Thursday, 22 September 2016

blogging question #1: on representation

Our upcoming sequence of classes on XML and TEI will lead us into the topic of using digital technologies to create representations of existing artifacts (like digitized books), as distinct from born-digital artifacts like video games and hypertext novels. This week's blogging question is designed to get us thinking about representation, digital technologies, and what's at stake in their relationship.

At the beginning of one of our readings for this coming week, Michael Sperberg-McQueen starts with a counter-intuitive claim: "Texts cannot be put into computers. Neither can numbers. ... What computers process are representations of data" (p. 34). This helpful reminder serves to point out the paradox of the term digitization: when we say we're digitizing a book, we're not actually doing anything to the original book (usually; there are exceptions). What we're really doing when we digitize is to create a new, digital representation of the original. Yet the English word digitization, and its grammatical form of an action (making digital) performed on an object (something not digital), can lead us to forget the act of representation that underlies all digitization.

Why is this important? Well, Sperberg-McQueen's answer is that "representations are inevitably partial, never distinterested; inevitably they reveal their authors' conscious and unconscious judgments and biases. Representations obscure what they do not reveal, and without them nothing can be revealed at all" (p. 34). This line of argument leads to a deceptively simple consequence for everyone involved in digitization: "In designing representations of texts inside computers, one must seek to reveal what is relevant, and obscure only what one thinks is negligible" (p. 34). All digitizations, being representations, are choices -- so we'd better learn how to make good ones. That's why Mats Dahlström and his co-authors make a distinction between mass digitization and critical digitization in one of our upcoming recommended readings.

This week's blogging question starts by asking you to find an example that helps us think critically about digitization. Can you think of some specific instance of digitization -- it could be anything: an image, an ebook, digital music, you name it -- where an originally non-digital object or artifact, broadly defined, has been digitized in ways that reveal interesting (or controversial, or funny, or illuminating) representational choices. I'm not asking for examples simply of digitization getting something wrong, as fun as those may be. Rather, I'm asking you to unpack examples where a choice made in digital representation illuminates some quality of the original thing that we might otherwise take for granted, or some revealing aspect of digitization itself -- or possibly both. Your example might arise from digitization gone wrong somehow, but I'd like us to look beyond basic error-identification for this question.

The next question, then, is this: what does the error—or simply the choice—in representation teach us about the original or about the act of representation itself? 

Digitized books are good places to explore this question, but you could draw on other kinds of media and other kinds of texts (in D.F. McKenzie's broad sense of the word text; see our recommended reading from last week titled "The Broken Phiall: Non-Book Texts."). For example, if you bought the Beatles record Sgt. Pepper's Lonely Hearts Club Band on vinyl LP when it was first released in 1967, you'd experience it in at least a couple of different ways than if you bought it on iTunes today, or on CD in 1995. For one, an LP listener would need to flip the record over partway through, which may or may not give the impression of the whole album being divided into a 2-part thematic structure: some bands exploited this imposed division of records into Sides 1 & 2, but not all did. More to the point, an LP listener reaching the very end of the record, in which the song "A Day in the Life" ends on a long E-major chord that would just keep on resonating in a continuous loop until one lifted the needle from the record's run-out groove. A CD track or MP3 file can't (or simply doesn't) do this. What is the representational choice here, and why does it matter? I'd offer the answer that the original design of the Sgt. Pepper LP involves the listener bodily in the music, in that "A Day in the Life" only ends when you chose to lean over and stop the record. That effect is lost in the digitized version of the album -- or is it replaced by different effect that influences how we'd interpret the song? (I like to imagine that somewhere in the great beyond David Bowie and Prince are having this conversation with John Lennon and George Harrison, while Jimi Hendrix and Lemmy are playing air-hockey nearby...)

This might not seem to have much to do with books, but being able to unpack this kind of representational choice, in which form and meaning become intertwined, is exactly what bibliographers and other textual scholars do -- not to mention text encoders who are concerned with critical digitization, not just mass digitization. Your example need not be as involved as the one I've spun out above: the point is to get us thinking about how representation works, and what's at stake.


Dahlström, Mats, Joacim Hansson, Ulrika Kjellman. "'As We May Digitize' -- Institutions and Documents Reconfigured."Liber Quarterly 21.3-4 (2012): 455-74.

McKenzie, D.F. "The Broken Phiall: Non-Book Texts." In Bibliography and the Sociology of Texts, 31-54. Cambridge University Press, 1999. []

Sperberg-McQueen, C.M. "Text in the Electronic Age: Textual Study and Text Encoding, with Examples from Medieval Texts." Literary and Linguistic Computing 6, no. 1 (1991): 34-46. []