The following discussion is an archived debate of the proposed deletion of the article below. Please do not modify it. Subsequent comments should be made on the appropriate discussion page (such as the article's talk page or in a deletion review). No further edits should be made to this page.

The result was merge to Principle of maximum entropy. The keeps failed to satisfy the content fork concerns. \ Backslash Forwardslash / (talk) 01:47, 17 October 2009 (UTC)[reply]

Law of maximum entropy production[edit]

Law of maximum entropy production (edit | talk | history | protect | delete | links | watch | logs | views) – (View log)
(Find sources: Google (books · news · scholar · free images · WP refs· FENS · JSTOR · TWL)

Not notable. POVed. Ϙ (talk) 18:51, 30 September 2009 (UTC)[reply]

If you look at the edit history, I tagged it speedy a few minutes after creation and was soundly rejected. I have since found a bunch of stuff on the talk page which should serve as a good reference for those interested in this topic. There seems to be a few Swenson proponents with few additional edits except related topics that ref Swenson. The work the proponents want to make the focus minor or fringe by most obvious criteria. I think the term may be notable but article needs a lot of work and we now have some interested folks. FWIW Nerdseeksblonde (talk) 18:57, 30 September 2009 (UTC).[reply]
yeah, well the special issue highlights Dewar who is marginalized here as having shown to been "Wrong." Additionally, I'm not sure if contributor or Swenson is making inconsistent or at least imprecise claims or making unhelpful comparisons and assertions. I keep asking on talk page for sources but without luck in many cases. The one ref that treats Swenson in any depth seemed a bit circular with maybe only one citation on gscholar. Nerdseeksblonde (talk) 22:46, 30 September 2009 (UTC)[reply]
CounterPoint: If you take out the self-cites for Swenson, IIRC you only have 2 in which he is singled out to any great extent. Of these, one is only mentioned once on gscholar. Sure the volume of stuff is there, but you have to look at it a bit. Keep in mind, I'm defending more inclusiveness on Creation Science due to certain definitional issues. And I would support this article if there is some community of MEP-ers even if their work has no scientific merit. We need notability, not nobel prizes but given the number or, eh transient and unorthodox theories this areas draws such as perpetual motion machines, you may be able to impose a greater bar on notability before merging. Personally I think the topic can stand but Swenson would be a small part of a larger taxonomy of things with the same basic name. Nerdseeksblonde (talk) 00:20, 1 October 2009 (UTC)[reply]
This article is distinguished by its use of the word 'Law' to refer to the assertion that it discusses. It seems that the main author of the article thinks that use of the word 'Law' is very important. Mostly the mainstream talks of 'a' or 'the' 'principle' of maximum entropy production. About the status and generality of the assertion that is made, whether 'law' or 'principle', the mainstream is not remotely as confident as Swenson and the main author of this article. The mainstream is avowedly struggling to provide a really sound thermodynamic definition of entropy for a non-equilibrium system, but such problems are not serious, it seems, for Swenson and the main author of this article. It is one thing to announce or "give" or "postulate" that an assertion is a law; it is another to prove its generality or validity. From the viewpoint of the mainstream literature, the 'empirical proof' of the assertion offered by Swenson and the main author of this article is jejune.
This article is really about the work of Swenson, not about a mainstream concept or study of the possibility of a principle of maximum entropy production in thermodynamic terms. The above "keep" comment by Cyclopia that this article is "well referenced" is valid only on the premise that the article is about the work of Swenson; it would not be valid on the premise that the article was about mainstream ideas on maximum entropy production in thermodynamics. Perhaps the enthusiasm of the main author of this article could be kept alive by giving the present article a new name that makes it explicit that it is about the work of Swenson, not about the mainstream thermodynamic concepts and problems of extremal principles for non-equilibrium processes. How important for the Wikipedia is the work of Swenson is another matter, which I am not qualified to assess.
If the present article title is left to stand, the present content of the article should be replaced with mainstream discussion of maximum entropy production in thermodynamics, with perhaps a very brief mention of and reference to Swenson, a very brief mention. It seems that the replacement article content should not be written by the main author of the present content.Chjoaygame (talk) 03:23, 2 October 2009 (UTC)[reply]
Can we have perhaps some more evidence that Swenson himself is Thermohistory and DrProbability? If so, then it will be hard to avoid the idea that this page should be scrapped entirely. I agree with Q's comment (under the heading "Delete this page" in the discussion tab of the page), that the present "entire piece is just trying to promote Rod Swenson's papers". I do not go quite so far as he does in saying that those papers are "entirely vapid" but I would agree so far as to say that they are largely vapid or jejune. Fiddling with the present version of the page is not a good way to solve the problem I think.
After the present page is scrapped, if someone else likes to write afresh another page on extremal principles for non-equilibrium thermodynamics, well and good; with luck such a new author will do a good job. The new author will presumably make some introductory statement more or less along the following lines: Onsager (1931)[1] wrote: "Thus the vector field J of the heat flow is described by the condition that the rate of increase of entropy, less the dissipation function, be a maximum." Careful note needs to be taken of the opposite signs of the rate of entropy production and of the dissipation function, appearing in the left-hand side of Onsager's equation (5.13) on page 423. The possible new Wikipedia page on extremal principles will probably talk about both maximal and minimal principles, and about both entropy production and the dissipation function. This is a task for a serious expert, who may not be available for it. The Swensonist is not in the race for this task. If the Swensonist likes to start a page about Swenson, I suppose he is free to do so, provided that what he does be subject to the usual Wikipedia customs. Chjoaygame (talk) 22:34, 3 October 2009 (UTC)[reply]
I can't get full text of Linsker paper but apparently the issue with Dewar is confined to something to do with SOC but not sure what else it addresses. AFAIK, these principles are all imprecise and seem to be yielding to fluctuation theorems. I think the Kirkoff law example was mentioned by Landauer as being wrong when circuit isn't isothermal or some other simple counterexample. There may be some rules of thumb buried in here but these have a lot of problems as dynamics would seem to be a microscopic issue and indeed many macroscopic quantities like temperature may not even have obvious definitions depending on mechanisms available for thermalization/equillibration of different pars of the system. But, there is a bit of discussion so I imagine the topic is notable but it would probably include more history, including some Onsanger as you suggest, and discuss the fluctuation theorems as Dewar has done. Issues of trying to relate subjective "order" to something precise like entropy should generate a lot of text but maybe not much science. In any case, current discussion relating min and max entropy producers doesn't seem consistent with current form in article from Team Swenson. Nerdseeksblonde (talk) 01:45, 5 October 2009 (UTC)[reply]
There are two issues in Grinstein and Linsker 2007. The more important for our present concern is that they say that the Dewar 2005 proposed proof of a principle of maximum entropy production relies unnoticed on an assumption of linearity, while the main point about the principle is that it should extend to cover non-linearity. They say that this leaves the validity of the principle undecided, not that it refutes the principle. As you note, the other issue in Grinstein and Linsker 2007 is about SOC. I haven't seen a reply from Dewar.Chjoaygame (talk) 03:44, 5 October 2009 (UTC)[reply]
Well, I guess the Team Swenson assessment of Dewar is a bit too negative then and I was just parroting that. AFAIK, these are principles or almost platitudes, slightly above Murphiy's Law, designed to give guidance but not trajectories to the systems' analyst. As such, a missing proof is probably not too damaging any more than Landauer's ( ca 1975) counterexample of the non-isothermal circuit. In any case, the Swenson citations for MEP as opposed to other things he may have done seem to be limited and going forward the Dewar approach with segue into flucutations seem to be more popular. Nerdseeksblonde (talk) 14:47, 5 October 2009 (UTC)[reply]

Oh, dear, I wouldn't like to think these principles were almost platitudes! I am not sure if they are really well defined or true or general. I think the serious experts find them problematic. To read Swenson you would think they are platitudes, but that's just because he doesn't seem to care about accuracy, or doesn't understand the problems. I think it really matters whether Dewar's proof is valid or not. I have an idea that perhaps he may be able to remedy the defect that Grinstein and Linsker 2007 found. I think accuracy matters a lot here. I would like to read a paper that made the physical meanings of the various principles and quantities clearer.Chjoaygame (talk) 15:20, 5 October 2009 (UTC)[reply]

I'm not saying it is irrelevant, but again consider the existence of counter examples or "judgement" in looking for constraints etc- did you find the Landauer example and determine which laws that addresses?I guess that circuit example is done in detail, you could probablyjust put in different temps and see what happens but note that V=IR doesn't have a tmperature term other than R(T). A single counter example would mean at best it is somekind of approximation. The dynamics are determined by microscopic issues- you can probably construct 3 level systems and do it almost analytically ( use Stirling, make up some interaction between particles, make a non-equillibrium situation, and write expressions for entropy and temperature etc). Nerdseeksblonde (talk) 15:46, 5 October 2009 (UTC)[reply]

Hmm.Chjoaygame (talk) 21:10, 5 October 2009 (UTC)[reply]

I have now read Swenson and Turvey 1991. I am reaching the conclusion that Swenson's work is a pollyanna wish list rather than a scientific achievement list. The notion of entropy arose in thermodynamics where it has a precise meaning. The stretched notion that entropy and 'disorder' are closely related is a vulgar error, as explained clearly in Entropy and the Time Evolution of Macroscopic Systems (Oxford, 2008, ISBN 9780199546176) by W.T. Grandy Jr, who there in secction 4.3 examines the serious literature about it. As explained by E.T. Jaynes in many places (for example, E.T. Jaynes (1965), Gibbs vs Boltzmann entropies, American Journal of Physics 33: 391-398), entropy in thermodynamics is about experimental reproducibility, not about 'order', whatever that might mean. Pace Swenson, the entropy of thermodynamics is not a prime candidate to explain the origin of species. Swenson takes entropy far beyond what its thermodynamic basis will carry. Swenson does not really understand the thermodynamic meaning of entropy. Perhaps the Swensonist would like to try to refute this statement of mine: the Wikipedia page on the "Law of Maximum Entropy Production" is not the place for him in the meantime to claim the benefit of what he would perhaps call 'the doubt'; Swenson has not made a good enough case to earn a place in the Wikipedia as a serious contributor to understanding of thermodynamic entropy production, which he is ostensibly claiming.

That is not to say that the origin and survival of species is not explicable in physical terms, but it is to say, contrary to Swenson, that thermodynamic entropy is not the main relevant physical concept. Also it is not to say that irreversible processes and entropy production are irrelevant to the origin and survival of species, but it is to say that they are not the main explanation. Entropy is not a force of nature, it is an explanatory concept, linking explanatory ideas, not describing a physical force. Swenson's work is really a grandiose and vapid misuse of a word, which apart from his misuse of it, has a well established precise meaning; the misuse works by baffling and impressing the ignorant. Swenson, like many others, would like to see that precise meaning extended to a wider domain, but such an extension is easier wished for than achieved. Swenson and Turvey 1991 set up a straw man: "... ordered states ... are the inexorable products of natural law rather than miraculous debt payers fighting against it." No scientist thinks that ordered states, nor living organisms (which, by the way, Swenson would like us to equate to 'ordered states'), are "miraculous debt payers fighting against [natural law]". To accept that living organisms are not "miraculous debt payers fighting against [natural law]" is not to accept that the principle or "Law" of maximum entropy production, with entropy defined as in thermodynamics, is the only explanation left. "There are more things in heaven and earth, Horatio, than are dreamt of in your philosophy." The straw man is testimony to the rhetorical skills of Swenson and Turvey, not evidence of their contribution to the understanding of the thermodynamic concept of entropy.

Entropy production is the path to empirical reproducibility, not 'order', and evolution is the creative advance of nature into novelty, not its inexorable advance into mere colourless 'order'. This is far from saying that natural creation disobeys physical causality, but it is saying that the path of evolution is unique and not reproducible in any empirically verifiable sense. Getting down to hard brass tacks, Swensonism is nonsense clothed in fancy dress. The Wikipedia is not the place to invite novices to struggle through this kind of critical reasoning. An Wiki article on Swenson would be better written as showing how even today someone can get away with using pseudo-science to bluff the ignorant, than as an article about how entropy is a useful thermodynamic concept.Chjoaygame (talk) 04:48, 6 October 2009 (UTC)[reply]

But is it notable? I guess my point is many or all of these princinples really can't hope to be much more than this, probably even the Noble Prize winners, as AFAIK they generally have simple counter examples. Even science fiction has been known to spur scientific advancement, anecdotes and folklore distilled into a one liner do have merit sometimes. And, again, these basic issues come up in most hard core sciences- try following biotech stocks LOL. If someone can find a one line equation to translate some subjective order into a quantity great but again we can't care much about merit. There is probably even less scientific regard for Creation Science but I would continue to defend a comprehensive article on this topic for a variety of intellectual reasons. So, I again think a purely one-author article does not fit this topic but a bio on Swenson or maybe even a Swenson MEP page, given the 100 possible g-scholar hits may be something to debate elsewhere ( do you want a page on every perpetual motion machine? Thermo has had a lot of non-notable or trivial entrants. ). Nerdseeksblonde (talk) 11:26, 6 October 2009 (UTC)[reply]

Thermodynamics is an exact science. "Principles" that have simple counterexamples are not part of it. An approximation in an exact science is distinguished by its stating exactly how far it can be relied upon.

Cyclopia writes above: "Article is problematic but subject is clearly academically notable and well referenced." This shows how easy it is for pseudo-science to be taken as if it were real science. Work is not academically notable to Wikipedia standards unless it is referenced by reliable secondary sources, which Swenson's is not, whether Cyclopia thinks so or not. Swenson is perhaps the only writer (or one of a fringe few) to propose that title "Law" for a principle of maximum entropy production. This is evidence of his boldness, but not of his correctness.

The Wikipedia does not aim to promote pseudo-science as if it were real science. The Swensonist wants to promote the work of Swenson as if it were part of thermodynamics, when in fact it is pseudo-science. An article in the Wikipedia is not the place for him to do so, nor for an accumulation of more or less related statements that might be mistaken for a scientific debate. Chjoaygame (talk) 22:23, 6 October 2009 (UTC)[reply]

Chjoaygame, I have two unrelated comments:
1)I have said that it is notable and referenced, not that it is true or meaningful. If it is pseudoscience, so be it, and let it be described as it, but it has nothing to do with it staying or not on WP; after all we have an article on flat earth.
2)For sure thermodynamics is not my cup of tea, yet I feel a bit worried by your explanation. Entropy is maybe not strictly "disorder", but isn't the Boltzmann definition of entropy a direct function of the accessible microstates of the system? And what do you mean by saying that it is an "explanatory concept"? Isn't it a physical quantity? I am worried that we're seeing conflicting POVs on the thing. --Cyclopia - talk 00:01, 7 October 2009 (UTC)[reply]
As originally written, notability was established by Team Swenson self-cites. The one ref that could claim intellectual independence itself has one passing ref on gscholar and that scarity of usage is ( IMO ) likely due to lack of merit. If you see the talk page, I think Team Swenson further claimed Dewar to have been shown to be invalid or unworthy of more than passing mention but AFAIK right now the proof is questionable and no affirmative proof of badness has been shown and AFAIK Swenson's Law is not even really testable. So, Swenson may be notable for an MEP notion and probably deserves mention somewhere but his proponent(s) here haven't been too helpful. Nerdseeksblonde (talk) 00:17, 7 October 2009 (UTC)[reply]

Cyclopia: Thank you for your two comments. As to the first: I am pointing out that at present one might very easily get the impression that the article on Law of Maximum Entropy Production was an article about an area within thermodynamics; but it is not so. As to your second comment: the Boltzmann definition of entropy is far from the only one. Boltzmann's definition is not suitable for defining the time rate of entropy production: Boltzmann's definition is very strictly an equilibrium definition, as usual for classical thermodynamics, while entropy production is specifically a non-equilibrium concept. If you would like to make thermodynamics your cup of tea, you might, I suggest, do well to read the references to Jaynes and Grandy that I gave above. For me, classical thermodynamics is perhaps best set out by the masterly E.A. Guggenheim in Thermodynamics: An Advanced Treatment for Chemists and Physicists, North Holland, Amsterdam, 1987, seventh edition ISBN 0444869514 (paperback). You are right that there are conflicting points of view about the statistical mechanics of thermodynamics. It is not an easy subject. When one is confronted with pseudo-science, one can be led into endless arguments, and that is not our purpose here. I suggest you carefully examine the content of Swenson's work for the sake of getting a clear idea of its status.

Nerdseeksblonde: Thank you for your comments. Swenson was bold to assert as a Law what others had long considered a possibility, but Swenson seems hardly to understand why they were not asserting it as a law; the use of the word "independent" seems intended to indicate his originality, but it does not establish the validity of his work as a part of thermodynamics. Is the use of the word "independent" intended to mean that he had not read the long-standing literature? Swenson's Chapter 6 of Cybernetics and Applied Systems (1992) cites not Rayleigh, Onsager, Ziegler, Gyarmati, Prigogine. Swenson's 2000 paper in the Annals of the New York Academy of Sciences cites 13 papers by Swenson including Swenson and Turvey 1991; it cites Schroedinger 1945; from the just-previous list it cites only Prigogine 1977. Swenson and Turvey 1991 cites 11 papers by Swenson; it cites Haken 1983, Malkus 1954, Nicolis and Prigogine 1989, Schroedinger 1945, Thompson 1852, but not others on the just-previous list. For thermodynamics, the problem is to define the precise meaning and range of applicability of the principle.Chjoaygame (talk) 09:26, 7 October 2009 (UTC)[reply]

I guess the question then is if Swenson is even notable enough to be included here or to get it's own article? If S has 100+ gscholar hits that are him, plus his Plasmatics work ( no, this has nothing to do with the non-equillibrium distribution of bond angles in polyethylene ROFL) he made be a reasonable BLP candidate. If the LMEP comes up in two places, then would it just be easier to make its own article and link to it from the two places? I don't think anyone beliefs he is the most prominent player in this topic other than Team Swenson. In response to other comments, all of this is about making up macroscopic summary numbers that tell you most of what you need to know about the system. Instead of chasing around individual particles, temperature may tell you enough. So, in reality all of these drop a bit of detail but somewhere between a microscopic trajectory and Murphy's Law an approximation or rule of thumb can have merit. And, sure, entropy production is precise, dS/dt but then you need to define S still. Note some terms from literature like "non-negligible" etc and think about what is accessible etc. Nerdseeksblonde (talk) 11:42, 7 October 2009 (UTC)[reply]

Relisted to generate a more thorough discussion so consensus may be reached.
Please add new comments below this notice. Thanks,  Sandstein  06:03, 8 October 2009 (UTC)[reply]
Ziegler, e.g.,[3], a contemporary of Prigogine, but whose work had previously gone largely unnoticed, has recently received recognition as person who believed the idea of entropy production maximization had an important role to play in physics early on. His work went mostly unnoticed for a number of reasons. One of them was that he did not come at the problem from the view of trying to solve the problem of why the world inexorably and opportunistically produced order, with life as a manifestation, but instead as a theorist in the specialized field of continuum mechanics. There was no suggestion by him that entropy production maximization might play a role in solving these larger theoretical problems, and thus his work escaped the attention of those with these broader, more general theoretical interests. Another reason is that while he stated a maximum entropy production principle, it was never stated as a universal law although it would seem he intuited this. His principle which he demonstrated using the geometry of vector space relied on a “orthogonality condition” which only worked for systems where the velocities define a single vector or tensor. It was thus, in his own words “impossible to test by means of macroscopic mechanical models”[27] and was invalid in “compound systems where several elementary processes take place simultaneously”. Such limitations are severe, but although Ziegler thus only went a certain distance with his intuition, he deserves credit for being there early at a time when the tide was running the other way.

This is a travesty of the meaning of Ziegler's conclusions on 346-347 of the second revised edition of An Introduction to Thermomechanics, North-Holland 1983. The travesty is slanted, perhaps unintentionally, to make it look as if Swenson has discovered and revealed to the world something that Ziegler missed. The reality is that Ziegler found that the principle of maximum entropy production does not make sense for and cannot be applied to compound processes, because the latter are not of essentially thermodynamic nature. Swenson may wish they were of essentially thermodynamic nature so that the principle of maximum entropy production might apply to them, but he is not remotely able to show that it is so, because in fact, as pointed out by the thoroughly expert Ziegler, it is not so. Swenson's arguments do not come within 100 miles of dealing with these problems with the precision that Ziegler has mastered. Glansdorff and Prigogine 1971 on page 15 make a point close to that just quoted from Ziegler: "Let us emphasize from the outset, that the local equilibrium assumption implies that the dissipative processes are sufficently dominant to exclude large deviations from statistical equilibrium." This means not that some better definition of entropy will make the principle apply, as Swenson evidently hopes and asserts without justification, but it means that the principle is undefined or often violated when the required conditions do not hold; but Swenson just shuts his eyes to this and asserts the principle as a "law" regardless, and congratulates himself on his boldness of intuition. The Swensonist condescendingly congratulates Ziegler on his intuition: no, Ziegler does not rely on intuition, rather he relies on a seriously rational scientific approach. If at some future time, a further extension of the principle of maximum entropy production should be validated, Swenson cannot claim that his work anticipated it; he has just guessed blindly.

It is also the case that the Swensonist, perhaps unintentionally, has previously seriously misrepresented the statement of Mahulikar and Herwig 2004 about "The major revolution in the latter half ... etc". And that he has, perhaps unintentionally, seriously misrepresented what Grinstein and Linsker 2007 had to say about Dewar's purported proof.

There is a fair amount of literature on entropy production, and different experts, including two Nobel Prize winners, have tackled the problem in different ways. It is not at all easy for a non-expert to write a fair sampling of the literature. It is not an easy subject. The degree of difficulty is illustrated by the absence of references to Rayleigh, Onsager or Gyarmati in the present article, and by the leading place given to the article by Mahulikar and Herwig 2004 (which, by the way, is cited once without listing Herwig as an author and a second time with his name misspelt). And while we are on spelling, principle is so spelt, not as principal.Chjoaygame (talk) 15:54, 8 October 2009 (UTC)[reply]

Well, I have to admit I had the same impression reading that as an attempt to include Zeigler as at least one other author who did single out Swenson did so for admirable intuition and independent rediscovery of existent notions. How about if we just blank the page and try again? I was trying to document the "undiscovered" work of Zeigler with a year-by-year citation count but I couldn't find a good tool and gave up, I'm sure our conclusion is right but I thought it would be easy to get numbers to prove it. Nerdseeksblonde (talk) 18:18, 8 October 2009 (UTC)[reply]


I have followed this up.Chjoaygame (talk) 07:56, 9 October 2009 (UTC)[reply]
Yes, we have been talking. I would note you moved the page to a more general topic and started a more comprehensive outline. It looks like the article will become pretty large and the entropy principles could be a topic in themselves and the terms are used in literature. If you can source your more general term that would help I guess but otherwise looks fine, we can argue over scope later. Nerdseeksblonde (talk) 11:16, 9 October 2009 (UTC)[reply]
I think it would be hard to present the subject without a fair amount of talk about the dissipation functions.Chjoaygame (talk) 21:02, 9 October 2009 (UTC)[reply]
Well, you could also define entropy and temperature but we have other pages for that and that was the path this article was going down. I would just avoid too much detail on things that have their own pages. If dissipation is that big a deal, create a page for it. Nerdseeksblonde (talk) 21:31, 9 October 2009 (UTC)[reply]
It isn't a fork see my OMG response, as the article does say or said somewhere, this is related to the entropy production rate and not the final entropy value. The topics are somewhat different. Nerdseeksblonde (talk) 11:45, 11 October 2009 (UTC)[reply]
The above discussion is preserved as an archive of the debate. Please do not modify it. Subsequent comments should be made on the appropriate discussion page (such as the article's talk page or in a deletion review). No further edits should be made to this page.
  1. ^ Onsager, L. (1931). Reciprocal relations in irreversible processes, I, Physical Review 37:405-426