Bright's Anglo-Saxon Reader

Available here is the entirety of Bright's Anglo-Saxon Reader (1912).

• The entire book is available as scanned TIFF and PNG images.
• The glossary is available as text in both HTML format and an unvalidated approximation of XML. The glossary comprises pages 241-385 of the text.
• July 2004: Keith Briggs typeset the glossary using T_EX; the result looks just great! The PDF file is available here.

About the glossary

This glossary was one of my earlier projects. I tried performing OCR on the glossary, but the results were so degraded that I considered them to be unusable (I would probably have better luck now with better know-how and better software).

Since OCR did not seem to be an option, I decided to recruit volunteers to type the text. I wrote to the ANSAX-L mailing list and then snail mailed xeroxed pages to the volunteers. Then I regularized the character encoding and reassambled the submissions into a single online document. This strategy worked splendidly. A hearty thanks to these volunteers for making this reference available to all!

The original text distinguished e-with-hook from plain e, and o-with-hook from plain o. I told the volunteers to ignore this etymological distinction and just type e and o. Likewise, the original glossary contained page references back to the main text; since it would be a lot of work to work to key, I told the volunteers to skip the page numbers. My thinking has evolved as I've gotten more experience at these things; I now try to represent every bit of information which was present in the original, because it can often come in handy in unanticipated ways. Still, this glossary should be useful for most purposes.

After reassembling the typed text, I wrote a Perl script to parse the glossary and output an XML version to show the structure of the document. For convenience, there is also an HTML version derived directly from the XML version.

What I haven't done yet is to perform any kind of systematic automated check on the contents of the fields in the tagged glossary. For example, I need to make a histogram of all the words occurring within <tags> ... </tags> tags, i.e. part of speech abbreviations; typos and mis-parsed elements will be readily apparent in such a list. Spell-checking on the definition fields will probably shake out some further infelicities, etc. Ultimately, all of the cited Old English forms will need to be automatically checked against a corpus of Old English to confirm that the reported forms do in fact exist.

None of these steps have been taken yet (I can't say how soon I'll get to them); it is therefore nearly certain that there are undetected errors in this glossary. Please report any problems you find.

Revision history

1 March 2003: Removed an extraneous control character from the text version of the glossary. Thanks to Steven Krise for pointing out this glitch.

Abbreviations in the glossary

WI, WII, etc. are weak verb classes. 1-7 are strong verb classes.

Germanic Lexicon Project main page

Contact