...In our view, part of Halio's distress arises from a wider cultural anxiety about iconic representations. We have tended to forget that graphic comes from a Greek root meaning "to inscribe or write." As W. J. T. Mitchell argues, there is no inherent difference between depiction and description. Yet despite their inseparability and despite centuries of productive combinations, words and images have always been the scene of intense ideological struggle. Since the beginning of the print revolution, publications have incorporated graphic elements not merely as decorative illustrations but as substantive and irreplaceable content. In addition to such visual conventions as titles and subheadings, maps and diagrams have always "gilded" the printed page, especially in technical and scientific subjects. Perhaps because graphics have always been more expensive to publish, however, they have never occupied the central position that words have held.

Before mechanical reproductions of texts, writers could illustrate their texts with relative facility (though dissemination remained a problem). This ease changed with the advent of mechanical writing systems. A pen can be used to draw as well as to write, but a typewriter cannot. But technologies now emerging allow the writer to control both the final appearance of the text and many other forms of graphic expression: line drawings and graphs as well as digitized photographs, video images, animations, and multimedia presentations. So the struggle between iconic and verbal expression has just now spilled over into curricular debates because technologies enabling graphic and typographic representations to coexist in a student writer's work are just now becoming available.

[return to top]This development raises new questions about what counts as knowledge and what sorts of representations certain kinds of knowledge might require. Composition teachers like Halio seem to believe that illustrations have no role in serious intellectual matters, but a research chemist would surely disagree. In one key research task in chemistry--finding what chemists call an analogous transformation--visual representations of molecular structures convey more meaning than the accompanying text. In fact, students in one study trying to complete this task were always unsuccessful unless they found the correct illustration in the materials through which they were searching (Egan). Advanced work on visualization of data also belies the logocentric viewpoint so evident in much discussion of the writing curriculum. In these cases, writing--by which we mean the interpretation and representation of ideas--necessarily involves more than mere words (Landow 49 52).

Not surprisingly, Halio has taken up arms on the side of the verbal, at the expense of the iconic, just at the moment when the evolution of electronic technology may enable valuable combinations of words and images, to say nothing of sounds and simulations. Hypertext and hypermedia systems have come into widespread use, and now software and hardware designers plan to add facilities for interactive video. In technical fields, graphic visualization programs for calculation and design have become important educational resources. [return to top]Some technological forecasters see these tools as the primitive elements of Gelernter's artificial reality, or as some call it, cyberspace (see Benedikt). But in what sense are the objects produced by these technologies still texts? To what extent are we still discussing writing?

Perhaps the communications activities we have described are indeed not "writing" since they are not purely typographic or even alphabetic. But we maintain that the representation of ideas through advanced technologies still requires the production of text; it is still composition . This type of composition situates language--spoken, written, and iconographic--in a much richer context than the typed or wordprocessed essay can provide. A course in multimedia composition would still do the work of rhetoric, the critical study of semiotics in action, but it would do so in a broader technological context.

Composition, as we envision its future, involves more than words plus pictures, or video, or threedimensional modeling programs and the like. As Jay David Bolter argues, writing is and has always been "topographic"--it is speech made visible and then arranged in a mental space. But until graphic user interfaces for computers became available, writers could not fully exploit the spatial and visual dimensions of texts.

[return to top]Computer mediated technologies like hypertext allow creators of texts to construct their discourses in multiple dimensions, exploring alternative pathways for traversal and development. This is the feature of hypertext that makes its spatiality most apparent. Working in a hypertextual writing space, an author employs visual as well as verbal codes to structure and represent knowledge. The arrangement of topics, their order and their relations to one another, can be mapped on a plane, displayed as a hierarchical tree, or represented in some other scheme. At least one researcher argues that this graphic schematization augments textual communication: "integrating additional information about the author's intentions and knowledge structure.... [i]mplies that documents produced with [hypertext] tools keep authors' knowledge structures alive by preserving their argumentation and rhetorical structures which then can be used for subsequent processing" (Streitz 343).

"Subsequent processing," however, implies that the knowledge structure revealed in topographic writing is neither definitive nor static, and this implication has important consequences for the future of composition. Although it is possible to conceive of hypertext conservatively as a simple extension of book technology (see Bolter), the more interesting and relevant applications, from the viewpoint of many professional communities, belong to what one theorist calls "constructive hypertext," a class of texts that allow collaboration and dynamic revision (Joyce 11). These texts blur or collapse altogether the distinction between reading and writing, defining the knowledge worker as always both a producer and a consumer of textual information. Although this collapse of distinctions has long been a concern of literary theory (see Barthes, Iser, Landow), it has recently become a key issue for nonliterary practice. Reviewing the applicability of hypermedia products, a team of industrial researchers contends that "a larger role for hypermedia requires eliminating the distinction between authors and readers. [return to top]We assume that all members of engineering teams will be able to create and access information in a shared, distributed environment" (Malcolm 15). The kinds of electronic texts envisioned by these researchers resist the closure and the formality that have characterized printed texts.Always open to further interventions and to new arrangements and relationships, such texts require us to rethink what we mean when we say "text," suggesting that what we mean is less an object than an articulated social practice.This difference in conception would require corresponding differences in our idea of rhetoric. All this will mean rethinking the project of writing education. (263-265)