Introduction

1. I began making "The Mirror of Simple Annihilated Souls" with two concepts in mind: One was the curious story of Marguerite of Porete, and the other was the image of a deteriorating space suit. Using both a moving image and a historical concept as placeholders, I wished to fill in the space this image and concept created with a personal essay that mirrored the technical challenges facing hypermedia writers today. The resulting piece combined text, image, Flash movie sequences, sound, and navigational structure to explore the implications of obsolescence.

2. When this "Mirror" (a Flash version: http://califia.hispeed.com/mirror/one.htm) was completed, I sent it to the Currents editors, John Slatin and Roger Rouland. Since the submission guidelines asked for compliance with W3C Accessibility Standards, I was concerned with the customary "accessibility" questions - would they hear the sound, would they have a Flash plug-in? The Currents editors, however, had something quite different in mind; they were accepting pieces, as per the W3C standards, that were accessible across a range of special reader conditions. John Slatin, for example, was familiar with screen readers to access the WWW. On a screen reader, he could not see the graphics, nor he could he view the Flash sequences; even the layout of the pages produced a nearly incoherent playback. Further correspondence with John and Roger led me to attempt to follow the current W3C* standards for Web Accessibility with a new version of "Mirror" (an accessible version). Since "Mirror" was about writing and technology, I felt that I could, by creating a W3C-compliant version, learn valuable information about the nature of hypermedia writing. The Accessible Version of "Mirror," then, would be designed to reach many readers with disabilities and/or limited access to the Web.
   
   
   

   The "Mirror"

3. The problems in creating a W3C-accessible "Mirror" were both technical and aesthetic - but in a very important way, these considerations were interdependent at almost every point. Priority 1 for the W3C states that certain conditions must be satisfied:

   Provide a text equivalent for every non-text element (e.g., via "alt", "longdesc", or in element content). This includes: images, graphical representations of text (including symbols), image map regions, animations (e.g., animated GIFs), applets and programmatic objects, ascii art, frames, scripts, images used as list bullets, spacers, graphical buttons, sounds (played with or without user interaction), stand-alone audio files, audio tracks of video, and video.

4. Hence, the first job was to survey the image, motion, and sound elements and determine how to replace each of them with a representation in a different medium. The sound was the easiest, so I began with that. But here, immediately, the aesthetic considerations arose. The MIDI sequence for "Mirror" is a 16th century air from the Court of Henry the 8th. While it was a simple enough matter to provide an explanation of the harpsichord-like sound and the articulated, antique notes, it was less clear that the "text" was in any way an "equivalent." Unless the reader could, in fact, imagine a dark-light melody from five hundred years ago, then the effect of the bright mirror, the cameo of Marguerite of Porete, and the use of manuscript markers such as "Recto" and "Verso" became much less resonant. I began to see, in a way that had not been evident to me previously, that multimedia writing depends on a carefully-constructed oscillation between the visual, the aural, and the textual.**  That is, each of the media elements provides a set of sensory suggestions that act in a rhythmic interplay. It wasn't so critical that a reader could actually hear the sound in and of itself, yet the sound acted in concert with other sensory signifiers. The problem I faced in making the sound accessible to the hearing impaired was that the images and text relied upon the sound for completion.

5. Technical problems arose, as well. Screen readers, for example, are aural devices, and the reader needs to be able to hear the text read aloud. The music, one media element that visually impaired readers could access and enjoy, interfered with the text reader - so these readers benefit from an easy way to stop the sound. Currently, each of the browsers handles the loading of MIDI files differently, and the creation of a Javascript that allows manipulation of the sound will not work cross-browser. The solution was to turn the sound down - and even that strategy tended to relegate the sound more to the background, muting the sense of oscillation between aural and visual signifiers. Since, finally, the reader can adjust the sound in a number of ways outside the piece itself (on the bottom Toolbar, or in the Control Panel), a low level of sound was an awkward, but workable, solution.

6. The persistence of the oscillation-balance problem continued as I moved to the simplest of images - the ones used for navigation. For "The Mirror of Simple Annihilated Souls" (Flash Version), I had collected an array of miniature mirror graphics, and these were to be used to indicate the main topics of the essay: Recto, Permanence, Impermanence, Alteration, Obsolescence, and Obliteration. As a visual statement, they were decorative in the sense that they suggested the intricacy of manuscripts of the Late Middle Ages. In this way, the elaborate, small mirrors echoed the notes of the music and suggested the subtleties of variety that so engaged the musicians and artists of that time period. However, while the small differences between the Mirror navigation graphics could be discerned by the eye, these nuances were nearly impossible to distinguish through textual descriptions - so, I eliminated the graphics in favor of word links that provided clear guidance in an aural format.

7. An interesting technical problem arose regarding the main topic links as well. The pages in the accessible version, "The Mirror of Simple Annihilated Souls" (Accessible Version), were arranged with a navigation column down the left side of the page. The sighted reader essentially disregards this information when it is reiterated on subsequent pages. A screen reader, however, interprets the text exactly as it appears in the code - and the repetition of the navigation options on each page is a time-consuming, and often unnecessary, operation. The workaround for this is to put a "skip navigation" option into the code - using a "bookmark" function. While this was not a difficult coding issue, it illustrates the way that accessible material must be thought through in order to provide a smooth reading experience for a wide spectrum of readers.

8. The larger and more content-centered graphics in "The Mirror of Simple Annihilated Souls" (Flash Version) were originally incorporated into Flash movies. Hypermedia pieces which use text, sound, and image often depend, as I have said, on an oscillation of attention between sensory methods of apprehension. The introduction of movement or animation may actually complicate the viewing process considerably, and it certainly makes the sequence of images more significant. The Flash movies in "The Mirror of Simple Annihilated Souls" (Flash Version) were, first, synchronized with the MIDI melody. While this effect could be detected even with the sound playing softly, it was not a feature that could be represented in a textual description for a screen reading machine. Second, the graphics in the movie were heavily layered with moving, transparent gifs; unfortunately, linear text descriptions are surprisingly inadequate for depicting two or more actions occurring at the same time, especially when the images are emergent, suggestive. Finally, these movies contained images that established cognitive associations with text on other pages. For example, the Flash sequence for The Mirror of Permanence depicts an expanding universe - stars explode and then fade back into the mirrored background. This image is echoed in another section, The Mirror of Alteration, where the reader finds a segment on WordStar. The connection between the two "star" signifiers is not made explicitly anywhere in the text; it resides, rather, in that oscillation between visual and textual symbolism, a poetic rather than a logical association.

9. The technical solution to rendering the Flash sequences accessible was to actually create separate pages that showed a still image of the Flash movie, one moment in time, and provide a detailed description of the movement of the graphics. On the main pages, labeled thumbnails now provide links to these explanatory pages, and the sidelining of the main-graphic focus materially changes the effect of the text page, alters the rhythm of the reading, and re-structures the design. Nevertheless, the experiment did lead me, with Roger's generous editorial assistance, to see the kinds of additions that should be made to a text that would need to function as both a part of an oscillating, mediated system and as a linear argument. It is, then, not so much the presence or absence of any one media element that defines the dynamic of hypermedia; rather, it is the oscillation of the reader's attention between image, sound, and text that characterizes the multi-linear, multimedia creation. Conversion of such work to W3C standards is possible, technically.

10. In my own work and experience I have come to believe that the function of each sensory element does not reside in the "object" but in the relationship between sensory elements. As a result, it seems to me that finding an "equivalent" for many of the combinations of media will likely remain problematic in electronic literary works because "the relationship" cannot be duplicated. Finally, in new hypermedia forms - electronic environments where signifiers can be rendered in several different ways - each decision involves a trade-off between optimal technical/aesthetic harmony and the widest possible readership. Poets and fiction writers in this new and developing field might then find it necessary to continue to experiment with issues of equivalency.
    
    
    

    Conclusion

11. And here, of course, I return to the essence and intent of the W3C accessibility standards. The standards legitimately request a "text equivalent" for each of the elements of a Web construction that uses media other than text. They also mandate this: "Use the clearest and simplest language appropriate for a site's content." W3C Accessibility Standards, then, are drafted to cover a wide variety of public and private entities - everything from government agencies to on-line department stores. Poetry, fiction, and literary criticism are relatively insignificant in the scope of all Web material - still, we have a clear challenge to include as many readers as possible in the future of electronic literature. It may be the case, though, for the present, that equivalency cannot be entirely satisfied by current W3C standards when those are applied to electronic fiction and poetry.

    
    

    *The complete text of the W3C Accessibility Guidelines is available at:  <http://www.w3.org/TR/WAI-WEBCONTENT/full-checklist.html>

    **For the concept of oscillation in hypermedia texts, I am indebted to Katherine Hayles and Stephanie Strickland.  Katherine Hayles' landmark work, "Virtual Bodies and Flickering Signifiers," October 66 (1993) 69.91, first suggested that media could be both present and absent at the same time.  Stephanie Strickland, in a talk given at Siggraph 2001, "Moving Through Me As I Move," discussed the oscillation between text, sound, and image.  Her full paper will appear in the forthcoming MIT Press volume, First Person: New Media as Story, Performance and Game, edited by Noah Wardrip-Fruin and Pat Harrigan.