Papex

Open problem · Physics · Open

Arrow of time

Why does time have a direction when the basic laws of physics work almost the same forwards and backwards?

Summary

Entropy increases, we remember the past and not the future, and causes precede effects, yet the underlying microscopic laws are nearly time-symmetric; how these arrows of time arise and relate to one another is unresolved.

Why it is open

It is almost universally acknowledged, the philosopher Craig Callender writes, that the time asymmetry of thermodynamic phenomena is captured by the second law of thermodynamics, the statement that entropy increases [12]. Yet the laws of the four fundamental forces are time symmetric, with one caveat: a very weak asymmetry in the weak interaction, which George Ellis argues "does not have enough purchase on the rest of physics" to be the origin of the asymmetry we see at macroscopic scales [7].

Callender separates two problems [12]. The first is where the thermodynamic asymmetry comes from. The usual answer adds an initial state of the universe with very low entropy [1, 12], a posit with its own page, low entropy of the early universe. The second, the subject of this page, asks whether the other arrows hold "in virtue of the thermodynamic arrow of time (or what grounds it)" [12].

The other arrows are the epistemological arrow (we know the past in a different way from the future), the mutability arrow (the future feels open, the past fixed), the psychological arrow, and the linked arrows of causation, explanation and counterfactual dependence [12]. Physics adds the radiative arrow, in which waves spread out from sources rather than converge on them, and the asymmetry of quantum measurement [7, 12]. Which arrow depends on which is itself disputed, and the answer "will depend, to a large degree, upon one's general philosophical stance" [12].

No body sets a bar for a resolution. An answer would have to show, for each arrow, whether it holds in virtue of the thermodynamic arrow or independently of it [12]; for the psychological arrow, Meir Hemmo and Orly Shenker propose two necessary conditions on the workings of the brain that any account must satisfy [11].

What is known

The thermodynamic arrow (1999–2019). In the statistical account that Joel Lebowitz reviews, entropy increase rests on the great disparity between microscopic and macroscopic scales, initial conditions, and probability, since only an overwhelming "majority" of microscopic states obey the second law [1]. A far-from-equilibrium initial macrostate is needed, but no special initial microstate [1]. Standard thermodynamics also presupposes that interacting systems start uncorrelated [10]. In 2019 Kaonan Micadei and colleagues prepared two quantum-correlated nuclear spins in local thermal states at different effective temperatures and observed heat flowing spontaneously from the colder to the hotter spin, made possible by a decrease in their mutual information, so that the second law still held for the isolated pair [10]. They conclude that the arrow of time "is not an absolute but a relative concept that depends on the choice of initial conditions" [10].

The radiative arrow (1909 onwards). Maxwell's equations allow "retarded" solutions, fixed by sources in the past, and "advanced" ones, fixed by sources in the future, and physicists routinely discard the advanced ones [12]. In a debate in 1909 Walther Ritz is said to have held that the radiation asymmetry is a law from which the thermodynamic one follows, and Albert Einstein that irreversibility rests on probability; the actual exchange was "far more complicated" [12]. Most agree that the asymmetry of spreading waves is "ultimately 'thermodynamic'", Callender reports, but emission versus absorption and the damping of accelerated charges remain contested [12].

Weak-interaction T violation (2012). Violation of charge–parity (CP) symmetry, first seen in neutral kaon decays, is expected to imply violation of time-reversal (T) symmetry, because quantum field theories of the usual kind are invariant under the combined CPT transformation [6]. In 2012 the BaBar collaboration compared transitions of B mesons with their time-reversed counterparts and observed T violation directly, at a significance equivalent to 14 standard deviations [6]. What it does not explain: these violations "don't appear to be responsible for thermodynamic behavior", Callender writes [12], and Leonard Mlodinow and Todd Brun note that in most practical applications C and P violations "play no role", so that the dynamics can be treated as invariant under T [8]. CP violation has also been studied in relation to the excess of matter over antimatter [9], a separate problem, the baryon asymmetry.

Quantum measurement. Interpretations in which the wave function collapses have so far always made the collapse time asymmetric: "there are collapses but no 'uncollapses'" [12]. Yakir Aharonov and Lev Vaidman argue that the asymmetry may lie in the formalism rather than in nature: their two-state vector formalism, which originated in work by Aharonov, Bergmann and Lebowitz, describes a system with states fixed by measurements both before and after, and makes the same predictions as standard quantum mechanics [2]. How measurement works is disputed, the measurement problem [12].

Records and memory (2009–2022). Lorenzo Maccone argued in 2009 that in quantum mechanics processes in which entropy decreases leave no information that they happened, which would make the second law "a mere tautology" [3]. David Jennings and Terry Rudolph gave a three-qubit counterexample in which local entropy falls while classical correlations, the record, increase [5]. Maccone acknowledged the flaw, defended his core argument and added that "it seems necessary to separately postulate a low entropy initial state" [4]. Mlodinow and Brun argue that any memory, even a reversible one, generically aligns with the thermodynamic arrow, because that arrow is robust to small perturbations and a memory should not have to be fine-tuned to what it records [8]. Hemmo and Shenker argue the reverse: in their account the psychological arrow is fundamental, and the direction of entropy increase is derived from it [11].

Main approaches

One source for all arrows. A philosophical programme based on statistical mechanics. David Albert and Barry Loewer derive the thermodynamic, epistemic and causal arrows from one package, the "Mentaculus": the low-entropy initial state, a uniform probability over the microstates that realise it, the present macrostate and the dynamical laws [12]. Strongest argument: one assumption, needed anyway for the second law, would account for several arrows [12], and physical models of memory agree [8]. Main difficulty: many find the claim that statistical mechanics settles nearly every generalisation that supports counterfactuals "going too far"; the mutability and psychological arrows have been discussed less in the programme [12]; and Hemmo and Shenker argue that the low-entropy past is "neither necessary nor sufficient" for the psychological arrow [11].

Independent arrows written into law. A proposal about laws. Oliver Penrose and I. C. Percival's 1962 Law of Conditional Independence requires the effects of interactions to come after them, building the causal arrow into physics; on the common account Ritz made the radiation arrow a law [12]. Strongest argument: for the radiation arrow, Callender's survey notes that electromagnetism has several arrows besides spreading waves, so that making the thermodynamic arrow rule them all is, in its word, "premature" [12]. Main difficulty: Huw Price and Lawrence Sklar hold that such a restriction is unwarranted or unexplanatory, and Callender notes that Ritz's position implies that gases of neutral particles would not tend to spread out [12].

Time-asymmetric dynamics. Physical proposals for new laws. In the Ghirardi–Rimini–Weber (GRW) collapse theory, which Albert has studied, entropy is overwhelmingly likely to increase towards the future without being likely to increase towards the past [12]. Joan Vaccaro argues from a sum-over-paths model that T violation may be what makes matter evolve in time at all and what gives time a direction [9]. Strongest argument: GRW avoids the problem in its usual form [12], and T violation is observed [6]. Main difficulty: GRW does not explain how the universe came to be out of equilibrium [12]; Vaccaro's model is deliberately simple, and further work is needed to devise feasible experiments [9].

Directed time and top-down conditions. Metaphysical and cosmological positions. Some philosophers hold that time itself is directed; Robert Weingard proposed a time-ordering field that could act as a "master arrow" [12]. Ellis argues that a past condition cascades down from cosmological to microscopic scales and sets even the quantum arrow, in an "evolving block universe" in which the future does not yet exist [7]. Strongest argument: a purely local mechanism gives no reason for the arrow to point the same way everywhere [7]. Main difficulty: no mechanism yet couples a time-ordering field to thermodynamic processes, so the idea remains "interesting yet embryonic" [12]; Ellis's account of the quantum arrow depends on how state-vector reduction works, which is unresolved [7].

Open sub-questions

  1. Correlations at scale. How does the reversed heat flow that Micadei and colleagues saw between two correlated spins scale with the size of the systems, a question the authors leave open [10]?
  2. Quantum records. For which classes of three-party states can local entropy fall while the classical mutual information with a memory rises, as in the counterexample of Jennings and Rudolph [5]?
  3. Memory without fine-tuning. In a simulated reversible memory that meets the conditions of Mlodinow and Brun, does a record of an entropy-decreasing process survive, against Maccone's argument [3, 8]?
  4. Emission and absorption. Can the emission–absorption and damping asymmetries of electromagnetism be derived from the thermodynamic arrow without assuming retarded fields [12]?
  5. Testing T-violating time evolution. Which observable, such as a departure from exponential decay in T-violating matter, would separate Vaccaro's model from standard quantum mechanics [9]?
  6. The brain's arrow. Which measurable brain process could meet Hemmo and Shenker's two necessary conditions for the psychological arrow [11]?

Work on this in Papex

If you have a testable answer to one of the sub-questions above, you can draft it as a private hypothesis in Papex, or prepare the draft from Claude or ChatGPT with the Papex connector. You review the exact version, sign it and decide whether to publish it; a public record that names this problem is listed on this page. A Papex record shows who signed which version and when. It is not peer review and does not mean the hypothesis is correct.

Key references

  1. Lebowitz, J. L. (1999). Statistical mechanics: a selective review of two central issues. Reviews of Modern Physics 71(2), S346–S357. 10.1103/RevModPhys.71.S346 · Checked 15 Oct 2026
  2. Aharonov, Y. & Vaidman, L. (2008). The two-state vector formalism: an updated review. In J. G. Muga, R. Sala Mayato & Í. L. Egusquiza (eds.), Time in Quantum Mechanics, Lecture Notes in Physics 734, Springer, 399–447. 10.1007/978-3-540-73473-4_13 · Checked 15 Oct 2026
  3. Maccone, L. (2009). Quantum solution to the arrow-of-time dilemma. Physical Review Letters 103(8), 080401. 10.1103/PhysRevLett.103.080401 · Checked 15 Oct 2026
  4. Maccone, L. (2009). A quantum solution to the arrow-of-time dilemma: reply. Preprint, arXiv:0912.5394 (version 1, 30 December 2009), not peer reviewed. Source · Checked 15 Oct 2026
  5. Jennings, D. & Rudolph, T. (2010). Comment on "Quantum solution to the arrow-of-time dilemma". Physical Review Letters 104(14), 148901. 10.1103/PhysRevLett.104.148901 · Checked 15 Oct 2026
  6. Lees, J. P., Poireau, V., Tisserand, V., Garra Tico, J., Grauges, E., Palano, A. et al. (BaBar Collaboration) (2012). Observation of time-reversal violation in the B⁰ meson system. Physical Review Letters 109(21), 211801. 10.1103/PhysRevLett.109.211801 · Checked 15 Oct 2026
  7. Ellis, G. F. R. (2013). The arrow of time and the nature of spacetime. Studies in History and Philosophy of Modern Physics 44(3), 242–262. 10.1016/j.shpsb.2013.06.002 · Checked 15 Oct 2026
  8. Mlodinow, L. & Brun, T. A. (2014). Relation between the psychological and thermodynamic arrows of time. Physical Review E 89(5), 052102. 10.1103/PhysRevE.89.052102 · Checked 15 Oct 2026
  9. Vaccaro, J. A. (2016). Quantum asymmetry between time and space. Proceedings of the Royal Society A 472(2185), 20150670. 10.1098/rspa.2015.0670 · Checked 15 Oct 2026
  10. Micadei, K., Peterson, J. P. S., Souza, A. M., Sarthour, R. S., Oliveira, I. S., Landi, G. T. et al. (2019). Reversing the direction of heat flow using quantum correlations. Nature Communications 10, 2456. 10.1038/s41467-019-10333-7 · Checked 15 Oct 2026
  11. Hemmo, M. & Shenker, O. (2022). The second law of thermodynamics and the psychological arrow of time. The British Journal for the Philosophy of Science 73(1), 85–107 (published online 2019). 10.1093/bjps/axz038 · Checked 15 Oct 2026
  12. Callender, C. (2001, substantively revised 12 June 2026). Thermodynamic asymmetry in time. Stanford Encyclopedia of Philosophy. Source · Checked 15 Oct 2026

Page reviewed 15 Oct 2026

Source

Open, as listed by Wikipedia · Unsolved problems in physics. Checked 08 Oct 2026.

The status is the source’s. Papex does not decide whether a problem is solved.

Records on this problem

No public record addresses this problem yet.

More in Physics