|
Qu’est-ce qu’une constante logique ?| old_uid | 1102 |
|---|
| title | Qu’est-ce qu’une constante logique ? |
|---|
| start_date | 2006/04/24 |
|---|
| schedule | 17h30-19h30 |
|---|
| online | no |
|---|
| summary | The standard semantic definition of logical consequence rests on the distinction between logical and extra-logical expressions. How should we draw the line? According to the so-called Tarski-Sher thesis, an operation is logical iff it is invariant under permutations or bijections. The motivation for the thesis comes from a conceptual analysis according to which permutation invariance adequately captures the generality and the formality of logic. Unfortunately, the thesis has been widely criticized for counting too many operations as logical and yielding a trivial identification of logic and mathematics. The problem is therefore both to identify what has gone wrong in the conceptual analysis and to provide an alternative and stronger formalization of logicality in terms of invariance under a given class of transformations.
I will discuss the classes of transformations that could be chosen. More specifically, I will suggest that pushing further the analysis of what it is for logic to be general and free from content leads to two natural constraints on transformation classes, namely closure under definability and absoluteness. On this basis, I will argue that invariance under potential isomorphism is a good candidate to capture logicality. |
|---|
| responsibles | Pataut, Dubucs, van Atten |
|---|
| |
|