Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Transactional Interpretation, space and information

  1. Oct 27, 2006 #1
    While trying to wrap my head around the Transactional Interpretation, I wondered why we need signals from the future. For the electron about to be ejected, the surrounding space is already filled with information about everything around, and if the momentum conditions assign it to strike beyond the double slit pane, both paths will already be cleared, signaled as those of least action. Am I overly naive in thinking so?
  2. jcsd
  3. Oct 29, 2006 #2
    My understanding is that you need both past and future boundary conditions so as not to interpret there to be a random probabilistic collapse of the wave function.
  4. Oct 29, 2006 #3
    As far as you know, the Transactional Interpretation is really able to solve all known paradoxes and is really free from limitations in describing QM results?
    What do physics community think about it?
  5. Oct 30, 2006 #4
    Well, not necessarily, but more words are needed to explain the correlation in delayed choice versions of Bell test experiments. (This is something I've also been thinking about lately)

    Like, let's say we are measuring a polarization of two entangled photons flying to different directions, and we choose which way to align the measurement device A while the photon A is already on its way.

    While it is conceivable that the photon A can meet the information about the measurement device on its way and align itself accordingly, we cannot say how the photon B receives information about any of this; it will hit the measurement device at its end before any local information reaches it.

    And how do we know that B receives any information about how the A was measured at all? The experimental data seems to suggest it does; if the measurement device A "forces" a certain polarization to photon A that is diagonal to the directions that device B is measuring, it really looks like photon B is hitting device B in such diagonal angles (as interpreted from the correlation between A & B measurements)

    This is a simplification of real experiments of course, but anyhow...

    It seems to me that it is not very popular interpretation, which is odd because if you believe into Relativity (which most of the physics community does :), transactional interpretation pretty much naturally follows. Yet, physics community tends to talk in terms of newtonian time when they talk about "delayed choice experiments" and what not. Even the name "transactional..." suggests whoever named that interpretation was basically thinking in terms of evolution over time; only that evolution moves forward and backward.

    The odd thing with spacetime, or at least the simplest ontological flavor of it, is that it does not evolve in any sense. The routes taken by the photons do not really evolve one way or another according to how the measurement device is rotated "while they are on the way". The route of the photon just exists in static manner, and in the "inertial frame" of the photon, time does not pass "during flight". The measurement device is already measuring the photon "at the same time" as the photon begins its journey. And it seems fairly straightforward to assume that when the entangled photons are emitted, their "future measurements" exist in the future only in our inertial frames, and as such these measurements are bound to affect the way the photons are emitted in the "semantical past".

    It is kind of funny that I don't personally give much weight to this sort of ontology of spacetime, but most of the physics community seems to think it is reasonable interpretation of relativity. Yet the same community refuses to look at reality this way when it comes to quantum mechanics, and rather resorts to concepts where evolution over time can be considered "real", such as in MWI.

    I guess we are too used to think in terms of motion to really understand what it would mean if motion did not exist in metaphysical reality. This is quite evident to me every day, and notice that even Einstein didn't realize how trivially spacetime resolves his EPR-paradox.

    Take a look at a related post here:
    (Spacetime interpretation is basically the same as "transactional interpretation", I just didn't know of such a thing when I wrote that, and as I said, "transactional" suggests metaphysical motion which is not coherent with spacetime. Also notice that I do not particularly advocate this interpretation as being ontologically very valid, but it can still be useful idea)

    So, while "transactional interpretation" might seem as completely looney, it is not more looney than the idea that there really is a metaphysical spacetime.
    Last edited: Oct 30, 2006
  6. Oct 30, 2006 #5
    Thank you to everybody who replied. I lack the knowledge to understand some things, but I do understand the link you draw between the TI and GR spacetime, AnssiH. In such a case the whole notion of causality has to be revised - the notion that the future evolves in the transactions seems a fallacy. One way out would be to add a second time dimension along which spacetime constructed and deconstructed itself.:bugeye: What would be the rules of construction and evolution of spacetime along that "hypertime"? What would entanglement mean in that logic? I guess I'll go through a few headaches.
  7. Oct 31, 2006 #6
    Yeah, you are approaching the problems with our ideas of motion and spacetime. They don't mix very well. Adding another time dimensions gets you forward by allowing motion to spacetime itself, but it tends to end up into infinite regress when you get deeper into the properties of that second time dimension, and also many features of "semantical motion" become nonsensical when imagining motion onto spacetime itself, for example the speed of this motion causes no observable effects at all.

    This is not necessarily a problem of transactional interpretation though, since it is trivial to just use such terminology where the spacetime exists in static sense, but it can be a problem to the spacetime interpretation of relativity. There are some very good reasons to assume it could be motion after all that is more "real" in metaphysical sense, as oppose to spacetime as a static construction.

    Quite a few threads lately seem to touch that subject more or less (hmmm, perhaps I am to blame...), so I'll just link to one recent post:
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook