Sharp split on the time issue (Rov. vs Smo.)

[marcus]

Carlo Rovelli and Lee Smolin have recently taken diametrically opposed positions on whether time is a basic part of reality or merely an emergent feature----like for example temperature, or the surface of a liquid.

The surface of a liquid is well-defined at our macro scale, but if you look closely it does not exist.
Likewise the temperature is only meaningful at macroscopic scale, when the system has a lot of degrees of freedom. It is said to be an emergent property of matter, not a fundamental property.

By analogy, time might be emergent. It might not exist at a fundamental level. Rovelli thinks that at the fundamental level there are only correlations between observables----several of these observables can be used as alternative clocks, and the others (like position etc.) can be correlated with them. But he questions whether there is a ONE RIGHT CLOCK. Maybe there is no one unique big clock that we can distinguish as the master timekeeper. And so time, in the abstract, does not exist. There are only many more or less good quantum observables that you can use, and correlate other observations to them, to keep track.

Rovelli develops this viewpoint in an August 2008 essay that FQXi has put online.
"Forget Time"
http://fqxi.org/community/forum/topic/237
when you click on PDF you get the full 7 page essay
http://fqxi.org/data/essay-contest-files/Rovelli_Time.pdf

This idea that time is not fundamental, does not exist at the level of basic description of nature, seems pretty reasonable. But Lee Smolin argues persuasively against it in this video lecture.

The Reality of Time and the Evolution of Laws.
Here is the main page, with links to the online video lecture and the slides PDF
http://pirsa.org/08100049/
Here is just the slides PDF, if you want to read the gist of the talk.
http://pirsa.org/index.php?p=media&url=/pdf/files/68863c39-f22d-430f-991d-bb8d224440ad.pdf

He agrees that you can get rid of time for describing all subsystems of the universe, but you cannot, he argues, dispense with time in a fundamental description of the universe as a whole. This is a subtle and interesting point. As long as there is an outside observer looking at an isolated subsystem, as if in a laboratory, there is always the observer's clock---the clock on the laboratory wall. But how to deal with the situation where there is no isolated experiment, no outside observer, no setup with initial conditions, no reliable eternal laws but instead the natural history of this one unique world with more or less regular patterns of behavior. If we picture this as resembling the laboratory experiment, which we can repeat with different initial conditions, then (suggests Smolin) we are fooling ourselves. So, he argues, we cannot get rid of time in the way we are used to, as if inside the box of a subsystem.

 

[MTd2]

Marcus, did you read the discussion on this thread https://www.physicsforums.com/showthread.php?t=239414 ?

 

[marcus]

That thread refers at one point to a 1996 paper of Rovelli on Relational Quantum Mechanics. that is a bigger topic, by far, than what he has to say specifically about time in the essay we are talking about here.
As I recall we have had several threads here at PF about RQM. Especially triggered by Rovelli's paper with Smerlak from around 2 years back. I started some of them
But I don't think it is focused enough on the topic of this split between Rovelli's and Smolin's views.
To get back on topic, how do you come down on this issue? Do you side with Rovelli or with Smolin, on the question of the reality vs emergence of time?

 

[Fra]

To throw in my personal view on this.

First it's difficult to deny that both Smolin's and Rovelli's reasoning contains attractive points. Both of them are IMO good thinkers and inspiring writers.

I don't think that it has to be either or though, perhaps they both see the issues from different points, and that there is a way to both have emergent time but also have evolving and emergent laws. This is what I personally think.

I think Smolin's argument in favour of evolving law is good in general. From an information point of view, information about law and information about initial conditions really must be treated on the same footing. Here I am totally with Smolin.

But I don't share his reasoning that this implies that time must be some objective succesion of moments.

I can share some concept that time in a certain sense, must be fundamental, but the question is if he is picturing a global objective succession of moments, or some subjectively defined time? If he talks about some objective succession of moments I disagree with smolin and more like Rovelli's argument.

I get the feeling that both can make sense if you make the following connection.

The fundamental need for time, to parameterize the evolution of law smolin argues in favour for, could IMHO be subjective. Ie. it's fundamental once you single out an observer. And since the observer is part of the universe, there is always and observer. And this the symmetry idea implicit in the timeless thinking is broken by the choice of observer. And this is unavoidable.

Then the connection to Rovelli could then possibly be that this now apparent sea of possible times depending on the choice of observers, develop a relation, and this RELATION is only emergent in the process of communication/interaction with the observers (RQM style communication). And the result of this communication is that there is feedback that provides a evolution and selection among the population of observers. Thus the result is emergence among the population (observers) - and thus LAW, because the concept of LAW is manifest in the properties of the observer-population.

As I understand it, this is not how neither smolin nor rovelli picture it, but it's my interpretation of how I think this works, and it at least partially bridges smolins and rovelli's views. So that both their points are partially valid, but there is a bigger picture wher the relations are seem more clearly.

/Fredrik

 

[MTd2]

To get back on topic, how do you come down on this issue? Do you side with Rovelli or with Smolin, on the question of the reality vs emergence of time?

If you think the world emerges from an eucledian-like lattice, then:

In that lattice, I think the time should be absolute, parametrical, that is, there is reality to it. In that case, I tend agree with Smolin.

The emerged world is essentialy timeless, it doesn't have a time dimension, but 4 dimensional axis, like in GR, but are due to a minimized action, you have world line, in which you can define a clock. In this case, I tend agree with Roveli.

The tick-tack of a clock is local, and you know that in the real world, in that frame, and just in that frame, you relate the tick tack of a clock, the time taken for light to cover a constant and linear path, with the tick tack of an atomic clock, that is, the average time it takes to the system to update itself, that is, to make a constant averaged great number of state transitions.

I don't think both of them are wrong. I guess each other helps to understand wilczek view of things, although I haven't read his book, so thiss is just a guessig.

 

[atyy]

Rovelli's essay seems quite readable, but I don't understand Smolin's argument at all. It seems that Rovelli bases his argument on the coordinate time in General Relativity not necessarily having the meaning of a universal direction of time against which everything evolves. It's almost as if he already considers Newtonian time emergent. If that is the case, isn't Smolin's argument already wrong, since General Relativity seems to be capable of describing our cosmos, of which there is only one? Does he think that General Relativity is fundamentally unsatisfactory because of its singularities, or is there a different idea of emergent time?

 

[p-brane]

They`re both wrong, as usual.

 

[marcus]

Rovelli's essay seems quite readable, but I don't understand Smolin's argument at all. It seems that Rovelli bases his argument on the coordinate time in General Relativity not necessarily having the meaning of a universal direction of time against which everything evolves. It's almost as if he already considers Newtonian time emergent. If that is the case, isn't Smolin's argument already wrong...?

I'm glad you read Rovelli's essay. What about watching the video?
http://pirsa.org/08100049/
I think that is probably the only way to see what Smolin's argument is, so that one can compare the weight of the arguments and judge which is the more persuasive.
Reading quickly thru the slides would give some idea, but he has additional stuff to say that connects the slides together.

========================
I'd like to read your comments on Smolin's talk after you have watched it. In the meantime, I think what you say about Rovelli's essay is right: in the real world there are only more or less inaccurate real clocks, which themselves are quantum observables. One is forced to choose some particular observable to be the clock and then to correlate other observables (like position) with whatever one has chosen to be the clock. There is no absolute time in nature. So in a fundamental theory there can not be any absolute time either. Am I missing somethng? I will go back and take another look.
(And then there is the business of how some useful ideas of time can emerge in certain situations, like the way temperature emerges though it does not exist at the micro fundamental level.)

I think you are right that Rovelli's essay is readily accessible. Smolin's argument is more difficult to grasp---one has to work a bit more---but still I think interesting, and the conclusions could turn out to be valid.

 

[atyy]

I'm glad you read Rovelli's essay. What about watching the video?
http://pirsa.org/08100049/
I think that is probably the only way to see what Smolin's argument is, so that one can compare the weight of the arguments and judge which is the more persuasive.
Reading quickly thru the slides would give some idea, but he has additional stuff to say that connects the slides together.

Yes, I'll watch the video, hopefully that'll be more understandable. One thing that bothers me a bit about Rovelli's argument (if I've understood it) is that the light cone structure at every spacetime point is absolute, and doesn't that provide an absolute notion of time, or at least of causal structure?

 

[Fra]

One is forced to choose some particular observable to be the clock and then to correlate other observables (like position) with whatever one has chosen to be the clock. There is no absolute time in nature. So in a fundamental theory there can not be any absolute time either. Am I missing somethng?

When reading the various comments the following strikes me, that there is a difference between, the "rate of time", and the "direction of time".

The rate of time as measured by a choice of clock and is hardly fundamental, not even in the subjective/observer-dependent view - I agree. But the direction of time(ie the arrow of time) OTOH is fundamental in the subjective view of an observer, but still not fundamental in the observer independent view - in this view, the choice of the observer is "arbitrary".

But the real question IMHO, is wether this observer independent view is a sensible physical view? Because clearly, everything is seen from the perspective of SOME observer. There really isn't a way around it, is it? right? And in that sense, the arrow of time is there to provide the direction of evolution of law.

This is how I personally make partial sense of smolins argument.

The problem with GR, is that what seem to correspond to the "physical observables", are "arbitrary" from the point of view of the invariants of GR. The problem seems to be that the observables are a matter of choice of observers.

This was also elaborated in Gambini's paper.
http://fqxi.org/data/essay-contest-files/Gambini_essay.pdf

"In technical terms, what one can consider as physical observable in general relativity is a quantity that is left invariant under the symmetries of the theory, or in the canonical language, that commutes with the constraints. Since one of the constraints is the Hamiltonian, physical quantities do not evolve. Therefore they cannot work as clocks"

So the paradox is that what is fundamental in one view(ie what real observer actually measures) is arbitrary choice in another view (the view of objective relations between the arbitrary choices).

/Fredrik

 

[Demystifier]

They`re both wrong, as usual.

You must be a string theorist.

By the way, does anybody know a true string theorist (Smolin is only a half string theorist) exploring the nature of time on a deeper level?