home *** CD-ROM | disk | FTP | other *** search
- Xref: sparky sci.physics:14725 sci.math:11524 sci.logic:1392
- Newsgroups: sci.physics,sci.math,sci.logic
- Path: sparky!uunet!haven.umd.edu!darwin.sura.net!wupost!usc!snorkelwacker.mit.edu!galois!riesz!jbaez
- From: jbaez@riesz.mit.edu (John C. Baez)
- Subject: Re: Computability of the universe
- Message-ID: <1992Sep15.221629.14794@galois.mit.edu>
- Sender: news@galois.mit.edu
- Nntp-Posting-Host: riesz
- Organization: MIT Department of Mathematics, Cambridge, MA
- References: <1992Sep10.183603.2357@sei.cmu.edu> <1992Sep11.013141.28378@galois.mit.edu> <1992Sep15.090238.186@nntp.nta.no>
- Date: Tue, 15 Sep 92 22:16:29 GMT
- Lines: 30
-
-
- Someone writes (I'm sorry not have the attribution, but you folks don't
- know what contorted convolutions I have to do these days to post news
- from here at UCR where nothing works):
-
- > The so-called "three body problem", computing the trajectory of
- >.three (yes only 3) bodies under mutual gravitational (Newton-style)
- >interaction, is not computable.
-
- [in classical mechanics].
-
- Do you have a reference? To be frank, I don't believe you; I think
- you are mixing up uncomputability with the nonexistence of a
- closed form solution. It's an old result, due to Poincare I believe,
- that this problem is not completely integrable. But that has
- practically nothing to do with uncomputability.
-
- Indeed, in a problem such as this a large amount of work goes into
- stating precisely what one means by computability. The point is that
- there are collisions, so one could only hope for a global solution
- on an open dense set of initial data at best. If this is the case
- one says one has "asymptotic completeness". Asymptotic completeness
- was an open problem for a long time in the classical n-body problem
- with inverse square force law. I forget if it has been solved. If it
- were, one would need to state precisely the sense in which this open
- set was or was not computable, and in what sense the solution was or was
- not computable for data in this open set.
-
- In a sense it doesn't matter since I showed the corresponding *quantum*
- problem has computable unitary time evolution. But it's interesting!
-