Since the subject of processing progressive forms in corpora has come up
recently, I must say that I agree with Chris Tribble concerning the
application of Mike Scott's WordSmith. It has also helped me to establish
this type of POS tagging with "*V*G>*ING" as a search string. The
identification of such patterns in specific corpora says a lot in terms of
stylistic analysis, and that is highly valued when you come to face up to
language description for teaching purposes. I don't think there is a safer
method from this perspective to cling to empirical data (at least for us
language teachers not knowing programming).
All the best
Alejandro
This archive was generated by hypermail 2b29 : Wed Feb 23 2000 - 12:41:32 MET