ï~~Bulletin of the American Society of Papyrologists 44 (2007) 245-247
Melissa M. Terras (Chapters 4, 5, and Appendix A co-authored with
Paul Robertson), Image to Interpretation: An Intelligent System to Aid
Historians in Reading the Vindolanda Texts. Oxford: Oxford University Press, 2006. xii + 252 pages. ISBN 978-0-19-920455-7.
In this unusual little book, papyrologists are put, as it were, under glass.
The general aims of the book are twofold: (1) to create a scientific (i.e. computer-understandable) account of how papyrologists read an ancient text, and (2)
to test in concept whether this activity can be emulated by a computer using
artificial intelligence techniques. The Vindolanda tablets comprise the sample
set, however, and thus more specific goals intrude, which align to but do not
exactly match the general aims: (1) to create an account of how papyrologists
deal with the Old Roman Cursive script as represented on the Vindolanda
tablets, and (2) to explore how automated decipherment of character strokes
could be used to assist an expert in reading these texts. Papyrologists familiar
with the difficulties of reading the Vindolanda tablets will see at once that the
holy grail here is to enable the reading of the texts that have proved intractable
even to the genius of Bowman and Thomas. In particular, the researchers wish
to tackle the difficulties of reading the "stylus tablets," those wax tablets whose
written remains are the faint scratches on the wood where the point of the stylus penetrated the wax in the writing. Only a very few of the 200 Vindolanda
stylus tablets have been read to date.
It must be said at once that the holy grail is not reached, nor is that claimed.
The book amounts, in effect, to a status report on an ongoing project of considerable complexity. The first part of the book, in which the reading and analysis
behaviors of the papyrologist are specified, is a typical product of communication studies, in which "knowledge elicitation" by a "knowledge engineer"
(the author) of three "domain experts" (papyrologists) using "Think Aloud
Protocols" results in a detailed description of how papyrologists go about the
trade. I wish I could say that the results are fascinating, but they are not. The
author in conclusion (p. 83) admits that "some of the findings regarding the
papyrology process may seem obvious" and adds - in a startling rhetorical
break from a steady diet of scientific modes of speech - the self-deprecating
remark that "the more difficult and deteriorated a document, the more the experts pay attention to the features of the document: who would have thought?"
Still, there is something of interest in having the trade subjected to this sort of
unfamiliar objectification and even quantification; and for those unfamiliar
with papyrology, there may be something of value in the details.