If you want to make the stems dynamic (i.e., you don't want to define them before-hand), you could do some looping to determine for each word if it is a substring of another word (use the Index function) and there are no cases where the word is a substring of another word. Also, use the Lower function first so you don't have to worry about case. So, "manage" would be a substring of "manager" and "managed", but never a 'superstring' of any other words. You'd therefore classify "manage" as one of your stems. Repeat for each word. This would be a two-level loop (quadratic), so it might be slow. Sorry, I don't have time to work out the syntax right now.
Now, if you wanted to have "appreciat" as the stem for both "appreciated" and "appreciating", this would be even more complex because "appreciat" is not a word in the corpus. You could do another loop in which you look at all possible substrings of each given word to see if they are in other words, but this would be very cumbersome and inefficient. But, since its in the load script, if you don't mind a long run time (if there are a lot of words), you could give it a try.
.png "jonvitale"){kind=avatar}
