I do the most traditional of humanities work, transcribe texts for the purpose of editing them, and yet I find that digital humanities is less interesting as an idea about algorithmic criticism than about the very matter that Borges’s “Pierre Menard, Author of the Quixote” has as its root concern, the radical instability of texts even when they are (in the sense Marche would have it) the same text.
But there are many ironies herein. Stephen Marche in his recent shot across the bow of digital humanities (“Literature is not Data: Against Digital Humanities”) appears to take his text of “Pierre Menard…” from Labyrinths, which has the translation by James E. Irby instead of, say, the translation by Robert Bonner in Ficciones. Of course, since Marche is publishing in the L.A. Review of Books rather than in a scholarly journal, he in his context will exclude a bibliographical citation in which he would specify the source of his quotation.
That Marche used Bonner’s translation can be ruled out because Bonner transcribes Borges’s Spanish—presumably Bonner seeks to assure the reader that his identical translations of the two passages (Cervantez’s and Menard’s) mirror Borges’s two identical Spanish versions. These are just details, but then one point of Marche’s post is that context matters when details are the same.
But if Marche is quoting Irby’s translation, as I believe he is, then why did Marche not honor Irby’s choice to print each quotation from Borges as a separate block paragraph? And why did he omit the ellipses? More details, I suppose. But it may be that Marche or his editor do not think the ellipses and block paragraphs in Irby’s translation are important.
Perhaps the choice is related to the appearance of this review article in digital format.
Might it be that block quotes within block quotes do not render well in the digital environment?
Maybe the ellipses were inadvertently omitted, or maybe the the ellipses in the default web font are considered distracting by Marche or the editors. And so some human decided not to honor the block quotes in the source. Would Marche continue to hold that the “text” is still the same—format and ellipses be damned? Or could inaccurate quoting be the first step in the digital fall? I hold out hope that it’s another instance of a delicious Borgesian twist, that Marche or his editor has mindlessly altered Bonner’s translation to more closely match Irby’s—it’s that fascistic demon tradition having its way with Marche’s attempt to transcribe. Or maybe there’s another ineffable digital text that is Marche’s Borgesian source?
It would be too delicious were the latter so, and so I will suppose that it is not. But it’s so not because such details don’t matters or because context is everything. Context is only the first thing missing from Marche’s discussion of digital humanities. The subjects of humanities studies have been born digital works of literary art and games (see the work of MITH), are being archived in digital format (see any research library or scholarly archive), and are being produced in digital format as first-generation material (see almost every present-day author or scholar). Let us even take up Marche’s most frightening example, that a handful of scholars wish to (and are capable of) applying algorithms to collections of digital texts and thus deriving readings. Marche says, in his most laughably ambitious claim, that “algorithms, exactly like fascism, work perfectly, with a sense of seemingly unstoppable inevitability, right up until the point they don’t.”
That’s just ridiculous: no scholar who wishes to remain a scholar believes that algorithms work perfectly. Scholars may be susceptible to fascism, but algorithms are not the reason. Marche then exhibits his own unexamined belief, that texts are just there whereas meaning is handmade. This naive belief is as dangerous a matter as an algorithm. Bruce Manning Metzger in The Text of the New Testament (1964) has as such an example Erasmus’s edition of the Greek New Testament, which excited the ire of a competing editor, Stunica, who participated in the preparation of the Complutensian Polyglot. Stunica’s complaint was that Erasmus’s text lacked the
Trinitarian statement concerning ‘the father, the Word, and the Holy Ghost: and these three are one. And there are three that bear witness in earth’ (1 John v. 7-8, King James version). Erasmus replied that he had not found any Greek manuscript containing these words, though he had in the meanwhile examined several others besides those on which he relied when first preparing his text. In an unguarded moment Erasmus promised that he would insert the Comma Johanneum, as it is called, in future editions if a single Greek manuscript could be found that contained the passage. At length such a copy was found—or was made to order!”—apparently by a Franciscan friar who prepared a new-old Greek manuscript on the basis of the Latin Vulgate (101).
Church authority trumped transcription and manuscript provenance, and would seem to suggest that such things may matter, yet Marche continues to espouse that present-day pipe dream, the “complete, instantly accessible, professionally verified and explicated, free global library.” We can only hope that Marche invokes this utterly ridiculous prospect in jest as no less a Borgesian dream, with its ever-present teasing wink, than his “handmade insight.”
That’s all now. I must go back to mulling over a space in a printing of Stowe’s Uncle Tom’s Cabin or a handwritten letter by Walt Whitman’s mother and try to decide whether I should transcribe the space in a digital representation so as to separate one word from another. And the word count, should it be 1 and then 2 words? or just 1? True, an algorithm may lead someone into faulty interpretation. But it won’t be the first time—or the last.