Lars Henrik Mathiesen wrote: > I'm all for building elegant, concise, parsimonious and beautiful > theories. And if you're doing mathematics, that's the end of it. > But if someone goes on to let their theory guide their analysis of > further data, and then claim that that analysis represents some sort > of fact, they tend to lose my sympathy and interest. I understand what you're trying to say, but I don't really see what it would mean to have an analysis which was not guided by theory. It's literally impossible to do any kind of analysis without making theoretical assumptions, however mundane. As I remarked in a previous post, even such humdrum linguistic concepts as "word", "syllable", "consonant", "prefix", "verb", and "past tense" are theoretical constructs. Matt.