Having studied the Web since 1996, I've been bothered by how Web discourse often seems to follow wherever Web technology leads it. One familiar example is the blink HTML tag, which thankfully is now rarely used. This tag offers a superficial technological solution to the rhetorical problem of drawing readers' attention to important bits of text, but it's a cure worse than the disease, and it not only distracts readers but also distracts authors from solving their rhetorical problems with rhetorical means. So I'll be exploring how Web discourse, rather than functioning as rhetoricians might have hoped—to achieve meaningful purposes with, perhaps, world wide audiences—seems instead, like the dog being wagged by its tail, to be significantly wagged by the Web "look," the Web form.

In this presentation, I . . .

1. start by introducing a terminological metaphor I'll be using to illustrate my argument,
2. offer a brief theoretical account of how form can shape content, and show how such a theoretical model applies to the Web;
3. explore a few cases of such shaping that I discovered in an empirical study I conducted;
4. point to the implications of this shaping in teaching, and in professional Web authoring.

Let me start with a terminological distinction. The relatively new concept of "information architecture" has, with its validating metaphor of architecture, accorded some needed legitimacy to a newly recognized field of Web authoring. By contrast, the relatively less prestigious field of writing, of composition, has contributed relatively few metaphors. For instance, in current usage, "Webmasters," not Web "authors," "construct," not "compose," for the Web, and constructing Web "sites" sounds more ambitious than constructing Web "pages." The more prestigious metaphors of architecture and construction have been crowding out potential alternate metaphors from writing.

Along these lines, to capture the current state of Web discourse, I'd like to introduce what I think is an unfortunately suitable guiding metaphor, revealingly drawn again from "architecture": "archi-texture." I will speak of the technology and tools of the Web as the technologically inspired architecture and of the discourse of the Web as the derived archi-texture. The Web's composition, the Web textuality, has, to a significant degree, emerged out of its architecture, to the extent that, metaphorically, Web archi-texture is an echo of, a rhyme of, almost a homonym of, Web architecture. Archi-texture thus captures the techno-textual duality, the derivative nature of Web textuality, and its subservience to Web architecture.

Such a leading role for architecture has been proposed by Richard Coe (1987) in his work on form. Coe characterises form as heuristic, as "a motive for generating information," and he explains that . . . "Faced with the emptiness of a form, a human being seeks matter to fill it" (p.18; italics in original). Likewise, we could deduce that Web form, most explicitly it's HTML formating and its multimodality and functionality, is a motive for generating Web information, Web content. And likewise, faced with the emptiness of a blank .html file, a novice Web producer seeks matter, appropriate matter, fitting matter, to fill it. Rather than symbiotically supporting each other, the form of the Web page interface may thus precede its content, and perhaps, with its heuristic potency, thereby lead and shape Web content.

A specific form's heuristic potency could be acculturated through what Coe describes as "a tacit process of ‘indwelling'" (p.19), a means of acquiring some of the ways of functioning discursively in an environment. On the Web, such an "indwelling" would occur through the experience of surfing. For my research, I surveyed 106 personal homepage authors and analysed their homepages. The majority of respondents reported visiting scores, indeed hundreds, of Web sites before beginning to compose their own. Moreover, about 90% of them reported that their exposure to other sites had a role in their compositing of their own sites, in particular by shaping their sense of an effective Web "look," and through that look, of what opportunities the Web offers, or of how Web discourse best works.

For example, one writes,

Mostly, I could see what worked and what didn't work, and I could use ideas and styles that I liked in other sites.

I learned what to do and what not to do. I saw what got peoples [sic] attention, and what didn't.

Keeping in mind Coe's perspectives, let's move towards how novice Web authors are indeed seeking matter to fill the form, seeking furnishings to fill the architecture of their Web homes. The subclass of Web users who actually go through the trouble of constructing a Web home would perhaps be more motivated than others by the rhetoric of public discourse surrounding the Web: the "Information Revolution" and the "Information Superhighway."

Now, Richard Coe characterises information as "data . . . in formation" (1987, p.16). The HTML basis of the Information Superhighway's architecture, betraying its inspirational roots as a medium for the transmission of scientific information, is well endowed with coding that supports structures among pieces of data.

One common manifestation of the prestige of information on the "Information Superhighway" is the frequent use of HTML formatting tags most amenable to marshaling "data in formation": lists and tables. Compared with print, the computer screen seems especially well suited to such discursive formats. According to William Horton, a specialist in technical communication, the two media—print and screen—invite different textual strategies, print being better suited for long flowing passages, for text mostly in paragraphs, whereas the screen is a suitable medium for short (modular) chunks, text mostly in lists and tables (1994, p.46). Similarly, studies of readability by John Morkes and Jacob Nielsen (1997) have found that Web-based texts should be "scannable," containing formatting features, such as bulleted lists, that convey textual information to the scanning, not reading, eye.

Personal homepage authors would be indirectly influenced by the manifestations of these and similar sentiments through their "indwelling" process, their exposure to and use of professionally developed Web sites. And, indeed, among the sites in my study, there are a high number of cases in which the representation of material appears to be motivated in part by the technical capacity of the Web as the medium for what resembles, in its archi-texture, information.

For instance, one participant's site is devoted to Scotch whisky. The longest page on the site features an extensive bulleted list of 109 Scottish distilleries, each identified by name and township (Scottish shire), all organized in alphabetical order by their name. Another page offers tongue-in-cheek instructions for making Scotch whisky, listed in seven steps. But this site does not discuss what is compelling about Scotch whisky, as opposed to other alcohol or other pursuits. The site does not discuss what motivates the author in his fairly sober interest in Scotch whisky. Judging by his name and location, which he provides in a table along with a few other demographic bits of data about himself, he is not Scottish and is not, geographically, all that close to Scotland. What is the history and nature of his incongruous passion? We receive no answer. Aside from a poem by Robbie Burns, the longest passage of prose on the site is a page with three short paragraphs, totaling only seven sentences, offering what appears to be a verbatim copy of a tourist guide's description of what my participant identifies as his favorite Scottish distillery. He doesn't tell us, however, why it is his favorite. The architecture does not seem to have granted him a place for his own voice.

Another participant devotes his site to hyperlinks and a couple of pages about the rock group Rush. The links pages are very short and comprise only lists of a few links each, with few annotations. One of the Rush pages features a table of 11 Rush bootleg tapes with six neat columns. Below that is a squat table of two Rush video bootlegs, only two rows deep and four columns wide. Above the first table are four links to Rush-related sites, set in a list. The page is introduced by a five-sentence paragraph that opens by welcoming surfers, closes by soliciting input from surfers, and, in the middle, reveals the author's discovery of and enthusiasm for the group's music in a sentence that includes, "I've been hooked on the style, music, lyrics and also the talent [of the group Rush]." The author says no more, however, about the group's style or music or lyrics or talent, or of what these features do for him or might do for us. The other Rush page on this site features a numbered list, 1 to 23 down the page, of the 23 tours the group has made between 1973 and 1994; each line identifies the name of the tour, the album it was designed to promote, the duration of the tour, and the number of concerts the group played. At the bottom of the page, we are told the print source from which this list was copied. Aside from one line at the top introducing the list, that bottom acknowledgment comprises the only full sentence on the page.

These two sites are typical of many of the sites in my sample. They include items, not paragraphs, items whose main feature is their place within a surrounding list or table, within an architectural framework. This Web architectural "look," supported by its distinctive HTML construction materials, seems to create an apparent exigence for the corresponding textual furnishings, Web archi-texture, to consummate that look. Ideologically, to have on the Web an arrangement of data, to have "data in formation," is to have something seemingly legitimate to say on the Information Superhighway. These structures serve not necessarily to add insight into our world but perhaps to validate their own textuality as significant looking Web signs. (These structures are thereby the alchemy by which trivia is transformed into information.)

In their preoccupation with that Web look, these homepage producers may be overlooking the kinds of discourses not foregrounded by the novelty of Web architecture, those demanding a textuality of prosody, of development, narrative, elaboration, sustained argumentation, an organic Aristotelian beginning, middle, and end. The prompts for such discourses, what we might recognize to be meaningful representations of and reflections on our world, are weaker because they are without a readily visible Web architecture. The Web, though nominally a medium of free speech, does not in practice invite socially meaningful free speech to the degree that we might have hoped for.

To recap. . . .
I've discussed how Web architecture, a Web "look," seems to play a leadership role in Web textuality, what I call the archi-texture, that emerges on personal homepages. In short, Web form does not just provide supporting background structure but significantly precedes, prompts, channels, and thereby perhaps overlooks, potential Web content.

Let me conclude by expanding my argument to suggest its implications to wider domains beyond personal homepages per se: teaching and professional Web authoring.

I have encountered challenges similar to these when I teach Web authoring. In one such course, students proposed and constructed Web projects and sometimes, despite my best efforts, Web-looking projects, skewed by an extreme vision of the architecture of a data-preoccupied medium. Despite my having carefully defined the assignment to discourage "data-dumps," I still had to confer with students who had proposed, for this professional writing assignment, typing in a telephone directory, typing in retail price lists, typing in a databank of Canadian airline flights, and other similar data processing projects. Some students seem to have become overwhelmed with the informational structures of the medium and, to the extent that their Web sites may be said to have "voice," it is a voice that seems sometimes to echo off the functional and multimodal surfaces of the medium instead of probing and analyzing their topics in depth. Such teaching experience has taught me the importance of modeling, for students, the kinds of challenging site development that achieve meaningful purposes with real audiences and that would also offer scope for the development and articulation of their ideas.

As for professional Web authoring, my argument parallels concerns increasingly raised by Webmasters and Information Architects themselves. For instance, David Ceolin, the CEO of a Canadian computer company, observes in a recent newspaper column, "[M]ost corporate sites are a virtual snoozefest in terms of engaging users. Too much is spent on building the infrastructure of the house and not enough on making the house inviting to visitors."

One such case is the Web site of the Philips Design group, the industrial design component of the Philips consumer goods and electronics company. To enter the Philips site from its flash-enhanced front page, surfers are asked to drag and drop a small pulsating circle into a larger pulsating circular hole. Doing so expands your screen to full size and plays a flash movie which ends with your screen showing a larger pulsating circle surrounded by 5 smaller pulsating circles representing hyperlinks, none of which are labeled, however.

This use of flash recently sparked a spirited thread of over 40 posts on the Computer Human Interaction Web listserv of the Association for Computing Machinery. The thread began with a posting from Donald Norman, professor emeritus and author of The Design of Everyday Things. He complained of having his screen kidnapped by this flash movie, which does not offer a clearly visible way of escaping—he had to control-alt-delete—and he ended with the declaration, "Designers of the world: stop thinking of yourselves -- think of your viewers."

While some on this listserv defended this newfangled design as appropriate to its avant-garde audience, most contributers, who are largely practicing Information Architects, criticized the site design. The essence of their critiques was how technology seemed to take precedence over other concerns like purpose and content. One poster explained:

"[W]hat annoys me most about this website [is] that it exemplifies the frightening new idea that being a designer means . . . allowing your software to do the thinking for you, and having no regard whatsoever for the purpose of what you're designing. Real design is a problem-solving activity."

So let me tie all this together by recalling Donald Norman's plea to the designers of the world, CEO David Ceolin observation about "[t]oo much [being] spent on building the infrastructure of the house and not enough on making the house inviting to visitors," and my own observations about discourse on personal homepages. In sum, these point to an unfortunate early Web leadership of the techno over the textual.

Coe, Richard M. (1987, January). An apology for form: or, Who took the form out of the process? College English, 49 (1), 13-28.

Horton, W. (1994). Designing and writing online documentation: Hypermedia for self-supporting products (2nd ed.). New York: John Wiley & Sons, Inc.

Morkes, J. and Nielsen, J. (1997). Concise, SCANNABLE, and objective: How to write for the Web. [On-line]. Available: http://www.useit.com/papers/webwriting/writing.html. Cited 1998 July 17.

[Others to be added.]