Project Goals

Many readers encountering Virginia Woolf’s To the Lighthouse for the first time find it a challenging, “difficult” novel. But what makes it “difficult”? It is not the language or the diction: the novel uses mostly plain, simple, everyday words. It is not the subject matter: it is about a family and their friends, about everyday events like knitting socks, about the death of a mother. It is not the range of references: unlike other celebrated modernist works like T. S. Eliot’s The Waste Land or James Joyce’s Ulysses, To the Lighthouse makes few direct references to other works, and generally makes very modest demands on its reader’s knowledge of literature, history, or geography.

The source of difficulty in To the Lighthouse is something else: it is simply that, in a novel that takes the reader into the intimate thoughts of so many characters and seldom explicitly marks the transition between the thoughts of one character and another, it is often very hard to know who is speaking.

The technical reason for this is Woolf’s employment throughout the novel of the narrative technique free indirect discourse (FID) — a method for blurring the boundary between character and narrator. Because Woolf uses FID so widely (it is employed in as much as 90% of the novel) and because she uses it to introduce the words of so many different characters, there is very seldom a “correct” answer to that vexing question, “Who is speaking?”

Questions without straightforward answers, however, are often the most intriguing ones. This website is devoted to exploring the question of voice in To the Lighthouse, fully aware that final, definite answers are not only impossible, but decidedly undesirable. Its overall goal is twofold: to engage you in the process of trying to untangle the complex vocal knots of To the Lighthouse, and to help you to see that, however attached you may be to your own interpretation, countless other, equally valid readings of the text are also possible. Inhabiting this paradoxical readerly position — engaged, active, yet humbly resolved to uncertainty — is, we believe, something like the “key” to being a good reader of Woolf — and perhaps of modernist literature generally.

Scholarly interventions

Besides the broad goal of helping readers better understand To the Lighthouse and modernist literature, our work intervenes in academic debates in several areas.

Its contribution to the field of modernist studies is to insist, using a combination of qualitative and quantitative methods, on the crucial importance of dialogism to any understanding of modernist literature. To our minds, modernism is dialogism — dialogism understood not just as a stylistic feature, but as one of modernist literature’s most important political interventions. For more on this, see our essay “FID and the Politics of Modernist Dialogism.”

We propose to the field of literary criticism a method for “crowdsourcing literary theory.” While literary theory is traditionally premised on the a priori, anecdotal speculations of individual critics, this project seeks answers to a fundamental literary question by looking inductively at how readers actually respond to a literary text. While our essay “What is Free Indirect Discourse?” offers a series of traditional definitions of the device in order to get you on your way, we are also interested in working in the other direction: in generating a definition of FID based on the 320 student annotations of To the Lighthouse gathered in the “What the Class Said” section of this website, as well as on and those submitted via the “Have Your Own Say” section of this website. Pursuing this method has already yielded some interesting results. For more on this, see the “What the Computer Said” section of the website.

Our work also contributes to the “mainstreaming” of Digital Humanities approaches in literary research. In the field of English literature, many so-called “traditional humanists” remain skeptical of the value of applying systematic, rigorous, unambiguous computational methodologies to the analysis of something as deeply, necessarily, and proudly ambiguous as literature. Building on Stephen Ramsay’s idea that “A scientific literary criticism would cease to be criticism”(489) — because “criticism,” the task of humanists, doesn’t seek answers — our work aims to use digital approaches not to resolve literary cruxes but rather to highlight and describe “cruxiness.” Willard McCarty has argued that computational “modeling problematizes” (26); following McCarty, our goal is to problematize modernist literature with digital methods. Rather than telling you the truth about To the Lighthouse, we seek to provide you with the tools and resources to explore for yourselves its knottedness, its cruxiness, and its irresolvability — without seeking to “disambiguate” this complexity in any definite direction. For more on this, see the paper we presented at HASTAC 2013, “Problematizing Literature with Digital Methods.” (Slides available here.)

Our contribution to Computational Linguistics (a subfield of Computer Science) follows from the above. Computational Linguistics is still at the early stages of engaging seriously with literature. (We are happy to be involved in this process, having participated at the two inaugural Computational Linguistics for Literature (CLfL) workshops at the NAACL-HLT in 2012 and 2013). The field continues almost uniformly to treat ambiguity as a problem to be resolved; its focus remains on disambiguation, with the assumption that one true, correct interpretation exists. Such an approach would have little use in analyzing To the Lighthouse, a deeply ambiguous text in which “correct” interpretations are in very short supply. Our effort to develop an algorithmic method of detecting FID has proceeded by challenging many such Computational Linguistics assumptions. FID an inherently unstable, uncertain device — so that our computational task is paradoxically to identify uncertainty. Our work introduces a further level of ambiguity by seeking a algorithmic understanding of FID based in part on the messy, conflicting annotations of human readers. For more on this, see our CLfL position paper, “A Tale of Two Cultures: Bringing Literary Analysis and Computational Linguistics Together.”

Works Cited

McCarty, Willard. Humanities Computing. New York: Palgrave Macmillan, 2005.

Ramsay, Stephen. “Algorithmic Criticism.” A Companion to Digital Literary Studies. Eds. Susan Schreibman and Ray Siemens. Oxford: Blackwell, 2008.