From Irene Eckstrand, Jan 21, 1999

OK, here are some thoughts.  Like you, I am surprised that it takes fewer words
to describe a blurry picture.  I am interested that so many of us resorted to
metaphor when it became clear that one could not precisely describe a
phenomenon.  Are there analogues to this in science?  Are our models the
scientific equivalent of metaphors?

An exercise we often do in science education is to put people in groups, each
with a bowl of peanuts.  Each person writes a description of one particular
peanut and puts it back in the bowl.  Then the groups trade bowls and
descriptions.  The groups then try to find each described peanut.  What they
usually find is that the more quantitative the description, the easier to find
the peanut.  On the other hand, someone who can draw a good peanut picture is
also pretty successful.  I don't know that metaphors have ever worked well.

I hope you will have some undergraduates actually try to draw the pictures from
the descriptions.  That would be a very useful exercise in determining if
metaphor is helpful in describing blurry phenomena - and if so, what kinds of
metaphors work well.
--------------------------------------------------------------------------------