In this sense, it seems to me both true that mediational means transform
affordances and that they afford. In other words, the unit of analysis is
something more like a notion of functional systems as dynamic, fluid,
distributed, and situated constructions (I'm thinking here particularly of
Hutchin's use of functional systems in _Cognition in the Wild_). In this
sense then, you can't *really* separate out the activity setting from the
persons from the tools, although our language and our need to analyze
activity may encourage or require us to do so.
------
I recommend Ed Hutchin's book-- and the symposium about it
in MCA-- most highly for discussion of these issues. However,
its interesting the Ed does not mention Gibson and I do noit
recall him using the notion of affordances.
mike