Skip to ContentSkip to Navigation
About usFaculty of PhilosophyNewsEvents

Oldenburg-Groningen Workshop in Formal Epistemology

Workshop organized by the Department of Theoretical Philosophy


  • 11.00-12.00 Jakob Koscholke: Against relative overlap measures of coherence
  • 12.00-13.00 Jan-Willem Romeijn (partly joint work with Olivier Roy): Pooling as a Bayesian update: relations with Condorcet and Aumann
  • 13.00-14.00 Lunch break (Academiegebouw)
  • 14.00-15.00 Benjamin Bewersdorf: Bayesian Epistemology, Logical Omniscience and the Objects of Beliefs
  • 15.00-16.00 Jos Uffink: Phase transitions and infinite idealizations
  • 16.00-17.00 Mark Siebel: Needs-Based Justice Mathematised. Some Difficulties with Applying Jasso's Theory of Justice
  • 17.15 Dinner and drinks (Land van Kokanje)


Jakob Koscholke: Against relative overlap measures of coherence

Coherence is the property of propositions hanging or fitting together. Intuitively, adding a proposition to a set of propositions should be compatible with either increasing or decreasing the set’s degree of coherence. In this paper we show that probabilistic coherence measures based on relative overlap are in conflict with this intuitive verdict. More precisely, we prove that (i) according to the naive overlap measure it is impossible to increase a set’s degree of coherence by adding propositions and that (ii) according to the refined overlap measure no set’s degree of coherence exceeds the degree of coherence of its maximally coherent subset. We also show that this result carries over to all other subset-sensitive refine- ments of the naive overlap measure. As both results stand in sharp contrast to elementary coherence intuitions, we conclude that extant overlap measures of coherence are inadequate.

Jan-Willem Romeijn (partly joint work with Olivier Roy): Pooling as a Bayesian update: relations with Condorcet and Aumann

Opinion pooling offers a natural way to accommodate the opinions of others. This paper relates pooling to two other models of epistemically interacting agents, in particular Condorcet's jury theorem and Aumann's agreement theorem. Specifically, a Bayesian representation of pooling allows us to identify the trust parameter from pooling with the so-called truth-conduciveness of jurors. Furthermore, it can be shown that consensus based on iterated pooling coincides with a series of Bayesian updates in an Aumann structure. The iterated exchange of probabilistic opinions from pooling can thus be viewed as an approach to common knowledge. Particular attention will be given to the relations between categorical and probabilistic beliefs, as they show up in the two contexts.

Benjamin Bewersdorf: Bayesian Epistemology, Logical Omniscience and the Objects of Beliefs

The problem of logical omniscience is often considered to be a major challenge for Bayesian epistemology. It is argued that Bayesianism requires of an agent to be certain of all logical and mathematical truths and to be aware of all deductive inferences. Since no real agent could live up to these requirements, Bayesianism is dismissed as a theory too remote from reality to be useful. Bayesians often agree that Bayesianism demands more than can be expected of a real agent. They respond by either trying to contain the problem to a subset of logical and mathematical truth, by pointing out that Bayesianism is not a descriptive theory of real agents but a normative theory about how agents should ideally reason, or by arguing that Bayesianism is not a theory of agents at all, but rather a theory of pure logic. Whether or not these replies succeed in repelling the problem of logical omniscience I think that they already concede too much to the critic. Bayesian epistemology in itself does not imply anything about the question to which degree of belief an agent should believe a necessarily true sentence to be true. It does so only on the additional assumption that the objects of belief are sentences, an assumption denied by most Bayesians today in favor of a propositional account of belief.

Jos Uffink: Phase transitions and infinite idealizations

Macroscopic physical systems appear in several phases (solid, liquid, vapour, etc), depending on their temperature and pressure. A 'phase transition' is the usual term for the well-known phenomenon that such systems can change their phase at characteristic, sharply defined values of the temperature and pressure. A special kind of phase transitions (so-called second order or continuous phase transitions) is reserved for the behaviour around critical points. In the philosophy of physics, the efforts to explain phase transitions within statistical mechanics has recently attracted much attention. A variety of authors have argued that any such explanation of phase transitions must essentially employ models consisting of infinitely many particles, even though such models are idealized: actual macroscopic physical systems, after all, always contain a finite number of particles. A particularly strong case for this argument can be made for the explanation of critical phenomena by the renormalization group method by Wilson and Kadanoff. Some authors, like Bob Batterman, argue that this case shows the failure of the traditional Hempel-Nagel view on theory reduction. In this talk I will present and defend Craig Callender's analysis of the situation, and argue that the infinite idealization of physical systems is not essential for the explanation of critical phenomena and that this explanation is not problematic for a (suitably generalized) Hempel-Nagelian view on theory reduction.

When & where?

Thursday, 25 June 2015
Faculty of Philosophy, Room Beta

Last modified:15 June 2015 11.31 a.m.