THE RHETORIC OF PERFORMATIVE MARKUP

In the classic account first proposed by DeRose et al. (1990) and subsequently developed by Renear et al. (1996) and finally by Renear (2001), text markup of the sort typically practiced by humanities computing scholars is a reflection of reality. It seeks to express observations about the nature of the text, rather than giving orders to a processor. Markup of this sort—labeled as “logical indicative” markup in the most recent formulation—is widely familiar in the scholarly community, instantiated
in markup languages like TEI, EpiDoc, and other
languages intended for the transcription and preservation of primary sources.
What we think of as logical indicative markup, however,
is almost never that simple. Although markup in the
indicative mood claims to advance a simple statement of fact (“This is a paragraph”), the actual intellectual activity being undertaken is often much more complex. Despite the early claims of the OHCO theorists that markup observes what text “really is”, it is only under tightly constrained circumstances that indicative markup truly makes something approaching factual observations concerning textual features. Within a given disciplinary community the identification of a paragraph or a line of verse may be uncontroversial (so that one would appear
to be quibbling if one paraphrased as “I believe this to
be a paragraph”). Within a digital library context the
encoding may be so slight that its claims about the text carry almost no information (so that marking something
with may mean only “this is a block of text”) and
hence no information with which it would be possible to
disagree. But if we broaden the context at all—using markup to communicate between disciplinary groups, or to describe more complex documents—we enter a very different and less factual terrain. For any early work where the modern generic distinctions are not yet solid, identifying a passage of text as, variously, a paragraph, a verse line, an epigraph, or some other less determinate segment is not an act of factual observation and correct
identification, but of strategic choice. The question
is “how does it make sense to describe this textual
feature?” rather than “what is this feature?” And implicit in the idea of “making sense” are qualifiers such as “for me”, “now”, “for my present purposes”, “here at this project”, “given my constraints”, and others that can readily be imagined. The choice of the phrase “making sense” is not casual here: the act of encoding is indeed an
act of making sense, creating conditions of intelligibility.
Renear, in his essay “The Descriptive/Procedural
Distinction is Flawed,” extends the earlier taxonomy of markup types by adding the dimension of “mood”, by which markup may be characterized as indicative,
imperative, or performative. Where indicative markup is the kind described above—making factual statements about the textual world—and imperative markup is
the kind that issues a command (for instance, to
formatting software), performative markup is a less familiar domain, which Renear identifies with authoring. As his phrase “markup that creates” suggests, this domain has to do with calling text into being and in particular with naming and effecting the structures through which that text expresses meaning. It is tempting to make a clean distinction between this kind of “authorial markup” and the more familiar indicative markup on the basis of the type of document concerned: authorial, performative
markup being what we use when we write new
documents, and ordinary indicative (or perhaps
“editorial”) markup being what we use when we transcribe
existing source material. However, having noted this
distinction we must immediately trouble it: first, because
these categories are often intermingled (for instance, annotations and commentary in a scholarly edition are “authored” in this sense). But more significantly, even with content that is not “new”, markup does not exist solely to name what is there, but also serves to express
views about it, and the expression of these views
constitutes an authorial act just as surely as the
generation of a sentence of commentary. Authorial markup brings a structure into being just as writing brings
words into being, and in some cases the two may be
isomorphic. Adding the TEI element amounts to the same thing, informationally speaking, as adding a note whose
content is “this sentence is unclear in the source; we
believe the reading to be X.” To extend Renear’s terms, can be either an indicative, editorial statement (“this passage
is unclear”) or a performative, authorial statement
(“I make this assertion about unclarity”, “I create this meaning with respect to the unclear reading”).
This authorial dimension to markup systems like the TEI is unfamiliar, little used, obscure. But it crucially amplifies our understanding of the rhetoric of markup,
and of the kinds of meaning it can carry. Most
importantly, it suggests that Jerome McGann’s assertion
that text markup cannot represent “the autopoietic
operations of textual fields—operations specifically
pertinent to the texts that interest humanities scholars” (2004, 202-3) reflects a very limited sense of the potential rhetorical operation of markup. In characterizing the TEI as “an allopoietic system” which “defines what it marks...as objective”, McGann draws on markup’s own conventionalized account of itself. This account,
which as we have seen locates systems like the TEI
firmly within the indicative realm, deals solely with the editorial rhetoric of statement and description—not with
interpretation and certainly not with performance. It ignores the extent to which even this indicative markup can make statements which are not simply factual: which represent local knowledge, perspective, contingency,
belief, positionality, uncertainty, purposiveness, and even
deception.
Most importantly, it ignores the authorial quadrant of
Renear’s grid: the space of performative logical markup, in which an author brings meaning into existence either by creating new marked content, or by adding markup to an existing text and performing upon it a new set of
meanings. This latter case would in fact resemble
performative instruments like the Ivanhoe Game, which represent for McGann the archetypal scholarly textual
activity: a performative apparatus, in effect, through which scholars express interventions in a textual field: “readings”, commentary, textual engagements that
inflect the object text rather than simply standing apart
from it. The extended version of this paper will expand on this point, exploring how performative or authorial markup might enact the kinds of textual engagements
that McGann calls for as constitutive of humanistic
textual study.
McGann is correct in identifying the predominant use of markup systems like TEI as “coding systems for storing and accessing records” (202). But this predominant use does not define the limits of capability for such markup systems, let alone for text markup in general. Our choice to use markup in this way derives from the collective sense, within the humanities disciplines, that this is what
markup should be for: a technological tool external
to ourselves, rather than an expressive medium. In
expecting, as McGann does, that markup cannot be made of the same stuff as poetry, we create a self-fulfilling
prophecy. In fact, we make markup in our own image—in the image of our own fears.
References
DeRose, Steven, J., David Durand, Elli Mylonas and Allen H. Renear. “What is Text, Really?,” Journal of Computing in Higher Education. 1:2 (1990).
Liu, Alan. “Transcendental Data: Toward a Cultural
History and Aesthetics of the New Encoded Discourse.”
Critical Inquiry 31:1 (Autumn 2004), 49-84.
McGann, Jerome. “Marking Texts of Many
Dimensions” in A Companion to Digital Humanities, ed. Susan Schreibman, Ray Siemens, and John Unsworth. Blackwells, 2004.
Renear, Allen. “The Descriptive/Procedural Distinction is Flawed.” Markup Languages: Theory and Practice
2.4 (2001): 411-420.
Renear, Allen, Elli Mylonas, and David Durand.
“Refining our Notion of What Text Really Is: The Problem of Overlapping Hierarchies.” Research in Humanities Computing, ed. Nancy Ide and Susan Hockey. Oxford University Press, 1996.