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

Physics law evo book (Lee Smolin Bob Unger, in prep)

  1. Oct 6, 2008 #1


    User Avatar
    Science Advisor
    Gold Member
    Dearly Missed

    Smolin and Unger are preparing a book titled Can the Laws of Nature Evolve?
    What are in effect skeleton chapters of the book IMHO are available here. I don't know which chapters so I will call them A and B.
    Chapter A:
    Chapter B:

    The first of these would be the general theory of physics law evolution and the second would be the special case where the ideas are applied to landscape theories, making them testable.
    In each case you can go there and click where it says PDF and download the slides to the talk so you can study.

    The book will have a major effect IMHO and is worth thinking about already. Here are direct links to the PDF slides:



    There are certain main examples to assimilate and get in your head to make the whole thing easy to understand. Here is one.

    Since Newton, if not earlier, it was a universal law of nature that space geometry was orthodox 3D Euclid. Geometry was a reliable rigid given. After 1915 we realized that geometry evolves. the laws about triangles and stuff are dynamic and in process of evolving.

    Here is another case.

    Since Newton, we have the idea of conservation of energy, momentum, angular momentum. But Emmy Noether taught us that those arise from space and time symmetries. So their applicability is contingent on space having expanded enough to approximately even out. The conservation laws are contingent, approximate, and the result of the process mentioned before---the dynamical evolution of spatial geometry noticed in 1915.

    There are a lot of other cases discussed in the slides. There are 72 slides in the Chapter A set that you download. These examples of physics law that we already know have evolved begin around slide #44 or #45, if you want to look it up.
    Last edited: Oct 6, 2008
  2. jcsd
  3. Oct 6, 2008 #2


    User Avatar
    Science Advisor
    Gold Member
    Dearly Missed

    I should mention another basic example to have in the back of your mind (help as something to hang SmolinUnger ideas on). Recall that 1905 Special seemed to take away the idea of simultaneity.

    (Though I can view 1905 Special is a contingent approximate symmetry that depends on the universe having flattened and smoothed, so that Lorentz symmetry could be applicable locally.)

    In any case 1915 General gave back at least in an approximate sense an idea of absolute simultaneity and a universal present moment. All observers at rest with respect to the CMB for whom the CMB is 2.725 kelvin belong to the present. Conventional cosmology is based on this kind of timeslicing. The standard metric and the Hubble Law are formulated in those terms. Hubble law is a statement about the present ratio of the present distance between observers at rest with respect to CMB and the present rate of increase of that distance. In normal cosmology there is a succession of present momnts consisting of cohorts of stationary observers for whom the CMB sky has the same temperature.
    Cosmologists sometimes refer to it as "universe time"

    This isnt an example from the talk, it is something I am tossing in because having it in mind helped me to understand the talk.

    Another point is that in some quantum gravity models space is emergent but time is fundamental. When space does not exist there is no issue of Lorentz symmetry, violate or inviolate, because Lorentz symmetry is meaningless at that point. After space has emerged from more fundamental degrees of freedom and has expanded some, so there is an idea of locality, then Lorentz symmetry can evolve locally. I'm thinking in particular of Renate Loll triangulation model, where the sequence of successive momnts is fundamental and explicit while a smooth spatial geometry is emergent---arising dynamically out of microscopic chaos. This is just a side remark. I will get back to talking about the slides in a minute.
    Last edited: Oct 7, 2008
  4. Oct 6, 2008 #3


    User Avatar
    Science Advisor
    Gold Member
    Dearly Missed

    Another example they give is electroweak symmetry breaking

    When you unify two forces what you may end up doing is you contrive a mathematical formulation that works for both and then you need a mechanism to explain how it happens that the two forces are many orders of magnitude different, different strengh, different range etc. Then various mechanisms are proposed (to "break the symmetry") to explain why the unification isn't observed.

    Also it may happen that a proposed explanation sometimes depends on something accidental that just happened the way it did at one point in the cooling of the early universe----that is just how the cookie crumbled.

    Because the pencil balancing on its point had to fall down in some direction and just happened to fall that way.

    This is another example of real or imagined dynamical evolution of physics law.

    The slides give a lot of examples, starting at page 44 or 45. There are a lot of ways where we already know of physics laws evolving dynamically.
    Last edited: Oct 7, 2008
  5. Oct 6, 2008 #4


    User Avatar
    Science Advisor
    Gold Member
    Dearly Missed

    I'm gradually chipping and gnawing away at the 72 pages of the Chapter A slides.

    there is an important idea here which is the Newtonian schema----how you model a subsystem separate from the observer, using a combination of eternal laws with initial conditions. The laws are assumed to be eternally true and the initial conditions are subject to determination by the experimenter.

    in that case the experimenter is outside the box and there is a clock on the wall of his lab. time as a succession of present momnts exists for the experimenter

    but in the mathematical description of the subsystem happening inside the box, time does not exist this way, it is geometrized as if on an x-t graph plot, a trajectory in the phase space or unchanging space of possible configurations. with time geometrized there is no longer any way to distinguish a present moment---the stuff in the box has no special present, only a trajectory. Quotes here from Einstein, who worried about this.

    The name of the game for the experimenter is to describe and predict and explain what happens inside the box, using some eternal laws and whatever initial conditions he set up when he prepared the box.

    Time as a succession of present momnts has no place in the Newtonian schema that is used to describe the subsystem. This time-honored schema is a pragmatic success in dealing with isolated subsystems separate from the observer.

    A point made in Chapter A is that the schema leads to irrationality when applied to the whole universe.

    What were eternal laws supposed to be doing before the universe existed?
    How were initial conditions selected?
    What is the status of the points in phase space which will never be visited? Of trajectories that remain forever untraveled, universes that might have been but aren't? etc etc.
    Why these particular laws? Why these particular parameters?
    What is the practical difference between a law and an initial condition?

    Irksome and vexing questions are listed on page 32.

    Chapter A blows away the Newtonian setup when applied to the universe as a whole.
    Page 34:
    " We take attempts to apply the Newtonian scheme to cosmology as fallacious, as being attempts to take a theory outside of the domain in which its elements have operational significance.

    We call this the cosmological fallacy, namely:
    to apply to the whole universe a methodology that is only suitable for studying small subsystems of the universe."
    Last edited: Oct 7, 2008
  6. Oct 7, 2008 #5


    User Avatar
    Science Advisor
    Gold Member
    Dearly Missed

    I'm going to switch over now and work a while on Chapter B. I can see a bit of how the book is going to work and it looks like it will be a major influential book, and provocative, because it will take clear positions on puzzling issues that many are not clear about.

    Just looking at the Chapter B slides (see earlier post for links) we have the observation on page 8 that physics used to be about finding what are the laws? but now it is shifting towards being about why these laws? and why these parameters?
    What chooses the values of the parameters?
    What chooses the gauge groups?
    What chooses which fermion representations? (page 10)

    and page 11 mentions the problem of extreme ratios, ("hierarchy" puzzle)
    these approximate equalities (could write ~ instead of =)
    protonmass/planckmass ~ 10^-19
    hGLambda ~ 10^-120
    topquarkmass/electronmass ~ 340,000

    page 12 mentions that there is only a small volume of parameter space where the parameters give you a broad range of stable structures such as for example:
    long-lived stars
    ~ 100 stable atomic nuclei
    complex chemistry

    page 13 gives some details such as we must have
    neutronmass - protonmass < 18 MeV
    alpha < 1/10
    alpha_strong at least half the present value

    pages 13-15 give some 20 inequalities like that, showing how restricted the volume or what the extent of fine tuning is

    He then outlines FOUR possible responses
    1. unification will solve everything:
    there is a unique theory which will unify all the forces and it will specify that all the standard model parameters have to be what they are, and it will make unique predictions of new phenoms that will allow it to be tested---a unique testable TOE will answer our questions.

    2. and 3. are not predictive. he describes them in some detail. they are not the usual testable empirical science

    4. involves evolution of the parameters of physical law--apparent finetuning explained by reproductive fitness, a kind of optimality condition.
    the optimality hypothesis is to some extent falsifiable, certain discoveries could show non-optimality and shoot it down
    and it can explain some of those 20-odd tuning inequalities (why the parameters happen to fall in the ranges they do.)

    Page 68 summarizes the derivation of predictions, and gives three testable predictions, from the #4 option here.
    Here is a related PF thread
    discussing a paper which appeared earlier this year.
    Matrix universality of gauge field and gravitational dynamics
    Last edited: Oct 7, 2008
  7. Oct 7, 2008 #6


    User Avatar

    Thanks for digging up the additional links! I skimmed chapter B, the mp3 and the slides. I didn´t realise from the other talk there is actually a book coming.

    The whole concepts of evolving/emergent laws are closey related to my personal take on things which is why I enjoy reading about these ideas. I conclude though that my personal angle on this is somewhat different than Smolins, though I see many analogies.

    To me the whole concept of evolution, is very close to the concept of inductive reasoning. And the logic of evolution bear some analogies to the logic of guessing and the logic of correction.

    In a typical evolutionary context, survival and reproduction is what provides selection. Smolins ideas is to consider the generation of black holes a trait and that universes that spawn alot of new universes are the viable ones and thus become to dominate?

    I think that is an interesting idea, although my personal idea of selection is that of consistency. Consistent "systems" are preserved, and reproduction can be thought of as indirectly broadcasting your "opinion" into the environment, causing a selection on other systems to "comply". Inconsistency is thus relative to the environment. Here Smolin is suggesting a more concrete idea.

    My vision would treat the evolution of laws (universes) and dynamics in hte universe (as per given laws) on a uniform footing, the different could like in an assigne inertia. The inertia would work like confronting eitwo inconsistent viewpoints, either you revise your opinion to comply to your local environment or you are powerful enough to partially distort the local environment. Here the black hole concept seems to correspond to a special case.

    So Smolins idea seems interesting although I suspect there is different possbile directions along the spirit of evolving laws, possible differing in selection principles. The black hole generation seems like a very complex assumptions to be a fundamental principle?

    I think the evolutionary-law idea can also fit nicely into the idea of the laws of nature beeing self-preserving structures, that can provide evolution even withing a single uinverse, without the need to spawn universes in black holes to make progress. I am personally leaning closer to thta idea.

  8. Oct 7, 2008 #7
    Thanks for the links Marcus
    I'm intrigued Lee Smolin is looking in this direction.
    It seems he has recognized the incredibly sublime error that has persisted and permeated all of physics since Newton - Physics has failed to understand there is a larger dynamic that reveals nature's plot, and instead has spent many years fascinated with each frame of the movie.
    Time is an integral, dynamical and conjugate property of physicality.

    There is an underlying philosophical distinction necessary to understand RSPM as a framework for physical theory.
    The "reality" of the succession of present moments is somewhat of a self contradiction unless it is understood that the "reality" is in the "collective succession" not in any one, mathematically structured "moment" or "slice" or "layer".

    Neither kinematics nor dynamics have lawful "meaning" in the singular "slice" or "layer" in which they are mathematically framed except with respect to any other singular "slice" or "layer"(enter Lorentz). As strange as that may sound it is IMHO the crucial realization necessary to finding cosmologically relevant and persistent laws.
    In analogy: When framing laws in mathematical statements as "slices" at, or differentials of time, Physics is looking at a chess game "AS" and "VIA" snapshots of the board between each move.
    A limited relation of each piece to its past can be found by this method, a finite probability in the number and proximity of squares,(enter QFT) but the actual "LAW" of each move dictating the position of each piece can never be understood from these snapshots. The physicist must understand the dynamics "OVER" time "AS" time, to understand the "LAWs" of motion for each piece before they will ever have any idea "HOW" to play chess and most importantly the objective of chess.

    The problem of time is that it has yet to be understood as a fundamental dynamic.
    Slide #16 - Einstein - "... there is something essential about the Now which is just outside the realm of science."
    That is the problem in a nutshell.
    What is just outside the realm of science, is outside because science frames nature as lawful in the "Now". It is "Outside of Science" because science has insisted on doing physics "IN" the Now.
    All equations hold good in the "slice" of time they are framed. But no deeper, fundamental realization of principle will come until science recognizes the need to frame laws "OVER" time not "IN" time.
    Time is, Time persists, and all that persists, persists and arises from the dynamic of Time.
    The question is not what is nature in a slice of time, but what is time that it frames nature?
    When we answer this, we will be prepared to understand nature.
  9. Oct 8, 2008 #8


    User Avatar

    If I read you right(?) I agree here.

    I think of time as a relative measure of distinguishable change. I am somewhat doubtful in the uniqueness of global fundamental time or universe time that Marcus mentions. I can´t help thinking that this so called global time is emergent. I don´t see a fundamental reason in this conflicting with the evolutionary idea, as long as you note that any measure of time is relational.

    If i understand smoling right, he seems to simply note that looking at some of the more popular attempts so far, those making use of a kind of fundamental time has in his view been more successful. That might be, but I don´t think that a lone is a convincing argument. There are many things that IMHO suggests a subjective observe dependent time where the objective connection (that we see in GR) must be emergent. This becomes IMHO even more plausible when you point out that the only thing we can speak of regards our KNOWLEDGE of any possible objective connection. Sure one can picture that there is one (in some classical realism sense) but that is ugly logic IMO.

  10. Oct 9, 2008 #9


    User Avatar
    Science Advisor
    Gold Member
    Dearly Missed

    Fra, Crisc,
    the thought occurred to me today that we could be considering eventual pedagogical consequences of this evolutionary point of view. Like when a young person learns TRIGONOMETRY, should they also be told that this nice useful rule about 180 degrees was the result of evolution?

    In the early universe, with more spatial curvature, maybe it wasn't true that the angles of the triangle always would add up to 180 degrees. Probably we can't be sure how they would add up on any given day of the week.

    But by now there has been so much expansion that it always works perfectly, or nearly so.

    The same with conservation of energy and other nice laws.

    Shouldn't we begin to give little hints to our children that some of these are not eternal absolute Laws, but instread are practical and useful regularities which have come about by some dynamical process which can be understood and talked about?

    Sometimes I think of the analogy that we are like the fleas on an antelope who hop around studying and making maps of the antelope---they try to understand why she is built like that, and why does she have such long legs. And one of these fleas, a kind of Darwin, has the insight that the antelope must be built for speed---and he infers that we fleas must live on a typical antelope because the antelopes that can run fast have been more successful in replicating themselves. (reproductive success is something that fleas understand very well, so in effect he just applies this well-known idea to the universe as a whole. only in this case the universe is an antelope.)
    Last edited: Oct 9, 2008
  11. Oct 10, 2008 #10


    User Avatar

    I for one think the concept of learning and evolution are naturally combined and it´s even my personal main source of inspiration since I think that the logic of reasoning and the logic of physical law and it´s evolution does have common denominators.

    So some increased awareness of that all knowledge is evolving is I think a great idea to to teach. But maybe it´s difficult to know at what age a typical kid is mature enough to understand these higher abstractions.

    As far as pedagogics i concerned my personal experience has always been that the number one factors is motivation. A teaching procedure where the students are presented tentative answers before they are presented the main question is I think not optimal. And in this sense, the apparently universal motive not only for humans but also for physical systems and matter to survive, improve and persist is something that is the ultimate motivator. This motive is someone closely relate to the apparent never neding "why" questions, which if properly interpreted is not necessarily a meaningless philoosophy question, it can be interpreted as a affinity for further improvement. To understand this in the deepest sense, one might try to abstract what this improvement is.

  12. Oct 10, 2008 #11


    User Avatar
    Science Advisor
    Gold Member
    Dearly Missed

    Hmmm, you may be right about the pedagogical point. First teach the laws as if they were absolute and eternal---because for all practical purposes they are that.

    Only later introduce the idea that they are contingent and the result of some dynamical process. That sounds better, now that you point it out. Don't make the teaching too complicated at the beginning.
  13. Oct 10, 2008 #12


    User Avatar

    I am doing some weird thinking at this point, but I am not sure I would have appreciated this myself when I was say 15. I probably wasn't mature to appreciate the complexity.

    the logic of guessing - this is in essence the problem of given what you "know" or at least what you "think you know", what actions to take - it's like a action formulation based on the present state of affairs

    the logic of correction - this is the problem of how to revise the state of konwledge in response to feedback. Here the evolutionary perspective comes to rescue, give that you know something, and then suddently there is new evidence thrown into your face, inconsistent with the previous state. Then at this point, the main question is not to conlude that the previous opinion was wrong, the main problem at hand is howto merge the previous opinon with the new information.

    Sometimes this correction can be made by adjusting say parameters in an existing model, but what happens when there is not parametrization that yields consisntecy? Then one needs a strategy to deform the "state space itself".

    It was first when I have studied the first courses in QM, and started to see the simplistic realism of classical mechanics that I personally started to elaborate on some deeper abstractions. Maybe those "basic classes" was the food for thought I needed to grow my own questions. If someone had pointed this questions out to me prior to that, I am not sure I would have understood them.

  14. Oct 10, 2008 #13


    User Avatar

    I think this illustrates the downsides of two extremes.

    OTOH, some people argue that if the state space needs to expand of change, then one could simply from the beginning picture a bigger "superspace". But then the obvious problem of performance enters. You easily get lost on this superspace if you do random walks. This measn that it isn´t effieicent.

    There is also the other obvious problem of representation. From the perspective of the holographic principle, how does a bounded observer represent a giganti superspace of possibilities? one can imagein that this easily require more information than what possibly constitutes the observer itself. This to me suggests an upper bound to the "complexion number" of any statespace.

    The other extreme, to never extrapolate, means that on EACH feedback the opinion needs to be severly revised, and that basically every input is "unexpecte". Thus the other extreme gives no predictive guides.

    I think there is a balance. And this is the essence of evolution. The balance between representation and fitness. A supermassive structure is for the above reason not coherent, it will "get lost" to use a metaphor and thus destabilise.

    This is the game I am trying to understand. How to formalise this. I think the holographic principle and information bounds together with an evolutionary vision are hot leads here.

  15. Oct 10, 2008 #14
    Marcus, I think you're correct in assuming the eventual pedagogical consequences will inevitably redefine the nature of physics and therefore how it is taught.
    No matter how isolated we attempt to keep the focus and scientific method of physics, as a truly grand theory emerges, there is no falsifiable way to isolate the laws from the immediately obvious "fact" that we, as humans are the stuff of the subject we aim to define. The classification of sciences is easily sustained between chemistry - biology - physiology - psychology - etc, but when one arrives at physics - the buck stops here so to speak.
    There is no other science in a hierarchy of knowledge that takes the baton from physics, so whatever we posit as the first manifestation of physical reality above metaphysics must logically lead to the intelligence of life.

    I have also done a great deal of thinking in the last few days on the premise of Smolin's and Unger's book.
    What struck me more than any other point they make was the concept of eternal, immutable law.
    How incredibly naive we've been to fall into the same trap over and over again. Today we find an almost childlike
    perspective in the notions of "absolute" space and time that were so well received in Newton's work.
    Yet now we see once again the "absolute" nature of the laws we defend as our highest order of accomplishment in physics.
    I look at all the laws over and over again and I see an as yet undefined common framework.
    It is as if physics has been cultivated in a language insufficient to express itself.
    All of the laws and all the principles supporting them must ultimately be realized as small incoherent phrases, individual stabs at defining a fully comprehensive sentence. No law can stand independent of another, no principle is unique, they are each analogous to the childlike attempts to express an idea of intuitive comprehension but lack the maturity and experience to assimilate into a comprehensible statement.
    The framework I see is very much like that discovered by Noether. Each law is an extension or small handle on a core concept that frames them all. Conservation and symmetry are expressions of the same thing. What is that "thing"?
    IMHO unification will not be found in the geometrics of quantization, but in the same manner of insight Emmy Noether had. Unification will come from realizing the framework for the unification of the principles and laws, the model will follow.
  16. Oct 10, 2008 #15
    Excuse me If this comment is too far off point, ignore it if it is. It almost sounds as if this is edging towards a "Bohmian" type interpretations or redefinition of principles. If so how does it avoid drifting into unchartable (so to speak) waters ?
  17. Oct 11, 2008 #16


    User Avatar

    It is difficult to convey a way of seeing the problem, but I do not adhere to any Bohmian interpretations. However, if you insist, there might be a remote connection to a sort of bohmian type of "hidden degrees of freedom" and "subjective degrees of freedom".

    But I never ever even remotely see myself as bohmian.

    My vision is the opposite to hidden variables though, but I take serious the fact that the degrees of freedom is relative. But my motivation is totally different from the bohmian ideal who seems to try to restore realism. I am suggesting the even the realism present in normal QM is an illusion. QM is still a kind of deterministic theory since the wavefunction is evolving deterministically.

    The difference is that while bohmian introduces the notion of "hidden variables" I instead see it the way that hidden degrees of freedom are NO degrees of freedom. But OTOH the degrees of freedom are only the distiniguishable part of the world, and the degrees of freedom is only a relation to an observer. So I am suggesting that there is no objective definition of universal degrees of freedom.

    On the contrary to I take the charting process seriously, and ponder, what the physical constraints on the observer is, on the charting process and the process of constructing distinguishable measures.

    So the impression of bohmian stuff is due to my poor way of presenting my reasoning. I´ve got no secret bohmian dreams :)

  18. Oct 11, 2008 #17


    User Avatar
    Science Advisor
    Gold Member
    Dearly Missed

    Both video lectures have a key quote from Charles Sanders Peirce including this sentence:

    "Now the only possible way of accounting for the laws of nature and for uniformity in general is to suppose them results of evolution."

    I tracked this down and got a slightly longer excerpt than what Smolin uses.

    "To suppose universal laws of nature capable of being apprehended by the mind and yet having no reason for their special forms, but standing inexplicable and irrational, is hardly a justifiable position.

    Uniformities are precisely the sort of facts that need to be accounted for.

    That a pitched coin should sometimes turn up heads and sometimes tails calls for no particular explanation; but if it shows heads every time, we wish to know how this result has been brought about. Law is par excellence the thing that wants a reason. Now the only possible way of accounting for the laws of nature and for uniformity in general is to suppose them results of evolution."

    This is quoted on slide #3 of the set I am calling chapter A
    and on slide #44 of the set I refer to here as chapter B.
    Last edited: Oct 11, 2008
  19. Oct 11, 2008 #18
    My mistake. I was overgeneralizing. I was more generally referring and comparing to later (70's onward) Bohm attempts at a meta-structure ie through holographic projection of a deeper (seemingly undetectable) ordering. It just struck me that it sounds as if Smolin is flirting with similar meta-ideas? I have to save my speculation until I have an opportunity to read it.
  20. Oct 11, 2008 #19


    User Avatar
    Science Advisor
    Gold Member
    Dearly Missed

    a civilized degree of thoughtful speculation can help stimulate conversation. Not always bad. In this case you may want to look at the video lectures, or the PDF slide sets, that are linked to in the first post, and not wait for the book. The book is in progress but I understand it will be a while yet. I'll post as soon as I know more.
  21. Mar 18, 2010 #20


    User Avatar

    Does anyone have any updated detailed expectations on when this new book will be released?

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