Talk:Statistical physics

From Wikipedia, the free encyclopedia
Jump to navigation Jump to search
WikiProject Physics (Rated Start-class, Top-importance)
WikiProject iconThis article is within the scope of WikiProject Physics, a collaborative effort to improve the coverage of Physics on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.
Start-Class article Start  This article has been rated as Start-Class on the project's quality scale.
 Top  This article has been rated as Top-importance on the project's importance scale.


The much more extended article in the french version of wikipedia has been proposed for translation into english on the platform. (talk) 20:18, 8 November 2013 (UTC)

-- Seemingly linked discussion thread--

Is »statistical physics« any different than »statistical mechanics«? Does it indeed deserve a separate entry? --Peterlin 14:16, 9 Feb 2004 (UTC)

I'm not an expert, but I would say that from reading this article statistical mechanics is quite possibly a subdiscipline of statistical physics. That is, there are many physical phenomena which can be analyzed using probability methods, not all of which are mechanics problems. moink 16:35, 9 Feb 2004 (UTC)
As a physicist, I can attest that the terms "Statistical mechanics" and "Statistical physics" are used interchangeably, but that "Statistical mechanics" (or "Stat mech") is preferred. Furthermore, "Statistical mechanics" is not restricted to refer only to the application of probability to problems in classical mechanics. For instance, the book Statistical Mechanics by Kerson Huang devotes a significant amount of space (and a chapter title) to quantum statistical mechanics. I think that these pages should be merged. The Wilschon (talk) 23:02, 14 May 2008 (UTC)
  • I think statistical physics is something wider, and therefore different. For example, percolation models aren't necessarily about statistical mechanics; nor I think are eg the power-spectral properties of random processes that are the subject of the Wiener–Khinchin theorem. There are random processes in physics which don't reduce to the application of probability to either classical or quantum microscopic dynamics. Jheald (talk) 16:45, 19 May 2008 (UTC)
Interesting in this regard might be the topics list for the conference StatPhys 23 held last year. While a majority of the topics probably could fit under the heading "statistical mechanics" (broadly interpreted), I'm not sure it would necessarily include #4 (Pattern formation in systems out of equilibrium), #5 (Dynamical systems and turbulence), #10 (Biologically motivated problems), or some of #11 (Interdisciplinary topics in statistical physics: networks, econophysics, traffic flow, algorithmic problems, astrophysical applications, etc).
I'd tend to identify statistical mechanics with the application of statistical methods to molecular-scale models. Some of the Statistical Physics topics I think are broader than that, and in some cases not even analogous to established molecular-scale results. Jheald (talk) 20:19, 19 May 2008 (UTC)
To restart this discussion, I've posted some suggestions about a general merger of this topic at Wikipedia_talk:WikiProject_Physics#Statistical_physics_.2F_mechanics_.2F_thermodynamics. Djr32 (talk) 13:17, 22 November 2008 (UTC)
I'd also like to add my weight to a discussion that has been stale for too long. Firstly the article from the French wikipedia simply seems to be a translation of statistical mechanics, so I am not sure how useful it is. On to the main subject...
I agree with much of what was said above by Jheald and moink especially; however I want to say it in a different way. Mechanics is the conceptual starting point of physics (in general, physics theories tend to begin from theories of mechanics), but the two are not the same. In particular, classical mechanics is contained in, but not synonymous with classical physics. In this sense statistical mechanics is the basis of statistical physics. Stat. phys. may be thought as the application of stat. mech. to traditional physics (or chemistry) problems (phase transitions, turbulence, nonlinear models, spin ice, strongly interacting systems (which stat. mech. mostly neglects) e.g. plasmas and ionic solutions, even gravity i.e. in astrophysics) or even outside physics proper (e.g. network theory, graph theory, neural networks, cellular automata, excitable media etc some of which can even be described by a partition function). In toto, hopefully this will be another step towards a WP:consensus that the two are different, and that the future direction of this article should reflect that. (talk) 07:20, 11 December 2012 (UTC) PS Note also (although admittedly, this does not strictly support my argument) that the branches of physics template has both statistical mechanics and statistical physics, so this discussion about separateness may be academic.
The discussion Djr32 was referring to is now located at Statistical physics / mechanics / thermodynamics. (talk) 14:05, 15 December 2013 (UTC)

speak in natures own words when discussing - not to become vague[edit]

First I think that the exact content of the text about "statistical physics" is not very important because how ever you formulate it - it would result in rather vague texts. This may be because all questions put forward (such as this one - what is staistical physics ?) by the investigating human beings are not optimised to get clear cut answers.

When that is the case you would get a more consice and meaningful text if you stick to very simple statements about nature, and make a text about them, letting the reader know that you are talking about a vaguely defined term.

For instance, In nature occurs phenomena such that we may describe as "statistical" either due to the appearance and interactions of many objects in a physical system, or due to quantum mechanical effects...

Then build up a popularized text according to this. Then you may possibly add a discussion of why subjects like "econophysics" and similar have more in common with statistical physics compared to theories of neural networks. Do they?

Regards Erik Eneroth, Askersund! —Preceding unsigned comment added by (talk) 18:46, 10 June 2008 (UTC)

Strongly advocate merging with statistical mechanics[edit]

  1. This "statistical physics" page is simply not useful as it stands. The bulk of the discussion concerns "statistical mechanics", which has its own page - why duplicate here?
  2. The current box urges translation from French, and in the French wikipedia there is NO "statistical mechanics" page ("mécanique statistique") only the linked statistical physics page, which is largely about only EQUILIBRIUM statistical mechanics.
  3. The French page links to a separate discussion of non-equilibrium statistical physics, which sounds like a good idea to me; the techniques for handling non-equilibrium are very different.
  4. I taught statistical physics/mechanics 25 year ago as a grad student; in my experience the term "statistical mechanics" has never meant just applying statistical concerns to mechanical systems, but covers every application of the statistical many-body problem.
  5. The ONLY thing that occurs to me as might be covered by "statistical physics" but not by "statistical mechanics" as the term is commonly used is the more traditional application of statistics in analysis of observations, where the number of instances N is closer to 1 than to Avogadro's number. That subject is covered just fine by pages like statistical hypothesis testing so if we really want a "statistical physics" page maybe it should be limited to a disambiguation page between that (and any other similarly appropriate pages) and statistical mechanics.

Anyway, I strongly advocate eliminating this page in favor of the "statistical mechanics" page with either a redirect or a disambiguation replacing this. ArthurPSmith (talk) 20:44, 31 January 2014 (UTC)

I thought this debate could use some more information: I surveyed some textbooks, looking into the introduction and the index for definitions of statistical physics vs. statistical mechanics (my notes here). I found that the two terms are almost entirely conflated in introductory (and advanced) textbooks. Many books titled 'Statistical Physics' use only the term 'statistical mechanics' on the inside. Most books titled 'Statistical Mechanics' don't mention 'statistical physics'. The basic definition given is roughly the same, whether the book is calling it 'statistical physics' or 'statistical mechanics'. Almost no authors explicitly talk about a difference between the terms. Quantum statistical mechanics is still called statistical mechanics.

Personally I agree with User:Ema--or and would like to see 'Statistical Physics' used for some applications of statistics that are beyond statistical mechanics. I've seen research in biology, sociology, etc called statistical physics. However, it seems if we go with the consensus/default of the field they are two terms for the same thing. I think it would require at least some (newer?) sources explicitly explaining this conception of 'statistical physics' to justify having a wikipedia article about it as a distinct concept. Cyrej (talk) 16:40, 17 February 2018 (UTC)

Verdict, 24 October 2019, moved from here to there. Ema--or (talk) 21:00, 16 February 2021 (UTC)

Well, well, well, well (well x4). Have a look at that. This old chestnut. Our faithful friend, (who earlier appeared here). Ema--or (talk) 20:44, 16 February 2021 (UTC)

Ain't that something? We even got it elsewhere... Ema--or (talk) 21:40, 18 February 2021 (UTC) 'N even 'ere. Ema--or (talk) 21:45, 18 February 2021 (UTC)

I need to make a section on this. Lol Ema--or (talk) 20:53, 19 February 2021 (UTC) Persistent problem. Systemic, even maybe. Ema--or (talk) 21:10, 19 February 2021 (UTC)

I’ve also decided to make connections btw this discussion and others, as I see fit. Ema--or (talk) 21:16, 19 February 2021 (UTC) Arghhhh. Its's everywhere. There’s an avalanche (nearly mispelt avalance). An avalanche


Sorry, been away for a while; but don't worry, here's a warm up to my view ...

Thermal qft, quantum stat mech and stat field theory:

  1. are they the same, and if not what's the difference? Ema--or (talk) 03:48, 28 January 2018 (UTC)
  2. are they part of stat mech? (And why?) Ema--or (talk) 04:11, 28 January 2018 (UTC)
  • Sorry I had forgotten about this. The response to my query... no, yes and no; more precisely 1)No,2)yes, and 3)no. Why? Because 1 and 3 are field theories, and thus not mechanics branches while 2 is just the quantisation of stat mech essentially. Indeed, one might argue that mechanics and field theory are completely opposite and complementary parts of physics. These should be correctly, viewed as statistical field theories or applications of qft in this subfield area. Or as an extension of stat mech to field theoretical regimes. cheers! Ema--or (talk) 01:17, 7 July 2018 (UTC)
    1. And of course, all three are not the same. The difference between tqft and sft is to me though, formally unclear. So, what is it? Any help from experts? Ema--or (talk) 01:23, 7 July 2018 (UTC)

New one. Quantum thermodynamics ...? Ema--or (talk) 20:54, 17 February 2021 (UTC)

List of analogiesEma--or (talk) 16:56, 20 February 2021 (UTC) page becoming a mess. Sorry abt. this.

Possible branches of stat phys[edit]

statistical dynamics, statistical electrodynamics, statistical dynamics (again) Ema--or (talk) 02:50, 19 October 2017 (UTC)

scattering statistics Ema--or (talk) Ema--or (talk) 04:02, 28 January 2018 (UTC) wave stats Ema--or (talk) 01:44, 7 July 2018 (UTC)

statistical physics of waves Ema--or (talk) 20:58, 16 September 2018 (UTC)

Been a long time coming. 

Here is my verdict, where I bring all these things together. There's a textbook called statistical optics, fields called statistical electrodynamics (more commonly called/aka statistical electromagnetics), statistical dynamics, statistical kinematics. Statistical acoustics is relatively rare, but it seems to exist. Statistical fluid dynamics (and subbranches) is well known to exist. Statistical field theory is a field theory approach in opposition to the particle view of normal statistical mechanics.

All the above are branches of physics, they not? IE minus the 'statistical'? I think statistical physics is a subset of math phys applying statistics and probability. Ema--or (talk) 23:59, 24 October 2019 (UTC) Appl. stat. & prob.? Ema--or (talk) 22:04, 17 February 2021 (UTC)

More [1], [2]1 Ema--or 20:25, 17 February 2021 And more [3] Ema--or (talk) 19:53, 18 February 2021 (UTC) [4] [5], [6] , [7] (maybe makes it stat mech huh?) Ema--or (talk) 20:05, 18 February 2021 (UTC)

Nuclear? Ema--or (talk) 21:06, 19 February 2021 (UTC)

Probabilistic physics? Ema--or (talk) 21:17, 19 February 2021 (UTC) Even probabilistic electrodynamics, online search.

translation template[edit]

I removed the template recommending that content from the french fr:Physique statistique should be moved here. As evidenced by this talk page, there's a lot of debate and uncertainty about statistical physics vs statistical mechanics and how to organize this. The french version is different: there is only one article, fr:Physique statistique, and it's about (equilibrium) statistical mechanics. Since they don't directly correspond and where the content goes is still very debatable, a translator should not blindly move content from the French article to this page. Cyrej (talk) 12:28, 17 February 2018 (UTC)

- Agree!! Ema--or, 01:34, 7 July 2018 (UTC)