Which Predicates, which Properties for Better Best Systems? Professor Markus Schrenk - Philosophy Departmental Research Seminar

Our departmental research seminar talks will take place on Wednesdays during term time from 15:00 to 16:30. 

Each week an email will be circulated to students with full details of about speaker and the talk taking place (including whether it will be online or in person).  Up to date details and information can also be found via the Philosophy Student SharePoint, events section (link here).

This week's speaker is Professor Markus Shrenk from the Heinrich-Heine-Universität Düsseldorf. Markus' research and teaching focuses on metaphysics, especially metaphysics of science (laws, dispositions, causation, modality), epistemology, philosophy of language and science. Markus is interested in Schopenhauer, Wittgenstein, logical empiricism, philosophy of art and analytic existentialism.

He give lectures and seminars within the above mentioned fields, and especially the recurring beginners lectures Philosophy of Language, Metaphysics, and Logic.

Which Predicates, which Properties for Better Best Systems?

Many advocates of the Better Best System Account (BBSA) of laws of nature suggest that Lewis-style best system competitions (BSA) can be executed for any arbitrary but fixed set of predicates/properties. This affords the possibility to launch system analyses separately for each of the special sciences (e.g. Cohen & Callender 2009, Author 2008). However, predicates/properties of these sciences can cause trouble. In Lewis’s original best system analysis, predicates refer to perfectly

natural properties only, i.e. we have canonical language-to-world fit. This possibility of a smooth transition is taken for granted in most formulations of the best system idea. Yet, when turning from Lewis’s BSA and his natural properties to the BBSA, the transition from the BBSA’s (non-natural, scientific) predicates to the respective properties is not straightforward. Indeed, it is surprisingly hard to find a semantics for the predicates of the special sciences that suits the purposes of the BBSA. In this paper, I will consider (i) semantic externalism, (ii) reference magnetism, and (iii) description theories of reference and will find all of these theories wanting for the purpose of the BBSA: The first tends to run against the Humeanism that is at the core of BBSAs, and the second is prone to collapse into Lewis’s BSA. While the third option is the most promising, it puts the BBSA in danger of being circular: in a description theory of reference it’s the predicates’ intensions, notably causal/ nomological roles, that belong to their semantic content and that fix their extensions. Yet, if this reference-fixing mechanism chooses the predicates’ extensions already for the nomological roles they fulfil then the BBSA seems to be obsolete: the BBSA was meant to deliver the nomological facts, yet, they are already present prior to the BBSA. I propose three answers to this problem: (1) While the intensional roles attached to predicates discern their extensions, it is only these extensions, i.e. the ‘naked’ properties void of such roles, that are systematised in BBSAs. I.e., the intensional roles are mere reference-fixers. The scientific predicates’ intensions can be treated as epistemic agents’ nomological conjectures. Not these ‘hypotheses’ count as the ‘real’ laws, only those nomological roles the BBSAs will deliver do. (2) Relatedly, as scientific progress shows, some of the predicates’ intensions (nomological role hypotheses) might well be wrong: scientists will probably err. That such discrepancies between prior conjectured roles and anterior ideal BBSA outcomes are likely diminishes the danger of predetermined outcomes. (3) The predicates’ intensions are, finally, most probably not exhaustive. The best system might well list some additional axioms or theorems involving global matters which are not yet captured locally by the predicates’ prior intensions. Still, the reference fixing roles of the predicates’ intensions do at the very least introduce a bias into the mosaic of objects the BBSA is supposed to systematise, however small it might be. For BBSAs, an innocently given mosaic is a myth.