Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Could someone explain this entropic force business to me?

  1. Feb 15, 2010 #1
    So I don't understand what all the craze is about entropic forces because as far as I can tell, they're nothing new.

    All forces as we understand them are caused by a large number of particles. The electric force is due to the interactions of large numbers of photons; the weak force due to large numbers of W & Z bosons; etc. We have their distributions, and we can calculate their entropies. These entropies carry laws of their own that we can derive using statistical mechanics.

    If we look at the entropies of the standard model, don't we find "entropic forces" there as well?

    Am I missing something? Is there something that these entropic forces do that can't be done with statistical mechanics that forces us to adopt a new picture?

    I've heard the area laws of gravity are weird, but... at worst that should tell us that we should be counting things differently..
  2. jcsd
  3. Feb 18, 2010 #2


    User Avatar

    Right, they are nothing new as concepts as such. What's new is trying to explain some of the fundamental forces in terms of a similar abstraction.

    Classical gravity, as newtons gravity or GR, are not on par with the generalised thermodynamics forces that can infer from differences in chemical potential. Think osmosis for example.

    The current understanding can not infer say electromagnetism in terms of something analogous to entropic forces due to differences in statistical potentials.

    The fact that one sometimes can see the force mediating bosons as a gas doesn't quite mean that it is possible to infer the corresponding interactions just from some generalised gas laws. OR at least, that is not how the standard model is made.

    If that could be done - ie, one or more of the 4 forces could be "explained" as expected interactions in analogy with say osmosis, where different types of information flows between coding structures, resulting in an effective interaction forces due to the equilibration, then that WOULD be NEW.

  4. Feb 20, 2010 #3
    I was unsure just how much "entropic forces" could currently explain and if a quick Wikipedia check is any indication, the answer seems "not much"...


    For discussion purposes I'll disagree with Fra's opening comment (but I think I know what he means) :

    and say, instead, oh yes they are new concepts!!...or at least new interpretations.

    On the other hand, I agree completely with his idea here:

    Suppose we could show unequivocally that all forces derive from entropy!!! Like maybe via chaos ideas; of Verlinde's approach; WOW, that would be a monumental achievement..

    Our understanding of the relationship between "fundamental forces" and any other physical entities such as entropy, information, mass, time, space is pretty sparce so far ....perhaps something like the standard model (of particle physics) where we have glued together a lot of slick interpretations, can explain a lot, but are so far left with a hodgepdge of ideas and inputs and have no absolutely idea why an electron, for example, has the charge it does.

    Nor why time "flows" in one direction....but it seems related to increasing entropy....so one thing that is universal, entropy, would be a great entity to better understand. And if it can be shown that any of the basic entities of physics derive from entropy, HOLY COW, that is massive,impressive and could forge the link among all those constituents...

    Another way to say it is that maybe the one thing that comes from a bang origination for our universe is ENTROPY...and forces and everything we are, comes from a single overarching entity, entropy...forces are a derivative of entropy, as might be all other constitutents...even "strings" for example might be a form of entropy, maybe information...

    So anyone who can directly link entropy, or its evolution, with any other physical entity, like force, might have taken a step as big as Einstein's when he decided "the speed of light must be fixed".... His ideas gave us some insight, a peek, into some relationships between space,time and speed...and changed our understanding forever. Maybe that's Verlinde can do.
  5. Feb 21, 2010 #4


    User Avatar

    Yes, with not new I of course meant the concept of entropic force.

    It's application to inertia and gravity though, is new

    I share the vision you describe, this "big step" is the one step we need. And I agree verlinde is going in the right direction.

    To point to probably the key question arising from verlindes papers, is to also try to understand the origin of the holographic principle itself (to just refer to results from black holes is not really quite satisfactory; I'm sure Verlinde would agree - he admits that his paper feeds many questions, all he insists on I think is that they are important, and I agree there).

    In particular to _separate_ the holographic principle referring to spacetime degrees of freedom, from general degrees of freedom. Howevever, complex as it maybe I think think "problem" is also indeed the key to eventually getting ALL forces in under the same umbrella.

    I think the next step has to include a better understanding of the holographic principle.

  6. Feb 21, 2010 #5
    yes of course, why didn't I think of that???..entropy, information...geometry....
  7. Feb 21, 2010 #6


    User Avatar

    Smolin has some thoughts on this that I think partly are in the right direction.

    In Smolins reflections on Verlindes paper

    Newtonian gravity in loop quantum gravity (http://arxiv.org/abs/1001.3668)

    he writes during comments & conclusios

    "Nonetheless, it is intriguing to wonder if the relationship between area and entropy
    is even more fundamental than the notion of geometry itself. Could there be a more
    fundamental picture, before spacetime emerges in which area has the fundamental
    meaning of the capacity of a quantum channel by which information flows[38]?"

    Reference [38] is this

    Holography in a quantum spacetime by Fotini Markopoulou and Lee Smolin
    "We propose a formulation of the holographic principle, suitable for a background independent quantum theory of cosmology. It is stated as a relationship between the flow of quantum information and the causal structure of a quantum spacetime. Screens are defined as sets of events at which the observables of a holographic cosmological theory may be measured, and such that information may flow across them in two directions. A discrete background independent holographic theory may be formulated in terms of information flowing in a causal network of such screens. Geometry is introduced by defining the area of a screen to be a measure of its capacity as a channel of quantum information from its null past to its null future. We call this a ``weak'' form of the holographic principle, as no use is made of a bulk theory."
    -- http://arxiv.org/abs/hep-th/9910146

    Now, seeing the holographic principle as relating to communication channels in general is IMO a step in the right direction. There are still more questions but it's sniffing is the right way.

    Now, if we can combine a generalization of this (holography in terms of general interaction channels) with a entropic force defined relative to each screen in this "multi-screen" ~ "multi-observer" idea then I think it's starting to get really nice.

    Last edited: Feb 21, 2010
  8. Feb 23, 2010 #7


    User Avatar

    Some further interesting ket pointers from Smolin in that paper goes onto Fotinis "quantum causal histories" approach and Louis Cranes abstractions of categories of observers and observations. I like parts of this, but still feel that there is still alot of baggage that is coming from forcing the desired results. I like to pick some raisins out of the bun here.

    These are to me some key traits

    Smolins comment on Quantum Causal histories:
    "...any physically meaningful observable corresponds to some observer (represented as an event or collection of events) inside the closed universe. (In the classical case, see [15]). As each observer receives information from a distinct past, the algebra of observables they can measure, and hence the (finite-dimensional) Hilbert spaces on which what they observe can be represented, vary over the history. Consequently, the algebra of observables of the theory is represented on a collection of Hilbert spaces. These replace the single wavefunction and single Hilbert space of other approaches to quantum cosmology"

    Smolins comment on Cranes ideas:

    "Crane proposed that in quantum cosmology, Hilbert spaces should be associated not with the whole universe, but with any choice of a boundary that splits the universe into two parts. The idea was that the observers and their measuring instruments live on one side of the boundary, and they observe the quantum gravity dynamics on the other side by means of measurementsmade on the boundary, and recorded in a boundary Hilbert space."

    IMHO, is should also be obvious that since entropy as a measure of disorder and thus related to the "a priori" probability of a particular probability distribution in a space of such, or in the discrete case a probability measure of the macrostate given ignorance of the microstate IN A GIVEN microstructure, that entropy is a strictly relative concept. IT does not make sense to talk about the entropy as an objective property of a system.

    This fits in the above context that explains that entropy can only be defined in the decomposition of the universe in two parts: the observer and it's environment. Or put differently a measure of the expected future, relative to a constrained history of this observer.

    The question should IMO be, how the action of a system, coding an expectation on it's environment depends on this expectation, and then the hard problem - what relations that emerge in a network of such abstractions?

    I like to think of this as a purified question, and I think some approaches use global external constraits that force the emergent structures to have certain structures. The nature of these assumption are disturbing me as I've read a number of ideas related to this.

    I think it's wrong to think of them as "constraints", I prefer to think of them as "expected constraints" that refer to an expected equiblirium. At first the difference is only interpretational but I think it also makes a difference on wether you try to hide in some structural realism or not.

  9. Feb 23, 2010 #8
    "Now, seeing the holographic principle as relating to communication channels in general is IMO a step in the right direction."
    yes but that doesn;t seem new here...Susskind, for eample, discussed that elsewhere...in his holographic considerations I'm pretty sure....

    I like your comments in post #6 and read the introduction to the paper you reference there...Remarkably I stopped and reread "Smolin's comments on quantum causal histories" which I tehn saw you quoted in post #7, so maybe I understood a touch more than I thought!!!! But not as much as I'd like.

    The conclusions/implications the authors seem to make from various statements do not seem to flow logiocally for me...not that I disagree, I just am unable to see the flow:

    so they need a causal framework, and apparently liked one already defined..ok

    Now I start thinking, why closed? what does that mean? is that a critical, obvious unavoidable consideration (a raisen) or just an unnecessary piece of fluff (some dough)??


    Makes sense, different observers see different histories....

    Why Hilbert? because others used it and it seems to fit the requirements? So here it sounds like prior views required all observers to see identical histories?? I can't even begin to imagine the implications between the current paper and this single wavefunction formulation....sounds like a version of "relativity of simultaneity" for example....

    Anyway, I'm not seeking answers necessarily, just commenting that the links/logic between the authors thoughts I think elude me.....
  10. Feb 23, 2010 #9


    User Avatar

    I do not see a complete logic either because IMO it's just not there yet :wink:

    That's why I just try to highlight and combine what I think are good points and try to see what alternative routes one may see.

    The core issue is how to view/handle one of hte observer paradoxes

    a) we want the laws of physics to be independent of observer choice
    b) all observations are unavoidable observer dependent

    Noone suggests that all observers see the same history, since the escape is that there exists relations between the set of observations made by the set of observers, and these represent the invariants.

    This is what's done in relativity. The set of observers is associated to reference frames.

    That sounds very easy, and it is easy in that classic cases but it has the similarly obvious problem that these invariants still need to be inferred by an inside observer, unless you think scientific processes and representation of the laws of physics somehow exist in a mathematical external reality. That's what some of the wavefunction of the universe ideas does. That's the objection I raise - there is not really a physical observer corresponding to that wavefunction, then the obvious question is what the heck IS it? IMHO it's a complete rape and misapplication of the philosophy behind a sensible "measurement theory" to try to keep the abstractios and do away with the context. I've raised the same opinon numerous times. I think it's what happens when people use mathematical formalisms in absurdum without considering what it means and in what context it belongs.

    There are different philosophies here, each of them having their own problems to solve.

    "Doing away" with the observer is one way, and see the invariants as a form of structural realism of physical law. The problem here is that this does IMO not fit at all in an intrinsic measurement theory where all information is the result of interactions. It contains massive realist elements which breaks the logic.

    Others, I belong to this camp, thinks that there is no way to do away with the observer dependence. I also think there is no acceptable form of realism applying to physical law. In this alternative view, the natural escape is to consider that "in principle" each observer "sees" potentially different laws of physics. The problem here is to make this predictive, and to explain the #effective objectivity" that we unquestionably see. I think this is best and naturally accomplished by an evolutionary scenario. Evolution of law is really the evolution of observers, but it's definitely not the antrophic arguments, it's a more proper darwinian mechanism.

Share this great discussion with others via Reddit, Google+, Twitter, or Facebook