Fundamental Physics, Partial Models and Time’s Arrow

This paper explores the scientific viability of the concept of causality—by questioning a central element of the distinction between “fundamental” and non-fundamental physics. It will be argued that the prevalent emphasis on fundamental physics involves formalistic and idealized partial models of physical regularities abstracting from and idealizing the causal evolution of physical systems. The accepted roles of partial models and of the special sciences in the growth of knowledge help demonstrate proper limitations of the concept of fundamental physics. We expect that a cause precedes its effect. But in some tension with this point, fundamental physical law is often held to be symmetrical and all-encompassing. Physical time, however, has not only measurable extension, as with spatial dimensions, it also has a direction—from the past through the present into the future. This preferred direction is time’s arrow. In spite of this standard contrast of time with space, if all the fundamental laws of physics are symmetrical, they are indifferent to time’s arrow. In consequence, excessive emphasis on the ideal of symmetrical, fundamental laws of physics generates skepticism regarding the common-sense and scientific uses of the concept of causality. The expectation has been that all physical phenomena are capable of explanation and prediction by reference to fundamental physicals laws—so that the laws and phenomena of statistical thermodynamics—and of the special sciences—must be derivative and/or secondary. The most important and oft repeated explanation of time’s arrow, however, is provided by the second law of thermodynamics. This paper explores the prospects for time’s arrow based on the second law. The concept of causality employed here is empirically based, though acknowledging practical scientific interests, and is linked to time’s arrow and to the thesis that there can be no causal change, in any domain of inquiry, without physical interaction. I. Temporal symmetry and fundamental physics It is prevalently held that the fundamental or primary laws of physics are symmetric in time and all-encompassing. Part of the plausibility of this under This paper was first presented at the 2015 Conference on Model-based Reasoning held in Sestri Levante, Italy; and it will be published by Springer, 2016 in the Proceedings of the Conference edited by Lorenzo Magnani. I would like to thank Professor Magnani and the anonymous referees for helpful comments. 1. See, e.g., Davies, Paul 1977, The Physics of Time Asymmetry, p. 26, who puts it this way: “All known laws of physics are invariant under time reversal,”—though noting the singular exception of processes involving K-mesons and the weak force. Cf. Greene, Brian 2004, The Fabric of the Cosmos, p. 145: “...not only do known laws fail to tells us why we see