In this (traditional) presentation I will discuss Ramsay’s algorithmic analysis of Virginia Woolf’s novel, The Waves. In so doing, I will argue that traditional literary criticism is not so different from scientific inquiry. While critics like Jonathan Gottschall assert that “literary studies should become more like the sciences”, both disciplines already revolve around the ‘primacy of pattern’: what Stephen Ramsay identifies as “the basic hermeneutical function that unites art, science, and criticism” (4; xi). When we analyze a text, we search for frequencies: whether of themes, tropes, or individual words. In a critical paper, we marshal those frequencies as evidence of overarching trends or patterns—extrapolating about the use of words in a text, and moving from individual citations to “the grander rhetorical formations that constitute critical reading” (Ramsay 17). Pattern-making, in other words, is where paper-based literary criticism and computational analysis overlap, since “critical reading practices already contain elements of the algorithmic” (Ramsay 16).
As literary critics, our brains are finely honed text analysis tools: “We read out of order, we translate and paraphrase, we look only at certain words or certain constellations” (Ramsay 48). By adopting a theoretical standpoint, we select a set of meanings from a field of possible meanings, just like an algorithm. As such, literary criticism is algorithmic in that it is a process by which we apply certain limits or constraints to our analyses. Where the paper-based literary critic searches manually for water imagery in The Waves, the computer-based critic uses an algorithm capable of identifying every instance of the word “water” in Woolf’s novel. With the application of text analysis tools, tables of frequencies, charts, and graphs generated by the algorithm replace marginal notes generated by the paper-based literary critic.
Thus computer-based criticism is not so different from paper-based criticism in that both involve the identification of patterns and the application of rules or constraints. For all intents and purposes, the computer just speeds up the process: it “revolutionalizes, not because it proposes an alternative to the basic hermeneutic procedure, but because it reimagines that procedure at new scales, with new speeds, and among new sets of conditions” (Ramsay 31). Using TAPoR to find water imagery in The Waves is basically the same thing as using a concordance, but “at a different scale with expanded powers of observation” (Ramsay 17).
Yet if computers can answer all of our questions, what makes literary studies ‘special’ or distinct from scientific inquiry? If criticism is, by nature, ambiguous, how do we quantify our qualitative evidence? Should we? The problem is with analyses like Miriam Wallace’s:
There is no experiment that can verify the idea that Woolf’s ‘playful formal style’ reformulates subjectivity or that her ‘elision of corporeal materiality’ exceeds the dominant Western subject. There is no control group that can contain ‘current feminist reconfigurations.’ (Ramsay 7)
Algorithmic criticism, then, cannot turn literary criticism into a science—nor should it seek to do so. Instead, it advocates the use of scientific method to unleash the latent potentialities of a text, and takes already existing hermeneutic practices to new heights and speeds. Text analysis tools like Voyeur, HyperPo, Wordhoard, MONK, and TAPoR are no different than psychoanalytic or Marxist approaches: they allow for alternate perspectives on canonical works, and enable us to test old theories and new hypotheses (they can also reveal a lot about your own writing; try plugging your blog summary into HyperPo). By running Woolf’s novel through TAPoR, we are not trying to solve Woolf: we are “trying to ensure that discussion of The Waves continues” (Ramsay 15).
The Discussion, continued:
❧ Is art quantifiable? Does algorithmic criticism implicitly favour quantitative over qualitative evidence and quantity over quality?
❧ Is there a risk in non-reading and will this trigger a “slow-reading” movement?
❧ If you could design a text analysis tool to answer any question about any literary work, what would it be?