Leopold and Molly Bloom live at 7 Eccles Street, Dublin. Why might that choice of street number matter? In Ulysses by Numbers, Eric Bulson gives us a few options. Perhaps Joyce chose the number seven because of its symbolism, with the days of the week and the biblical Creation story in mind. But the number also reflects the number of years Joyce spent writing Ulysses, or at least the number he tells the reader he spent doing so. Then again, we could use this number to map out places in the novel: a computer could bring together the addresses in Ulysses and create all sorts of graphs and data sets. Or perhaps we can read it as just one more detail in the intricate web that makes up a realist novel.

Bulson sets out with an ambitious goal: to combine the kind of numerological reading that has been around for millennia (think of the mystical importance attached to numbers in the Bible, or Dante, or Spenser, and readings that find hidden meanings in line numbers) with computational literary analysis (CLA), which treats literature as data, using computer programs to analyse style, syntax and genre. Bulson wants to make the graphs, but also to combine them with close reading, always remembering that he is studying a novel, not a spreadsheet. Ulysses, he argues, is uniquely positioned for such a marriage of opposites, because it “holds them both in suspension”: its numbers invite both numerological and computational analysis. But as he picks apart word counts, paragraph lengths and lexical density, readers may wonder if this tells us anything meaningful about the novel or its creator.

Bulson is certainly aware of the distinction between Joyce’s conscious counting – his meticulous planning of episodes, characters and correspondences – and the unconscious patterns, coincidences, miscounts and errors that populate Ulysses. But when the evidence is unclear on intentionality, he is inclined to see authorial decisions where they may not exist. Bulson counts 106 characters in episode 10, for example, but suspects that “getting to one hundred characters in a single episode would have been particularly meaningful for Joyce”. He continues, “Did he want 100 and end up with 106? … Was 106 the plan all along?” Bulson later acknowledges that the process threatens to become “an exercise in critical desire” rather than quantitative analysis, but offers no alternative to this line of questioning.

For all the discussion of “miscounts”, “preconscious forces” and even a “numerical unconscious”, psychoanalytic readings barely get a look-in, when they could represent an avenue for Bulson to connect the computational and the literary. They do make one appearance, in one of the strongest chapters, “One or How Many?”, in which Bulson asks how many characters there actually are in Ulysses, and discovers along the way the infinite gradations of characters Joyce crafted: human and animal, alive and dead, real and imagined. Gilles Deleuze and Félix Guattari’s essay “One or Several Wolves?” – which reads a numerical error by Freud as evidence of his inability to see the unconscious as multiple – is used by Bulson to illustrate the networks of characters contained within the minds of Stephen, Leopold and Molly by virtue of interior monologue. But rather than linger on these shades of personhood, or on the unconscious motivations behind Joyce’s own numerical errors, Bulson counts up and splits the characters into four distinct categories: “major”, “major minor”, “minor” and “minor minor”, a quantitative solution that elides the qualitative complexity he has already beautifully described.

Ulysses by Numbers sings when it does manage to combine quantitative and literary analysis, when it remembers the text is both a collection of data and a living, breathing world of its own. Bulson’s analysis of word counts, for example, leads him to think deeply about the functioning of time in Ulysses. If each episode describes a single hour, what happens when one hour takes 4,300 words and another takes 19,214? This kind of questioning showcases the potential of Bulson’s particular brand of quantitative analysis to open up new critical readings. And as he traverses character, style, book history, reception history and compositional history, his enthusiasm is infectious.

This is undoubtedly an important contribution in Joyce studies. Bulson is right that in the age of PDFs, Kindles and Ctrl+F, “the literary object has changed irrevocably”, and it may be the case that our methods of analysis must change, too. But I’m not sure we’ve arrived at the marriage of the computer and the critic just yet. The number on the door of 7 Eccles Street is all but forgotten when we smell the kidney burning inside.

Laura Hackett is assistant books editor of the Sunday Times

Browse the books from this week’s edition of the TLS at the TLS Shop