Why do stringy people don't like LQG?

[marcus]

Sorry, I do not understand you. Have you really read my previous posts? .

I must apologize! Looking back at my post I see that it seems as if I objected to your idea in post #29. This was not intended.

 

[tom.stoer]

I agree that it would be very helpful for us if there would ba a sound discussion of string- with loop-oriented insiders in this forum (or at a different place in the web). Perhaps the meeting in Marcus' new thread is a good starting point. In addition I would really like to see something like the "outside-view" - "inside-view" discussion Nicolai and Thieman had on LQG.

Every couple of years it's time to put the most relevant topics and open issues on the table and try to collect opinions from a different "school" or different perspective.

Then I agree that different perspectives dio not only provide different approaches and insights but additional technical issues as well. The measurment-problem or the time-problem is certtainly a conceptual one in all QG approaches, but the correct Hamiltonian of LQG is very specific to LQG. In any discussion those issues should be kept seperately.

 

[Fra]

The measurment-problem or the time-problem is certtainly a conceptual one in all QG approaches, but the correct Hamiltonian of LQG is very specific to LQG. In any discussion those issues should be kept seperately.

Perhaps in some ways of analysing this, the separation into separate discussions differ too :)

Now I don't represent any major program at all, but from my perspective the various problems of measurement, time and establishing state spaces and evolution operators are related, I don't see them as cleanly separable.

So different schools of thought (not just LQG and string), might not always agree on the separation of subquestions. I personally question the physical (observable) basis of these constraints, and this is I think related although not directly to the measurement and time problems.

To me the construction of physically justified (construtcible) measures/observables, is one of the keys. And I don't understand how that process can separate to for example the understanding of time. But then I don't share rovelli's logic. I think there is confusion between "timless" notions as in the objectively of this timeless notion. Most objectivity seems to be attached to realist, imaginary constructs, that are not subject to scientific inquiry. Some conceptual framework that is assume to just be.

I think it's true that there is no objective time, but this does IMHO NOT imply that there exists a objective notion of "timelessness". Rovelli's logic doesn't attach a physical status to the timeless information. In other words, it's a realist type of information, needed, but that is not subject to measurement processes. I find this to be inconsistent reasoning which is why I don't prefer it.

/Fredrik

 

[tom.stoer]

I don't say that the problem of the Hamiltonian and the problem of time are not related. All what I want to say is (and perhaps this was not clear from my post) that the specific technical details of the LQG Hamiltonian (quantization-, regularization-and ordering-ambiguities) are essential topics for the LQG approach - they don't affect in any way other approaches (as far as we can see). Another example could be string theory: if the finiteness of the perturbation series for one specific string model cannot be proven this affects only string theory - nothing else.

 

[Fra]

I don't say that the problem of the Hamiltonian and the problem of time are not related. All what I want to say is (and perhaps this was not clear from my post) that the specific technical details of the LQG Hamiltonian (quantization-, regularization-and ordering-ambiguities) are essential topics for the LQG approach - they don't affect in any way other approaches (as far as we can see). Another example could be string theory: if the finiteness of the perturbation series for one specific string model cannot be proven this affects only string theory - nothing else.

I guess what you say is that since the choice of question posed, are part of the various programs, it is not straightforward to draw implications of any answers to other programs where the questions asked are somewhat different. This is exactly in line with what I said too.

This is exactly why I think that before we invest in answering a question, it might be a sensible to first discuss the construction and justification of the question itself (before we can answer it). In this respect I feel that some reasonings, ignore what should be important questions, and I see that as very speculative. The risk is that you spend 40 years trying to answer the wrong question, and as per the logic above, the answer to one question is not easily decoded into answering a almost logically unrelated question from another program. That's why I can not overlook these details. I tend to think they might be very important. This is why I first of all evaluate the framework and the way the basic problem is formulated. Here I do not like the way Rovelli reasons. Since I don't share the framework from which he phrases the technical problems of LQG, then the technical details somehow becomes deattached to my reality.

The first time I started reading his book, I read it in a very positive way, and interpreted it to it's advantage, but I noticed that the picture I made for myself - regarding the physical meaning and interpretation of the spin networks (I thought of them first as encoding the observers microstructure) simply wasn't what rovelli was doing.

I have similar views on string theory, where a string ca be constructed in a high complexity limite as the simplest possible measure. But in that interpretation there is no external space the string is embedded in, instead the space is emergent as relations between different strings. i still think if string theory ever makes it out of mathematical toyery something in this direction is the way I can make sense of it. But if that is true, strings are not fundamental, and probably the paramters of the string would be understood from an evolutionary point and there should be no landspace problem.

Although I can relate to both approaches, my different angle makes my reasoning so different that the wayt of reasoning typically seen in those papers, even more so the plain technical papers focusing on isolated mathematical problemes withing the program, simply are not motivated for me.

If LQG or string theory will be part of the future big theory, then I think there certainly are some similar issues that need to be addressed. But I want to start at what is for me, the right end.

I guess I misinterpreted you before.

/Fredrik

 

[Demystifier]

I think it's true that there is no objective time

Do you also think that there is no objective space? Or do you think that the difference between time and space is much more than a difference in the sign of the metric?

Anyway, since you like to see how other people think, you might find this interesting:
http://xxx.lanl.gov/abs/0905.0538 [to appear in Phys. Lett. B]

 

[Fra]

Do you also think that there is no objective space?

That's right. I don't see the physical basis for objective space either.

Or do you think that the difference between time and space is much more than a difference in the sign of the metric?

Yes it's a bigger difference. To me the two questions you ask are not XOR, I say Yes on both.

In my view, time is related to evolution of information. Somehow time is simply a distinguished direction in the "data processing" hierarchy of evolving memory records. But the distingusihed direction is observer dependent, but I think not in a way that allows a timeless transformation. The "timeless structures" rovelli uses are emergent in my view.

My idea (under development) is that the first glance "inconsistency" that is the result of denying objective time and space, are exactly where interactions are encoded and classified, and moreoever the "evolving perspective" is the supposed key to unification. The rational reasoning behind that is simple; actions of each subsystem are depending only on the local information, thus any incompleteness does result in a different action. The problem with this is that the theory is always in motion, you can not carve it in stone. That's where thus microinformation exchange connects to the cosmological scale and the universe at large. I think this connection is there from square one, and I dno't know to pretend it's not.

Anyway, since you like to see how other people think, you might find this interesting:
http://xxx.lanl.gov/abs/0905.0538 [to appear in Phys. Lett. B]

Thanks, I'll check that later!

/Fredrik

 

[Fra]

My idea (under development) is that the first glance "inconsistency" that is the result of denying objective time and space, are exactly where interactions are encoded and classified, and moreoever the "evolving perspective" is the supposed key to unification.

This is clearly similar to the idea behind gauge symmetry, but with the major difference that there is no objective/fixed/universal gauge symmetry that we can use for reasoing without the normal constraints of measurement. It's all about perspectives and there is no master perspective, and thus no universal action.

In short the usually symmetry arguements, as a means of reasoing, and inferring interactions from a master symmetry is still there, BUT, this master symmetry is no master symmetry. It is instead an evolving structure, but there are not metalaw governing this, it's just a sea of opinons of how this evolves, which is exactly what drives interactions and evolution. To understand how that evolves, is I think also the key to unification.

From this I think it's easy to see why I for example like parts of string theory, but also why I really don't like other parts.

Neither string theory nor LQG implements, or even acknowledges these issues, but there are components all over the place. Smolin in his arguments on evolving law does seem to acknowledge at least a part of the problem, which is why I like him, although the solution isn't yet there. But relative to the major programs at least he is IMHO sniffing in the right direction from where we are.

/Fredrik

 

[tom.stoer]

I agree to most what you are saying.

... where a string ca be constructed in a high complexity limite as the simplest possible measure. But in that interpretation there is no external space the string is embedded in, instead the space is emergent as relations between different strings

Is there an article on this idea somewhere?

There is another issue regarding LQG that might show up in string theory as well (it could be a kind of pattern universal for quantum gravity): Whenever you try to develop such a theory your starting point is a classical theory with certain classical structures (e.g. symmetries). Having defined you quantum theory it seems that your classical structures have changed or even dissappeared. In LQG what dissappears are the manifold and the infinitesimal diffeomorphisms. The problem is that these structure (especially the latter one) are at the very core of the theory! The derivation relies on a structure that does not exist in the final theory! In string theory the same pattern might occur after having defined M-theory in a fully background idependent way.

The ultimate building says that its fundaments do not exist.

So my conclusion is that the naive way of quantizing a classical theory may no longer be valid, but one has to do it the other way round: start from a well-defined quantum theory w/o making any reference to a "classical theory". Unfortunately (as far as I know) such an approach does not exist. No quantization w/o starting from a classical theory is know.

 

[Fra]

Demystifier, I printed your paper yesterday but I haven't gotten around to actually read it yet, but I'll let you know when I have.

Is there an article on this idea somewhere?

If you wonder if *I* have written an article on what I mean with this then, No. But if everything goes as intended, at some point there will be. Also the technical context can not be justified in isolation from the larger picture.

I apologize in advance but here is a vauge rambling on this... just to give a hitn, that's all it's meant. Because I think we're talking about approaches and ways of reasoning here, and comparing programs:

I mentioned this as an illustration that I can see grains of hope in for example string theory, even though I reason completely different.

The very basic idea is that I aim to reconstruct the continuum. Going to back to square one, starting from distinguishable discrete events, characterizing an observer inside view, the continuum has IMHO not yet a physical justification. In particular continuum measures, like probability measures and thus probability spaces.

In this reconstruction, follows together a reconstruction of the continuum mathematics, probability theory (from combinatorics to continuum), and the idea is self-organising measure-complexes, that in my mind are the prototypes of "observers" - which ultimately of course is to explain matter. Ie. matter content of the universe is simple it's population of "observer complexes". The physical action of the matter systems, are inferred from the information processing properties of the observer complexs, which in turn are justified as a result of evolution. The rules for selection are contained here, and even the rules of selection are explained as the only constructive ones. Mass/Energy are related to the "complexity" of the observers. Ie the number of distinguishable internal states - basically memory size.

The simplest (lowest complexity) continuum-like measure might be a string like thing. And the idea is that the strings "mass" are simply distinguishable events, and an distinguihable ordering of the events can emerge as a resolution of fitness selection, this defines let's call it the string position. The size of the string would then somehow be observer dependent. For example one can imagine one string observing another string, and interesting interactions will take place. It might well be that a string observing another string, will end up seeing a membrane instead.

I'm not prepared to explain this now though. For simplicty let's just call it a wild idea I've had for a while, that has developed as My attempt to make any sense of string theory.

My reasoning is strictly an imagine "inside view", where the measures and operations are evolving as a result of the measure-complexes interacts with the environment. Starting from the simplest possible measure-complex, a boolean bit, and then scaling up the complexity, at some point the sea of distinguishable states will be forced to either disintegrate or form structures in the index space. A one dimensional distinguishability-index, can be more unstable than a two-dimensional index, when you understand how the existence of theose structures as "information processing structures" sits in the interaction with an unknown environment.

So in my view there is something that might with some stretch of imagination can come close to a string, but there is no such thing as a fundamental continuum string. Instead the string is composed of discrete degrees of freedom, that mathematically "looks like string" if you don't look close enough.

I picture these processes as a game. As a player in that game I can only try to form my own views of the effective rules, and confirm them by my own success. I think the logic is the same for elementary particles.

So I start with no space. I just start with a distinguishability notion, which if you prefer is 1-D. This structure is uncertain - this defines a direction of dissipation - time. But as this structure is subject to an environment that continously challanges it's very consistency, things will happen to this structure - it evolves.

That's just a hint, if that makes not sense just forget it, it is not meant to be conveyed yet.

But OTOH, it is equally possible to make an interpretation that looks like spin networks to this. Then the spin network could represent the microstructure of the measure-complex, and one could "imagine" two spin networks observing each other. And since a finite spin-network has a finite encoding capacity, evolution will select the fittest one, which sometimes may or may not yield string-like things. I think the action of a spin network as decomposed in the action coming from the inside acting on the environment as one thing, this contains random elements but depends only on the spin network itself. But the reaction from the environement, is the feedback that is principally undecidable and that is the cause of evolution and selection. A system can not maintain an inconsistency against a massive environment indefinitely. It's this destabilisation process I aim to understand, and how it's predicts the actions.

But all this is my, looking at both these programs foundations from a totally different view. I presume a try LQG theorist or a true string theorist would not acknowledge any of this.

Most the technical parts I've done so far are combinatorics, and expressions for actions and information measures, and these mathematical expressions are "solutions" to optimation problems. So a particle, would in my world, represent a measure-complex, who is justified as beeing a solution to a particular optimation problem (data compression effiency).

Another way of putting it, to me physical interactions are reduce to compressed data systems interacting and competing for control, of outcoding the other one, and conquer his memory.

In that space are simply a preferred structure that has formed in the memory structure. The bits of space are simply distinguishable events, no more no less. The fact that different structures encodes different sometimes inconsistent structures, implies that there will be an interaction force between them. (just like in gauge theory)

But there is no objective setting to describe this interaction (here I strongly deviate from the gague interaction logic), instead the incompletness and undecidability here imples a simple uncertainty. This is also in my view, the origin of quantum mechanics.

I do not even think in terms of "classical system" and "quantum". I think there is a common framework, where both of them are natural extremes.

/Fredrik

 

[Fra]

Anyway, since you like to see how other people think, you might find this interesting:
http://xxx.lanl.gov/abs/0905.0538 [to appear in Phys. Lett. B]

Hrvoje, I looked at your paper and what I like about several of your papers is that they are usually fairly clearly written and almost a big pedagogical. I don't think I said that before, but that struck med before with other papers of yours!

But from your paper it seems fairly clear that you have a very different attitude and way of reasoning than I do. When I read that papers the first things that point out to me, are not technical, they are your choice of scientific approach to this entire thing.

Just for fun, I'll comment briefly on the things your wirte that first hits me.

- Your blockuniverse view

This is in direct contrast to my inside view. In short think the root of how it is possible that we see this so differently, lies in our view of science. To me, scientific inquiry are closely related to measurements, and information is also depending on measurements.

I simply don't see the physical basis of the information in the universe implicit in the block universe idea. Who is possessing, and acquiring this information? I object to this abstraction since you seem IMHO to use abstractions that I fail to see how to ever realize in a physical subsystem.

This relates to the philosophy of science in the sense that, sure we consider ourself theorists, but still to me, every element in the model must be at least possibly realisable and testable in a real physical implementation. Otherwise for me it somehow falls outside of science.

Information always has a context, and placeholder. The information in a computer for example can not be maintained if the memory hardware is removed.

If you talk about the wavefunction of the entire universe, I simply doesn't understand what that is, in the context of science. This birds view or gods view, is exactly what I oppose to.

To me the only birds view that can be realized, when one observer, observes the only thing it can observe - it's environment, and tries to understnad other subsystems as part of his environment, and in that perspective the first observer has a constrained sort of external view of the other observers. But this view, is not an birds view existing in a matematical universe, it is a highly physical view, constrained by the physical universee as everything else.

Since we are looking to extend QM into QG, it is not obvious that the meaning of wavefunction remains the same, so here I'll take wavefunciton to be a metaphor for the "microstate of the observers microstrucutre", or simply the INSIDE information an observer has about it's own information, that is encoded in the observers SELF.

In this sense, to me it's obvious that information is something that the wavefunction AND the microstructure itself (the physical placeholder of the wavefunction - the memory hardware) is constantly challanged and evolving. Time is then a parameterisation of this flow.

The problem with unitary evolution is that - here I speak from my view - it applies only to the EXPECTED evolution. The actual evolution contains reactions from the environment, and wether they are as per expectations or not depends on how well equilibrated the observer is in it's environment (ie how well evolved it is).

The paradox is is when one things that the actual evolution must conincide with the expected one. The actual evolution is not necessarily a simple unitary one. However, there can still be a sound logic for the deformation of the placeholder for hte wavefunction so that the unitarity is sort of always true in the differential sense.

So to me, the picture you dry to draw, is painted in a scientific context (birds view) that does not have a physical justification IMO. That in short, is how I see your reasoning :) But it's nevertheless interesting to try to understand your reasoning, because there is no doubt that you have your own rational arguments to justfy it.

But let me guess that you have a view somewhat similar to Tegemark, which pretty much places the picture in a mathematical universe context? Ie. an eternal context that exists independent of the physical universe? Then your ideas appear rational also to me.

But I think the Tegemark style of replacing a physical scientific context for a mathematical universe is taken us away from good science, rather than getting closer? The reason why I think so, is because all history suggest that "scientific information" is changing all the time, rather than "accumulating". Sometimes we have revolutions that not only update the information relative to a framework, but sometimes not state change required by new data can maintain the integrity of the framework, and the framework itself breaks down.

That's the interesting situation when the trick is to survive a framework breakdown. One can not afford to start from scratch. Such situations are real, and occur in nature, not just biology but I think also in physics. For example, an observer can evolve as a reusult from interactions, and this is to me the one and same logic as the previously mentioned.

The mathematical universe or block universe idea instead picutres a god domain, where the truth is written in stone. But the question is, if the scientific method is the process by which an frog like us, are probing for this information, then strange questions appears. How can there be interaction between an inside observer, and this gods-domain?

Block universe, mathematical universe, universal symmetries, timeless eternal law, they all have this something in common which I object to.

/Fredrik

 

[Fra]

So my conclusion is that the naive way of quantizing a classical theory may no longer be valid, but one has to do it the other way round: start from a well-defined quantum theory w/o making any reference to a "classical theory". Unfortunately (as far as I know) such an approach does not exist. No quantization w/o starting from a classical theory is know.

I agree with your desire here. To start with a classical theory, and then add some - more or less - ambigous quantization procedure is not the satisfactory level of understanding I am content with.

I would label what I am trying to start with as a sort of intrinsic measurement theory, but where the theory itself, is not just a theory sitting outside the system describing it. This kind of gods view is the root of many problems. Instead, the theory itself is pretty much encoded in the physical systems. So the theory evolves as the observers does.

I think of this as a new framework, from which quantum actions will follow. It's known from normal game theory that quantum games are more efficient, and somewhat similar a system implementing quantum logic is more fit (though it depends on the environment) than a classical system.

The emergence of various logics, is something I'm working on, but I believe that it at least partially is in the direction you look for. Ie. there is no "classical system" to start with. There is just information encoded and manifested by real observer. the classiifcation of classical or quantum is depending onthe action beeing implemented in the system, and in my view there is no fundamental action, instead the action is a result of optimations done during evolution. It's simply the most stable and successful actions that we see in nature. If we can infer the actions we sene in nature, from such a model it would be quite spectactular. I think it can be done.

I believe there is yet a logic to this, yet to understand that will shed new light on the foundations of QM, as well as it's connection to the rest of the physics, in particular mass and inertia (and for sure GR).

/Fredrik

 

[Demystifier]

Block universe, mathematical universe, universal symmetries, timeless eternal law, they all have this something in common which I object to.

Thank you for the clarification of your point of view, which is best summarized with your sentence cited above.

 

[apeiron]

I believe there is yet a logic to this, yet to understand that will shed new light on the foundations of QM, as well as it's connection to the rest of the physics, in particular mass and inertia (and for sure GR).

/Fredrik

Do you not have any cites for your approach?

I think I mentioned CS Peirce and his bootstrapping view of reality. Pansemiosis some now call it.

It is also a logic to be found in systems science, hierarchy theory and dissipative structure theory to various degrees.

We could describe this approach generally as reality emerging from a "sum over observers".

And a key point you may or may not have considered is that this would be a scale-free equilbrium balance. So the "information" is not encoded in local observers (nor globally either) but instead is spread smoothly (that is, fractally) over all spatiotemporal scales.

Where is the observer that is dissipating the uncertainties, dissipating the histories, and so firming up the system? Everywhere and thus nowhere. Equilibrated across every scale and so not existing at any particular scale.

 

[Fra]

Do you not have any cites for your approach?

I think I mentioned CS Peirce and his bootstrapping view of reality. Pansemiosis some now call it.

It is also a logic to be found in systems science, hierarchy theory and dissipative structure theory to various degrees.

Apeiron, of course there are a number of papers worth reading, that have components of relevance to this!

But there since I'm not a professional, and I've had a 10 year break from physics I resume only some years ago, I have not yet published anything myself on this. Also I don't have any incentives to publish anything because I have to. If I think I solve a real problem I will publish. Many research papers concern isolated technical probllems within research programs, that interest everyone in that program and not anyone else. Since I am thinking of something that would be a new way of reasning, it would qualify as a new "program" and since this program doesn't yet exist, there is no context to publish isolated technical issues which anyone would care about.

It's harder, I have to produce a convincing overall picture, and solve at least one real open problem to convince anyone.

I think you are one of a few on here that seem to be somewhat on the same page.
Yes Peirce reasoning is good (although I am certainly no exper on Peirce! I think you know Peirce far better than I do; I am not sure i agree with Peirce all the way). Lee Smolins evolving law is good. Olaf Deryers internal relativity reasoning is interesting. Ariel Catichas idea to infere the laws of physics from the rules of rational inference is also good. The early part of Rovelli's RQM paper is also excellent (but not the finish).

But of course none of them have - to my current knowledge - produced any papers that implements more than what's fragments of what I'm looking for. But they are all worth mentioning as an indication of what I'm talking about.

I think you get it reasonably well Apeiron, and we reached a partial agreement about the general direction I think in previous threads. And some of hte unclairities left, and things that simply need more more work to sort out. The technical work I have done is under progress and I see no meaning in publishing fragments that only makes sense in a larger context. It will be published when I have a strong case, not sooner. Given how critical I am on others work I expect others to be critical on this as well.

/Fredrik