0.10 Give us editors! re-inventing the edition and re-thinking the  (Page 9/18)

Some sources will always have greater appeal than others—Vergil will always attract more attention than your random nineteenth-century dissertation in Latin—but it also recognizes that much, indeed most, of our new interest in Greek and Latin may emerge from improved access to a rich body of sources that were physically accessible in only a few research libraries and archives. Even when sources outside of the traditional canon were physically available, their intellectual context and their idiosyncratic language made them intellectually inaccessible to all but those few readers who had in fact developed an advanced knowledge of Greek and Latin.

Students of Greek and Latin need to think about breadth as well as depth. We can treat our core authors as corpora on whom we can continue to lavish an extraordinary amount of labor. Many of the most heavily studied authors are relatively small: the Homeric Epics, Greek Drama, Catullus, Vergil, and Horace constitute approximately one million words. Authors such as Plato (600,000 words), Aristotle (1.1 million words), Cicero (1.1 million words), and Livy (570,000 words) are, however, considerably larger and do not lend themselves so readily to the same intensive methods that we can apply to the 35,000 words of Aeschylus. The TLL began work in 1894 on its ten million words from the most heavily studied Latin authors. With a staff of twenty Latinists, the TLL has completed approximately two-thirds of its task. As we move beyond the classical corpus into much larger collections where we lack the editions, commentaries, indices, specialized lexica and other scholarly infrastructure available for the heavily studied authors, we need different methods.

Digital philology and, indeed, much of modern editing, must depend upon two new disciplines with deceptively similar names, often lumped together but complementary in principle and separate in practice: computational and corpus linguistics.

We need computational methods as we confront tasks that overwhelm manual methods. This issue is more thoroughly explored in Bamman and Crane 2009. As we confront the challenge of editing billions of words, we need editors who can apply automated methods, measure the results by analyzing randomly sampled subsets of the data, and provide large bodies of textual data, of known accuracy, useful for most automated purposes and ready for others to refine, in whole or in part, as time allows and interest dictates.

But important as computation may be, the most important new discipline for classical—and any philology—is corpus linguistics. For a good overview of the relationship between corpus linguistics, computational linguistics and historical corpora, see Lüdeling and Zeldes 2008. All students of historical languages are, in some sense, corpus linguists, for they are studying corpora that are fixed—we may recover new papyri and inscriptions, but the surviving linguistic record, whether discovered or not, can never expand.

Editing in classics

Editors of classical texts have for the most part focused on the challenge of reconstructing an original copy text. All scribes make errors and even relatively small error rates add up in a single manuscript. As texts are copied, they accumulate new errors. In some cases, relatively large texts survive with relatively few obvious errors—the 600,000 words of Plato present relatively few problems. In other cases, much shorter texts can become garbled in transmission, especially when only a few or even simply one manuscript survive—the 35,000 words in the seven surviving plays of Aeschylus are notoriously problematic. When print allowed scholars to publish hundreds and thousands of identical copies, transmission stabilized and scholars set out to undo the damage. Classicists published thousands of editions from the editions principes of the late fifteenth century through the great systematic editions of the nineteenth and twentieth centuries. For more on the tradition of creating critical editions in classics see Bodard and Garcés 2008 and Monella 2008.