Piecing together and tearing apart:

finding the story in afternoon

by Jill Walker

This paper was presented at the ACM Hypertext 1999 conference where it won the Inaugural Ted Nelson Newcomer Award. It's available in print in the conference proceedings, which are held by most university libraries.

If you prefer, you can read the essay as a pdf file or see more info about it in ACMs digital library.

[cmc.uib.no/jill]

[jill.walker@uib.no]

ABSTRACT: This paper is a reading of a classic of hypertext narrative: Michael Joyce’s afternoon, a story. Several writers have discussed afternoon previously [1, 3, 4, 9]. However I have chosen to explore afternoon from a different angle by using theories of narratology, especially Genette. In this reading, I explore ways in which the text confuses the reader but also the many stabilising elements that aid the reader to piece together a story.

KEYWORDS: Criticism, theory, rhetoric, hypertext structure, hypertext fiction, hyperreading.

introduction: a story?

Stories are one of the most basic and familiar ways in which we understand our world: from the fairytales of childhood to the news stories, soap operas, novels and films we are constantly fed as adults. To appropriate afternoon as a story or as a narrative is not obvious. For instance, Rimmon-Kenan’s definition of narrative fiction can be interpreted to rule out a hypertext fiction like afternoon: “By narrative fiction I mean the narration of a succession of fictional events.” [13] In contrast, afternoon can be read in several different orders, and the succession of events is not always clear.

Several critics have discussed whether a hypertext can have a story at all, or if it has many stories [1,9,11]. Espen Aarseth argues that afternoon and other hypertext “fictions” are something other than narrative, they are ergodic: “non-trivial effort is required to let the reader traverse the text.” In addition to narrative's description and narration, afternoon has ergodics, in the reader's choices. [1] In afternoon, these choices certainly demand more than a trivial effort from the reader, yet a patient reader can find many clues to guide her reading.

putting the pieces together

The first time I read afternoon I clicked my mouse haphazardly on any old word, and quickly grew disoriented. Realising I was lost, I began to carefully choose which words to click, but I usually couldn’t understand the connection between the word I had chosen and the node to which it led me. I never worked out what was going on, who was narrating what and which names belonged to whom. After an hour or so of frustration I gave the whole thing up.

It took several months before I got up the courage to have another go. This time I read the instructions provided instead of jumping straight into the text, and was surprised to discover that a “default” path existed through the nodes. The default can be followed by pressing return at each node instead of clicking on a particular word. If you “answer yes at the beginning and page through on a wave of Returns,” (“In my mind” ) the nodes make up a fairly traditional narrative. From this narrative, a reader can quite easily construct a story. The narrator (who is nameless in this default path, but whom we in other readings find is called Peter) sits through a lunch with his employer, Wert, worrying that his ex-wife Lisa and son Andy may have been involved in a car accident he drove past that morning. He makes a series of phone calls, trying to find out where Lisa and Andy are, but cannot track them down. He also revisits the scene of the accident. He then decides to call Lolly (nothing in this sequence indicates who she is) – but suddenly my eager pressings of the return button only result in one of those irritating digital beeps, and will lead me nowhere. From this last node (“I call”) I can only move on by actively clicking a word.

This default sequence eases the new reader into reading afternoon. Reading this sequence gave me enough background information to start enjoying the leaps between story lines, and to understand connections where I’d earlier only been confused.

The default sequence is for the most part chronologically narrated – so this gentle entrance to afternoon uses time as it is used most often in story-telling. There are many other chronological sequences in afternoon, where node follows upon node with no more feeling of jumps than between paragraphs in a traditional novel. Several of these sequences are “locked”; there is only one possible path from each node. Others can be escaped (or lost, as a reader-detective searching for clues may feel) by clicking certain words instead of pressing return.

So for me to enjoy reading afternoon, I needed to give up my reader’s choice and instead follow the author’s arranged default reading, which was structured in the most conventional way a story can be told: chronologically. But after grasping a minimal version of the story, which the default reading gave me, I could fit new nodes into my constantly changing picture of the story.

The way readers construct a story in hypertext fictions is much like that which has been described by reader response theorists like Wolfgang Iser [6], who wrote about codex literature:

As we read, we oscillate to a greater or lesser degree between the building and the breaking of illusions. In a process of trial and error, we organize and reorganize the various data offered us by the text. These are the given factors, the fixed point on which we base our ‘interpretation,’ trying to fit them together in the way we think the author meant them to be fitted. (page 222)

Iser describes the reading of a traditional narrative, but the work performed by the reader is very much the same in afternoon. The difference is that the process of interpreting the text, trying to construct a story from the nodes available to us, is made more obvious to the reader of a hypertext. The need to physically press buttons in order to read on functions as a Verfremdungseffekt causing the reader to pay more attention to her own act of reading.

stories and chronology

Definitions of stories, such as Rimmon-Kenan’s that I quoted above, usually require that a story has events ordered in a time sequence. But stories are often told in an order that is different from the order in which the events supposedly happened – thus breaking with traditional chronological narration. This is where narratology becomes useful for studying hypertext. After discussing some of these breaches with chronology as described by the narratologist Gerard Genette, I will show how they are used in afternoon.

Genette describes narrative fiction as having three levels: histoire (story), récit (narrative) and narration (narrating) . Story is defined as “the signified or narrative content” (page 27) [5], or as glossed by Rimmon-Kenan, story

designates the narrated events, abstracted from their disposition in the text and reconstructed in their chronological order, together with the participants in these events. (page 3) [13]

The narrative is “the signifier, statement, discourse or narrative text itself,” while narrating is “the producing narrative action” (page 27) [5]. We can therefore separate the order in which events “happen” on the level of story from the order in which these events are told on the level of the narrative.

Most novels and short stories are basically chronologically narrated, so the narrative will tell of events that are supposed to have happened on Monday before events that are supposed to have happened on Tuesday. In chronological narration, story order is the same as narrative order. If the narrative does depart from the story’s chronology, the departures are usually clearly marked by phrases like “before that” or “she remembered last summer, when…” or “little did she know she would soon become queen.” These anachronies, as Genette [5] calls such narrative departures from story chronology, are clearly marked as looking backwards or forwards; they are analepses or prolepses. (page 40) The first two of my examples above are retrospective, or analepses in Genette’s terminology. The last is an anticipation of the future, or a prolepsis.

A reader of a novel where anachronies are clearly marked by phrases like these can easily reconstruct a story from the narrative, as Rimmon-Kenan describes in the quotation above. Reconstruction of the story is more difficult, perhaps impossible, when narrative breaks from story chronology are not marked. Genette calls such unmarked departures from chronological narration achronies. These are narrated events with no temporal reference, events that cannot be placed in temporal relation to other events in the story.

To be unplaceable they need only be attached not to some other event (which would require the narrative to define them as being earlier or later) but to the (atemporal) commentorial discourse that accompanies them. (page 83)

In hypertext, events in one node can be “attached” by links to several other nodes, and several other events. It is easy to assume that achrony must be the norm. However, as we shall see, the events in afternoon are very often anachronical, not achronical, and they are usually clearly marked as such.

Another form of achrony is the narrative ordering of events by a system other than their temporal succession. Genette calls all such “anachronic groupings governed by one or another kinship (spatial, temporal or other)” syllepses:

Geographical syllepsis, for example, is the principle of narrative grouping in voyage narratives that are embellished by anecdotes, such as the Mémoires d’un touriste or Le Rhin. Thematic syllepsis governs in the classical episodic novel with its numerous insertions of “stories,” justified by relations of analogy or contrast. (page 85)

These sorts of atemporal syllepsis are often used in hypertext fictions.

piecing together the default

Most nodes in afternoon contain stabilising elements which more or less situate the when, where and who of the event or situation being narrated. I will show how these elements, which I will call markers, assist the reader to orient herself in the default sequence I described at the beginning of this paper.

The first node presented in afternoon, asks a question: “Do you want to hear about it?” (“begin”) If the reader answers “yes,” she is led to a new node, consisting of only one sentence: “I want to say I may have seen my son die this morning.” (“I want to say”) In a sense, this sentence is a condensed version of the rest of the text. It gives the content: the narrator’s son may have died, and the narrator may have witnessed his death. The phrase “I want to say” suggests that the emphasis is more on the telling, or on the possibility of telling, than on the son’s possible death. The sentence also sets a time: the words “this morning” and the use of the past tense tell us that the narrator is speaking in the afternoon or evening of the day the son may have died.

If we continue reading this default sequence, pressing return at each node, we pass through two nodes with commentorial discourse (a comment about Werther and a memory) and then arrive at “asks.” “asks” is the first node in a long sequence where Peter (the main narrator) and his employer Werther have lunch together. This sequence functions as an arrival and departure point for other nodes. “asks” is one of the many nodes in this sequence which consists only of dialogue and the narrator’s thoughts:

He asks slowly, savoring the question, dragging it out devilishly, meeting my eyes.

<How... would you feel if I slept with your ex–wife?>

It is foolish. She detests young men.

At first sight this node seems to tell us little about the narrated event's context. But an observant reader can glean a wealth of information from it. The main marker in this node is “ex-wife.” A new reader merely stores this information. At this stage in the default sequence it only tells us that the narrator, who is not explicitly named in this node, is a divorced man, probably not “young” since he thinks his ex-wife “detests young men.” If you have read other nodes, however, the word “ex-wife” is a potent marker. It tells us that the narrator is Peter, since he is the only character in afternoon mentioned in connection with an ex-wife. The next node in our default reading (“yesterday”) gives more concrete markers, mentioning “this heat,” “the humidity,” “Wert” and “as we sit detached by the restaurant.” Already the situation is defined, even for a new reader: the male, divorced narrator is sharing a restaurant meal with someone called Wert; it is hot and humid, probably summer. These markers are repeated and refined in a sentence a few nodes later, “I have been employed here three years now, lunched with him over three summers” (“He, he says.”) This information further fixes our image of the situation. But before this node, there is a brief departure from the chronology of the lunch sequence, in a node called “Die”:

I felt certain it was them, I recognized her car from that distance, not more than a hundred yards off along the road to the left where she would turn if she were taking him to the Country Day School. Two men stood near the rear of the grey buick and a woman in a white dress sprawled on the wide lawn before them, two other men crouching near her. Another, smaller body beyond. (..)

Read after “I want to say,” this node functions as an analepsis in relation to the lunch sequence. At this stage of the reading I interpret this node as describing an accident scene which the narrator has driven past, and in which he fears that his ex-wife and son may have been involved. It could logically follow directly from “I want to say,” as a closer explanation of what “I want to say,” but there is no such link. The repetition of the lunch sequence markers in “He, he says,” makes it clear that “Die” was a short detour from that sequence, which at this stage of the reading is a sort of home base for the reader.

The rest of the default sequence is mostly chronologically narrated. The few anachronies to be found here are clearly marked for a careful reader. An example is the ellipsis (an unnarrated passage of time) between the lunch sequence and the nodes where Peter makes phone calls to find out where Lisa and Andy (his ex-wife and son) are, where the first node in the sequence following the ellipsis starts:

And yet he has given me means, if no end. We live, or did live, comfortably, and my child, our son, attends the County Day School, and so when I call I am shunted – as all divorced parents are – to the Assistant Headmaster. (“no end”)

Here, the pronoun “he” refers back to Wert in the previous node. (If it is possible to reach this node from elsewhere, “he” might be read to refer to someone else. That could change the meaning a lot.) This opening elegantly moves the narrative from a lunch to a sequence of phone calls, even giving us a little background information on the way.

In the last node of our default reading, the time is set when the offices that Peter calls are closed: “There are no humans after five o’clock.” This sets an approximate chronology for all the events we encounter in afternoon: The accident happens in the morning. Peter has lunch with Werther at noon, then makes phone calls and visits the site of the accident in the afternoon. He calls Lolly after five o’clock. These, and Peter’s meeting with Lolly after the phone call, are in fact the only events at the level of the “first narrative” in Genette’s sense.

Based on the events I found to be in the first narrative in the paragraph above, I have sorted the nodes in afternoon into five groups. The first group consists of nodes narrating events in the first narrative, as described in the paragraph above. The second consists of analepses in the form of Peter’s memories from four different periods: life with Lisa, his first meeting with Werther when he interviewed for the job at Werther’s firm, his first meeting with his colleague and later lover Nausicaa, and a date with his son’s head-mistress. The third group of nodes consists of commentorial discourse, musings, and short recollected anecdotes narrated by Peter. Three sequences of nodes narrated respectively by Lolly, Nausicaa and Lisa make up the fourth group. These sequences function both as characterisations of the women, and as alternative points of view to the events narrated by Peter. The fifth and last group of nodes forms an outer narrative layer, a meta-level commenting upon the nodes which form the story proper. Among these nodes are quotations from various books, many of which comment upon the form of the text. There are also comments from a narrator calling himself “Michael Joyce.” These groups of nodes aren't necessarily connected by links (which I would situate at Genette's level of narration), but are connected in the structural level Genette calls story.

is the puzzle complete?

After putting together the pieces like this, the puzzle seems completed. We have found our story, reconstructed the events in sequence as Rimmon-Kenan requested. Similarly, J. Yellowlees Douglas [4] has succeeded in finding closure in her readings of this hypertext:

I find my desire for closure sated by [my] fourth reading, which satisfies the tensions that originally gave rise to the story and also resolves or accounts for the greatest number of ambiguities in the narrative.

But how can we resolve the ambiguities of this hypertext, where all the nodes are present, and none have been read once and for all and finished with? Critics of hypertext literature have often demonstrated a strong urge to find features of traditional literature in hypertext, as I myself started out to find story and Douglas found closure. But to force a hypertext like afternoon into ready-made moulds is to overlook the difference of hypertext, the new qualities which hypertext gives literature. This is a danger of using theories of conventional narrative to understand hypertexts, as I have done in this paper. And yet hypertext is not entirely new, of course. As I have shown in the above, afternoon uses many strategies that are well known from conventional narratives.

Although afternoon has many stabilising markers, allowing me to sort into categories and pieces of a puzzle, it contains as many disorienting factors. I will discuss some of these below.

destabilising narrators

An element that strongly works against the reader’s (re)construction of a story in afternoon is the uncertainty in regard to who is narrating. If you read closely, and have a basic idea of the story, for instance after reading the default, you can usually figure out who the narrator is. For a new reader, however, it can be nearly impossible. It’s easiest to keep track of the narrators if you use the default option extensively, which of course down-plays the hypertextuality or at least the reader’s choice.

Switches between narrators in the various nodes are used to confuse and to surprise the reader. Such a switch occurs from Peter in “What’s datacom? I ask.” to Nausicaa in “Negative Values”. Since most nodes are narrated in the first person it is very disorienting to discover that “I” in the current node is someone other than “I” in another node. Yet even here there are markers to aid the reader in solving the puzzle. For instance, in “Negative values” the narrator speaks like a woman (or as a man thinks a woman would speak,) for instance saying “Surely you know men like that.” The previous narrator was a man; we know this because he had an ex-wife. The switch is made even clearer when the previous narrator is referred to by name, or indeed given a name other than “I.” One of the times I read afternoon, I came upon a node (“no in thunder”) where Wert tells the first person narrator (whom I at that stage didn’t know was Peter) that he’s a Lutheran. Having this information, the lines in a later node:

I remember when I told him Wert was Jewish.< He told me he was a Lutheran.> Peter said. (“Negative Values”)

told me that the previous narrator must be called Peter, and that there was a new narrator now, since Peter is not the person referring to himself or herself as “I.” I first thought Nausicaa was the narrator in this sequence, but in the node “Dream pools” (after several other nodes), there seemed to be another narrator, on a higher level:

Do you know what I think of when I think of Dataquest or Star Wars? Sometimes a trick wants to come on you, spread his semen across your breasts, or face, or ass, or have you collect it in your hand... I know that the standard interpretation is that these are gestures of conquest and humiliation, and that may be, but I can tell you from experience (she laughs) they are like boys afterward... I think they merely want to see it with their own eyes, the waste, the result of dollars spent and minutes gone.

Who is this new narrator, who reports Nausicaa’s monologue and observes “she laughs”?

Narrators change gradually in these nodes. From Peter to Nausicaa to someone relaying Nausicaa’s words. A few nodes later it seems that the new, reporting narrator, is Lolly, the psychologist, who has been telling us about a therapy session she has had with Nausicaa. The title of the node, “Lolly’s monologue” is an obvious marker telling us who is speaking, and the first sentence of the node, “Can’t or won’t? Nausicaa will not say.” confirms that “I” in the previous nodes was Nausicaa. Often, as here, the title is a key to the node’s place in the jigsaw of the story, although it is easily overlooked due to its physical separation from the rest of the text: in the title bar in the “frame” of the window.

Readers may come across another example of this confusion of narrators in a sequence which belongs to the fifth category I sorted nodes into above, that is the outer narrative layer. This sequence is a discussion about afternoon which seems to be between “Michael Joyce” and Lolly. Most of the nodes in this sequence are pure dialogues, with no marked narrator. The few narrative comments make it clear that this is a discussion between “she” (Lolly, I deduce from later nodes) and “I”, who speaks and is spoken to as if he is the creator or the writer of afternoon. Lolly fumes at this writer-narrator, criticising the text she is a part of:

<Ha! – she scoffs – (..) This whole electronic circus, this literary pin-ball machine, is nothing less than wish fulfilment and fantasies of domination... It’s not just the foolish obsession with writing as if you were a woman (something any woman would see through—remember Nora Joyce!), not just the episodes upon episodes of erotic confusion and quasi-earthmotherish psychobabble (even this!).... No, the whole thing stinks, its all a fraud: the illusion of choice wherein you control the options, the so-called yielding textures of words... All of it typical, control-oriented male fantasy...!> (“dialectic”)

Pressing the return button (asking for a default) sends us on to a node telling us about Lisa, starting with the sentence “She’d prefer that little be said about her.” (“ex-wife”) If you have read the Michael Joyce-Lolly sequence quoted from above, this node has only one exit, which leads to a continuation of the sequence, starting:

< “She’d prefer that little be said about her,” indeed! It’s utter nonsense! For all your supposed variations, you’ve written nothing but the same old patterns: the wooden wife, the receptive whore, the all-accepting female mind! (…)> (“gift of hearing”)

At this point the sequence converges with a node where Lolly is acting as a therapist. The reader may already have read this node as part of Peter’s dialogue with Lolly, rather than “Michael Joyce’s” conversation with her:

<Fuck this! —I say—I don’t need this...> I stop this short. It is what she wishes me to do.<Look—she says—I’m sorry. That was wrong. You have to understand I am not used to this, to dealing with men...> (…) (“salt washed”)

So this sequence from the outer narrative layer of afternoon, the self-referential layer where a first person narrator called “Michael Joyce” discusses afternoon itself, converges with Peter’s therapy; the meta-narrator converges with the main narrator of the first narrative. The confusion of who is narrating is augmented by this confusion of the real author Michael Joyce, the fictional character “Michael Joyce” who discusses the text the real Joyce has written with one of the fictional characters in it, and the first narrator, Peter. The fact that Peter is a writer himself, and is telling the story, hardly hinders the identification between Peter and the real and the fictional Michael Joyces.

nietzschean repetition?

Repetition is one of the most important rhetorical figures in afternoon. We re-read nodes, re-interpret nodes read in a new context or we try frantically to escape from relentless loops of re-reading. The node “Salt washed,” that I discussed above, reads differently depending on whether you come to it after reading Lolly’s dialogue with Peter or Lolly’s dialogue with “Michael Joyce.” Re-reading nodes in new surroundings is a form of repetition typical of hypertext. Often, re-reading a node invests it with new meaning.

Another shift of meaning caused by having new information occurs when we read Lolly’s interpretation of Peter’s idea of the accident:

We can grant the truth as Peter conceives it. Let us agree, with him, that he was concerned about Andrew and distracted because the school said they could not locate Lisa. Let us stipulate that, in his anxiety, he might have lost concentration—perhaps spilling something on himself—at exactly the spot in the road where he sees Wert’s truck and her in it. (“1/”)

This node leads on (through a few others) to “white afternoon,” which may be a retelling of the scene described by Peter in “Die” (cited above):

The investigator finds him to be at fault. He is shocked to see the body so beautifully there upon the wide green lawn. The boy is nearby.

Pressing return, I default on to “4 what I see,” Peter’s narration of his return to the scene of the accident, after everything has been cleaned up. I have already read this node earlier in this reading. So I re-read a node after having read another interpretation of the same event.

These re-readings are examples of what J. Hillis Miller calls Nietzschean repetition. In contrast to Miller’s concept of Platonic repetition, where each repetition is a copy of an ideal original,

the other, Nietzschean mode of repetition posits a world based on difference. Each thing, this other theory would assume, is unique, intrinsically different from every other thing. Similarity arises against the background of this “disparité du fond.” It is a world not of copies but of what Deleuze calls “simulacra” or “phantasms.”(..) This lack of ground in some paradigm or archetype means that there is something ghostly about the effects of this second kind of repetition. It seems that X repeats Y, but in fact it does not, or at least not in the firmly anchored way of the first sort of repetition. [12]

This describes the experience of re-reading nodes well. The knowledge that you are reading the same node as before gives a feeling of materiality to this repetition that is not exactly repetition, but re-vision. It is being re-read, but not re-narrated. This is a paradoxical materiality, since the text is nothing but electrical impulses. But this re-vision is different from the repetition of a phrase or even a paragraph in a codex book, which is being re-narrated as well as re-read.

Repetition is important in conventional narratives as well as in hypertext. Genette [5] describes frequency of narration, but not of reading. So

a narrative (..) may tell once what happened once, n times what happened n times, n times what happened once, once what happened n times (page 114 )

Let us enlarge Genette’s model of frequency. A reader of hypertext fiction may read once what is told once, n times what is told n times (this is how repetitions in codex books are read), n times what is told once (this describes the re-reading of one and the same node in afternoon) or once what is told n times. The last possibility could occur if a section of text was duplicated, but the reader only ever came across one of the nodes containing the text. I will discuss how something similar to this occurs in afternoon below.

The re-interpretation of the same node when it is re-read seems a perfect example of Nietzschean repetition. Not only does the node seem the same on the surface, it is the same more deeply than a traditional codex repetition can be. And yet it is different, changed, as the unchanged words of “4 what I see” are not the same when read after Lolly’s interpretation of the accident rather than without that background.

After reading for a while, you expect to come across nodes you’ve already read. Joyce’s use of false repetitions breaks this automatisation, subtly changing phrases while keeping them similar enough to a well-known node to disturb the reader. A reader who sees one of these nodes, but never reads its false repetition, will very nearly “read once what is told n times,” as in my model of the frequency of reading above.

Let me give you an example of this: the node “begin”, is the first node you see after the cover page, and many paths lead back to it so that a reader reads this node several times:

I try to recall winter. < As if it were yesterday? > she says, but I do not signify one way or another. By five the sun sets and the afternoon melt freezes again across the blacktop into crystal octopi and palms of ice—rivers and continents beset by fear, and we walk out to the car, the snow moaning beneath our boots and the oaks exploding in series along the fenceline on the horizon, the shrapnel settling like relics, the echoing thundering off far ice. This was the essence of wood, these nodes say. And this darkness is air. (..)

Do you want to hear about it? (“begin”)

Seeing the first words and the layout of the node “false beginning,” a reader may well assume she has read this node before, and skim on to a new node. But more than a speedy skim will show how different “false beginning” actually is:

I try to recall yesterday. < As if it were winter? > I say, but she does not signify one way or another. By five the sun rises and the night freeze melts again across the blacktop into crystal rivers—octopi beset by fear, and we walk out to the car, the snow exploding beneath our boots and the oaks moaning in series, echoing off far ice. This was poetry, she says, without emotion, one way or another. Do you hear it? (“false beginning”)

“False beginning” is almost an inversion of “begin”. “She says” becomes “I say”, “winter” and “yesterday” swap places, it is dawn (five a.m.) not afternoon, or sunset (5 p.m.). “False beginning” comes after “begin”, after a long afternoon and night of anxiety.

Another example of false repetition can be seen in the two nodes “Die” and “die?” These two nodes contain exactly the same text, but have slightly different titles. The question mark in the second of these nodes suggests the importance of the word (or command) die’s double meaning. The arbitrariness of the roll of the die, and the uncertainty that is introduced by the question mark behind the near-duplicate node may be at least as relevant as the fact of death; which may not be a fact.

Repetition is also used more traditionally in afternoon as repetition of a phrase within different nodes. For instance, the sentence “There is no simple way to say this.” occurs in at least four different nodes: “Work in progress,” “I see,” “you have no choice” and “WUNDERWRITER.” Phrases from the dense imagery of “begin” are also repeated in other nodes. These repetitions function similarly to repetitions in traditional novels, and strengthen the feeling of repetition already instated by the re-reading of whole nodes. This is the traditional codex form of hypertextuality: where a word or phrase is read in different “surroundings.”

Repetition is used both to disorient the reader and to help the reader find patterns. Nietzschean repetition is particularly suited to the disruptive role of confusion. But repetition can also help fight this disorientation, by establishing rock-solid resting places, familiar nodes that the reader can use as landmarks for navigation. [2]

the puzzle will always be incomplete

The reader’s search for pattern, for a story, is acknowledged in “calm,” where Lisa ends her monologue with an apology to the reader for finishing telling her side of the story, which she knows the reader has longed for:

I’m sorry (I shouldn’t keep saying I’m sorry I know—even Lolly told me that the one time I saw her—and I shouldn’t say that either, not “even Lolly,” she’s really quite good at what she does, you’ll see, the others depend on her...), but I am sorry that I’ll have to end this now.

I do know what you feel. You make some choices, you begin to see a pattern emerging, you want to give yourself to believing despite the machine. You think you’ve found something. (..) That’s why I’m sorry I have to end it for you so soon.

The only path on from this node is to a blank node titled “&Mac176;”, which leads to the node “closure” and then restarts with “begin” (the first node in each reading).

The disorienting and clarifying forces in afternoon are tightly connected with the content of this story. The confusion we feel as readers trying to piece together the story is very like Peter’s frenzied hunting for his son [4, 8] – although the readers search probably isn’t as self-delusory as Peter’s may be. Less noted, and perhaps therefore perhaps more interesting, are the constant markers stabilising the text, and working against the confusion.

My analysis of afternoon makes use of several tools that are commonly used in the analysis of conventional literature. I have used terms from Genette’s narratology, I have studied the narrators in the text and the ways in which repetition is used. These tools have been useful in seeing how the hypertext is structured. But I hope to see many more readings of afternoon and of other hypertexts. Different approaches will illuminate other aspects of the text.

references

1. Aarseth, Espen 1997. Cybertext. Perspectives on Ergodic Literature. The Johns Hopkins UP, Baltimore and London, p 2.

2. Bernstein, Mark 1998. “Patterns in Hypertext” in Proc. Hypertext 98, ACM (Pittsburgh June 1998) ACM Press, pp 106-112.

3. Bolter, Jay David 1991. “Interactive Fiction” in Writing Space. The Computer, Hypertext, and the History of Writing. Lawrence Erlbaum Ass., Hove and London, p 124

4. Douglas, J. Yellowlees 1994. “How Do I Stop This Thing?”: Closure and Indeterminacy in Interactive Narratives” in Landow, George P. (ed.) 1994. Hyper/Text/Theory. The Johns Hopkins UP, Baltimore & London, 172.

5. Genette, Gerard 1980. Narrative Discourse. An Essay in Method. Cornell University Press, NY.

6. Iser, Wolfgang 1988. “The reading process: a phenomenological approach.” in Lodge, David. Modern Criticism and Theory. A Reader. Longman, London and New York.

7. Joyce, Michael 1987. afternoon, a story. Massachusetts: Eastgate Systems (on diskette.)

8. Keep, Christopher and Tim McLaughlin and robin 1995. The Electronic Labyrinth. http://jefferson.village.virginia.edu/elab/elab.html Accessed October 1997.

9. Landow, George P. 1992. “Reconfiguring Narrative” in Hypertext. The Convergence of Contemporary Critical Theory and Technology. The Johns Hopkins UP, Baltimore and London, 117.

10. Landow, George P. (ed.) 1994. Hyper/Text/Theory.: The Johns Hopkins UP, Baltimore and London.

11. Liestøl, Gunnar 1994. “The Reader’s Narrative in Hypertext” in Landow, George P. (ed.) 1994. Hyper/Text/Theory. The Johns Hopkins UP, Baltimore and London, p 97.

12. Miller, J. Hillis 1982. Fiction and Repetition. Seven English Novels. Blackwell, Oxford, 6.

13. Rimmon-Kenan, Shlomith 1983. Narrative Fiction. Contemporary Poetics. Routledge, London, p 2.