This file should not be here.
More specifically, the information in this file, this node, should be present, but it should be in a link rather than a node. To fit my conception of the project, the particular subject of my chosen writing tools should not be in a separate node like any other subject but in the linkspace. It is the connective tissue, not the organ; the gravity, not the planet. It is in the links. The problem is that in choosing to do my project on the Web rather than some other hypertext environment, I chose (all knowingly) a tool which currently makes no allowance for cramming such complex and abstract information into a link. Other systems do (or could). But the standard version of HTML when I began writing (2.0), and the browsers I've been using (Netscape 1.1 when I began; generally 2.0 or some slight variation as I finish) only make allowance for one piece of information in a link: GO!
Choosing to write this project on the Web, then, was in fact a conscious choice rather than a given. Many other hypertext systems--not networked so broadly, granted-- exist in which I could have constructed both my story and my critical web on disk. Such choices are a crucial aspect of writing in hypertext: technology affects not merely the form of publication, but the actual artistic form and even the content.
What follows is not an in-depth comparison of the Web versus standalone hypertext environments, but rather a short list of what I gained and what I made unavailable to myself by choosing to do my project on the Web.
The capability to publish the project in a networked environment was more important than any other single consideration. I could have created the project in Eastgate's Storyspace system--it offers many advantages over the Web, such as ease of linking and visualization of the structure of the hypertext, adding information to the links (rather than just the nodes) such as names, dates, relationships, use of guard fields to shape readers' possible paths through the text. I could then have given copies of the disk to anyone who was interested, just as I could with a printed copy of a conventional text. I could even have ported the Storyspace version into HTML in some semblance of its original structure. But it seemed an unnecessary step. Many of the subtleties possible in Storyspace would be lost on the Web version, and I would have had two things: the true version in Storyspace and a ghost version on the Web. So I chose to work directly on the Web and let that be the one and only version of the project. Currently the story is listed in my Web site, Hyperizons; I have not yet listed the theoretical part of the project anywhere but on my personal Web page (although of course anyone who reaches the story can reach the rest of the project).
Many writers and theorists speak of hypertext as a spatial method of writing and reading. Jay Bolter, author of the influential Writing Space, has dubbed writing in hypertext as "topographic writing," a phrase he derived from the word topography: "Electronic writing is both a visual and a verbal description. It is not the writing of a place, but rather a writing with places, spatially realized topics." (Bolter, 25). Many hypertext systems incorporate this literally into the features they provide both for readers to view the relationship between nodes and for writers to create and alter links. For readers, this often comes in the form of a viewable map or web that shows the various nodes as boxes linked by lines. A similar map can be called up by writers. For example, in the Storyspace hypertext environment, the writer can also change the linking patterns between nodes by moving the boxes on the screen, adding new pointers between them, or deleting or adding boxes. The maps can be quite simple, showing only the relationship of nodes in the immediate neighborhood, or can be dense with diagrammed nodes and their overlapping links if a larger web is viewed.
This cannot be done with HTML and current Web browsers. (I should note here that I recently learned of a new HTML editor, GNNpress, that allows the author to see a "miniweb" view of his linked files. This still does not help the reader, and is not part of HTML specifications. It is quite interesting, though) Thus audiences whose only experience of hypertext is the World Wide Web may surf the Web for a long time, even construct their own home pages, without thinking of hypertext spatially. Following links appear on the surface only to lead one through an infinite series of files or, as they are more commonly called, pages. The phrase "page" itself is an obvious holdover from print, much lamented by enthusiasts of other hypertext systems.
There are, of course, ways around this problem, convoluted though they may be, as I discuss in detailing how I created navigational maps for "Holier Than Thou."
In Hypertext and Hypermedia, Jakob Nielsen provides succinct descriptions of several types of useful links not currently available on the World Wide Web. One is the fat link (also called the multi-tailed link--e.g., see David Stotts). As described by Nielsen (1990, 109)) a fat link could open several windows simultaneously with one click of the mouse. This is not provided in HTML as currently specified. Netscape's proposed frames extensions to HTML, introduced while I was working on this project, do make this possible, though I did not have the time to work it into "Holier Than Thou" (if it were even appropriate). Here's a rudimentary example:
"Once outside the restaurant, Jim, Melanie, and Carlos said good-bye and parted."While I did not work this technique into my project, it does seem to have potential benefits for such a web. In particular, critical discussion in one frame could contain links that bring up the text under discussion in a frame beside it. The critical writing would navigate the fiction not only intellectually but literally.
A second kind of link Nielsen mentions is the smart link. Smart links contain information rather than being just dumb pointers: e.g., creator, date of creation/revision, etc. In a fiction, what if, for example, you wanted to search for links that follow the imagery of silver in Conrad's Nostromo? Trace the movement of Decoud through the plot? A system could theoretically extract a Decoud web from the many webs Conrad constructs in his mythical Costaguana.
The annotation is another useful link currently unavailable on the Web. Annotations allow the reader to click to see a small note without being taken away from the main text--that is, a note that would appear in a pop-up window or frame only for as long as the mouse button was held down. A hypertextual implementation of this with which many readers will be familiar is the pop-up notes in Windows 3.1 help files. For an example of how this lack affects writers and readers on Web, see my comment on "teasing" in the Abstract. To make the comment, which is merely an aside, I have to take the reader to a separate file.
Neither is there a facility for a reader to add notes to her own copy in the margin (a la HyperCard, which can supply a narrow column in one margin or other to add notes or a gloss). Note that here I am talking about a reader adding notes to a copy that would reside on her own computer, not adding notes to the original, which would bring in serious security issues.
A frame is a subsection of a window. Whereas the size of windows is always controllable by the viewer, frames, at least as implemented in Netscape 2.0, give the author a little more control. The author may decide whether a frame is going to be of fixed or variable size. Most Web browsers are still window-based, and it should be noted that Netscape's windows quickly lost popularity after the first few months. They cause problems for readers without frames-capable browsers, and even for those with the right version of Netscape, they cause some navigation problems. The one I've encounted most frequently is when I click in a frame containing a link that takes me to another site using frames. It the HTML tagging is done improperly you can end up with a sort of receding mirror effect in which you get smaller and smaller frames within frames. Still, there are some interesting possibilities for frames, as I discuss in combination with the fat link.
When I first began making "Holier Than Thou" into a hypertext, HTML 2.0 was the standard. Now it's at version 3.2. In addition, some influential companies, particularly Netscape and Microsoft, are introducing "extensions" to HTML that work only with their browsers. Some of these extensions are mere gimmicks, while others have interesting potential for hypertext fiction. I took time to work one of them, Netscape's client pull feature, into "Holier Than Thou." Another, frames, was tempting but ultimately not used, as I discuss above.
This disadvantage is not specific to the Web, but applies to any hypertext system I may have chosen. I don't own any of the software or hardware, or have any of the network connections necessary to working on the World Wide Web. The same would be true of Storyspace or any other system. Without my university connection as a student and employee, I could not afford to be working with hypertext. It would be pointless to be thinking about it, (if I even knew of its existence). I examine this in more detail elsewhere in my discussion of writing spaces.
Abstract | Future Directions | Bibliography | Glossary | "Holier Than Thou" | Project Entrance
Initial release: April 2, 1996
Last update: August 4, 1996