Skip to main content

On the Very Idea of Distant Correlations


Contemporary debate over laws of nature centers around Humean supervenience, the thesis that everything supervenes on the distribution of non-nomic facts. The key ingredient of this thesis is the idea that nomic-like concepts—law, chance, causation, etc.—are expressible in terms of the regularities of non-nomic facts. Inherent to this idea is the tacit conviction that regularities, “constant conjunctions” of non-nomic facts do supervene on the distribution of non-nomic facts. This paper raises a challenge for this conviction. It will be pointed out that the notion of regularity, understood as statistical correlation, has a necessary conceptual component not clearly identified before—I shall call this the “conjunctive relation” of the correlated events. On the other hand, it will be argued that there exists no unambiguous, non-circular way in which this relation could be determined. In this regard, the notion of correlation is similar to that of distant simultaneity where the necessary conceptual component is the one-way speed of light, whose value doesn’t seem to be determined by matters of (non-nomic) facts.

This is a preview of subscription content, access via your institution.

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8
Fig. 9
Fig. 10


  1. X is sometimes called “sample space.” Note that the elements of the sample space should not be confused with the elements of the statistical ensemble. The elements of the sample space are event types, while the elements of the statistical ensemble are token events. In the two coin example, the sample space we consider is \(\left\{ \left( H,H\right) ,\left( H,T\right) ,\left( T,H\right) ,\left( T,H\right) \right\}\); while the statistical ensemble consists of particular instances of double tosses, different elements of which may realize one and the same element of the sample space.

  2. Conditions (1)–(3), defining the notions of independence and correlation, are not expected to hold exactly in general, but only with some finite precision depending on the size of the statistical ensemble. In this sense there is no sharp boundary between cases of statistical independence and correlation, and the question of independence versus correlation should be seen as a matter of degree rather than a matter of kind. Note however that the choice of the conjunctive relation may alter whether a pattern of data is seen as evidence of independence or maximal correlation (with some finite precision), which are the two extremes on the correlation scale.

  3. This condition may or may not be specifiable merely in terms of \(\mathscr {F}\) and \(\mathscr {F}'\).

  4. Alternatively, one can take a finite partition of the unit square into small cells of equal size, and identify the ensembles with the finite collections of these cells.

  5. It must be noted that the Clauser–Horne inequalities are only sufficient in conjunction with the following supplementary inequalities (as also noted by [7, p. 293]):

    $$\begin{aligned} \begin{array}{ccc} p\left( A_{i}\wedge B_{j}|a_{i}\wedge b_{j}\right) \le p\left( A_{i}|a_{i}\right) \\ p\left( A_{i}\wedge B_{j}|a_{i}\wedge b_{j}\right) \le p\left( B_{j}|b_{j}\right) \\ p\left( A_{i}|a_{i}\right) +p\left( B_{j}|b_{j}\right) -p\left( A_{i}\wedge B_{j}|a_{i}\wedge b_{j}\right) \le 1 \end{array} \end{aligned}$$

    These however follow from (12)–(16) and thus are respected by all three conjunctive products.

  6. A local hidden variable model provides a common causal explanation. However, after modifying the conjunctive relation the pairs of events in the two wings matched together may no longer be spatially separated and so explaining the statistics relative to the modified ensemble may also invoke direct causal connection between the two wings. Nevertheless, what we have demonstrated here is that an explanation purely in terms of common causes is also available.

  7. There do exist experiments in which particle pairs can be identified independently of their detections. However, in none of these experiments they succeeded to ensure spatial separation of the two wings [19, Sect. 3].

  8. The thought that the particles’ time of arrival at the detectors may depend on the hidden variable is not new. This is the main idea of hidden variable models based on the so-called coincidence loophole [6, 8, 20, 21]. In these models one builds on the fact that in real EPR-type experiments pairs of detection events in the two wings are counted as “coincident” if they happen within a given time window; if the retardation of one detection with respect to the other is larger then the prescribed amount the data is simply ignored and will not contribute to the statistics. If, as it is assumed in the models in question, there is a systematic connection between the value of the hidden variable and the detections that are counted as coincident, then the statistics over the subensemble of coincident detections may exhibit violations of Bell’s inequalities even though the statistics over the original ensemble, being produced by a fully local mechanism, adhere to them.

    There is another type of “loophole” in Bell’s theorem related to the fact that the statistical data is given as a temporal series of outcomes. In the original Bell analysis events in different runs of the experiment are treated as independent. However, since different runs of an EPR-type experiment are timelike separated, earlier values of outcomes, settings and the hidden variable can in principle influence the later ones. For example, if the consecutive measurement choices in one wing are correlated, one can imagine a mechanism whereby the local measurement apparatus predicts the current remote measurement choice by registering the earlier ones. This way the remote measurement setting could influence the local outcome, resulting in a violation of Bell’s inequalities, by means of a fully local mechanism. This problem is sometimes referred to as the memory loophole [2, 9,10,11, 14].

    While both the coincidence and memory loopholes are related with the role of time in the EPR–Bell problem, and thereby they have some similarities with the idea of the time-delayed models discussed here, it must be emphasized that the main thought behind the time-delayed models is different from the problems discussed in the literature. The reason is that all existing hidden variable models assume that it is the statistics of the simultaneous measurements that is to be explained by a local mechanism, while the essence of the time-delayed model idea is to question that the simultaneous statistics exhibit real correlations in need of explanation.

  9. Here we assume that a token event can be individuated without referring to the event types it falls under. Such an identification seems possible, for example, by referring to the spatiotemporal location of the token events.

  10. It must be emphasized that the physical realizability of the simultaneous pairing has nothing to do with the issue of whether simultaneity is factual or conventional, in the sense as it arises with regard to the problem of distant clock synchronization. What gives physical objectivity to the simultaneous pairing is not that the local clocks in question are synchronized or that they measure the “same” (temporal) order—in fact no such an assumption has to be made. What is important here is that the counters that we call clocks—synchronized or not—are identically constructed and therefore there is a physical correspondence between their readings, based on their similarity. This physical correspondence is what grounds the reality of the simultaneous pairing.


  1. Aspect, A., Philippe, G., Roger, G.: Experimental tests of realistic local theories via Bell’s Theorem. Phys. Rev. Lett. 47, 460 (1981)

    ADS  Article  Google Scholar 

  2. Barrett, J., Collins, D., Hardy, L., Kent, A., Popescu, S.: Quantum nonlocality, Bell inequalities, and the memory loophole. Phys. Rev. A 66, 042111 (2002)

    ADS  Article  Google Scholar 

  3. Clauser, J.F., Shimony, A.: Bell’s theorem experimental tests and implications. Rep. Progr. Phys. 41, 1881 (1978)

    ADS  Article  Google Scholar 

  4. Einstein, A.: On the electrodynamics of moving bodies. In: Lorentz, H.A., et al. (eds.) The Principle of Relativity: A Collection of Original Memoirs on the Special and General Theory of Relativity. Methuen and Company, London (1905)

    Google Scholar 

  5. Einstein, A., Podolsky, B., Rosen, N.: Can quantum-mechanical description of physical reality be considered complete? Phys. Rev. 47, 777–780 (1935)

    ADS  Article  Google Scholar 

  6. Fine, A.: Correlations and physical locality. In: Asquith, P., Giere, R. (eds.) PSA 1980, vol. 2, pp. 535–562. Philosophy of Science Association, East Lansing (1981)

    Google Scholar 

  7. Fine, A.: Hidden variables, joint probability, and the Bell inequalities. Phys. Rev. Lett. 48(5), 291–294 (1982)

    ADS  MathSciNet  Article  Google Scholar 

  8. Fine, A.: Some local models for correlation experiments. Synthese 50, 279–294 (1982)

    MathSciNet  Article  Google Scholar 

  9. Gill, R.D.: Accardi contra Bell (cum mundi): the impossible coupling. In: Moore, M., Froda, S., Léger, C. (eds.) Mathematical Statistics and Applications: Festschrift for Constance van Eeden, IMS Lecture Notes-Monographs, vol. 42, pp. 133–154. Institute of Mathematical Statistics, Beachwood (2003)

    Google Scholar 

  10. Gill, R.D.: Time, finite statistics, and Bell’s fifth position. In: Foundations of Probability and Physics 2 (Växjö, 2002). Mathematical Modelling in Physics, Engineering and Cognitive Science, vol. 5, pp. 179–206. Växjö University Press, Växjö (2003)

  11. Gill, R.D., Weihs, G., Zeilinger, A., Zukowski, M.: No time loophole in Bell’s theorem: the Hess–Philipp model is nonlocal. Proc. Natl. Acad. Sci. USA 99, 14632–14635 (2002)

    ADS  MathSciNet  Article  Google Scholar 

  12. Hájek, A.: ‘Mises Redux’–Redux: fifteen arguments against finite frequentism. Erkenntnis 45, 209–227 (1997)

    MathSciNet  MATH  Google Scholar 

  13. Hájek, A.: Fifteen arguments against hypothetical frequentism. Erkenntnis 70, 211–235 (2009)

    MathSciNet  Article  Google Scholar 

  14. Hess, K., Philipp, W.: A possible loophole in the theorem of Bell. Proc. Natl. Acad. Sci. USA 98, 14224–14227 (2001)

    ADS  MathSciNet  Article  Google Scholar 

  15. Hofer-Szabó, G., Rédei, M., Szabó, L.E.: The Principle of the Common Cause. Cambridge University Press, Cambridge (2013)

    Book  Google Scholar 

  16. Hume, D.: An Enquiry Concerning Human Understanding. Oxford University Press, Oxford (2007)

    Google Scholar 

  17. Immerman, N.: Computability and complexity. In: Zalta, E.N., (ed.) The Stanford Encyclopedia of Philosophy (Spring 2016 Edition) (2016).

  18. Janis, A.: Conventionality of simultaneity. In: Zalta, E.N., (ed.) The Stanford Encyclopedia of Philosophy (Fall 2014 Edition) (2014).

  19. Larsson, J.-Å.: Loopholes in Bell inequality tests of local realism. J. Phys. A Math. Theor. 47, 424003 (2014)

    MathSciNet  Article  Google Scholar 

  20. Larsson, J.-Å., Gill, R.D.: Bell’s inequality and the coincidence-time loophole. Europhys. Lett. 67, 707–713 (2004)

    ADS  Article  Google Scholar 

  21. Pascazio, S.: Time and Bell-type inequalities. Phys. Lett. A 118, 47 (1986)

    ADS  MathSciNet  Article  Google Scholar 

  22. Rynasiewicz, R.: Simultaneity, convention, and gauge freedom. Stud. Hist. Philos. Mod. Phys. 43, 90–94 (2012)

    MathSciNet  Article  Google Scholar 

  23. Salmon, W.C.: The philosophical significance of the one-way speed of light. Noûs 11, 253–292 (1977)

    Article  Google Scholar 

  24. Sober, E.: The principle of the common cause. In: Fetzer, J. (ed.) Probability and Causality, pp. 211–229. Reidel, Dordrecht (1988)

    Chapter  Google Scholar 

  25. Szabó, L.E.: Critical reflections on quantum probability theory. In: Rédei, M., Stoeltzner, M. (eds.) John von Neumann and the Foundations of Quantum Physics, pp. 201–219. Kluwer, Dordrecht (2001)

    Chapter  Google Scholar 

  26. Szabó, L.E.: The Einstein–Podolsky–Rosen Argument and the Bell Inequalities, Internet Encyclopedia of Philosophy (2008).

  27. Szabó, L.E.: Meaning, Truth, and Physics. In: Hofer-Szabó, G., Wroński, L. (eds.) Making it Formally Explicit, European Studies in Philosophy of Science, vol. 6, pp. 165–177. Springer, Berlin (2017)

    Chapter  Google Scholar 

  28. Winnie, J.A.: Special relativity without one-way velocity assumptions. Part I and II. Philos. Sci. 37, 81–99 (1970)

    Google Scholar 

Download references


This work has grown out from discussions with László E. Szabó on his physicalist account of scientific theories [27], in which the semantics of a theory is analyzed in terms of the “distant correlation” of facts of the world and “facts” of the theory. I wish to thank László for these discussions. I am also grateful to Gábor Hofer-Szabó and Balázs Gyenis for valuable comments on earlier versions of the paper. The research was supported by the (Hungarian) National Research, Development and Innovation Office (NKFIH), No. K115593

Author information

Authors and Affiliations


Corresponding author

Correspondence to Márton Gömöri.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Gömöri, M. On the Very Idea of Distant Correlations. Found Phys 50, 530–554 (2020).

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI:


  • Regularity
  • Correlation
  • Causal explanation
  • Common Cause Principle
  • EPR–Bell
  • Distant simultaneity