Program to count word-repetitions?

[Eric Van Tassel 101233.342@xxxxxxxx]

One of my major vices as a writer is the over-use of certain favorite
words, and I was once delighted to see my editor employ a program to help
me avoid this pitfall.

A couple of years back, when I wrote a piece for the N.Y. Times (Sunday
Arts & Leisure), the article was set using a program that seemed to be
XyWrite-related. When I saw first proofs of the article, each place where I
had re-used the same word was flagged, with a code indicating the number of
times that word had been used.

Does anyone know how it's done? Is there an XPL routine I can use, or
adapt, to perform this function?

It would be especially helpful, though not absolutely essential, if the
program linked up uses of the same word in different parts of speech (e.g.
"find/found"). Perhaps this could be dealt with (albeit tediously) if I
compiled, over time, my own dictionary of words and expressions that I have
a particular weakness for. If the program used such a dictionary, it might
also employ (though at this moment I don't see how I would write such a
routine) cross-links between syntactical variants of the same root word.
But such a refinement would be a luxury: it would be nice if I could just
catch the crudest direct repetitions.

Cheers
Eric Van Tassel


PS: I've just seen a possible way to simplify the creation of the
dictionary of my own particular "dangerous" words. Even if the program
isn't sensitive to syntactical variants, perhaps it could be asked to save
a list of all the words that it has had to alert me to in a particular job.
I could merge the word lists from several jobs, to build the core of my
personal pitfall dictionary; then all I'd have to do would be to
(presumably manually) add variant parts of speech.
 But (as Richard III says) I'm running before my horse to market: first I
need the program itself.