Thanks to a generous grant from the President's Office of the University of Massachusetts, the preliminary portions of the engine have been produced. Each interconnected module is designed to fulfill a need in the teaching or research of Old English. Because of the modularity of the engine, it can be applied to any language, including Latin and Greek. Current modules are 1) proximity search engine, and 2) on-line grammar exercises.
Module 1: Proximity Search
This module has been built in PERL to interact with a CGI and web interface. It scans files (or streams) built of single lines of Old English poetry. (There is a model in C which can be compiled for the Mac or for PC.)
Using grep patterns, users can search for single words, or search for words in proximity to one another. For example, one might want to find instances where a particular adjective modifies a particular noun ("god" and "cyning," for example) in a sentence irrespective of words or clauses interposed between them.
It also searches for single words or parts of words.
Module 2: Exercises
A second module is under development. Using the same technology as Module 1, this module will deliver sentences to a web page, leaving blanks where, for example, pronouns should be. Students fill in the blanks or choose from a pull-down list, and the engine checks the answer. The benefit of this approach is that students learn Old English grammar from actual and random Old English sentences. Hopefully, the module can be employed to search prose texts as well.
We are currently at work on a morphology engine, which plugs into the current configuration. It differs from standard parsing tools in a number of ways. At the moment, we are still in development.
**If you have any suggestions, please feel free to contact me.