The fourth FQXi essay contestWhich of Our Basic Physical Assumptions Are Wrong? is underway.


Wittgenstein knew precisely which one; I am all about that one as you may know, but we are idiosyncratic, of German culture, and incomprehensible to many. Thus I humbly ask you to return the sweat I have put into this already and please criticize the following constructively in order to help me to get the difficult message across. Tell me where the text starts to sound awkward / idiotic / unintelligible/ hopelessly nonsensical; tell me in the comments or privately. All suggestions are welcome (Title stupid? Structure upside down? Figure suggestions? …).


The first half of the total essay “Direct Realism falling in Wittgenstein’s Silence: Accelerating the Paradigm Change that Renders Relativistic Quantum Mechanics Natural” has already been revised with lots of helpful hints from readers (thank you all very much indeed). Here is the second half:

Spooky Non-Locality more Unreal than Modal Realism

  All the above is crucially relevant to the advancement of QM, which can be didactically most easily clarified with the Einstein Podolsky Rosen (EPR)paradox [4]. EPR’s violation of John Bell’s famous inequality [5] is usually presented as leaving two options, namely either so called non-locality or modifying realism. Non-locality is usually seen as less suspicious, because realism is widely confused with, for example, the scientific method generally. This situation is aggravated by the difficult to fully grasp the profound nature of QM non-locality. Even after having studied Bell’s proof [6] against hidden variables and all of that, many physicists still opine that the indicated non-locality is merely a ‘really complicated’ correlation, but in the end not profoundly different from the correlation that ensures Alice getting the left sock of a pair if Bob gets the right sock. “Really complicated” feels not as ‘ghostly’ as anti-realism.

  Einstein,although he could not find the solution in his lifetime, already understood the problem much better, and therefore he did not just say “well, so it is non-local and I am fine with that”. He called it “spooky” {Footnote a} for good reasons. Einstein would not casually throw away relativistic micro causality{Footnote b}, the arguably most successful ingredient in all of modern physics still today, just in order to prop up a kind of realism which then, via non-locality, becomes a ghost story nevertheless. QM non-locality destroys DR anyway, as shall become ever more obvious as you read on. You are far more conservative if you accept modal realism – still a realism after all.

{Footnote a: In 1947, Einstein wrote to Max Born that he could not believe that quantum physics is complete "because it cannot be reconciled with the idea that physics should represent a reality in time and space, free from spooky actions at a distance." (Emphasis added)}

{Footnote b: “Relativistic micro causality” is very roughly that stuff needs to bump against each other, that there is no mechanism for instantaneous interaction at a distance (in a direct realism (DR)). This limit derives from how one can measure.}

  Today it is known why QM non-local correlations are so spooky: They are correlations with non-actualized alternatives! It is the entanglement with counter factual possibilities, which has been empirically verified for instance in interaction-free bomb-testing [7,8]. This very core of QM is precisely what makes QM correlations stronger than any classical hidden variables could possibly provide. QM correlations can be profoundly more correlated than even complete classical correlations could ever deliver, because QM correlations are moreover/additionally fixed between the many alternative worlds that classical physics ignores.

 Variations of Bell’s inequality [9] have been violated by diverse experiments, most impressively by closing the so called “communication loophole” [10] and quite recently again by confirmation of the Kochen-Specker theorem [11]. Desperate attempts at saving unmodified realism try to exploit the so called “detection loophole”, but they have by now retreated to claiming what Shimony called a conspiracy {Footnote c} – one not much different from a creator god planting a fossil record to actively deceive us. As Einstein said: spooky!

{Footnote c: “… there is little that a determined advocate of local realistic theories can say except that, despite the space-like separation of the analysis-detection events involving particles 1 and 2, the backward light-cones of these two events overlap, and it is conceivable that some controlling factor in the overlap region is responsible for a conspiracy affecting their outcomes.” [Abner Shimony,]}

  Two clarifications should be mentioned right away in order to understand QM non-locality and the confusion around it (although these clarifications will be much better understood later on):

  1) If you consider a direct reality suitably misinterpreted as a lonely world being ‘really out there’, locality is crucially relevant. If space is a box with all objects ‘really’ at certain locations inside, localism is implicit. Non-locality in a direct realism (where spatially separated events outside of a past light cone are merely unknown yet not fundamentally undetermined/indeterminate) requires superluminal interaction. The huge success of relativistic micro causality argues against such, but is not as relevant as the following: Locality is implicitly assumed precisely because non-locality observed in an internally relativistic, micro-causal billiard table, implies that there is something else, something spooky that interferes in the mere mechanical, supposedly dynamically self-sufficient, independent ongoing of a classical box! Correlations that are proven to be faster than the fastest velocity inside the game board (its ‘velocity of light’) are a sure sign of players (gods) messing around with the game pieces {Footnote d}.

If however you modify realism, you start undermining the tacit assumption about ‘space really being out there’. In a modal realism, localism is not implicit. Therefore, if for example Smerlak/Rovelli, Bousso, Deutsch, Zeh, and so on state that QM is fundamentally local, do not confuse them with those who desperately cling to naïve directly real (DR) models and hidden variables.

{Footnote d: No ‘new atheism’ here, but a fundamental description of the totality of possible creators/creations (e.g. QM decoupling of ‘creations’ from ‘creators’) is beyond the scope of this work. All I say here is that such ghostly issues afflict any non-local relativistic DR.}

  2) The local/non-local distinction is almost precisely parallel to the determinism versus indeterminism one: QM is fundamentally a determinism (~ unitarity), and if it were not, then a more fundamental theory would be anyway, because totality is totally determined as all that there is (including all times etc.). However, this very fact allows the phenomenal world that we perceive to show in-determinism. Modal realism (MR) assures our experiencing QM in-determinism. The same goes through for locality: QM is fundamentally Einstein-local (micro causality, no correlations faster than light), but MR allows the classical worlds that we seemingly find ourselves in to appear QM non-local. Admittedly, without an intuitive model, I would find this last statement suspiciously mysterious. I therefore promote Many World (MW) models which greatly clarify these issues in visually intuitive ways.

Everett Relativity: Einstein could have Known

  Quantum physics is our starting to find the mathematical description that does not neglect any possibilities. At first it was thought such encompasses only the alive and dead Schrödinger cats in our particular universe, but this immediately includes all the ways a universe may be described to unfold internally as its own Schrödinger box, so it contains anyway all possible universes; everything possibly phenomenal is included in the ultimate description, else it is not the ultimate description. The core insight is not quantum; already with classical determinism would an ultimate description contain all possible worlds. "Many worlds" are tautologically true. The core of QM is the interference (entanglement) between alternatives. Ignorance about non-actualized alternatives is no longer an option. But is it news?

 Theoretical quantum gravity considerations have lead Gibbons and Hawking to propose an ‘anti-realistic’ observer dependent definition of particles already in the seventies, clearly endorsing “something like the Everett-Wheeler interpretation of quantum mechanics” [12]. Zeilinger stresses anti-realism for a number of years, recently with a novel setup [13]. Black hole complementarity (holographic horizons), and the cosmological measure problem all point to local, observer dependent ‘causal patch’ descriptions that crucially deny ‘real reality’ to what lies beyond horizons. The horizons are wrapped ever closer around the observer. Why do we still largely ignore the paradigm change? Answering that is a different essay, but for now, the message is that perhaps the following should not surprise:

 Einstein could conceivably have anticipated QM long before even the Everett relative state description in 1957 [14]. Although EPR is commonly misunderstood as a clash between relativity and QM, Everett relativity is only suspect without special relativity {Footnote e}. Special relativity (SR) is more than merely a ‘temporal modal realism’. SR already deconstructs the classical world into a collection of past light cones, which each are an ‘observer’s’ individual determined past. Assuming otherwise immediately implies a fully pre-determined, directly real block universe where any phenomenal indeterminism is divine pre-arrangement. SR already demanded MR to enter physics. Einstein locality and micro causality are yet more important than already widely recognized. They prepare QM, which merely (see VI first part) adds correlations between alternative “determined pasts”.

{Footnote e: A branching MW model illustrates why: A non-relativistic universe would have to branch everywhere at once, into infinitely many different ones, all the time. SR is thus prerequisite for understanding QM, because apparent “world branching” only occurs at observation events (interactions).}

  In fact, I expect that the following argument will be made rigorous: SR and the demand that at least some possibilities should be unobservable, both together imply that there is some “mechanics” that lets different alternatives interact: Quantum mechanics’ interference. In this way, Einstein could have conceivably understood QM and resolved the EPR paradox right away.

Many World Models – Just Models (!) – That Though Immensely Clarify

 Relational Quantum Mechanics (RQM) [15] provided a resolution [2] of the EPR paradox. Its abstractness lacks an intuitive, didactic picture and also does not yet, so it seems to me, aim at deriving the QM Born probabilities (which I believe can be done by considering the fact that Alice and Bob, the two famous participants in the EPR setup, must exchange information about the relative angle setting every time, not just the outcome of the spin measurement). My own work [16] has contributed many world (MW) models which clarify vital aspects in an intuitive way – you can kind of sort of “really see it” (again, it is modal realism). Allow me to convey a few lessons that such models taught me, as I constructed them somewhat by accident and they subsequently indeed taught me these lessons with a clarity that I did not expect to be possible before.

 However, I must first stress with the strongest possible emphasis that I do not imply any ontology, do not imply that MW ‘branch counting’ does not mess up normalizations in the measure problem, and do certainly not claim that any particular model is on the same level as tautological truths. The model I refer to never claimed more than proving that 1) MR resolves EPR while preserving Einstein locality and 2) that the QM probabilities become possible with the precise step that destroys the DR description. I suspect a certain way in which the model offers to derive the correct QM probabilities instead of just allowing violations of Bell inequalities, but even that was never claimed yet.

  It is far beyond the current work to introduce whole MW models, but I think it is not impossible to understand the following even without having a model, which look initially like sausages incidentally, in front of one’s inner eyes.

  1) One aspect is that MW and MR can be classical. The initially classical model already has many worlds, but they do not interfere. They are literally parallel worlds (but of course QM orthogonal states). Therefore, a little arrow, literally a little arrow labeled “DR”, can pick out the one world that a direct realist believes to be the only real one.

  2) The nature of the QM non-locality is illuminated: The mere classical non-locality in the form of the (anti)-correlations are already present in the classical model, even at all relative angles between Alice’s and Bob’s measurement crystals. However, the Bell inequality cannot be violated, because it is still a classical model. The deeper QM non-locality arises at the very point, again literally the very space-time point, at which the number of worlds in the MW model branches according to how different alternative worlds correlate (interact) locally. Let me refer to this vital step as the “last local branching”. Thus, the QM non-locality is very clearly separated into its classical part and the QM component, the latter being necessarily correlations between alternative worlds.

  3) QM indeterminism: Looking at the model, all possible alternative outcomes are already there {Footnote f} and thus everything is determined. You can point arrow DR to any of the resulting worlds where certain outcomes are known. However, if the last local branching is QM, there is no preferred path that DR can follow from one initial world to a certain final one.

{Footnote f: Given a random angle setting w.l.o.g.; multiplying by all angle alternatives would be very bothersome. Also: those are not equally distributed. All aliens throughout the universe experiment at the Bell angles (n Pi /8) where it makes most sense to do so.}

  4) QM probability (empirical, Bayesian) is very different from the classical, circular “fair randomness” and its regress error: DR cannot, say via hidden variables, be initially rigged to end up in any certain one of the new worlds (their numbers depend on the relative angle that is not yet determined in the beginning). It may, via randomness added from outside of the model, be made to point to a particular new world, but such interference does not touch the classical probabilities. The classical measure stays the same while the empirical probabilities of observers violate the Bell inequality. Inside of an MW model, observers have neither access to the classical probability measure nor can they directly count other worlds. But they remember and add to their records, and those predominantly reflect the QM probabilities.

  5) In a modified realism, QM non-locality comes for free: The expected phenomenal observations, i.e. the empirical data that the different worlds contain, depends primarily on the number of “new” worlds (especially the “extra branching” {Footnote g} that Wallace so despises [17]). Therefore, once MW and thus MR is in place and you, the god of the model, add branching of worlds into new worlds, you may change the model’s parameters to let the inhabitants find locality in their perceived worlds, or non-locality, or even, if you micromanage enough to change the numbers of worlds at will at every measurement separately, that the moon is made from green cheese! If the realism is so modified that all is 'only in your head' instead of “out there” to such a degree that the moon may as well become green cheese dancing over the mountains, obviously it does no longer matter whether the moon “is really at a certain locality”. The distinction between local and non-local lost its relevance, because the model can adjust it arbitrarily; it can tune down the QM non-locality until it vanishes.

{Footnote g: “Extra growth” does not necessarily mean that the total number rises (new worlds), because the changing of microstates means that information is also “forgotten”, i.e. distinctions not relevant to the measurement in question disappear.}

  6) In nature (not a model on paper), you cannot change the parameters at will. Some consistency fixes the tuning so that we find non-locality instead of locality or cheese moons. But what consistency? I suggest two ways to think about it:

6.1) The Born probabilities in the EPR setup are constrained by the consistency between the quantum (~ microcosm) and the classical (~ macro world) descriptions. In the former alone, photons may as well just go 50-50 into the two different beam splitter channels without leading to Bell inequality violation, but in the latter, you know from playing with polarized sunglasses that polarization filters act on the light’s electromagnetic vectors, thus they diminish intensity via a sine-squared law – nothing else would make sense in a macro world. (This point 6.1 should possibly be put before mentioning MW models.)

6.2) The MW model shows that the numbers of new worlds are proportional to the dot product between Alice’s and Bob’s measurement axes in order to get probabilities consistent with experiments. This suggests that the neglected microstates which carry the information about the relative angle setting are what distinguish these worlds (because the resolution of an angle measurement depends on the total spin available).


What do you think will be the conclusion? Or do you think I should not add one and instead add another 2000 words to make the above clearer? Please let me know.


I am sincerely grateful for the many crucial questions and revision suggestions that readers of the internet draft have contributed in order to make this text as accessible and clear as possible.

[1] Ludwig Wittgenstein: “Tractatus Logico-Philosophicus”, Routledge and Kegan Paul, London (1922)

[2] Smerlak, Matteo, Rovelli, Carlo: “Relational EPR." Found. Phys. 37,427-445 (2007)

[3] Lewis, David Kellogg: “On the Plurality of Worlds.” Blackwell (1986)

[4] Einstein, A., Podolsky, B., Rosen, N.: “Can Quantum-Mechanical Description of Physical Reality be Considered Complete?” Phys. Rev. 47(10),777-780 (1935)

[5]  Bell, J.S.: “On the Einstein Podolsky Rosen paradox.” Physics 1(3), 195–200 (1964); reprinted in Bell, J.S.Speakable and Unspeakable in Quantum Mechanics. 2nd ed., Cambridge: Cambridge University Press, 2004; S. M. Blinder: “Introduction to Quantum Mechanics.” Amsterdam:Elsevier, 272-277 (2004)

[6]  Bell, J. S.: “On the problem of hidden variables in quantum mechanics.” Rev. Mod. Phys. 38,447–452 (1966)

[7] Elitzur, A.C., Vaidman, L.: “Quantum mechanical interaction-free measurements.” Found. Phys. 23, 987-97 (1993)

[8] Paraoanu, G.S.: “Interaction-free Measurement.” Phys. Rev. Lett. 97(18),180406 (2006)

[9] Clauser, J.F., Horne, M.A., Shimony, A., Holt, R.A.: “Proposed Experiment to Test Local Hidden-Variables Theories.” Phys. Rev. Lett. 23,880-884 (1969)

[10] G. Weihs, T. Jennewein, C. Simon, H. Weinfurter, A. Zeilinger: “Violation of Bell’s inequality under strict Einstein locality condition.”Phys. Rev. Lett. 81, 5039-5043 (1998)

[11] G. Kirchmair, F. Zähringer, R. Gerritsma, M. Kleinmann, O. Gühne, A.Cabello, R. Blatt, C. F. Roos: “State-independent experimental test of quantum contextuality.” Nature 460, 494-497 (2009)

[12] Gibbons, G. W., Hawking, S. W.: “Cosmological event horizons,thermodynamics, and particle creation.” Physical Review D 15(10), 2738–2751(1977)

[13] Radek Lapkiewicz, Peizhe Li, Christoph Schaeff, Nathan K. Langford, Sven Ramelow, MarcinWiesniak, Anton Zeilinger: “Experimental non-classicality of an indivisible quantum system.” Nature 474, 490-493 (2011)

[14] Everett, Hugh: “‘Relative State’ Formulation of Quantum Mechanics.” Rev Mod Phys 29, 454-462(1957), reprinted in B. DeWitt and N. Graham (eds.), The Many-Worlds Interpretation of Quantum Mechanics, Princeton University Press (1973)

[15] Rovelli, Carlo: “Relational Quantum Mechanics.” Int. J. Theor. Phys. 35,1637-1678 (1996)

[16] S. Vongehr: “Many Worlds Model resolving the Einstein Podolsky Rosen paradox via a Direct Realism to Modal Realism Transition that preserves Einstein Locality.” arXiv:1108.1674 [quant-ph] (2011)]

[17] D. Wallace: “Quantum probability from subjective likelihood: Improving on Deutsch’s proof of the probability rule.” SHPMP 38, 311-332 (2007)