I Are there any accepted relativistic interpretations of quantum mechanics?

[gentzen]

• Retrocausality/transactional

Because I wanted to better understand a sketch/drawing in chapter 4 of a "strange book" I was reading

(I currently read chapter 4, it doesn't seem to be as badly wrong as chapter 3, but already well on its way deep into esoterics.)

It is not pseudo-science, but it is not science or popular science either. It would say it is scientifically inspired speculation.

I started to dive a bit into the transactional interpretation (TI). I quickly learned enough to understand that sketch and finish chapter 4 (after which I decided to stop reading that book for the moment).

But the TI was still unclear to me, and I also wanted to understand it, because of its greater time-symmetry compared to most other interpretations. I had already learned that there was disagreement between Ruth Kastner and John G. Cramer about "some details" of TI, and that Tim Maudlin had "disproven" Cramer's original version of TI. Kastner's version seems closer to what I am looking for, but Cramer's version seemed less abstract and easier to visualize. I wondered whether Cramer's version could really work, so I read his 1988, 1986, 1983, and 1980 papers. I actually started with his 1986 paper, but it provided insufficient detail in crucial places, and talked about so many vaguely related stuff (not necessarily wrong, but somehow it was "all over the place"). In the end, I was seriously underwhelmed, and I better appreciated why Tim Maudlin decided to "disprove" TI.

And I also learned that I am not alone in being underwhelmed by Cramer's elaborations:
https://www.amazon.co.uk/gp/customer-reviews/R202O26FF2W72V

I've been fascinated by the Transactional Interpretation of QM since reading In Search Of Schrodinger's Cat many years ago - treating quantum processes as bidirectional interactions in time seems to make all that is mysterious in the quantum world comprehensible. [...]

Unfortunately, when the TI is laid out in detail it disappoints. Although Prof. Cramer seems to believe that the TI is nothing but the mathematical formalism of QM converted into words, a number of additional mechanisms are bolted on to explain how a transaction arises in response to offer and confirmation waves (the interpretation of the Schrodinger wave equation in TI). This part of the process seems to be very unclear, with the description being almost anthropomorphic. Apparently the source of an offer wave selects from the time-reversed confirmation waves it receives according to some sort of hierarchy whilst making sure that conservation laws are upheld.

This seems to require a larger set of rules and mechanisms bolted on top of the QM formalism than even Copenhagen or MW require - and if there is a proposed mechanism for how the selection process happens then its description here is so vague as to be incomprehensible.

So it seems that the Transactional Interpretation is not and cannot be the "correct" interpretation of Quantum Mechanics because at the critical moment it waves its hands and mumbles, leaving gaping questions open about what happens and how it happens during a transaction. Maybe Cramer has a vision in his mind, but he has not been able to set it down on paper here.

I started to read Ruth Kastner's stuff more seriously. This made me realize the "conceptual enormity" of the picture TI is forced to paint, if it wants to have any chance to reproduce the predictions of QM. So now I am completely convinced that Cramer's version never had any chance to be correct, and that Tim Maudlin's "disprove" points into the right direction why it had to fail.

 

[physika]

But the TI, because of its greater time-symmetry compared to most other interpretations.

And regarding the similarity with the Two-State Vector Formalism, what is your opinion?

​

 

[gentzen]

And regarding the similarity with the Two-State Vector Formalism, what is your opinion?

I hope that Kastner's version will be similar to it (that's what I meant by "Kastner's version seems closer to what I am looking for"). The Two-State Vector Formalism is fine for weak-measurement, but it is not obvious to me how to model normal "perturbing" measurements. The time-symmetric versions of CH might allow to model them, but probably in an ugly way that reduces the time-symmetry somewhat:

There is a wikipedia article for the two-state vector formalism. It says:

I got the impression that Reznik and Aharonov came up with a completely different approach, namely instead of two independent hermitian positive semidefinite matrices $\rho_i$ and $\rho_f$, the uncertainties in the initial and final state can be correlated:

But they were well aware of time-symmetric versions of CH:

 

[gentzen]

And regarding the similarity with the Two-State Vector Formalism, what is your opinion?

I hope that Kastner's version will be similar to it

and Cramer's version is very different from it:
https://arxiv.org/abs/1503.00039

The assertion that quantum wave functions cannot be considered to exist in normal space and must be viewed as existing only in an abstract higher-dimensional space, of course, creates a severe roadblock for any attempt to visualize quantum processes. (We note that Ruth Kastner’s “Possibilist Transactional Interpretation”[15, 16] adopts this point of view and treats quantum wave functions as being real objects only in an abstract multidimensional Hilbert space, from which transactions emerge in real space. The possibilist approach is not incorrect, but we consider it to be unnecessarily abstract.)
The “standard” Transactional Interpretation, with its insights into the mechanism behind wave function collapse through multi-vertex transaction formation, provides a new view of the situation that make the retreat to Hilbert space unnecessary. The offer wave for each particle can be considered as the wave function of a free particle, initially free of the constraints of conservation laws and indepent of the characteristics of other particles, and can be viewed as existing in normal three dimensional space. The connections between an ensemble of such free particles is only established when the multi-vertex transaction forms.

 

[WernerQH]

I started to dive a bit into the transactional interpretation (TI).

When I first encountered the transactional interpretation (~1986) I found it very attractive. Its main problem, as I see it now, is its ontology. What exactly is a "transaction"? And I think those "offer" and "confirmation" waves can't be real, neither in real (3d) space, nor in configuration space.

But essentially the same time-symmetric idea is already implicit in Schwinger's closed time-path formalism (1961). The "waves" are just the propagators connecting the vertices, and they "propagate" on the forward or backward time branch, respectively. For me they aren't real, but just elements of a mathematical apparatus that allows us to calculate the probabilities of patterns of events ("histories", if you like, but I don't think of them as continuous, but sequences of discrete points in spacetime).

 

[Fra]

There are also Dirac and Klein-Gordon equations for many-particle states. See e.g. my https://arxiv.org/abs/1205.1992 Sec. 8.3.3.

Not sure why i got back to this old thread, but regarding the OT and "relativistic interpretation of QM" I think indeeed the introduction of fermions (which are as we know hardware to understand classical, what "IS" a fermion really?), and it's relations to bosons (and thus the KG -> Dirac) relations is very central to the OT.

Relating to the qbism comment in post 6, one can reflect over werther the classification of boson vs fermion, is in the eye of the beholder? ie. the agent? After all the KG -> Dirac transformation is a sort of cahnge of variables that we can try to "interpret", does this transformation have a physical meaning that we can try to understand - or not?

/Fredrik

 

[PeterDonis]

werther the classification of boson vs fermion, is in the eye of the beholder? ie. the agent?

It isn't. Boson vs. fermion has experimental consequences.

 

[Fra]

It isn't. Boson vs. fermion has experimental consequences.

Yes agreed, and we can leave it at that, I guess we also get no further clues to the interpretational issues of relativity. I was trying to elaborate a bit as i think the KG/Dirac that is the standard example of introducting relativistic QM in a heuristic manner, with a tang that never left...

Relativity is after all about an equivalence class of observational frames; related by transformations, though referring only to spacetime. The problem is that way we understand this equivalence class in classical physics (SR), is problematic in QM. As QM is primarilty constructed in the observer bias, and this is more so emphasises more strongly in the qbist perspective. But the objectivity is restored if the observations in different views are related by transformations.

This is conceptually what i mean in the "eye of the beholder".

But part of the interpretational issues of QM so some at least are related to the unification to relativity. Many of the deep conceptual problems there are IMO clearly rooted in how we understand this "equivalence class" AND still insist of the observer-frame-bias that is built into the QM foundations.

The interesting thing is that many symmetries or maybe better to say dualities seem to come with the same conceptual problem. And always felt they are related... (meaning extendion spacetime transformations, to transformations that mix also internal degrees of freedom, which is a BTSM topic, but still part of interpretational issues imo)

So in a way I think the interpretation of QM as is, and the "interpretation of observer equivalence" and their relations are interwined, and still unresolved.

We can think of them as unrelated and get on with other things, but I feel that is letting go of a clue we need.

/Fredrik

 

[PeterDonis]

The problem is that way we understand this equivalence class in classical physics (SR), is problematic in QM.

I don't think it's problematic in quantum field theory, since QFT is explicitly modeled using spacetime. The question would be which QM interpretations take account of QFT and base their interpretation on it, instead of on non-relativistic QM.

 

[Fra]

I don't think it's problematic in quantum field theory, since QFT is explicitly modeled using spacetime.

To me it's problematic because this "spacetime" implies an external observer. It's not a problem for any particle physics experiment, but it is conceptually a problem for unification questions.

The question would be which QM interpretations take account of QFT and base their interpretation on it, instead of on non-relativistic QM.

I figured, but what i described would reverse of that logic. Ie. the qbist stands imo more naturally suggests that a qm interpretaiton of the observer invariants is the "wrong way" conceptually, the alternative is to take the observer-biased views more seriously, and allow them to interact, and then find the QFT and the effective "equivalence" class as emergent.

This does not change anything regarding experiments, nor claim that any of the effective theories are "wrong" in any way, it's only a different way of thinking. The main differences is in the view of fine tuning, nature of symmetries etc.

/Fredrik

 

[PeterDonis]

this "spacetime" implies an external observer

No, it doesn't. Classical relativity has no problem at all modeling the observer as being in spacetime, not outside it. The simplest such method is simply to model the observer as having a particular worldline in the spacetime.

conceptually, the alternative is to take the observer-biased views more seriously, and allow them to interact

Do you have any references that expound this kind of model?

 

[Fra]

No, it doesn't. Classical relativity has no problem at all modeling the observer as being in spacetime, not outside it. The simplest such method is simply to model the observer as having a particular worldline in the spacetime.

I meant, the whole construction of quantum mechanics, with state preparations and repeated statistics, and tomography IMO requires a "background" where the units of spacetime, including clocks are defined. IF you change this background context, then the whole quantum phenomenology as deformed (this of course is because we have no unification with Gravity and cosmology). I think this is not controversial, if there is a disagreement i supposed it is because I expressed myself wrong.

In classical relativity we don't have the same issue with non-commuting measurements, so it's not a problem in the same way.

Do you have any references that expound this kind of model?

There is to my knowledge not anywhere near something worthy to be called an explicit model to expound. I was just making a plausible conceptual extrapolation of a flavour of subjective qbist interpretation, not an explicit theory. That in the context of what is relevant for "interpretations of relativisti QM", from my perspective.

But lacking something to "expound", the the conceptual overeally model theoretic context of the is ...

The notion if agent/observer-centerd view contrasts to the system dynamics. But they are not in contradiction, they are just two ways of modelling, having pros and cons.
https://en.wikipedia.org/wiki/Agent-based_model

The "interacting agents" is in AI or computer science called "Multi-agent reinforcement learning"
https://en.wikipedia.org/wiki/Multi-agent_reinforcement_learning

The idea is that "self-organisation" and the "emergence" of effective objectivity may follows from this. some interesting modelling along these lines are

Modeling Others using Oneself in Multi-Agent Reinforcement Learning​

"Self Other-Modeling (SOM), in which an agent uses its own policy to predict the other agent's actions and update its belief of their hidden state in an online manner."
-- https://arxiv.org/abs/1802.09640

Applies to the "agents/obsevers" in physics, this essential would suggest that each part of the universe, encodes a "view" or "effective theory" of it's own environment (which is of course fellow agents). Also not that this seems conceptually close to a hologram.

Autonomous Agents Modelling Other Agents: A Comprehensive Survey and Open Problems​

"A core area of research in modern artificial intelligence (AI) is the development of autonomous agents that can interact eectively with other agents. An important aspect of such agents is the ability to reason about the behaviours, goals, and beliefs of the other agents. This reasoning takes place by constructing models of the other agents."
-- https://arxiv.org/abs/1709.08071

One missing big part is the abstraction of how cow would "inanimate matter" in it's fundamental forms, encode/model/execpt things about it's inaminate environment??

We don't know of course but, one can couple the above idea also to ideas similar to this paper

Law without law: from observer states to physics via algorithmic information theory​

"instead of a world or physical laws, it is the local state of the observer alone that determines those probabilities. Surprisingly, despite its solipsistic foundation, I show that the resulting theory recovers many features of our established physical worldview: it predicts that it appears to observers as if there was an external world that evolves according to simple, computable, probabilistic laws."
-- https://arxiv.org/abs/1712.01826

But as is clear this just dresses up and elaborates the possible promises I see in the "qbist" interpretation, which was the point of hte original post I wanted to comment on.

/Fredrik

 

[physika]

The idea is that "self-organisation" and the "emergence" of effective objectivity may follows from this. some interesting modelling along these lines are

I think otherwise.
(the inverse)

And:
"self-organisation" of what ?
what is "What"
Entities ?
Entities of ?

 

[PeterDonis]

how would "inanimate matter" in it's fundamental forms, encode/model/expect things about it's inaminate environment?

Your use of the phrase "inanimate matter" is a circular argument. The whole point is that if "agents" are made of matter like everything else, then at the very least the matter that "agents" are made of is not "inanimate".

 

[Fra]

"self-organisation" of what ?
what is "What"
Entities ?
Entities of ?

Your use of the phrase "inanimate matter" is a circular argument. The whole point is that if "agents" are made of matter like everything else, then at the very least the matter that "agents" are made of is not "inanimate".

It is true that there is an apparent "circularity" here, but it is indeed part of the open problem. Especially when you try to summarize this briefly, as to keep this on topic. There is a chicken and egg situation. What comes first, the observer or the observed etc. Do you explain one in terms of the other or vice versa? Is one view more fundamental or not?

Some short comments to try to explain without diverging..

I referred conceptually to self-organisation of the ambigous subjective views. The idea is the the self-organisation tames or cures the PROBLEM with the ambigous views. The resulting complex group-dynamics will by assumption work in parallell with the evolution among the views (agent population), so that there will be a steady state where the agent-population matches matter phenomenology, and the group-dynamics matches the matter interactions.

Needless to say this is an open problem, but the benefit if this works in the future is that instead of a finetuning of mathematical modells with parameterspaces (at the system dyanmics level) that has beeen made up or heuristically constructed without insigiht, we will have a evolutionary mechanism that from (agent model level) probably from a managable parameters space (with less fine tuning) that can provide explanation (in terms of the mentioned mechanisms) of the effective structures of system dynamics.

The term inanimate could be questioned yes, but what I meant was to distinguish between the persistent idea of many physicists that "theories/expectations/observation" requires some conscious or higher order context. But the challenge is indeed to understand how to explain the emergence of these apparently "animate" or "living" systems, if they are built from "simple" elementary building blocks, that themselves are not usually though of as "animate". But indeed, one could argue that even a single bit, flipping it's state in response to some enviromnet or data stream is "animate"? If so, you are right, all is animate. It's hard to choose the right terms.

/Fredrik

 

[PeterDonis]

There is a chicken and egg situation.

Only if you assume that "observer" and "observed" are fundamentally different kinds of things. Which is a circular argument. Similar remarks apply to terms like "conscious" vs. "unconscious" or "animate" vs. "inanimate".

Also, you are getting very far afield from QM and into philosophy, which is off topic here. This thread is about relativistic interpretations of QM, not about philosophical questions.