Epistemology
Justified Belief and Legal Proof
Mario Günther (MG) developed a theory that reduces justified belief to credence and utility. The notion of justified belief is suitable for a probabilistic account of legal proof that solves the problem of statistical evidence and unifies the standards of legal proof, such as beyond a reasonable doubt and preponderance of the evidence.
​
-
"Probability of Guilt" (Canadian Journal of Philosophy)
-
"Legal Proof Should Be Justified Belief of Guilt" (Legal Theory) refines the notion of justified belief, in particular by a decision theoretic argument for finding the appropriate credence thresholds for justified belief.
​
MG argued that a notion of epistemic sensitivity is better suited than a factive one to discern individual from merely statistical evidence.
​
-
"Epistemic Sensitivity and Evidence" (Inquiry)
MG provided a Bayesian Justification for the scenario approach to legal proof—an approach that is applied in the Netherlands.
​
-
"A Bayesian Justification for the Scenario Approach to Legal Proof" (CEUR Workshop Proceedings)
Learning Conditional and Causal Information
Mario Günther (MG) developed a theory of learning conditional and causal information. The theory fills a lacuna in Bayesian Epistemology by satisfying the desiderata postulated by the philosopher and cognitive psychologist Igor Douven. Notably, the theory resolves Bas van Fraassen’s Judy Benjamin Problem.
​
-
"Learning Conditional Information by Jeffrey Imaging on Stalnaker Conditionals" (Journal of Philosophical Logic)
2022 Wolfgang Stegmüller Award
This paper generalizes David Lewis's updating method called "imaging" to "Jeffrey imaging" and proves a theorem which shows that the generalization has the desirable properties for a theory of learning conditional information.
-
"Learning Conditional and Causal Information by Jeffrey Imaging on Stalnaker Conditionals" (Organon F) extends the theory to the learning of causal information.
​
Afterwards, MG argued that orthodox Bayesians, whose learning rule is compatible with Jeffrey conditionalization, still don't learn from conditionals.
​
-
"Bayesian's Still Don't Learn from Conditionals" (Acta Analytica, with Borut Trpin)
​
