Expanding universe and the Boltzmann brain problem

[mfb]

If by "result" you mean "single microstate of the system", yes, this is true.

That is exactly what I mean.

With distinguishable particles, "all particles in the upper 1/100 of the room" (which looks very odd) is as likely as "the first 1% of the particles in the upper 1/100, the second 1%in the second 1/100, ..." (which doesn't look odd). With indistinguishable bosons the first case is more likely.

Macrostates are convenient descriptions for macroscopic systems. But they are groups of microstates, and those groups are somewhat arbitrary.

Unlikely microstates happen all the time - for any sufficiently large system, every microstate is unlikely. A Boltzmann brain is a small subset of all microstates, but there is no law saying that some set of microstates is possible and some other set is completely impossible.

A Boltzmann brain is certainly possible. It is just extremely unlikely to occur within spacetime volumes comparable to the observable universe.

 

[Chronos]

I'm saying a Boltzmann brain is a special arrangement of matter that is highly improbable because it is not among the infinitude of configurations experience tells us are common. The point is a random sequence of events has an outcome. Multiplying a ridiculously low probability by a ridiculously huge number yields nonsense. Science suggests the observable universe is not infinitely old and never will be, so an infinitude of possible outcomes have never and will never be realized. Events beyond the observable universe cannot causally affect the observable universe.

 

[PeterDonis]

those groups are somewhat arbitrary.

Coarse graining of the phase space, as is done in thermodynamics, is somewhat arbitrary, yes, but only somewhat. I don't think the inevitable "fuzziness" of phase space boundaries changes the fact that human bodies are much further from thermodynamic equilibrium than a parcel of gas at atmospheric temperature and pressure.

 

[analyst5]

That is exactly what I mean.

With distinguishable particles, "all particles in the upper 1/100 of the room" (which looks very odd) is as likely as "the first 1% of the particles in the upper 1/100, the second 1%in the second 1/100, ..." (which doesn't look odd). With indistinguishable bosons the first case is more likely.

Macrostates are convenient descriptions for macroscopic systems. But they are groups of microstates, and those groups are somewhat arbitrary.

Unlikely microstates happen all the time - for any sufficiently large system, every microstate is unlikely. A Boltzmann brain is a small subset of all microstates, but there is no law saying that some set of microstates is possible and some other set is completely impossible.

A Boltzmann brain is certainly possible. It is just extremely unlikely to occur within spacetime volumes comparable to the observable universe.

Again, this is just a speculation and you avoided my question.

Sure, it's mathematically possible but we don't know if it's physically realizable. Not all things that are mathematically possible are necessarily physically realizable in nature.

The key is not that it's just one member of a much larger set of possible states

The key is in the transition from a high entropy state to a low one. It is very debatable that this is indeed possible physically.

Your logic is something like this case. I will quote the great R. Feynman:
'You know, the most amazing thing happened to me tonight. I was coming here, on the way to the lecture, and I came in through the parking lot. And you won't believe what happened. I saw a car with the license plate ARW 357. Can you imagine? Of all the millions of license plates in the state, what was the chance I would see that particular one tonight? Amazing!'For the second time, is it both extremely unlikely for you to stay in your position right now or for you to quantum tunnel through the wall? That's the point of the argument. Do you seriously believe that quantum tunneling for macroscopic objects is physically possible? That is a good analogy for a fluctuation to yield a organized structure that we see on an everyday basis.

Regards
Analyst

 

[Stephen Tashi]

The thought that brain is an improbable state of a gas seems to imply that we have some idea of what type of structure is needed to form a brain. A devil's advocate type of argument is that there might be brains emerging from gases all the time, but we cannot detect or communicate with them.

Discussions about non-biological brains usually assume they would have to share certain properties with biological brains. It's worthwhile to list some of these assumptions

1) A brain must consist of components that are somewhat similar in structure e.g. We can't have a brain where one artificial "neuron" is a lamp post and another neuron is a planet and another neuron is an art museum.

2) Each component of brain has a direct means of communication with only a limited number of other components of the brain - e.g. One could argue that in a tank of gas molecules, each molecule "affects" each other molecule indirectly, but we don't count this as a form of communication that could implement a brain.

3) Communication between the components of a brain occurs at approximately the same time scale throughout the brain. (This assumes we have some way of defining the speed and duration of communication).

I don't see any way to prove these assumptions from more fundamental assumptions. We could add more assumptions if we want to insist that non-biological brains are even more like biological brains.

 

[mfb]

The point is a random sequence of events has an outcome. Multiplying a ridiculously low probability by a ridiculously huge number yields nonsense.

All macroscopic thermodynamics is based on such a multiplication. Do you claim that all macroscopic thermodynamics is nonsense?

Science suggests the observable universe is not infinitely old and never will be

It is sufficient if it keeps getting older.

Again, this is just a speculation

It is not speculation, it is basic thermodynamics.

Do you seriously believe that quantum tunneling for macroscopic objects is physically possible?

It won't happen even once in the observable universe within the lifetime of macroscopic objects in the universe, but there is an incredibly small probability. You can calculate it.

 

[analyst5]

It won't happen even once in the observable universe within the lifetime of macroscopic objects in the universe, but there is an incredibly small probability. You can calculate it.

Give me evidence that it's possible. Do unicorns exist if the universe is extremely large?

As far as I know experimental physicists agree that it's impossible no matter what conditions are in questiom.

There exists a borderline between theoretically possible and physically realizable. And we aren't even sure if the things that we consider theoretically possible are really theoretically possible.

Physics is based on observation and experimental evidence, saying that small probability events represent something real and labeling everything as equiprobable so your argument makes sense is as speculative as it gets.

 

[durant35]

There are two issues here.

First, it is possible that dark energy is something other than a cosmological constant. In that case, it's conceivable that it will eventually dilute away.

Second, the end state of our universe with a cosmological constant, which only has a cosmological constant and no matter fields, is called de Sitter space. de Sitter is a stationary state that has no fluctuations at all. This seems to indicate that the temperature of the space is an illusion.

The thing about the de Sitter that you referred to - are you implying that the temperature of the space generally is an illusion or only the temperature of the space in de Sitter is an illusion?

I've read Sean Carroll's blog posts about no fluctuations in de Sitter and the reason he gave is that because there is no decoherence involved. Are you suggesting there is another reason for no fluctuations in those conditions?

Also, I didn't understand are the vacuum energy and vacuum temperature one and the same thing, e.g. if there is no temperature of vacuum there is no energy and no fluctuations can be produced. Is the cosmological horizon radiation the only source of possible fluctuations or does the vacuum itself fluctuate and if there was no Gibbons-Hawking radiation we would get fluctuations?

 

[mfb]

Give me evidence that it's possible.

You are reversing the logic.

Give me evidence that there are planets orbiting stars in some galaxy far away. You cannot (as of 2017) - we do not have the telescopes to see them. But we can use our knowledge about the universe to predict that they exist. There is no reason to expect that some galaxy doesn't have planets.

You are suggesting that some states, despite having the same probability as other states (that happened in our universe) in thermodynamics, do not happen at all. You suggest that the laws of physics are violated. Where is your evidence?

Do unicorns exist if the universe is extremely large?

Something that can be described as unicorn: Sure.

 

[analyst5]

You are reversing the logic.

Give me evidence that there are planets orbiting stars in some galaxy far away. You cannot (as of 2017) - we do not have the telescopes to see them. But we can use our knowledge about the universe to predict that they exist. There is no reason to expect that some galaxy doesn't have planets.

You are suggesting that some states, despite having the same probability as other states (that happened in our universe) in thermodynamics, do not happen at all. You suggest that the laws of physics are violated. Where is your evidence?

Give me one example of a macroscopic event with the same probability as quantum tunneling that happened in the history of our universe.This is fundamentally different than searching for planets orbiting galaxies. According to your logic, just because the keyboard I'm writing on has one of million possible configurations, cube-shaped planets are possible.

It is absolutely ridicuolous.

What you are mentioning is a priori probability which bears no physical significance, the only thing that it is useful is so you can theoretize and make this argument.

I'm not saying that the laws of physics would be violated or that something cannot exist, I am simply implying that while it can, it doesn't have to.

Something that can be described as unicorn: Sure.

Is this physics forum or fiction forum?

 

[mfb]

Give me one example of a macroscopic event with the same probability as quantum tunneling that happened in the history of our universe.

Give me one example of a planet more than 1 billion light years away.
Why do you ask for macroscopic events? Do you suggest laws of physics that determine the path of particles depending on exclusively macroscopic quantities? You have to know if something looks like Obama to determine the trajectory of an atom? Where is this Obama-law in particle physics?

According to your logic, just because the keyboard I'm writing on has one of million possible configurations, cube-shaped planets are possible.

I never suggested anything like that.

Is this physics forum or fiction forum?

It is a physics forum, and I'm still waiting for your evidence that the laws of physics, which clearly predict that Boltzmann brains are possible (just unlikely), are violated.

 

[Grinkle]

It won't happen even once in the observable universe within the lifetime of macroscopic objects in the universe

@mfb Do I understand your argument correctly?

I think you argue that the universe exists without end, and events that can occur will eventually occur. Above you allow that some conditions in the universe required for some events to occur have a finite life, and that bounds what will be observed for some classes of events. For example, you allow that events related to macroscopic objects can only occur during the portion of the universes life during which macroscopic objects exist. You see Boltzman Brain formation as a series of events that has no egg-timer on it. The conditions required for a BB to exist never expire, if I understand the portions of this thread discussing what I think of as heat death.

This reasoning leaves no way I can see to put bounds on anything - how do you dis-allow the spontaneous re-generation of the universe as-we-know-it from occurring, or the occasional spontaneous appearance of the odd macroscopic object or any other event one might imagine?

 

[analyst5]

Give me one example of a planet more than 1 billion light years away.

Give me one example of an unicorn anywhere in the universe. Which seems more realistic? I expect you to say something like "Atoms don't care!'

Why do you ask for macroscopic events? Do you suggest laws of physics that determine the path of particles depending on exclusively macroscopic quantities? You have to know if something looks like Obama to determine the trajectory of an atom? Where is this Obama-law in particle physics?I never suggested anything like that.It is a physics forum, and I'm still waiting for your evidence that the laws of physics, which clearly predict that Boltzmann brains are possible (just unlikely), are violated.

No, but you completely eliminated the 2nd law of thermodynamics in conjucture with macroscopic behavior from your conclusions.
I agree that the current laws of physics make that possibility unlikely and not impossible. But on what basis should one expect an extremely unlikely event to occur?

Note that the 2nd law allows us to expect the events which you call "unlikely" to happen. It also says that there might be rare fluctuations. But that isn't an absolute statement. While it might happen, nothing in nature forces it to happen.

Read the chapter 7.4. from George Ellis "On philosophy of cosmology": http://www.math.uct.ac.za/sites/def...George_Ellis/Overview/philcosm_18_04_2012.pdf

 

[analyst5]

I never suggested anything like that.

1) Each microstate is unlikely.
2) Cube-shaped planets are also unlikely.
3) Our choice of calling collection of atoms macroscopic is arbitary, so each state that we perceive and expect is also unlikely.
4) Therefore, we might expect cube-shaped planets, because we are in contact with low-likelihood events anyway.

This is your reasoning. True or false?

 

[mfb]

Give me one example of an unicorn anywhere in the universe.

I think we established that we both cannot give examples.

Which seems more realistic?

You missed my point: "I cannot give an example" does not imply "We should doubt its existence". You seem to go even further: You claim its non-existence based on its non-observation (where the non-observation so far is consistent with our expectation). That does not work at all.

I agree that the current laws of physics make that possibility unlikely and not impossible. But on what basis should one expect an extremely unlikely event to occur?

Elementary logic. If we wouldn't expect it to occur at all (given sufficient time or space), it would not be unlikely, it would be impossible.

Cube-shaped planets: Our current definition of planets requires them to be round, but that is a technical detail. Cube-shaped planet-like objects are like Boltzmann brains, just larger (which means even less frequent). I'm highly confident we don't have them in our current observable universe, but again: Take a spacetime volume large enough and you can find one. This has nothing to do with your keyboard, however.

I think you argue that the universe exists without end,

I don't know if it does. My posts are all "assuming that the universe doesn't have an end, and that expansion stays accelerated, then ...". If this is true, and if our understanding of quantum mechanics in expanding space is not completely wrong, then the universe will approach some finite temperature, making Boltzmann brains (and everything similar to that: Obama, cube-shaped planet-like objects, ...) possible forever. A "universe-as-we-know-it" appearing would require a different spacetime structure, I don't see how that would work.

 

[analyst5]

You missed my point: "I cannot give an example" does not imply "We should doubt its existence". You seem to go even further: You claim its non-existence based on its non-observation (where the non-observation so far is consistent with our expectation). That does not work at all.

No, I'm completely agnostic to the possibility of its existence. You are extrapolating things without a doubt where there should be plenty of reasons for doubt, not including empirical evidence.

Elementary logic. If we wouldn't expect it to occur at all (given sufficient time or space), it would not be unlikely, it would be impossible.

Cube-shaped planets: Our current definition of planets requires them to be round, but that is a technical detail. Cube-shaped planet-like objects are like Boltzmann brains, just larger (which means even less frequent). I'm highly confident we don't have them in our current observable universe, but again: Take a spacetime volume large enough and you can find one. This has nothing to do with your keyboard, however.

I think nobody except you expects cube-shaped planets to exist, and that speaks enough about your 'elementary logic'. Again, your logic is pretty naive: If the universe is infinite, everything that can happen will happen. I have doubt that the universe works that way.

I don't know if it does. My posts are all "assuming that the universe doesn't have an end, and that expansion stays accelerated, then ...". If this is true, and if our understanding of quantum mechanics in expanding space is not completely wrong, then the universe will approach some finite temperature, making Boltzmann brains (and everything similar to that: Obama, cube-shaped planet-like objects, ...) possible forever. A "universe-as-we-know-it" appearing would require a different spacetime structure, I don't see how that would work.

That is exactly what modern cosmology tries to avoid and what is regarded as a fatal flaw in the flat-lambda model. So these are pretty unevidenced assumptions to begin with, and a lot of if's to begin with. If you don't find the scenario controversial, then I'm afraid I have nothing else to say to you.

 

[Grinkle]

assuming that the universe doesn't have an end, and that expansion stays accelerated, then ...". If this is true, and if our understanding of quantum mechanics in expanding space is not completely wrong

That helps, thanks.

A "universe-as-we-know-it" appearing would require a different spacetime structure

I'll change my extreme case to observable universe, then, of which the spontaneous appearance, I think, is only different from a human or a planet in likelihood, not in principle.

I do not have anywhere near your depth of understanding of quantum mechanics so me saying 'your logic is sound' would be silly, I can't assess the adequacy of your assumptions, but for what its worth I credit you with not being wrong in your list of assumptions and the conclusion certainly follows.

My reaction is that you are pointing at a free lunch allowed by current theories, or perhaps an infinite buffet of free lunches, and I expect that if we humans can refine our knowledge of physics further we will find that physics does not predict that everything happens sooner or later. I find @analyst5 's arguments that not being able to establish a zero probability for an event should be insufficient to conclude the event will happen compelling.

 

[analyst5]

That helps, thanks.
I'll change my extreme case to observable universe, then, of which the spontaneous appearance, I think, is only different from a human or a planet in likelihood, not in principle.

I do not have anywhere near your depth of understanding of quantum mechanics so me saying 'your logic is sound' would be silly, I can't assess the adequacy of your assumptions, but for what its worth I credit you with not being wrong in your list of assumptions and the conclusion certainly follows.

My reaction is that you are pointing at a free lunch allowed by current theories, or perhaps an infinite buffet of free lunches, and I expect that if we humans can refine our knowledge of physics further we will find that physics does not predict that everything happens sooner or later. I find @analyst5 's arguments that not being able to establish a zero probability for an event should be insufficient to conclude the event will happen compelling.

It's not a problem, according to mfb once we die, we will all be resurrected many, many times and our observable universe also. And we will spend time with unicorns and Obama's twin in thermodynamical equilibrium.

Not bad for a physics expert and mentor which should brighten the mind and give clarity to many young physics enthusiasts like me and supposedly you.

 

[mfb]

No, I'm completely agnostic to the possibility of its existence.

I got a different impression from your previous posts (e. g. "If a theory predicts this kind of stuff, it must be false").

I think nobody except you expects cube-shaped planets to exist

I don't have a survey on cube-shaped planets, obviously, but I would expect the majority of physicists to assign them a probability of existing, with the implication (by definition of probability) that, given enough time or space, that there will be one somewhere with near certainty.

So these are pretty unevidenced assumptions to begin with

It is the easiest model consistent with observations, but it doesn't matter: If we discuss implications of the model we have to assume that it is true. Other assumptions will lead to completely other consequences, and should be discussed in separate threads.

 

[durant35]

It is the easiest model consistent with observations, but it doesn't matter: If we discuss implications of the model we have to assume that it is true. Other assumptions will lead to completely other consequences, and should be discussed in separate threads.

I agree with you on this one, but the fact that it has a temperature is controversial. As member Chalnoth already said, de Sitter doesn't have temperature. So if you're going to discuss an eternal expanding universe it is quite plausible that it does not have temperature and that it is static.

 

[Stavros Kiri]

So then the total number of unsatisfied possibilities continually increases the larger the universe becomes. Again, pardon my ignorance, but I don't see how we can expect to satisfy every possibility even if x is an infinite number.

Careful that there are two types of infinite (with substantially different topology for each one): discrete and continuous.

I think what you are all trying to investigate and interpret here is a topological issue of probability theory, and I think mfb is correct (not to say that others aren't ...) , + see my next following post.

Unlikely microstates happen all the time - for any sufficiently large system, every microstate is unlikely. A Boltzmann brain is a small subset of all microstates, but there is no law saying that some set of microstates is possible and some other set is completely impossible.

A Boltzmann brain is certainly possible. It is just extremely unlikely to occur within spacetime volumes comparable to the observable universe.

every microstate is unlikely

So a better way of responding to mfb's claim is to say that, for example, the phase space volume in which the state "Barack Obama is in front of me" occurs is much smaller than the phase space volume in which the state ...

I think that what mfb may be trying to say is that Improbable is not Impossible. Even an event with zero probability can happen!
Believe it or not, we see this all the time!
(Examples upon request).

[And this is true even in classical statistical physics, which offers a mathematical explanation of thermodynamics ...]

 

[analyst5]

I got a different impression from your previous posts (e. g. "If a theory predicts this kind of stuff, it must be false")

If a theory predicts that this stuff dominates, than it is almost certainly false. Only you think that there's nothing unusual being in a minority of non-Boltzmann brains so you base your conclusions based on a premise that is false and bizzare.

On the other hand, if a theory predicts this stuff ocasionally, it is certainly weirder than everything than we have observed so far from the beginning of the universe.

Again, I've red many opinions from experimental physicists which consider macroscopic quantum tunneling a fiction which is impossible no matter what circumstances are involved, no matter if the universe is infinite in size and so on. Only you give a realistic shot for that to occur somewhere.

I don't have a survey on cube-shaped planets, obviously, but I would expect the majority of physicists to assign them a probability of existing, with the implication (by definition of probability) that, given enough time or space, that there will be one somewhere with near certainty.

If you want to I'll start a new thread which will be named "Do you believe cube-shaped planets and unicorns exist somewhere?" or create a poll of the same name so we could get opinions of many smart people on this forum. Again, nobody really believes that oddities actually exist. Boltzmann brains are a statistical tool which is used as a reductio ad absurdum, which is similar to the Schrodinger's cat scenario. Only you find it extremely plausible.

It is the easiest model consistent with observations, but it doesn't matter: If we discuss implications of the model we have to assume that it is true. Other assumptions will lead to completely other consequences, and should be discussed in separate threads.

I don't disagree with the model, the problem is your interpretation of the model which includes Boltzmann brains. Sean Carroll, Don Page and many other authors have spent years with different hypothesis to make the flat lambda compatible with the fact that there is a small number or none odd structures. Again, only you don't have a problem with nothing of what I mentioned.

Maybe you should read some metaphysics before going into drastic extrapolations. And I don't want to change the subject to philosophy because I am strictly interested in science and physics, but I feel that if you have the right to come to conclusions such as "Unicorns exist" that I may advice you to read something concrete which will make you realize how bizarre these extrapolations are.

I'm sure you haven't read a letter from the George Ellis book I sent you, but that's ok, who needs opinion on metaphysics of cosmology from one of the greatest cosmologists ever, you've got infinity, small probabilities and unicorns to develop your opinions.

 

[mfb]

Again, I've red many opinions from experimental physicists which consider macroscopic quantum tunneling a fiction which is impossible no matter what circumstances are involved, no matter if the universe is infinite in size and so on. Only you give a realistic shot for that to occur somewhere.

Check the statements again. Quite sure they talk about things we'll observe. Not about things that are possible with but so unlikely that we won't see them here on Earth.
We also have a FAQ article about it. Same conclusion: The probability is so tiny that we can safely neglect it here on Earth. But it is not zero.

If you want to get sarcastic and personal: feel free to, I don't care, but then continuing the discussion is pointless.

 

[analyst5]

Check the statements again. Quite sure they talk about things we'll observe. Not about things that are possible with but so unlikely that we won't see them here on Earth.
We also have a FAQ article about it. Same conclusion: The probability is so tiny that we can safely neglect it here on Earth. But it is not zero.

If you want to get sarcastic and personal: feel free to, I don't care, but then continuing the discussion is pointless.

Fair enough. Just tell me, why did you avoid commenting on any of the implications of your statements on my previous post?

 

[mfb]

Because you got sarcastic and personal, and I don't think that allows a reasonable discussion.

 

[analyst5]

Because you got sarcastic and personal, and I don't think that allows a reasonable discussion.

You're right, my bad. I agree that this disallows a reasonable discussion. I apologize.

If you're okay with the fact that I would like to continue the discussion, I hope that you can write your opinion because I respect it despite occasional disagreements.

 

[Chronos]

I agree with MFB, never say never, although IMO Boltzmann brains, Boltzmann unicorns, etc., are so egregiously improbable they never have nor will occur within the lifetime of the observable universe.

 

[Stavros Kiri]

Boltzmann brains, Boltzmann unicorns, etc., are so egregiously improbable they never have nor will occur within the lifetime of the observable universe

How do we know that God isn't a Boltzmann brain?
[[But] In that case He was created by the Universe, and not vice versa ... (?)]

 

[Chronos]

There is at least a miniscule probability some things will remain forever unknown.

 

[analyst5]

Where is the unexplainable position? In a universe that produces some "normal" brains and then Boltzmann brains later, the normal brains should not rule out that they live in such a universe, because they would be wrong.

There is also the probability that you are a Boltzmann brain.

Especially if both the number of normal and Boltzmann brains is infinite, probabilistic considerations stop making sense.

This is what is obviously wrong, and not only in my opinion but in opinion of many who have studied the subject precisely.

If there are infinite BBs in the future, normal brains would have no logical reason to believe their observations because each normal observation would be vastly outnumbered by a BB observation of the same kind. There would be only an infinitesimal chance that that they are correct to rule out that they are not BBs.

 

[windy miller]

The Botlzman brian problem assumes that a single brain is more likely to fluctuate into existence than a single universe , as a single universe ( which contains brains) is less likely than a single brain.
But I think inflation and Darwin have undermined this claim. Inflation implies a sub atomic sized seed of inflationary matter is all you need to make an entire universe this universe can then evolve and over millions of years and with Darwinian evolution brains are created form other simple beginnings. The seed and the Darwinian process are not necessarily less likely than the Boltzman brain and therefore I don't see who the BB is a well posed problem.

 

[durant35]

They should be possible. Assuming I am not a Boltzmann brain, the universe is so young that regular brains should be much more common, but in the very distant future Boltzmann brains could dominate. Most of them will have weird inconsistent memories, but some will have a brain like we have. This is incredibly unlikely for a given place and time, but if you have a finite chance for it and infinite time and/or infinite space, it wil happen.

https://arxiv.org/abs/1702.00850 [Sean Carroll - Why Boltzmann Brains are bad]

Read this.

 

[nikkkom]

then eventually there will be an infinite number of such brains, regardless of how absurdly rare they are.

Not true. A Boltzmann brain is not indestructible. In fact, its destruction due to a random quantum event (a change which makes it non-operational, just a lump of dead machinery) is very, very much more likely event than creation.

IOW: Boltzmann brains "decay" and have some "half-life". Since they are created astoundingly rarely, ones which _are_ created eventually break down. Every individual brain is unlikely to ever see even just one another (functioning) brain.

 

[Chalnoth]

Not true. A Boltzmann brain is not indestructible. In fact, its destruction due to a random quantum event (a change which makes it non-operational, just a lump of dead machinery) is very, very much more likely event than creation.

They should be destroyed as quickly as they're created. I don't mean an infinite number at once, but rather an infinite number created in total.

 

[rootone]

I think I understand why Boltzmann gave it up now.
The issue is not that the existence of such observers is ruled out by data, but that the theories that predict them are cognitively unstable: they cannot simultaneously be true and justifiably believed.