Reading Machines: Toward and Algorithmic Criticism

5. ’Patacomputing

5 ’PATACOMPUTING A few years ago, Martin Mueller, the animating force behind the text analysis system WordHoard, decided to perform what we might call an experiment but would better be thought of as the fulfillment of a brief moment of curiosity. Using the system’s powerful word-counting and lemmatization features, Mueller was able to create lists of the most frequent words in Homer and Shakespeare: Homer Shakespeare man (ἀνήρ) lord ship (ναὗς) man god (θἐος) sir heart (θῦμός) love hand (χείρ) king son (υἱός) heart horse (ἳππος) eye father (πᾶτήρ) time word (ἒπος) hand companion (ἑταἷρος) father I call attention to the spur-of-the-moment character of his investigations, not to suggest that one could not conduct elaborate experiments involving word frequency (many have done so), but simply to point out the ways in which this operation has been made virtually effortless by digital technology. The creation of WordHoard required many hours of careful programming and design, but that effort is minimal in comparison to what was required of those who created the first dictionaries and concordances—or even, for that matter, the first books. If Alfred North Whitehead was correct, the result of such effortlessness represents not just the formation of a new convenience but also an epochal moment for civilization, which, he thought, “advances by adding to the number of important operations which we can perform without thinking about them” (42). 70 chapter 5 It is appropriate, of course, that we ask whether Mueller’s lists are the result of an “important operation.” They do not contain anything that one might call, at first glance, an astonishing result, and one immediately intuits facile lines of argument proceeding from them. Homer’s Iliad, after all, is a book “about ships” only in the way that the old bromide figures Moby Dick to be a book “about fishing.” It comes as a surprise to no one that God and man are subjects of concern to both Homer and Shakespeare. The lists seem “right,” but the knowledge we gain by their revelation might be thought of as confirming what we already know. For Mueller, it is both more and less than what we expect: “Given the fact that writers spend endless hours putting their words into the right order, it is disconcerting that a list of their most commonly used nouns will tell you quite a bit about what they are up to. This grossly reductive model of a text works much better than it should. . . . Homer is about ‘man, ship, and god’ in that order. Stop reading right there” (“Digital Shakespeare” 12). That we can arrive at a three-word précis of the Iliad “without thinking” seems of little moment if the result of such critical operations leads to “grossly reductive models.” In every way, the un-deformed text seems preferable to its atomized form. Still, it is unlikely that a human being, even if asked to name only the top three words in each text, would produce these lists precisely as the machine gives them to us. And for this reason, our explanations must assume the character of narrative. Perhaps “eye” figures prominently in Shakespeare because, as Maurice Charney observed, “Love enters through the eye” (122), with the result that romantic protagonists in Shakespeare are no longer masters of themselves but instead subject to forces not of their own making—forces different from, but related to, the forces that perturb the lives of Achilles, Hector, and Odysseus. One might also observe that the “heart” of Homer is not at all the “heart” of Shakespeare. Thumos is soul, vitality, spiritedness—the animus of the body. It is also the will to distinguish oneself—the horse that runs alongside eros in the Plato’s Phaedrus and that will appear centuries later in Evagrius’s description of the “passions.” At some point we must contend with the anomalies. Why is “hand” so prominent in both lists? “Love” and “time” seem right for a Renaissance dramatist, but the explanation of both words will require much more elaborate explanations. Mueller notes that “‘love’ and ‘time’ mark out important semantic domains of Shakespeare and indeed of ‘modern’ literature as it has been theorized since the Quarrel of the Ancients and Moderns” (“Digital Shakespeare” 13). But what are these semantic domains? Other algorithmic processes suggest themselves. WordHoard can unerr- ’Patacomputing 71 ingly (and instantaneously) locate every instance of “love” and “time” in both texts—even distinguishing the noun from the verb...