Reductive Reading: A Syntax of Victorian Moralizing by Sarah Allison (review)

To begin with a reductive reading of a book about the promises of reductive reading: Sarah Allison’s Reductive Reading: A Syntax of Victorian Moralizing is about how to read at the level of the sentence, the line, or the clause for features of the Victorian novel usually mapped at the level of theme, plot, or the developmental arc of character. In other words, what Allison calls reduction many readers of her work will recognize as the sheer closeness of very close reading. And yet the power of the book lies in precisely this defamiliarizing move, by which Allison shows us how digital methods usually associated with distance and data-crunching can also allow a kind of closeness sometimes unavailable to our ordinary reading practices. Allison’s method derives as much from Roland Barthes’s S/Z (1970), with its fine-grained sentence-by-sentence parsing of Honoré de Balzac, as it does from the field of computational literary analysis. I would encourage skeptics of the digital humanities above all to pick up Reductive Reading without delay. One of the great pleasures of reading Allison’s work for non-digital scholars who do not know their types from their tokens is that the method is laid out so persuasively, with an eye to the tight relationship between computation and close reading.

The book purposefully makes argument and method difficult to disentangle. Its central question as a work of nineteenth-century novel criticism has to do with “how style relates to moralizing.” The question is disarmingly spare, but for that very reason it is “impossibly complicated,” as Allison points out, demanding the kind of reductive reading not merely allowed for, but required by, computational analysis (3). In her introduction, Allison explains how her desire to read for style “at the level of the grammatical clause” requires a rigorous practice of reduction, one that is modeled most effectively by a methodological norm of computational literary analysis: “the practice of beginning each part of the study with a clear justification for reducing a body of texts to specific features.” Deciding on what kinds of syntactical form to use as a “proxy” for the style of a particular author—and indeed the style within which is embedded a particular mode of moralistic argument—is Allison’s challenge, and this problem itself can work as a kind of proxy for debates about close reading writ large (1). Do not all of us, computationally or otherwise, perform such selections and such reductions? What changes when we use a computer program like Microsoft Excel to assist us in our practice of cutting up a text into the parts we need, reassembling them to make our interpretations hum?

In chapter 1, Allison mounts a more extended defense of computational methods for literary study, turning more fully to the method debate. It is a cogent and persuasively argued defense of the idea that “reduction is valuable rhetorically and argumentatively,” and it makes an original argument about the ways in which polemic in particular—in a dazzling array of examples from George Eliot, Bruno Latour, Eve Sedgwick, and Audre Lorde—relies upon reductive reading in order to generate new and complex ideas (15). Chapter 2 then turns to an extended reading of the Victorian literary review as a periodical genre characterized by the kind of polemical simplification that Allison theorizes in chapter 1, “a reading practice that is designed to counteract the stylistic complexity of a text and thereby help readers see its morals more clearly” (39). Victorian reviewing shows us the ways in which Victorian readers might have unmoored moralistic argument from plot and character, understanding it as a stylistic feature that is obvious, inscribed for all to see at the level of the sentence.

In chapter 3, Allison isolates what she calls the “commentative clause,” “the ‘which’ clause that shifts a sentence from past to present tense,” as a pervasive feature of Eliot’s style (61). This chapter is...