Corpora: undergraduate seeks help on treebanks and transformation grammar ...

[Tony Perretta]

Hello all

Apologies for the urgent tone in the subject ...

Normally I would wait a few days before posting to a new list but my time is
severely constrained. I am finishing my final yr project on the subject of
statistical parsing. I am a computer scientist but my project was set by a
linguist and as such his interest in this project centres around Zellig
Harris's transformational grammar (on which he has published papers at Oxford
university). I have completed most of the project but have found very little
about Harris's grammar and statistical parsers.

Harris was Chomsky's teacher. I am sure that you are all aware that Chomsky
was to discredit the use of corpora but with the use of technology we are
increasingly seeing a convergence between corpus-based linguistics and
Chomskyan formal grammars.

My question is this: Is there a treebank and/or a statistical parser that is
implemented using a Harris style transformational grammar? Where can I find
out about the marriage of the two to ascertain whether this is even of value?

Any help from you experts out there would be greatly appreciated.

TIA

Tony Perretta