Questions re Cramer's Transactional Interpretation?

[cesiumfrog]

(Continuing above post)

I'd like to emphasize the importance of reading (at the very least) Cramer's 1986 before forming an opinion or attitude towards TI. If you have a specific question about, or criticism of, something in that or any other paper on TI, I'll try to address it.

Cramer's 1986 paper does address Schroedinger's cat, but his discussion is restricted to (what MWI might term) only the parallel cats that would result in differing detection outcomes, ignoring the issue of cats that can no longer be distinguished at the detection event.

 

[rkastner]

I understand that there are superpositions of the O&C waves. I'm asking about the trajectory of the particle itself, between emission and detection.

If TI does say that a particle (like a photon, complex atom or buckyball) ultimately was actually super-positioned as it passed the slits, then does TI not also say that an entire lab can be isolated and prepared such that it (and its occupants, particularly any cats) will develop in a superposition

No. You seem to have forgotten about absorbers here, and that's the advantage of TI. The radioactive atom sends out a weak OW which is absorbed by the geiger counter detector which generates CW in response. There may or may not be a transaction at this level, so there is no need for endless superpositions of cats, Wigner, Wigner's friend, lab, etc. The 'buck stops' between the emitter and the absorber: a transaction may or may not occur within a given time frame, so the cat may be either alive or dead after that time based on whether or not it has occurred. Superpositions of OW are not maintained beyond the level of their absorption (unless they get identically re-generated, e.g. as at a mirror).

Re why QFT is not a full and complete theory, see, e.g.,
http://plato.stanford.edu/entries/quantum-field-theory/

 

[rkastner]

(Continuing above post)

Cramer's 1986 paper does address Schroedinger's cat, but his discussion is restricted to (what MWI might term) only the parallel cats that would result in differing detection outcomes, ignoring the issue of cats that can no longer be distinguished at the detection event.

Under TI there would be no cats that can't be distinguished. After a given period of time there is either a live cat or a dead cat, based on whether or not there was a transaction in that time. There are no superposed cats in TI.

 

[rkastner]

For those of you who adhere to the MWI picture, TI is a bit of a paradigm switch, because absorbers enter on an equal footing (almost) with emitters, and play a crucial part in the dynamics. This is how TI can account for the determinacy of cats and of the Born Rule for probabilities of outcomes, the latter remaining a deep conundrum in the MWI picture, where advocates are currently attempting to explain it in terms of 'rational observers' and social theory,which seems to me to be getting a bit far afield from what ought to be a physical theory.

 

[Dmitry67]

If you mean to suggest that emitters and absorbers are arbitrary or primitive in TI, this is not the case. The reflecting action of a mirror under TI can be modeled just the same as in standard qm , in which the quantum state (OW under TI) undergoes reflection due to an infinite potential barrier. Alternatively, if you want to take into account the specific interaction with the metal, the mirror is an absorber that generates a new OW identical with the one received by it. Thus there is a fully consistent account of emitters and absorbers as quantum objects, and there is no 'magic' involved. The photoelectric effect is not the same phenomenon as perfect reflection from a mirror even in standard qm, (the metal ejects an electron, not an incoming photon beam), so it wouldn't be appropriate to expect that they should be treated the same by a qm interpretation. But thank you for this example, I may want to include it in my presentation in order to help make the distinction.

The question neither TI nor CI can answer is "what is an absorber" (measurement device - CI). What *some* configuartions of atoms are NOT absorbers (measurement devices) while some other configurations are? Can we look at some configuration of atoms (without knowing what is the purpose of that thing) and calculate the value of the function IsAbsorber(parameter system)?

Of course, neither CI nor TI provide such answer, because there is no one.

The photoelectric effect is not the same phenomenon as perfect reflection from a mirror even in standard qm, (the metal ejects an electron, not an incoming photon beam)

On the high level, it is the same. The evolution from some state into another state, not taking care of what exactly is emitted (photon, electro, or something else).

So again, I provide you a definition of some tiny systems with 1000000 atoms. You don't know in advance if it is mirror or say, photo-sensitive cell in my photo camera. Can you provide a formal procedure to determine if such system is a measurement device (or, in TI, will it send a back wave, or how is it called...)

 

[Dmitry67]

Note that there is an some equivalence between CI and TI: if you can tell absorber from non-absorber, you can tell measurement device from what is not the one, and vice versa.

 

[Demystifier]

An emitter is just another dynamical object, and as such is also described by quantum theory. E.g., you can have an offer wave as an emitter: an OW for an electron can emit an OW for a photon, and another electron OW can act as an absorber for that photon. So it would be incorrect to think that emitters and absorbers are undefined or primitive, as is the case with an "observer".

But the processes of absorption and emition cannot be described by the Schrodinger equation (or a couple of Schrodinger equations). Right?
So either
1) TI replaces Schrodinger equation(s) by a different (set of) equation(s), or
2) TI does not provide a mathematical description of absorption and emition

Now please tell me which is correct: 1) or 2)?

 

[cesiumfrog]

Under TI there would be no cats that can't be distinguished. After a given period of time there is either a live cat or a dead cat, based on whether or not there was a transaction in that time. There are no superposed cats in TI.

I think this confusion is coming from Cramer's contention that MWI http://www.npl.washington.edu/ti/TI_app.html#A.4" time-symmetric.

After the transaction (between the OW from the apparatus that prepares the cat-box and the CW from in the apparatus opening the box) is formed one might suppose that the experiences of the cat (over the duration of time that this completed-transaction spans) were now determined uniquely by which memories the cat has afterward (and similar identifiers). (Or in MWI one might suppose that the observation result singles out just one cat, for the entirety of the time-period that the box was closed, from a clutter of parallel cats.) But obviously not every difference possible in the cat's experience would result in such distinguishability afterward. MWI's parallel worlds can merge: the process is called interference. To claim otherwise would be to claim that some fundamental difference exists between consciousness (of the cat) and any mechanistic micro-abacus (or C60, to pick something with thousands of parts, where superpositions of the whole have been confirmed). To be explicit, TI implies that the cat will have always been definitely alive (or dead and not both) but that it could nonetheless have been in a superposition of having forgotten dreaming about mice and having forgotten dreaming about fish.

You seem to be claiming to escape super-positioned cats by not treating the cat as one whole, and by instead considering many shorter transactions between each microscopic part separately. Isn't that inconsistent since the same method wouldn't be applied to C60 experiments?

 

[rkastner]

Note that there is an some equivalence between CI and TI: if you can tell absorber from non-absorber, you can tell measurement device from what is not the one, and vice versa.

The 'equivalence' you suggest is illusory, since CI does not admit that absorption has any role in collapse. That's why you get an endless proliferation of branches in CI (I assume you mean here orthodox interpretation), accompanied by hand-waving accounts of "when a measurement is complete" which have no clear answer. In contrast, In TI an absorber is characterized unambiguously in terms of interacting particles. E.g., any time you have states in > different states out, there has been absorption of the incoming particles and emission of new particles. So this is an objective matter describable in terms of particle states and not handwaving as in the usual 'when is a measurement complete'. In cases where one doesn't bother to go into these details, as in a mirror, it is because the outgoing states are essentially indistinguishable from the incoming states or otherwise trivial. Note that for such interactions there is always a probability amplitude which must be squared. Under TI, the squaring is done because a CW is generated during the absorption/interaction process. Under all other version of qm, you just square the amplitude because 'Born said so'. TI physically characterizes the absorption process in terms of the generation of a CW along with any interaction which transforms incoming particle states into outgoing particle states. This is implicit in Cramer's 1986, Fig. 3, which shows an interaction consisting of an excited particle emitting a photon OW, which is received by an unexcited particle, upon which the second particle generates a CW, resulting in a transaction which leaves it with increased energy. Physics is full of examples like this with amplitudes that must be squared. The point is that any time you square the amplitude it is because there has been a CW generated and a transaction with that probability may occur.

 

[JesseM]

rkastner, do you think the criticisms of Cramer's interpretation made by Tim Maudlin on pages 197-200 of Quantum Non-Locality and Relativity, which can be viewed on google books here, are unfair or misguided? In particular note the section on page 200 dealing with the ambiguity in how the interpretation is supposed to work when the position of the absorber in the future depends on which events occur in the past...

 

[rkastner]

But the processes of absorption and emition cannot be described by the Schrodinger equation (or a couple of Schrodinger equations). Right?
So either
1) TI replaces Schrodinger equation(s) by a different (set of) equation(s), or
2) TI does not provide a mathematical description of absorption and emition

Now please tell me which is correct: 1) or 2)?

Why? You have a false choice here, since it implies that all QM is is the Schrodinger Eqn. Non-relativistic QM is the Sch Eq plus the Born Rule which allows us to make empirical contact. Standard QM has no physical referent for the Born Rule; TI does.

Furthermore, as I've noted elsewhere, the process of emission/absorption would need to be treated in detail in terms of qm for interactions, which includes the relativistic domain, while the Sch eqn describes persistent particles in the nonrelativistic domain, so again your question is misleading, presuming that NRQM must be the whole story, which it isn't. (This is made evident in Cramer's 1986; have you read it?) TI is relativistically consistent, and applies anytime we need to square amplitudes to get probabilities. TI is perfectly capable of doing (2) since it is a time-symmetric account of standard qm which treats such cases; see my other replies. It simply interprets Born Rule squaring process as resulting from CW and OW. Anytime you take an amplitude and square it in standard qm, you are doing it because "Born said so"; in TI you do it because there are CW as well as OW.

In doing philosophical inquiry it helps to have an open mind.

 

[rkastner]

rkastner, do you think the criticisms of Cramer's interpretation made by Tim Maudlin on pages 197-200 of Quantum Non-Locality and Relativity, which can be viewed on google books here, are unfair or misguided? In particular note the section on page 200 dealing with the ambiguity in how the interpretation is supposed to work when the position of the absorber in the future depends on which events occur in the past...

Thanks, good question. Three different authors have addressed this argument by Maudlin. He raised an interesting challenge but I think seriously overstated his case. My arxiv paper

http://arxiv.org/ftp/arxiv/papers/1001/1001.2867.pdf

mentions this and provides reference to the 3 replies to this challenge, which is absolutely NOT fatal for TI.

 

[rkastner]

Dear Thread Participants:

In view of the continuing unfriendly tone and lack of preparation of some posters (by no means all!), I will no longer be regularly checking this board as of 2/24/10. If you have read some of the published material on TI and have an open-minded and sincere question, you could send me a private message. I cannot promise to reply but I will do so if possible.

Thanks for the many interesting questions and comments!

R. E. Kastner

 

[rkastner]

Concerning this question: " You seem to be claiming to escape super-positioned cats by not treating the cat as one whole, and by instead considering many shorter transactions between each microscopic part separately. Isn't that inconsistent since the same method wouldn't be applied to C60 experiments?"

I thought I posted an answer but don't see it.
So here is my answer again:

I would put it differently: in MWI one is failing to escape superposed cats by failing to notice that the cat (actually the geiger counter) is a collection of absorbers, anyone of which can give rise to a transaction. In physics there are just collections of interacting particles (quanta). 'Cat-particles' don't care that they happen to be part of a cat; they are capable of emission/absorption and as such can trigger collapses. Of course, before that, the geiger counter detector particle(s) are absorbing the atomic decay product OW and generating a CW, so the collapse takes place on a level before any 'entanglement' with a macroscopic object like a cat. Offer waves and confirmation waves certainly can be superposed; but their sums provide for transactions (collapses) that result in determinate situations, like either an alive or dead cat.

As for superpositions of mesoscopic objects (I don't know about the specific object here but I assume that's what it is); this is all perfectly consistent. Detecting such a superposition would be analogous to detecting a photon landing in an interference pattern in a two-slit experiment (as if it went through both slits). Similarly, with a more complicated object you could get a transaction reflecting some exotic superposed state. TI allows for that. It's just that, the more complicated the object, the less likely such superposed states are; objectively, their squared amplitude gets smaller. This is where decoherence is relevant.

 

[rkastner]

The question neither TI nor CI can answer is "what is an absorber" ...

Of course, neither CI nor TI provide such answer, because there is no one.

Of course it can. A quantum system in its ground state, or in any energy state capable of being excited, is an absorber. Why the alleged big mystery here about absorbers? It's standard physics.

It's trying to decide when a measurement is complete without taking into account absorption that is the problem.

 

[Demystifier]

Why? You have a false choice here, since it implies that all QM is is the Schrodinger Eqn. Non-relativistic QM is the Sch Eq plus the Born Rule which allows us to make empirical contact. Standard QM has no physical referent for the Born Rule; TI does.

None of this answers my questions.

Furthermore, as I've noted elsewhere, the process of emission/absorption would need to be treated in detail in terms of qm for interactions, which includes the relativistic domain, while the Sch eqn describes persistent particles in the nonrelativistic domain, so again your question is misleading, presuming that NRQM must be the whole story, which it isn't.

This has a potential to be the answer, but it is not. So let me rephrase my questions: In my questions, replace the word "Schrodinger" with "Dirac" or "Klein-Gordon" or something of that sort. What would be the answers in that case?

(This is made evident in Cramer's 1986; have you read it?)

No I didn't, that's why I ask you.

In doing philosophical inquiry it helps to have an open mind.

I certainly agree with that. And you should know that I am not a fan of the Copenhagen interpretation. If that helps, I am a fan of the Bohmian interpretation. But I am open minded and before philosophy I want to understand what new TI says on the mathematical level. So let me rephrase my question again: Does TI contains any EQUATION that is not contained in standard QM?

 

[Dmitry67]

...

To get rid of ambiguity, I will try to use only Yes/No questions.

1. A chunk of carbon (monocrystal containing 1'000'000'000'000'000'000 atoms) is an absorber for light in terms of Ti or not?
2. Monocrystal containing 1000'000 atoms is an absorber or not?
3. C60 molecule is an absorber or not?
4. Single C atom is an absorber or not?
5. single electron, u and d quarks are absorbers or not?

To symplify, all monocrystals are cooled to near zero temperature so they have regular structure.
Thank you in advance.

 

[Demystifier]

Good question, Dmitry! I also want to see how will he deal with it.

 

[cesiumfrog]

Ruth, would you clarify, do you agree or disagree that TI allows Schroedinger's cat to have actually been in a superposition of "having forgotten dreaming about mice" and "having forgotten dreaming about fish"? (Presuming we observe it to have survived, I'm sure we both agree TI also says the cat was not ever in a superposition with a dead version of itself, but that is unconnected to my query.)

It seems to me that a slippery slope argument can be demonstrated using double slit interference patterns: I think you agree that TI says a photon really is in a superposition as it travels between the source and detector? So I presume TI also says that a single atom can have been in a superposition (in the same kind of interference experiment). And the same will go for progressively bigger objects, composed even of far more fundamental parts than an atom is. So then the same even for cats (or people), fired by cannon through a partition with two gaps, provided that the room can be made dark and cold enough to avoid decoherence issues etc.

To me, it seems that accepting this conclusion should be anathema to any proponent of TI over MWI, but I also don't see how it can legitimately be avoided. I would elaborate, but after reading your posts I still don't yet understand exactly what your position is regards the above.

 

[rkastner]

Re recent questions/intended challenges: please read the relevant literature before asking questions. No need for me to repeat what is already available out there in print. My purpose was to address questions arising from those papers. But as I said, I cannot promise to check this regularly due to other pressing commitments, and won't feel inspired to do so if there continue to be questions from people who refuse to read the basic literature before presenting alleged challenges. Thanks! :)

And of course, I have already answered in very specific terms what counts as an absorber in TI--the same things that count as absorbers in standard physics.

 

[rkastner]

Ruth, would you clarify, do you agree or disagree that TI allows Schroedinger's cat to have actually been in a superposition of "having forgotten dreaming about mice" and "having forgotten dreaming about fish"? (Presuming we observe it to have survived, I'm sure we both agree TI also says the cat was not ever in a superposition with a dead version of itself, but that is unconnected to my query.)

It seems to me that a slippery slope argument can be demonstrated using double slit interference patterns: I think you agree that TI says a photon really is in a superposition as it travels between the source and detector? So I presume TI also says that a single atom can have been in a superposition (in the same kind of interference experiment). And the same will go for progressively bigger objects, composed even of far more fundamental parts than an atom is. So then the same even for cats (or people), fired by cannon through a partition with two gaps, provided that the room can be made dark and cold enough to avoid decoherence issues etc.

To me, it seems that accepting this conclusion should be anathema to any proponent of TI over MWI, but I also don't see how it can legitimately be avoided. I would elaborate, but after reading your posts I still don't yet understand exactly what your position is regards the above.

Look at decoherence. The amplitude for a coherent superposition of a macroscopic object like a cat and its alleged memories (I suppose you mean based on brain states) is vanishingly small, and so would be the probability of the relevant transaction. The problem with a decoherence-only account of this is that it doesn't explain why there IS a determinate outcome (see, e.g., Jeff Bub's 1997 book, Interpreting the Quantum World). Decoherence can explain why the probability of a superposed cat is extremely small, but still cannot explain why there IS one outcome or another--i.e., the mixture of possible observable outcomes in an 'improper' one which can't be interpreted as epistemic uncertainty. In TI, the same mathematics obtains--an extremely small prob. of a superposed cat--and you have an account of determinate outcome, because collapse is triggered by transactions resulting from the availability of CW, which are missing in the standard account.

 

[Dmitry67]

Good question, Dmitry! I also want to see how will he deal with it.

:)
That was easy :)

 

[Demystifier]

And of course, I have already answered in very specific terms what counts as an absorber in TI--the same things that count as absorbers in standard physics.

That's enough for me, I don't need to ask any further questions. So far I was not sure about that, but now I am: TI makes no sense to me. :zzz:

 

[Dmitry67]

Well, even rkastner obviously does not understand the severity of the issue with 'absorbers', which is the central question, the TI itself is not bad. It least it is nto worse then CI. Just take CI, and buy locality for the price of retro-causality.

 

[Demystifier]

Just take CI, and buy locality for the price of retro-causality.

Does it mean that TI is local? I think that wave functions in TI represent an objective reality, so isn't it in contradiction with the Bell theorem which asserts that no theory of local reality can be compatible with QM?

 

[Dmitry67]

No, CI is not local of course
I meant, take CI (which is not local), then, if one prefers locality he can get locality sacrificing causality.

CI: nonlocal, causality
TI: local, retrocausality

 

[Demystifier]

No, CI is not local of course
I meant, take CI (which is not local), then, if one prefers locality he can get locality sacrificing causality.

CI: nonlocal, causality
TI: local, retrocausality

Dmitry, I made a typo in my post above. Please respond again, now on the corrected version.

 

[Dmitry67]

oops, me too.
Of course, it is not about locality, it is about realism!

So, TI can recover the objective properties of entangled particles, because these particles from the very beginning 'know' the angles of the polarizers Alice and Bob will use in the future!

 

[Frame Dragger]

I don't mean to sound in the least insulting, but this thread has really closed the book for me on the TI, and I used to appreciate it... I thought.

I'm getting very tired of theories which add nothing to SQM's predictive capacity. We have enough ad hoc explanations for what we're already seeing; let's get a little predictivite power in the THEORY, or at least testability. Feynman and Wheeler would have 'le mot juste' for this contortion of their theory.

 

[Demystifier]

I see. Thanks, Dmitry!