Rene Thom gives some interesting arguments from mathematical topology
about properties of 3-space that are quite basic to our descriptions of
form and change processes, even quite abstract ones. (See _Structural
Stability and Morphogenesis_).
Finally, I do not believe that even the notion of a Latourian network is
not fundamentally a spatial topology. What Latour argues against, and I
agree with him, is a particular sort of spatial metaphor, a common
default notion of spatial systems as what I have called, appropriating a
mathematical metaphor via Thom, systems of co-dimension 3 in 3-space.
These are the nested spheres within bigger spheres view that we commonly
assume about all relations of scale in space. But Latour's nets are of
co-dimension 1 (but still in 3-space), so that points along a net are
close, but otherwise 3-space near point that are not on the net are
effectively 'far' (which contradicts codimension 3 intuitions). There
are also codim 2 possibilities. I will try to post something on this
soon, but people can look at my chapter in the Kirschner and Whitson new
volume, _Situated Cognition_, at the end of the chapter.
I do not go quite so far as Lakoff in privileging spatiality in our
semantics, but clearly there is something basic we need to understand
about its role.
JAY.
-- JAY L. LEMKE CITY UNIVERSITY OF NEW YORK JLLBC who-is-at CUNYVM.CUNY.EDU