The Block Universe - Refuting a Common Argument - Comments

[PeterDonis]

PeterDonis submitted a new PF Insights post

The Block Universe - Refuting a Common Argument

Continue reading the Original PF Insights Post.

 

[andrewkirk]

I don't know what to make of things that are described as 'arguments for (or against) the block universe'. To me it is like having arguments for or against ice cream, or Justin Bieber. The way I see it, metaphysical theories such as block universe are just stories that humans tell themselves and each other in order to come to terms with the world and to imagine it in a satisfying way.

One can imagine the world as a fixed block (McTaggart's B Theory, aka blockworld), or as a growing block (McTaggart's 'A theory', same ref) or as a sliver ('presentism'), according to what one prefers. It is possible to devise metaphysical hypotheses for each of these that are consistent with known physics.

Personally, I find the B Theory preferable on the grounds of Occam's Razor, based on the following argument. If we imagine the world as a growing block then we can always take the union of all such blocks, over all time, to get a fixed block, in which everything is 'certain'. A pan-dimensional being that could see that fixed block could also see all the growing blocks arrayed before them in their different stages of development. Further, one needs the 'current universe' to extend at least a little into the future (ie for every point in it to have a neighbourhood that overlaps with the future) in order for the Riemann, metric and stress-energy tensors to be defined.

Given then, that the growing block implies an ultimate fixed block, and requires a block that extends at least a little into the future, why not just discard the intermediate stage from one's picture and a have a simpler model in which we only imagine the fixed block?

But to repeat, I see this as entirely a matter of taste.

 

[PeterDonis]

I don't know what to make of things that are described as 'arguments for (or against) the block universe'.

In the particular case I describe in the article, the argument is that SR, specifically relativity of simultaneity, "requires" the block universe interpretation (in the sense that that interpretation is logically necessary given the postulates of SR). That is the argument I am refuting. More general or philosophical claims about whether the block universe is "real" are out of scope, at least for this particular discussion. I was only discussing a particular logical argument and why it is invalid.

I see this as entirely a matter of taste.

As far as "interpretations" in general go (not just of SR but any physical theory), I entirely agree. But again, I would like to keep this thread focused on the specific logical argument I refute in the article.

 

[andrewkirk]

I think that the 'argument' you are refuting is - in Pauli's immortal words - 'Not Even Wrong', and hence not amenable to refutation.

The argument of the necessity of a blockworld is simply incoherent, because it assumes that the key terms like 'certain' (and its putative antonym 'uncertain') mean something, without saying what it is that they mean. When one focuses the microscope on those terms to try to pin down a clear meaning, they dissolve into nothingness.

The sort of constructions that Putnam et al get up to with 'arguments' like that are the sort of thing that JL Austin demonstrated to be mere word games in developing his ideas of Ordinary Langauge Philosophy.

 

[dl58]

It is a reductio ad absurdum.
1. Assume the future is uncertain.
2. Show the contradiction "according to one of them, the decision lay in the uncertain future, while to the other, it lay in the certain past."
3. Conclude the future is certain (no destinction between the uncertain future and certain past)

The problem with your past light cone approach is that it requires a preffered event. Which event decides what is real and what is not. Consider 3 events. 1. happens in 1950. 2. happens on 1960. 3. happens in 1970. 2 is real according to 3 and unreal according to 1. The alternative to a preffered event would be to declare reality to be relative.

 

[PAllen]

It is a reductio ad absurdum.
1. Assume the future is uncertain.
2. Show the contradiction "according to one of them, the decision lay in the uncertain future, while to the other, it lay in the certain past."
3. Conclude the future is certain (no destinction between the uncertain future and certain past)

The problem with your past light cone approach is that it requires a preffered event. Which event decides what is real and what is not. Consider 3 events. 1. happens in 1950. 2. happens on 1960. 3. happens in 1970. 2 is real according to 3 and unreal according to 1. The alternative to a preffered event would be to declare reality to be relative.

Or conclude your boundary of past certainty is wrong if it is based on a purely conventional, unobservable construct. Note that this notion is a carryover from pre-relativity, where simultaneity is absolute, so it makes sense to consider this the boundary of certainty. Also note that there is no general way to carry simultaneity boundary of certainty to GR, since there is no generally preferred simultaneity possible.

 

[PeterDonis]

The problem with your past light cone approach is that it requires a preffered event.

Or it requires admitting that what is "fixed and certain" is relative; it is different for different events.

The alternative to a preffered event would be to declare reality to be relative.

"Reality" is not a precise term, and I don't want this thread to get sidetracked on philosophical issues like what is "real". Saying that what is "fixed and certain" is relative--that it is different for different events--as I did above, is more precise and avoids all the baggage around words like "real" and "reality". (It still does carry some baggage since the words "fixed and certain" are not completely precise either; but I think those words are easier to give a reasonably rigorous meaning in the context of a physical theory.)

 

[dl58]

Consider the following scenario. A star explodes. Bob stands closer to the star then Alice. Bob sees the explosion while it is still outside of Alice's light cone. The explosion is an observed reality for Bob. Is it unreal for Alice? Consider a third person Tom. The previous scenario is in Tom's past light cone. Tom sees the light hit Bob and then Alice. What was real for Bob turned out to be real for Alice. This can be repeated by Tom. What's real for Bob will also be real for Alice. Is there really a good reason to believe reality is relative?

 

[PeterDonis]

Consider the following scenario. A star explodes. Bob stands closer to the star then Alice. Bob sees the explosion while it is still outside of Alice's light cone. The explosion is an observed reality for Bob. Is it unreal for Alice?

There is no scientific way to answer this question, since there is no experiment you can run that will tell you whether the answer is yes or no. So the question is off topic for this discussion. Again, please do not get sidetracked on "reality" or "real" or other unscientific questions. Please focus on the specific argument I refute in the article.

 

[dl58]

The explosion may be uncertain to Alice but the example with Tom shows it is fixed. By fixed I mean Alice will not see something which contradicts what Bob saw such as the same star burning for another 100 years(assuming Bob and Alice are standing close and not moving at a high speed with respect to each other.). This experiment can be observed and repeated. Your refutation requires what is fixed to be relative not just unknown. The andromeda paradox describes a scenario where two people speak of a past event where one persons fixed past was another's uncertain future. They now both agree that one persons uncertain future was the others fixed past. The contradiction is avoided by eliminating the destinction between uncertain future (not fixed) and certain past (fixed)

 

[PeterDonis]

The explosion may be uncertain to Alice but the example with Tom shows it is fixed.

You are assuming your conclusion. You can't just state that it is "fixed" in an absolute sense. My contention is that "fixed" is relative; it depends on which event you pick. You can't refute that by just claiming that "fixed" is absolute. You have to show how SR requires "fixed" to be absolute, and you haven't done that.

By fixed I mean Alice will not see something which contradicts what Bob saw

That's because you constructed the scenario that way: you stipulated that the events of the explosion, Bob seeing it, and Alice seeing it are all in the past light cone of Tom at some event. So of course they're all "fixed", by my definition.

But now suppose this: Bob sees the star explode. Bob also sees an image of Alice at that same event (i.e., the light from that image of Alice reaches him at the same event on his worldline as the light from the exploding star). Bob predicts, based on those images, that Alice will see the exploding star at the event on her worldline that intersects his future light cone at the event where he sees the exploding star and the image of her.

However, Bob's prediction turns out to be wrong: what he does not know is that an alien spacecraft , coming in at high deceleration, took Alice aboard and flew off with her at high acceleration in the opposite direction, at an event on her worldline just outside Bob's past light cone at the event where he sees the exploding star and the image of her. The alien ship's acceleration is high enough, in fact, that both the past and the future light cone of Bob at the event where he sees the exploding star are behind the alien ship's Rindler horizon. That means that no light from the approaching alien ship had reached Bob at that event, and the light from the exploding star that is passing Bob at that event will never reach Alice, because it can't catch up with the alien ship.

You may object that you didn't include all this in your scenario. But in the real world, you don't get to choose the scenario. You picked a scenario in which nothing of interest happens except the exploding star; but in the real world, you don't get to pick what things of interest happen. The point is that both "futures"--both sets of events involving Alice, the one you gave where she sees the exploding star and the one I gave where she gets taken away by the alien ship and never sees it--are consistent with what is in Bob's past light cone at the event when he sees the light from the exploding star. And that will be true of any event. Even at the event where Tom has all of this in his past light cone, so he knows which of the "futures" that were consistent with Bob's past light cone actually came to pass, there are still an infinite number of "futures" that are consistent with what is in Tom's past light cone at that event, and there is no warrant for claiming that anyone of them is "fixed and certain".

 

[dl58]

A. Your scenario was explicitly forbidden in my scenario.
B. Alice still would not see anything that contradicts what Bob saw.
C. You have only pointed out that the explosion may be unknown to Alice in a alternative scenario.

If Bobs fixed past is not fixed for Alice then it would be possible for Alice to observe something which contradicts Bobs observation.

My scenario shows Alice will observe what Bob sees or she will remain outside of the light cone. She will not observe anything that contradicts what Bob saw.

There is no evidence showing that if Bob sees something then Alice will see that which contradicts Bobs observation.

It can be experimentally observed and repeated that what Alice sees will either agree with Bob or the light may never reach her. Your counter argument seems to be "anything's possible" That argument works on anything.

You said:
"There is no warrent for claiming that anyone of them is "fixed and certain"

This is not a preffered event. 100 years later somebody may observe that only one of the infinite possible futures have come to pass.

 

[PeterDonis]

Your scenario was explicitly forbidden in my scenario.

So what? As I said, in the real world you can't specify the scenario. And you certainly can't make absolute claims about what is "fixed and certain" based on a particular scenario that you specified. You need to make an argument that applies in any scenario that is consistent with the laws of physics.

Alice still would not see anything that contradicts what Bob saw.

This will be true in any scenario whatsoever. No two observers will ever see things that contradict each other; that would violate the laws of physics. But that says nothing about which of all of the physically possible things will happen, and it says nothing about what is or is not "fixed and certain".

If Bobs fixed past is not fixed for Alice then it would be possible for Alice to observe something which contradicts Bobs observation.

Non sequitur. As above, it is not possible at all for Alice to observe anything that contradicts what Bob saw; that would violate the laws of physics. But that fact is not a logical consequence of anything about what is or isn't "fixed" for a particular observer. It's a logical consequence of the fact that the laws of physics are self-consistent.

Your counter argument seems to be "anything's possible"

No, it is that, from the standpoint of any given event, anything that is consistent with the laws of physics and the events in the past light cone of that event is possible. So what is possible is relative: it depends on which event you pick.

100 years later somebody may observe that only one of the infinite possible futures have come to pass.

No, they will observe that, within their past light cone, only one of the infinite possible futures with respect to the past light cone of some previous event has come to pass. But there will still be an infinite number of possible futures that are consistent with the past light cone of the new event, and there will still be no warrant for claiming that anyone of them is fixed and certain.

This is not a preffered event.

Exactly; which means that my argument, that at a given event there are an infinite number of possible futures that are consistent with the past light cone of that event, applies to every event.

 

[Smattering]

It is a reductio ad absurdum.
1. Assume the future is uncertain.
2. Show the contradiction "according to one of them, the decision lay in the uncertain future, while to the other, it lay in the certain past."
3. Conclude the future is certain (no destinction between the uncertain future and certain past).

I never read the term "block universe" until very recently. So please allow me a naive question:

Is there actually any significant difference between believing in the block universe and believing in determinism? It sounds to me as if both was pretty much the same.

If the future is "certain", then it is fully determined, isn't it? And if it is fully determined, then it is certain. And of course, everyone who believes in determinism, implicitly believes in a static 4D universe where every event is fully determined.

So again: Where is the difference? Or is "block universe" only a modern word for the very old idea of determinism?

 

[Dale]

To me it is like having arguments for or against ice cream, or Justin Bieber. The way I see it, metaphysical theories such as block universe are just stories that humans tell themselves and each other in order to come to terms with the world and to imagine it in a satisfying way.

+1 on this.

Any time I see someone ask "what REALLY is ..." I immediately suspect that they are looking for a nice bedtime story.

 

[dl58]

"This will be true in any scenario whatsoever. No two observers will ever see things that contradict each other; that would violate the laws of physics. But that says nothing about which of all of the physically possible things will happen, and it says nothing about what is or is not "fixed and certain"."

The paradox states that if A observes an event then it is fixed for B."No, they will observe that, within their past light cone, only one of the infinite possible futures with respect to the past light cone of some previous event has come to pass. But there will still be an infinite number of possible futures that are consistent with the past light cone of the new event, and there will still be no warrant for claiming that anyone of them is fixed and certain."

There will be an infinite number of future light cones who disagree. Why would they be any less correct?

 

[eltodesukane]

I don't know what to make of things that are described as 'arguments for (or against) the block universe'.

In the particular case I describe in the article, the argument is that SR, specifically relativity of simultaneity, "requires" the block universe interpretation (in the sense that that interpretation is logically necessary given the postulates of SR). That is the argument I am refuting. More general or philosophical claims about whether the block universe is "real" are out of scope, at least for this particular discussion. I was only discussing a particular logical argument and why it is invalid.

I see this as entirely a matter of taste.

As far as "interpretations" in general go (not just of SR but any physical theory), I entirely agree. But again, I would like to keep this thread focused on the specific logical argument I refute in the article.

"I see this as entirely a matter of taste."
"As far as "interpretations" in general go (not just of SR but any physical theory), I entirely agree."
Interpretations are irrelevant if they have no consequences.
But interpretations with consequences are.
Newton interpreted gravity as an instantaneous action at distance, while Einstein interpreted it as local spacetime curvature propagating at the speed of light. Those two models, or interpretations, have different consequences, so we are able to choose the better one according to experimental results.

 

[PeterDonis]

Is there actually any significant difference between believing in the block universe and believing in determinism?

As a general thing, I would say no, there probably isn't. But please bear in mind that in this thread, we are focusing on a specific claim (which my article refutes), which is that SR, and specifically the relativity of simultaneity, "requires" the block universe interpretation, which for our purposes here is the claim that the entire 4-d spacetime is "fixed and certain". Whether or not this is equivalent to determinism is out of scope for this thread.

 

[PeterDonis]

The paradox states that if A observes an event then it is fixed for B.

Do you mean that if A observes an event on B's worldline, then it is fixed for B? Fixed at what event on B's worldline?

Once again, you keep on using the word "fixed" as it if has some absolute meaning--but that is precisely the point being disputed. So you can't just assume it; you have to show that SR requires it. Continuing to make arguments that assume it does not do that; it just argues in a circle.

Note that your proposed "reductio ad absurdum" argument makes the same assumption implicitly. Your first premise is "assume the future is uncertain". But if the "future" is different at different events, then "the future is uncertain", as it stands, is not well-defined, so your argument can't even be made. So before you can even set up your "reductio ad absurdum" argument, you need to first show that "the future is uncertain" makes sense without specifying a particular event. You haven't done that.

There will be an infinite number of future light cones who disagree.

"Disagree" with what? I've already said the laws of physics are self-consistent, so given the past light cone of one event, there can be no past light cones of any other events which "disagree" with that data; that would violate the laws of physics. So I don't understand what point you are trying to make.

 

[Smattering]

The paradox states that if A observes an event then it is fixed for B.

I think this wording is somehow misleading as it implies a *causal* relationship between A's current observation and B's future observation. But A's current observation is *not* the root cause for what B will observe. As long as we are not talking about quantum entanglement, but rather macroscopic observations (supernova, spacefleet leaving Adnromeda, etc. pp.), an observation does not fix anything that has not been fixed before.

 

[jtbell]

Newton interpreted gravity as an instantaneous action at distance, while Einstein interpreted it as local spacetime curvature propagating at the speed of light. Those two models, or interpretations, have different consequences, so we are able to choose the better one according to experimental results.

Correct. This means that they are not "interpretations" in the sense that we are discussing here. In the context of this thread, two interpretations of the same theory make the same predictions for anything that we can test experimentally, and are therefore indistinguishable experimentally. This is not the case for the Newtonian and Einsteinian theories of gravity.

 

[Smattering]

If Bob observes a super nova then it is fixed for Alice.

Can we please leave away this term "to fix something" and discuss this in strict terms of cause and effect?

Bob's observation does not cause Alice's observation, and Alice's observation is not an effect of Bob's observation.

 

[Torbjorn_L]

Oy. I saw comment boxes on Mark Stuckey's "blockworld" series and prepared a lengthy comment in response to his ideas. But those boxes were not functional, comments were turned off for the whole series. Since the comment is related to the topic of this article and indeed refer to it, I post the comment here instead. If it is too irrelevant, feel welcome to delete it.
****

I have some problems with [Mark Stuckey's] article series. No doubt the block universe works in some generic schemes. But until it can be shown that static models can be as rich as dynamic models (doubtful, see e.g. differential equations of dynamics vs equations of statics), it seems to me to be a problematic proposition.

And how would we test it? It is easy to observe and test time (clocks) as well as GR. To point to GR as a test of a "spatiotemporal global constraint" is all well and good, but the present [sic!] theory of time predicts more.

I am also wary of the prodigious use of philosophy. For the problems with claiming "real" in general, I refer to Peter Donis's excellent article. [ https://www.physicsforums.com/insights/block-universe-refuting-common-argument/ ]

For the specifics, first the discussion of 'real' in relativity of simultaneity. The latter says simply that you would expect only an agreement on physical laws. The existence of time is such a law.

Then as for for 'local realism' it is more precisely a tested hypothesis of no hidden variables. (Tested locally in Bell tests but holds globally over the light cone in gauge theories.) All that it says is that some quantum systems aren't local in the sense of correlations, and that the philosophic idea of 'real' - as a superfluous add on on nature - doesn't hold for quantum properties between observations. Or if you insist they do, you have to adopt theories like Many Worlds that makes correlations _really_ non-local (holds over many universes).

The discussion of cosmology and CTCs seems confused, with all respect for that the author has a PhD in general relativistic cosmology and I have not.

Shrinking the cosmological scale factor to zero makes no sense, and it wouldn't be "one point". (I don't think that is what the author wanted to say however.)

Inflation makes the era before the Hot Big Bang of indefinite duration, which accords with the problem of deriving a zero scale factor or a singularity out of semiclassical worldlines. [If you like the latest Planck data release, it may be that we are probing the inflation era so successfully that we can say inflation had a finite duration. The more or less model less fits of potentials sees them being steep.]

CTSs are generally invalid solutions to GR, are they not? No one claims that GR solutions always are physical, no more than they claim quantization of fields are automatically such. They have to be tested for relevancy. Specifically here CTSs implies time travel, which seems to be forbidden. I refer to CS Scott Aaronson's paper on this: if time travel was physical, the universe would have such systems already and/or we could make time travel computers. Then the complexity measures of CS falls, and all physics is simple. But we observe it is not, hence no time travel - or CTCs - exist.

 

[PeterDonis]

Bob's observation does not cause Alice's observation, and Alice's observation is not an effect of Bob's observation.

Maybe not in the ordinary language sense of "causality", but as the scenario was specified (the original version, not my alternate version with the alien ship), Bob's observation is in the past light cone of Alice's observation, so with respect to the causal structure of spacetime, Bob's observation is in the causal past of Alice's observation. This is true even if, in ordinary language terms, Bob's observation did not "cause" Alice's observation. The causal structure of spacetime is the important point for this discussion.

 

[nikkkom]

"I see this as entirely a matter of taste."
"As far as "interpretations" in general go (not just of SR but any physical theory), I entirely agree."
Interpretations are irrelevant if they have no consequences.
But interpretations with consequences are.

If two "interpretations" predict different results for the same experiment, they aren't interpretations anymore, they are bona fide physical theories.

BTW, this makes these discussions not entirely devoid of meaning - by looking at various possible scenarios, we may realize that we do have such an experiment.

 

[Smattering]

Maybe not in the ordinary language sense of "causality", but as the scenario was specified (the original version, not my alternate version with the alien ship), Bob's observation is in the past light cone of Alice's observation, so with respect to the causal structure of spacetime, Bob's observation is in the causal past of Alice's observation.

O.k., at least I get the point now.

So the claim is that my own past light cone contains third-party observations of events that I myself have not yet observed.

Interesting, but are there any concrete examples how far this can be pushed? As I understand, this is also your counter-Argument: It is unclear if this really can be pushed so far that an event early in the history of the universe will determine the future of the entire universe completely.

 

[PeterDonis]

So the claim is that my own past light cone contains third-party observations of events that I myself have not yet observed.

No, at least not in principle. Any event in your past light cone is an event you could have observed, since by definition light from that event could have reached you. If you happen not to have paid attention to those events, that doesn't matter for the causal structure of spacetime.

As I understand, this is also your counter-Argument

You've lost me here; what counter argument do you think I am making?

 

[Smattering]

Any event in your past light cone is an event you could have observed, since by definition light from that event could have reached you.

Yes, I am aware of this. But this refers to events that I myself could have observed.

If you happen not to have paid attention to those events, that doesn't matter for the causal structure of spacetime.

I thought the claim in this Andromeda paradox was that some other observer can inform me of an event before I could have observed it myself.

You've lost me here; what counter argument do you think I am making?

That these kind of "pre-notifications" by other observers do not necessarily "fix" (I hate this term) the complete future of the entire universe.

 

[SlowThinker]

It seems these arguments are just too complicated.

In my opinion, the alternative to Block universe is:

Events can only be observed in their past light cone.

and the author of the original series just does not have the imagination to understand it.
I haven't seen any other argument for BU in that series or this thread.

BU stays a possibility but not a necessity.

 

[PeterDonis]

I thought the claim in this Andromeda paradox was that some other observer can inform me of an event before I could have observed it myself.

Nope. Nobody has ever claimed that information can travel faster than light. That is not what the argument I'm refuting says.

It's true that, once you realize the argument I'm refuting doesn't say that information can travel faster than light, i.e., it agrees that nobody can inform you of any event before you could, in principle, have observed it yourself, the argument starts to look pretty silly. That was part of my point in the article.

 

[Smattering]

Nope. Nobody has ever claimed that information can travel faster than light. That is not what the argument I'm refuting says.

Then, I do not know what you were trying to express with the past light cone stuff.

 

[PeterDonis]

I do not know what you were trying to express with the past light cone stuff.

Do you mean in this thread, or in the article?

 

[meviccar]

Peter,

If, hypothetically, you could prove that the present (surface of simultaneity/3D world) could be directly observed, wouldn't that negate your claim that your alternative premise 'accounts for all observations'?

I just can't shake the feeling that you're argument basically hinges on the technicality that laws of physics prevent one from proving that the present exists. Sure, reality is unintuitive, but that feels like a stretch. And yes, I recognize the irony in my statement.

 

[meviccar]

Peter,

If, hypothetically, you could prove that the present (surface of simultaneity/3D world) could be directly observed, wouldn't that negate your claim that your alternative premise 'accounts for all observations'?

I just can't shake the feeling that you're argument basically hinges on the technicality that laws of physics prevent one from proving that the present exists. Sure, reality is unintuitive, but that feels like a stretch. And yes, I recognize the irony in my statement.

you're/your...

 

[SlowThinker]

I just can't shake the feeling that you're argument basically hinges on the technicality that laws of physics prevent one from proving that the present exists.

This "technicality" is the core of the dispute.
Either you accept it, and BU does not make sense, or you don't accept it, and BU is the only resolution.