Obligatory mention of Parse::RecDescent. Not that I'd do it that way; the stop list pruning and destemming Kingpin pointed out is the way to go. I am evaluating search engines at the moment and the vast majority of the ones that claim to do 'natural language processing' or 'semantic recognition' are just doing this. And the ones that do more don't produce any better results. I was recently copied on a document that had been written in MS Word and exhibited no squiggly underlines. It was full of spelling and grammatical errors, however. I keep it as evidence of the general pointlessness of computerized English grammar parsers. -- Peter Scott Pacific Systems Design Technologies ==== Want to unsubscribe from Fun With Perl? Well, if you insist... ==== Send email to <fwp-request@technofile.org> with message _body_ ==== unsubscribe