The footnote, once considered the treasure of a special artistic talent, has sunk in reputation to a bauble at the bottom of a scholar's page. And yet a footnote--an expository addendum to the text, an annotation, a reference--exhibits a fascinatingly complex set of possible relationships with the text to which it refers. When a text like James Joyce's Ulysses enters the picture, and digital hypermedia is the medium, the entire question of footnotes and annotations needs to be rethought.

Undergraduate English students probably first encounter annotations and footnotes in The Norton Anthology of English Literature or another college textbook. They squint to decipher the tiny bits of information at the bottom of the page and then return to the text at the top either appreciative or frustrated. They seem to form a judgment quickly that footnoted annotations are either useful places to find what they need to know or repositories of strangely arcane and irrelevant displays of knowledge.

Graduate students learn that creating footnotes as references to, and also as extensions of, the main text's argument is an essential part of scholarly work, and the first stage in my professional relationship with footnotes consisted of becoming proficient at writing them. The second stage involved reading other people's footnotes with interest and also scepticism, occasionally reading the footnotes before and maybe even in place of the main text. In the third stage, I stopped reading footnotes entirely, simply skipping them whenever a text included them. In the fourth (current) stage, I try to write whenever I can without using them.

Because this is an essay about annotation, however, I will include a few notes. Doing this in a print article is a routine activity, but it is a different matter in a digital text. On a screen, for one thing, I can't provide footnotes at all. The bottom of the screen doesn't mean the same thing spatially as the foot of a page (text on top, note at bottom). I could provide notes at the end of long electronic "page," but that would be just a screen imitation of a printed text. Since notes can appear as new screens, no longer at the bottom or end and no longer necessarily subordinate to a "primary" text, they do not have to be short in order to meet printing requirements or to avoid overwhelming the main text.

* * *

In The Footnote: A Curious History, Anthony Grafton documents how, for a long time, the writing of footnotes involved particular, uncommon skills and was even considered a special artistic talent. For the most part, this isn't the case now, as several examples illustrate. Gérard Genette, appropriately in a footnote, quotes a clever disparaging remark about footnotes from the French writer Alain: "A note is the mediocre attached to the beautiful."
2 In his poem "The Scholars," William Butler Yeats contrasts "young men, tossing on their beds" and writing poems inspired by passionate emotions with the "old, learned, respectable bald heads" of their editors and annotators. And, sustaining the unlikely combination of footnotes and beds, Grafton relates a wonderfully witty quip from Noel Coward--to the effect that having to read a footnote resembles having to go downstairs to answer the door while making love. (In a footnote--where else?--Grafton notes that Coward attributed a stronger version of the remark to John Barrymore.3)

Scholars have talked about footnotes in various ways. Patricia White uses Gaston Bachelard's phenomenological account of space, in which the attic represents pattern and framework and the cellar stands for irrationality, to account for our tendency to value the text on the top of a page to the detriment of the footnote at the bottom (84-85). John Lavagnino neatly captures the sense of frustration that footnotes can elicit when he talks about what he calls commentary's "social ineptitude": it is never there when you want it, invariably there when you do not (82). Demonstrating how hot and controversial a topic annotation currently is, in a recent issue of the Chronicle of Higher Education Rodger Beehler gives several examples of intrusive footnotes as part of his claim that modern annotation "gives new meaning to the idea of wrestling with a text" (B14).

In the eighteenth and nineteenth centuries, writers sometimes included their own footnotes and marginal commentaries in their novels and poems for serious, comical, or satirical purposes, as in Coleridge's Rime of the Ancient Mariner, Fielding's Tom Jones, Sterne's Tristram Shandy, and Pope's Dunciad. In the late twentieth century, they appear in novels usually to be ridiculed or caricatured, often to stunning or hilarious effect.
4 In Vladimir Nabokov's Pale Fire (1961), for example, a novel in the form of an annotated edition of a poem, the annotator desperately tries to commandeer the poem to give it a meaning the poet refused to provide. In one chapter of Jonathan Coe's novel The House of Sleep (1997), a sleepy film-journal editor inadvertently omits one footnote number from a filmmaker's annotated memoir, causing all the subsequent notes to refer, sometimes scandalously, to the wrong cue in the text (note 4 is supposed to annotate what is erroneously numbered 3 in the text, etc.).5 And, an extreme case, a mock mathematical-theory article that occupies twelve pages of Gilbert Sorrentino's novel Mulligan Stew (1979) contains 114 footnotes, all of which are essentially non sequitors.

In a text destined for print publication such as this one, I can refer to footnotes without reproducing most of them because we are all so familiar with footnotes and annotations that we know what is meant. But mistakes such as those in The House of Sleep don't make sense in a digital environment, where there is no need to number notes. (A parallel error in a digital text might involve misconnected links.) I recently sold my two-story house and bought a one-floor apartment, so I have to relate even to the part of the Coward quip about going downstairs to answer the door from memory and no longer from direct experience. Talking about annotating Ulysses in hypermedia is about changes of these kinds.

* * *

Annotations can seem to exist at a level of fact, but Martin Battestin lists three variables that always affect annotation: the assumed audience, the nature of the text being annotated, and the nature of the annotator (4). Traugott Lawler emphasises the second variable, the text being annotated and its presumed attitude towards annotation, when he suggests that "perhaps the central question to ask, before we start annotating a text, is whether the text itself embodies an attitude to annotation" (97). Several critics have attempted to describe this complex relationship. For example, Peter Cosgrove remarks that a footnote leads a "double existence," both outside the text giving information and inside it hindering its progression (148). John Lavagnino considers ways in which this double existence can be seen in terms of conflict: the text vs. the commentary or the commentary (as supposedly objective "fact") vs. the more valued act of criticism. Less neutrally, Ralph Hanna characterises annotation in terms of power and aggression both towards the author and towards the presumed audience (181-82).

Jacques Derrida captures these paradoxes neatly when he describes the "double bind" of annotation: the text says to read it in silence but also at the same time cries out for response from the reader:

[W]e see how [the] law text, which makes the law, produces at the same time a double bind: it says to the reader or auditor, 'Be quiet, all has been said, you have nothing to say, obey in silence,' while at the same time it implores, it cries out, it says, 'Read me and respond: if you want to read me and hear me, you must understand me, know me, interpret me, translate me, and hence, in responding to me and speaking to me, you must begin to speak in my place, to enter into a rivalry with me.' (202)

This response, of course, often appears in the form of commentary and annotation in footnotes.

* * *

Ulysses is an excellent test case of Lawler's suggestion that an annotator should always ask about the text's attitude to annotation. Texts are exhibited throughout Ulysses, and its characters respond to, comment on, even actually annotate texts. An entire chapter features Stephen Dedalus's theory of Hamlet, in which Stephen interprets detail after detail in Hamlet and Shakespeare's other works in the light of biographical information. Less loftily and less aggressively, Leopold Bloom annotates an ad for Alexander Keyes, Tea, Wine, and Spirit Merchant as he talks to the editor of the Evening Telegraph about placing the ad in the newspaper. And the newspaper editor recounts how a Dublin newsman cabled classified information about a local murder to a New York newspaper by providing an elaborate annotation of an ad which his New York colleague had in front of him.

In many different ways, then, Ulysses is a work in which texts exist to be commented on and to be annotated. Because so much information in the book is obscure--Dublin details from 1904, specific information from Irish history, allusions to popular culture from the late nineteenth and early twentieth centuries, references to Homer's Odyssey and phrases in Latin and Greek (which maybe readers of Ulysses could have been expected to know when the book was published in 1922 but which are increasingly beyond the experience of its readers now)--readers have often looked for help in annotations, and such information has been provided for them. Four collections of annotations exist. Two are book-length: Weldon Thornton's Allusions in "Ulysses" (1968) and Don Gifford and Robert Seidman's "Ulysses" Annotated (1988). The other two are editions of Ulysses, the Oxford World's Classics paperback with notes by Jeri Johnson and the Penguin Twentieth-Century Classics Annotated Student's Edition with notes by Declan Kiberd, both of which offer endnotes to accompany the text.

All these notes have proven both useful and frustrating. In different ways, they provide much valuable information and leave a great deal out that a reader might want, they make mistakes along with providing reliable information, they mix interpretation with more factual details. To use them, a reader must either keep a separate book next to Ulysses or flip to the back of a large paperback. The annotations in the Ulysses editions are, understandably, aimed at students and provide the kinds of information that beginning readers probably need and want. The Thornton and Gifford books provide information that more advanced readers as well as beginners might look for.

These annotators necessarily take a pragmatic approach: they have a job to do, and they set out to do it. They do not worry much about the relationship of the notes to the text-although Gifford offers the bizarre suggestion that a reader might want to look at his notes first and then turn to Ulysses (xvi). But, as soon as the medium changes, and especially when, as in a hypermedia presentation, the notes can begin to occupy the same visual space as the text, such questions have to be addressed.

* * *

I began thinking about issues of annotation as Director of "James Joyce's Ulysses in Hypermedia," a presentation of Joyce's book in an electronic, hypermedia format. A major component of the project, which is now called "Digital Ulysses" and which I am doing in collaboration with the Poetry/Rare Books Collection, University at Buffalo, State University of New York, is a multilevel, multimedia annotated text of Ulysses. A great deal of material will be part of the presentation, including:

* the full text of Ulysses, in more than one version

* specific words and phrases linked to definitions and annotations and to extended analyses and commentary

* photographs of places and people

* maps

* audio versions of songs and other music mentioned or sung in the book

* audio readings from quoted or cited works

* an oral pronunciation guide

More extended sections of Ulysses will be linked other kinds of materials, all coordinated to the passage of Ulysses that is on the screen:

* an archive of published scholarship

* biographical and historical background material

* newly written hypertext criticism

* oral readings of Ulysses

* literary works that are quoted or referred to or echoed

* filmed excerpts of Ulysses

Readers will be able to bookmark their place and take notes for future use. The presentation is designed to allow the simplest possible ways of navigating through the vast amounts of available material. As with any hypertext, readers will be offered multiple pathways through the materials, and they can choose which information and how much detail they want to see.

* * *

Presenting any print-based text in digital format inevitably changes it, and those changes affect annotation in important ways. In the decentred writing that is hypertext, there is nevertheless a kind of centre to a structure built around a pre-existing text like Ulysses. Reading this kind of hypertext, you move out from and back to the central text. You might stay away from that text for quite a while. You might even read all of the Odyssey or follow several newspapers for June 16, 1904 or watch a slide show of photographs from the Dublin of that day. But most likely you will eventually go back to Ulysses, to wherever you were in it.

Ulysses on the screen, and Ulysses as part of an electronic hypertextual network, can never be the same as Ulysses in the pages of a book. The digital text loses all sense of the book's physical pages and also of the bulk that makes the printed versions so distinctive and also so imposing. More important, the text cannot exist in isolation, separated by its covers from other books on the shelf. The book on the shelf is like a house in a neighbourhood. The digital text is more like an apartment in a high-rise complex.

If it is part of a hypertext system, the words of Ulysses (or of any other work) will be linked to all kinds of other material, including to other parts of itself. George Landow has characterised a print-based work like Ulysses as "incomplete" in a digital environment (79), which for Landow also means an intertextual network. On one level, of course, Ulysses is as complete as any literary work; it is even what Richard Ellmann has called "one of the most concluded books ever written" (xiv). But it is problematical whether a book called Ulysses, which uses Homer's Odyssey as a structural grid, can ever be considered "complete" in itself. Readers hardly ever approach it that way: they almost always accompany their reading with secondary books such as Gifford and Seidman's "Ulysses" Annotated or the other available annotations or any of the hundreds of books of criticism and analysis. They bring their previous reading and cultural experiences with them; they store information about Ulysses in their heads or in notes; they annotate the page margins of their books. If Ulysses can be called "complete" in print form, that largely means that its pages can exist within covers that contain no other works. A digital hypermedia Ulysses has to dispense with the covers and the boundaries they provide and with the physicality of the pages, but as compensation it can provide a great deal of flexibility and openness in the way it can present secondary information.

* * *

A passage from Ulysses can help to illustrate the issues about both annotations in general and Ulysses in hypermedia that I have been talking about. The three variables for annotation that Martin Battestin mentions--the nature of the text, the assumed audience, and the annotator--all figure into the example. So does the question of an annotation's social ineptitude or grace that John Lavagnino raises: is the annotation needed or unnecessary? does it say too much or too little? does it try to answer the right questions or the wrong ones? does it provide the information that its presumed audience wants and needs? is it appropriate to the text it is annotating, that is, in this case, to Joyce's Ulysses?

The passage I want to focus on occurs at the end of the fifth episode, "Lotus Eaters." Leopold Bloom is walking along a street in Dublin after spending a few minutes in a church, and he is approached by Bantam Lyons, a man he knows slightly and likes even less. Bloom mainly wants to get rid of Lyons as quickly as he can, and after a brief conversation Lyons walks away. Bloom is pleased with himself for extricating himself from Lyons so easily and effortlessly.

He strolled out of the shop, the newspaper baton under his armpit, the coolwrapped soap in his left hand.
    At his armpit Bantam Lyons' voice and hand said:
    --Hello, Bloom. What's the best news? Is that today's? Show us a minute.
    Shaved off his moustache again, by Jove! Long cold upper lip. To look younger. He does look balmy. Younger than I am.
    Bantam Lyons's yellow blacknailed fingers unrolled the baton. Wants a wash too. Take off the rough dirt. Good morning, have you used Pears' soap? Dandruff on his shoulders. Scalp wants oiling.
    --I want to see about that French horse that's running today, Bantam Lyons said. Where the bugger is it?
    He rustled the pleated pages, jerking his chin on his high collar. Barber's itch. Tight collar he'll lose his hair. Better leave him the paper and get shut of him.
    --You can keep it, Mr Bloom said.
    --Ascot. Gold cup. Wait, Bantam Lyons muttered. Half a mo. Maximum the second.
    --I was just going to throw it away, Mr Bloom said.
    Bantam Lyons raised his eyes suddenly and leered weakly.
    --What's that? his sharp voice said.
    --I say you can keep it, Mr Bloom answered. I was going to throw it away that moment. Bantam Lyons doubted an instant, leering: then thrust the outspread sheets back on Mr Bloom's arms.
    --I'll risk it, he said. Here, thanks.
    He sped off towards Conway's corner. God speed scut.
    Mr Bloom folded the sheets again to a neat square and lodged the soap in it, smiling. Silly lips of that chap. Betting. Regular hotbed of it lately. Messenger boys stealing to put on sixpence. Raffle for large tender turkey. Your Christmas dinner for threepence. Jack Fleming embezzling to gamble then smuggled off to America. Keeps a hotel now. They never come back. Fleshpots of Egypt. (70; 5:517-44)

First-time readers are unlikely to pay much attention to Bloom's repeated words, "I was just going to throw it away" and "I was going to throw it away that moment." Bloom himself does not consider them to be anything special. They are just part of his attempt to shoo Lyons away. But Lyons seems to treat them as very meaningful, although the text doesn't say why or even which of the words he responds to.

Three of the four sets of notes that I've mentioned annotate these lines. Don Gifford and Robert Seidman, in a first, long note to the race itself, mention the event that Lyons thinks Bloom is alluding to:

Ascot. Gold cup: The Gold Cup, one of the two main annual events of the British racing calendar, was to be run that day at Ascot Heath, twenty-six miles from London, at 3:00 p.m. "The Gold Cup, value 1,000 sovereigns with 3,000 sovereigns in specie in addition, out of which the second shall receive 700 sovereigns added to a sweepstakes of 20 sovereigns each . . . for entire colts and fillies. Two miles and a half. The field: M.J. de Bremond's Maximum II; age 5. Mr. W. Bass's Sceptre; age 5; A. Taylor. Lord Ellesmere's Kronstad; age 4; J. Dawson. Lord Howard de Walden's Zinfandel; age 4; Beatty. Sir J. Miller's Rock Sand; age 4; Blackwell. Mr. W. Hall Walker's Jean's Folly; age 3; Robinson. Mr. F. Alexander's Throwaway; age 5; Braime. M.E. de Blashovits's Beregvolgy; age 4. Count H. de Pourtale's Ex Voto; age 4. Count H. de Pourtale's Hebron II; age 4. M.J. De Soukozanotte's Torquato Tasso; age 4. Mr. Richard Croker's Clonmell; age 3." "Selections for Ascot Meeting. Gold Cup-Zinfandel." "Tips from 'Celt': Gold Cup-Sceptre." (as reported in the Freeman's Journal, 16 June 1904, p. 7.) The race was won by the dark horse Throwaway, a twenty-to-one shot; see [reference to a note to a later passage] (98; the ellipses and italics are Gifford's)

In a follow-up note to Bloom's words, Gifford and Seidman emphasise a second time why Lyons thinks the words are significant:

throw it away: See preceding note. The point is that Bloom has just unwittingly given a tip on the Gold Cup race. (99)

Declan Kiberd in his note combines some factual information with an interpretation of Joyce's passage:

throw it away: later, when a horse called Throwaway wins the Ascot Gold Cup, Lyons will circulate a rumour that Bloom has won money on the bet: another example of the treachery of misunderstood language. The newspaper-phallus is now 'thrown away' by Bloom before he opts for his narcissistic bath. (979)

In contrast, Jeri Johnson tries to limit her brief note to the race to factual information:

Ascot Gold Cup: an important event for Dubliners in Ulysses; the actual Gold Cup, an annual event, was run in 1904 at Ascot (in England) at 3:00 p.m., 16 June. (801)

Some clear and certain statements--factual and with an unproblematical relationship to the passages--can be said about Bloom's two remarks, "I was just going to throw it away" and "I was going to throw it away that moment." There really was a horse race, the Gold Cup, at Ascot in England on June 16, 1904. Lyons mentions this when he asks to see Bloom's newspaper, and Bloom recognises this, even though he seems to have no idea which horses are entered in the race. There really was a horse named Throwaway in the race, a horse that ran as a 20-to-1 outsider. Throwaway went on to win the race. All this becomes clear later in Ulysses. A group of Dubliners--all of whom bet on Sceptre (even Lyons, who temporarily followed up on what he thought was Bloom's tip but who was later dissuaded from the bet)--talk in a pub, and one of the men reports Bloom's conversation with Lyons. As a result, the men believe that Bloom, in what they see as typical of Jews, had an inside tip on Throwaway, won a pile of money on the race, and stingily failed to share his winnings with the other men by buying them a round of drinks. Bloom is almost injured in an attack as a result (274-75; 12:1548-58). He eventually reads a newspaper account of the race, but it is not clear even then if he connects the results of the race with his remark earlier in the day to Lyons (529; 16:1274-85).

An annotation can point out all kinds of details, including, as in the Throwaway passage, information that the book will reveal later on. But how much should it say? Should beginning readers, in a note to a passage at this early point in Ulysses, be told information that they will learn only later on in the book?

There is no clear answer to this question. People will respond according to their sense of the experience of reading a book like Ulysses, or what they think it should be--and this will affect their answers to questions of whether readers should or shouldn't be told things ahead of when they will encounter them in the book, or whether the task of guiding readers through the book sometimes involves violating some of Joyce's patterns of revealing information. My students tend to be divided, or, perhaps more accurately, conflicted: in the abstract, they say they do not think they should be told anything ahead of its appearance in the book, but, in practical terms, they welcome any information that gives them a sense of how this bewildering and mystifying book (which is how Ulysses appears to most of them as they read chapter 5) actually works.

Thus, even if annotators agree on an assumed reader who is experiencing Ulysses for the first time, they will not necessarily agree about how much information should be included in a note. The audience might be constant, but the annotators' sense of how Ulysses works will crucially affect how they construct the note. And what if the presumed audience changes? For second-time or more experienced readers, there is no need to maintain the secret of Throwaway. For these readers, the questions annotators have to ask themselves involve such issues as how much to say, in what order, and whether anything should be left out.

* * *

Thus, we are talking about a rhetoric and even an ethics of annotation: how to present information effectively, what order to present it in, how much to say and to omit, how to distinguish indisputable fact from interpretation. This is a matter both of accuracy and also of tact. A change in the medium can affect the way annotators might think about both. The general situation regarding annotation, and some questions about the activity (such as how much information to include and how to order the information), remain relatively constant no matter what the medium. They are equally relevant and problematical for print annotations and for digital ones. But, for several reasons, digital presentation opens up new possibilities and therefore new angles to the questions.

For one thing, there is much less restriction regarding the space that annotations can take up in digital presentations. Print annotations typically need to be kept short, or to take up less space than the primary text, or to fill up fewer than, say, five hundred pages, but these kinds of limitations disappear in digital presentations. Also, there is no need to posit only one kind of audience, whether first-time readers, or students plus some more experienced readers, or scholars. Because information can be doled out as users request it, a digital presentation can plan for several different levels of readers. In the Throwaway example, it is possible to posit a readership that includes both people who should not be told anything about Bloom's throwaway phrase, "I was just going to throw it away," and also people who will not be bothered or compromised by being told everything that can be said about the phrase.

For digital presentation, the question then becomes how to present the information, how to be tactful. I helped run and participated in a month-long discussion regarding annotation on the University of Utah's J-Joyce email listserv in November and December 1998, and we considered questions like this.6 One possibility is to construct the information in layers, so that a series of screens starts with basic factual and identifying information and then expands into more elaborate information and ultimately into various interpretations. Beginning readers can start with the basic information and, only if they want it, move on to more elaborate information. Advanced readers can choose to start with the more advanced information and skip the basic identifications that they probably already know.

A related question involves how much information beginners should be told. It is counterproductive (and one of the reasons that students often do not read printed annotations) to give beginners more information than they want. If they are looking for a simple definition of a word or identification of a detail, a long paragraph full of scholarly information is overkill--it might turn them off the annotations, and maybe even off the book itself. It seems most useful to make the first-level annotation as simple and short as is realistically possible (five words? ten? twenty-five? whatever will suffice in each case) and then give readers easy access to more information. If a reader wants only a quick fix on a word or phrase in order to keep reading, the least intrusive annotation is probably the most successful one.

* * *

From the Throwaway example, or from almost any other passage from Ulysses (or, for that matter, from any other work) that might be given--and with the differences between print and screen presentation in mind--several questions about annotation can be formulated. I will close this essay with seven questions and some thoughts in response to them:

* What should be annotated?

* What, if anything, should be not annotated?

* Should information be presented differently for first-time readers than for later ones? If so, how can this be done?

* How should the information be presented, given the many possibilities opened up by computer links?

* Can there be too much information? Too little? Is there a desirable mean?

* Is there a line between information and interpretation? If so, how do we proceed in order not to cross it? If not, how do we construct annotations?

* Do these questions change for different categories of information (historical, other languages, intertextuality)?

None of these questions have clear answers, and, in some cases, the possible answers for a digital presentation of the Throwaway example will be different from those for a print one.

* What should be annotated?

Clearly, something needs to be said about the conversation between Bloom and Lyons, which at first glance seems either inconsequential or bewildering. But, since the digital format permits a layered series of screens, it is possible, even desirable, to resist mentioning in the first instance the reason why Bloom's remark about the newspaper--"I was just going to throw it away"--becomes so significant. An initial note to those words might be a simple "watch-this-space" symbol, or the note might be limited to the mention in the text of the horse race and say something similar to Johnson's note, which simply indicates that there was a horse race in England on that day. A link from that initial note can take readers to more information.

* What, if anything, should be not annotated?

There probably is nothing that is known that should be kept out of the annotations, but most of the details should be reserved for second- and higher-level notes.

* Should information be presented differently for first-time readers than for later ones? If so, how can this be done?

In a hypermedia presentation the note aimed at beginning readers should present only the barest factual information. More elaborate options, such as highlighting the relationship of Bloom's remark to the horses in the race (as Gifford and Seidman's first note does) or indicating what Bloom has inadvertently conveyed with his remark (as does Gifford and Seidman's second note), or presenting the annotator's interpretive spin on the passage (as Kiberd does in his note), can be reserved for the notes aimed at more advanced readers.

* How should the information be presented, given the many possibilities opened up by computer links?

There are many options, each with advantages and disadvantages. The one option that should not be followed is the print one, in which whatever information is presented comes out as one short or long note. Since it takes little time or effort to get from one screen to another, the information can be parcelled out in increasingly complex layers, and different kinds of materials, such as words and images, can be presented on different screens. There are various options for presentation, such as putting the annotations in pop-up windows or in frames, indicating that certain words are triggers for links or hiding that fact, and requiring the reader to click the mouse or merely to hold the cursor on a word for a few seconds to call up an annotation.

* Can there be too much information? Too little? Is there a desirable mean?

Obviously, an annotator can present too little information. In print, space considerations sometimes force the annotator into this position. Too much? This is probably a matter of context. Beginning readers who want a quick definition of a word or identification of a person or place or historical event in order to understand a passage will probably find an annotation less than helpful, to the point of uselessness, if it goes on and on with all kinds of scholarly knowledge, especially if the information the readers want isn't easy to find within the exposition. And yet other readers will probably desire the extra information, and even some beginners will become interested enough in the topic to be inspired to stop their reading to learn more.

The best solution to this situation, and one that digital presentation can do well, is the method I've already outlined: give in the first instance only a short introductory note but provide clearly marked links in that note that lead to more information. Readers who do not want anything more can quickly move back to the text, and those who do can follow the links that are provided.

* Is there a line between information and interpretation? If so, how do we proceed in order not to cross it? If not, how do we construct annotations?

In one sense, any annotation is interpretation. Even if the information is completely factual (historically, there was a Gold Cup race run at Ascot in England on June 16, 1904), the decision to provide the annotation involves an interpretation regarding what matters in the text of Ulysses. Of the existing annotations to the Throwaway passage, Johnson's stays most strictly at what might be called a factual level. The first Gifford and Seidman annotation to some extent, and the second one to a much greater degree, move into the realm of interpretation by deciding, or assuming, that telling readers information that they will not encounter in the book for hundreds of pages is an acceptable practice. Kiberd's note rests at the other extreme on the fact-interpretation scale from Johnson's: it seems to exist mainly to provide the annotator's interpretation of the passage.

We are constructing the annotations to the digital Ulysses so that the ones at the first level come close to Johnson's practice. Annotations aimed at advanced readers can adopt procedures like Gifford and Seidman's, and eventually the annotations become fully interpretive. However, digital presentation does not need to remain single-voiced in the way a print book usually is, and so whenever an interpretation appears, a conflicting or complementary one, or one simply in a different voice, can also appear. Readers can choose to close off all interpretations except one, but the default position will be that Ulysses can be interpreted in many ways and in many voices and not by a single "authority."

* Do these questions change for different categories of information (historical, other languages, intertextuality)?

A short note with links to more detailed information might be a desirable norm for annotations, but not all details from Ulysses will lend themselves to such well-ordered annotations. Some passages cannot be discussed at all except at the level of interpretation. In the J-Joyce listserv discussion of annotation, Andrew Blom suggested one such detail: it involves an Italian exclamation, "Giŕ," near the end of episode 3 ("Proteus"--42; 3:392-96). Gifford and Seidman annotate the word as an adverb, meaning "already" or "Let's go. Let's go" (66). Blom suggests, however, that the word is an exclamation and derives from the German ja; he translates it as something like "yes," "sure," " of course," or "right," as in "That's me, all right. Yeah, sure." If neither translation is erroneous, then no single note can provide a quick first annotation. The initial note will itself be an interpretation.

A similar and even more direct situation exists regarding an incomplete phrase that Leopold Bloom writes in sand on a beach at the end of episode 13 ("Nausicaa"). He writes "I." and then "AM. A." before he covers over the letters with sand (312; 13:1256-69). The text never specifies what else Bloom intended to write, if anything. Annotators who choose to comment on "I. AM. A." can do so only at the level of interpretation, since any consideration of what Bloom might be writing can be only speculation.

The complications that this question opens up do not affect the appearance of the annotations or the ways in which readers call them up, but they do suggest that the separation between factual information and interpretation is a thin, perhaps nonexistent, one and also that some first-level notes will have to be content to say something like "Critics disagree on the meaning of this detail."

* * *

Early on in Ulysses, Bloom thinks about a time when his wife Molly asked him to make love with her. She said, "Give us a touch, Poldy. God, I'm dying for it" (74; 6:79-80). A reader can interrupt an experience of this memory by turning to one of the published sets of annotations to find "a touch" defined as "slang for sexual intercourse" (Gifford and Seidman 107) and might find the note helpful in increasing the erotic implications of "touch" or even in extending the response into the emotional ("touched," since Bloom is remembering the love-making that produced his and Molly's son Rudy, who died eleven days after he was born). Or the reader might find the note intrusive, socially inept, even "touched" in the sense of unbalanced or demented. Footnoted annotations, whether in print or in electronic hypermedia, can be all of these. It is our challenge in "Digital Ulysses" to demonstrate, perhaps in a new way, the value and utility of annotations.

Notes + Bibliography

© Michael Groden, 2004
volume 4, issue 2, 2003-4