Contemporary physics: realism or constructivism?

Thanks to Allison for a surveying Kosso for us, to Liz for coming along, and to all for a rich discussion. The issue was physics as a description of what is or a way of making sense of what has been observed, one world or many (in Goodman's terms), historically Einstein or Bohr.

Allison/Liz/Kosso all make the important point that contemporary physics does NOT assert that everything is relative (special and general relativity) or that everything is indeterminate. It instead suggests that SOME things are relative (space, time) while others aren't (the speed of light), and that some things are indeterminate (momentum, position) while others aren't (Planck's constant). To put it differently, current physical descriptions suggest that some things that might have been regarded as perspective-free/frame independent aren't in fact so but offers in that category others that are. Moreover, Liz made the additional point that the concept of frame independence (invariance) plays an important role in contemporary physics. The notion that physical law is that which is frame-indendent bears a close relation to symmetry considerations which in turn can yield many observed physical laws (time invariance implies conversation of energy, translation invariance implies conversation of momentum). All of this is consistent with physics as a description of what is. Particularly if it is coupled with a "transcendental argument": one has to work hard to make sense of things in any other way.

On the flip side, it was noted, and agreed to by all, that contemporary physics does not and cannot establish frame independence in the absolute. It can only explore what is or is not invariant over particular changes in coordinate frame. The upshot is that Planck's constant and the speed of light, for example, are constant over a particular set of perspectives/coordinates frames but may not be over additional ones. In short, contemporary physics can provide a way of describing an expanding set of observations but provides no way to establish whether any aspects of that description will necessarily hold from all possible perspectives. We remain free to treat science either as a description of what is, or as a way of making sense of what has been observed that may prove to be one of many adequate descriptions.

An intriguing set of questions that follows from this is how does one discrimate among "worlds"? between realism and constructivism? Does this require a still more expansive coordinate frame/perspective, or can it be done "locally", by simply noticing differences and choosing? Does "realism" facilitate the doing of science, at least of normal science? Does it facilitate a positivist/rationalist/enlightenment program? What claims might be made for a more constructivist position (see Mica and me in last week's forum)?