From: "Terry Kerr" <terry@bizarsoftware.com.au>

Is there any particular reason that a Text Index will index an object by acquiring the value to index? Shouldn't it not index at all if the value doesn't exist on the object?

If the value is aquired, then it means that this object has that value, aquired. I find it logical that the indexing will behave like that too, and I'm pretty sure I have used it in some application.

I believe KeyWord Indexes have the same problem

Why is it a problem? It sounds like a feature to me. :-) If you have an attribute that you don't want to be aquired, I guess you have to make a computed attribute out of it.