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

Lee Smolin's metaphysical principles

  1. Sep 19, 2009 #1

    I understand he is arguing against multiverses here, but is the tone of the article meant to be sarcastic?

    His four principles just seem to be obvious statements.

    I thought that the "timeless universe" idea was more meant as "a timeless model of the evolution of the universe idea."

    In relation to his analogy to chess, while any chess game is only occuring in a series of presents moments, chess games can be modeled successfully on how they evolve.

    What is his point here?
    Last edited by a moderator: Apr 24, 2017
  2. jcsd
  3. Sep 25, 2009 #2
    That time is fundamental.
  4. Sep 25, 2009 #3
    How could time be fundamental when time is non-sensical unless tied to a particular frame of reference? What happens to this alleged 'fundamental' time at the speed of light or at the centre of a black hole or at the planck scale? Do you have access to the whole article?
  5. Sep 25, 2009 #4
    What makes me, me and you, you is not that we occupy different space but the fact that we exist as different times, tied to two different frames of reference. Read your first question. Every fundamental particle is the center of its own frame of reference, with its own variation of time dependent on its own relative motion in its own present. Time is the variable in the form of one that makes a black hole possible. As for planck's scale, time is the smallest entity on this scale making it the least common denominator of reality. Time is also reality's largest common denominator with everything that exists within our visible universe nothing more than a subset of this global time. So you tell me what is not fundamental about time?
  6. Sep 25, 2009 #5


    User Avatar
    Gold Member

    I think you have to read this as part of the loops vs strings struggle. Smolin is saying timeless approaches like string with its platonic landscape are in principle untestable. Which means strings are already a dead-end scientifically. So everyone better come over to loops where realities evolve in time and so there may be only one reality and thus one testable package of laws.

    I didn't find Smolin's paper either clear or convincing. But I guess that is because time-ful vs time-less seems an obviously false dichotomy. I would expect time itself (along with space) to be part of any "lawful" package that develops.

    To have a physical reality, it seems we would have to have both stability and flux (a proper dichotomy, being a mutual, asymmetric, complementary parting of pure possibility). Or in other words space and time. Spatial dimensions are about the existence (or persistence) of stable locations and the temporal dimension is then a global measure of the range of change, the degrees of freedom, that still remains in a system with stable locality.

    So time need not be considered a simple dimension - a straight space-like line. Instead it could be something far more complex.

    Reading Smolin, it is like a philosophical discussion taking place in an old school Newtonian setting. Did time always exist or did it have a beginning? QM and GR already move us into a systems view of time - one spelling out the dichotomies (or complementaries), the other the equilibration of the separated facets of existence (space and time made one again).

    So as Wavejumper says, the Newtonian perspective that appears to inform Smolin's argument (against his string competition) does not seem sufficiently modern. It seems a backward step for someone who is good at pushing the new.

    From a relativistic perspective, for example, it would seem to me that "standing still, being located" is no longer about being at absolute rest within a global space of possible change (as in Newtonian space-time). Instead, it is lightspeed radiation which sets the baseline. And then mass appears as locations which create actual time - create real opportunities for change - by going slower than light speed.

    So the relativistic realm is vanilla - time has no rich meaning as everything happens everywhere at the same rate. And then when mass condensed out with cooling, suddenly you had new local degrees of freedom. You had massive locales that could travel at a full range of speeds (between zero and c). So within a cooler universe, you get a phase transition in which time, as a yardstick of change, suddenly becomes more richly developed. Less simple, more complex.

    We know this to be the case for our universe. So it seems fair to extrapolate it a general developmental principle to the origins of the universe and beyond.

    This would mean that the relativistic early universe with its simpler temporal structure would have arisen as a phase transition from a realm of even higher symmetry (greater spatiotemporal simplicity).

    So time, in some sense, always existed. There must be some version of space and time, stasis and flux, built into the beginnings of everything (technically: a potential or vagueness). But time would be something capable of development via phase transitions. And so debates about whether reality needs time as a pre-existent background, or whether realities are actually timeless, both miss the middle ground solution - the way time itself would develop as a crisp "dimensionality".

    Now the big question shifts to whether time had to develop to the one kind of time we see, or could many kinds of time evolved? If we asked the same of dimensionality, for example, it seems a legitimate question to ask why not two or 20 spatial dimensions, instead of three extended (and perhaps 6 compact) we find?

    Does a developmental approach produce one kind of time eventually (a platonic view), or do we have to get anthropic here? So perhaps - whether you stick with Newtonian simplicities or get more sophisticated - the essential issue raised by Smolin does not change.

    A background-producing theory is better than a background-independent one. And the reason is that a background-producing theory is going to also have to model the selection mechanism that constrains the development of fundamental features of a physical system, such as stasis and flux (or spatial and temporal degrees of freedom/degrees of constraint).

    The only way a background-independent approach can win is if it has only one possible solution - if it does not require the constraints of selection because only one solution can be constructed. And in its present state, strings is a landscape of choices.

    Anyway, I think Smolin's paper makes sense if you treat it as merely part of a local academic skirmish - a chance for sly public dig at his competition - rather than an attempt to present any really new ideas about the "time" in which realities must develop (and the way time itself must develop to become more "real")
  7. Sep 26, 2009 #6
    I guess if you can take one of our four dimensions out of the equation with strings then a theory that does not follow the mass as with the multiverse would be ok.
  8. Oct 22, 2009 #7


    User Avatar

    I have not read that specific article but I personally think the better focus is to say that his point is first of all an argument against the realist view of eternal physical law.

    As he has tried to explain reasonably well in various places the notion of timeless law works much better when we study subsystems from a controlled environment, than for studying an effectively open system.

    The reality of evolution of law, implicitly contains a evolutionary "time" line. But this should not be confused with newtonian style absolute clock-time and space. It is not what he is talking about.

  9. Oct 22, 2009 #8

    Physical law is not going to go away. Each universe might have local bylaws, but the multiverse itself would be govern by a metalaw, or a physical law.
  10. Oct 22, 2009 #9

    Give me a try:

    He is trying to hold on to the traditional notion of physical laws as dynamical laws.
    That is, what things do as things evolve in time. This is as oppose to finding the legrange, and minimizing the action, or finding transformational group for some ultimate theory such that the parameter of time would come out as a particular realization in a universe of an ensemble of universes. See the difference? In the latter case, time pop out as a particular realization of a particular equation.
    Last edited by a moderator: Apr 24, 2017
  11. Oct 22, 2009 #10
    Sort of, but as i was commenting in a dfferent thread, it seems like more of a semantic argument. It seems sort of a priori that you need time and causality for conscious observers. If there is some mathematical way to think of time that makes the rest of the universe make more sense, then so be it, but that doesn't change what time "is" from the human perspective.
    Last edited: Oct 22, 2009
  12. Oct 22, 2009 #11


    User Avatar

    IMHO - not necessarily.

    My personal opinion is that there is no metalaw, part of the evolution is constrained, but parts of if it is simply undecidable.

    About Smolin, I think he is inconclusive. In the talk he gave on this close to exactly a year ago:

    "On the reality of time and the evolution of laws"
    I would only guess that this is better than the physicsworld article.

    a guy in the audience raises the question wether there is a metalaw governing the evolution of law (or universes if you think in terms of one set of law = one universe) and he responds that he doesn't know.

    My opinon is that the answer must be no, there is no timeless realist type of metalaw.

    (physical law sure doesn't go away, but it evolves with the universe. Evolutionary evolution is different from deterministic dynamics). The "dynamics" of physical law is not known to an inside observer.

  13. Oct 22, 2009 #12
    Ok, say you have a quantum foam, or a random number generator, or a million monkeys on a million typewriters writing all the possible laws of the universe(s). There still has to be some generator of some kind. And impossible combinations would still produce impossible results. If there's no meta-laws even governing what makes sense, then we might as well stick to the logic of THIS universe which says there are universal meta-laws, because the universes that don't make sense but make sense if you were in them (or not) are completely inconceivable to us.
  14. Oct 22, 2009 #13
    No, law like regularities can` t just be explained away from physics.
  15. Oct 22, 2009 #14


    User Avatar

    Did you listen to the talk?

    I don't wnat to start a lenghty argument here but, from my point of view (forget smolin for a moment) the problem is the limits of the context, that you need to make an inference or abduction of laws. This means that it's not always possible to make a perfect deduction of a perfect regularity, this is IMO why the inferrable regularlities evolve.

    If you picture non-inferrable regularities or laws, then you have a realist view of law. I'd say that view is more common than mine, so there is no need to explain it. The problem I see is that only inferrable information influences the action of a system. So I really don't see the predictive power of a non-inferrable realist picture.

  16. Oct 22, 2009 #15
    I don't think you guys are disagreeing per ce, you'rem just talking about slightly different concepts.

    frA is arguing that the only meaningful laws are those that can be inferred, and that these will not necessarily hold true at all times in all such situations.

    vectorcube is saying that you can't do away with the notion of physical laws, perhaps thinking that was what fra was implying.
  17. Oct 22, 2009 #16
    Look, i know about the "Laws of nature". I know the "literatures", and modern debate on it.
    Laws of nature cannot be explained away from appeal to evolution.

    I am working under the assumption that you know the modern debate:

    As you know, the realist view of law comes in many variations. One of which is the Armstrongs view of law as a nomic relation N, between universals P, Q such that N(P, Q) relation hold. In the armstrong view, the relation N is a contingent, but necessary relation between universals. Suppose this is the case, then it follows that there is a possible W such that M(P, Q) hold, where M is not equal to N. This is enough to allow some philosophers to postuate all sorts of worlds, with all sorts of contingent, but necessary relations in those worlds to hold.

    In the discussion about modern conception of a multiverse, there would be still be regularities in the level of the multiverse, and that this regularies( we called physical laws) would hold in each individual universe. The question of why the physical law in our universe obtain, would shift to the question of why the regularies, and laws in the multiverse itself obtain. So, as you can see, the ultimate contingent of the world does not become smeller when we opt for a multiverse, but it only shifted the problem up a level.
  18. Oct 22, 2009 #17
    The only major non-realist view is david lewis`s view that nomic facts determind laws, or also known as the regularity view, or the best system view of laws of nature. It seems he is not saying that.
  19. Oct 23, 2009 #18


    User Avatar

    I see your reasoning here and yes there are different versions of realism or structural realism.

    But I'll just say that I don't share this reasoning. I see other possibilities. In my view the "shifting" to a larger level (multiverse level) increases the complexity, but if the overall inside view constrains the inferrable regularities, then what happens is that you loose decidability and it limits the configuration space so to speak.

    The next step for me, is that get consequences for the expected rational action of any system, and it thus makes a difference to interactions. The way a subsystem behaves (it's action) reflects that some things are undecidable from it's point of view.

    I personally think that smolins ideas are not very attractive if you really make the conclusion that there must be a meta law describing the evolution of law or universes. If you take other views, I think the ideas gets more interesting.

    This is why I wanted to note that this is not a unquestionable conclusion. As far as I know, Smolin was inconclusive, by my personal view is that there notion of metalaw is even inconsistent with the very spirit here. So either you reject all of it as baloney, which is fine, and alot of people do that, or you make some sense of it, but i think, thinking of metalaw is not the way to make sense of it for me at least.

  20. Oct 23, 2009 #19
    if you want to read about it, this is a good sources:
    http://www.st-edmunds.cam.ac.uk/faraday/resources/George%20Ellis%20Lecture/Ellis-Faraday.pdf [Broken]

    A personal reason why i opt for realist view of law is that there is a sense, or feeling that things could be different. I feeling that i can understand the world, and the underlying design. That the universe do not have to be the way it is. If physical reality is only, and only such and such constitutes, and laws, then it would be boring. I can ask the question of why those laws, and not others? Maybe other worlds could be defined by different constitutes, and laws. I think it is very cool that i can ask "what if". To me, physical reality is very boring if everything is described by a single equation. I am a modal realist in philosophy, and i do believe in max tagmark` s level lV multiverse.

    A non-realist view do not give you comprehensibility. If you deny the existence of laws, then what do you have in it`s place?
    We use laws all the time to make predictions. What are we left with if you substitute a vague description?
    Last edited by a moderator: May 4, 2017
  21. Oct 23, 2009 #20


    User Avatar

    Yes, I figured something your view is something like that from your responses. There are other tegermak fans on here as well and I don't want to start yet another argument against it.

    As you probably already guessed, I do not share tegemarks reasoning.

    I have tried to explain my conceptual position position to other realists on here before. But I definitely have answers to your questions of "what is in it's place" etc, they challange is to provide an argument in a form that is acceptable, from your point of view so to speak, which has proved tricky.

    I might comment more later, but not that I do not deny existence of laws, I just claim that from the point of view of something that makes a real difference - the action of a system - the only importance is the inferrable laws, and the inferrable laws are relative to the inference system, and thus observer dependent, and thus evolving with the observer.

    Objectivity is simply emergent consensus to the exten possible, in a local group of interacting observers, but there is always a residual disagreement, and this residual disagreement becomes not "relative law" but rather defining the symmetry groups and gauges that are part of describing interactins. Ie. the residual subjectivity is the set of observer frames, and the emergent objectivity is the symmetry transformation that allows for some kind of invariant formulation.

    But even this objectivity is evolving, since it's constrained to the inferrability condition.

    // Note that I am more radical, or perhasp more definite that smolins idea here, but I share the basic arguments of smolin against timeless law

    The predictive power in my view is the fact that the action of such a system, depends on these things. The fact tht a system can not infer a particualr regularity in it's environment, means that systems action is invariant to the choice of regularity.

    I think if you insist in tegemarks terms, my objections would amount to insisting that his mathematics is uesless or lacking predictive power unelss the computability and representation capacity is taken into account. Can you even define the say the set of all mathematics? it makes no sense ot me. Also a mathematical system tht isn't computable by a computer at hand (rather than imaginary computers larger than the universe itself) doesn't offer predictive power. The predictive power comes only when you actually complete a computation. Also the computational time is an issue.

    With inferrability law, I more or less talk about the constraint on "mathematical regularties" that come from the limtis of complexity of the inside observer.

    The size of somebodies brain pretty much sets a limit of what can be comprehended. I think there is an analogy to physical law, to the extent what regularities in the action one system can infere from another system.

    Sorry to not have more at the moment. This is a bried motivation only. The ultimate exposition of this in terms of something that is doing real predictions is still owrk in progress in my part. I'm not aware of anyone else either that has more than fragments implemented.

    As for Tegemarks resoninng, I would similarly ask how tegemarks idea can help solve open problems in physics, and the unification of QM and GR etc. This is the real question. all this prior to that are just motivation in different directions.

    Not sure if that makes anything clearer, probably not :biggrin:

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