Beginning and Ending a Hyperbook: Possibilities for Authors

Where should the story begin? How will it end? These are two of the primary questions an author must answer when creating any fiction. Hypertext foregrounds these questions of boundaries; in this non-linear environment, the author has the freedom to discard old structural conventions and traditional ideas of closure.

It is perhaps more difficult to abolish the beginning than the end. When the software is launched, there must be a screen the reader sees first. The author has several options:

1. Present a title page with author and publication information. Some might say this privileges the meta-textual information over the contents of the work itself, but in practical terms the perception of the work is changed very little. After all, we must assume that the reader knows what work they are about to read. However, this option begs the question of what element of the actual fiction will appear first.
2. Start with the first page of the first chapter, or whatever node would be its equivalent in a linear text. We could call this the traditional option.
3. Start with a graphic concept map, table of contents, or other metastructural navigation aid. Depending on which display format the author chose, this would reinforce a linear, random, clustered, or other perception of the work.
4. Start with a node selected at random. Of course, this makes some assumptions about the narrative content of the text. Will the work "make sense" if read starting from the middle? Do we want it to?
5. Start with a search dialogue or other mechanism which forces the reader to "seed" the work. One might type "tree," and be sent to the first node containing that word. In essence, this is a pseudo-random beginning.

Important to one's perception of the hyperbook is whether the reading experience is repeatable or not. All of the options except the random beginning are repeatable.

The ending of a hyperbook necessarily problematizes the very notion of narrative closure. At the simplest level, however, a hyperbook can be said to have ended when the reader exits from the program. The following are the most obvious ending options:

1. The reading path terminates at a node which announces "the end" and does not allow further readings. Of course, there may be more than one such node. This is the method used by most interactive fictions.
2. The path never ends, but eventually begins recycling nodes. After a time, the reader will tire of the repetition. This option is put to effective use in such hyperfictions as Victory Garden.
3. The path returns to the first node, completing an obviously cyclic narrative.
4. The path concludes by oscillating between two nodes (as in Hopscotch) or cycling among an obviously finite number of nodes. Borrowing the language of chaos theory, we will call this the strange attractor method.

A reader's perception of concluding a work is strongly influenced by how much information they have at their disposal. A table of contents or other map allows a reader to ensure that they have read all of the text, thus "finishing" the book. At the other end of the spectrum is a hyperfiction like Afternoon, which has branching reading paths and no maps by which a reader can measure their progress.