ABSTRACT. — In Carnapian explication, we engineer new conceptual schema in order to improve our epistemic lots in reasoning about the world. Such explications are satisfactory, inter alia, to the extent that they provide us with precise concepts that can be fruitfully applied toward achieving some epistemic end. To defend “conceptual engineering” as a useful methodology, it is thus crucial to demonstrate that there are meta-evaluative tools allowing us to test and compare meaningfully the fruitfulness of candidate explications. This talk discusses ways in which computer simulations can fill this role. As a case study, I compare Bayesian and Explanationist logics. Such inductive logics offer explications of salient epistemic concepts and judgments. And they are developed with the aim of improving our abilities to reason well with uncertainties. Bayesian and Explanationist logics overlap both in the concepts that they are meant to explicate and in the epistemic aims that they are engineered to help us attain. The question thus arises as to which logic is more fruitful. I argue that computer simulations are extremely helpful in helping us answer this question. Computer simulations demonstrate that, across different cases, both Bayesian and Explanationist logics can offer superior explications. Which logic is best in any specific case depends on certain factors. These simulations enable us to gauge the respective merits of our candidate explicata. But this work also reveals that simulations can help us in conceptual engineering in more subtle ways. For example, in the cases covered in this talk, simulations clarify for us senses in which seemingly competing, newly engineered conceptual schema can all be appropriately adopted, relative to distinct explicanda and epistemic goals. 

ZOOM INFO

• Meeting ID: 892 5895 0975
• Password: ACEW21
• Invite link: here