2011-03-06, 18:06
#1
Jag läser tredje året på gymnasiet på samhällsvetenskapliga linjen och har förmodligen tagit mig vatten över huvudet, angående mitt val av projektarbete. Jag ska i arbetet diskutera determinismens kompabilitet med i huvudsak två vetenskapsparadigm; Newtons klassiska fysik och kvantmekaniken. Som samhällsvetare har jag endast läst Matematik C och Fysik A, och mina nästan obefintliga förkunskaper har, föga förvånande, visat sig bli en begränsing i mitt skrivande av arbetet. Jag hittade litteratur som tar upp ämnet, John Earmans bok "ASPECTS OF DETERMINISM IN
MODERN PHYSICS" visar diskuterar utförligt den ovan nämnda kompabiliteten. Om jag bara hade förstått vad han skrev hade allt varit frid och fröjd, men det gör jag naturligtvis inte:
"Classical physics is widely assumed to provide a friendly environment for determinism.
In fact, determinism must overcome a number of obstacles in order to
achieve success in this setting. First, classical spacetime structure may not be
sufficiently rich to support Laplacian determinism for particle motions. Second,
even if the spacetime structure is rich, uniqueness can fail in the initial value
problem for Newtonian equations of motion if the force function does not satisfy
suitable continuity conditions. Third, the equations of motion that typically arise
for classical particles plus classical fields, or for classical fields alone, do not admit
an initial value formulation unless supplementary conditions are imposed. Fourth,
even in cases where local (in time) uniqueness holds for the initial value problem,
solutions can break down after a finite time."
1) Vad är en rik rumstidsstruktur?
2) "uniqueness can fail in the initial value
problem for Newtonian equations of motion if the force function does not satisfy
suitable continuity conditions"?
3) "the equations of motion that typically arise
for classical particles plus classical fields, or for classical fields alone, do not admit
an initial value formulation unless supplementary conditions are imposed"?
Vidare skriver Earman:
"Here is the (naive) reason for thinking that neither Laplacian determinism nor
any of its cousins stands a chance unless supported by enough spacetime structure
of the right kind. Assume that the (fixed) classical spacetime background is
characterized by a differentiable manifold M and various geometric object fields
O1,O2, ...,OM on M. And assume that the laws of physics take the form of
equations whose variables are the Oi’s and additional object fields P1, P2, ..., PN
describing the physical contents of the spacetime. (For the sake of concreteness,
the reader might want to think of the case where the Pj ’s are vector fields whose
integral curves are supposed to be the world lines of particles.) A symmetry of the
spacetime is a diffeomorphism d of M onto itself which preserves the background
structure given by the Oi’s — symbolically, d∗Oi = Oi for all values if i, where
d∗ denotes the drag along by d.17 By the assumption on the form of the laws, a
spacetime symmetry d must also be a symmetry of the laws of motion in the sense
that if M,O1,O2, ...,OM, P1, P2, ..., PN satisfies the laws of motion, then so does
M,O1,O2, ...,OM, d∗P1, d∗P2, ..., d∗PN.18
Now the poorer the structure of the background spacetime, the richer the spacetime
symmetries. And if the spacetime symmetry group is sufficiently rich, it will
contain elements that are the identity map on the portion of spacetime on or below
some time slice t = const but non-identity above. We can call such a map a ‘determinism
killing symmetry’ because when applied to any solution of the equations
of motion, it produces another solution that is the same as the first for all past
times but is different from the first at future times, which is a violation of even
the weakest version of future Laplacian determinism."
4) Hur kan rumstiden vara symmetrisk, och alltså ha två lösningar?
Jag vore oerhört tacksam om någon kunde hjälpa mig att förstå detta åtminstone lite grann, annars måste jag överväga att välja ett annat ämne till mitt projektarbete.
Hela John Earmans text finns att hämta här: http://www.lps.uci.edu/malament/prob-determ/Earman.pdf
MODERN PHYSICS" visar diskuterar utförligt den ovan nämnda kompabiliteten. Om jag bara hade förstått vad han skrev hade allt varit frid och fröjd, men det gör jag naturligtvis inte:
"Classical physics is widely assumed to provide a friendly environment for determinism.
In fact, determinism must overcome a number of obstacles in order to
achieve success in this setting. First, classical spacetime structure may not be
sufficiently rich to support Laplacian determinism for particle motions. Second,
even if the spacetime structure is rich, uniqueness can fail in the initial value
problem for Newtonian equations of motion if the force function does not satisfy
suitable continuity conditions. Third, the equations of motion that typically arise
for classical particles plus classical fields, or for classical fields alone, do not admit
an initial value formulation unless supplementary conditions are imposed. Fourth,
even in cases where local (in time) uniqueness holds for the initial value problem,
solutions can break down after a finite time."
1) Vad är en rik rumstidsstruktur?
2) "uniqueness can fail in the initial value
problem for Newtonian equations of motion if the force function does not satisfy
suitable continuity conditions"?
3) "the equations of motion that typically arise
for classical particles plus classical fields, or for classical fields alone, do not admit
an initial value formulation unless supplementary conditions are imposed"?
Vidare skriver Earman:
"Here is the (naive) reason for thinking that neither Laplacian determinism nor
any of its cousins stands a chance unless supported by enough spacetime structure
of the right kind. Assume that the (fixed) classical spacetime background is
characterized by a differentiable manifold M and various geometric object fields
O1,O2, ...,OM on M. And assume that the laws of physics take the form of
equations whose variables are the Oi’s and additional object fields P1, P2, ..., PN
describing the physical contents of the spacetime. (For the sake of concreteness,
the reader might want to think of the case where the Pj ’s are vector fields whose
integral curves are supposed to be the world lines of particles.) A symmetry of the
spacetime is a diffeomorphism d of M onto itself which preserves the background
structure given by the Oi’s — symbolically, d∗Oi = Oi for all values if i, where
d∗ denotes the drag along by d.17 By the assumption on the form of the laws, a
spacetime symmetry d must also be a symmetry of the laws of motion in the sense
that if M,O1,O2, ...,OM, P1, P2, ..., PN satisfies the laws of motion, then so does
M,O1,O2, ...,OM, d∗P1, d∗P2, ..., d∗PN.18
Now the poorer the structure of the background spacetime, the richer the spacetime
symmetries. And if the spacetime symmetry group is sufficiently rich, it will
contain elements that are the identity map on the portion of spacetime on or below
some time slice t = const but non-identity above. We can call such a map a ‘determinism
killing symmetry’ because when applied to any solution of the equations
of motion, it produces another solution that is the same as the first for all past
times but is different from the first at future times, which is a violation of even
the weakest version of future Laplacian determinism."
4) Hur kan rumstiden vara symmetrisk, och alltså ha två lösningar?
Jag vore oerhört tacksam om någon kunde hjälpa mig att förstå detta åtminstone lite grann, annars måste jag överväga att välja ett annat ämne till mitt projektarbete.
Hela John Earmans text finns att hämta här: http://www.lps.uci.edu/malament/prob-determ/Earman.pdf
