Science a GoGo's Home Page Forums General Discussion General Science Discussion Forum A Single Universe or a Multiverse?

Ibliss: ?Well, that's even more reason to abandon the idea of ''real'' physical universe.?

You have lost me here. OK, I understand that you are a proponent of a mathematical reality replacing the good ol? physical reality, but the arguments I offered do not justify in the least your statement above. Let me try another avenue, clearer I hope. Do you think that any and every observation regarding the surrounding nature can be translated into a mathematical model? At least for the time being the answer is no. So for the time being, any comprehensive mathematical model will necessarily be an incomplete description of the nature.
Now, is this a trend that is likely to be universally valid (that some observations will never be expressible in mathematical language)? I have no ideea, and no one has any ideea. Are there more or less educated guesses? Not at this time. We only have speculations. What could actually help us answer these questions? The physical reality for sure. The mathematical reality never, unless you compare it to the observations, i.e. with the physical reality. Sure, there is the possibility a la Umberto Eco?s ?Foucault?s Pendulum? to actually devise all the possible models (there is still the question how you could even do this practically with only partial knowledge available), and then compare them with observations to see which one(s) match. But you still cannot rid yourself of the physical reality, even in this case. You need the standard for comparison!


Pasti: ?And the evidence for your assumption is...? Of course, you can make this assumption, but then how is your scheme better than the reductionist scheme above that you don't agree with? You have given up a concept only to postulate another.?


Ibliss: ?No evidence for it, but also no evidence against it, so why not try to do without? Now you do have to adopt certain postulates to make progress that aren't explicitely present in single universe theories (what is the measure on the set of all possible universes, e.g.). However, as I tried to explain , all physicists make implicit assumptions about this when interpreting experimental results (Occam's Razor).?

OK, let?s suppose indeed that it is worth to try your approach, and let?s forget about the evidence and arguments against it. And now let?s suppose that you develop your multiverse theory, and you find (and you will necessarily find so) that there are several ?universes? infinitesimally close to the observations, within your measure. I.e. you will find several universes, with different degrees of theoretical complexity that fit your observable universe. What will you do then? Which will you choose, with only partial knowledge of the ?whole?? So you still need Occam?s Razor, and not only that, you will also need all the assumptions one makes in a multiverse, for the obvious reasons. So if indeed you are true to Occam?s Razor principle, you have just developed a theory requiring a much larger number of assumptions than the one that is already ?on the market?, so by Occam?s razor, you should drop it and take the minimal one. Of course, unless you find an effect describable only in your multiverse theory, which unfortunately, by design, is not testable. So what have you gained? Insight? Not any more insight than you could get from the theory on the market. Intelectual satisfaction? Definitely. But unfortunately, the latter is not essential in understanding nature?


Pasti: ''I am not postulating anything. I am observing, interpreting, observing again, etc. There is no postulation (in the mathematical sense) in physics beyond the mathematical formulation. The "postulates" in physics are merely the conclusions of observations, and they are believed to hold true only until they are contradicted. Think about it.''

Ibliss: ?It's precisely the ''mathematical formulation'' that you mentioned.?

And how about those observations that cannot be put in mathematical language, even if this is only for the time being, so to speak?

Ibliss: ?We have found so far is that nature is so extremely regular that a small set of postulates suffices to describe the outcome of experiments below, say, 100 GeV. So, you have a handfull of (as far as we know) arbitrary constant, that defines the standard model and general relativity. Even though we already know that these theories are not consistent, we nevertheless claim that experimental outcomes are correctly described by these theories provided we don't go to too high energies.?

OK, so you have a handfull of constants. According to the viewpoint, this may or may not have any significance whatsoever (you know, the fundamental constants are just unit matching fudge factors and so on and so forth ? I don?t subscribe to the ideea and I am not sure if you do). But nevertheless, significance of the constants aside, you still have more than only two models, that cannot be unified. What makes you think that there will be such a GUT, even in this context? Sure, people hope, but even if their hope will become reality, the unification of physics with biology has not even begun, and so on and so forth. So what makes you think that such a comprehensive model could exist?


Ibliss: ?No, because the Ar interacts with the walls of the chamber and thus the initial state of the Ar cannot be recovered from the final state of the Ar alone.?

Well, there you go. No matter the reason, the situation I presented you is a perfectly valid physical situation, and in the process you have lost information.


Pasti: ''I am afraid you misunderstood me. Think of my statement as of the Maldacena conjecture:works in simple/particular cases but not in the general case.
The comparison you made with conservation of momentum is not exactly to the point, no offence. Conservation of momentum, of course within experimental errors, has been confirmed not only for simple systems, but for all systems where such conservation is relevant. This is the reason why it is believed to be a "fundamental law" until the contrary is observed. But this is not the case for the conservation of information law that you mentioned, which has been only theorethically developed for very simple geometries, and has not been tested experimentally beyond such simple cases, it tested at all.''

Ibliss: ?But what about unitary time evolution?? Have any experiments detected violations of this rule for isolated systems??

Not to my knowledge, but then why is this relevant to anything? The observable universe is NOT isolated (according to modern observations, the cosmological horizon has roughly 3000 MPc/H, and as for the entire universe, we know even less than about the observable one?.
So where does unitary evolution come into play?


Pasti: ?And the proof of your statement in regard the the "...can only be further explained..." is where? ?

Ibliss: ?I meant this. When you abandon the notion of a (God given) single physical universe, you are left with an ensemble of all possible universes, each one ''as physical'' as another one.?

Nothing ?God given? about such an assumption. It is based on observation/observational capabilities. As soon as we can observe something that contradicts this assumption, it will change accordingly. If you give up this assumption, as you suggest, you are left with whatever you can imagine, including your multiverse theory.

Ibliss: ?So, you will now have to explain why we live in this particular universe instead of another. There is no reason to assume that the answer to such questions can be found, but at least such questions do make sense in this setting. In single universe theories it is not clear that the question has any meaning at all.?

I agree that you can do all you say in your setting. But how is this relevant to anything? Let?s back up a bit, and review the idea. You give up a concept backed up by observation in favor of a concept with no observational support whatsoever, you create a conundrum stemming directly from the introduction of the unsupported concepts, and than you claim that the answer is somehow relevant to the phenomenology in a single universe? Sounds quite fishy to me.
The question of why do we exist in this universe makes sense in single universe theories, there is no doubt about this issue. But I once again doubt that we will be able to find any cogent answer to it at this moment in time, with the little we know. We don?t know yet what life is and how it has appeared, and by going to multiverse theories you won?t solve these problems either. They must first be solved in a single universe, and only after they are solved we can attempt to ask questions of the nature that you mention. Otherwise you will solve only artificially created conundrums, which by design cannot elucidate question to which you don?t know the answers (you should remember that the way you construct a model in physics is by knowing the answers beforehand, from observations, and not the other way around; when you extend the model infinitesimally beyond the known answers, you are looking in fact for ?experimental? questions and hoping that the model already includes in its built the right answer that will come from observation).

Ibliss: ?The very act of interpreting means that you are searching for a model. Now, in may cases that model manifests itself in a very compelling way, so you may say that no ''searching'' is involved. However, that brings us back to the question why nature is so regular.?

No Ibliss, a theory is not necessarily fully translatable in a mathematical model. Ideally, people hope that it would be, but then historically sometimes it was necessarily to invent the language necessary to translate the theory into (Newton and Leibnitz), sometimes it was not necessary to translate anything into mathematical language to develop a theory (Faraday), and sometimes, there are observations that we don?t even know if they are translatable into mathematical language (life).

Ibliss: ?Explaining something means that your model contains less information than is contained in the original experimental data. Otherwise it would be trivial curve fitting. This approach has been successful and there is no way conventional physics can explain why.?

Of course it can. It is called symbolism/symbolistic representation, and it?s been around since almost prehistory. All symbols contain much more information than the representation itself, from cave paintings to hieroglyphs to pictography to whatever else you like. Mathematically, if it pleases you, you have the theory of hierarchies and categorization, applied to symbolistic language/representations.
Ibliss: ?Another example. Fermi's theory of the Weak interaction is now known to be the low energy limit of electroweak interaction. However, there are an infinite number of (ugly) non-renormalizable theories that have the correct low energy limit. According to your arguments they could all have been correct and there would be no preference for the theory we now know to be correct.?

Ibliss, you are giving the wrong counterexample. At the time the electroweak theory was developed, renormalizability was already considered to be a sine qua non requirement for any quantum theory, due to the earlier success of QED. Renormalizability is a requirement (albeit purely mathematical in character) that has been observed to hold valid in constructing models consistent with experimental observations (you will have to agree with me that recovering from a mathematical model a value matching the observation up to the 14th decimal is a heck of a consistency, which is the case for QED). Is it empirical? Sure.Has it been observed to be working? Sure again. Could it fail, say for gravity? Of course.
If the renormalizability requirement would have been absent, then the situation would have been drastically different, and indeed one should have had to use Occcam?s razor criterion, unless something else could have qualified as a selection criterion.


Ibliss: ?I agree that they could in principle all have been correct, but that the electroweak theory was more likely to be correct, because it contains fewer arbitrary parameters.?

Well, not necessarily because of that. In your argument, the criterion was renormalizability, and not necessarily simplicity.

And yes, in developing a theory one expressly includes the ?additional? Occam?s Razor ?principle?. But then, you will ask why should such a principle be included in such a task. The answer is simple enough, in principle at least. Once again it has been observed that roughly speaking, more complicated theories ?contain? more complicated phenomenology, which should be observed if the theories are correct. If some of the predictions of such theories yield unobserved phenomenology, or incorrect phenomenology, then these theories are discarded. If you apply this ?selection? method to all the theories that are candidates for a given set of observations, maybe not surprisingly, you will end up with the simplest theory fitting the data. This is the essence and the reason for Occam?s Razor ?principle?. The practice however, turns out to be much more difficult, if you know what I mean.

Ibliss: ?Within the framework of your ''philosophy'' the verification of the electroweak theory raises more questions: Why this particular one with 2 new parameters, and not one of the trillions of others which contain zillions of arbitrary parameters? ?

Oh, come on Ibliss. You know exactly what I was trying to say in my ?philosophy. Using any other selection criterion or criteria leads you to the same questions. Even after choosing somehow the best model, the question of why does the best model for the Omega theory contain exactly x parameters and not y remains a more or less xalid question, and you know it.