the way you would approach putting together a cartoon or a movie than a simple piece of writing. Or as Pat said, the first step is graphically representing the web site as a whole and figuring out where the pages fit within the text and the layout (see Appendix A). Often in the classroom, in my first experiences teaching web design before I knew enough to suggest storyboarding, mind-mapping, or sketching, I would observe students instinctively drawing diagrams on paper interspersed with words.

My own experience hyperwriting is very much the same. Although I usually first give some consideration to my target audience, and mull over my purpose, or conduct some preliminary research, when I actually put pen to paper or fingers to keyboard it is to create some kind of visual schema. Unlike other writing tasks where I may begin in any number of ways be it by brainstorming, outlining, or even getting right down to writing an entire section of a document, when it comes to hypertext I need to visualize the bigger picture. I storyboard before I can move on to any of the particulars of content. Mary put it this way: I have to grasp the interface before I push the material.

My own habit is to rely much less on paper and pen now than I used to, going almost immediately to my favourite web authoring software, FrontPage 2000, which incorporates a storyboarding function (see Appendix B). I find that this feature is much more flexible than pen and paper, and working in an electronic environment seems to facilitate my hyper-writing process. Among my informants, most started on paper and some stayed there longer than others. At one extreme, Lee preferred to create her entire site on paper before she inputted it into an electronic environment, stating I can design a site on paper, and somebody else can publish it, and it'll still appear the way I intended it to. Nonetheless, Lee's process was still very visual incorporating both images and words, and envisioning words within a grid dominated by spatial logic (see Appendix C). Others, like Pat and Don—who both, coincidentally, used FrontPage 2000—moved onto the computer fairly early in the process. This prompts the question:

The acute need to visualize continued for most informants, including myself, as they worked at the macro level. Either simultaneously or after having worked out a rough overall design, it seems that their attention was drawn to the "look" of the site which, of course, meant the look of each page on the screen. Mary emphasized that you need to be aware of what the screen will hold. I find it ironic that although we could, theoretically, scroll down or up a web page indefinitely, hyper-writers seem to be far more concerned about what a screen will hold than about what a typical printed page will hold. Don described how he actually made decisions about what the screen would hold. I think it's a combination of good graphical design with colours that go well together where the pages are divided into separate zones. There's an obvious grid and it's very well-structured. You piece the page together like a puzzle. Don further mentioned that in his estimation the sense of a grid behind the page layout is one of the major differentiating factors between a professional site and a personal or amateur site. Mac had a rule of thumb for dividing a screen saying, no more than one-third of a screen should be devoted to text; the rest should be visual, and by that I mean even white space. In my own hyper-writing, I have come very much to visualize each separate page within a web as divided into zones or a grid. I normally consider those zones in terms of what is likely to be seen onscreen at any given point, and how that relates to a need for the reader to scroll down or up. Furthermore, as several informants mentioned, special attention needs to be paid to what will likely appear on the screen when it first downloads.

All these elements contribute to what several informants called "the look" of their web sites which might be viewed as a combination of aesthetics and visual logic—all, ideally, based on a gathering theme or subject. Most informants, in one way or another, acknowledged the importance of how a site looks from a reader's perspective and expressed a need to visualize that look early in the process. No doubt they had acquired this sensitivity from their habits as hypertext readers. As Don said, I think first impressions are very important. When you look at a site and your first impression is oh, that's a good thing, then you stick around and notice other features. In my experience as a writing instructor and an editor, I have noticed that writers working in a traditional printed medium normally give far less thought to how their writing "looks" on the page. We have been trained to expect words to carry the whole message. Even though this may not be entirely the case in reality, writers seem much more aware of the non-word elements when composing in hypertext.

My informant Pat, however, added an important insight that should perhaps inform the way in which hyperwriters make decisions based on the look of their sites. He declared that the needs of the viewer should supersede the look. I interpret this to mean that look without functionality will not be enough. Ideally, functionality and aesthetics co-exist in a complementarity from which the value of the interface derives. Of course, any web site is, in fact, a layering of interfaces. There is the browser interface that the viewer uses to navigate the Internet and web site at large, and there is the look and function of the individual website. In other words, there is an interface within an interface. This embeddedness could be compounded even further if an external link were to introduce another site within the frame of the website under view. If, as Nicholas Negroponte asserts, "the secret to interface design [is to] make it go away," how might the perceptions of hyper-writers about the "look" of their webs be judged relative to that axiom? Perhaps we need to inquire:

Furthermore, the need for writers who have previously composed in a text-only writing space to now develop visual or "artistic" skills alongside their language skills is unquestionably challenging. According to Steven Johnson, "we will come to think of interface design as a kind of art form—perhaps the [emphasis in original] art form of the next century" (213). This realization gives rise to the very immediate concern:

The look of his page actually preoccupied my Don in many ways, and he had a way of dealing with it that was apparently unique among my informants. He simply surfed the Internet until he found a site that looked generally like what he envisioned. Then he would just lift the source code—fire up an editor, and start modifying. Don claimed that everyone does it and I suspect he may be right, but to varying degrees. Pat noted that he regularly browsed through similar sites to get ideas, and perhaps he or others excerpted sections of a page, but neither he nor any of my other informants, except Don, claimed to actually lift source code in its entirety as a way of developing a web site almost from scratch. In this way, Don was able to compensate for his lack of expertise with the visual design elements of his site. The salient point, however, is that Don, like others, needed to visualize continually and he accomplished this to some degree by examining and playing with other sites. This, of course, brings up the interesting and difficult tangential question:

Although Rose, my "creative writing" informant was not concerned about the "look" of her page in the same way as my other informants, she did go through a decidedly visual stage in her process. She described how she worked to transform her short story, "Girl Birth Water Death," into hypertext: I divided the story into segments. I laid every segment printed out on paper on my living room floor so I could see how everything did or did not link, given any path a reader might take. The big difference in Rose's case unlike most of my other informants is that her narrative was essentially finished before she worked on the visual and physical structuration of the hyperlinked segments. From Rose's point of view, the hyperlinked structure is already built into the story. The hypertext lets it come out like it could not before.

Among my informants, there was, overall, a decided need to visualize all components of a docuverse in spatial terms during the hyper-writing process. I suspect that further research would show that this need for spatial visualization is decidedly more pronounced in hyper-writing than in traditional writing. I further suspect that this kind of spatial visualization may be a particularly crucial element of the early stages of hyper-writing. Many questions arise from the visual dimensions of hyper-writing. Researchers might ask specifically:

We must also allow for the possibility of a reciprocally energizing relationship between electronic media and print media as "written (printed) texts are also increasingly becoming multi-semiotic texts, not only because they incorporate photographs and diagrams, but also because the graphic design of the page is becoming an ever more salient factor in evaluation of written texts" (Fairclough 4). The point Fairclough makes here in 1995, before the Web could have had much if any influence on his thinking, suggests that the rise of the Web may be more a result of social than technological determinism. Perhaps the concern with "the look" of their websites, expressed by so many of my hyper-writers, is less in response to hypertext and more a result of changes in literacy that were already taking place when they came to hypertext. Hypertext may simply have offered my informants a convenient venue in which to expand their repertoire as senders of information to include the high degree of visual expression they had already become accustomed to as receivers of information.

This visual aspect of hyper-writing also extended to include a form of audience analysis that a number of informants expressed in different ways. They expressed a desire to envision the various paths their readers might take through their hypertext compositions. Jay described this kind of visualization of audience as a need to anticipate the kinds of twists and turns that your readership might take. Jay's and others' similar descriptions suggest a sense of audience that text alone does not provoke as they try to visualize their readers physically navigating their site. Somehow, a consciousness of the reader seems to be more present. Perhaps this is because the writer can see imaginatively how the reader might move physically through a hypertext, whereas traditional reading is more a mental activity with very little physical movement. Thus, again, visualization appears to be a key mental function embedded in the hyper-writing process.

According to Jay David Bolter, "In this shifting electronic space, writers will need a new concept of structure. In place of a closed and unitary structure, they must learn to conceive of their text as a structure of possible structures" (144). My informant, Jay, said more or less the same thing in a different way: I think the web allows for many different paths through a body of information. Formerly, I think a lot of things which weren't taxonomies were sort of retrofitted into a taxonomy to make them easier to process. Mary saw structuration as the writer's job to make a path through that information with logical connections. She elaborated on how a hyper-writer might conceive of this "structure of possible structures." I think you still have to give readers structure and closure, but you give them in different ways. Those ways could be navigational elements that always appear in the same place. It could be colour coding that lets them know they're here or there. If I'm going to leap off somewhere, I must give readers a way of getting back and I must bring this experience together for them. I think that it should be obvious; there should be a connection. There is a structure but it's a different structure. It's a hypertext structure. Landow echoes Mary's observation that hypertext need not lack the coherence which structure provides, noting "coherence may appear in new and unexpected forms" (186).

What Mary described is the need to create a navigational system; that is, to devise ways for readers to move among the various hyperlinked pages—both internal and external—that are the basic operatives of electronic hypertext. Writers must decide not only where to create and place links, but how to help readers to move comfortably among the linked information. Hyper-writers must also decide how much relative freedom to "allow" a reader in moving among those links.

Jay elaborated how the need to create this kind of navigational structure had affected his consideration of audience. I had to determine how people might look through the material and what people looking through the material might be searching for, and what kinds of questions they might have, to first and foremost anticipate the kinds of twists and turns that my readership might take—and you have to go there before they do. This is different, this trying to anticipate the thought processes of other people. This is a fundamental difference, and I think I only now understand just how fundamental a difference it was. I agree with Jay that there is truly a fundamental difference between audience anticipation in hyper-writing and in traditional writing. In printed text we know relatively surely the reading patterns our audience will likely follow and, commensurately, we control those patterns with relative surety. What Jay describes of his process is similar to Janet Murray's contention that "authorship in electronic media is procedural. Procedural authorship means writing the rules by which the texts appear as well as writing the texts themselves" (152). In straining to anticipate the kinds of twists and turns that my readership might take, Jay likely relied to a significant extent on his own hyper-reading habits.

Where to place internal links within a "text" becomes one of the most crucial decisions in structuration. For example, in this article I had to decide whether to link all in-text citations to "Works Cited." After careful consideration, I decided it would be an unnecessary distraction given that "Works Cited" is a major link in the navigation bars both top and bottom of the page. I was reasonably sure that any reader of this text would know enough to consult "Works Cited" for further details. I also felt that linking in-text citations would needlessly confuse readers: how would they differentiate between those links that pointed merely to bibliographic information and those that led to substantive supporting information? Also, should I provide links to individual researchers' homepages within the text, or should I provide these links only in the "Works Cited"? Again, wary of unnecessary and potentially annoying alternative pathways, I chose to provide those links only in the "Works Cited." At all points in the decision-making process regarding link placement, my notion of the reader's needs and preferences was the controlling factor. Thus, a logical point for further examination is:

How much control a hyper-writer can actually exert over a reader by manipulating structure is questionable. Certainly it appears that writers exert more control in a printed text because the intended structure is set within the confines of the physical artefact. Still, a reader is not forced to read any text in the order in which it is presented, but there is nonetheless a prescribed order. Hypertext, however, not only permits but invites the reader to take control of the structure of the text. Thus, we might agree with Bolter that "an electronic text permits the reader to share in the dynamic process of writing" (Bolter 5). Johnson-Eilola finds that a range of authorial control can be exercised even within a hypertextual structure. He speculates that

The general traits of hypertext organization . . . are not necessarily offered in all hypertexts. At one end of the spectrum, a hypertext may be so open—interconnected and reader-controlled—that a reader might be overwhelmed by the sheer number of choices offered by the text. At the other end of the spectrum, a hypertext might be so restrictive that readers find that they have no more—or perhaps less—navigational choice than they would have with a linear version of the text. ("Reading and Writing in Hypertext" 197)

The article you are reading now is certainly an example of a hypertext that is more towards the restrictive end of the spectrum. Individual sections can be read out of any particular order simply by clicking on a link in the navigation bar or in the site map, or individual pages may be accessed directly by their own URLs (Internet addresses). There is, however, an implied order that suggests the intended progression of an "argument." I suspect that because of the way I have structured this hypertext, most readers will, more or less, follow the suggested order. Have I created this structuration more because I sense it is expected of me in an academic milieu or because it really suits what I wish to say? I'm not sure; however, my ambivalence is not for want of asking myself the question.

In conceptualizing structure, a number of informants echoed Lee's observation that you have to think of your work as hierarchical. Generally, informants who emphasized hierarchy also stressed the importance of the home page. Pat noted, it's the home page that gives people the utility to move within your site. It's sort of a central area where users can go. Several envisioned the home page not only as the top of the hierarchy, but as a logical starting point for content development. There were others, however, like Mac who placed no particular value on hierarchy, and who pointed out that the home page may actually be the last thing you do.

In many ways, Pat represented those who tended to conceive of a web site's structure as hierarchical. He envisioned his information divided and linked in such a way that someone drills down from very summarized information at the top, down deeper and deeper into more and more concrete information. He saw it as his responsibility to make explicit the pathways that would lead the reader in that direction. In addition to hierarchy, Pat used the term layering to describe the way he went about integrating this structure into his docuverse.

Douglas Hesse distinguishes "the difference between summarizing and discussing" in somewhat the same way Pat sees "drilling down" from one layer to the next to retrieve ever deeper, more elaborate information. Hesse suggests that "the difference between summarizing and discussing versus presenting and linking is the difference between electronic texts being essayistic or not." Hesse further speculates on how the home page might function similarly to an essay since "like essays, home pages have the function of organizing and presenting a view of something (in this case, other documents, sites or images), and like essays they are developed and managed by a single author (or authorial entity), which distinguishes them from other types of electronic discourse" (Passions and Pedagogies 42). At the other end of the spectrum, Mac envisioned hypertextual structure as less hierarchical and more web-like where, although each individual segment requires structure, the overall structure and its various pathways allow readers to move freely through a thicket of material imposing their own structure as they proceed. Each of these views implies a different envisioned relationship between the writer and reader with corresponding differences in the writer's and reader's responsibility for meaning.

In my own experience, I find that I do envision a site as having a certain amount of hierarchy, but more among the higher level pages than the lower level ones. But I believe that even though an overall hierarchy may create a certain kind of macro level organization, there is definitely an associational organization that is superimposed upon that hierarchy—even in the most linear sites—that is not typically found in Gutenbergian print. Process-wise I find that I must think of my structure both hierarchically and as a web of associated links as I write, and that I work recursively among pages including the home page to coordinate all the links. Like Jay, I try to anticipate the twists and turns that readers minds might take as I create hyperlinks and build in navigation devices that underlie the structure of my site.

Take this page, for example; I had to decide whether to leave it as a single page or sub-divide it into individual pages based on each of its sub-topics. I felt that in this instance I wanted the kind of integrated sequential discussion that only a single page could render. I did not want to encourage readers to read only an individual section of this page because of how closely the sections relate to each other. However, I have created a certain amount of sub-divided structuration by inserting near the top of this page a table of hyperlinks that link directly to the sub-headings on this page. I've also inserted "back to top" hyperlinks at logical points along the way. And, of course, readers can always come and go as they please by using the main navigation bar at the top and their own bookmarks. I think structuration also depends to some extent on the nature of the writing task. For example, I am far more concerned with hierarchy in this sort of academic exercise than in other web sites I work on. I am very conscious that my readers may have varying comfort levels and expectations in hypertext and need a navigation system that presents a structure that is closer to a traditional one.

Informants often struggled with the challenge to create a navigation system that was both functional and aesthetically pleasing, suggesting an inherent complementarity between structure and "look," or what one might call embedded structure and apparent structure. Don's experience with one site illustrates this challenge. Don observed, one of the most challenging parts of doing a site is the navigation system. And in terms of knowing where you are in a site, that's one of the most important things in being able to find your way back and forth through the site. In this instance, Don noted that I first took it to an extreme by finding the most functional navigation system that closely resembled the Microsoft Windows directory structure. But in using that I totally sacrificed the look and feel of the site (see Appendix D). Relying on his habit of "recycling," Don then created a new navigation system by using a sysytem from another site in which he changed the colours a bit and then thought about how I was going to incorporate it into this site (see Appendix E). One can easily see from comparing the navigation system in the first iteration with the one in the second iteration how, even though both are probably equally functional, the second iteration is much more aesthetically pleasing and in keeping with the "look" of the site.

According to Bolter, "for writers of the new dialogue, the task will be to build, in place of a single argument, a structure of possibilities. The new dialogue will be, as Plato demanded, interactive" (119). Perhaps Mac best sums up the nature of the hyper-writer's challenge to manipulate structure when she notes that it's the overall structure that's different, and the way one moves through that structure. It is much closer to the way human beings think naturally, as opposed to the way they are trained to think, at least in this culture. In this culture we are trained towards linearity and logic. We are trained to trim down the extraneous, but I think we naturally tend to be curious and more encompassing, and this sort of structure allows you to digress, to enjoy side pathways that might not be able to be incorporated in a more traditional—at least Western traditional—work. A number of questions are immediately suggested by the challenge of structuration in hypertext:

Although Rose, as a storyteller, did not conceive of structure in the same way as most other informants, it was clear that hypertext's "structure of possible structures" was what drew her to hypertext. As she observed, the hyperlinked structure is already built into the story. The hypertext lets it come out like it could not before. Rose put her story "Girl Birth Water Death" in a drawer for about a year until she learned about hypertext in1993 because she didn't believe her narrative could be properly told in linear print. There are countless examples of "ur-hypertext," particularly in postmodern fiction, where writers have tried to superimpose a kind of hyper-linked reality upon the physical confines of the linear printed page. José Luis Borges's short story, "Garden of the Forking Paths" is a frequently cited example. As Lanham observes, "hypertext, a border and genre-crossing mode of writing, inevitably stitches together lexias written 'in' different modes, tones, genres, and so on" (260). Some very exciting research possibilities exist in probing the question:

It may well be that, as Myka Vielstimmig suggests, when it comes to hypertext "the coherence here is performative" (101. Vielstimmig's observation echoes Bolter who notes that "electronic text is the first text in which the elements of meaning, of structure, and of visual display are fundamentally unstable. . . . a kind of controlled movement. . . . a kaleidoscope of relationships" (Bolter 31).

As my own experience and the experiences of my informants show, it is very easy to get lost in the demands of technology or what a site looks like. And despite the oft cited aphorism that "the medium is the message," Mary pointed out what probably should be a kind of golden rule for any hyper-writer: Content is king, and content does not come from technology; it comes from the brain, the imagination, the ability of people to synthesize what they know and to create something. But to most writers content has always meant words. In the medium of hypertext, content is certainly still the most important consideration, but the role of words may be changing.

Mary's further observation alluded to that change. You write as though you were producing a drama or a film script. The writing has to be colourful. It has to be succinct and concise. You must make every word count because if you are going to show a picture - a good picture - you don't need to describe it in words. Your whole role, the whole role of text, changes. Since before the advent of hypertext, semiotician Umberto Eco was writing and lecturing about the need to develop the best complementarity possible between text and image rather than to view them as competing (http://www.italynet.com/columbia/internt4.htm). However, in the art or craft of what has been known until now as "writing," the image has been unquestionably subordinate to the word. (Nowhere is this more true than in most scholarly discourse.) If the experience of most of my informants is any indication, hyper-writing has already changed that relationship drastically and irrevocably.

As a creative writer Rose, however, had a singularly different perspective. She would certainly have agreed with Mary that content is king, but to Rose content was unequivocally words regardless of the medium. As she stated flatly, I need words. Rose would probably agree with Eco about the need to develop the best possible complementarity between words and images, however, since she noted that creative writers must be careful that the words and images work together. But Rose unquestionably saw the role of other elements as subordinate to words for, as she perceived, as a rule, creative writers are more concerned with words and sentences than other writers.

Don's attitude, however, reflects the apparent understanding of most of my informants who sooner or later came to the conclusion that normally in hypertext the operative principle is to minimize words. Many informants spoke about the need to write short(er) sentences or to use point form. Don said that he decided how much of his message could be conveyed graphically and how much textually depending on the type of message you're trying to get across. I know from my work what I've tried to do is condense the text. I try to get as straight to the point as possible in terms of text because I don't like reading paragraphs of text online. I don't think people are there to read. They're there to get information. They want their information as quickly as possible. And the graphics and audio supports that. I think that if a graphic enhances a text, or the text describes the graphic and gets the message across, that's good. But my principle is to minimize the text. I think in terms of space. I have to fit it [the text] into that little cell.

Don even manipulated his text to fit into his site after everything else was done, rather than starting with what he wanted to say in words and then figuring out more or less how to adapt the space to the message. He began by designing the layout space and then more or less said to himself, well here's what I've got in terms of space so whatever I want to put into words will have to fit into that space. Don readily admitted that he was much more concerned with how the text looked on the screen than what it actually said. He observed, I find that my biggest concern when I'm putting text into pages is how it looks rather than the content it contains. I've got to get it into that specific size cell in the table. And I'll sacrifice whatever it takes to get it in there, which is not good. Don was, perhaps, the most extreme among my informants in making the words absolutely subordinate to the amount of onscreen space he allotted for them; however, his experience is illustrative of a trend predicted by Bolter that, "texts written explicitly for this new medium will probably favour short, concentrated expression. . . . Electronic writing will probably be aphoristic rather than periodic" (ix).

Even Rose, the "storyteller," observed that she thought her writing had been getting noticeably shorter as a result of her experience with hypertext. I think I tend now towards being shorter—both the stories and the sentences and words in them. But who knows? It's hard to say. The dialogue in TV sitcoms may also have been a factor. Everything is short and clipped. Ya gotta be quick these days. Rose makes a very good point that provokes the question:

As a number of other informants either said or implied, words are almost like a last resort or are there more to support an image than replace it. Furthermore, words seem to be the last element with which most hyperwriters deal. As Mary said, "writing" is the last thing I do whereas before it was the first. The general opinion among informants was that if you can communicate your message in an image, still or moving, or an audio clip, then that's a better alternative in this multimedial medium. Often, as Pat remarked of his sites, the words become more plentiful and appropriate as the reader drills down in a site in search of more detailed or concrete information. It appeared there does come a point where a picture or sound, as appropriate as it might be, simply cannot say enough about a given subject to satisfy a reader's needs. Rose expressed a similar sentiment as it might apply to creative writing when we discussed how the role of words might change for a "creative" hyper-writer. She said, it is not enough to look nice, you have to do something with the story. You have to take the reader somewhere. One cannot help but wonder, to what extent can the role of words ultimately be replaced by other forms of media?

Furthermore, many informants appeared to approach language differently in terms of "correctness." The need that they expressed to fit words in a given space, or to condense meaning so that a reader is not unnecessarily burdened by onscreen reading, will likely have long term implications for our understanding of what constitutes "acceptable" literacy. Mary reflected on some of these new writing habits. When I'm actually writing language I use many fewer compound-complex sentences. I tend to write subject-verb-object. I tend to write in a much less formal way. I will use contractions. I will use incomplete sentences—whatever. I will use things I would never—as a former English teacher—have used. If I had been marking an essay like that before, I would have been appalled. But now in this medium I find that my writing has changed. And it has to happen. Indeed, if a former English teacher can accept this kind of dynamism in language, how can notions of what is acceptable not change?

The other area where content seems to be changing is a writer's need to generate the content in the first place. As Mary said, I'm not going to rewrite everything because there are those huge resources out there that are a link away. But even though the resources available on the Internet may relieve a writer of some responsibility in one area, they create new responsibilities in another. Mary elaborated, saying that the biggest difficulty is finding quality resources and you need to know enough about your subject to validate the links you choose. And then you have a whole new set of problems that were never there before in that you have to check those links every month. And then I'm always on the hunt for newer or better ones. And that is enormously time consuming.

The hunt that Mary refers to is a large part of what Pat described as the ability of digitalization to make finished unalterable text a thing of the past; this means that a piece of hypertextual writing can be forever in process. Not only do external links need to be updated, but the plasticity of the medium also creates the impetus to continually revise and update. Although the idea that "a text is never finished, but at some point the writer decides to quit" (Faigley, Fragments of Rationality 37) is not new to process theory, the idea that a writer might not be able to quit takes the concept of the "never finished text" to a whole new level. As Lee exclaimed when I asked her how she knew when she was finished with a hyper-writing task, well, you know you're never really finished! It seems that process and product are, in many ways, inherently fused in hypertext.

It is apparent that, at least in the experience of the vast majority of my informants, hyper-writing requires more than language skills and content is defined by more than words. Perhaps these additional skills are ones we all possess but which we have not developed as greatly as we might have because of our traditional habits of literacy. Perhaps these new skills are more finely developed in younger literates who have grown up in a more visually oriented environment or at least adults in a profession where a high degree of visual and aural literacy is required. What may be entirely new in hyper-writing, however, is the way language skills—or to use Howard Gardner's term, linguistic intelligence—must be combined with other skills or intelligences to maximize our effectiveness as hyper-writers. We need to re-examine the relationship between words and other forms of symbolic representation, and to evaluate the new role of words in this new writing space. Finally, we need to define how best to deal with a writing process that is so open-ended. To phrase these theoretical issues as a series of research-oriented questions:

As we struggle with the role of words in the content of hypertext, it is important to remember that as Lanham, among others, points out "although the Web has an increasing amount of images and sounds . . . its fundamental currency remains the written word" (152). It is a question of redefining the role of words. However, we must also recognize that the role of the word as symbol may be changing at a very fundamental level; we probably need to give particular consideration within hypertext to the the question Jay David Bolter poses of all electronic writing:

As with any new medium, certain conventions begin to develop over time. Using the printing press as an example, we know that many of the features we take for granted today such as pagination, indexing, standardized spelling, and so forth developed over more than a hundred years. These conventions generally evolved in response to both receivers' and senders' needs, and the ability of the technology to perform certain functions. Hypertext, especially in the context of the Internet, is no exception.

With the notable exception of creative writer Rose, for whom the issue of conventions simply never came up in the context of creative writing, many of my informants spoke about the need to include or arrange particular information within a site or on a page. There are certainly particular conventions we have come to expect, for example, on the home page of a site. Therefore, hyper-writers are increasingly aware of the need to include conventions such as a main menu on the home page. Likewise there are conventions we are now beginning to find on every page of a site such as an email link to the "webmaster," the date the site was created and its most recent update, and a navigation bar to mention a few.

Perhaps, as a teacher of hyperwriting at the university level, I am in the best position among my informants to comment on how these conventions that have come to be expected affect the hyper-writing process. I have surfed the Internet, read books, and tried to think of what I really want and need as a reader in order to create a list of common conventions for my students to use as a checklist when working on their web assignments. As a reader, I certainly appreciate the need for these conventions as an aid to accessing commonly required information, navigating through a site, and critically evaluating the information on that site.

You will notice at the top of this page, for example, that in addition to the journal's navigation bar there is a title that indicates the immediate focus or subject of the page. This is somewhat analogous to chapter or section headings. Immediately below that main heading is a navigation bar unique to this article that provides links to all the main pages in my article, including the home page. In addition to the journal's main nav bar, this information is included at the top of every page of this article in an area of the page known as the "header." I could have chosen to place my own navigation bar on the left-hand side of each page. I decided not to do so because this would have been best accomplished using frames. I did not want to use frames because not all browsers support frames and because a frame occupies relatively more space on the screen than a simple navigation bar at the top of the page. Nonetheless, the presence of a navigation bar at the top or to the left of the screen has increasingly become an expected convention.

At the bottom of this page is the "footer" which is also repeated on every page in this site. The footer contains a second navigation bar that matches the one at the top. I decided to include this bottom nav bar to save readers from having to scroll back to the top at the end of a long page should they wish to access the navigation bar. I could have simply included a "back to top" icon; however, I felt that including the entire navigation bar would better orient and serve my readers' needs. Additionally, the bottom of each page contains information provided by the journal, The Writing Instructor. This information is consistent across all pages of the journal's website with small modifications as necessary from article to article. Typically, information expected in a footer may now include dates published and updated, relevant contacts with email links, and links to feedback or discussion mechanisms. Finally, you will note that the copyright symbol (©) is also included. (The whole issue of copyright, which Negroponte calls "a Gutenbergian artifact" (58), urgently requires research.) These kinds conventions are becoming increasingly common and, therefore, expected on all web sites. As these kinds of expectations of conventions increase, including this information becomes critically important to enhance the credibility and readability of a site.

I have found that because so many of these conventions are detail oriented, it is very helpful for me as a writer to have a checklist to which to refer during my process. I have also found, as both a writer and a teacher, that because the medium is still so much in a state of flux I must constantly monitor and revise my checklist. New functions and thus conventions develop on an almost daily basis. Steven Johnson predicts, for example, that the time is not far off when hyper-writers will have "more than one type of link" (111) and "a better trail-building device to add to their repertoire. Two very practical questions to arise out of this issue are: