Do TOE candidates predict SM parameters?

Dmitry67

I agree with you, but the main point was about "free parameters". TOE with "baby universes" should not have any free parameters. Of course, it doesn't predict the values of SM parameters in *our* universe - it might be based on AP.

So may be it is a problem of terminology, parameters which appear to be "free" on the scope of individual baby universe, are not free on the scope of the full multiverse. My point was that on multiverse level TOE can't contain any free parameters. Do you agree?

nicoo

To be a potential ToE, it should be proved that ST contains the vacuum we are living in (even among 100^500 others), which AFAIK is not (yet) the case...

tom.stoer

No. It may contain some free parameters which you can fix via a small number of experiments.

Again: it seems that we do not agree an "ToE". A theory describing all known phenomena consistentyl is a ToE. There can be more than one ToE and there can be free parameters in a ToE (to be fixed in the above mentioned sense).

There need not be a unique ToE; different ToEs may exist, and experiment will select the 'correct' one. It's called 'Theory of Everything' = ToE, not 'Unique Theory of Everything' = UToE.

tom.stoer

In order to be a potential candidate ToE it need not prove that ;-)

But I think what we are discussing here is irrelevant; it's only the meaning of "ToE". It is much more interesting if a certain theory is ab le to describe nature better than other theories. If ST is better than the SM it's fine and nobody cares whether it is a ToE, a candidate ToE, a potential ToE, a candidate for potential ToEs, ...

friend

You are entitled to your opinion, of course. But a Theory of Everything should at least explain everything physical. If parameters can only be "explained" by measurement, then there is no theory for those parameters and thus no TOE. If you have a theory that explains those parameters as being proababilistic, then at least there is a theory for them, but no way to confirm that probabilistic nature by comparing instances where the parameters are different. A complete TOE would not only predict the value of those parameters but would also explain where QM an GR come from and how they are connected. And as far as explanations go, I think that a complete TOE would have to be derived from logic itself, else you are left wondering about why some part of nature is the way it is.

Dmitry67

friend, this is exactly what I wanted to say. Only such TOE "derived from logic itself" can be compatible with MUH. I had so got used to MUH that it became my "hidden assumption", when I was talking about TOE, I was actually talking about MUH-compatible TOE.

If tom.stoer is right, and there is a non-MUH TOE, where one or more "free" parameters just have these values, just remember these values - we can't derive them - well... then it would be soooooo sad.... Because it that case Nature is not beautiful

God doesn't like free parameters (c) me :)

haushofer

That's one thing we should have learned from string theory 25 years ago: to rethink our expectations of what a theory of physics should be capable of. Especially crackpots are fond of theories which can explain "why the universe is as it is", probably because of bad popular science. I think for that you quickly run into a discussion of what the precise relation is between mathematics and the world surrounding us. Somehow I like to compare the idea that our mathematical notions are capable of explaining "why the universe is as it is" with the idea that our earth is the centre of the universe. Why should mathematics do that? Because it is so highly succesful in the natural sciences?

I think the most down-to-earth way of shedding light on this question is actually trying to construct extensions of the standard model, or writing down theories of quantum gravity, and see how far one can get ;)

tom.stoer

Dmitry67 is right when considering MUH-ToEs and non-MUH-ToEs. And haushofer is right when preferring down-to-earth-theories.

Suppose we had a class of theories like SUGRA with some special properties:
1) countably many (or even a finite number)
2) perturbatively renormalizable (or even finite)
3 with some very specific and testable predictions (perhaps even in the low-energy = non-Planck-regime)
4) with a few free paramaters which can in principle be determined by a finite number of experiments
Suppose that there is no underlying uniqueness principle, no MUH-like reasoning; ...
nevertheless b/c of (1-4) I would be rather happy and would call this a great success

haushofer

Me too! :D I'm not that much into phenomenology, but can people make sense of N>1 SUSY or SUGRA phenomenologically due to the non-chiral nature of these theories? As far as I know it's only for N=8 SUGRA that we still don't know if it's renormalizable or not, but things don't seem to look that good.

On the other hand, if it would be renormalizable but phenomenologically "incorrect", would it be such a huge important succes (apart from all the technicalities one has to invoke to come to such a result)? It's nice to write down a theory of QG which is renormalizable of course, but then?

tom.stoer

The problem-setting is multiform
1) we have to have some sort of internal consisteny by construction
2) we have to to have some predictive power in principal
3) we have to have some phenomenological success

In the case of the standard model criterion 1) is not really satisfied; we know that the theory is renormalizable, but we know that the perturbation deries is not convergent, we know that we miss non-perturbative effects, but we do not have a rigorous proof (disproof) that the SM does exist (does not exist) mathematically.

But we are satisfied with 2-3) -- and b/c we do not expect the SM to meet criteria for a ToE -- whatever this means in detail -- 1) is of minor importance.

Shifting or interest from phenomenological successful theories to ToE-like models we focus more on 1) than on 3) Why? b/c we know that there is no phenomenological reason to be interested in anything else but SM + GR! So the reasons to be interested in ToEs is due to elegance of unification, internal consistency, uniqueness, etc.

That does not mean that 2-3) become irrelevant, but they are secondary criteria. We are in the situation that looking for new theories is neither necessary by 3) nor can the new theories we construct be tested against 3) So there must be other guiding principles, e.g. in the spirit of 1)

It's a philosophical questions whether this means MUH-ToEs, string-like ToEs, xyz-ToEs or whatever.

Back to you question: having a consistent but phenomenological wrong theory would mean that it's ruled out as a theory of nature. But after decades of dominationof theories which are neither phenomenological successful (neither right nor wrong -- not even wrong ;-) nor provable consistent (nor inconsistent) this would be step forward: it would be the first tangible phenomenological result in quantum gravity!

friend

Yes, that's some pretty scary territory. On the one hand, we'd like to develop a ToE from the requirement of logical consistency alone. Then no one in any way could possibly argue with it, and we'd be assured that everything is absolutely reasonable as we suspected all along. On the other hand, what if it's wrong, and we make measurements that are not compatible with our theory of reason. That would be a dilemma that probably persuades some to not even look for such a theory. I think we need to have more faith than that.

Dmitry67

Another scary option - what is there are more then 1 TOE compatible with our experiments? What if the difference would be only at Planck energies, so we won't have a chance to tell what theory describes our Universe?

As an example, there are 2 theories of gravity compatible with our experiments: GR and Einstein–Cartan theory, and there is no chance to test the difference experimentally. We hope that future TOE would rule out one of them, but what if TOE itself is not unique in the same way?

Almost ANY mathematical theory is extensible... Of course in many (but not all) cases these extensions come with the price of some sacrifices and artifacts (like "dividers of zero" which come with quaternions), and there is no guarantee that TOE is not extendable in that way...

tom.stoer

I agree with everything except for

Quaternions are a division algebra, so there always exist a unique inverse element and there are no two quaternions p,q ≠ 0 such that pq = 0 or qp = 0; but quaternions are non-commuative, i.e. pq ≠ qp in general

Dmitry67

My bad, octonions...

tom.stoer

no, not even octonions ;-)

octonions are a divison algebra w/o zero divisors; but they are not associative

Dmitry67

Another try. Sedenions :)

tom.stoer

yes, they have zero divisors and are not a division algebra!!