Saturday, February 27, 2010

Some of my summer faculty/student research this year
(if we get the funding)

Lexomic Analysis of ‘Winchester Vocabulary’ Texts

We propose to use Lexomic methods of analysis to try to determine relationships among prose texts from the Anglo-Saxon period. Lexomic methods, developed by Wheaton Professors Mark LeBlanc, Mike Kahn and Mike Drout, employ computer-based statistical techniques to find the structures within and the relationships between texts. The Lexomic Project at Wheaton has already published significant research on the interrelationships of Anglo-Saxon poems and the significance of divisions within poems. This summer’s project will focus on the next frontier: the large corpus of Old English prose texts, in particular those associated with the tenth-century Benedictine Reform.

The culture of tenth-century England was shaped in large part by the political, cultural and religious movement called the Benedictine Reform. A small group of monks led by Dunstan (eventually Archbishop of Canterbury) and Æthelwold (eventually Bishop of Winchester), assisted by a series of sympathetic kings, took over religious life in England. In the 1970s scholars Helmut Gneuss and Walter Hostetter identified the “Winchester Vocabulary,” a set of words that the Benedictine Reformers used in specific semantic contexts. Hostetter showed that the Winchester Vocabulary spread from Æthelwold’s Winchester to other major reform centers and eventually ended up influencing late Anglo-Saxon throughout England. In the late 1990s, Mechthild Gretsch demonstrated that the roots of the vocabulary came from Glastonbury in the 940s, when Æthelwold and Dunstan, in internal exile, spent years studying the work of the early Anglo-Saxon writer Aldhelm, in particular his de Virginitate. Gneuss, Hostetter, Gretsch and other members of the “Munich school” of Anglo-Saxonists identified multiple Winchester Vocabulary texts, but their methods are extremely time-consuming, somewhat subjective, and require many educated or even inspired guesses. Lexomic methods are not subject to these same constraints. The computer does not get tired or bored, and our screening methods can process many texts to look for subtle clues which we then follow up using traditional philological methods.

But there is a significant hurdle we must pass before we can use Lexomics to examine the full range of the Winchester vocabulary. Unlike the poetry, nearly all of the major prose texts exist in multiple copies. When the manuscripts were edited to produce the versions that are now used in electronic corpora like the Dictionary of Old English corpus, editors produced “best texts” by collating manuscript witnesses. Thus the electronic edition of an Anglo-Saxon prose text can often represent no single manuscript. Our lexomic methods are considerably hampered when they do not have the raw material of spelling and grammatical variation to work with. Thus the edited prose texts are difficult to use, as editors have squeezed our variability in order to make single, consistent texts.

Phoebe’s work this summer will be to mitigate this problem by converting the “best-text” electronic editions to multiple versions that are consistent with the manuscripts. To do this she will begin with the electronic file of a given text and the scholarly edition from which it is drawn. Then, using the apparatus criticus of the edition, she will modify the file to make it like one of the major manuscript witnesses using a set of mark-up conventions developed by Prof. Scott Kleinman of Cal State Northridge. Phoebe is familiar with mark-up conventions from Prof. LeBlanc’s “Computing for Poets” course, and she understands the medieval cultural context of the Winchester vocabulary texts from Prof. Drout’s “Medieval Literature” course. Once the files are marked up to be consistent with their original manuscript witnesses, Phoebe will work with Prof. Drout, Prof. LeBlanc and Prof. Kahn as part of the Lexomics Project. She will attend our weekly or twice-per-week meetings, suggest and run experiments, and help write up the final results in one or more jointly authored papers.

Student’s Statement ###

No comments: