Is the multiverse cosmology or metaphysics?

[yrreg]

There have been a few threads of late on the multiverse concept in cosmology, and whether it can be viewed as a viable, albeit currently underconstrained cosmological theory that is leading us to demonstrably correct discoveries about our universe, or if it is essentially a fairly arbitrary metaphysical conviction that is masquerading as science. I'd like to advance the latter thesis, and central to my argument is the Popperian stance that if, as Feynman said, science should be a way to keep us from fooling ourselves, then we need theories that make "risky" predictions-- predictions that, were we to be skeptical of the theory, we would expect to fail. A theory that only makes predictions that no one can expect to fail, even if they discount the theory, is more like a technique for performing rationalizations than it is a technique for making predictions.

[...]

"There have been a few threads of late on the multiverse concept in cosmology, and whether it can be viewed as a viable, albeit currently underconstrained cosmological theory that is leading us to demonstrably correct discoveries about our universe, or if it is essentially a fairly arbitrary metaphysical conviction that is masquerading as science."

So, what is the finding of the published savants of astrophysical and sub-atomic cosmology?

Have they taken a vote on the multiverse whether it is science or masquerading as science?



Yrreg

 

[twofish-quant]

Forgive the interruption again, but tell me: Did the math come after some people thought up the idea of the multiverse, meaning, the math was engineered in order to give substance to the idea of the multiverse or to prop up the multiverse?

Nope. People weren't looking for multiverses, it just comes out of the math.

Let me talk about one specific example, which is the one that I know the best. The eternal inflation model.

There is a lot of evidence that the universe underwent a huge amount of expansion in the early universe (see wikipedia article on inflation). So then theorists think about what could have caused this expansion. It turns that's easy, you can invent a lots of ways to have the universe suddenly expand. The hard part isn't figuring out how to make the universe expand. It's to try to make the universe stop expanding.

OK, so the theorists go into the back room and figure out ways to get the universe to stop expanding. In the case of inflation, the conditions are extreme, but they aren't so extreme so that our theories will totally fall apart, and people have come up with various mechanisms to stop the universe from expanding rapidly and to get the universe to settle down to "normal" expansion.

But we have another problem. Most of those mechanisms are 100% effective. Which means to say that you can stop rapid expansion in *part* of the universe, but you aren't stopping expansion in *all* of the universe. At that point a theorist points out that this will still work. You don't have to stop rapid expansion in all of the universe, you just need to stop rapid expansion in the part of the universe that we see, which means that mechanisms which stop rapid expansion in an area of the universe that is larger than we see are still possible.

But... Suppose you have a mechanism that stops rapid expansion in 99.999999% of the universe. Most of the universe stops rapid expansion. A tiny part doesn't. Very quickly that tiny part is going to keep expanding and much up most of the universe. Now most of that section may eventually stop expanding, but all you need is a tiny part of that doesn't, and the process ends up going on forever...

At this point you have the eternal universe picture. Now we aren't exactly sure what is the nature of the thing that caused the universe to inflate. There are about hundreds of different ideas. However, you can mathematically show that most of those will lead to the situation in which some part of the universe doesn't stop inflating.

Now as we get more information, we can reduce what is possible. It's possible that once we go through all of the data that the mechanism that works will stop inflation in 100% of the universe. But a lot of the things that are possible right now, will just stop inflation of part of the universe, that that leaves the rest of it expanding. The other thing is that the different mechanisms are the result of high energy physics, so it's possible that particle experiments will give us the "right equation."

I should point out that it may make more sense to stop using the term "multiverse". In the inflationary picture, you have "part of the universe that we can see and that has stopped inflating" and "part of the universe that we can't that hasn't."

Or there was established math going on in the mathematical exposition of a physical phenomenon, and some people could not help but come to posit the idea of the multiverse?

If you start the universe inflating, and the mechanism you use to stop it inflating isn't 100% effective, then you will get large parts of the universe that keep inflating. That's a direct consequence of the math.

But you don't need the "stopping mechanism" to be 100% effective to work. You just need it to work in the parts of the universe that we can see, and then it could work. It's a lot easier to come up with a stopping mechanism that is say 10% effective or even 99.99999% effective than 100% effective, and if the stopping mechanism is anything less than 100% effective, you are going to have "parts of the universe that don't stop expanding" and the one part of the universe that doesn't stop inflating is going to make up most of the universe.

And they continued with the established math without engineering or if I may inventing a new math, in order to expound on the multiverse?

No one invented anything particular to get "multiverses." In fact, for inflation, it probably makes more sense to talk about "different parts of the universe" rather than "multiverses."

 

[Ken G]

But there is still "unknown physics". For example, in trying to figure out how supernova work, it's unlikely that we will find out that the speed of light is not constant or that gravity doesn't work through GR. However, putting the pieces together still involves "unknown physics."

It is important to distinguish between physics as we know it, playing out in as yet unknown ways, versus inventing completely new physics-- especially in regard to whether or not there remains an xunsatisfied need for making "risky" predictions.

In the case of inflation, the high energy physics is known well enough so that you are limited in what you can "make up" and so it's a matter of connecting the dots.

Personally, I'd say that anyone who imagines that the story of inflation is going to be just "connecting the dots" of current physics is not being at all honest to themself, but only time will tell who is right-- and probably a lot of time. I'll settle for, if we are on our death beds, and the "dot connecting" isn't done yet by then, then we can agree it had to be something more than dot connecting after all.

Also, the definition of "risky prediction" is problematic. What's the difference between a "risky prediction" that has already been made with "non-risky prediction" that has already been made.

Not "problematic" at all: A risky prediction is a prediction that anyone who didn't understand the theory that made it would have no reason to expect to hold.

It's also known from high energy experiments that the effective coupling constants do change at high energies.

I certainly wasn't aware that non-perturbative effects observed in the strong force is evidence for a multiverse, rather than simply evidence for non-perturbative effects in the strong force. I daresay you are getting a bit caught up in the string theory hype now. I haven't seen how string theory has fulfilled the grandiose claims we often see about it, and the way it doesn't constrain the coupling constants is usually seen as a serious weakness of string theory-- not a strength that is explaining the observations.

It's a *prediction* of string theory.

That's just not what a prediction is, period. A prediction always looks like this: experiment A will come out X.

If you accept string theory then you end up with multiple universes in much the same way that if you accept the Coprenician model of the solar system, you quickly end up with exoplanets.

The analogy is not good, because establishing exoplanets involves predictions of exactly the form I just said. Even Newton could have thought of a dozen ways to directly detect them, in his sleep. He just wouldn't have had the technology to do it.

I'm not a string theory expert, but this is what the experts tell me. If you don't believe that there are multiverses, then string theory is wrong.

I do know a string theory expert I can ask that. I expect he'll say you can accept any theory without buying the metaphysical baggage (you know, "shut up and calculate", and all that).

The same is true with exoplanets. Once you accept the Coprehenican model of the solar system and Newtonian physics, then you pretty much had to accept that other planets existed even if you couldn't see them.

In many ways the Copernican principle is the opposite of the multiverse, as the latter asserts that we are living in a very special universe, to account for its seemingly finely tuned attributes.

If you reject the concept of multiverses, then you *must* logically reject string theory, which is what a lot of people have done.

I don't believe that. I'm sure there's no difficulty rejecting the existence of multiverses, and accepting string theory anyway, as an empirical prescription for making predictions (i.e., as physics, not metaphysics). The same could be said for quantum mechanics-- if one requires that all time evolution be unitary, then one must hold to some kind of many worlds, but one can simply view unitary evolution as a provision of some aspect of the predictive machinery of quantum mechanics, rather than some metaphysical claim on reality-- and still use quantum mechanics quite easily, with only one world.

To put it crudely as to why, string theory works by symmetry, and in order to have a symmetry that explains the current universe you have to set certain quantities as "quantum mechanically random".

It's fine to invoke the symmetry, and set the randomness. That's all a prescription for making testable predictions, not a metaphysical claim on reality. The virtual particle example comes to mind again. Symmetries are made to be broken in reality, they are just mathematical devices. Does someone have to believe symmetries really exist to use them in deriving constants of the motion in classical physics? Only the devoted rationalist actually believes their own postulates (and quite against the weight of historical evidence, I might add).

If something is quantum mechanically random, that means that different observers in different parts of the universe will "flip the Schroedinger coin" differently and come up with different numbers for those quantities. At that point you explain away the fact that we *don't* see things like the fine structure constant changing in our universe by the fact that some sort of inflation expanded the universe so that the parts that we see are all parts where the "coin got flipped the same way" but that must mean that there are parts that we don't see in which the "coin got flipped differently."

That's all well and good, and so let's observe those other regions where the constants are different. Not possible? Then it ain't science. What is the technology that would be required to detect those other universes where the physical constants are different? The purpose of good science is to answer just that kind of question, not to make metaphysical claims on existence.

It's not as if people like the idea of multiverses and added it because it was fun.

I understand, but what I'm saying is the multiverse was added by a strong desire to rationalize issues that we really just have no explanation for. Those who tend to think highly rationalistically, and think the universe "is governed by laws", as if the universe looks up a law or makes a calculation every time it does something, will always be looking for ways to rationalize what happens. But they have to be honest to themselves and ask-- are they just spinning a tale, like a creation myth, because it helps them believe in their rationalistic picture? Or do they really have scientific evidence that their claims on reality are getting supported by successful risky predictions?

People generally hate the idea of multiverses, and only add them because the math gives them no choice in the matter.

Just look at those words-- the math gives them no choice. Quintessentially rationalistic! This is my whole point, the math doesn't rule the universe, the universe, as analyzed by our ability to perceive it, gives rise to the math we put into our theories. Math involves idealizations and approximations, it doesn't make claims on what exists-- unless you adopt rationalist metaphysics. Take the Greek belief in orbits as perfect circles, that was the math leading what exists. But the empiricist works the other way-- the observations are crudely consistent with circles, circles are simple, build a model using circles, but never make claims on what exists based on the math of circles, because you know it's going to be an idealization from the get go, and you know it is going to break down at some level of precision.

(Also, I'm not a string theorist, so if I've messed up the explanation, corrections are appreciated).

Your descriptions actually sound pretty reasonable, I think you are doing a good job of presenting their case, kudos to your descriptions-- but you are missing how rationalistic and metaphysical that case is, and how extraneous it is to the actual predictions of string theory (if and when there really are any such thing), as per the "shut up and calculate" approach to physics.

No. It's not. Define "exist" and then problem goes away. Do numbers "exist"? Does anything outside my head, "exist"? Do I "exist"?

Those are the questions of metaphysics. Obviously metaphysics goes away if you resolve metaphysics, but you can't, which is why it's metaphysics.

I don't see how the "existence" of "virtual particles" are any more problematic than the "existence" of "financial risk."

Financial risk is not an ontological claim, virtual particles are to some, and not to others, so they argue the metaphysics of them. You might not like to enter into metaphysical arguments, which is fine, but that doesn't make it not a metaphysical argument-- like the existence of the multiverse.

Personally, I'd argue that if you want to argue philosophy, it's best to do that with examples that you are familiar with.

I agree, which is the entire reason I started this whole thread with an analogy that we are all familiar with! But no one has chosen to avail themselves of the analogy to argue their point.

No you can't. (At least that's what the string theorists tell me.)

I don't believe them, I think they are not using their imaginations. I can easily think of a way to use probability distributions on the coupling constants without requiring that any more than one of them is ever instantiated. You need to observe the others or else they are just a mental construct, the "math" telling us what is, like virtual particles.

If you have one uranium atom and you watch it decay, you *can* create a set of interpretations so that there isn't some "phantom" atom in an alternative universe.

The trouble is that if the decay of atoms is quantum mechanically random then if you have two uranium atoms next to each other then will probably decay at different times.

Yes, but you have two uranium atoms. We don't have that, we have one universe to observe, and we have no idea if there even is another "uranium atom"-- it's pure rationalization that the other one exists. Make a risky prediction that requires that other uranium atom, and then you have something, not just some math that can be rationalized that way (like universal unitary evolution in quantum mechanics).

So what happens is that if force coupling constants are quantum mechanically random (and string theory says they are), then when they get set up at the early universe, you are going to end up with different values for different parts of the universe. You can "fix" the problem by blowing up the universe so that we can see only one value, but that creates "other parts of the universe" with different values.

Well, the mathematical model "creates" that, but creation doesn't actually occur by mathematics-- unless you are a devoted rationalist. Many who are not of that metaphysical bent see that as reversing the correct logic.

I'd like to see you try, because that will help me to explain chaotic inflation to you.

It's easy, you only think it's hard because you have limited the metaphysical options. Let's say you give me a theory that says space spontaneously inflates, at random, and with random coupling constants. That's a form of chaotic inflation. You use the theory to make risky predictions (let's pretend that was actually possible, maybe it will be someday), which pan out, so we like your theory.

Now, to you, it makes sense to say that this implies the existence of multiple universes, because inflation keeps happening. I say that your theory is a theory devised by an intelligent being that is capable of interacting with its universe. Of course it is not what the universe is "actually doing", that's an absurd rationalist pipe dream that has never actually been true in the entire history of physics. So instead, we quite reasonably interpret your theory as a kind of idealized toy that makes nice predictions. Then we ask, does this imply the existence of other universes?

Well, we have no way of knowing if the inflation in your model has ever happened before, or if we are the first event. Can we say which is more likely? Well, if your theory only has meaning for describing how an intelligence interacts with its environment, because that's really where the meaning in any physics theory lies, and quite demonstrably so, then it has no sway at all over any situation that is logically inconsistent with the presence of intelligent beings, like coupling constants that cannot be inferred by any intelligence because they are inconsistent with the presence of intelligence. So there is no "fine tuning problem", because a universe that has intelligence in it to interact as we do must infer the constants we infer, and a universe that cannot have intelligence in it to interact with that universe is a meaningless construct that we have no language to even describe, let alone calling it "another universe." Thus, we cannot claim that our universe is unlikely so must be one of many-- go back to the analogy in the OP, I introduced it precisely to handle this situation. As easily as we take the metaphysical stance that chaotic inflation (a random deal) implies there must have been many other deals, we can take the stance that the rules of the game were devised expressly to make the hand we were dealt special in exactly the way we find it to be special. Or, we can just as easily say that the deck from which the cards were dealt (that "random" quantum mechanical distribution) was stacked to produce a hand that could infer a theory like that, given that theories can only be inferred by intelligent beings. So it's all a question of stepping away from the assumptions that highly rationalist thinkers make, those who tend to see the laws as being separate from our involvement in arriving at them.

Now, you may object to my alternatives on the grounds that they seem like less natural assumptions than a multiverse. I might counter that they aren't unnatural at all if we say that theories are beholden to our own ability to arrive at them. But that debate is a metaphysical debate, proving the point that the existence of the multiverse, even for chaotic inflation models themselves, is an inherently metaphysical issue.

So you are claiming that you can take any model with eternal inflation and turn it into a model without multiverses with exactly the same predictions.

You betcha. The analogy in the OP was meant to convey that.

Define "really happening" and "physical existence."

That is a metaphysical exercise. If you argue that is necessary to understand if the multiverse exists or not, I say, "yup, you got it."

I've never been to Paris, but I can figure out that it exists.

Indeed, so let's look at how you figured that out. Was it an interpretation of some law of physics that led you to that conclusion? I'll bet it was observations that led you to that conclusion-- you observed photographs of Paris, you observed people talking about Paris, you observed books that detailed Paris and also talked about other things that you did find to pan out, so you trust the book. Basically, you amassed a wealth of direct empirical information that Paris exists, information that you would have no reason to expect to observe if you were skeptical about the existence of Paris. The existence of Paris makes risky predictions!

Talking about parts of the universe we can't see isn't much different from talking about places in the world that I can't go.

Then what are the risky predictions, that's what I keep asking.

One reason that I tried to avoid getting into philosophy is that I thought that we could *avoid* the philosophical issues once I established *why* cosmologists are talking about multiverses. You take the physical laws as we think they exist or that could reasonably exist, and you end up with large parts of the universe that we can't directly see.

If you think physical laws can dictate to what you can't see, which is rationalistic. The empiricist points out that to avoid fooling yourself, you still have to support your stance with successful risky predictions. So it has been true since the ancient Greeks, who were masters of rationalization-- and masters of fooling themselves.

It's not.

And this is the sole substantive issue that has emerged-- whether or not chaotic inflation makes risky predictions that have been supported. Gaussian noise has supported the idea that the noise is quantum mechanical, which supports the inflation phenomenon. That's all I've seen referred to, in places like the WMAP website-- support for inflation, not a multiverse. But even if the single-shot models of inflation that people were toying with did not make that prediction, causing chaotic forms to be currently in vogue, these models are still so vague that it's not clear much more can be said. Those watching the signposts might say they are pointing toward chaotic inflation models, but it remains a leap of faith that that flavor of model can work, because none have yet. There is no model of inflation that is currently indicated as the proper model, which makes the whole issue highly hypothetical at present. But as I stressed above, even for chaotic inflation, the claim that this implies that a multiverse exists is a metaphysical claim, because other ways to interpret that very same model, as in the card game analogy, simply express different metaphysical assumptions-- and result in different conclusions about the need for a multiverse.

The other thing is that we are making some progress. There's been a lot of work on "decoherence" and it's pretty clear that "naive Copenhagen" isn't going to work. The problem with "naive Copenhagen" is that you collapse the wave function by waving a magic wand, but there are lots of systems in which it's not obvious when you should wave that wand.

This is off topic, but the only thing "naive" about the Copenhagen interpretation is most people's understanding of it. What Bohr was actually talking about was the fact that physics involves an interface between the physicist and her environment, and collapse occurs somewhere in that interface. It was never a requirement of the Copenhagen interpretation that that collapse happen at any given place in that interaction, Bohr knew that he simply didn't have enough constraints to say more. Decoherence doesn't change that in the least, the Copenhagen interpretation has no difficulty whatsoever in accounting for gradual decoherence, because even the experiments that study gradual decoherence always involve "crossing the Heisenberg gap" at some point along the way-- they still represent the interaction of the physicist with his/her enviroment, so Copenhagen still says the collapse is inherent in that interaction, it's just collapsing an incompletely decohered system.

I've just spend a few pages trying to explain why you've got your physics wrong.

You might imagine that's what you've done, but I still do not see a single statement I made in this entire thread that could be considered wrong physics. Please quote the statement if you think otherwise, rather than making vague and opinionated remarks.

 

[twofish-quant]

So, what is the finding of the published savants of astrophysical and sub-atomic cosmology?

The published savants of astrophysical and sub-atomic cosmology talk about the issue to try to avoid "unnecessary philosophy."

A lot of the math is hard, so much of what gets published is of the form "if you assume X, then Y is going to happen." If you don't like Y, then X is wrong. If you can't accept multiverses, then you can't accept string theory as it currently exists.

Have they taken a vote on the multiverse whether it is science or masquerading as science?

As it is discussed in the professional papers, it's clearly science. One problem is that what is presented to the public tends to oversimplify things, and it's different from what's in the professional papers.

If you ask me about whether "multiverses" exist, the short answer is "I don't know." The medium answer is what I'm trying to represent here, and it's boring to a lot of people. The long answer involves lots and lots of greek letters. It's not going to make for an interesting video on Youtube.

Whereas if I start saying that "multiverses exist and God doesn't" that gets people's attention, and the stuff goes virial.

One thing that I appreciate about your asking these sorts of questions is that it shows that you (unlike most people) have the patience to listen in what's really going on, rather than just go for the "gee-whiz, quick answer" stuff.

 

[Chronos]

The thing that bothers me about 'parallel' universes is the lack of empirical evidence. I agree the mathematical basis for a 'multiverse' is solid, but, unlike Tegmark, I am unwilling to concede their existence without empirical evidence.

 

[Ken G]

What's more, I'd say, and indeed have said, that Chronos' empiricist standpoint is extremely common among professional astronomers and educators. The one place where it is much rarer is among cosmology theorists, which goes a long way to explain the divide that is being explored in this thread, as well as the importance of guidance from thinkers like Popper to help avoid the rationalization problem.

 

[ImaLooser]

If no one is listening, then it's a waste of my time.

I'm listening and finding it very interesting. I had no idea about this.

 

[twofish-quant]

The one place where it is much rarer is among cosmology theorists, which goes a long way to explain the divide that is being explored in this thread

It's actually not rare. The problem with asking for empirical evidence is "what constitutes empirical evidence?" If we want to know if eternal inflation is true or not, then what exactly is the observational signal that we should be looking for?

The job of the theorist is to come up with the answer to this question. If multiverses exist, then what are the observational consequences? Trying to figure this out doesn't mean that the person thinks that multiverses exist.

 

[Ken G]

The job of the theorist is to come up with the answer to this question. If multiverses exist, then what are the observational consequences? Trying to figure this out doesn't mean that the person thinks that multiverses exist.

I don't think that it does, and yet, it is all too clear that the "persons" we are talking about do not so limit their rhetoric, which is the whole issue. If cosmology theorists limited their language to statements like "my model, motivated by multiverse thinking, predicts this," there would be no motivation to pose the question of this thread-- a theorist making predictions is clearly doing good science, and is allowed to cite any motivation they like. Instead, what we see are highly rationalistic arguments like "the math leads us to conclude this", or arguments along the lines of, since this very simplified class of models, involving scalar potentials that have never been observed in any experiment, require a chaotic form of inflation to get the observations, we can conclude that inflation is chaotic and our universe is one of many. It's similar to statements like "to accept string theory, you have to accept a multiverse"-- they are simply not statements like "expect to observe A if there is a multiverse, and not A if there isn't." That is why, at the very least, we can say the case has not been made that a multiverse is not a primarily metaphysical construct. That may not always be the case, but it appears to be the current state of affairs. Perhaps it goes too far to therefore assert that the multiverse is not science, but it also seems to go too far to assert that anyone who demands an empirical foundation for statements of what exists should hold that the multiverse does.