This review appeared in International Studies in Philosophy , vol. 22, no.  3 (1990), pp. 149-150.

NORMAN SWARTZ, The Concept of Physical Law, New York and Cambridge, Cambridge University Press, 1985, xi + 220 pp.

A necessary condition for a proposition's being a physical law  is that it be a contingently true universal conditional, (x)(Ax ⊃ Bx), with purely descriptive predicates. But is this sufficient? Are all such propositions physical laws; or can we distinguish between mere accidental regularities of the above sort, which are not  laws, and propositions which also have the additional modal property of nomic necessity ?

This book is a relentless defense of the "Regularity Theory" of physical law against the "Necessitarian Theory" (or theories.) In a manner reminiscent of Berkeley's attack on the notion(s) of material substance in Three Dialogues Between Hylas and Philonous, Swartz dismembers the various arguments in favor of various forms of "natural necessity" or "nomicity" as a further characteristic of physical laws. He finds no coherently defensible difference between mere regularities and laws – nothing over and above the conditions laid out in the first sentence of this review. His hunt takes him through all the usual places: possible world theory, counterfactuals, potentialities, coins in your pocket, etc.

For example (alas, all that is possible in a brief review), the ability to support counterfactual conditionals is sometimes offered as a mark of the "nomicity" being sought. Swartz's strategy here is to agree that some regularities do  get used to support counterfactual conditionals – those regularities which are believed to be "pervasive," i.e., independent of various conditions. This is reminiscent of Brian Skyrms's idea that "resiliency" (roughly, invariance under changes of physical circumstances) can distinguish between laws and mere regularities. (See my review of Skyrms's book, Causal Necessity , in this journal, Vol. xiii, 1981, No.2.) But Swartz is interested in the metaphysics of science, not its epistemology. We seek  pervasive (unconditional) regularities, and our current best guesses as to what they are are what he calls "scientific  laws." And we do use these to support counterfactual conditionals. Swartz is interested in the way the world is  – its physical  laws." Physical laws, however, are true  (unconditional), whether we know them or not. Pervasiveness (unconditionalness, resiliency) is in any case not a modal property – it cannot be identified as the sought-after nomic necessity. As part of the epistemology/methodology of science, pervasiveness can be sought by regularists and necessitarians alike, but not in order to lay before us some additional characteristic of laws: their physical necessity.

As in the case of Berkeley's attack on substance, one may still be left with the nagging feeling that there is something more than (unconditional) regularity in physical laws. There is still the intuition that there is a "mustness" that has been left out of the analysis. Swartz locates the origin of this intuition in a now defunct third view of physical law: that laws are prescriptions – edicts issued by God. There is still a tendency today to view laws as prescriptions, but without a prescriber – a remnant of the God-as-law-giver view. Laws prescribe how events must  unfold. The world rolls on in accord with  laws. Events are as they are because of  physical laws (which they obey.) Laws determine  the way events are related. It is hard to shake this view and replace it with the view that the events come first, as it were – that physical laws are determined logically) by the array of singular facts of the world, rather than the laws physically determining  (how?) the constellation of singular facts.

I think we have here a picture  (in Wittgenstein's sense), a ghost that is difficult to exorcise (like viewing space in a Newtonian absolutist fashion as a dark empty room with flat walls, or Berkeley's Hylas viewing substrate  as something holding together the properties of things, without which things would disintegrate.)

Necessitarians ought to read this book. But I suspect that few of them will be convinced. Finally, I will just mention Swartz's discussion of predictability, the uniformity of nature, and determinism (in which be embeds a defense of compatibilism in the free will vs. determinism controversy); his discussion of statistical laws, and a brief but interesting treatment of miracles.