Raine Koskimaa: Reading Victory Garden 

1 - 2 - 3 - 4 - References

Hypertextual Structure in Victory Garden

As said, Victory Garden offers twenty preordered paths through which the reader can traverse simply by pressing the return key after each lexia. These paths differ greatly for their length and coherence. The shortest path is just four lexias long, while "The Grand Tour" comprises some four hundred of the 993 lexias. "NORMAN - the path of glory" is a very coherent, one is tempted to say traditional, narrative, where Emily is the main character. The scenes from Saudi Arabia and Tara are in balance, and letters from Tara to Emily, as well as letters from Emily, are also in an important role. Even though absent from Tara, Emily is clearly the central figure so that events in Tara are all somehow related to her. 

The paths have two ways to end. The first way is to deny the default link - at some point pressing the return key will produce only an audible beep. This is as close as hypertext can come to an ending but this is no definite ending in any way (Jane Yellowlees Douglas especially has dealt with the question of ending and closure in hypertext fiction: see Douglas 1992, 118-152; 1994). First, there may be other links from the lexia, even though there are no default links. Thus, by double clicking yielding words or activating the link list, the reader can still find links leading further from the provisional ending. And even though there were no links at all, it is always possible to backtrack one's way and make a different choice somewhere at an earlier stage and that way totally avoid the dead end. So, after a path of lexias connected with default links, the author may suggest a possible ending by denying any further default links, but this is no more than a suggestion which the reader may or may not accept. 

The other way to provide "a sense of an ending" is to make an infinite loop - from one point the story returns to an earlier phase in the path, from which it then continues exactly the same way until it reaches the point from where it is once again thrown back to the earlier phase. Thus, the path does not exactly end, but there is nothing new happening anymore; using Douglas' distinction we can say that in a case like this there is no ending of a conventional kind, but some kind of closure is nevertheless provided. Naturally, the same precautions hold here as with the no-default option: the reader may, at any point in the infinite loop choose an alternative link leading her out of the loop. 

We can think of a third solution for ending, too, even though I find it much more problematic than the previous two. The reader may always take it as her task to exhaust all the possibilities in the hypertext. In a more or less systematic way she can try to reach each and every single lexia in the hypertext. Even though there may always be readers finding this the best attitude towards hypertexts, there may not be too many hypertext fictions offering even a theoretical possibility to succeed in this attempt. With texts like Afternoon and Patchwork Girl (or, Reagan Library and Hegirascope) it does not take too much time and effort to read the lexias in their totality - it can even be argued that even though these texts quite naturally do invoke the idea of parallel or intersecting alternative possible stories, deep inside they also suggest precisely this exhaustive (or nearly exhaustive) reading as the "real story" - Douglas' description and interpretation of Afternoon illustrates quite convincingly this point. But then, this has to be just one very special sub-class of hypertext, intentionally limited to a relatively small and dense web of storylines, and also, I would like to argue, intentionally subscribing to the aesthetics of traditional narrative form.

But there is a mathematical logic behind the lexia-link structure making it unavoidable that a relatively small number of lexias and a relatively restricted number of links between them, in any case, produce quite a great number of possible paths through the hypertext web. Because of this, the whole of hypertext is, quite soon, out of full authorial control. The order in which story segments are read, the local context for each lexia, determines to some extent how they are interpreted, and in some cases the cumulative effect of unpredicted interpretations may really produce a new story altogether. Of course, at least since Stanley Fish's essay "Interpreting the Variorum", we have been confronted with the idea that any stable, printed text may be interpreted in totally different ways in different interpretative communities (Fish 1976; reprinted in Fish 1980). With hypertexts the situation is different, though: even if we belong to the same interpretive community and thus share the same interpretative conventions, we can still end up with radically different interpretations. The difference is located in the functioning of the text itself, not in the interpretative strategies. 

This discussion of the possibility of exhausting a hypertext, and of totalistic control of hypertext narrative, is highly relevant for Victory Garden. With as many as 993 lexias and 2804 links the task of reaching and reading every lexia is a serious challenge - since a hypertext fiction, in Espen Aarseth's terminology, is not random, but restricted in access, many lexias require the reader to go through a certain path of lexias before it can be reached. (Thinking in other terms, it is the equivalent of some 200 hundred print pages, which is still, in novelistic terms, quite a modest number.) Because of that, reading all the lexias in Victory Garden means that one has to read, or at least skip through, many of the lexias several times. Douglas estimates that reading a five-hundred-lexia hypertext requires as much as seventy hours, and compares that to the six to twelve hours required for an average reader to consume a three-hundred-page novel. (1992, 50). Trying to comprehend all the possible permutations of orders in which the lexias may be read, is well beyond the capacity of any individual reader. Thus, even though still a limit case, the Victory Garden is clearly pointing towards the kind of hypertext fiction which, because of its size, is theoretically and practically, inexhaustible. Because of this, I will rule out the exhaustive reading as a definitive ending for Victory Garden (there are other reasons too, to be dealt with later). 

Singular Loops 

One of the hypertextual structures in Victory Garden is what I call a singular loop (as opposed to an indefinite loop). In singular loop the reader is taken back to a previous point in the path she is reading, but the next time around not. There is a loop, a sequence of lexias read twice, but after that the path continues forward: 


This particular device is used in a few places in Victory Garden, but usually there is no clear motivation for this. One explanation could be - and this is related to one interpretation I'll give for the whole text - invoking a certain sense of malfunctioning, of an unintentional lapse in the running of the narration. With loops, both infinite and singular, the question of repetition is foregrounded. Jill Walker has written about the "Nietzschean repetition" (following J. Hillis Miller), repetition with difference in Afternoon (Walker 1999). With loops the difference follows from the backward movement in reading - it is like a concretization of Peter Brooks' idea of "anticipating retrospection" in reading: we read expecting to get a thorough understanding of a situation, trusting that that understanding will help us better comprehend the things we are currently reading (Brooks 1984). With loops in hypertext, in the occasion where some (local / temporary) conclusion is reached, the reader is taken back to read the previous moments anew in the light of that conclusion. 

There is one particularly interesting occasion mixing loops and repetition with variation. The moment when Urquhart - after the runaway from police - arrives at the Observatory and meets provost Tate, in one path, will be repeated not once but twice, and in each case with concrete difference, not just difference produced by the repetition. The titles of the lexias in that sequence are as follows: 

"Ring" -> "Help Mister Wizard" -> "Fool's Errand" -> "Ring Cycle" -> "Errant Fool" -> "In Need of Help" -> "Ring Around" -> "Arrant Fool" -> "Helpful"

The beginning paragraphs of each starting lexia seem to be commenting on this exact cycle structure:

"Ring": U ran through the dark field, slipping and scrambling on the dry ground. He knew Madden was behind him somewhere. He did not look. […]

"Ring Cycle": Once more U ran through the dark field, slipping and scrambling on the dry ground. He knew Madden was behind him somewhere but he did not look. […]

"Ring Around": U is once again still always running through that dark field, slipping scrambling through his own footprints on the dry ground. He knows Madden is behind him somewhere but he doesn't dare look. You've had this dream before, you know.

This sequence comes (in some of the preordered paths, at least) right after a long sequence where Urquhart is fleeing first from agent Madden, then from the highway patrolmen, and then from Madden again - so even if you happened to read only the third variation of the "Ring" (starting with "Ring Around") you could easily interpret the beginning in the light of these previous events, totally ignorant of the other variations. Other variations during this cycle are less significant:

"Help Mister Wizard": […] U picked up the book. It was quite heavy. Ponderous.

"In Need of Help": […] U picked up the book. It was very heavy. Voluminous.

"Helpful": […] U picked up the book. It was very heavy. Massive.

The title "Ring Cycle" invokes in quite an ironical way (this is no real loop, or, ring, at all since the lexias are not the same but just closely resemble each other) an allusion to Wagner's opera cycle, and also to the idea of Gesamtkunstwerk. This again can be related to a more general interpretation, where Gesamtkunstwerk could seen as an early version of virtual reality (and multimedia). What is important here is that this time, when the repetition is only approximate (the three instances differing in details), it has a totally different effect than with real loops: this time it is more like different drafts, or adjustments, trying to find the exact atmosphere. 

This again nicely demonstrates the difference between hypertextual semantics and narrative semantics: narratively speaking we certainly are dealing with repetition; the hypertextual structure here is plainly a linear path (as shown in the title list above). 

1 - 2 > 3 - 4