accidentals and substantives

conjectural criticism

I've blogged a fair amount of my teaching the last couple of years, but have said far less about my research. In an effort to counterbalance that lopsidedness, I thought I'd post a formal abstract of my dissertation, which is nearly done. Three looong chapters were recently broken down into four. Here's how things now stand.

Conjectural Criticism: Computing Past and Future Texts
Kari Kraus, University of Rochester

Broadly conceived, my dissertation re-imagines the role of conjecture in textual criticism--the particular branch of literary studies in which I intervene--at a time when computers are increasingly pressed into service as tools of reconstruction and forecasting. The kinds of examples that are of interest to me occur in humanistic contexts: the computational modeling of the evolution of a literary text or the slow erosion of a statue or the descent of a natural or invented language. Conjectural criticism is thus concerned with issues of transmission, transformation, and prediction. It has ancient parallels in divination and modern parallels in the comparative methods of historical linguistics and evolutionary biology. It also stands in contradistinction to current practices of archival or documentary reading, which foreground the material specificity of texts. My project thus emphasizes the need for alternative ways of conceptualizing conjecture other than as a balm to help heal a "maimed" or "corrupted" text. The pathological metaphors long ago ceased to strike a chord in editorial theory. Instead, conjecture might be better thought of as a form of subjunctive knowledge, knowledge about what might have been or could be or almost was. The object of conjecture is notional rather than empirical; possible rather than demonstrable; counterfactual rather than real. This subjunctive mode, I contend, is not antithetical to the humanities, but central to it. Whether it is a student of the ancient Near East deciphering a fragmented cuneiform tablet or a musician speculatively completing Bach's unfinished final fugue or a literary scholar using advanced 3D computer modeling to virtually restore a badly damaged manuscript, the impulse in each instance--vital and paradoxical--is to go beyond purely documentary states of objects.

Chapter one, "The Beauty of Innuendoes," draws critically on work by Jerome McGann, Nelson Goodman, Stephen Wolfram, N. Katherine Hayles, and others to develop a computational model of textuality, one that better supports conjectural reasoning, as a counterweight to the pictorial model of textuality that now predominates. I understand "computation" broadly to mean the systematic manipulation of discrete units of information, which, in the case of language, entails graphemes (writing) or phonemes (speech). Computation is thus not a notion confined to electronic machines, but something much more general that has been practiced on clay tablets, papyrus, and paper for millennia. I propose that one of the rhetorical forms divinatio takes in computation is the selection control structure, a series of if/then branching structures that dictate the conditions under which a particular instruction or operation may be carried out. Nearly every programming language features an array of such conditionals, such as if/then, if/then/else, if/unless, if/unless/while, and so forth. These statements can be nested one inside another to create complex and subtle logic architectures. Such structures are a mainstay of software engineering, and without them computational modeling and prediction simply would not exist. Drawing analogies between manual and mechanical computation, ancient and modern conjecture, I illustrate how some of the earliest inscribed prophecies of the ancient Near East, whose influence on early textual scholarship of the Hebrew Bible is well established, make extensive use of the same conditional blocks to control the flow and "output" of the mantic code.

Chapter two, "The Word as Image," shows how the fortunes of conjecture are tied to the contest between words and images in textual criticism. I argue that is possible to draw a distinction between linguistic and pictorial textuality along the faultline of codicology and bibliography or, what amounts to almost the same thing, the editing of classical texts in manuscript and vernacular texts in print. The editing of Greek, Latin, and biblical literature to this day takes place within the fiercely linguistic (and as a corollary to that, conjectural) tradition of Richard Bentley and A. E. Housman; the editing of modern vernacular literature, within the material tradition first associated with Greg-Bowers, and more recently with figures like Jerome McGann and Randall McLeod. I thus narrate a version of mid-twentieth-century textual scholarship whose dominant theme is the progressive eviction of the conjectural from physical bibliography. My argument is that this eviction eventually gave rise to the paradigm of "unediting" in the last quarter of the twentieth-century. I try, in other words, to demonstrate a clear continuity between two editorial traditions often assumed to be antithetical, and to reappraise some of our core beliefs about the Greg-Bowers school of editing. From the moment New Bibliography made the semiotic shift from digital (or logical) to analogue (or physical) codes, the death of editing in any interventionist sense was a fait accompli.

Chapter three, "Image Reconstruction," examines some of the problems posed by documenting and comparing variants across different states of the "same" illustrative print and the problems of reconstructing text illustrations in antique manuscripts. To the degree that we assume texts to be images, we inherit these challenges. The text becomes semantically richer, but syntactically poorer, with implications for comparison, reproduction, and conjecture.

Chapter four, " 'Probably Arboreal in its Habits': Trees, Networks, and Conjectural Units," focuses on the parallel importance of the tree paradigm in textual criticism, historical linguistics, and evolutionary biology. In general, my themes are the interdependence and correspondence of computer-assisted methods for reconstructing relationships among texts, languages, and genomes. More than a century after the publication of Darwin's Tree of Life in On the Origin of Species, tree methodology remains at the center of some of the most ambitious and controversial conjectural programs of our time, including efforts to reconstruct macrofamilies of languages, such as Proto-Indo-European, and the Last Universal Common Ancestor (LUCA), a single-celled organism from which all life putatively sprang. At the outposts of the biological and linguistic sciences, conjecture is prospective as well as retrospective: synthetic biologists and language inventors are known to experiment with modeling future states of genomes and languages, respectively.

An important strand of the dissertation asks whether conjecture applies to the future as well as the past. Are there predictive traditions in the humanities? If so, what are they and where are they to be found? How does one "predict" a future version of a text? Why would one want to? I look to language inventors--a curious assortment of linguists, linguaphiles, poets, literary scholars, computer programmers, science fiction aficionados, and others--as one possible source for thinking proleptically about texts. Conlangers, as they are called on the internet, often project future versions of languages, into which they translate literary texts.

The intended audience for the work includes textual scholars, specialists in the digital humanities and new media, and others interested in visual literacy and the technology of the written word.

Copyright (C) 2006 Kari Kraus

Update: the following abstract more accurately summarizes the final version of the dissertation, which I defended on 28 July 2006:

In textual scholarship, conjecture is predicated on the idea that words are always signs of other words; the received text also harbors the once and future text. When David Erdman alters a line from William Blake's Jerusalem so that it reads "The Universal Conclave raged" instead of "The Universal Concave raged," he is practicing the art of conjecture. The mechanisms of inference that underlie this art and the properties of the sign systems that support it are the subjects of this dissertation. Adopting Nelson Goodman's terminology in Languages of Art, I refer to those systems that are most susceptible to conjectural permutation as "allographic." It is the discontinuous or digital syntax of an allographic medium that enables conjecture, comparison, reproduction, translation, and transformation. These are interdependent procedures, none of which can be understood apart from the others. By "conjecture," then, I mean the considered manipulation or processing of digital signs with the goal of either recovering a prior configuration or predicting a future or potential one. This dissertation was written with the conviction that a cogent theory of conjecture is a desideratum of textual studies.

Chapter one, "The Beauty of Innuendoes," draws critically on work by Jerome McGann, Nelson Goodman, Stephen Wolfram, Alan Liu, N. Katherine Hayles, and others to develop a computational model of textuality, one that better supports conjectural reasoning, as a counterweight to the pictorial model of textuality that now predominates. I understand "computation" broadly to mean the systematic manipulation of discrete units of information, which, in the case of language, entails the grammatical processing of strings rather than the mathematical calculation of numbers to create puns, anagrams, word ladders, and other word games. Computation is thus not a notion confined to electronic machines, but something much more general that has been practiced on clay tablets, papyrus, and paper for millennia. I propose that a textual scholar endeavoring to recover a prior version of a text, a diviner attempting to decipher an oracle by signs, and a poet exploiting the combinatorial play of language collectively draw on the same library of semiotic operations, which are amenable to algorithmic expression and computational simulation.

Chapter two, "The Word as Image: A Narrative of a Metaphor," shows how the fortunes of conjecture are tied to the contest between words and images in textual studies. I argue that is possible to draw a distinction between linguistic and pictorial textuality along the faultline of codicology and bibliography or, what amounts to almost the same thing, the editing of classical texts in manuscript and vernacular texts in print. The editing of Greek, Latin, and biblical literature to this day takes place within the fiercely linguistic (and as a corollary to that, conjectural) tradition of Richard Bentley and A. E. Housman; the editing of modern vernacular literature, within the material tradition first associated with Greg-Bowers, and more recently with figures like Jerome McGann and Randall McLeod. I thus narrate a version of mid-twentieth-century textual scholarship whose dominant theme is the progressive eviction of the conjectural from physical bibliography. My argument is that this eviction eventually gave rise to the paradigm of "unediting" in the last quarter of the twentieth-century. I try, in other words, to demonstrate a clear continuity between two editorial traditions often assumed to be antithetical, and to reappraise some of our core beliefs about the Greg-Bowers school of editing. From the moment New Bibliography made the semiotic shift from digital (or logical) to analogue (or physical) codes, the death of editing in any interventionist sense was a fait accompli.

Chapter three, " 'Probably Arboreal in its Habits': Trees, Networks, and Conjectural Units," focuses on the parallel importance of the tree paradigm in textual criticism, historical linguistics, and evolutionary biology. In general, my themes are the interdependence and correspondence of computer-assisted methods for reconstructing relationships among texts, languages, and genomes. More than a century after the publication of Darwin's Tree of Life in On the Origin of Species, tree methodology remains at the center of some of the most ambitious and controversial conjectural programs of our time, including efforts to reconstruct macrofamilies of languages, such as Proto-Indo-European, and the Last Universal Common Ancestor (LUCA), a single-celled organism from which all life putatively sprang. At the outposts of the biological and linguistic sciences, conjecture is prospective as well as retrospective: synthetic biologists and language inventors are known to experiment with modeling future states of genomes and languages, respectively.

The conclusion underscores a recurring theme of the dissertation, which is that we are unwittingly abdicating disciplinary knowledge by failing to recognize that textual criticism is no longer exclusively, perhaps even predominantly, the province of literary scholars. More and more, the study of textual transmission and its attendant issues is carried out under the sign of genetics, bioinformatics, information science, linguistics, and cognitive science. We cannot remain indifferent to these developments if we wish to advance the cause of textual scholarship and ensure that it continues to have a viable future in literature departments. We need a curriculum for textual studies that synthesizes the knowledge and expertise found in the diverse fields enumerated here. This dissertation, then, is an incitement to systematically venture beyond the conventional bounds of our discipline so that we might avail ourselves of the insights, methods, and tools of our scientific colleagues, applying them to traditional problems of textual descent and variation, as well as acquaint evolutionary biologists, historical linguists, and cognitive scientists with a branch of literary studies that is critically relevant to their own research programs.

An important strand of the dissertation asks whether conjecture applies to the future as well as the past. Are there predictive traditions in the humanities? If so, what are they and where are they to be found? How does one "predict" a future version of a text? Why would one want to? I look to language inventors--a curious assortment of linguists, linguaphiles, poets, literary scholars, computer programmers, science fiction aficionados, and others--as one possible source for thinking proleptically about texts. Conlangers, as they are called on the internet, often project future versions of languages, into which they translate literary texts.

The intended audience for the work includes textual scholars, specialists in the digital humanities and new media, and others interested in visual literacy, the technology of the written word, and the emerging field of biohumanities.

Posted by karik at January 7, 2006 12:21 PM | TrackBack