home *** CD-ROM | disk | FTP | other *** search
- Xref: sparky sci.physics:21035 sci.math:16807
- Newsgroups: sci.physics,sci.math
- Path: sparky!uunet!enterpoop.mit.edu!snorkelwacker.mit.edu!galois!riesz!jbaez
- From: jbaez@riesz.mit.edu (John C. Baez)
- Subject: Re: Why Wave Functions aren't Physical (Was: Re: QM non-causal?)
- Message-ID: <1992Dec13.204233.10835@galois.mit.edu>
- Sender: news@galois.mit.edu
- Nntp-Posting-Host: riesz
- Organization: MIT Department of Mathematics, Cambridge, MA
- References: <1992Dec8.180855.22727@newshost.lanl.gov> <1992Dec10.192451.9924@lmpsbbs.comm.mot.com> <1992Dec11.220139.14066@newshost.lanl.gov>
- Date: Sun, 13 Dec 92 20:42:33 GMT
- Lines: 54
-
- In article <1992Dec11.220139.14066@newshost.lanl.gov> jlg@cochiti.lanl.gov (J. Giles) writes:
- >In article <1992Dec10.192451.9924@lmpsbbs.comm.mot.com>, bhv@areaplg2.corp.mot.com (Bronis Vidugiris) writes:
- >|> [...]
- >|> P2: Mathematically, if some measurable phenomenon C affects the
- >|> probabilities of both A and B, one can mathematically state that
- >|>
- >|> P(A) = intergal of Pa|c() * delta P(C)
- >|> P(B) = intergal of pb|c() * delta P(C)
- >|>
- >|> where Pa|c is the 'conditional' probability of A given C
- >|> (I belive this is known as Baynes thereom)
- >
- >There are several problems with this. One is that the formal mathematical
- >properties of probability are not necessarily applicable to sub-QM physics.
- >The second is that the definition of integration is not clear when the
- >underlying functions are not continuous - or, at least, the number of
- >discontinuities are infinite (and what proof have you that they are
- >continuous?). There are many other assumptions buried in the math.
- >The idea that conventional mathematics is necessarily applicable to
- >these physical domains is pure speculation. Even the definition of
- >*multiplication* is a human invention which we have no reason - beyond
-
- The relationship between probability theory and quantum mechanics is far
- better understood that Giles seems to imply. Quantum mechanics can be
- thought of as a generalization of classical probability theory, and it's
- easy to check in many cases which theorems of classical probability
- theory extend to the quantum case. Irving Segal (who was a statistician
- before he went into mathematical physics) developed the formalism of
- C*-algebras and noncommutative integration theory in the 40's and 50's
- for exactly this reason (among others). Bayes' theorem holds, in
- particular. (But Vidugiris has not stated Bayes' theorem... maybe he
- really DID mean Baynes' thereom! :-) Bayes' theorem is that the
- probability of F given E is the probability of E given F times the
- probability of F divided by the probability of E.)
-
- Giles' fears about discontinuities and such are uncalled for; measure
- theory doesn't really care about discontinuities. Of course, learning
- probability theory, measure theory and C*-algebras takes work, and it
- *is* wise to be cautious about this stuff before one has studied it.
-
- So where to learn this stuff? Bratteli and Robinson's 2-volume book is
- excellent but perhaps a little intimidating. One could always go to the
- locus classicus:
-
- Segal, I. E. A noncommutative extension of abstract integration, Ann.
- Math. 57 p. 407, 58 p. 595 (1953).
-
- Abstract probability spaces and a theorem of Kolmogoroff, Am. J. Math.
- 76 p. 721 (1954).
-
- A note on the concept of entropy, Jour. Math. and Mech. 9 p. 623 (1960).
-
-
-
-