Better Best Systems - Too Good To Be True

Abstract

Craig Callender, Jonathan Cohen and Markus Schrenk have recently argued for an amended version of the best system account of laws -the better best system account (BBSA). This account of lawhood is supposed to account for laws in the special sciences, among other desiderata. Unlike David Lewis's original best system account of laws, the BBSA does not rely on a privileged class of natural predicates, in terms of which the best system is formulated. According to the BBSA, a contingently true generalization is a law of a special science S iff the generalization is an axiom (or a theorem) of the best system relative to the set of predicates used by special science S. We argue that the BBSA is, at best, an incomplete theory of special science laws, as it does not account for typical features of special science laws, such as attached ceteris paribus conditions and the idealized character of law statements in these disciplines.1 This paper is fully co-authored.

How to Be Humean about Idealization Laws

“…Something like this problem has been observed in a more general assessment of Humean approaches to special science laws by Backmann and Reutlinger (2014) who identify both 'Lange's dilemma' (after Lange 1993) between laws' triviality and falsity, and 'Cartwright's dilemma' (after Cartwright 1983) between laws' vacuity and falsity. Both dilemmas, they claim, undermine the possibility of the best systems' approach accommodating special science laws, which are invariably idealising or require ceteris paribus clauses.…”

Do Statistical Laws Solve the ‘Problem of Provisos’?

“…The mentaculus 16 Lewis's (1994) best system account is not adequate for macro-probabilities (Callender and Cohen 2009: section 4.2). The discussion provided here generalizes to so-called "better best system accounts" as advocated by Cohen 2009, Schrenk 2007 and this volume, Unterhuber this volume (Backmann and Reutlinger 2014). According to Cartwright's dilemma, idealized laws are either literally false or they lack empirical strength, because they are vacuously true.…”

“…If the law is taken to be literally false, then it is not a good guide to the actual frequencies in the world; if the law is taken to be vacuously true then it does not represent actual frequencies at all and, hence, lacks empirical strength (cf. Backmann and Reutlinger 2014). 15 I anticipate a question at this point.…”

“…I agree with the critics that the concept of approximation is not helpful to improve our understanding of idealizations, because the notions of approximation and approximate truth stand in as much need of clarification as the notion of idealization does (Strevens 2008). Backmann and Reutlinger (2014) provide an in depth discussion of approximation and other accounts of idealizations. cal discussion of hypothetical frequentism including objections targeting the idealized assumptions of this view).…”

Berkeley’s Best System: An Alternative Approach to Laws of Nature

“…There is a closely related dilemma posed by Hempel (1988) and Lange (1993). 42 Backmann and Reutlinger (2014) provide a thorough catalog of attempts to deal with the problem of ceteris paribus laws within the context of the Best System Account. I find their critiques of Cohen and Callender's 'better best systems' account (2009) and Marcus Schrenk's (2007) proposed solution, among others, persuasive, and will not rehearse them here.…”