dub@pur-phy.UUCP (Dwight U. Bartholomew) (04/06/84)
I am not an expert in quantum field theory, but I do know that in the recent past people working in the field DID have annoying singularities in their calculations that made their calculated numbers worthless. Then came Renormalization theory!!! It said that some of those singularities (infinites, to those not into the jargon) were different "kind" of singularities and by redoing the calculations one could make these "odd" singularities drop out. The end result is that quantum field theory can predict numbers like the fine structure constant to zillions (a slight exaggeration) of decimal places. Dwight Bartholomew UUCP: {decvax,ucbvax,harpo,allegra,inuxc,seismo,teklabs}!pur-ee!Physics:dub INTERNET: dub @ pur-phy.UUCP
clt@pur-phy.UUCP (Carrick Talmadge) (04/06/84)
Also, continuing Dwight's defence of Q.E.D., one must distinguish between infinities which arise as a consequence of the field theory, and infinities which arise as a consequence of the method by which we obtain solutions to that field theory (namely via the quantum perturbation method). Infinities may be obtained even in classical field theories, such as General Relativity, when one does a perturbation expansion ( the usual example of this is in solving the two body problem to determine the rate of precession of the perihelion as in the case of the Sun-Mercury system). These infinities are eliminated via "renormalization" just as in the case of a quantum field theory. Carrick Talmadge UUCP: {decvax,ucbvax,harpo,allegra,inuxc,seismo,teklabs}!pur-ee!Physics:clt INTERNET: clt @ pur-phy.UUCP
gwyn@brl-vgr.ARPA (Doug Gwyn ) (04/08/84)
Then there are REAL infinities caused by conceptual problems, such as the self-energy of the electron in classical electromagnetism. The problem with QED renormalization is that it smacks of an ad-hoc method to avoid a problem. The basic idea in simple terms is to consider the terms in the formulas as pertaining to an (unknown) "bare" quantity (mass or whatever) and reinterpret the theory so the "real" quantity is predicted. Unfortunately this pushes the difficulty into the unmeasurable "bare" quantity, much like sweeping dust under the rug. (To forestall flames: yes, I know about renormalization groups and the associated paraphernalia, but I don't want the critical concepts to get lost in a morass of technical details. I think this happens far too frequently these days.)