0VERTURE
1. Hugh Kenner, “Shem the Textman,” in Languages of Joyce: Selected Papers from the 11th International James Joyce Symposium, Venice 1988, ed. Rosa Maria Bosinelli (Philadelphia: John Benjamins, 1992), 146. Kenner’s computational optimism runs contrary to Roland Barthes’s more pessimistic poststructuralist claim that “the TEXT is not to be thought of as an object that can be computed.” See Roland Barthes, “From Work to Text,” in Image-Music-Text, trans. Stephen Heath (New York: Hill & Wang, 1977), 155–64.
2. Joyce, Letters of James Joyce, vol. 3, ed. Richard Ellmann (New York: Viking, 1957–1966), 58. Hereafter referred to as L and identified by volume (I, II, or III). In a personal email (February 8, 2019), Daniel Ferrer mentioned that he never noticed it before, but pointed out that the double numbering was occasionally done at the end of the nineteenth and beginning of the twentieth century. A good example, one mentioned by Daniel Ferrer to Michael Groden in an email exchange, can be found on the cover of La Nouvelle Revue Française: 78, Rue d’Assas, 78.
3. Consider as well that the correct numbering of the address occurs on the announcement and subscription forms for the novel: 12 rue de l’Odéon, Paris—VIe. There is no 12 in the second impression done for the Egoist Press in October 1922 (because the address had changed), and the second 12 is dropped in the 1924 edition and after. It’s worth noting that the first and second 2s, taken from 12, occupy the same position: 10 + 2. This correspondence with the date would not work, for example, if they were positioned in 20 (2 tens) or 230 (2 hundred plus 3 tens).
4. Guy Davenport, “Joyce’s Forest of Symbols,” Iowa Review 6, no. 1 (Winter 1975): 75–91. Joyce’s idea for twenty-two (4, 15, 3) episodes was divulged to his brother on June 16, 1915. See Richard Ellman, ed., Selected Letters of James Joyce (New York: Viking, 1957), 209. Hereafter referred to as SL.
5. When serialized in the Little Review, each episode was identified by a roman numeral in the contents page. In the fair copy manuscript Joyce copied himself, each episode has an arabic numeral but each section a roman one.
6. Richard Ellmann, “Ulysses” on the Liffey (London: Faber & Faber, 1974), 2; Vincent F. Hopper, Medieval Number Symbolism: Its Sources, Meaning, and Influence on Thought and Expression (New York: Columbia University Press, 1938), 4, 5, 11.
1NTRODUCTION: ULYSSES BY NUMBERS
1. Appearing six times each, 133 and 102 are the numbers with the highest frequency.
2. Weight reported by Sylvia Beach, Shakespeare and Company: The Story of an American Bookshop in Paris (New York: Harcourt, Brace, 1959), 86. The dimensions provided by Sisley Huddleston in Robert H. Deming, ed., James Joyce: The Critical Heritage, vol. 1 (London: Routledge & Kegan Paul, 1970), 242. Bulk provided by John Ryder, “Editing Ulysses Typographically,” Scholarly Publishing 18, no. 2 (1987): 119.
4. Thomas Kuhn, “The Function of Measurement in Modern Physical Science,” in The Essential Tension: Selected Essays: Selected Studies in Scientific Tradition and Change (Chicago: University of Chicago Press, 1977), 183.
5. T. S. Eliot, “The Function of Criticism,” in T. S. Eliot: Selected Essays, 1917–1932 (New York: Harcourt, Brace, 1932), 21.
7. Alan Turing, “On Computable Numbers, with an Application to the Entscheidungsproblem,” Proceedings of the London Mathematical Society s2, 42, no. 1 (1937): 230–65.
8. Friedrich Kittler, “Number and Numeral,” Theory, Culture, & Society 23, no. 7–8 (2006): 55.
9. Stuart Gilbert, James Joyce’s “Ulysses”: A Study (New York: Vintage, 1955), 3, 8–9.
10. I’ve chosen to refer to the eighteen episodes of Ulysses by number and not Homeric title for the simple reason that these titles did not exist in the original 1922 edition. In fact, the episodes did not contain any numbers at all. But to impose a Homeric title retroactively not only presupposes a specific interpretation but also has the effect of further erasing the presence of numbers that were meaningful to the specific episodes in question.
11. John Kidd, “Editor’s Preface” for an unpublished Dublin edition of Ulysses, 1994.
13. For some of the history of numerology see Hopper, Medieval Number Symbolism; Tobias Dantzig, Number: The Language of Science (New York: Plume, 2007); E. T. Bell, Numerology (Baltimore, MD: Williams & Wilkins, 1933).
14. Dante Alighieri, Vita Nuova, trans. Mark Musa (Bloomington: Indiana University Press, 1973), 62.
15. As Dante outlines it here, that number only becomes legible because he is able to recognize God’s design. Nine is literally embodied in the original event (written in the stars), but it also gives shape to the structure of the work in which the entire experience is then narrated. The twenty-eight poems that appear in between the prose sections are also punctuated by three canzone. The 10 I 4 II 4 III 10 arrangement can also be subdivided into 1–9 I 1–3 II 3–1 III 9–1, each breakdown of four and ten necessitated by the introduction of a new theme or topic. What it all reveals, according to Mark Musa, is not only Dante’s obsession with nine, and its square root but also the prominence of the second canzone placed in the middle, which looks forward to and backward on Beatrice’s death.
16. See John A. Shawcross, “The Balanced Structure of Paradise Lost,” Studies in Philology 62 (1965): 697–718; Alastair Fowler, Spenser and the Numbers of Time (London: Routledge, 1964); Clyde Murley, “The Structure and Proportion of Catullus LXIV,” Transactions of the American Philological Association 68 (1937): 3305–17.
17. R. G. Peterson, “Critical Calculations: Measure and Symmetry in Literature,” Publication of the Modern Languages Association 91, no. 3 (May 1976): 373.
18. Alastair Fowler, Spenser and the Numbers of Time; Fowler, Triumphal Forms: Structural Patterns and Elizabethan Poetry (Cambridge: Cambridge University Press, 1970); A. Kent Hieatt, Short Time’s Endless Monuments: The Symbolism of Numbers in Edmund Spenser’s “Epithalamion” (New York: Columbia University Press, 1960); Maren-Sofie R⊘stvig, Fair Forms: Essays in English Literature from Spenser to Jane Austen (Cambridge: Cambridge University Press, 1975); Gunnar Qvarnström, Poetry and Numbers: On the Structural Use of Symbolic Numbers (Lund, Sweden: Lund C. W. K. Gleerup, 1966); Christopher Butler, Number Symbolism (London: Routledge & Kegan Paul, 1970).
19. Peterson, “Critical Calculations,” 374.
20. Franco Moretti, Canon/Archive (New York: N+1 Foundation, 2017), 98.
21. James Joyce, Ulysses, prepared by Hans Walter Gabler (New York: Random House, 1986), 224, 621. Hereafter referred to as U. Other editions will be identified when necessary; my emphasis.
22. Fowler, Triumphal Forms, 202.
23. György Lukács, Theory of the Novel (London: Merlin, 1971), 88; Ernst Robert Curtius, European Literature and the Latin Middle Ages (Princeton, NJ: Princeton University Press, 504; Kidd, “Editor’s Preface,” 50.
24. Pound, How to Read, 27.
25. To prove his point, Pound directs his readers to the alleged “40 books” that belonged to Chaucer and the “half dozen” or so of Shakespeare.
26. As conveyed to me by Moretti in an email. Virginia Woolf, “How It Strikes a Contemporary,” Common Reader (London: Hogarth Press, 1948), 297.
27. Nan Z. Da includes a variety of computational approaches under this label including cultural analytics, literary data mining, quantitative formalism, literary text mining, computational textual analysis, computational criticism, algorithmic literary studies, social computing for literary studies, and computational literary studies. See “The Computational Case against Computational Literary Studies,” Critical Inquiry 45, no. 3 (Spring 2019): 601.
28. Richard So and Hoyt Long, “Literary Pattern Recognition: Modernism between Close Reading and Machine Learning,” Critical Inquiry 42, no. 2 (Winter 2016): 235–67; On Paragraphs: Scale, Theme and Narrative Forms, pamphlet 10 (Stanford, CA: Stanford Literary Lab, October 2015); Ted Underwood, David Bamman, and Sabrina Lee, “The Transformation of Gender in English-Language Fiction,” Journal of Cultural Analytics (February 2018) https://culturalanalytics.org/article/11035-the-transformation-of-gender-in-english-language-fiction; Andrew Piper, “Characterization,” in Enumerations: Data and Literary Study (Chicago: University of Chicago Press, 2018), 118–46.
29. On this point, then, I am not in agreement with Underwood’s claim that “even when an argument relies on a single data set, quantitative researchers implicitly understand its claims as part of a larger conversation, organized by a goal of replicability that entails comparisons between different kinds of samples.” Underwood, Distant Horizons, 176.
30. And this emphasis was further encouraged by the publication of Matt Jockers, Macroanalysis: Digital Methods and Literary History (Champaign: University of Illinois Press, 2013).
31. Ted Underwood, for instance, mentions that his own investigations of the novel do not move beyond 1800 because the data is harder to gather. Taken another way, you could say that his inquiry is restricted by the availability of the resources. This restriction shapes the scope of the investigation.
33. For a representative sample, see Jacob Leed, The Computer and Literary Style: Introductory Essays and Studies (Kent, OH: Kent State University Press, 1966).
34. Leo Spitzer, “Linguistics and Literary History,” in Representative Essays, ed. Alban K. Forcione, Herbert Lindenberger, and Madeleine Sutherland (Stanford, CA: Stanford University Press, 1988), 22.
35. “The object of study in literary science,” Roman Jakobson writes, “is not literature but literariness, that is, what makes a given work a literary work.” Quoted in Boris Eichenbaum, “The Theory of the ‘Formal Method,’” in Russian Formalist Criticism: Four Essays, trans. Lee T. Lemon and Marion J. Reis (Lincoln: University of Nebraska Press, 1965), 107.
36. Franco Moretti and Oleg Sobchuk, “Hidden in Plain Sight: Data Visualization in the Humanities,” New Left Review 118 (July–August 2019): 86.
37. P. Shannon et al., “Cytoscape: A Software Environment for Integrated Models of Biomolecular Interaction Networks,” Genome Research 13, no. 1 (November 2003): 2498–504.
38. R. P. Blackmur, “A Critic’s Job of Work,” in The Selected Essays of R. P. Blackmur, ed. Denis Donoghue (New York: Ecco Press, 1986), 45.
39. More than simply providing the factual information about where and when a particular title appeared, the paratext, he explains, also establishes how it can be read. Title, copyright, and limitation pages, Genette explains, are “the conveyor of a commentary that is authorial or more or less legitimated by the author, constitut[ing] a zone between text and off text, a zone not only of transition but also of transaction: a privileged place of a pragmatics and a strategy, of an influence on the public, an influence that…is at the service of a better reception for the text and a more pertinent reading of it.” See Gérard Genette, Paratexts: Thresholds of Interpretation, trans. Jane E. Lewin (Cambridge: Cambridge University Press, 1997), 1–2.
40. Kidd, “Editor’s Preface,” 52–53.
42. These “hidden line counts” belong to Kidd, “Editor’s Preface,” 51, 52.
43. Kenner, Ulysses (Baltimore, MD: Johns Hopkins University Press, 1987), 156.
44. Unsigned Review, Evening News (London), April 8, 1922; Arnold Bennett, Outlook (London), April 29, 1922; quoted in Deming, James Joyce, vol. 1, 194, 220.
45. Edmund Wilson, “Ulysses,” New Republic, July 5, 1922. See also Ezra Pound (who gets the page numbers right) “James Joyce et Pécuchet,” Mercure de France 106 (June 1, 1922): 307–20.
46. Hugh Kenner, Flaubert, Joyce, and Beckett: The Stoic Comedians (London: Dalkey Archive Press, 1962), 34.
47. Maurice Darantière to Sylvia Beach, October 17, 1921. The Darantière letters to Beach are in the James Joyce Collection at the University at Buffalo.
48. It is likely that this calculation was intended for Harriet Shaw Weaver. On October 7, 1921 Joyce wrote: “I expect to have early next week about 240 pages of the book as it will appear ready and will send on [i.e. through the end of ‘Cyclops’].” L I, 172.
49. Maurice Darantière to Sylvia Beach, April 18, 1921.
50. Maurice Darantière to Sylvia Beach, September 30, 1921.
51. An additional 9.806,25 francs were added for corrections (fr. 4,75 per hour). Compare that with the original estimate of 10.349, 64 francs for the paper required for one thousand copies under six hundred pages in 10-point Elzevir. The final paper cost with 732 pages and 11-point Elzevir was 14.060,85 francs.
52. Thanks to Ronan Crowley for providing clarification here (and elsewhere).
53. Ryder, “Editing Ulysses Typographically,” 111.
54. Wim Van Mierlo, “Reflections on Textual Editing in the Time of the History of the Book,” Journal of the European Society for Textual Scholarship 11 (2013): 144–45.
55. On March 16, once the copies of the first edition were all printed and shipped, Darantière wrote to Beach: “Il faut s’en tenir au nombre de cette première édition ou bien couler le volume en entier. Un tirage sur clichés vous permettrait d’augmenter l’édition de 100 ou 200 exemplaires à votre convenance.”
56. “The book grew by one third in proof”: Richard Ellmann, James Joyce, rev. ed. (Oxford: Oxford University Press, 1982), 513; “Joyce had told me that he had written a third of Ulysses on the proofs”: Beach, Shakespeare and Company, 58; Stacy Herbert, “Composition and Publishing History,” in Joyce in Context, ed. John McCourt (Cambridge: Cambridge University Pres, 2009), 10; “In all, Ulysses grew approximately one-third longer from additions Joyce made on the proofs”: Luca Crispi, “Manuscript Timeline, 1905–1922,” Genetic Joyce Studies 4 (Spring 2004) https://www.geneticjoycestudies.org; “Joyce enlarged Ulysses by approximately one-third through additions to the proofs”: Derek Attridge, James Joyce’s Ulysses: A Casebook (Oxford: Oxford University Press, 2004), 9; “For Ulysses’ 732 pages forty-six gatherings were required: somewhat more than the thirty-seven figured in the printer’s castoff—i.e. in the initial typographic estimate to work out the space copy would require. Such an underestimation (155 pages short) goes some way to substantiating the familiar claim that Ulysses ‘grew by one third in proof’ ”: Luca Crispi and Ronan Crowley, “Proof^finder: Page Proofs,” Genetic Joyce Studies 8 (Spring 2008) www.geneticjoycestudies.org.
57. Thanks to Hans Walter Gabler for saving me countless hours by running his typed version of the Rosenbach manuscript against the 1984 edition. Here’s his explanation (July 21, 2019): “Roughly, it [Rosenbach Manuscript] runs to 220,000 words. The word count for the 1984 edition, according to the Handlist, is approx. 264,500 words. The difference then is approx. 44.500 words by which the text of the Rosenbach was eked out. Which—I admit—approximates 20 percent augmentation, not 33 percent. There goes another almosting.”
58. A note handwritten by Beach on the back of a letter typed by Darantière dated August 31, 1921, indicates that he did know what would constitute one-third of his book: “15th June Joyce gave him 1 3rd of book which he has now printed (in 3 months) with a few pages of the second part.”
60. From a letter to Weaver, it is clear that Joyce was aware of the relative value of pages, L I, 196: “It [Ulysses] is an extremely cheap (considering paper and type) edition of the equivalent of 8 English novels of standard size (75,000 to 80,000 words) at the normal selling price today or less, as 8 10/- novels would sell at 270 francs a copy.”
61. In James Joyce’s Ulysses: A Study, 31, Gilbert explains that the symmetry is contained within a prelude of three episodes, a central section of twelve, and a finale of three.
62. I am grateful to Aaron Matz for leading me here.
63. Hopper: “The great astrological numbers being 4, 12, 7, the age of the world was known to be divided into 4 periods, into 12 parts, or into 7,000 days, 70 generations or 7 weeks.” Hopper, Medieval Number Symbolism, 30–31.
64. Kidd: “It remained to be discovered decades later that the book is arranged and proportioned in accordance with a concealed numeric pattern. Ulysses, meticulously adjusted during months of revision in proof, is as consummately ordered as Dante’s Divine Comedy or Spenser’s Faerie Queene. In the editions since the printings of 1922–1925, the geometric form of Ulysses has been obscured beyond recognition.” Kidd, “Editor’s Preface,” 49–50.
65. Kidd’s theory as laid out by D. F. McKenzie, Bibliography and the Sociology of Texts (Cambridge: Cambridge University Press, 1999), 58.
66. When would it have been likely for Joyce to know the middle? And how likely is it that this would have been possible given the chaotic nature of the revision in the fall of 1921? Here’s Joyce to McAlmon: “I have now written in a great lot of balderdash all over the damn book and the first half is practically as it will appear.” September 3, 1921, L III, 48.
67. John MacQueen, Numerology: Theory and Outline History of a Literary Mode (Edinburgh: Edinburgh University Press, 1985), 4; emphasis added.
68. Robert Spoo, “Copyright Protectionism and Its Discontents: James Joyce’s Ulysses in America,” The Yale Journal 108, no. 3: 633-667.
69. James Joyce, Ulysses: A Digital and Critical Synoptic Edition, ed. Hans Walter Gabler, Ronan Crowley, and Joshua Schäuble (New York: Garland, 1984), http://ulysses.online/index.html. For more background on the process and its implications, see Hans Walter Gabler, “Seeing James Joyce’s Ulysses into the Digital Age,” Joyce Studies Annual (2018): 16–18.
70. Kidd’s Dublin edition was going to retain the original pagination and line counts in order to restore the novel’s “geometric form.” He writes: “This is the only edition to account for number symbolism more often associated with classical, medieval, and renaissance artists than with modernists.” Kidd, “Editor’s Preface,” 50.
72. For an in-depth account of the steps involved in this process, see Hans Walter Gabler “Towards an Electronic Edition of James Joyce’s Ulysses,” Literary and Linguistic Computing 15, no. 1 (2000): 115–20.
73. Hugh Kenner, Joyce’s Voices (Oakland: University of California Press, 198), 166.
74. Hugh Kenner, “Reflections on the Gabler Era,” James Joyce Quarterly 26, no. 1 (Fall 1988): 18, 19.
75. Kenner, “Reflections,” 19.
76. Hanley never includes the word counts, but by Zipf’s estimation, there are 28,899 unique words in 260,430. Compare with Gabler’s version of 39,801 and 264,448.
77. Charles Kingsley Zipf, Human Behavior and the Principle of Least Effort: An Introduction to Human Ecology (Cambridge, MA: Addison-Wesley, 1949), 546n3.
78. Zipf, Human Behavior, 24, 35.
79. See Zipf, Human Behavior, 40–47.
80. Zipf, Human Behavior, 51.
81. I’m using the term law here realizing that it is neither a law, because of the deviations in most languages and its nonuniversality, nor entirely Zipf’s, because it was already discovered by Jean-Baptiste Estoup in 1912. See Erez Aiden and Jean-Baptiste Michel, Uncharted: Big Data as a Lens on Human Culture (New York: Riverhead, 2014), 249n34.
82. René Wellek, A History of Modern Criticism, vol. 5 (New Haven, CT: Yale University Press, 1986), xix.
83. Pound uses “dateline” in an essay by the same name, but I take it here from A. Walton Litz in the opening line to The Art of James Joyce: Method and Design in “Ulysses” and “Finnegans Wake” (Oxford: Oxford University Press, 1961), 1.
1. MAKING STYLE COUNT
1. See, for instance, James Shapiro, Contested Will: Who Wrote Shakespeare? (New York: Simon & Schuster, 2010).
3. As reported to me by Kevin Dettmar, who heard it firsthand from an unnamed woman waiting in line at the British Library.
4. Frederick Mosteller and David L. Wallace, “Inference in an Authorship Problem,” Journal of the American Statistical Association 58, no. 302 (June 1963): 275–309.
5. A. Q. Morton, “Who Wrote St. Paul’s Epistles,” Listener (March 21, 1963): 489ff. See also Morton, “The Authorship of Greek Prose,” Journal of the Royal Statistical Society, series A, 128, no. 2 (1965): 169–224.
6. A. Q. Morton and J. McLeman, Christianity in the Computer Age (New York: Harper & Row, 1964).
7. McCandlish Philips, “Computer Flouts Test by Another: Study on St. Paul’s Epistles Questioned at Yale Parley,” New York Times, January 23, 1965, 17. The talk in which Ellison made these comments appeared in an edited collection from the conference proceedings. See “Computers and the Testaments,” in Computers for the Humanities: A Record of the Conference Sponsored by Yale University on a Grant from IBM, January 22–23, 1965 (New Haven, CT: Yale University Press, 1965), 64–74.
8. See, for instance, G. Udney Yule, “One Sentence-Length as a Statistical Characteristic of Style in Prose: With Application to Two Cases of Disputed Authorship,” Biometrika 30, nos. 3/4 (January 1939): 363–90; C. B. Williams, “A Note on the Statistical Analysis of Sentence-Length as a Criterion of Literary Style,” Biometrika 31, nos. 3/4 (March 1940): 356–61; Franco Moretti, Sarah Allison, Marissa Gemma, et al., Style at the Scale of the Sentence, pamphlet 5 (Stanford, CA: Stanford Literary Lab, June 2013), https://litlab.stanford.edu/LiteraryLabPamphlet5.pdf.
9. Here’s how they describe it in their results: “Multifractals are not simply the sum of fractals and cannot be divided to return back to their original components, because the way they weave is fractal in nature. The result is that in order to see a structure similar to the original, different portions of a multifractal need to expand at different rates. A multifractal is therefore non-linear in nature.”
11. T. S. Eliot, “Lettre d’Angleterre: Le Style dans la prose anglaise contemporaine,” La Nouvelle Revue Française 19 (July–December, 1922): 751–56. The landmark book on the subject is Karen Lawrence, The Odyssey of Style in “Ulysses” (Princeton, NJ: Princeton University Press, 1981), with the focus on the narrative discontinuities that modify readerly expectations over the eighteen episodes.
12. T. C. Mendenhall, “A Mechanical Solution of a Literary Problem,” Popular Science Monthly 60 (1901): 97–105.
13. Throughout this chapter, I am using the paragraph counts from the xml files that were prepared for the Ulysses: A Digital and Critical Synoptic Edition. These files are based on the reading text of the novel that was generated by the TUSTEP program for Gabler’s original 1984/1986 edition. The online edition, DCSE, is being overseen by Ronan Crowley and Joshua Schäuble. There are some discrepancies in the paragraph breaks between the Little Review episodes and the 1922 edition that Gabler corrected, consulting typescript and other drafts when possible. Those interested in these differences are encouraged to consult Philip Gaskell and Clive Hart, “Ulysses”: Review of Three Texts (Gerrards Cross, UK: Colin Smythe, 1989).
14. Gertrude Stein, “Lectures in America,” in Gertrude Stein: Writings, 1932–1946, ed. Catharine R. Stimpson and Harriet Chessman (New York: Library of America, 1998), 218.
15. Franco Moretti, Ryan Heuser, Mark Algee-Hewitt, “On Paragraphs: Scale, Themes, and Narrative Form,” pamphlet 10 (Stanford, CA: Stanford Literary Lab, October 2015), 69.
16. Fredric Jameson, “Joyce or Proust?” in The Modernist Papers (New York: Verso, 2007), 181.
17. Jean-Michel Rabaté and Daniel Ferrer, “Paragraphs in Expansion (James Joyce),” in Genetic Criticism: Texts and Avant-Textes, ed. Jed Deppman, Daniel Ferrer, and Michael Groden (Philadelphia: University of Pennsylvania Press, 2004), 132–52.
18. Hugh Kenner, “The Most Beautiful Book,” English Literary History 48, no. 3 (1981): 604.
19. Michael Groden, Ulysses in Progress (Princeton, NJ: Princeton University Press, 1977), 160.
20. Groden, Ulysses in Progress 160.
23. For one of the few books of criticism devoted entirely to the subject, see Jean Châtillon, ed., La Notion de paragraphe (Paris: Éditions du Centre national de la recherche scientifique, 1985).
24. Jason Pamental, “Life of ¶” Print 69, no. 4 (2015): 24–26.
25. Robert Bringhurst, The Elements of Typographic Style (Vancouver: Hartley & Marks, 1992), 80.
26. Henri Mitterand, “Le paragraphe est-il une unité linguistique?” in Châtillon, La Notion de paragraphe, 85.
27. Thanks to Jonathan Reeve for giving me a record of the conversation that took place between the various members working with Gabler on the digital edition.
28. Kidd, “Dublin Edition,” 50.
29. Kenner, “Reflections on the Gabler Era,” 17.
30. Joyce, Finnegans Wake, 123. I say “came to think” here because we know from a letter to Budgen that Joyce had in mind sentences as he was writing episode 18: “There are 8 sentences in the episode.”
31. Moretti et al., “On Paragraphs,” 5.
32. See, for instance, E. H. Lewis, The History of the English Paragraph (Chicago, IL: University of Chicago Press, 1894), 173.
33. Lewis, History of the English Paragraph, 156–57.
35. On Paragraphs, 93. That wasn’t exactly the plan when they started. The Literary Lab was originally hoping that the unit of the paragraph could provide a way into thinking about style as they discovered earlier in the discrete elements of the sentence. It did not. But that, they came to realize, was because paragraphs are not sentences, and we should not assume just because they are made up of sentences that they would function according to the same rules. See Moretti et al., “Style at the Scale of the Sentence,” pamphlet 5 (Stanford, CA: Stanford Literary Lab, 2013), 58–59.
36. Adrienne Monnier, The Very Rich Hours of Adrienne Monnier, trans. Richard McDougall (Lincoln: University of Nebraska Press, 1996), 115.
37. S. L. Goldberg, Joyce: The Man, The Work, The Reputation (Edinburgh: Oliver & Boyd, 1962), 90.
39. Theodor Adorno, “Punctuation Marks,” in Notes to Literature, vol. 1, trans. Sherry Weber Nicholsen (New York: Columbia University Press, 1991), 91.
40. Wolfgang Iser, The Act of Reading: A Theory of Aesthetic Response (Baltimore, MD: Johns University Press, 1978), 113.
41. For a fascinating account of the subject, see Rudolf Arnheim, “The Images of Thought,” in Visual Thinking (Berkeley: University of California Press, 1969), 97–115.
42. Stéphane Mallarmé, Divagations, trans. Barbara Johnson (Cambridge, MA: Harvard University Press), 298. French reads: “Mobiliser, autour d’une idée, les lueurs diverses de l’esprit, à distance voulue, par phrases: ou comme, vraiment, ces moules de la syntaxe même élargie, un très petit nombre les résume, chaque phrase, à se détacher en paragraphe gagne d’isoler un type rare avec plus de liberté qu’en le charroi par un courant de volubilité. Mille exigences, très singulières, apparaissent à l’usage, dans ce traitement de l’écrit, que je perçois peu à peu: sans doute y a-t-il moyen, là, pour un poète qui par habitude ne pratique pas le vers libre, de montrer, en l’aspect de morceaux compréhensifs et brefs, par la suite, avec expériences, tels rythmes immédiats de pensée ordonnant une prosodie.”
43. Mallarmé may have singled out the press for this discovery of prose distillation and layout, but it’s also worth noting that Joyce actually used the abbreviated term par as journalist slang for paragraph. And when the word appears in Ulysses, Bloom uses the unabbreviated version but in reference to an advertisement he hopes to place in the newspaper (see epigraph). “Get a par,” he once ordered his brother when trying to publicize a business venture in Dublin back to his adopted home in Trieste. L II, 277.
44. The recent discovery of the earliest known draft for this episode would seem to support this decision: it includes fifteen different sections (beginning on page 5 of the notebook) separated by a line of asterisks, with dialogue dashes extending outside of the paragraphs, thereby keeping narration and dialogue on the same level. A digital copy is publicly available from the National Library of Ireland. See http://catalogue.nli.ie/Record/vtls000357767. Critics are already in the habit of referring to these different sections as fragments, but that obfuscates what is otherwise an important point: Stephen is already thinking in complete paragraphs. See, for instance, Sam Slote, “Epiphanic ‘Proteus,’” Genetic Joyce Studies 5 (Spring 2005); Luca Crispi, “A First Foray into the National Library of Ireland’s Joyce Manuscripts: Bloomsday 2011” Genetic Joyce Studies 11 (Spring 2011). Daniel Ferrer wonders if Joyce considered using a “fragmentary form” for this episode, something more in line with episode 10. Whatever happened along the way, it wasn’t fragments that Joyce collected for episode 3, it was paragraphs, then arranged in a sequence, which would allow him to weave together the movement of thoughts in Stephen’s head with the movement of his body along the beach. Daniel Ferrer, “What Song the Sirens Sang…Is No Longer beyond All Conjecture: A Preliminary Description of the New ‘Proteus’ and ‘Sirens’ Manuscripts,” James Joyce Quarterly 39, no. 1 (Fall 2001): 53–67.
45. Gerhard Dohrn-van Rossum, History of the Hour (Chicago: University of Chicago Press, 1996).
46. At 376 paragraphs in 11,439 words, episode 11 would seem an outlier but only if we don’t factor in the fifty-nine bits of text, each one its own paragraph, contained in the overture.
47. Kenner, “Gabler Era,” 17.
48. A. M. Klein, “The Oxen of the Sun,” Here and Now 1 (1949): 28–48.
50. In this table, I am using the breakdown provided by Jeri Johnson in her edition of the 1922 Ulysses (Oxford: Oxford University Press, 2008), the one that she checks against the information already provided by Robert Janusko, Sources and Structure of James Joyce’s Oxen (Ann Arbor, MI: UMI Research Press, 1983); and Stuart Gilbert’s “Ulysses”: A Study.
51. Saintsbury, History of English Prose Rhythm, 53.
2. WORDS IN PROGRESS
1. “Graphic unit” is the term preferred by Gabler and his team because it emphasizes the fact that words, though made up of strings of letters, can also be numbers, initials, symbols, sounds, etc.
2. Vladimir Nabokov, “Ulysses,” in Lectures on Literature, ed. Fredson Bowers (New York: Harcourt, Brace, Jovanovich, 1982), 285.
3. Only three issues of the Little Review appeared without any installments from Ulysses between March 1918 and December 1920. Because of the length, it is possible that episode 14 could have fit into three different installments, but based on the two or three long episodes that preceded it, I suspect that four would have been more likely.
4. See The Little Review “Ulysses,” ed. Mark Gaipa, Sean Latham, and Robert Scholes (New Haven, CT: Yale University Press, 2014).
5. Pound wrote one particularly animated letter complaining about the typos in the December 1918 number of the Little Review: “What the ensanguined lllllllllllllllllllllll is the matter with this BLOODY goddamndamnblastedbastardbitchbornsonofaputridseahorse of foetid and stinkerous printer???????????? Is his serbo-croatian optic utterly impervious to the twelfth letter of the alphabet ?????”; The Letters of Ezra Pound to Margaret Anderson: “The Little Review” Correspondence, ed. Thomas Scott, Melvin Friedman, and Jackson Bryer (New York: New Directions, 1988), 181.
6. For a discussion of these errors, see Amanda Sigler, “Archival Errors: Ulysses and the Little Review,” in Errears and Erroriboose: Joyce and Error, ed. Matthew Creasy, European Joyce Studies 20 (Amsterdam: Rodopi Press, 2011), 73–88.
7. The most comprehensive history of the episode can be found in Matthew Garrett’s Episodic Poetics: Politics and Literary Form after the Constitution (Oxford: Oxford University Press, 2014).
8. Forrest Read, ed., Pound/Joyce: The Letters of Ezra Pound to James Joyce (New York: New Directions, 1970), 143, 158.
10. Gabler’s theory of the lost manuscript is based on the careful comparison of extant manuscripts. The typescript for some episodes, then, may correspond with the Rosenbach manuscript, but there are too many differences to suggest that another source was being consulted. This point was clarified to me by Gabler via email after he read a version of this chapter when published in Representations.
11. Lost in the wood, I finally contacted Michael Groden, who, after patiently considering the possibilities, wrote to me in an email: “I’d stay with words.”
12. Anthony Burgess, not entirely thrilled with the whole idea of counting words, joked that “innumerable Ph.D. candidates yet unborn have their thesis subjects waiting for them—an Old Norse word-count of Finnegans Wake, II. iii.” See ReJoyce (New York: Norton, 1965), xi. He’s more than likely referring to Frank Budgen’s word-count free “James Joyce’s Work in Progress and Old Norse Poetry,” in Our Exagmination Round His Factification for Incamination of Work in Progress (Paris: Shakespeare & Co., 1929), 37–46. As far as I know, the subject is still wide open for any enterprising PhD candidates with an interest in the subject.
13. Hugh Kenner, “Reflections on the Gabler Era,” James Joyce Quarterly 26, no. 1 (Fall 1988): 14.
14. Hugh Kenner, Ulysses (London: George, Allen, & Unwin, 1980), 67; emphasis added.
15. Kenner, “Reflections,” 19.
16. For more on word counts, see Duncan Harkin, “The History of Word Counts,” Babel 3, no. 3 (1957): 113–24 and James E. DeRocher, The Counting of Words: A Review of the History, Techniques, and Theory of Word Counts with Annotated Bibliography (New York: Syracuse University Research Corp, 1973).
17. Miles Hanley, Word Index to James Joyce’s “Ulysses” (Madison: University of Wisconsin Press, 1937), 384. Another early application of the word count can be found in Erwin Steinberg’s The Stream of Consciousness and beyond in “Ulysses” (Pittsburgh, PA: University of Pittsburgh Press, 1973). See also Leslie Hancock, Word Index to James Joyce’s “Portrait of the Artist” (Carbondale: Southern Illinois University Press, 1967); Wilhelm Füger, Concordance to James Joyce’s “Dubliners” (Hildesheim: G. Olds, 1980); Clive Hart, A Concordance to “Finnegans Wake” (Minneapolis: University of Minnesota Press, 1963); and Ruth Bauerle, Word List to James Joyce’s “Exiles” (New York: Garland, 1981).
18. Wolfhard Steppe and Hans Walter Gabler, A Handlist to James Joyce’s Ulysses: A Complete Alphabetical Index to the Critical Reading Text (New York: Garland Publishing, 1986), vii.
19. Jacques Derrida, “Ulysses Gramophone: Hear Say Yes in Joyce,” in Acts of Literature, ed. Derek Attridge (New York: Routledge, 1992), 305–6.
20. Though it does include by breakdown: Yes (195), Yes (2), yes (161), and yes (1). In the Word-Index, Hanley lumps them all together under a single entry for “yes” and gets 354.
21. Pound to Joyce, March 1917, in Read, Pound/Joyce, 104. In December 1918, while still serving as foreign editor of the Little Review, Pound had the idea of starting up a quarterly, writing to Joyce that he wouldn’t be able to “serialize novels,” but he’d be willing to accept contributions of “up to let us say 10,000 words of story or essay. 10,000 not an absolute limit, but a general estimate”; Pound to Joyce, December 1918, in Read, Pound/Joyce, 104, 148.
22. Joyce to Pound, August 1917, SL, 227.
23. In Serial Encounters: “Ulysses” and the “Little Review” (Oxford: Oxford University Press, 2019), Clare Hutton argues that these negotiations were “vague,” which I assume means nonbinding and less than urgent (144). She rests her claim on two letters (one from June 1919, another from January 1920), but the first one was written after ten episodes had already appeared and the second when the thirteenth was drafted. It’s difficult to reconcile any broader claims about lack of urgency when more than half the novel was written.
24. On 27 November 1917, Joyce writes to Claud Sykes about episode 2 and episode 3: “In the other two episodes (which are not so long) it will depend on whether he [Rudolph Goldschmidt, the one they’re borrowing typewriter from] disposes of his place or not. I take it you could do what I sent quickly. There is no use losing time.” On December 12, 1917, Joyce reminds Sykes: “This episode [2] is only half as long as the other”; L I, 109; L II, 412.
25. In “The Development of Ulysses in Print, 1918–22,” Hutton also generates a “rough snapshot” of word counts (using Microsoft Word) but makes the erroneous claim that “the algorithm of what counts as a word and what counts as a space would be different in another programme.” Why? Computers count the spaces between and not the words as individual units. Dublin James Joyce Journal 6/7 (2013–2014): 121.
26. Pound was confused when he received the first seventeen pages of typescript, thinking it was intended for the first three installments instead of one. “Your 17 pages have no division marks,” he wrote Joyce on December 19, 1917. “Unless this is the first month’s lot, instead of the first three months, as you wrote it would be, I shall divide on page 6. After the statement that the Sassenach wants his bacon. And at the very top of page 12.” Read, Pound/Joyce, 129. An unpublished letter reveals that Joyce was, in fact, to blame because of he repeatedly elided episode and chapter: “The first chapter of Ulysses is being typed in Zurich and will send as soon as possible. There are three episodes in it. To save time I shall send on each as it is finished. This chapter should run for three months—the Telemachia. There are in all three chapters in the book, the other two being Odyssey (in twelve episodes) and the Nostos (in three episodes). The chapter of the wooers falls out. Of course, these names are not to be prefixed to the chapters”; Ezra Pound Papers, YCAL MSS 43, Box 26, folder 1112, Beinecke Rare Book and Manuscript Library, Yale University. There’s no date on the letter but it must have been written some time in November or December 1917 when Joyce was staying at the Pension Daheim in Locarno.
27. For a fascinating discussion about line length in all three canticles of the Commedia, see Joan Ferrante, “A Poetics of Chaos and Harmony,” in The Cambridge Companion to Dante, ed. Rachel Jacoff (Cambridge: Cambridge University Press, 1993), 181–200. For an early account of the length of Shakespeare’s plays, see A. Hart, “The Length of Elizabethan and Jacobean Plays,” Review of English Studies 8, no. 30 (April 1932): 139–54. For details regarding installment size in Charles Dickens, see Robert L. Patten, “The Composition, Publication, and Reception of Our Mutual Friend,” UC Santa Cruz, Our Mutual Friend: The Scholarly Pages (website), March 20, 2012, http://omf.ucsc.edu/publication/comp-and-pub.html. For a more general discussion of word count as a tool for literary analysis, see Mendenhall, “A Mechanical Solution,” 97–105.
28. Budgen: “As a work of reference for his Ulysses he used the Butcher-Lang translation of the Odyssey.” “James Joyce and the Making of Ulysses,” 323. I am very grateful to Jim Shapiro for suggesting that I go back to Homeric translations Joyce consulted for this information.
29. Stanislaus Joyce’s comment was relayed by W. B. Stanford, The Ulysses Theme (Oxford: Blackwell, 1954), 76. Kenner agreed with this assessment in 1969 but recanted a decade later on the grounds that Stanislaus would have had access to Joyce only before Zurich, since he was interned during the war, therefore making it possible that his brother would have consulted other editions, the Butler translation included. See Hugh Kenner, “Homer’s Sticks and Stones,” James Joyce Quarterly 6 (1969): 285–98 and Joyce’s Voices (Berkeley, CA: University of California Press, 1978), 110–11.
30. Henry Fielding, Joseph Andrews, in The Works of Henry Fielding: With a Life of the Author (London: Otridge & Rackham, et al., 1824), 5:91. “Numbers” might have also signified metrical verse to Fielding’s contemporaries.
31. An interesting point of comparison: the average word count per line in Elizabethan blank verse was eight. See Hart, “The Length of Elizabethan and Jacobean Plays,” 141.
32. Franco Moretti: “Adventures expand novels by opening them to the world: a call for help comes—the knight goes. Usually, without asking questions; which is typical of adventure, the unknown is not a threat here, it’s an opportunity, or more precisely: there is no longer any distinction between threats and opportunities.” See “The Novel: History and Theory,” New Left Review 52 (July–August 2008): 5.
33. There’s another possibility: I also suspect that Joyce may have sent along his typescript to Pound, waiting for the arrival of the issue of the Little Review before moving ahead. Since his issues, which Pound passed on to him from London, were always delayed, it is more than likely that he would have been receiving them up to half a year after he sent the typescript to Pound, sometimes more when the publication of an issue was delayed. “As ‘them females,’” he wrote to Joyce on November 24, 1919, “dont [sic] send me the little Review I cant [sic] forward it to you I will ask Rodker to do so, and suggest that a note from you to N.Y. office might help”; Read, Pound/Joyce, 162.
34. Episode 8 could easily have fit into a single issue. It’s unclear who decided where it should be divided. Pound and Jane Heap, once again, are the most likely candidates.
35. He never mentions this in his letters to Margaret Anderson, and there are no distinguishable marks on the batch of Little Review typescript pages that survive: seventeen and a half of them from episode 14 that begin with page 10. A copy of the first ten typescript pages can be found in the copy used by Maurice Darantière when preparing the proofs for Ulysses and reveal that the first serial installment of episode 14 was made from seven pages (with about three lines continuing on to page 8), making it possible to generate three more at roughly the same length. See The Letters of Ezra Pound to Margaret Anderson; James Joyce Archive (New York: Garland, 1977), 14:141–59, 171–98.
36. On August 6, 1919, he explained his method to an increasingly frustrated Harriet Shaw Weaver: “I understand that you may begin to regard the various styles of the episodes with dismay and prefer the initial style much as the wanderer did who longed for the rock of Episode 17. But in the compass of one day to compress all these wanderings and clothe them in the form of this day is for me possible only by such variation which, I beg you to believe, is not capricious.” Joyce, SL, 242.
37. Groden, Ulysses in Progress, 33.
38. Wyndham Lewis, “The Revolutionary Simpleton,” in The Enemy: A Review of Art and Literature (January 1927). Reprinted in Wyndham Lewis, Time and Western Man, ed. Paul Edwards (Santa Rosa, CA: Black Sparrow Press, 1993), 27–30. Referring specifically to a single sentence from episode 8, Kenner was in agreement. “‘The year marked on a dusty bottle,’” he argues, “seems too many words for Bloom’s swift thought; seven words have evidently been supplied for what passed before Bloom as a visual fancy.” Seven words for a single impression may have been too long for Kenner, but he was quick to remind everyone that Bloom is “a creation of words, which sometimes overlay and amend ‘his’ words and are also sometimes shut off, as when he came through his front door and later could not remember what the words did not record, what he had done with his hat.” Kenner, Ulysses, 70.
39. Gilbert, James Joyce’s “Ulysses,” 26.
41. Groden, Ulysses in Progress, 115.
42. Paul Ricoeur, Time and Narrative, vol. 2, trans. Kathleen McLaughlin and David Pellauer (Chicago: University of Chicago Press, 1984), 78.
43. A quantitative analysis of sentence length in Ulysses still remains to be done.
44. For a more detailed discussion of this middle stage, see Groden, Ulysses in Progress, 37–52.
45. Groden, Ulysses in Progress, 49.
46. Pound to Joyce, June 10, 1919, in Read, Pound/Joyce, 158.
47. For some reason or other, Joyce informed Budgen on January 3, 1920 that episode 13 was “not so long as the others.” Because he uses the plural here, “others,” I’m assuming he had episode 11 and episode 10 in mind as well. He was well off the mark: episode 13 is longer than both of them by more than 2,000 words; Joyce, SL, 246.
48. These word counts vary slightly from other published counts due to the ambiguities in the language and typesetting. In every case, though, the differences are very minor, between two and twenty words, with only one of them reaching seventy. These small differences do not affect the ratios I am presenting here.
49. Aristotle’s Poetics: The Argument, trans. Gerald F. Else (Cambridge: Cambridge University Press, 1957), 286.
50. In the revised 1986 Handlist, the number of graphic units, which was generated automatically from the TUSTEP typesetting program, is 264,448. In the word counts for the 1984 edition personally given to me by Gabler, the total is 266,601. Since I do not have access to the TUSTEP program, I am unable to explain the reason for these differences. That said, they are very slight and do not affect the point I am making here. However, the differences I keep encountering in word counts of Ulysses as a book and as a collection of serialized episodes reveals that a final, exact number remains to be found.
51. Groden, Ulysses in Progress, 169.
52. For these numbers, I compared the serial word counts I compiled using Word Counter with the book word counts provided to me by Gabler.
53. It’s a response the man himself had, making sure that everyone knew that the serial version was rotten to the core. James Pinker, for instance, was informed in December 1919 that “the text hitherto published in Little Review is not my text as sent on in typescript,” and on March 11, 1920 John Quinn was told that the “the version in The Little Review is, of course, mutilated”; L II, 456, 459.
54. Driver points out that the greatest number of additions were reserved for episode 5, episode 8, episode 13, and episode 12; Clive Driver, “Bibliographical Preface,” in Ulysses: A Facsimile of the Manuscript (New York: Octagon, 1975)
55. To Harriet Shaw Weaver on November 10, 1920, Joyce described it as “about twice as long as the longest episode hitherto, Cyclops [episode 12].” To Claud Sykes in early 1921, he explained that it would run “to nearly 200 pages of typescript,” adding that “it is as intricate as it is long.” L I, 149, 158.
56. Joyce to Robert McAlmon: “I shall give Molly another 2000 word spin, correct a few more episodes and write all over them and then begin to put the spectral penultimate Ithaca into shape.” Joyce to Budgen on August 16, 1921: “Penelope is the clou of the book. The first sentence contains 2500 words. There are 8 sentences in the episode.” Joyce to Weaver on August 7, 1921: he had “written the first sentence of Penelope but as this contains about 2500 words the deed is more than it seems to be”; SL, 285; L III, 48; L I, 168. It appears Joyce knew early on that episode 18 would be shorter than episode 15. On November 24, 1920, he wrote to Quinn: “I must have a few month’s leisure after January to write Ithaca and Penelope episodes [17 and 18] which, however, have been sketched since 1916 and are very short in comparison with the Circe episode.” L III, 31.
57. Ezra Pound, Literary Essays (New York: New Directions, 1968), 407.
3. ONE OR HOW MANY?
The epigraph to this chapter comes from Balzac’s preface to his 1842 edition of the Comédie humaine (10). I am grateful to Peter Brooks for reminding me of the number (c. 2400) contained in the six-volume catalogue compiled by Anatole Cerfberr and Jules François Christophe. See Repertory of the Comédie humaine: Complete A-Z (Philadelphia: Avil Publishing Co., 1912).
1. See Groden, Ulysses in Progress, 181; Character list reproduced in Joyce’s “Ulysses” Notesheets in the British Museum, ed. Philip Herring (Charlottesville: University Press of Virginia, 1972), 526–27.
2. And in a long line of characters not on the list is Mina Purefoy’s newborn baby boy from episode 14, Mortimer Edward Purefoy.
3. In Joyce’s “Ulysses” Notesheets, Herring confuses the “1” with an apostrophe, transcribing it as “ ’38” instead of “138” (366).
4. Pedro J. Miranda, Murilo S. Baptista, and Sandro E. de Souza Pinto, “Analysis of Communities in a Mythological Social Network,” ArXiv abs/1306.2537 (2013), n. pag. Curtius, European Literature, 365. Moretti points out that there are over “600 characters in The Scholars, 800 in The Water Margin and the Jin Ping Mei, 975 in The Story of the Stone.” See Franco Moretti, “History of the Novel, Theory of the Novel,” Novel: A Forum on Fiction 43, no. 1 (2010): 5.
5. All of them listed in Anatole Cerfberr and Jules François Christophe, Repertory of the Comédie humaine: Complete A–Z (Philadelphia: Avil, 1912).
6. Balzac, Comédie humaine (1842), 18.
7. Franco Moretti, Atlas of the European Novel, 1800–1900 (London: Verso, 1998), 68.
8. The original line reads: “Most novels are in some sense knowable communities.” See Raymond Williams, “The Knowable Community in George Eliot’s Novels,” Novel: A Forum on Fiction 2, no. 3 (Spring 1969): 255–68.
9. Even influencing the naming of children in ancient Rome (first four sons in a family get names, the rest numerals). See Georges Ifrah, The Universal History of Numbers, trans. David Bellos, E. F. Harding, Sophie Wood, and Ian Monk (London: John Wiley & Sons, 2000), 9.
10. Ifrah, Universal History, 7.
11. Richard Kain, Fabulous Voyager: A Study of James Joyce’s “Ulysses” (New York: Viking, 1959), 252.
12. Shari Benstock and Bernard Benstock, Who’s He When He’s at Home: A James Joyce Directory (Urbana: University of Illinois, 1980).
13. All of them are listed in Benstock and Benstock, Who’s He When He’s at Home, 373.
14. In the Legend, the Benstocks identify the need to include some animals and objects as characters: “Although this directory is limited to ‘human’ characters, the unusual nature of Ulysses dictates that (1) all dogs (like Garryowen) and horses (like Throwaway) must necessarily be included, provided that they have names, (2) that those inanimate objects that have speaking roles in Circe be included—as hallucinations.” Benstock and Benstock, Who’s He When He’s at Home, xii.
15. Alex Woloch, The One vs. the Many: Minor Characters and the Space of the Protagonist in the Novel (Princeton, NJ: Princeton University Press, 2003).
16. In order for these divisions to work at all, the reappearance of characters in episode 15 from previous episodes does not count as a reappearance. This is particularly relevant to a large quantity of characters who would fall in the minor minor category. In addition, by using speaking as part of the criteria for distinguishing types of characters, I am not including those characters who shout or yell unless they are addressing someone in particular. There will be places where arguments can be made for the appearance of characters in more than one category, but this is what I’ve come up with based on these four criteria. The most important distinction involves identifying the different between minor and minor minor for the analysis that follows.
17. Woloch, The One vs. the Many, 122, 117, 124.
18. The data for the network visualizations would not have been possible without the directory the Benstocks compiled. Like so many others, I am grateful to them for this valuable service even though I never expected to need it long ago when I first acquired a copy.
19. Bloom: “ ‘Twentyfour solicitors in that one house” (U, 237). This line was immediately followed by “Counted them” but did not appear in the 1922 edition. It did, however, appear in the Rosenbach manuscript and was reinserted by Gabler.
20. “The importance of the number [12] is made evident by the choice of another apostle to complete the number when one was lost. Or again, Noah, the tenth generation, had 3 sons, but one fell into sin to preserve the number. The tribe of Levi also made thirteen tribes in actuality, but they are always spoken of as 12, thus demonstrating that the idea of the number is more important than the actuality.” Hopper, Medieval Number Symbolism, 86.
21. And there’s also the added irony that the twelve characters may correspond with the twelve disciples, but there’s really no room for the Jew.
22. Joyce put Paddy on his list along with Rudy Bloom and Dodd & Son.
23. Curtius, European Literature, 509; Gunnar Qvarnström, Poetry and Numbers: On the Structural Use of Symbolic Numbers (Lund, Sweden: Gleerup, 1966), 36.
24. “Between the original work on Cyclops in summer 1919 and the pulling of the episode’s first proofs in October 1921, he began to write a new kind of book.” Groden, Ulysses in Progress, 158.
25. Michael Groden, Philip Herring, David Hayman, and Karen Lawrence.
26. Four of them begin with the adverb so, two with the conjunction and, one with a car pulling up, and another with a character popping in.
27. David Hayman, “Cyclops,” in James Joyce’s Ulysses: Critical Essays,” ed. Clive Hart and Hugh Kenner (Berkeley: University of California Press, 1974), 253.
28. Hayman,”Cyclops,” 253. See also Phillip Herring, Joyce’s Notes and Early Drafts for “Ulysses” (Charlottesville: University Press of Virginia, 1977), 146.
29. Groden, Ulysses in Progress, 135.
30. “Scenes and Fragmentary Texts for ‘Cyclops’,” NLI II. ii. 4 ‘Partial Draft: Cyclops”. MS 36, 639/10. There’s a line preceding the lineup that reads: “Michan’s: Bloom + ? [question mark Joyce’s] Woman.” This list is different from the one Groden discusses from later in the process (Ulysses in Progress, 122).
31. See Parliamentary Papers House of Commons and Command, vol. 68 (Dublin: Cahill & Co., 1901), 1.
32. Clive Hart and Ian Gunn, with Harald Beck, A Topographical Guide to the Dublin of Ulysses (London: Thames & Hudson, 2004).
33. In his excellent article on the census, Sam Alexander arrives at the number 108. Using the phrase “by my count,” shows that he is acutely aware of the difficulty of figuring out who counts but also how to count (several figures together are one?). See “Joyce’s Census: Character, Demography, and the Problem of Population in Ulysses,” Novel: A Forum on Fiction 45, no. 3 (2012), 448.
34. See Leo Knuth, “A Bathymetric Reading of Joyce’s Ulysses, Chapter X,” James Joyce Quarterly, vol. 8, no. 4 (Summer 1972), 405–22.
35. Curtius, European Literature, 505.
36. Stanislaus Joyce, Recollections of James Joyce, trans. Ellsworth Mason (New York: James Joyce Society, 1950), 19.
37. Gilles Deleuze and Félix Guattari, “One or Several Wolves?” in A Thousand Plateaus: Capitalism and Schizophrenia, trans. Brian Massumi (Minneapolis: University of Minnesota Press, 1987), 29.
38. In an interesting twist, M’Coy also asks Bloom to add his name to the list at the funeral, even though he plans to go to Sandycove to wait and see if the body of the drowned man is found. After his departure, Bloom’s thoughts, inspired by M’Coy’s reminder about Dignam, turn to his own dead father, leading him to remark, “Wish I hadn’t met that M’Coy fellow” (U, 63).
39. In their quantitative analysis of character networks in a corpus of sixty different nineteenth-century British novels, Nicholas Dames, David Elson, and Kathleen McKeown were curious to know if urban novels were more crowded than rural ones. They were surprised to discover that urban settings did not correspond with an increase in the number of characters. Instead, they concluded that the narrative standpoint from which they were imagined (whether omniscient or not) was a more significant factor in the population increase. “Extracting Social Networks from Literary Fiction,” Proceedings of the 48th Annual Meeting of the Association for Computational Linguistics, Uppsala, Sweden, 11–16 July 2010, 138–47.
40. Groden, Ulysses in Progress, 55.
41. None of the characters from episode 16 appear.
42. For these numbers, I am relying on Benstock’s lists, which include both a straightforward “Directory of Names” but also an Appendix with “Molly’s Masculine Pronouns” and another decoding the many “Anonymous Listings.” Benstock and Benstock, Who’s He When He’s at Home, 229–33, 169–216.
43. It was the first edition of Thom’s Irish Who’s Who: A Biographical Book of Reference of Prominent Men and Women in Irish Life at Home and Abroad (Dublin: Alexander Thom, 1923). The rough estimate, “upwards of 2,500,” is given in the preface.
44. February 26, 1923, L III, 73.
4. GIS JOYCE
The epigraphs to this chapter are from two of the earliest reviews of Ulysses: Sisley Huddleston, in the March 1922 issue of Observer; Joseph Collins in a May 1922 issue of the New York Times. Reprinted in James Joyce: Critical Heritage, vol. I, ed. Robert Deming (New York: Barnes & Noble, 1970).
1. The last documented copy was sold to Lewis Galantiere on July 1, 1922 (Beach writing “Only the Dutch and verge d’Arche lists are complete and have been kept up to date until July 1, 1922”), but others were still being made available into the fall of that same year, which indicates that some were held back. I am very grateful to Ted Bishop for sharing his materials with me for this chapter and his many valuable insights in “The ‘Garbled History’ of the First Edition of Ulysses,” Joyce Studies Annual 9 (1998): 3–36. Thanks also to Laura Barnes for generously answering my questions about the messy business of documenting the who, what, and where of this first edition.
2. The original number Joyce reported to Weaver in April 1921 was 1,020. L I, 162. As late as February 3, when the final bill was presented to Beach, Darantière was planning to print 1,047 copies (770 vergé à barbes, 155 arches, 102 hollandes, and 20 ordinaires) but never followed through, in part, because it was already too costly, and a second edition was already in the works for October. The number 1,020 was eventually arrived at, but the story is not so straightforward. On March 16, 1922, Darantière explained to Beach that he was unable to print all twenty press copies because of the ones that were damaged along the way. His revised total broke that number down to 754 vergé à barbes, 151 arches, 100 hollandes, 13 exemplaires de presse along with “deux paquets de defets.” In the catalogue she compiled for Glenn Horowitz, Laura Barnes, using Beach’s notebook, pins it down to an extra nineteen copies (fourteen for review and five for family and friends) but there’s another “unnumbered press copy” that she discovered in the Bibliothèque Nationale de France that was deposited by Darantière for copyright. In addition, there are the retained printer’s copies, an unnumbered one given to Henry Kaeser by Joyce from the one hundred pile and another one inscribed to Darantière by Joyce on October 12, 1922 from the 150 pile. Some of the numbered copies were erased and rubber stamped with “Unnumbered Press Copy,” including the three mentioned by Slocum and Cahoon, A Bibliography of James Joyce: From the Yale University Libraries Collection (New Haven, CT: Yale University Press, 1953), 25.
3. Ten percent then 20 percent discount was a problem. The cheaper edition sold out before the more expensive ones, but there were also some booksellers who used the discount to their advantage.
4. As reported by Sylvia Beach on August 7, 1922: “I got all the copies of ‘Ulysses’ to the subscribers (in all states except New York) through the post without any trouble.” The Letters of Sylvia Beach, ed. Keri Walsh (New York: Columbia University Press, 2010), 101. A majority of the subscription forms, along with the relevant address books and correspondence, can be found online in the Sylvia Beach Papers at Princeton University: https://findingaids.princeton.edu/collections/C0108/. Some of them can also be found in the Sylvia Beach papers in the Poetry Collection, Capen Library, University at Buffalo, New York.
5. I say allegedly since Joyce, prone to exaggeration in these matters, was the one to report it: “A second edition of Ulysses was published on the 12 October. The entire edition of 2000 copies at 2.2.0 [missing pound sign] a copy was sold out in four days” (L I, 191). In a letter dated October 9, 1922, Darantière informed Beach that 108 copies were being sent.
6. The names and addresses used in these maps come from a variety of sources that include address books, subscription forms, lists, cards, letters, and censuses. Beach was very inconsistent with her tracking methods, so a lot of this material had to be checked against multiple sources whenever possible. Only confirmed subscriptions are included in this data set, though others will certainly be found. What makes all of this even more difficult is the fact that some subscribers ordered through bookstores and agents, and Beach was less than thorough with her documentation of the copies that would have sold through her store for those who never actually filled out subscriptions. Anyone interested in the subscription forms and orders received can access the digital reproductions. See “Slips with Names of Subscribers,” Sylvia Beach Papers, C0108, Manuscripts Division, Department of Rare Books and Special Collections, Princeton University Library.
7. Beach uses this phrase in “The Ulysses Subscribers,” an essay she submitted to two different journals in 1950, but it was turned down. That same year she signed a contract for the book version with Harcourt, Brace, and Jovanovich, and portions of this essay were revised for her memoir. See “The Ulysses Subscribers” Sylvia Beach Papers, C0108, Manuscripts Division, Department of Rare Books and Special Collections, Princeton University Library. A digital reproduction can be found at https://figgy.princeton.edu/catalog/11e13587-f477-48f8-9b31-3ae7623d96e0
8. The total was divided in different amounts between December 1921 and June 1922. The first installment required using any profit from the business, and possibly loans, but subsequent ones were to be covered by subscribers. “My business is going well and I could tackle the carpenter bill if I didn’t have to put every single centime aside to pay the printer of Ulysses five thousand francs on the 1st of December. He requires it and naturally the cheques from subscribers will not arrive in time for that first payment.” Beach, Letters of Sylvia Beach, 88; emphasis added.
9. Rachel Potter, Obscene Modernism: Literary Censorship and Experiment (Oxford: Oxford University Press, 2013), 73.
10. Egmont Arens was a notable exception, paying only 1,620 francs out of 3,000. Beach sent her father to pick up the balance four years later (in 1926) without interest or the addition of shipping costs. Bishop, “The ‘Garbled History,’” 20. For a history of the common reader, see William Powell Jones, James Joyce and the Common Reader (Norman: University of Oklahoma Press, 1955). For an account of the common reader after Hans Walter Gabler’s 1984 edition, see Julie Sloan Brannon, Who Reads “Ulysses”? The Rhetoric of the Joyce Wars and the Common Reader (New York: Routledge, 2003).
11. Potter, Obscene Modernism, 62.
12. Information about the ban in the United States was used for the promotional material, including the prospectuses announcing that it was “suppressed four times during serial publication in The Little Review.”
13. Potter, Obscene Modernism, 65.
14. Lawrence Rainey argues that the subscriptions were one way to make Ulysses a desirable commodity for collectors, but Potter explains that he downplays the novel’s status as pornography that would need subscriptions to circulate. “It is only in retrospect,” she writes, “that the distinctions between Ulysses and pornographic texts, and the literary market and the pornography trade, are straightforward. It did not look so clear-cut to lawyers, judges, vice crusaders, publishers, customs officials, or others at the time.” Potter, Obscene Modernism, 69.
15. For the “unnumbered press copies” of the second edition, Joyce instructed Weaver to have a stamp made that would be applied “in the space left for no…?” (L I, 194).
16. The most detailed census was sponsored by Glenn Horowitz, a rare book dealer in New York City, and overseen by Barnes. Cataloguing subscribers, not mapping them, was the primary concern. See note 3.
17. Read, Pound/Joyce, 123.
18. Beach, Shakespeare and Company, 48, 96.
19. Febvre and Martin, The Coming of the Book, 111.
20. Ezra Pound, The Selected Letters of Ezra Pound, 1907–1941, ed. D. D. Paige (New York: New Directions, 1971), 262–63.
21. On April 10, 1921, Joyce provides an account to Weaver with a breakdown of the edition, adding that there would be “1,000 copies, with 20 extra copies for libraries and press” (L I, 162). On April 18, 1921, Darantière sends a full cost breakdown for one thousand copies with “la justification du tirage.” On April 23, 1921, Beach mentions to her sister that “It [Ulysses] will appear in October—a thousand copies—subscriptions only” (85).
22. April 19, 1921, L III, 40.
23. Vergil to Dante: “and he pointed out and named to me more than a thousand shades” (Canto V, 67–68); See Hopper, Medieval Number Symbolism, 29.
24. L I, 196. In his journal, Stuart Gilbert also recalls Joyce throwing out a rough calculation about his losses to Roth: eight thousand copies at seven or eight dollars apiece of which an unspecified percentage should go to him. Quoted in Alastair McCleery, “The Reputation of the Odyssey Press Edition of Ulysses,” Papers of the Bibliographical Society of America 100, no. 1 (2006): 96.
25. Quoted in Robert Spoo, Without Copyrights: Piracy, Publishing, and the Public Domain (Oxford: Oxford University Press, 2013), 217.
28. If we follow the paid subscription trail between May 19, 1921 and July 1, 1922, we discover that with 142,000 francs Beach more than covered the costs of Darantière’s bill. If Beach sold every copy without a discount and did not give any away, the total would have been 185,000 francs. Maurice Saillet, “List of Subscribers, ca. 1921,” Harry Ransom Center, University of Texas at Austin, box 262, folder 10.
29. Beach, “The Ulysses Subscribers.” The American subscribers had to wait the longest. In a letter to Marion Peter (owner of the Sunwise Turn bookshop) dated August 7, 1922, Beach explains how copies can be brought by bookleggers to avoid customs officials, but that was six months after the printing was completed. See Letters of Sylvia Beach, 101.
30. The names and addresses for these letters were taken from two separate lists. The first, entirely in Joyce’s hand, the second also in Joyce’s hand but with additions from Lucia Joyce and notations from Sylvia Beach. Originals can be found in the James Joyce Collection at the University at Buffalo, XXVIII Miscellaneous Material Related to Joyce’s Works, “Ulysses: The First Edition,” folders 7 and 8.
31. I got this number from the orders logged in Darantière letters to Beach. With his usual exaggeration, Joyce informed Weaver on August 7, 1921 that “5000 prospectuses” had gone out with “the capture” of a mere “260.” Original letter in the Harriet Shaw Weaver collection at the British library.
32. Quoted in Beach, Shakespeare and Company, 85. Discussing her tactics for enlisting subscribers, Beach claimed that “no one escaped from the rue de l’Odéon without subscribing.” But there was also Robert McAlmon, who would collect subscriptions from night clubs, leaving many people “surprised to find themselves subscribers.” That may help to explain the large number that didn’t pay up when the copies actually arrived (50, 51).
33. The original plan was for seventeen uncovered press copies to be sent out: five for the Continent, six for England, and six for the United States. The two unnumbered copies retained by Darantière were signed by Joyce on October 12, 1922: one was given to the printer, the other to Henry Kaeser. See Sarah Funke, Glenn Horowitz, and Laura Barnes, James Joyce Ulysses: An Exhibition of Twenty-Four Inscribed Copies of the First Edition (New York: Glenn Horowitz Bookseller, 1998), 28, 34.
34. Ellmann, “Ulysses” on the Liffey, 531.
35. Four hundred subscriptions may have been sent in by November 1921, but in Beach’s papers, there are only twenty separate payments, some of them for multiple copies.
36. L II, 182; L II, 145; L III, 61. Pound repeated this exact number to Alice Corbin Henderson on March 12, 1922. See The Letters of Ezra Pound to Alice Corbin Henderson, ed. Ira Nadel (Austin: University of Texas Press, 1993), 224.
38. September 30, 1924, SL, 302.
41. One of them included a personal note asking if the edition was “in English” (George Daugherty).
42. Beach, “The Ulysses Subscribers”: “I was still busy delivering the 1st edition of Ulysses to places all over the globe.”
43. In one of her lists for potential subscribers she also included a “miscellaneous” section lumping together subscribers from Ireland, Scotland, Holland, Jerusalem, Belgrade, Spain, Italy, Czechoslovakia, Germany, India, Belgium, Sweden, Norway, Brazil, Argentina, Austria, Norway, Luxembourg, and South Africa. See note 4.
44. Beach, Shakespeare and Company, 62. Though there is evidence that a subscription form was sent to Sarawak Straits and orders came in from China for the second and third edition. There’s also a curious anecdote about his Chinese readers that Joyce reported to Weaver in October 1923: “A friend of his [John Quinn] told me that there is a club in the far east where Chinese ladies (not American as I supposed) meet twice a week to discuss my mistresspiece. Needless to say the said club is in—shavole Shanghai!” (L I, 206).
45. Kenner, Joyce’s Voices, 1.
46. I’ve taken this phrase from an unsigned article that appeared in the Nation, December 20, 1933.
47. There was also the pirated edition that Samuel Roth printed in the United States in 1929. David Magorlick points out though that a majority of the copies were personal imports from Paris: “Illicit copies of Ulysses bound in light-blue paper, became nearly as available in the late 1920s as bootleg gin. It became commonplace for Americans in Europe to bring copies back with them, and the supply was sufficient for T. S. Eliot to include it as required reading for a course on modern literature he taught at Harvard in the late 1920’s.” See David Magorlick, “Judge’s Ruling Still a Landmark 50 Years Later,” New York Times, December 6, 1983. A year later, Eliot would complain to Dorothea Richards (who kept a record in her diary) “how the young don’t now get anything out of Ulysses now its notoriety is gone.” Letters of T.S. Eliot, ed. Valerie Eliot and Hugh Haffenden, vol. 6, 1932–33 (New Haven, CT: Yale University Press, 2016), 697.
48. Forty to fifty booksellers also received a letter asking them a similar question: “If the book were declared legal would you stock it for sale, and cooperate in effecting its general circulation.” See Michael Moscato and Leslie Le Blanc, eds. The United States of America v. One Book Entitled “Ulysses” by James Joyce: Documents and Commentary—A 50 Year Retrospective (Frederick, MD: University Publications of America, 1984), (henceforth U v. u). For a discussion of the role librarians played in this case, see William Brockman, “American Librarians and Early Censorship of Ulysses: Aiding the Cause of Free Expression?” Joyce Studies Annual 5 (1994): 56–74.
49. Moscato and Le Blanc, The United States of America v. One Book Entitled “Ulysses” by James Joyce, 124; emphasis added.
50. U v. u, Ernst, Appendix 4, 421. The full results of this questionnaire, “Re: Ulysses, Answers from Librarians and Booksellers,” complete with the map and tally of answers collected by Ernst’s team, can be found in the Morris Ernst Papers at the Harry Ransom Center, University of Texas at Austin
51. U v. u, 141. Ernst had wanted the data presented in an “easy form” to the court, as he told Cerf, suggesting further that “they might even be analyzed as to number of states in which libraries are anxious for the book.” A carefully chosen sequence of “comments of librarians on Ulysses” was also used as Exhibit C, many of them with dramatic critical flourishes (196).
52. The map does not indicate the locations or libraries where copies already existed. Twenty-four of the librarians were willing to admit the presence of copies in their collections, only two of them from 1922.
53. Ernst, “Brief for Claimant Appellee,” appendix 3, 66.
54. Without citing any sources, Herbert Gorman claimed that the number climbed to 50,000 by the time he was writing his biography some time in 1937 or 1938 (323).
55. Edwin McDowell, “Publishing: Cheaper Corrected: Ulysses Awaited,” New York Times, June 22, 1984, 26.
56. An article in the New York Times provides a further breakdown of this number. For the Random House edition there were 544,000; in the Modern Library Giant Edition, 69,000; in the regular edition, and since 1967, 275,000 in the Vintage paperback. The Bodley Head includes the 350,000 hard-cover copies, including the original edition of one thousand, and a new edition in 1960, and soft-cover sales of 300,000 after Penguin bought paperback rights in 1967. See Richard Shephard, “U.S. Sales of the Book at 880,000,” New York Times, February 2, 1972, 41.
57. This dataset was collected from Darantière’s letters. Each shipment was accompanied by a summary of the contents, but there are copies missing and duplicates. At least thirty-three copies slipped by and were left unaccounted for. I haven’t come across any document on Beach’s side that would indicate she was keeping track. See the “Correspondence from Imprimerie Darantière to Sylvia Beach” in the James Joyce Collection at the University at Buffalo.
58. Gorman, James Joyce, 304.
59. Gorman, James Joyce, 306.
60. Noel Riley Fitch provided one of the earliest and most erroneous lists of first subscribers: “The book ledger for England, Ireland, and Scotland was slowly filling with familiar names: Bennett, two Huxleys, three Sitwells, Woolf, Churchill, Wells, Walpole, and Yeats.” Sylvia Beach and the Lost Generation: A History of Literary Paris in the Twenties & Thirties (New York: W. W. Norton, 1985), 105. Only Woolf and Yeats subscribed: the others may have sent in subscription forms without paying, but it’s doubtful.
61. Andrew Shyrock and Daniel Lord Smail, eds., Deep History: The Architecture of Past and Present (Berkeley: University of California Press, 2011), 247.
62. Febvre and Martin, “Geography of the Book,” in The Coming of the Book, 167–215.
63. Reinhart Koselleck, Futures Past: On the Semantics of Historical Time, trans. Keith Tribe (New York: Columbia University Press, 2004), 260.
64. Hans Robert Jauss, “Literary History as a Challenge to Literary Theory,” New Literary History 2, no. 1 (Autumn 1970): 12.
65. In her memoir, Monnier has a chapter devoted exclusively to “Joyce’s Ulysses and the French Public.” In Monnier, The Very Rich Hours of Adrienne Monnier, 112–26.
66. As indicated by Beach to Weaver on June 26, 1922: “Adrienne Monnier…took over a number and is beginning to sell them at Fr 500 (edition at Fr 150) and will give the proceeds to Mr. Joyce. Perhaps you will give that address to people who inquire where they can obtain ‘Ulysses.’” Letters of Sylvia Beach, 97.
67. Though Beach is often prone to exaggeration in her memoir, I find this particular account believable not least because of the degree of detail. See Beach, “The Ulysses Subscribers.”
68. Rainey identified the one hundred copies of William Jackson, a number that is found in Beach’s UK address book for the 150 trade copies, but there is no proof of payment. The census compiled by Laura Barnes, however, shows that at least three copies from the 350 version were purchased. What complicates it further is the presence of two payments in Maurice Saillet’s notebook (May 5 and June 8), each one for 5,200 francs. Given the total 10,400 franc payment, and the possibility of the trade discount from Beach, it appears that one hundred copies at 100 or 120 francs could have been the original plan—that is, until copies sold out, forcing Jackson to purchase the more expensive option. Proof that only eighty-seven copies were bought by Jackson can be found in the receipt housed in Beach’s papers identifying that he bought eighty-five at 150 francs, and two at 250 francs. See “Order Forms J,” in Beach Papers, Princeton University.
69. Lawrence Rainey, “Consuming Investments,” James Joyce Quarterly 33, no. 4 (1996), 541.
70. Edward Bishop, “The Garbled First Edition of Ulysses,” Joyce Studies Annual, vol. 9 (Summer 1998): 4.
71. Beach, Shakespeare & Company, 85.
72. Beach claimed that the first copies to arrive in the United States were “confiscated at the Port of New York” forcing her to suspend shipments. If true, this may suggest that the copies confiscated had been replaced by the ones already printed (86).
73. Margaret Anderson, My Thirty Years’ War: An Autobiography (New York: Covici, 1930), 175.
74. Locations of librarians in New York State on the map of library responses include Buffalo, Rochester, Ithaca, Binghamton, Schenectady, Albany, and New York.
75. In 1939, Vivian Mercier, who smuggled a copy past “two customs barriers” in 1922, provided one characteristic response: “It would be natural to assume that the book [Ulysses] was banned in the Irish Free State under the Censorship of Publications Act, 1929, but in fact it never was: even the customs prohibition order against it was withdrawn in 1932.” See “Literature in English, 1892–1921,” in A New History of Ireland, vol. 6, ed. W.E. Vaughan (Oxford: Oxford University Press, 1989), 505.
76. James Joyce, Exiles (New York: B. W. Huebsch, 1918), 38.
77. Beach, 86. The recipients include Mrs. Murray (unnumbered), John Joyce (unnumbered), Michael Healy (unnumbered), C. P. Curran (unnumbered), Charles Rowe (#?), MJ McManus (#313), John Nolan (#609), Thomas Kiersey (#?), Lennox Robinson (#751), L. A. Waldron (#378), Hodges and Figgis (seven copies), Fred Hanna (six copies), Irish Book Shop (thirteen copies), Combridge & Co. (four copies), National Library of Ireland (press copy).
78. James Douglas, “Beauty and the Beast,” Sunday Express, May 28, 1922, 5.
79. Beach had her own wish list, but it was complemented by the ones collected by Sherwood Anderson, Robert McAlmon, and Alfred Kreyemborg. In these counts, I have not included the names of Ezra Pound, John Quinn, Baron Ambrogio Ralli, and Harriet Shaw Weaver because they were crossed out. I suspect that indicates they were already on Beach’s list and/or had already committed. The original “List of Clients Interested in Ulysses” is housed in the Sylvia Beach Papers, Manuscripts Division, Department of Rare Books and Special Collections, Princeton University and has been digitally reproduced here: https://findingaids.princeton.edu/collections/C0108/c00747.
80. Beach, Shakespeare and Company, 62.
81. Spoo, Without Copyrights, 183.
82. Ellmann, “Ulysses” on the Liffey, 159.
83. At least that was the number originally planned on in April 1921. Larbaud to Joyce, as reported in interview with Hoffmeister: “He compared it to a starry sky: when you look at it long and carefully, it becomes more beautiful as new, uncounted stars reveal themselves.” L I, 162.
84. It was printed with a batch of one hundred others in vergé à barbes and 8 verge d’Arches a month after the first copy arrived and with hundreds of others still left to do.
85. Jauss, “Literary History as a Challenge to Literary Theory,” 9.
5. DATING ULYSSES
1. Dubliners, for instance, is dateless. And while it is true that Portrait also ends with a series of cities and dates (Dublin, Trieste, 1904–1914), there’s no evidence to prove that Joyce was responsible for the addition or the way it was formatted. It’s more than likely that it was taken for book publication from the final serial installment in the Egoist. Gorman, however, claims that Joyce “added the dates Dublin, 1904, Trieste, 1914,” but he also mentions that there was a finis at the end (which there is not). Gorman, James Joyce, 221.
2. September 7, 1915, Read, Pound/Joyce, 364; emphasis mine.
3. Seven is one option for counting 1914–1921, and it is based on a more generalized quantity since the composition does not include ALL the months in these years. Joyce keeps the ambiguity, but for those who work through the evidence, as I will in what follows, the beginning for him was in March 1914 and the end in October 1921. The fact that there are only years and not months or days is central to understanding that the concept of compositional time was elastic and could be bent in many directions depending upon the different meanings that Joyce and others have wanted to assign to it.
4. Joseph Priestley, “A Description of a Chart of Biography,” in A Chart of Biography (Warrington: William Eyres, 1764), 5.
6. Margaret McBride, “The Ineluctable Modality: Stephen’s Quest for Immortality,” in James Joyce, Harold Bloom, ed. (Philadelphia: Chelsea House, 2009), 117.
7. It was definitely a more common practice for poets who would use dates to mark the day or month of composition.
8. The 1932 Odyssey Press is an exception. “The End” was inserted after the final inscription.
9. Edward Said, Beginnings: Intention and Method (New York: Columbia University Press, 1985), 49; Frank Kermode, The Sense of an Ending: Studies in the Theory of Fiction (Oxford: Oxford University Press, 1967); October 7, 1921. L I, 172.
10. Edward Said, Beginnings: Intention and Method (New York: Columbia University Press, 1985), 49; Frank Kermode, The Sense of an Ending: Studies in the Theory of Fiction (Oxford: Oxford University Press, 1967); October 7, 1921. L I, 172.
11. Alain Robbe-Grillet, “The Use of Theory,” in For a New Novel: Essays on Fiction (Evanston, IL: Northwestern University Press, 1992), 10.
12. A wonderful coincidence that this particular letter was written on the same day that the plot of Ulysses is set. “Written in the land of the living, on the right bank of the Adige, in Verona, a city of Transpadane Italy; on the 16th of June, and in the year of that God whom you never knew the 1345th.” Quoted. in Anthony Grafton, Cartographies of Time: A History of the Timeline (Princeton, NJ: Princeton University Press, 2012), 16.
13. Other signoffs can be found as early as his 1903 “Epiphanies” (JAJ 13/2/03) and his 1904 “Portrait of the Artist” essay (Jas. A. Joyce 7/1/1904).
14. Critics have suggested the possibility of counting eight years, but that would require including ALL the months. So, for instance, if Joyce began writing on January 1, 1914 and ended on January 31, 1921, that would be eight years. Anything less would be seven plus the months added. Others count from 1914 to 1915, 1915 to 1916, 1916 to 1917, 1917 to 1918, 1918 to 1919, 1919 to 1920, 1920 to 1921. Based on comments from Joyce, Nora, and the others he primed with information, including Gilbert and Gorman, we know that Joyce preferred seven.
16. He compiled another holograph notebook entitled “Shakespeare Dates” between 1916 and 1917. It is housed in The James Joyce collection at the University at Buffalo, James Joyce Collection, V.A.4.
17. The Joyce Papers, National Library of Ireland, NB MS 36, 639/5/B, p. 10 [v.].
18. John Henry Raleigh, The Chronicle of Leopold and Molly Bloom: “Ulysses” as Narrative (Berkeley: University of California Press, 1978), 4–5.
19. Here are two of them: on August 16, 1921 Joyce tells Budgen that “Molly Bloom was born 1871” (L I, 170), and in episode 17 the Narrator identifies “1870” (U, 605).
20. As evidenced from a letter dated September 3, 1921, in which he prepares to “give Molly another 2,000 word spin.” L III, 48.
21. See Peter Costello, Leopold Bloom: A Biography (New York: Harper Collins, 1983); Raleigh, Chronicle of Leopold and Molly Bloom, 51, 97.
22. Ellmann makes a point of mentioning that Joyce is thirty-eight years old when arriving in Paris in 1920, partly to identify that he had matured significantly since his years in Trieste and Zurich. “Joyce, the artist,” he writes, “now thirty-eight, completed Circe and then the final three episodes of Ulysses.” Ellmann, James Joyce, 486.
23. Wyndham Lewis, Time and Western Man, Book One (London: Chatto & Windus, 1927), 83.
24. Henri Bergson, Time and Free Will: An Essay on the Immediate Data of Consciousness (Mineola, NY: Dover, 2001), 104
25. The connection between 1914 and World War I is suggested but not confirmed, as it would be with Lukács in his 1962 preface to Theory of the Novel: “The first draft of this study was written in the summer of 1914 and the final version in the winter of 1914–15…. The immediate motive for writing was supplied by the outbreak of the First World War.”
26. Rodney Wilson Owen, James Joyce and the Beginnings of “Ulysses” (Ann Arbor: UMI Research Press, 1983), 3
27. Gorman, James Joyce, 224.
28. November 1, 1921, L III, 52.
29. Gorman, James Joyce, 224
30. In one of his reminiscences of the Ulysses origin myth, Georges Borach recorded the following conversation with Joyce on 1 August 1, 1917: “I first wished to choose the title Ulysses in Dublin, but gave up the idea. In Rome, when I had finished about half the Portrait, I realized that the Odyssey had to be the sequel, and I began to write Ulysses” (Portraits of the Artist in Exile, 69–70). Crispi agrees: “Given its cast of characters and relatively simple style (before Joyce overhauled the episode in 1921), it is possible that as early September 1906 to 1907 he had written some sketches or one or more proto-drafts of at least some of the scenes that appear in this episode.” Luca Crispi, “A Ulysses Manuscript Workbook,” Genetic Joyce Studies, issue 17 (Spring 2017), https://www.geneticjoycestudies.org/static/issues/GJS17/GJS17_Crispi_Appendix.pdf
31. See L II, 168, Joyce to Stanislaus Joyce, Postmark September 30, 1906, when Joyce first mentioned his plans for a new short story. See also L II, 190, Joyce to Stanislaus Joyce on November 13, 1906; L II, 190, Joyce to Stanislaus Joyce on February 6, 1907; L I, 98, Joyce to Harriet Shaw Weaver on November 8, 1916; L II, 392, Joyce to C. P. Curran on March 15, 1917; Owen, James Joyce and the Beginnings of “Ulysses,” 3–6.
32. And even though it is the year in which he wrote the essay “Portrait of the Artist,” the actual writing of Portrait would not start until 1907. It’s more likely that 1904 is significant as a date because this is the year he left Dublin.
33. Ezra Pound, “James Joyce et Pécuchet,” Mercure de France, June 1, 1922: “Novelists like to spend only three months, six months on a novel. Joyce spent fifteen years on his. And Ulysses is more condensed (732 pages) than any whole work whatsoever by Flaubert; more architecture is discovered”; emphasis added.
34. Ellmann, James Joyce, 524.
35. Owen is skeptical that Joyce could have had done much writing at all during that busy year, guessing that he could not have written more “than two dozen pages.” The last word is a mystery as well, though Gorman claims that for Finnegans Wake “it was in mid-November, 1938.” Owen, James Joyce and the Beginnings of “Ulysses,” 111, 347.
36. Ellmann, James Joyce, 357, 383. Owen: “Of the seven letters that are implied by Pound’s correspondence, only two from Joyce survive, and neither mentions Ulysses; unfortunately missing are Joyce’s replies to Pound’s questions about the novel in August 1915 and 1916.” Owen, James Joyce and the Beginnings of “Ulysses,” 117. Some of these unpublished letters can be found in the Pound Collection at the Beinecke Library.
37. SL, 209. The first and earliest known reference to the day was made to Weaver on October 10, 1916: “I am writing it as well as I can. It is called Ulysses and the action takes place in Dublin in 1904…. I hope to finish it in 1918.” L II, 387.
38. July 14, 1915. This letter is unpublished. Joyce had sent it to Pound so that he could, then, pass it on to Yeats, who was helping to secure money from the Royal Literary Fund. The portion I am quoting here was not included in Pound’s edited version as reprinted in L II, 352.
39. As was the case on 12 July 1915, when he corrected Weaver: “May I point out that there is a slight mistake in your advertisement columns. My novel did not begin in your paper on the 1 March 1914 but on the 2 February 1914—which, strange to say, is also my birthday.” L I, 83.
40. September 14, 1916 (it is mistakenly dated 14 October at the top by someone else). Unpublished letter. Ezra Pound Papers, YCAL MSS 43, box 26, folder 1112, Beinecke Rare Book and Manuscript Library, Yale University.
41. Joyce to Pound, July 24, 1917, unpublished. Quoted in Ellmann, “James Joyce, 416. Ezra Pound Papers, YCAL MSS 43, box 26, folder 1112, Beinecke Rare Book and Manuscript Library, Yale University; Joyce to Weaver on November 8, 1916: “I began this in Rome six years [ago] (or seven) and am writing it now. I hope to finish it in 1918” (SL, 223). Joyce to Curran on March 15, 1917: “I am writing Ulysses which I began six or seven years ago in Rome and hope I shall be able to finish it in 1918” (L II, 392). The notes he made from this visit were incorporated into the final chapter of Portrait and various episodes of Ulysses, indicating that at one point, at least, this was reason enough to revise the date. In Jane Lidderdale’s Dear Miss Weaver: Harriet Shaw Weaver 1876–1961 (New York: Viking, 1971), Joyce’s confusing timeline about the time and place of the beginning of Ulysses remained intact: “she [Weaver] knew no more about the book than the odd scraps gleaned from letters: that Mr Joyce had begun it in Rome in 1909 or 1910” (146).
42. See Walton Litz, “Early Vestiges of Joyce’s Ulysses,” PMLA 71, no. 1 (March 1956), 59. Using evidence from the “Alphabetical Notebook” Joyce compiled in 1910 for Portrait, Luca Crispi even goes so far as to identify the episodes: episode 3, episode 5, episode 6.
43. Gabler, “Rocky Road,” 3.
44. Budgen, Making of Ulysses, 22.
45. He wanted Portrait published before the end of 1916 because it was his lucky year. Ellmann, “Ulysses” on the Liffey, 406. A much earlier precedent for this practice of counting the sum of yearly numbers can be found in Giovanni Boccaccio, whose Decameron is set in 1348 during the black plague: 1+8=9, the number of the trinity squared, and 3+4 add up to 7, the number of virtues and days of creation. See Aldo S. Bernardo, “The Plague as Key to Meaning in Boccaccio’s Decameron,” in The Black Death, ed. Daniel Williman (Binghamton, New York: Medieval & Renaissance Texts & Studies, 1982), 39–64.
46. Joyce to Weaver on April 3, 1921, L I, 160.
47. Joyce to Quinn on November 24, 1920, L III, 30.
48. Thomas Hardy, Poems of Thomas Hardy: A New Selection, ed. T.R.M. Creighton (London: MacMillan Press, 1977), 318; For a discussion of numerical thinking in Hardy, see Kent Puckett, “Hardy’s 1900,” Modern Language Quarterly vol 75, no. 1 (March 2014): 57-75.
49. As reported to Weaver on June 24, 1921. L I, 166.
50. Tom Stoppard, Travesties (London: Faber & Faber, 1975), 44.
51. L III, 51. He continued: “I have still a lot of proofreading and revising to do but the composition is at an end…. About 3 weeks more stewing over proofs and all is finished. So three cheers for Bloom!” (L I, 175; emphasis added). A day later, and he wrote to Larbaud: “I finished Ithaca last night so that now the writing of Ulysses is ended, though I have still some weeks of work in revising the proofs.” Another day later, and he informed Weaver as if it had just happened: “I have now finished the Ithaca episode and with that the writing of Ulysses” (L III, 52).
52. Gabler, Ulysses: A Critical and Synoptic Edition, 1890.
53. Richard Madtes, A Textual and Critical Study of the “Ithaca” Episode of James Joyce’s “Ulysses” (New York: Columbia University Press, 1961), 9.
54. Madtes, Textual and Critical Study of the “Ithaca” [Episode 17]. According to Madtes, eighty-three of the 309 questions were added after the basic manuscript was written. That makes up 9,380 of the 22,421 words in the episode. He also calculated the number of additions: twenty-four in the manuscript, thirty-one in the first typescript, two in the second typescript, four in the first galley proof, fourteen in the second galley proof, and eight in page proof (44).
55. Michael Groden first pointed out that two hundred pages of Ulysses still needed to be set in page proofs in the final month as the process carried over into 1922 (Ulysses in Progress, 191).
56. “He was doing something that he had not done in the case of his previous books, actually composing, for his voluminous additions were no less than that, while his manuscript went through the press and came out in the form of galley and page proofs” (Herbert Gorman, James Joyce, 288).
57. Darantière to Joyce, August 31, 1921.
58. The first flyer printed July 5, 1921, another one in September 19/20, 1921.
59. In a letter to Weaver dated November 1, 1921, he wonders “on whose [birthday] it will be published.” In a footnote Ellmann suggests that “Joyce had already made up his mind that it should appear on 2 February 1922, his own birthday,” but other letters written at the same time suggest that Joyce was still thinking the end of November, possibly early December (L I, 175). Joyce may have had fantasies about his book appearing on February 2, but he also was at the mercy of the printers, to whom he was sending copious corrections. This coincidence was a late development in the process. In fact, by January 23, he was still thinking that Ulysses would appear on the 28th (Saturday) or the 30th (Monday) of the month. L III, 57.
60. The last set of proofs for episode 18 was sent to Joyce between January 16 and 20.
61. Both Groden and Gorman use eight years when calculating the amount of time Joyce spent writing Ulysses (Ulysses in Progress, 3; James Joyce, 223). Adrienne Monnier arrives at the same count in her memoir: “we had been waiting eight years” (Very Rich Hours of Adrienne Monnier, 114). Gilbert Seldes, in his review of the novel, arrives at the same calculation: “It has stood for eight years as the pledge of Joyce’s further achievement.” Gilbert Seldes, The Nation, August 30, 1922, 211–212.
62. In an interview with Adolf Hoffmeister in 1930, Joyce identified a few historical precedents for his numerical superstitions: “Dante was obsessed by the number three: he divided his verse into three cantiche, each written in thirty-three cantos and in terza rima. Why do we depend on the number four? Four legs on a table, four legs on a horse, four seasons and four provinces of Ireland? Why are there twelve pillars of the law, twelve apostles, twelve months, twelve of Napoleon’s marshalls? Why was the armistice of the Great War sounded at the eleventh hour of the eleventh day of the eleventh month? Number as a measure of time is indeterminable. The ratio of these numbers is relative to place and content.” See Adolf Hoffmeister, “The Game of Evenings,” trans. Michelle Woods, Granta 89 (April 2005).
63. Kenner, Joyce’s Voices, 59–60. When revising page 77 in the proof stage, D. F. McKenzie points out that Joyce added the phrase “seventh heaven.” See McKenzie, Bibliography and Sociology of Texts, 59.
64. Don Gifford and Robert J. Seidman, Ulysses Annotated: Notes for James Joyce Ulysses (Berkeley: University of California Press, 1988), 194.
65. William Shakespeare, As You Like It, Act II, Scene VII, accessed on May 6, 2020 at https://shakespeare.folger.edu/shakespeares-works/as-you-like-it/. In his lecture on Shakespeare, Stephen adapts this very line: “Every life is many days, day after day. We walk through ourselves, meeting robbers, ghosts, giants, old men, young men, wives, widows, brothers-in-love, but always meeting ourselves” (U, 175).
66. Ellmann, James Joyce, 519. Gorman, James Joyce, 289. Ellmann, like Gorman before him, understood, and he even used Nora’s testimony as backup in his biography. She may have never read the novel, but she was certainly aware of how long the process took, even making the distinction between the time spent thinking (sixteen years) and the time spent writing (seven years) (James Joyce, 524).
67. Ford Madox Ford, Critical Essays, ed. Max Saunders and Richard Stang (Manchester: Carcanet, 2002), 217.
68. Virginia Woolf, The Diary of Virginia Woolf, vol. 5: 1936–41 (New York: Houghton, Mifflin, Harcourt, 1985), 353.
69. F. R. Leavis contended that the variety of styles in Ulysses may have been the reason for so much praise from the “cosmopolitan literary world,” but like Pound, concluded that this novel was a “dead end.” The Great Tradition: George Eliot, Henry James, Joseph Conrad (London: Faber, 2008), 36.
70. The Letters of Ezra Pound, 1907–1941, ed. D. D. Paige (New York: New Directions, 1971), 174.
71. In one of the great dating ironies in literary history, it’s also the case that October 29/30 anticipated by a full year, the first date of the Fascist calendar after Mussolini’s March on Rome.
72. Ezra Pound, “James Joyce to His Memory,” Ezra Pound Papers, Yale University, Beinecke Rare Book and Manuscript Library, YCAL MSS 43, box 31, folder 5540.
73. Peter de Voogd is more precise: it’s a “quadrilateral six-point typesetter’s ‘em.’” See “Joycean Typeface,” Aspects of Modernism: Studies in Honour of Max Nanny, eds. A. Fischer, M. Heusser, and T. Hermann (Tübingen, Germany: Gunter, Narr, Verlag, 1997), 213. Less concerned with the compositor’s dilemma, Joyce referred to it as “un point” when revising the typescript: “La réponse à la dernière demande est un point.” Ulysses: A Facsimile of the Manuscript, ed. Clive Driver, vol. 2, Episode 17 (New York: Rosenbach Foundation, 1975), 32.
74. In 1926, Joyce tells Weaver: “The book really has no beginning or end…. It ends in the middle of a sentence and begins in the middle of the same sentence.” L I, 246.
75. Contrary to what Margot Norris argues, Molly does not seem to be “drift[ing] off to sleep.” See Margot Norris, “Character, Plot, and Myth,” in The Cambridge Companion to “Ulysses,” ed. Sean Latham (Cambridge: Cambridge University Press, 2014), 78. Compare with Bloom’s drifting off at the end of episode 17.
76. In 1930, Joyce admitted to Adolf Hoffmeister, “I never finish any of my work” adding, “Ulysses is the most unfinished.” See Hoffmeister, “The Game of Evenings.”
77. 13. Rabaté and Ferrer agree: “It is likely that Ulysses would never have been finished, or at least that it would have been much longer than it is, if Joyce, by superstition, had not decided that it should be finished the day of his 40th birthday.” Jean-Michel Rabaté and Daniel Ferrer, “Paragraphs in Expansion,” 139.
78. See James Ramsey Wallen, “What is an Unfinished Work?,” New Literary History 46, no. 1 (Winter 2015): 125–42.
79. He continues: “Accordingly, all events in the history of a people, a nation, or a civilization which take place at a given moment are supposed to occur and there for reasons bound up, somehow, with that moment.” See Siegfried Kracaeur, “Time and History,” History and Theory 6 (1966): 66.
80. Budgen, James Joyce and the Making of Ulysses, 17.
81. Gorman, James Joyce, 289.
82. Joyce to Weaver, March 11, 1922, L I, 183. Nora was no exception. A frustrated Joyce kept close watch of her progress, informing his aunt that “She [Molly] has got as far as page 27 counting the cover.” Joyce to Mrs. William Murray, L I, 193. When the fourth edition came out in 1924, he tried again: “The edition you have is full of printers’ errors. Please read it in this. I cut the pages. There is a list of mistakes at the end” (SL, 299).
84. Claude Lévi-Strauss, La pensée sauvage (Paris: Librairie Plon, 1962), 342.
85. In Theory of Literature, René Wellek and Austin Warren describe the shift from calendar centuries to periods in literary history, but they argue that instead of following periods that mirror the “cultural history of man,” we’d be better served by establishing literary periods by literary criteria. “But our starting point,” they write, “must be the development of literature as literature…. A period is thus a time section dominated by a system of literary norms, standards, and conventions, whose introduction, spread, diversification, integration, and disappearance can be traced.” René Wellek and Austin Warren, Theory of Literature (New York: Harcourt, Brace, 1942), 255.
86. Lévi-Strauss: “Dates may not be the whole of history, nor what is most interesting about it, but they are its sine qua non, for history’s entire originality and distinctive nature lie in apprehending the relation between before and after, which would perforce dissolve if its terms could not, at least in principle, be dated.” See Claude Lévi-Strauss, The Savage Mind, trans. George Weidenfeld (Chicago: University of Chicago Press, 1966), 258.
87. See, for instance, Clément Moisan, Qu’est-ce que l’histoire littéraire? (Paris: Presses Universitaires de France, 1987).
88. Hayden White, “Interpretation in History,” New Literary History 4, no. 2 (Winter 1973): 289.
89. Grafton, Cartographies of Time, 19–20.
90. Alan Liu, Wordsworth: The Sense of History (Stanford, CA: Stanford University Press, 1989), 154.
91. Liu, Wordsworth: The Sense of History, 577n44.
92. He returns to the problem thirty years later in Alan Liu, Friending the Past: The Sense of History in a Digital Age (Chicago: University of Chicago Press, 2018).
3PILOGUE: MISCOUNTS, MISSED COUNTS
1. William James, The Meaning of Truth (New York: Longman, Green, 1911), 92–93.
2. Ted Underwood, Distant Horizons: Digital Evidence and Literary Change (Chicago: Chicago University Press, 2019), 182.
3. Underwood, Distant Horizons, 183.
4. Margaret Cohen, The Sentimental Education of the Novel (Princeton, NJ: Princeton University Press, 2018), 23.
5. Gertrude Stein, Everybody’s Autobiography (New York: Random House, 1937), 100.
6. T. S. Eliot, Selected Prose of T.S. Eliot, ed. Frank Kermode (London: Faber & Faber, 1975), 75.
7. Kenner, Dublin’s Joyce, xi.
8. Large discrepancies exist between the document counting the responses as drafted by Ernst’s team and the actual responses. In the Ernst tally, for instance, Texas, Iowa, and Utah, are added as yeses to the question of whether or not there is demand for Ulysses, even though the column in the original is blank. There are also two fewer yeses for New Jersey and six more for New York. See “United States v. One Book Entitled Ulysses by James Joyce,” Morris Leopold Ernst Papers, box 364. 2–3, Harry Ransom Center, University of Texas at Austin.
9. Alexandre Koyré, “Du monde de l’à-peu-près à l’univers de la précision,” in Études d’histoire de la pensée philosophique (Paris: Gallimard 1971), 342.
10. Koyré, “Du monde de l’à-peu-près à l’univers de la précision,” 359–60.
11. Thomas Kuhn, “The Function of Measurement in Modern Physical Science,” in The Essential Tension: Selected Studies in Scientific Tradition and Change (Chicago: University of Chicago Press, 1977), 194.
12. P. H. Smith Jr., “The Computer and the Humanist,” in Computers in Humanistic Research: Readings and Perspectives, ed. Edmund A. Bowles (Englewood Cliffs, NJ: Prentice-Hall, Inc., 1967), 18.
13. Da, “The Computational Case against Computational Literary Studies,” 634.
14. Ezra Pound, “Date Line,” Literary Essays of Ezra Pound, ed. T. S. Eliot (New York: New Directions, 1968), 84.