Is the multiverse fake physics?

[kodama]

woit over at his blog not even wrong considers string theory based multiverse to be fake physics. he cites sean carroll as others on his blog as examples of fake physics.

is string theory based multiverse fake physics?

 

[Orodruin]

It is customary to include a link when discussing blog posts or similar.
http://www.math.columbia.edu/~woit/wordpress/
Not everyone reading will know the blog.

I think he has a clear point. Large parts of theoretical physics has become untestable and therefore unscientific. Trying to sell this as facts or success is misleading.

 

[ohwilleke]

I agree that the multiverse (not to be confused with the many worlds interpretation of QM) is fake physics.

 

[Orodruin]

I agree that the multiverse (not to be confused with the many worlds interpretation of QM) is fake physics.

To be honest, I would still place QM interpretations into the same category. It is also a subject that is given far too much attention in popular discussion on QM.

 

[haushofer]

I agree that the multiverse (not to be confused with the many worlds interpretation of QM) is fake physics.

Can you elaborate?

 

[kodama]

Can you elaborate?

ohwilleke is replying to my thread which is itself based on the claims of peter woit on not even wrong that
the multiverse is fake physics in analogy to fake news. even when claims of multiverse comes from
a cosmologist and phd like sean carroll or brian greene it is still fake physics.

woit says claims of the multiverse is based on string theory unification, which he regards to be a failed
research program, therefore physicists like sean carroll and several others he mentions are not discussing
genuine physics to a lay public.

what do the resident string theorists here say to this claim, i.e urs shcrieber et al

 

[haushofer]

ohwilleke is replying to my thread which is itself based on the claims of peter woit on not even wrong that
the multiverse is fake physics in analogy to fake news. even when claims of multiverse comes from
a cosmologist and phd like sean carroll or brian greene it is still fake physics.

woit says claims of the multiverse is based on string theory unification, which he regards to be a failed
research program, therefore physicists like sean carroll and several others he mentions are not discussing
genuine physics to a lay public.

what do the resident string theorists here say to this claim, i.e urs shcrieber et al

Well, I've read a fair amount of string theory and my own research involved it, so let me give my 2 cents.

The motivation for the multiverse from string theory is shaky. It involves an interpretation of the landscape, which is on its own quite ad hoc. I think the motivation from inflation is stronger: to get the right amount of inflation results quite naturally in eternal inflation. Let's say that at some time in the future we are very confident that inflation happened, and that our model implies eternal inflation. Philosophically, we entered then a boundary of our understanding: we have no way to falsify the implied multiverse, but as a consequence of our model we should take it seriously. There is nothing fake about that. It would be fake to reject it because we don't like the idea that we bumped into an end of our understanding. Physics thaught us that we should take our equations seriously. If we are just one bubble in a multiverse and we cannot enter those other bubbles, then that's the way reality is; nature doesn't care about being falsifiable.

But often people combine this 'inflationary multiverse' with the 'string landscape multiverse'. I don't see how that happens, and it is far from clear what the relation (if any) is. Let alone the multiverse from the MWI-interpretation. It is only in those cases you enter these "theory of anything"-objections agains the multiverse.

To be honest, if I had to choose, I would say that the MWI-interpretation of the multiverse is much more shaky than the inflationary multiverse. I've never understood why people like MWI, but that could also be my lack of deep knowledge of it.

 

[haushofer]

To be honest, I would still place QM interpretations into the same category. It is also a subject that is given far too much attention in popular discussion on QM.

It depends on how you view science and physics. I strongly disagree with your view. Ontology, and as such interpretational issues, should imo be considered as a part of physics, and not 'just' philosophy.

 

[Orodruin]

It depends on how you view science and physics. I strongly disagree with your view. Ontology, and as such interpretational issues, should imo be considered as a part of physics, and not 'just' philosophy.

As you have guessed, I strongly disagree with this. If there are no testable differences, I see no scientific point in debating the issue. Just pick whichever interpretation you fancy (if you must) and nobody can disagree with you. To me this violates the very core of empirical science.

 

[ShayanJ]

As you have guessed, I strongly disagree with this. If there are no testable differences, I see no scientific point in debating the issue. Just pick whichever interpretation you fancy (if you must) and nobody can disagree with you. To me this violates the very core of empirical science.

What if in 10 years, people still have no explanation for fine tuning? In 50 years? In 100? Even then you don't think we should consider the multiverse?

 

[Orodruin]

What if in 10 years, people still have no explanation for fine tuning? In 50 years? In 100? Even then you don't think we should consider the multiverse?

No. It is a concept similar to God (I am using God to cover three of the main world religions - anyone may substitute for personal preference). I give you the two hypotheses:

1. Fine tuning exists because of the multiverse.
2. Fine tuning exists because God wanted it so.

I ask you the question: What testable differences exist between those two hypotheses?
If neither hypothesis can be falsified, I simply chose the agnostic path and do not include any of the hypotheses in my scientific description of how the world works. (I also have the same approach to religion by the way, but that is a different matter.)

 

[ShayanJ]

No. It is a concept similar to God (I am using God to cover three of the main world religions - anyone may substitute for personal preference). I give you the two hypotheses:

1. Fine tuning exists because of the multiverse.
2. Fine tuning exists because God wanted it so.

I ask you the question: What testable differences exist between those two hypotheses?
If neither hypothesis can be falsified, I simply chose the agnostic path and do not include any of the hypotheses in my scientific description of how the world works. (I also have the same approach to religion by the way, but that is a different matter.)

The difference is that eternal inflation may be a good theory with strong observational support. We may find out that we can't come up with a better theory. Why should we throw out a prediction of our only good theory?
Its like gravitational waves, people didn't doubt their existence. No one actually was surprised when they were discovered because GR is a very well established theory.

 

[Orodruin]

The difference is that eternal inflation may be a good theory with strong observational support.

Please show me this observational support.

We may find out that we can't come up with a better theory. Why should we throw out a prediction of our only good theory?

What do you mean "better"? If two theories makes the exact same predictions apart from things that cannot be tested - they are the same theory for all practical purposes and you can never find out which is preferred. Predictions that cannot be tested are not predictions because predictions by definition makes statements about things that will occur or not.

Its like gravitational waves, people didn't doubt their existence. No one actually was surprised when they were discovered because GR is a very well established theory.

This is different. Gravitational waves were a priori testable and based on the current dominating theory - which had already made a large amount of verified predictions that differed from its predecessor (Newtonian gravity) - they should be there. So I ask you, what predictions of the multiverse have been observed that cannot be explained without it?

 

[ShayanJ]

Please show me this observational support.

I said maybe. I meant in the future, it may turn out to be so!

What do you mean "better"? If two theories makes the exact same predictions apart from things that cannot be tested - they are the same theory for all practical purposes and you can never find out which is preferred. Predictions that cannot be tested are not predictions because predictions by definition makes statements about things that will occur or not.

In the future, hypothetically, just imagine, that eternal inflation explains all observations really well and any other theory that doesn't predict a multiverse, is much more complicated. Imagine its like the choice between SR and aether theories. So people choose eternal inflation. They have observed all its predictions except for the multiverse. Why should they assume its not true?

 

[Orodruin]

In the future, hypothetically, just imagine, that eternal inflation explains all observations really well and any other theory that doesn't predict a multiverse, is much more complicated. Imagine its like the choice between SR and aether theories.

But this is the other way around! You can do away with the aether in LET and just be left with SR because there is no need to assume the aether of LET. In the same way, there is no need to assume an eternal inflation to explain what is going on in the observable universe.

Of course you can try to extrapolate a theory, but doing so you must be aware of that this is what you are doing - just as we know that we are extrapolating GR to a domain that by definition of the theory itself is untestable when we try to describe the interior of a black hole. If the theory is true in that regime - then you can never know that it is.

 

[kodama]

But this is the other way around! You can do away with the aether in LET and just be left with SR because there is no need to assume the aether of LET. In the same way, there is no need to assume an eternal inflation to explain what is going on in the observable universe.

Of course you can try to extrapolate a theory, but doing so you must be aware of that this is what you are doing - just as we know that we are extrapolating GR to a domain that by definition of the theory itself is untestable when we try to describe the interior of a black hole. If the theory is true in that regime - then you can never know that it is.

where does string theory and its prediction belong on this list?

 

[FactChecker]

As you have guessed, I strongly disagree with this. If there are no testable differences, I see no scientific point in debating the issue. Just pick whichever interpretation you fancy (if you must) and nobody can disagree with you. To me this violates the very core of empirical science.

Are you saying that something must be testable in order to be true? Or is your objection more about calling it physics?
What about a modern theory that is compatible with all we know today? Doesn't that imply that it passes the "test" of all our current knowledge?

 

[haushofer]

As you have guessed, I strongly disagree with this. If there are no testable differences, I see no scientific point in debating the issue. Just pick whichever interpretation you fancy (if you must) and nobody can disagree with you. To me this violates the very core of empirical science.

A physical theory must do more than just 'agree with observation' in view of future development. An example is given by the epicykle theory, which reproduces observations concerning our solar system perfectly in a geocentric model. Mathematically we now understand why: given any irregular shaped closed curve, we can always use a finite number of epicykels to explain such a curve.

That's why I don't understand how some people can just put ontology into the 'philosophy'-corner and pretend it is not part of physics. It is. E.g., one reason why quantum gravity is not well understood could be in a similar fashion as the epicykels: we can calculate with it, but by regarding QM just as a bookkeeping device, we could overlook crucial hints which are important when considering gravity.

Similary, nowaydays the discussion about the nature of spacetime and the meaning of coordinates can be regarded as 'mere philosophy', but it troubled Einstein so much that for two years he abandoned general covariance. And I think that general covariance is still not thoroughly understood, by the community and also by a lot of professional physicists.

 

[Orodruin]

A physical theory must do more than just 'agree with observation' in view of future development.

Yes, it must do so without introducing additional arbitrary epicycles that can fit any observation. Furthermore, epicycles would have lost out to Occam's razor long ago.

we can calculate with it, but by regarding QM just as a bookkeeping device, we could overlook crucial hints which are important when considering gravity.

Introducing different equivalent interpretations is not going to help you achieve this. Thinking about different and a priori distinguishable theories just might.

 

[rkastner]

To be honest, if I had to choose, I would say that the MWI-interpretation of the multiverse is much more shaky than the inflationary multiverse. I've never understood why people like MWI, but that could also be my lack of deep knowledge of it.

There are good reasons to be concerned about the unitary-only MWI approach. If one looks deeply into the foundations of MWI, one finds circularity and arguably worse logical fallacies. See, e.g, https://arxiv.org/abs/1406.4126 and https://arxiv.org/abs/1603.04845

 

[AlexCaledin]

...the two hypotheses:
Fine tuning exists because of the multiverse.
Fine tuning exists because God wanted it so.

― but they both might very well be true together ― just imagine the Great Supernatural Computer simulating all the Multiverse, including all those quantum Many Worlds in the whole space-time, all at once, as the Potentiality. Then, what remains to happen, is the actual choices of the observed events, made by the supernatural players of the great computer game (the first move is the choice of the good fine tuned Nature). It's called the "process formulation":

"The process formulation of quantum theory contains no explicit dependence on human observers: it allows quantum theory to be regarded as a theory describing the actual unfolding or development of the universe itself, rather than merely a tool by which scientists can, under special conditions, form expectations about their observations. The quantum theory of process is in general in accord with the ideas of the physicist, logician, and process philosopher Alfred North Whitehead. In particular, the actual is represented not by an advancing, infinitely thin slice through the space-time continuum, but rather by a sequence of actual becomings, each of which refers to a bounded spacetime region: event number n is represented, within physical theory, by a restriction on the set of classical fields allowed in the bounded space time region R(n). We have, therefore, neither becoming in three-dimensional space nor being in the four-dimensional world, but rather becoming in the four-dimensional world. "

Henry Stapp, "Physics and the Ultimate Significance of Time", page 267.

 

[mfb]

A finite number of epicycles cannot match the predictions of GR exactly - they cannot take the mutual attraction between planets into account, for example. You can always find a set large enough to be consistent with current observations, of course, but hundreds of epicycles clearly lose compared to GR.

Where is the relation to the idea of multiple universes? I'm not aware of any hypothesis that would introduce hundreds of arbitrary parameters tuned to some observations. The fundamental laws of the set of universes in those multiverse approaches are usually extremely simple. They do not necessarily make predictions about the fundamental constants of our universe - but we have many theories that do not make those predictions. The Standard Model is a well established theory, although it has free parameters it cannot predict, and its predictions are limited to particles we already know. How would a theory that implies multiple universes with different values for the free parameters of the SM be worse than that? It can predict the same as the SM, and it can be falsified in the same way (here: by finding violations of the SM). If it makes predictions beyond that, e.g. some structures in the CMB, neutrino or GW background: even better.

 

[Ernest S Walton]

Large parts of theoretical physics has become untestable and therefore unscientific. Trying to sell this as facts or success is misleading.

In other areas, when the accepted methodology generates scenarios which appear incompatible with that methodology, it is the methodology which comes under scrutiny. Similarly, if the search for truth leads to scientifically untestable theories, this may indicate a problem with Science, rather than the search for truth.

 

[eltodesukane]

Earth used to be The planet, The Universe.
Is the multi planet a fake? The multi solar system? The multi galaxy?
If, somehow, some entity exist, why not exist twice? Many? Infinity? Why not multi universe?
I think multiverse make sense in many many ways.

 

[Orodruin]

Why not multi universe?
I think multiverse make sense in many many ways.

Because it is inherently untestable. It might exist but you have no way of knowing that it does. Other planets, stars, and galaxies were never untestable - it was just a matter of constructing a large enough telescope.

 

[Cal ben]

I think the question about "fake science"has more to do with this: at what point should this stuff be publicly made known to be a best guess rather than truth.
Science is conducted for the benefit of mankind, right? But it sure isn't presented to mankind, in general, in a way that can be understood. Most theories seem to be presented to the general public as truths. If you want to know the real truth, you really have to dig and find scientific journals that are written for other scientists.
For instance, I'm relatively new to studying Astro and quantum physics. So a while ago I picked up Hawking book. Interesting read. He's an inspiring writer. Seems like he spent most of his ink going on about how M theory proved the existence of the multiverse and unified all quantum theories and the fine cogs of all reality is an open book to us now. No wonder this was a best seller. I was abuzz with visions of the vastness of the all of everything being intimately connected with little ol me. Only I'm not the type to stop there. This only sparks my interest and I start plowing into articles from scientific journals and really digging into what data has actually been accumulated and what all the other people smarter than me think. Come to find out that M theory is described by a vast number of mathematicians and physicists to be "a beautiful peace of mathematics" but doesn't actually mean anything, or worse, "a mathematical magic trick" and a bunch of other things like this. (I'm sorry I can't provide a reference, I read these a number of months ago. The language used made the statements stick in my head.) So the logical conclusion? Neat idea, maybe, maybe not. Can't say one way or another. Dig a little deeper and find out that even the most accepted quantum theories (string, SM) are just best guesses and my blossoming mind goes boom, fractures into a steaming pile of broken fantasies, and eventually crumbles back down to earth, back to the hard facts and studies that drew me to science in the first place. But the problem is that most people who picked up that book never looked deeper because, really, why would they. Us uneducated lay people are somehow convinced that Hawking is like some kind of physics Yoda and who are we to question Yoda? But, in reality, there are a ton of other Yodas out there just as smart who know it's just one idea out of many of how reality is pieced together. Not the absolute truth that it was presented as.
So is any theory fake science? Probably not. It's actually an important part of the process. Without theories, there can be no experimentation, no continuation of the search. What makes a theory, strongly or weakly supported, fake science is when it's presented as a truth when no theory has any business being called a truth.

 

[kodama]

Because it is inherently untestable. It might exist but you have no way of knowing that it does. Other planets, stars, and galaxies were never untestable - it was just a matter of constructing a large enough telescope.

true, but the string theory multiverse is based on string theory, which if you can construct a Planck scale collider is testable.

 

[member 563992]

 

[mitchell porter]

woit over at his blog not even wrong considers string theory based multiverse to be fake physics.

Peter Woit considers everything based on string theory to be fake physics.

Judging by a combination of empirical success and empirical predictiveness, I'd say holographic QCD is the best thing to come out of string theory, then conventional string phenomenology, with multiverse reasoning in third place. Holographic QCD seems to be on a par with many other approaches to approximating QCD. Conventional string phenomenology has great potential, but too many things still can't be calculated. And multiverse reasoning is like evolutionary psychology, it really can make predictions (if the Higgs mass had been 141 GeV, Hall and Nomura 2009 would be famous), but only at the price of numerous "what if" hypotheses about the structure of the string landscape and the implications of the anthropic principle.

 

[Haelfix]

Because it is inherently untestable. It might exist but you have no way of knowing that it does. Other planets, stars, and galaxies were never untestable - it was just a matter of constructing a large enough telescope.

Right but I mean if that's the criteria you use, then stars and galaxies beyond the cosmic horizon are also unscientific as well.

At the end of the day there is more to science than simply testability. There are also aesthetic and logical criteria as well. For that reason when people test inflationary models, simpler models are typically analyzed first. For instance you can always cook up some complicated inflationary potentials (for some choice of boundary conditions) that have predictions that can be made arbitrarily close to M^2 Phi ^2, but obviously if people discovered evidence that favored those classes of models the simplest one would be preffered.

Of course the multiverse would come along for the ride depending on the details, but then it's typically not very important unless you make some sort of additional anthropic argument (which is logically different) to explain some other phenomena.

Anyway, there are many very good physicists who take this material (in its multiple incarnations) seriously, so its surely not junk science.

 

[kodama]

Peter Woit considers everything based on string theory to be fake physics.

Judging by a combination of empirical success and empirical predictiveness, I'd say holographic QCD is the best thing to come out of string theory, then conventional string phenomenology, with multiverse reasoning in third place. Holographic QCD seems to be on a par with many other approaches to approximating QCD. Conventional string phenomenology has great potential, but too many things still can't be calculated. And multiverse reasoning is like evolutionary psychology, it really can make predictions (if the Higgs mass had been 141 GeV, Hall and Nomura 2009 would be famous), but only at the price of numerous "what if" hypotheses about the structure of the string landscape and the implications of the anthropic principle.

since the paper
A Finely-Predicted Higgs Boson Mass from A Finely-Tuned Weak Scale
Lawrence J. Hall, Yasunori Nomura

is falsified by LHC finding 126 gev, does this mean hypothesis of string landscape and multiverse is also falsified?

so string theory multiverse is fake physics in the sense it makes a prediction that has been falsified by the LHC

 

[Orodruin]

At the end of the day there is more to science than simply testability. There are also aesthetic and logical criteria as well. For that reason when people test inflationary models, simpler models are typically analyzed first

Simpler models are tested first because they are simpler to test. Beauty is not a viable scientific criterion. We may hope that Nature is kind enough to provide us with something beautiful - but you are never going to pick a beautiful theory that makes false predictions over an uglier one that makes correct ones. If in doubt - U(1)xSU(2)xSU(3) is not particularly beautiful.

but obviously if people discovered evidence that favored those classes of models the simplest one would be preffered

This is Occam's razor, which is essentially saying you should not overcomplicate things. Not that you should go out on a wild goose chase for something you find beautiful but can never test.

Anyway, there are many very good physicists who take this material (in its multiple incarnations) seriously, so its surely not junk science.

I do not think you can make this inference. By doing so you are essentially saying goid physicists cannot have any interests but physics. Of course they can! And a rather common one is the philosophy behind physics. Do not get me wrong, there is a point in thinking about such things and using them to pique people's interest.

 

[mitchell porter]

so string theory multiverse is fake physics in the sense it makes a prediction that has been falsified by the LHC

"Fake physics" is not physics that makes a wrong prediction, it's physics that makes no prediction at all, and is therefore not science. Woit doesn't say string theory has been falsified, he says it has "failed" because it predicts everything and nothing - everything because it has googols of possible vacua, nothing because you can't get numerical predictions from any of them. That would be his critique of conventional string phenomenology, anyway - the part of string theory that tries to guess the exact recipe of branes and dimensions that will give us the real world. The anthropic stuff (he would say) is even worse, since it doesn't even try to do that, it's mostly about producing dubious anthropic justifications for things already known.

But one goes too far by denying the status of science to string phenomenology and string multiverse reasoning. There are string phenomenology models and string multiverse hypotheses that do make predictions. It's just that those models and hypotheses are not as deductively clean as one would wish.

I can illustrate what I mean with another theory that you like, asymptotic safety of quantum gravity. Asymptotic safety is not a theory in the way that general relativity or supergravity is a theory. You can write an equation for them. Asymptotic safety is a property of a theory - it says that the theory has an interacting fixed point in the ultraviolet. The way it's supposed to work, you would write down the equation for quantized general relativity, and then you would simply deduce whether or not it has that infrared fixed point.

The problem is that no-one is able to do that. It's taken something like twenty years of work, just to show that the truncation of quantized general relativity to the first few terms of an infinite series expansion, has the property of asymptotic safety. What Shaposhnikov and Wetterich did, was to say, let's assume that quantum gravity is asymptotically safe, and let's assume a few other things (no new physics between the electroweak scale and the quantum gravity scale, and some technical assumptions about the gravitational contributions to the running of the Higgs couplings). From that they did deduce the Higgs mass. But note, they assumed asymptotic safety, they didn't prove that it's actually a property of quantized general relativity. And in fact they left it open as to whether some other theory of gravity, like conformal gravity or unimodular gravity, should be used instead.

Similarly, extra assumptions which ideally it would not be necessary to assume, are a commonplace in stringy models or hypotheses that actually makes testable predictions. The main exception to this might be when there is a qualitative prediction of something very exotic, like cosmic strings or particles with weird fractional charges - where you don't need to calculate much, and it's just obvious that the model contains such objects. But no such objects have been seen in the real world. So string phenomenologists are left with the task of picking a class of models that looks roughly like reality, and then working hard to identify the most promising members of that class, and to increase their quantitative understanding of how these models work.

It would be simpler if some string cosmologist could write a wavefunction for the universe, get a unique ground state, and then calculate the predicted particle spectrum. But it's not that simple. There's an enormous landscape of possible string vacua, there's no agreement about cosmology, and calculations range from difficult to impossible. So everyone just works on what they can and on what they think is promising.

 

[Dr_Zinj]

For something to be "scientific" it has to be testable to either confirm, or deny its existence. Since AFAIK the multiverse hypothesis isn't testable at this time, it doesn't qualify as science.

 

[mfb]

There is no single "multiverse hypothesis". There are models that lead to multiple universes, like eternal inflation, that lead to testable predictions about the early universe, for example.
See the example of galaxies beyond the edge of observable universe: We cannot see them. But we are sure they are there, based on a deeper model that predicts their existence. A universe that suddenly ends everywhere just behind the edge of the observable universe would be really unnatural.