Home › Forums › 2015 International Workshop on Quantum Foundations › Retrocausal theories › General "Block Universe" Discussion
This topic contains 19 replies, has 6 voices, and was last updated by Anonymous 4 years, 11 months ago.
July 12, 2015 at 11:34 am #2659
I’ve seen the issue of whether or not to adopt a “block universe” perspective come up in at least a half-dozen different contexts here, so I thought it might be useful to start a new topic. All interested parties can now have a common place to discuss this crucial issue.July 12, 2015 at 3:56 pm #2664
Taking off my moderator’s hat, maybe I can immoderately kickstart this discussion with Section IV of my latest FQXi essay. In that section I defend the block universe perspective, while also noting that some of our intuitions might be incorrectly biased against it.July 12, 2015 at 6:26 pm #2666
Michael B. HeaneyParticipant
How does the block universe perspective differ from Feynman’s space-time approach to QM?July 12, 2015 at 9:30 pm #2671
I see physics as trying to construct a single model (called reality) to account coherently for disparate subjective experiences; not all experiences of course, just those that are (assumed) common, (approximately) repeatable and can be represented by laws (rules of regularity). What does special relativity (SR) have to tell us about this endeavor?
Consider the two boys and three girls in the example attached (you don’t have to read it, just look at Figures 1, 4, 5, & 7). The boys say they are twins, but must conclude the girls are different ages, while the girls say they are triplets and must conclude the boys are different ages. Likewise, the boys and girls disagree as to how far apart they are in space. If you constructed a model to include a meaningful Now per each of these M4 foliations, the two foliated sets of experiences would be incongruous. The lesson of SR is — if you want to include Now in a meaningful way, you have to give up on the goal of doing so with a single, coherent model (of reality). If you don’t want to give up on that goal, you must abandon a meaningful notion of Now to accommodate both sets of foliated experience (relativity of simultaneity, Block Universe). So, SR forces you to choose between coherence and completeness. Most physicists avoid this dilemma because they don’t believe Now is of concern to physics. I think it’s one basis for peoples’ objection to Block Universe. Those who recognize Now as germane to subjective experience and believe physics is in the business of modeling such common elements of experience are going to object to Block Universe as incomplete.July 13, 2015 at 7:02 pm #2692
I confess I don’t know what a block universe is supposed to be. I looked briefly at the attachments at #2659 (Ken) and #2664 (Mark), and did not find them very enlightening. So maybe I’m just a blockhead. Anyway, here is my vague idea of what is involved, which then motivates a couple of questions which may be equally vague. Comments would be appreciated.
My assumption is that a block universe (blockworld) is an attempt to model events in spacetime by assigning to every point in spacetime a something, a state of affairs, assuming that one has introduced a suitable 4-dimensional coordinate system to label those points. These events at different points are then related to each other via some physical law or laws, probably deterministic, at least for classical physics. Since there are various ways to choose coordinate systems one also wants to have some idea of the rules for transforming from one coordinate system to another, but the transformation rules are then assumed to be ‘local’. I.e., given what is going on at or near some point in spacetime I can carry out the transformation, which will tell me what is going on at the same (but relabeled) point in the new coordinate system according to some well-defined rules, such as how to transform energy and momentum when one goes to a different Lorentz frame.
This is reasonably easy to visualize in the deterministic world of classical physics. However, if we allow some sort of stochastic time development where there is a probabilistic association between events at different times, the picture is not so clear to me: do we have a single block world filled with what actually occurs, or do we just say that there are a large number of different possibilities (different block worlds) to which we will assign probabilities? And, which is probably harder, how am I supposed to think about entangled quantum states in a block universe? Or is this last difficulty a good reason for abandoning the idea of a quantum block universe?
Bob GriffithsJuly 13, 2015 at 9:52 pm #2697
First let me say that I’ve been using term “blockworld” (BW) for years and was only this year told by my philosopher of science colleague that it’s now “block universe.” Since I’m speaking with a fellow physicist, I’m going to revert to BW 🙂
I received my PhD in general relativity (GR) and have taught it many times, so I perhaps take for granted that many (most?) physicists have no formal training in GR. But, what you said is right on the money – GR assumes a spacetime manifold with a certain topological structure upon which one associates a metric and stress-energy tensor (SET). Einstein’s equations then provide a “consistency criterion” that the metric and SET must jointly satisfy. Accordingly, there are many such “self-consistent” combinations of metric and SET, while physics only uses a few. Special relativity (SR) then applies in the locally flat (M4) regions of the curved spacetime manifold of GR (so that’s where you can apply your Lorentz transformations).
SR is where one encounters the relativity of simultaneity (RoS), although it can also be introduced to curved GR spacetimes that allow for global foliations (not all do). [In the case of curved spacetime however, it may be that observers on a surface of simultaneity are moving with respect to each other, e.g., surfaces of homogeneity in big bang cosmology models. And, such observers occupy different M4 frames, so SR can’t be used between them.] It’s RoS that implies (but not entails) BW, as I explain in my handout attached above (SR-Example-Phy200.pdf). That example is written for intro physics students, so if you read it carefully, I’m sure you’ll understand how RoS implies BW. If you want a layperson’s intro to BW, just watch the 11-min segment from 17:55 to 28:55 of https://www.youtube.com/watch?v=NcOBtnU-zSA
There are at least three general ways people accommodate the stochastic nature of QM in a BW, two are retro-time-evolved stories told from the perspective of an observer inside the BW (“dynamical” approach, Price’s “perspectival view”) and the other is an adynamical approach per a view “outside” of 4D spacetime using a “global constraint” of some sort. Examples of the two dynamical approaches are Cramer’s Transactional Interpretation http://arxiv.org/pdf/1503.00039 that invokes “pseudo-time” processes (processes in a time that isn’t housed in spacetime, whatever that means) and Kastner’s Possibilist TI http://www.ijqf.org/forums/topic/possibilist-transactional-interpretation which modifies TI so that the processes refer to proper time along observers’ worldlines, i.e., a time that actually resides in spacetime. RBW with its adynamical global constraint is an example of the adynamical approach, as is Price’s Helsinki (toy) model. I’m not sure how to classify the Two State Vector Formalism, they talk as if they’re using a “pseudo-time” a la Cramer, but Aharonov has proposed a different notion of time that he’s hoping is more along the lines of Becoming rather than BW http://arxiv.org/pdf/1305.1615v1.pdf.
Let me know if you require further clarification!
MarkJuly 14, 2015 at 1:15 am #2711
I agree with Mark that you have the right idea for a Block Universe, but Mark’s last paragraph doesn’t really address the key questions in that last paragraph of yours. (Mark: I would say that the whole point of the Block Universe is that everything ontological can be described “all-at-once”, so mentioning all of these different “processes” is obscuring the point, I would think.)
So to answer Bob’s final questions: a Block Universe could certainly be compatible with a stochastic theory. In that case “the Block” is a single world filled with what actually occurs. (An agent might not know which block is the real one, and might assign probabilities to various imagined blocks, but only one Block happens.) The big question is whether this “one 4D universe” is compatible with quantum theory, essentially the last questions that you lead up to.
Now, almost everyone will take a look at entangled states on one hand, and the Block Universe framework of SR/GR on the other, and decide that the “existence” of entangled states rules out any possibility of a Block Universe. But this section of this forum happens to be filled with the 20-odd people for whom retrocausality is a live option, which sheds some doubt on this usual conclusion. In principle, retrocausality provides a way to rescue the Block Universe (entangled states would have to be epistemic, not ontic, for starters, but this isn’t enough; retrocausality is needed to resolve the no-go theorems). Here’s a recent talk I gave about precisely this issue:
But… surprisingly to me, many of the people who are interested in quantum retrocausality do not like the idea of a Block Universe, and do *not* want to rescue this framework of SR/GR. Certainly Ruth Kastner would be in this camp, and perhaps also Eliahu Cohen and Avshalom Elitzur, based on their “meta-time” comments in this forum. I’m not sure about John Cramer (John, if you’re reading this, I’d like to hear where you stand on this…). I’m also not sure about Yakir Aharonov and other people involved with retrocausal interpretations of the two-state-vector formalism. The multi-particle states in that formalism continue to reside in configuration space, not spacetime, and I haven’t seen much discussion of whether they even want to resolve this issue.
Perhaps it would be good to hear from some retrocausal-minded people (Ruth?) who don’t particularly want to recover a Block Universe, and what they are hoping to recover instead.July 14, 2015 at 2:49 am #2716
I was trying to be exhaustive concerning the view of QM in a BW. To do that I have to include TI with its “pseudo-time” processes. You agree TI is in a BW, right? And having mentioned TI with its “pseudo-time” processes, I had to mention PTI with its proper time evolving BW (of sorts). This is precisely the point of PTI’s departure from TI.
To follow up on Ken’s post, rather than think about multiple BW’s each corresponding to a different experimental outcome, I just imagine subsets of one BW, each subset representing a different trial (with outcome) in an experiment that’s been repeated many times. Thus, the probability of QM is just giving a spatiotemporal frequency of occurrence for these subsets per normal physics.
Hope I’m not confusing the issue.July 14, 2015 at 11:18 am #2725
Mark: No, I’ve never been able to make sense of the Transactional Interpretation in a block-universe context… I know John Cramer is aiming to reduce entangled states to coupled-single-particle states, which would certainly help, but I haven’t seen a mathematical framework for how he hopes to do this. Even in the single-particle case, I’m not positive that the all-at-once account holds together properly, but it’s hard to tell because most of the discussion typically uses this “pseudo-time” account you mention.
In general, as I see it, any approach that *requires* some “pseudotime”, “meta-time”, etc., to make sense of it is not in a Block Universe. (I call that the “two time trap”, in that essay I linked to up above.)July 14, 2015 at 2:15 pm #2728
On p 8 of http://arxiv.org/pdf/1503.00039.pdf Cramer writes, “The transaction that forms after the emitter-absorber offer-confirmation exchange process goes to completion is the real object, what we would call the ‘particle’ that has been transferred from emitter to absorber.” Now look at Figures 3 and 5, and you’ll see that he’s using a BW.
The way I see it, he’s just like us in that he assumes a fundamental ontological entity that spans space and time then seeks an explanation for its distribution in the BW. In your case, that fundamental ontological entity is the classical field and in RBW it’s the spacetimesource element. In your case, the explanatory mechanism is L = 0, for RBW it’s the adynamical global constraint, and for TI it’s the “pseudo-time” process. [In the classification scheme I used above, I’d say you’re in the “global constraint” camp.] So, we have different fundamental ontological entities and different explanatory mechanisms, but we’re all playing in the BW.July 18, 2015 at 7:58 pm #2880
You are going to have to sell me on the block universe (again) next weekend when we tape the inaugural episode of Quantum Conversations. I can buy retrocausality on the micro level so I do not think the two ideas are necessarily connected. I think the block universe is merely a way to make retrocausality seem palatable on the macro level, or, rather, to make better sense of the quantum-classical contrast. But it assumes that space and time have some ontological status beyond mere relations.
IanJuly 19, 2015 at 1:42 am #2889
Hi Ian! Yes, I’m looking forward to talking about this in person…
I wouldn’t say that the Block Universe is particularly motivated by retrocausality — it’s best motivated by general relativity! (So yes, it does imply some ontological status of spacetime, modulo diffeomorphisms, etc.)
And curiously, many people interested in retrocausality don’t particularly like the Block Universe, a fact that I’m still struggling to wrap my head around.
I think I’ve heard you agree with Eddington’s perspective on this: that there must be something we’re missing in our physics of spacetime, because the Block Universe is somehow at fundamental odds with our sensation of “becoming”, or the “flow of time”. (Or, if not at odds with it, at least missing some crucial element that would permit such concepts to be discussed in a physical framework.) Is this accurate?
None of the “becoming” crowd has chimed in here, so if you’re willing to be a spokesperson for that perspective, I’d be curious as to what sort of things you think are missing from the Block Universe perspective. (And what you think of my anti-“flow of time” rant in section IV of my latest FQXi essay, linked to in the second post, up above.)
See you soon! -KenJuly 19, 2015 at 2:14 am #2892
Funny you should mention Eddington. I have gone back and re-examined some of his stuff and have come to the conclusion that he was actually an operationalist at heart. I don’t think he bought the ontological status of spacetime. In fact it is quite clear he didn’t. I just never thought about it in depth until last week’s RQI-N meeting.
Regarding the block universe, I am still at a loss as to how randomness can be accounted for in such a description. So more than the sense of “becoming” it seems to fly in the face of the way things actually work.July 19, 2015 at 11:43 am #2897
Ian: if you are just referring to the “Block” itself, see my #2711 above.
If you’re talking about classical randomness, or stochastic theories, another way to think about it is this: Most people’s trouble with the Block Universe has to do with the *future*. Very few people have trouble imagining a Block-year-1999. (Do you?) So any randomness that you associate with that year has a clear Block Universe representation. If you have trouble thinking the same way about the year 2999, you’re not alone, but I’d argue this is strongly linked with the “becoming” issue. (Why else would the future be thought different from the past?)
If you’re talking about *quantum* randomness, the situation is indeed trickier, but I also addressed this somewhat in #2711. My retrocausal models, including the ones in this forum, reproduce the observable effects of quantum randomness while being framed in a Block Universe. Still, this is quite a thorny subject — something we should definitely talk about!July 19, 2015 at 2:48 pm #2900
Like Ken, I don’t understand why people use (tacitly or explicity) a Block Universe (BW) for retrocausality then add a “pseudo-time” or meta-time to artificially create a dynamical notion of “causation.” If you want a robust Now/Becoming, it will cost you lots more formal machinery than a bare BW. But, if you’re willing to pay the price, you can have it, e.g., PTI.
Ian, I don’t understand why you think randomness is difficult to account for in a BW. I suspect you see something I don’t. I picture identical subsets of the BW corresponding to the repeated trials of an experiment. The goal is then to account for the distribution of outcomes shown in those subsets. The BW distribution function is found using probability amplitudes computed with the path integral (a BW computation). Can you explain what my understanding lacks that makes the process mysterious/problematic for you?July 20, 2015 at 8:31 pm #2926
Mark: well, I see two options here. Either randomness is merely a representation of our lack of knowledge (after all, the probability distributions you describe are only meaningful in the context of gaining knowledge, i.e. experiment), which would seem to suggest hidden variables are required to invoke a BW, or the multiverse exists in order to account for all possible outcomes simultaneously. The BW assumes that everything simply exists, i.e. time doesn’t flow, there’s no sense of “becoming” as Ken puts it. The only way I can see to reconcile that with quantum randomness is either hidden variables or the multiverse. Neither is particularly appealing to me.July 20, 2015 at 9:09 pm #2927
Ian; the multiverse isn’t a BW, which points to hidden variables. But one doesn’t need the sort of HV’s that bother you: all you need is that the *future* is a hidden variable, hidden from our knowledge, and you can have a BW theory that is fundmanentally stochastic. Surely you don’t have a problem with treating future events (events that lie the future of a given agent) as variables that are hidden from that agent, in an stochastic theory…?July 21, 2015 at 11:33 am #2934
What’s wrong with HV’s? What do you find objectionable about the spacetimesource element and adynamical global constraint of RBW, for example? Or Ken’s classical fields and L = 0 constraint?
Just curious 🙂July 25, 2015 at 1:23 am #2960
Well, I guess the fact that evidence continues to point to the impossibility of HVs. Every time someone thinks they find a loophole that allows for HVs, it is usually fairly rapidly closed.
The multiverse *could* be a BW (or a BU, I guess). There’s no reason it couldn’t be. In any event, I still see the “future events as HVs” scenario as being just a fancy way of saying everything is simply pre-determined.
Either way, let’s look at it this way. Suppose you are right and future events are simply hidden variables (or “governed by HVs” might be a more proper way of putting it). Under that scenario the apparent randomness of the universe corresponds to a lack of knowledge on our part. But if they are truly HVs then they are impossible to fully know. Hence the theory is untestable, i.e. there’s no way to prove such a theory is the correct theory of future events. If they’re not necessarily truly hidden, then where are they and how does a BW theory predict them experimentally?
So, in short, either give me something physically testable or else it’s just philosophy in my book. I’m sure that makes me look like an operationalist but I’m really not because I do believe that Occam’s razor can be employed to root out the simplest but broadest theoretical explanations whereas true operationalists don’t care.September 6, 2015 at 1:04 pm #3044
I have recently come up with a construction of a BW which is perfectly consistent with the statistical predictions of QM on the one hand, and with macroscopic classical mechanics and an arrow of time on the other hand.
Key features of the BW, overlooked in the above discussion, include:
– The BW must not admit a formulation in terms of as a Cauchy initial value problem, or else Bell’s inequalities would be respected.
– Quantum randomness arises naturally if one realizes that experiments are never repeated. Each repetition corresponds to a different slice of the BW, supported on different times, and QM merely describes the statistical properties of ensembles of such slices.
– The BW must include both a macroscopic radiation arrow of time and a local energy-momentum conservation law for it to be compatible with classical physics.
You must be logged in to reply to this topic.