In an oft-quoted 1912 essay entitled “On the Notion of Cause,” Bertrand Russell spoke for many twentieth-century philosophers in suggesting that classical physics had discovered the universe to be an acausal place: “The law of causality, I believe, like much that passes muster among philosophers, is a relic of a bygone age, surviving, like the monarchy, only because it is erroneously supposed to do no harm.” In Russell's austere view, the laws of classical physics are merely relations holding among various physical quantities at various times. The laws do not portray certain of these quantities as causes of certain others. To regard acceleration, for example, as caused by force and mass would be to regard force as like muscular push or pull. The bare equations of physics are cleansed of any such anthropomorphic vestige and (generalizing Hertz's famous remark regarding Maxwell's electromagnetic theory) a given physical theory is just its system of equations.

Many philosophers since Russell's day have found many reasons to agree with Russell's view. Ordinarily, we think of causal relations as asymmetric (if c causes e, then e does not cause c), as favoring a certain direction in time (the future does not cause the past), and as presupposing a distinction between causes and background enabling conditions (as when the lighting of a match is caused by its being struck, not by its being surrounded by oxygen). Many philosophers have argued that the fundamental laws of non-quantum physics exhibit none of these features. Causal relations, then, would seem not to belong to the world that physics aims to describe. Rather, causal relations seem to be projected onto that world by us—perhaps as a reflection of the limited range of observations we can make, or perhaps because we inevitably approach the world with the intention of bringing about various outcomes we desire.

In his exciting new book, Mathias Frisch opposes the view that causal relations play no role in physics. According to Frisch (a philosopher of physics at the University of Maryland, College Park), causal relations, far from doing no harm, are an essential part of the way that non-quantum physics represents the world. Frisch argues that many of the inferences routinely made in physics would be impossible without causal assumptions. For instance, we observe points of light on various nights and infer that these points are the light emitted by stars. We could not draw this conclusion by feeding our nighttime observations into the laws of light propagation, since that calculation would require us to have data along an entire cross section of a forward light cone centered on the star being inferred. Rather, our inference exploits the idea that correlations among events standing in no direct causal relation are probably explained by the existence of a common cause of those events; in this example, the correlations are among our observations of the night sky at different times and from different terrestrial locations, and the star is the putative common cause.

Whether this argument from Frisch demonstrates that causal assumptions play an essential role in physics depends on whether the inference to the star's existence, for example, depends on causal assumptions. The inference clearly depends on the assumption that certain correlations would be extremely unlikely in the absence of a star. It is not yet evident to me, however, whether this assumption would be unwarranted in the absence of specifically causal assumptions.

Similarly, Frisch argues that although Maxwell's equations can be solved by both retarded and advanced potentials, the latter are routinely rejected on causal grounds: the correlated events required for incoming radiation from different directions of space to converge coherently onto an antenna, for example, lack a common cause. Frisch's argument here seems to be twofold. First, to reject advanced solutions as improbable, without any causal rationale for doing so, would be unwarranted. Second, the fact that the delicately correlated initial conditions required for advanced solutions are so unlikely can be explained by causal considerations (namely, that the initial states of a system's components are distributed randomly when they have no common cause in their past). But without these causal considerations, the requisite correlation's unlikelihood would have no explanation. It would simply be a brute fact—an unexplained explainer. Frisch includes an extensive, historically rich discussion of the debate between Einstein and Ritz on the arrow of radiation.

Of course, if one were inclined to regard physics as not being in the business of describing the world's causal relations, then (since many ordinary scientific explanations appeal to causal relations) one might be inclined to conclude that the relations of explanatory dependence traced by scientific explanations are likewise not objective features of the world that physics aims to describe. On this spare picture of the physical world, it might not be so implausible for the unlikelihood of certain sorts of correlations to have no explanation (and to be responsible for the usefulness of causal concepts in representing phenomena).

Alternatively, an “initial randomness” assumption might have an explanation other than that the initial states of a system's components have no common cause. Perhaps the “initial randomness” assumption could be explained instead by certain features of the very early universe. Frisch explores and critiques recent attempts to elaborate such a neo-Boltzmannian view of the origin of thermodynamic asymmetry.

Frisch surveys and responds to a wide range of arguments drawn from the recent literature in the philosophy of physics. His book will serve as an excellent introduction to this literature for new readers as well as a valuable, original perspective on that literature for those already familiar with it.

Marc Lange is Theda Perdue Distinguished Professor of Philosophy at the University of North Carolina at Chapel Hill. He does research on the philosophy of science and is the author of An Introduction to the Philosophy of Physics: Locality, Fields, Energy, and Mass (Blackwell, 2002).