Philosophy

Lewis’s Best Systems Account (BSA) of laws was not much motivated by pragmatics. But recent commentary on his general approach to laws has taken a ‘pragmatic turn’. This was initiated by Hall’s defence against the ‘ratbag idealist’ which argued that best system accounts should be admired rather than criticised for the inherent pragmatism behind their choice of desiderata for what counts as ‘best’. Emboldened by Hall’s pragmatic turn, recent commentators have proposed the addition of pragmatically motivated desiderata to complement the canonical desideratum of pure pith. This, they hope, will allow their revisionary BSAs to respond better to various counterexamples against the original account. Here I problematise these revisionary approaches (though not the pragmatic turn itself). While actual laws may satisfy the newly proposed desiderata, they cannot be constitutive of laws. By comparison, the canonical desideratum appears to be relevant to explaining why and when the revisionary desiderata will reflect pragmatic features of the laws and better reflect the motives behind practitioners of fundamental physics. I therefore attempt to respond to the counterexamples on behalf of the Lewis’s ‘canonical’, purely pithy, BSA.

It is widely agreed that any theory of laws needs to be able to explain inferences from nomic facts to certain non-nomic facts, such as why ‘that it is a law that p’ entails ‘that p’. This challenge is often called the ‘inference problem.’ In this paper, we show that there is an additional ‘normativity problem’: theories of laws need to explain why nomic facts entail certain normative facts, such as what agents ought to do and believe. We will argue that solving the inference problem is not enough to solve the normativity problem and that neither non-Humean theories of laws nor the orthodox Humean best systems account can solve the latter. By contrast, we argue that recent pragmatist versions of Humean reductionism about laws have a ready solution to the normativity problem.

I will describe an account of laws of nature that is a descendent and improvement on Lewis’s Humean Best Systems Account (BSA) called “The Package Deal Account” (PDA). Lewis describes the BSA as a package deal since it proposes a package that includes both laws and chances.  But it  is a package in which the perfectly natural properties and space-time are not part of the package but are metaphysically prior to laws and chances. In contrast, the PDA is a package that includes fundamental properties, space-time, and laws and chances in one big package. Neither fundamental laws nor fundamental properties are metaphysically prior to the other. Because of this the PDA overcomes a number of objections to Lewis’s BSA and is an account that is closer to the practice of physics.

Can Super-Humeanism be legitimately regarded as a fully-fledged scientific realist view? Supporters of Super-Humeanism respond in the affirmative. Indeed, even though they take the properties figuring in the laws to be mere dynamical parameters that should not be interpreted literally, they have recently proposed a functionalist account to locate them in the primitive ontology and to vindicate their existence. I argue, on the contrary, that this functionalist strategy is dubious and that it is not enough to recover scientific realism. Instead, drawing on Varzi’s work on Humean Fictionalism, I propose that Super-Humeanism should be regarded as a fictionalist view. According to this view, the dynamical parameters appearing in the laws should be interpreted as face-value, but sentences about them should not be regarded as strictly expressing the truth. Since this fictionalist view only targets the dynamical structure of our theories, and is not properly ontological, it is still compatible with metaphysical realism and the primitive ontology approach. Moreover, given that it regards the dynamical structure of our theories as the best possible ‘fiction’ to make sense of reality, it also grounds and justifies our belief that reality should be structured with the laws that these theories propose. I will conclude that this view is not a scientific realist view. However, I will also suggest that the gap between such a fictionalist view and scientific realism is not as severe as one would think.

A pragmatist approach to laws, causation, counterfactuals and so on begins by following Wittgenstein’s advice to look for the use, not the meaning. So, how do we use the concepts of law, causation, counterfactuals and so on? In lots of ways. But one particular family of uses stands out, namely the ways in which we use statements about causal relations, law-statements, and counterfactual conditionals to block or permit attributions of responsibility for outcomes to agents. I show how to reductively analyze these natural modalities in terms of concepts pertaining to responsibility, and how to use a pragmatist theory of responsibility to generate pragmatist theories of the natural modalities.