SOPhiA 2019

Salzburgiense Concilium Omnibus Philosophis Analyticis

SOPhiA ToolsDE-pageEN-page

Programme - Talk

Meta-induction in a decision-theoretic setting
(Epistemology, English)

Meta-induction is a class of meta-inferences, which was developed by Gerhard Schurz (2008) in order to solve Hume's famous problem of induction. There are variants of meta-inductive inferences which are access optimal in all possible worlds. 'Access optimal' means, that there are no accessible alternative inferences, which are more successful than meta-inductive ones. This doesn't imply, that meta-inductive inferences always offer correct predictions. Still, there aren't accessible methods which have a higher predictive success-rate than these meta-inductive inferences and that, so it seems, gives us reasons to use these meta-inductive inferences. These properties of meta-induction can be proofed a priori and illustrated with computer simulations, so it won't be a circular inductive justification of induction. Instead, there is an a priori justification for these meta-inductive inferences and we can justify induction with these inferences, if we accept the a posteriori premise that inductive methods have been more successful than non-inductive methods in the past. So meta-induction offers a solution for some interpretations of Hume's problem of induction.

I want to apply the meta-inductive framework to a decision-theoretic setting and show what might be problems of adding utilities to the framework and how this affects the justification of induction by meta-induction.

First, I want to discuss several technical options of enriching the meta-inductive framework by decision-theoretic elements. Second, I want to present different interpretations which make certain embeddings plausible. My last step will be to outline for each option, whether the optimality of meta-induction can be preserved in the enriched setting, or whether it fails to preserve the optimality and its usefulness as a justification of induction.


Chair: Madelaine Angelova-Elchinova
Time: 10:40-11:10, 20 September 2019 (Friday)
Location: HS E.002

Maximilian Benito Seubold 
(Heinrich-Heine-Universität Düsseldorf, Germany)

Currently I am a student and student assistant at the university of Düsseldorf. I am interested in epistemology and philosophy of science. I am especially interested in different types of inferences and their justifications.

In my free time I like being outside, riding bicycles and doing sport.

Testability and Meaning deco