As I read the research so far it suggests that sometimes it helps to
be explicit about the elements or dimensions on which future
occasions/events/problems might be similar. I can relate to your
point about physicists and mathematicians, but I don't think that the
argument here is in favor of the "purely" abstract - it is that maybe
abstractions help mediate transfer or direct attention to the aspects
that are the focus of instruction (as in Gelman re: fractions??).
I just realized this resonates with a long ago xlchc discussion of
genre - and whether we are doing folks a favor by not introducing them
to priveleged or arcane genres - or _was_ that part of the discussion?
I didn't get into it at the time.
Fritz Mosher