Ulysses by Numbers

The previous chapters consider questions involving who wrote Ulysses, why it is as long as it is, how many characters there are, and who read the first edition, but there’s one more that needs answering: When did Joyce write Ulysses? Anyone who finishes the novel will claim that the answer can be found on the last page: 1914–1921 (figure 5.1). Though certainly the quickest answer, it’s not the most accurate, and that’s because this process of beginning, middling and ending was much less straightforward than Joyce is letting on here. By dating Ulysses, it’s as if Joyce needed to reaffirm the whole thing happened at all and that he was more like the wandering hero moving through the world than the “God of creation” hovering above it.¹ But what Ezra Pound first said when coming across the dates appended to the last page of A Portrait of the Artist as a Young Man applies here as well: “It is the [time] spent on the book…that counts. No man can dictate a novel, though there are a lot who try to.”²

Any hope of determining what the dates in Ulysses actually count means sifting through, and at times arguing against, a century’s worth of critical assumptions about where these numbers even came from, and it requires working with biographical hearsay, anecdotal invention, and evidence from the manuscripts, typescripts, and proofs. If to begin and end means to write the first and last word, page, or sentence, then these dates are certainly wrong, but any effort to correct them by moving the years forward or backward would inevitably disrupt the symbolic duration of seven.3

Getting that timeline adjusted to grapple with this ambiguity is not my only objective in this chapter. I also want to use it to examine, in more general terms, how compositional time can be measured, and in doing so think more broadly about the strategies used to access and visualize the creation and reception of this or any other literary work using numbers. In what follows, I’ll be considering Joyce’s dates both as an attempt to quantify time and challenge the idea that time can be quantified at all. Joyce may have left this timestamp behind for others to see, but it was one that foregrounds the complexity of seeing time as a line. And in this he is not alone. Joseph Priestley, the eighteenth-century polymath who revolutionized the design of timelines, summarizes just how strange the whole enterprise can be, writing in the introduction to his Chart of Biography:

But 1914–1921 doesn’t belong to the life of a man, or an empire, or a period: it is a literary timeline. This device is still getting used to organize writers, works, and movements across decades and centuries, mixing together biography, world history, with moments of creative invention. Frankenstein would be one example: the origin of the novel pinned down to a single evening on June 16, 1816 when Mary Shelley got her inspiration for the story whilst staring at the moonlight from her window.5 On a timeline, Shelley’s moment of creative inspiration could also be linked up with a more expansive set of circumstances involving the Enlightenment and the rise of Romanticism. In the process, Frankenstein becomes not only a point in time on the line, highlighting forces, ideas, and events converging at a specific moment, but it is also used comparatively to trace an evolutionary narrative at different scales (years, decades, centuries).

As pervasive as they may be, and whether as images on paper or in our heads, no one is ever entirely satisfied with literary timelines. They are either too general or too specific, too arbitrary or too deterministic, too schematic or too simplistic. But whatever the reason, the timeline is not something that most of us would want to stake literary history on. It’s too messy, too much overlap, too many belated events or writers, works that are forgotten and recovered later, and too many writers and works outside of their own time. To think of literature on a timeline would mean to think of literary history as a sequence. That’s where that timeline from Ulysses comes in. Instead of waiting for the professors, Joyce took the liberty of doing it himself. It was a willful act done to control the time of the novel in history. But let’s not forget it was done by someone who had just experimented with narrative temporality, using one novel to show how easily it can get manipulated. Yet, still he plants a timeline on the last page and critics take him at his word, many of them using “1914–1921” either to celebrate Joyce’s heroic resilience (as if he were the only novelist to take so long!) or to establish correspondences between the timing of the novel with the beginning of World War I and the founding of the Irish Republic.

But Margaret McBride is right to remind us that the “critical event…could be the novel’s composition.”⁶ Eventful as it may seem, Joyce was not the first or last one to put a date at the end of a novel.⁷ In fact, he may have come across an example in Lord Jim, though with a few telling differences: the dates in Joseph Conrad’s novel, which include the month and the year (“September 1899–July 1900”), appear awkwardly after the last sentence and before “The End.” In all of Joyce’s novels, there is a marked difference: the dates are the end, even though the temporal measurement is left more ambiguous. Joyce decides to insert a year and never a day or a month or an hour, as if the time of composition was not, unlike the plot of the novel itself, something that needed to be pinned down.⁸ The absence of such precise information is revealing, especially when you consider that in the course of writing Joyce actually did have exact months and days in his mind. And this attention to numbers would often lead him to find more cosmic correspondences between the publication dates of his books and the birthdays of him, his friends, and figures from the past whom he admired.

Conrad’s Lord Jim may be a convenient point of comparison, but it’s also necessary to establish from the start that the practice of dating books was not a convention at the Imprimerie Darantière. Between 1922 and 1925, William Carlos Williams’s Spring and All, Gertrude Stein’s The Making of Americans, John Rodker’s Montagnes Russes, and Adrienne Monnier’s La Figure all appear without dates on the final page, indicating that it was more of a personal choice than anything else. Joyce, unlike so many novelists before and after him, didn’t just want to have some vague “sense of a beginning” (Edward Said) or “sense of an ending” (Frank Kermode):9 he needed to leave behind a very public timeline and it was one that appears, ironically enough, after the end of an episode (18) with “no beginning, middle, or end.”¹⁰ But no matter how mystical, cyclical, abstract, or elastic time becomes in the plot of Ulysses, there was no getting around the fact that on the page everything had to begin and end somewhere. And since the time in the novel would eventually expire, the act of anchoring Ulysses in time was one way to anticipate what it might come to mean to literary history and, perhaps a more dramatic a claim, to world history.

By dating Ulysses, Joyce left behind a temporal measurement that would not have been discernible anywhere else in or on the book, and the significance of this gesture is far from obvious. Dates, of course, anchor Ulysses in time, thereby corresponding with Alain Robbe-Grillet’s suggestion that “the writer must proudly consent to bear his own date, knowing that there are no masterpieces in eternity, but only works in history.”¹¹ But as much as “1914–1921” may have been a note in a novel (like a message in a bottle) left behind for the future, it is also backward looking—there to signal that a significant event had already passed. These numbers are there as evidence that Ulysses is very much in and of its time, but like Petrarch with the dates attached to his imaginary letters to Cicero and Vergil, they also establish a distance from predecessors both ancient and modern and emphasize that even if he is in his time, he is not entirely of it.¹²

Joyce needed that autobiographical detail instead of a fictional finis, but what does it mean to think of Ulysses in this final moment not as a novel with words to be read but as a collection of numbers to be counted?13 Indeed, 1914–1921 is a date with a duration, but we might also think of it as a mathematical equation: 1914 + x = 1921 or, by reversing it, 1921 − 1914 = x. In both variations, x is the unknown quantity. It is the variable you need or the sum you get when adding or subtracting these numbers. And as simple as the equation might be, the act of counting, in fact, proves critical to the success or failure of this moment. Readers need to do the math, and what they discover is an unwritten seven.¹⁴ By leaving this number suspended on the final page, Joyce ensured that every copy would be numbered in perpetuity, each one counting the same no matter how many different editions appeared or copies sold.

Consider another possibility. What if the numbers of Ulysses existed before they were counted by Joyce or anyone else? This computational paradox is reminiscent of the one identified by William James when he explained how the very act of counting the known planets—seven of them, no less, when his argument was made—can make a manufactured truth seem as if it was always already there. In this “quasi-paradox,” as he calls it, “something comes by the counting that was not there before. And yet that something was ALWAYS TRUE. In one sense you create it, and in another sense you FIND it.”¹⁵ With Ulysses, something similar is happening. Seven years is a manufactured truth waiting to be found by readers, but it is also something already there. Joyce is not creating ex nihilo. Rather, he developed a strategy to reveal a truth about the time of composition that would have otherwise remained unknown.

And what’s more, that seven has a mystical correspondence with the six days of creation and one day of rest, but so too does the sum of numbers containing it. 1 + 9 + 1 + 4 + 1 + 9 + 2 + 1 = 28: it’s a perfect number built from the sum of its positive proper divisors (1 + 2 + 4 + 7 + 1 4 = 28), one that Pythagoras, Euclid, Philo of Alexandria, Nicomachus, St. Augustine, and others associated with the cycle of the moon. This presence of seven and twenty-eight, then, is one more signal that Ulysses exists beyond the limits of any mortal creation, thereby becoming part of a numerical system that organizes the universe while also fitting into the chronology of a much longer literary historical timeline reaching back as far as Homer and ahead to whoever might arrive in the future.

Dates, like so many of the other types of numbers I’ve already discussed in this book, can have a mystical charge that infuses the object or event with meaning. But let’s face it, they are also the data of literary history; the two terms sharing the same Latin root, dare (to give), but also dactylus, or “finger,” which is there to point. The dates in Ulysses are a given, but they’ve become signposts marking a significant shift not just in the history of the novel but of art more generally. To say that dates are data is to recommend that we consider what and how they are marking time, making it not only legible but also coming to influence how we continue to organize the past for the present.

I said this timeline is basic, but that’s in format only. All of the temporal data about the literary works is actually much more complex than it might seem. Of the eighteen (or nineteen) plays Shakespeare wrote, there is a discrepancy in every case among the composition, the performance, and publication (if they are even known at all). So, the 1595 of King John, the 1602 of Hamlet, and the 1606 of Macbeth are all misleading, since the first one is based largely on stylistic evidence (lacking external evidence and internal topical allusions), the second corresponds with the performance, and the third to a revision by Thomas Middleton and not corroborated by evidence from an original, and likely shorter, prompt book.

“Shakespeare Dates” is the title Joyce added to another notebook he used in 1917 when preparing the Hamlet chapter (episode 9) for serialization. Each year is given a separate page, and this time around he identifies whether the plays were recorded, revised, produced, staged, or begun. In addition, he intermittently includes a running log on the right hand side tracking the relative ages of Shakespeare and his family members (1593, 1596, 1601, 1613). In 1593, at twenty-nine years of age, Shakespeare publishes Venus and Adonis and “probably” begins his Sonnets. He returns to Stratford in 1596 at the age of thirty-two (after eleven years), turns thirty-seven in 1601 when his father dies, followed in 1613 by the death of his brother Richard when he is forty-nine.

The timeline Joyce made for Bloom in late 1921 works differently (figure 5.2).¹⁷ A sequence of years runs down the left hand side of a single page from 1866 to 1904, half of them containing facts about Bloom’s personal and professional life, most of them confined to the years after 1886 when he’s reached the age of twenty. The birthdates of Molly, Milly, and Rudy are included, along with the streets where the Bloom family lived, and the timeline ends with an entry that Milly Bloom is in Mullingar in 1904. Jotted in pencil on the right-hand side, are a series of running calculations comparing the respective ages of the Blooms in different years, the same kind he did for Shakespeare. He needed these calculations during the revision process as he went back to add memories to earlier episodes and implant others in the later ones. All of these numerical details were a way to give the Blooms a more substantial past, and John Henry Raleigh has pointed out that many of them were tweaked along the way proving that the timeline was made to be modified.¹⁸

Which gets us to one of the more puzzling entries: Molly Bloom’s birthday. The year 1871 includes a note reading: “M.B n [nata],” or, Molly Bloom born. It is one of the calendrical mysteries that have kept the critics busy, in part, because there is so much conflicting evidence in the novel, the notes, the letters, and the timeline.¹⁹ One thing everyone can agree on is the day: Molly’s birthday takes place on September 8. Eight is a number (vertically) and an infinity sign (horizontally) packed into one. But it’s the day/month/year that Joyce added in pencil next to Molly’s birthday that requires more scrutiny: 8/9/[1]921 marks Molly’s birthday all right, but that’s seventeen years after the plot ends, which should make us all wonder what this time-warped detail might be here (figure 5.3).

Here’s one possibility: Joyce, as he was finishing the final pages of episode 18 in early September 1921, realized that he was writing about Molly on what would have been her fiftieth birthday.20 That note, in other words, is there to mark the writing event itself. But there’s more. What if we also consider that 8/9/[1]921 represents the complete fusion of the biographical time of the character and the compositional time of the author: on September 8, 1921, Molly Bloom finally has a complete birthday because her episode is written. And though it may be true that Joyce shared the real year of Molly’s birthday with his friend Frank Budgen a few weeks earlier, this is the moment when the creation of Molly is complete. By seeing her birthday as if it was fifty years later is the surest sign that the fictional clock ticks in tandem with real time. His novel may have an end, but the characters still live on long enough to see their own story written down.

If Molly has a birthday without a definite year, the reverse is true for Leopold: he has a birthyear without a day.²¹ Thirty-eight years old when the novel opens, his age coincides with Joyce’s during the composition of Ulysses, but it’s also the age of Shakespeare when he wrote Hamlet.²² If, based on other evidence, we can deduce that the month for Bloom’s birth takes place in May, we will also want to remember it’s the name of Joyce’s mother: May Joyce is to James Joyce, then, as James Joyce is to Leopold Bloom. Try as we might, though, the month and the mother still don’t bring us any closer to pinning down an exact day.

And why should it? The year is all that matters when measuring Bloom’s life. I just mentioned that his age has an autobiographical correspondence, but in Ulysses, the mystery of his birthday is given one more twist. We never know a day but there is a period of time marked by the arrival of a star, one of “similar origin” to that appearing at the birth of William Shakespeare, another figure whose exact birthday remains unknown:

the appearance of a star (1st magnitude) of exceeding brilliancy dominating by night and day…about the period of the birth of William Shakespeare over delta in the recumbent neversetting constellation of Cassiopeia and of a star (2nd magnitude) of similar origin but of lesser brilliancy which had appeared in and disappeared from the constellation of the Corona Septentrionalis about the period of the birth of Leopold Bloom and of other stars of (presumably) similar origin which had (effectively or presumably) appeared in and disappeared from the constellation of Andromeda about the period of the birth of Stephen Dedalus, and in and from the constellation of Auriga some years after the birth and death of Rudolph Bloom, junior…(U, 575).

As with all the other answers in episodee 17, this one is scientific in sound and imprecise in sense. There is some astrological information that makes it possible to calculate the season, month of year—but try as everyone might, the day proves elusive. Even with the guidance of the stars, Bloom’s day remains unidentifiable. Bloom is an age instead of aged. Like the stars, his beginning happens at a different scale, one that resists any human desire for the precision of minutes, seconds, hours, and days.

That’s definitely not the design Joyce made for his dates (figure 5.4). Not only was he more liberal with the spacing between the years, but he also made the line in between a lot longer. In the subsequent revisions that took place during the proof stages, the question was never raised, and the hyphen as set by Darantière made it all the way through to the end. If we consider this first draft as an expression of what Joyce wanted, then we might try and imagine what that longer line means. More time, for one thing, but it also provides more space for an implied sequence of years: 1915, 1916, 1917, 1918, 1919, 1920.

Why stop here? That line could be made even longer, stretched in both directions so that it actually fills the same distance as the “Trieste-Zurich-Paris” looming above (figure 5.5). Time and place, then, would be in sync, the duration of the line corresponding with the wandering in and between the three cities. All of this might sound a little too playful perhaps, more of a fantasy scenario about what this detail in the novel could be instead of an assessment of what it is. But that’s precisely what I want to avoid: if all the other pieces of punctuation matter, many of them leading over the years to contentious debates about what Ulysses should look like, this line is one of the more provocative examples that has fallen entirely off the radar. So quick are we to calculate the number, that this small segment of line has been ignored.

Once you begin to imagine the line less as the final word (one marking the moment when time passed) and more as the material expression of time passing, then we’re getting closer conceptually to Henri Bergson’s durée. For Wyndham Lewis, “there would be no Ulysses” without Bergson.²³ He’s referring here to interior monologue, but this same sentiment can extend to the timeline. More specifically, I am referring to what Bergson says about time’s qualitative dimension. “Pure duration,” Bergson argues, “might well be nothing but a succession of qualitative changes, which melt into and permeate one another, without precise outlines, without any tendency to externalize themselves in relation to one another, without any affiliation with number: it would be pure heterogeneity.”²⁴ Joyce’s timeline has the effect of reducing a qualitative, creative experience to a quantitative one. The numbers speak but the emphasis falls on the final sum and not the relations, affiliations, and juxtapositions that are, in the end, so much a part of any creative process no matter how unidirectional or sequential we try to make it seem.

I bring this up if only to emphasize how we can imagine these dates as an intrusion, an attempt to control the time in which the work itself, as a creative composition, is being fashioned for the future. Joyce wants the compositional time of Ulysses to resemble a line when, as I’ll explain in what follows, it is not: dating Ulysses was a much more fluid, unfixed process with multiple possibilities. Not only does that line extend forward and backward, it tries to reduce the event to a succession all the while masking the question about what it means for this work, or any work or art for that matter, to begin and end. The timeline, you could say, provides a duration but masks a durée. Creation is never so clean and attempts to quantify it are one way to make the process less messy, as if the world within appeared by fiat and not, as we know, through the labor of thinking, writing, and revising.

The timeline in figure 5.6 is one attempt to represent the complexity of compositional time. We may have Joyce’s clean and compact version in the novel, but we also know from all the archival evidence that there are other possibilities, and they involve identifying earlier material that was reused, the erratic dating of Joyce himself, and the attempts of critics to identify a prehistory of the novel. This timeline is not complete by any means, but it gives us an opportunity to think seriously about what it might mean not only to imagine different durations for the composition but also different points of beginning and ending. This timeline may not provide a definitive answer to the question of when Ulysses began, but it forces us to consider the assumptions we make about the creative process when we go about looking for an answer and then trying to represent it.

So, the beginning: 1914 turned out to be a tragic year in world history, but when Joyce first started writing Ulysses, or eventually claimed to, nothing of the sort could have been known.25 In his early study of the pre-1917 Ulysses, Rodney Owen was the first to argue that this date (a year) was a gradually discovered fiction and not even firmly in place until 1920.²⁶ That, however, is not the story that Herbert Gorman, Joyce’s first biographer, recounts: “In the spring of 1914 he began Ulysses, setting down, first of all, the preliminary sketches for the final sections.”²⁷ This particular account may sound objective enough, but only if we are willing to disregard what he told Harriet Shaw Weaver in November 1921: “Ulysses began on 1 March (birthday of a friend of mine a Cornish painter) and was finished on Mr Pound’s birthday [October 30].”²⁸ Gorman had the year, but Joyce had an exact day and considering his superstition about numbers and birth dates, he probably an hour and minute. Evidence such as this certainly suggests that Joyce did end up having an exact moment in mind for the beginning and end of his novel, but it was not always fixed. In fact, even this one, as precise as it may be, gets contradicted in other letters addressed to Pound and McAlmon in which the beginning is moved a month earlier and the end a day earlier.

In Gorman’s version, one that was likely vetted by the author himself, Joyce begins “first of all” with “sketches” for the final episodes. The term sketch could refer to a rough draft, an outline, a series of notes or words, or a few lines or paragraphs, but in an effort to provide some clarification, Gorman goes on to explain that “It would be more precise to say that he began to put Ulysses on paper, for the theme as a whole, had been swelling in his mind for some years, growing secretly at first from the seed fecundated in Rome, then blossoming into his consciousness, and finally completely absorbing it.”²⁹ In this tortured organic metaphor, Gorman seems to suggest that beginning can certainly involve putting words down on paper, but it might also coincide with the moment when the seed for the story was first planted (not necessarily involving paper), which in this case would be the period when Joyce was living in Rome and first considered adding it as a short story to Dubliners.³⁰

So why did Joyce insist on using 1914 and not 1906 or 1907 (figure 5.7)?³¹ It wouldn’t be the first time he identified the date of writing a novel with the mere seed of an idea. After all, 1904 was not when he actually started writing the novel called A Portrait of the Artist as a Young Man: it was the year when he composed the essay “Portrait of the Artist.”³² Whatever the reason, Ezra Pound was not one to get misled, and in a review of Ulysses published in the Mercure de France, he told his French readers that “les romanciers n’aiment dépenser que trois mois, six mois pour un roman. Joyce y a mis quinze ans. Et Ulysses est plus condensé (732 grandes feuilles) que n’importe quelle oeuvre entière de Flaubert; on y découvre plus d’architecture.”³³ By 1922, Pound’s fifteen years puts the beginning of Ulysses back in Rome 1906. In an undocumented conversation that was supposed to have taken place on the day Ulysses was published that’s more in line with Nora Joyce’s claim, “he had thought about the book for sixteen years and spent seven years writing it.”³⁴

It was Joyce, not Pound, who got the last number, and he used Gorman’s biography in the 1930s to suggest, in retrospect, that the choice of a date actually had a lot to do with the good fortune he experienced that year with Portrait getting serialized, Dubliners finally published, and Exiles partially written. In 1914, Joyce was on a roll, and it marked an annus mirabilis in a literary career that had been more horribilis up until that point. The beginning of Ulysses, then, was bound up with a series of endings involving authorial victories that had so far eluded him, and if in this early biographical account Gorman is clear about where Joyce was in the novel before leaving Trieste for Zurich in June 1915, even going so far as to identify the very line where he stopped in episode 3, he says nothing about the first word, phrase, line, or set of lines, and no evidence has survived to clarify matters.35

Ellmann’s revisionary account doesn’t help much either. Like Gorman, he mentions that Joyce had been “preparing himself to write Ulysses since 1907” [imagining a “Dublin Peer Gynt”], and he even recycles the point that Joyce had reached “the first pages of the third episode by June [1915]” before leaving for Zurich. Unlike Gorman, however, Ellmann actually had access to letters, which reveal that even if Joyce had started writing sketches of the final episodes in the spring of 1914, he was not talking about beginning Ulysses until he finished the first two episodes in June 1915.36 In a letter to his brother Stanislaus, dated June 16 no less, he writes, “Die erste Episode meines neues Roman ‘Ulysses’ ist geschrieben” [“The first episode of my new novel Ulysses is written”] and to Pound a few weeks later, he reports, “I have written the first two episodes.”³⁷ This lag suggests that even if the writing had begun a year earlier, the novel itself was not ready for any official announcement until the actual beginning was in place.

Once there is a beginning to the plot of Ulysses, Joyce sets out to establish one that would coincide with the origins of the novel Ulysses. Having told Pound about his progress, he follows up a few weeks later asking if he should send “the first chapters of [his] novel” to Mr. Gosse, adding, “it began on 2 Feby 1914.”³⁸ It would seem logical to assume that in using the phrase “first chapters of my novel” Joyce is referring to Portrait, which started serialization in the Egoist on that date.³⁹ However, considering the fact that Pound was the one who had arranged for the serialization and saw each of the installments through, then it would seem strange for Joyce to remind him of the date, and, more to the point, it is worth emphasizing that Joyce informs Pound that he “began on 2 Feby 1914,” not began serialization on. Equally problematic is the fact that just two weeks earlier, he had already informed Pound that “the first two episodes” of his new novel were written, thereby making it likely that he was picking up the conversation where it had left off.

Whatever the case may be, though, this particular birthdate for the novel did not stick. By September 1916, Joyce was already recalibrating, telling Pound: “All the rest of my stuff is the bewildering mass of papers for Ulysses which I carry in a very large envelope that I possess [sic] for the last twelve years.”40 That same year, the timeline shifts again, making Ulysses begin somewhere around 1909, when he made his penultimate visit to Ireland and discovered both the address of 7 Eccles Street and the possibility of Nora’s betrayal (later proven false). That same year, Joyce tells Weaver and his friend C. P. Curran that he began “six or seven years ago” and Pound got a slightly different version shortly thereafter: “As regards Ulysses I write and think and write and think all day and part of the night. [It goes on as it] has been going on these five or six years. But the ingredients will not fuse until they have reached a certain temperature.”⁴¹

As usual, these dates are approximate, as if Joyce doesn’t really want to pin down the beginning. It’s always six or seven, five or six, but if you do the math that second number actually brings the beginning of Ulysses closer to 1912, the year, Owen suspects, when Joyce began to gather material for Stephen’s episodes (2, 3, 9, and possibly 7). Litz agreed: “The idea of Stephen’s participation in a later work to be called Ulysses must have been well advanced by 1911–12, and the outline of the whole novel was probably visualized as well since Gorman says that Joyce began to make preliminary sketches for the final sections of Ulysses early in 1914 immediately after finishing the Portrait.”⁴² Gabler takes this possibility one step further, speculating that 1912 marks the “material beginning” of Ulysses when Joyce began to separate scenes, characters, and even sentences for the two novels.⁴³ Indeed, 1912 might not have been an annus mirabilis like 1914, but it was the year Joyce left Ireland for the last time, the same year he delivered his lectures on Shakespeare.

So, why not 1912 (figure 5.8)? If we believe Frank Budgen, that was the year Joyce alluded to when they met in 1918: “In leaving the café,” he recalls, “I asked Joyce how long he had been working on Ulysses. ‘About five years,’ he said, ‘but in a sense all my life.’”⁴⁴ As usual, the number is left ambiguous: it’s about five years and there is never an official day or month. One likely reason: Joyce was still unpublished in 1912 and Portrait was unfinished, so beginning a new novel when the others were still in progress would end up disrupting the seamlessness of his creative chronology, one that he would eventually maximize to full effect with 1904–1914 (Portrait), 1914–1921, 1922–1939 (Finnegans Wake). Another reason: the act of gathering material for two different novels at the same time did not mean that a new project was officially underway. Cogitating, gathering, collecting, and rearranging, was not to be confused with writing. And, then, of course, there’s always Joyce’s numerical superstition. The year 1912 might seem harmless enough until you add the numbers up: 1 + 9 + 1 + 2 = 13.⁴⁵ If this kind of calendrical equation sounds paranoid, consider Joyce’s comment to Weaver in 1921 when Ulysses was set to appear: “It seems to me as if this year (1 + 9 + 2 + 1 = 13) is to be one of incessant trouble to me.”⁴⁶ It turned out that the numbers didn’t end up counting against him, and by 1921, when they had arrived again in a different order, the novel was almost finished and there was a definite end in sight.

Joyce, as I said, needed to have an end date in mind to establish once and for all when he actually got started. By February 1919, ten episodes of Ulysses had been serialized in the Little Review with an anticipated eight more to go, but that was enough for him to begin guessing when it would all be over. “This year or early next year,” he wrote Weaver on February 25, 1919; “will not be finished probably until the end of next year—if even then,” he told Quinn on August 3, 1919. Both estimates were way off. Almost a full year later, the situation was decidedly clearer. Ezra Pound, Carlo Linati, and Henry Davray were some of the first people to get the news and by late November 1920 the beginning of Ulysses was an established fact. “I began Ulysses in 1914…,” he tells Quinn in the bluntest way possible, adding “and shall finish it, I suppose, in 1921.”47 There it is, 1914–1921, a timeline meant to chart the development of a single novel’s becoming, one that so many of his contemporaries would identify with the end of an era.

After six years of writing, the dates of Ulysses were finally set, though the novel was still another year and two months in the making, with two copies appearing on an express train from Dijon on his fortieth birthday. Joyce was fond of staging publication coincidences like these, and they ended up regulating the entire life cycle of his writing. Contrary to what Thomas Hardy believed, the date of publication was not an “accident in the life of a literary creation.”⁴⁸ Rather, it was a deliberate astrological event with the power to connect the life of the author with the work he created. In an interesting twist, it was always years and never days that Joyce made public on the last pages of his novels. By the time he was busy calculating the number of hours spent on Ulysses (twenty thousand by his tally), 1914 would have meant more to the general public than Frank Budgen’s birthday.⁴⁹ It was a year of world historical significance, that, on a personal level, determined the direction of his own wanderings across Europe where he was deliberately avoiding any direct contact with the battlefield. This continued act of writing during the war and in spite of the war is what mattered most. “What did you do in the Great War?,” one of Tom Stoppard’s characters asks Joyce in Travesties. “I wrote Ulysses,” he responds, “What did you do?”⁵⁰

If, as I’ve been arguing, the beginning of Ulysses was a gradually discovered fiction, then so too was its end. The major difference is that we know significantly more about the middle and late stages of the composition process (1918 forward) than we do about the early ones (1912, or earlier, to 1917), in large part, because so much material was preserved, including the detailed correspondence through which Joyce kept friends and admirers abreast of his progress. “A few lines,” Joyce wrote to Robert McAlmon on October 29, 1921, “to say that I have just finished the Ithaca episode so that at last the writing of Ulysses is finished.”51 In this letter and others, Joyce selects an end date in which he distinguishes between writing and revising, and having finished his last rough draft of this episode, he believes that the work he has left to do is fundamentally different: revision and proofreading were not to be confused with composition. Looking through the available typescripts and proofs, however, Gabler and his team concluded to the contrary that Joyce’s work was far from over, and that episode 17, in particular, “was by no means ready to go to the printer.”⁵² In fact, in the following months, he would augment episode 17 by 34 percent, with 16 percent more text added to the notes and typescript and 18 percent to the proofs.⁵³ Richard Madtes, who first compiled these percentages, concluded that “composition may have been ‘at an end’ for most of the book, but certainly not for Ithaca [episode 17].”⁵⁴

This flurry of activity in the final months leaves us with an unresolved question: When did Ulysses end? The end of writing, as I mentioned, is not something that Joyce identified automatically with the act of revising.⁵⁵ They were different activities, the one involving creation from nothing, the other creation from something.⁵⁶ In addition, the dates that marked an end to writing were not to be confused with the publication. The original contract was negotiated in April 1921, but by the end of August, Darantière was already warning Joyce, “it will not be possible to finish Ulysses before the end of this year 1921” [“il ne sera pas possible d’achever ‘Ulysses’ avant la fin de cette année mil neuf cent vingt et un”].⁵⁷

The evidence from one of the fliers for Ulysses reveals that Joyce didn’t care.58 Along with the woodcut image of a meditative bard in the middle of the page is an announcement that “Ulysses by James Joyce will be published in the Autumn of 1921” with a correction to the date, “January 1922.” This move from season and year to month and year was obviously intended as a necessary point of clarification, but what’s so striking here is Joyce’s explicit acknowledgement that Ulysses was not going to be ready for publication before the year’s end. In fact, since this correction to the flier was made while the book was still being revised/written, it also suggests that the end of Ulysses was in sight though it was not yet an event that would coincide with his birthday.⁵⁹

None of this, however, was enough to make Joyce consider revising the one detail in his novel that could have used it: 1921.⁶⁰ Here the genetic evidence is hard to ignore. Joyce wrote a version of the final sentence of episode 18 between the end of August and late September when the final yes was still uncapitalized. But even then, with several months of revisions before him and episode 17 still unfinished, 1921 was going to be the year when Ulysses came to an end.

In these final months, Joyce was not just adding more material to his already massive novel: he was counting the time it took. And in doing so, he refused to keep the clock running as the revisions took him and his printers into another year. Indeed, Joyce may have designated October 30, 1921 as the official end, but the proofs are evidence enough that the work actually took three months more and up to January 31, 1922, two days before the first two copies were printed and delivered. All of this might seem trivial enough. Joyce’s distinction between writing and revising, one could argue, provides enough justification for this choice. Yet that’s only convincing if we are willing to ignore the fictionality of the beginning as well, a fictionality that actually determined when the novel itself could end in the first place. By November 1920, and with thirteen episodes already printed in the Little Review, there really was no choice for another end date. Joyce had to finish in 1921 in order for the numbers to add up. Instead of a Homeric ten years, Ulysses would take seven.⁶¹

Why seven?62 Kenner, as I mentioned in the introduction, was the first critic to pick up on the fact that Ulysses is a title made up of seven letters and is the “sole remark the author…permits himself amid a quarter-million words.”⁶³ In the novel itself, seven is the number of monsters Bloom must bypass (episodes 4, 8, 12, 11, 5, 9, 13). John Eglinton reminds Stephen, seven is “dear to the mystic mind” and Bloom, while farting, conflates Robert Emmet’s last words with the “seven last words” of Jesus Christ (U, 151, 238). This number appears again in a poem by Yeats, in reference to the number of known planets in 1895, and to identify the years D. B. Murphy has been at sea and the miles Bloom can stick his tongue up Molly’s “hole.” In Hebrew, Greek, Egyptian, and Eastern traditions, Seidman and Gifford tell us, it was a number representing, “the embodiment of perfection and unity, mystically appropriate to sacred things,” and for early Christian writers, it was “the number of completion and perfection,” corresponding as it did with the days of creation.⁶⁴ Of course, seven is also the number of signatories on the proclamation of the Irish Republic, the colors of the rainbow, the continents and seas, the notes on a numerical scale, the stars of the Big Dipper, the feelings of mankind (joy, anger, sadness, fear, love, liking, disliking), the wonders of the world, the cardinal and theological virtues (prudence, temperance, fortitude, justice, faith, hope, charity), the classical planets to the ancients, the liberal arts in the Middle Ages, and the days of the week.

On the last page of Ulysses, seven, as I mentioned, is not a number that gets mentioned explicitly. Rather, it is the sum derived from an equation that can involve either subtraction (1921−1914=7) or addition (1914+7=1921). Once discovered it can suggest unity and perfection while also corresponding with the life cycle of the novel, a novel, Joyce once claimed, that was an epic of the human body, which, if you take the analogy far enough, involves the inevitable aging process for author and book alike. Consider, for instance, his original, unrealized plan for Stephen Hero in 1903: an autobiographical novel that was to be comprised of nine sections each one made up of seven chapters. Gabler has argued that the sixty-three “seems to have been numerologically related to the periods of life of a man” that would include infancy, boyhood, adolescence, young adulthood, middle age, early old age, and later old age. As much as Ulysses was arranged by organ and divided by hour, it was also intended as a document of an entire human life, and each year of its becoming corresponding with one of these seven stages compressed into the hours of a single day. Considering the number of allusions to As You Like it, in fact, it’s impossible to believe that Shakespeare was not somehow behind the idea. “All the world’s a stage,” Jaques says to Duke Senior, “And all the men and women merely players;/They have their exits and their entrances,/ And one man in his time plays many parts,/ His acts being seven ages.”65

Seven ages, seven acts, seven stages. If Shakespeare’s stage (set in a Globe no less) could contain worlds and reveal the lives of so many players moving across it, then why not a novel in Dublin that begins and ends on 7 Eccles Street? The dates at the end are, in a sense, an attempt to reaffirm just that with the emphasis always on the time and place of creation. Ulysses, like every novel before and after it, must come to an end, but the seven years are there as evidence that something transcendent has happened in the process. No matter how many times this novel ends, no matter how many readers flip to that final page, something mystical takes place, an experience in fictional time and in a fictional world that leads outside of time, and much like Stephen and Bloom entering the cosmos with the final “Where?” of episode 17 (the first ending) or Molly Bloom passing into eternity after her emphatic “Yes” in episode 18 (the second ending), readers of Ulysses have no choice but to come face to face with a set of numbers waiting to be counted. Joyce, Ellmann tells us, looked forward to the “expiration of [his] seven years’ sentence,” picking up, no doubt from Gorman, who, in the first authorized biography, claimed “It was a long period—seven years, and it seemed even longer because of the hectic era that had been packed into those years.”⁶⁶

An age, a stage, but all together seven years that define an era. That helps to explain why none of these numbers matter much on their own. This paradoxical condensation and dilation of time requires the simple act of solving an equation, and the dates, the novel’s last word, are a way of summing up the larger point that the time of composition is strangely equivalent to the chronology of the plot. The labor of Ulysses might have continued up until the last day in January 1922 (with the first copy arriving at 7 a.m. on February 2), but in order for it all to add up, the novel itself had to end before the writing was done, and in doing so, Joyce put Pound’s statement about dates to the test. It’s the time spent on the novel that counts. If you take Pound at his word, it all adds up: 1 novel=7 years. Seven, it’s the number of perfection, unity, days of the week, the stages of life, the years Odysseus spends on Calypso’s island, and, from that moment forward, it is the amount of time Joyce took to write a novel that would mark the end of one era and the beginning of another.

The high watermark for modernism might be 1922, but Joyce’s contemporaries weren’t entirely sure if the arrival of Ulysses marked a beginning or an end to writing, to the novel as a genre, to modern literature, or even to civilization. “If it does not make an epoch—and it well may!” Ford Madox Ford wondered, “it will at least mark the ending of a period.”67 T. S. Eliot agreed, and his positive assessment recalled decades later by a more skeptical Virginia Woolf: “How could anyone write again after achieving the immense prodigy of the last chapter.”⁶⁸ For Pound, Ulysses represented a “dead-end” in the history of the novel, and he identified it more broadly with “the end of an era or a cycle of composition.”⁶⁹ “An END not a start,” he told Wyndham Lewis. This end may signify an irrevocable break with the past, but Pound also saw it as a unique opportunity to begin again, telling H. L. Mencken: “The Christian Era ended at midnight on Oct. 29–30 of last year. You are now in the year 1 p.s.U., if that is any comfort to you.”⁷⁰ p.s.U., as I mentioned in the introduction, was Pound’s acronym for “post scriptum Ulixes,” inspiring the calendar which he published in the spring 1922 issue of the Little Review.

In this incipit of a p.s.U. generation there lies another calendrical contradiction. Here, Pound identifies not just a month and a day but an exact hour for the transition from one era to another, but when, to return to the question I raised earlier, was Ulysses finished? In Joyce’s letter to McAlmon, October 29 was the day when he “just finished” episode 17, but in his subsequent letter to Weaver it was pushed back a day later (“Mr Pound’s birthday”). Given Joyce’s intense fascination with birthdays, it’s likely that the timing for the coincidence could be tweaked easily enough, but what’s looming behind it all is a fantasy that Ulysses was finished on the same day when Pound, the midwife for the novel if there ever was one, was born.⁷¹ Joyce may have stressed an actual birth day, but it was Pound who emphasized the diurnal transition so that the move from one day to another represented the dramatic shift from one era to another and the full meaning was less biographical (or biological) than it was literary and world historical.

Under this new calendar Ulysses was a morphological nightmare from which literary history would need to awaken, one that forced readers and writers back to the past to try and figure out where it came from, and for world history, it was a publishing event with the power to provide a glimpse into an uncertain future. In both cases, supporters and detractors alike were left trying to assess what the birth of Ulysses did to literature in general. Though opinions certainly varied, there was consensus from Joyce’s contemporaries that from February 2 forward the world of letters was never quite the same. It was Ulysses that made 1922 memorable and not the other way around.

The afterlife of Ulysses was not something Joyce could control, but that in itself can help us understand how we might continue to decode those numbers on the last page. They are, as I’ve already mentioned, one way to affirm the act of authorial creation by situating the novel’s production in time. But what’s left on the rest of the page matters as well. In the handwritten fair copy draft, there is no punctuation in his timeline. Instead, the bottom half of the page is mostly blank, which could signal creation ex nihilo and/or the destruction ex post facto. When Darantière’s printers decided to insert a period after 1921 in the typescript, however, they inadvertently changed what both dates meant, cordoning them off as if to stop the clock from ticking, and by the time the proofs arrived for the final episodes, that was it (figure 5.9). Ulysses was already a book with a clearly defined lifespan, those dates on the page reminiscent of the ones found on tombstones. A strange idea, maybe, but it’s one inspired by an observation Pound himself made in a 1941 radio broadcast comparing Ulysses to a “tomb and muniment of a rotten era portrayed by [the] pen of a master.”⁷²

In the final months, the periods in Ulysses, inserted to mark definitive ends, would cause a great deal of confusion with printer and author alike. Not only did Joyce have to remind Darantière and his fellow compositors that the enlarged em dash in episode 17 was the answer to the final question, but he also had to decide whether or not episode 18 should even have a period.⁷³ Joyce may have claimed that episode did not have a beginning, middle, or end, but he had still not arrived at the kind of inventive syntactical solution he found for Finnegans Wake, a novel that begins and ends in the middle of a sentence.⁷⁴

Putting a period at the end of episode 18 may have been necessary for Joyce to stop the flow of the eight paragraphs, but it also represents one of the novel’s least experimental moments, since the final period closes the novel but creates an artificial break in the narration. After her crescendo, Molly is wide awake and still walking down memory lane, but as readers we are cut off.75 Therefore, instead of beaming off into infinity, the novel closes with an abrupt stop, which should be a reminder that the end of Ulysses, like the beginning, was a fiction from the start, a set of points that had to be chosen so that a single day in Dublin could finally be contained.

Are these particular novelistic fictions so strange? After all, we don’t know when Stephen, Buck, or Bloom wake up in the morning, much like we don’t know when they fall asleep. And even if Molly is in the same place when the novel opens and closes, not even that final period in her monologue is enough to keep her thoughts from continuing, a point, which may help explain just why Joyce added dates after the fictional plot was presumably over. They are an afterword to the novel’s becoming instead of an afterward to the plot’s inevitable end.

1914–1921 may represent a grandiose intrusion by the author but considering what I just said about that last sentence (which really could culminate in an exclamation point or an ellipses), it reminds us that the composition of Ulysses, could have kept on going at least until Molly had fallen asleep. The actual publication of the novel functioned as an interruption in the same way as well. If you consider all of the possible beginnings for Ulysses, the book was eighteen, fourteen, twelve, nine, eight, or seven years in the making, and without the serialization, the legal battles, censorship, eye surgeries, search for a publisher, and an attenuated printing process, the composition could certainly have continued far beyond 1922.⁷⁶ Joyce may have finally stopped writing and revising, but I agree with Groden that he “never really ‘finished’ Ulysses.”⁷⁷ As much as the novel resembles a tomb, muniment (Pound’s phrase), body, machine, city, or cosmos, it is also a lot like non-finito sculpture, not because the artist didn’t have a vision for what he wanted to create but because the experiments could have continued if he didn’t finally put his pen down.⁷⁸

The dates of Ulysses have been convenient facts that we as readers latch on to try and make sense of what this novel is, where it came from, and what it did, and it involves attributing value to an arbitrary connection between real events and concrete numbers. We can continue to do so, of course, if we are willing to acknowledge that Joyce at least refused to let the arbitrariness of compositional time disrupt any correspondence with world history. In one of the more popular interpretations, which follows Siegfried Kracaeur’s reminder that “the date of the event is a value laden fact,” the beginning and ending of Ulysses coincide with the outbreak of World War I and Irish independence.79 This epic perspective on the timing of his novel is one that Joyce encouraged early on when he reminded Budgen that “the history of Ulysses did not come to an end when the Trojan War was over. It began just when the other Greek heroes went back to live the rest of their lives in peace.”⁸⁰ What the Trojan War and the journey back to Ithaca were for Odysseus and Homer, then, the Great War, which anticipated the Anglo-Irish war, were to Leopold Bloom and Joyce. But as much as readers are free to decide what does and does not happen in the novel, they can never choose the when. That was Joyce’s job, and as Gorman colorfully recounts it in his biography, the definitive moment when the pages of Ulysses were all set by Darantière and his compositors and unable to be revised further, the overworked author could finally “sit back and stretch his legs a bit and regard the world at his leisure.”⁸¹

His friends and family, it turns out, were not afforded the same luxury. Once the novel was in print, Joyce was anxious for it to get read, and the question “Have you finished Ulysses?” was a common refrain in the subsequent months, even though he was soon left wondering if the delays with reviews had something to do with the fact that his book was “so long.”⁸² By March 1923, though, he was at it again telling Weaver that he had written the first “two pages” of another novel.⁸³ When he revised the proofs for Finnegans Wake fifteen years later, there was no question about when it all began: 1922. At this point, it had to be if he wanted to imagine his life and work fused together. The clock of his career was already ticking, the sequence of his novels tocking, and it would all come to an end in 1939 just as he could have never counted on, with another world war erupting.

1914–1921 does not only belong to the production of a single novel: it is part of an ever-evolving literary-historical timeline. And what Claude Lévi-Strauss said about histoire (“Il n’y a pas d’histoire sans dates”) is true of histoire littéraire as well: there is no literary history without dates.84 Just try describing Dante’s Commedia, the age of the Victorians, the rise of the bildungsroman, or the end of the epic without identifying at least one or two numerical points of reference along the way.⁸⁵ It’s common enough to start with vague approximations about time when literary histories are involved, but eventually, there’s a moment when a before and after needs to be established, and that’s when the date emerges with all of its peculiar force.⁸⁶

Every major period in literary history is marked by one or another significant date, too many, in fact, to list here. It can be a day, but more frequently it’s a year or a well-rounded decade, and they all get used to define shifts in sensibility, innovations in technique, or generic inventions.⁸⁷ But as convenient as dates can be for anchoring chronologies, Hayden White, following Lévi-Strauss, has suggested that they are really problems that need to be solved.⁸⁸ What looks like coherence with dates is abstraction, and what seems like historical truth is something conceptually closer to myth.

Laurence Sterne responded to his own timeline-obsessed age by including a page filled with different squiggly lines in Tristram Shandy mapping out the plot of his novel. The linearity of the line in an otherwise fictional story made of digressions was the target of Sterne’s joke, but he was responding to the incredibly innovative work of contemporaries such as Joseph Priestley who were developing diagrams to visualize history, one of his most famous, a biographical chart (1765) laying out the lives and deaths of the men involved with the science of optics from 1000 C.E. to the late eighteenth century. It was intellectual history at a glance, each scientist building on the discoveries of his predecessor, the ideology of progress clearly on display for all to pick up on, Sterne included.⁸⁹ Priestley and others were playing with the illusion that time is quantifiable, something that can not only be measured but also made visible to document the forward movement of human progress.

With literary history, these ideologies of progress are certainly suspect. Time moves forward, but literature is not better or worse as a result. Still, there remains a fascination with the origins of literary works in their time, and it is one, at its very core, intended to help explain how a work of art happened. The when is a qualitative designation to help us understand the what. Alan Liu faced this challenge more than thirty years ago when he set out to create a “visual flowchart” plotting “the whole course of Wordsworth’s compositional history (using the best available dates)” and soon discovered that critical information was missing, eventually forcing him to abandon the original plan. In the process, he realized that there was still a “need for a visual representation to complement our current, elaborately detailed chronologies with all their repetitions, syncopations, overlaps, local mysteries, and strange loops.”90 When, Liu wanted to know, did The Prelude really begin? Was it the seed of the idea for the poem before a single word was even down on paper? Was it the moment when the first draft had been made or earlier when words, notes, and lines were collected.⁹¹ Instead of providing any real clarity, the collection of dates in his Wordsworth chronology reminded Liu more and more of the fruitless search for the “dream’s navel,” one involving an interpretive process in which dream thoughts cannot be traced to a single origin.⁹²

The same is true of Joyce’s dates on the last page of Ulysses. They are the riddle to the mystery of compositional time and not the answer. But what, finally, are they asking us? So far, we know that they are a highly personalized timeline with historical, biographical, and other symbolic correspondences, but we must also consider the fact that this is Joyce’s version, one with a whole range of distortions, omissions, and obfuscations that I’ve discussed throughout this chapter. As was true with the rest of the paratext, he wanted to manage the expectations of his audience as best he could, and this would include situating Ulysses not just in time but across time. This sequence of years is part of a much longer continuum absent from the page. That is how these dates work. They are an interruption in the timeline of literary history, but only temporarily, before the calibration process restarts and the weighing and measuring of reputation, influence, and achievement can take place. And you should feel free to tweak that line yourself: every direction you extend or shorten will modify what the numbers mean together and separately (figure 5.10). It is a simple act making you an active participant in the process and foregrounds how that qualitative dimension of those numbers can change.

Putting that date on the last page does not pull Ulysses out of history. Rather, it anchors Ulysses in history, one with a previously established chronology, as the definition of the term dating itself implies.93 But, in this case, it’s a chronology whose own beginnings will remain forever uncertain. Homer, whether an individual or the name for a collective tradition, may have been responsible for setting Western literary history in motion, but there was no book that he, or it, could ever put a date on. Turns out that wasn’t necessary. Joyce would do it instead. “The most beautiful book that has come out of our country in my time,” Buck Mulligan, (distorting a line by William Butler Yeats) jokes, before adding, “One thinks of Homer” (U, 178). The wink here is as much for Stephen as for us: the book is Ulysses, the time is 1922, and the whole process started thousands of years earlier long before the Republic of Ireland, books, Mulligan, or Leopold Bloom even existed.