Phenomenological theories: thermodynamics

[Signifier]

"Phenomenological" theories: thermodynamics

Hello, I was reading A. O. Barut's Electrodynamics and Classical Theory of Fields and Particles and was interested by a phrase he used, "'atomistic' physical theories - in contrast to phenomenological theories like thermodynamics - ...". What could be meant here by "phenomenological theories like thermodynamics"? I have a vague sense of what Barut meant, but any more insight would be very helpful.

 

[StatMechGuy]

I'm not entirely sure what he would mean by that. I guess thermodynamics is a theory that does not really discuss the microscopics, and it requires measurements to make further predictions (i.e. you cannot calculate U(N,V,T) purely from thermodynamic considerations except in very special cases), but at the same time thermodynamics is quite robust. It holds regardless of the system in question, quantum or classical, so in that case it's less "atomistic" than the other branches of physics.

Although I have no idea why he would divide physics up like that.

 

[cesiumfrog]

Statistical mechanics takes knowledge of the atoms of a system, and deduces from that some behaviour of the system as a whole.

But plain thermodynamics was phenomenological in the sense of describing just empirical results, without any other basis. In particular, the 2nd law of thermodynamics matches phenomena we measure, but almost contradicts our understanding of the underlying mechanisms by which atoms and so forth interact.

The phenomological approach is to measure some complex thing, plot the data, and try to guess an equation that will predict future measurements accurately. Fitting the equation involves choosing the best value for each parameter, with no prior reason or explanation to favour a particular value.

The atomistic approach is to study each element of the system in isolation, to learn how they work and interact. You can measure properties of each element, and then (by deduction) you write down the equation for the whole system. Hopefully you get the same equation as empirically, except now you also know exactly why each parameter has the value it does.

 

[dextercioby]

That's a nice explanation, cesiumfrog.

Daniel.