|
Degrees of Validity and the Logical Paradoxes| old_uid | 12341 |
|---|
| title | Degrees of Validity and the Logical Paradoxes |
|---|
| start_date | 2013/04/15 |
|---|
| schedule | 15h-17h |
|---|
| online | no |
|---|
| location_info | Grande Salle |
|---|
| summary | We traditionally accept a sharp distinction between deductive and inductive arguments. The former are taken to be undefeasible and thus we accept a principle that, roughly, goes as follows: (D) given a deductively valid argument with premises you believe, you should also believe the conclusion of the argument. In contrast, inductive arguments are defeasible and consequently we acknowledge that their validity admits of degrees: strong (“highly valid”) arguments can be “blocked” by equally strong or stronger (“equally valid” or “more valid”) arguments leading to the opposite conclusion. Accordingly, we accept something like this: (I) if it is important to make up your mind as to whether C or C, given an inductively valid argument A leading to C from premises you believe, then you should believe C, unless you know of another argument A' leading to C from premises you believe, such that A' is at least as valid as A. Now, logical paradoxes such as the Liar, Russell’s or Curry’s cast doubts on (D). For, at least prima facie, they are deductively valid arguments, but they can lead to any conclusion we please, either directly (as in Curry’s case) or via Ex Falso Quodlibet. Yet, in spite of them, we do not believe, e.g., that the moon is made of blue cheese. Traditional reactions to this problem question either grammar (e.g., by invoking type-theoretical distinctions) or logic (by regarding as not really deductive some inference rules that are traditionally taken to be deductive) so as to claim that the paradoxes fail to be deductively valid after all. Here I explore a different approach, based on the intuition that deductive arguments can be treated on analogy with inductive arguments, for in their case as well we can have two valid arguments that lead to opposite conclusions. For instance, Curry’s paradox can be used to show that the moon is made of blue cheese, but also to show that is not so made. This leads us to view deductive arguments as defeasible and suggests an extension to them of the notion of degrees of validity. In the light of this, (D) is dropped and (I) is generalized so as to cover both deductive and inductive arguments.
(Certains caractères ou symboles ne s'affichent pas correctement. Veuillez nous en excuser) |
|---|
| responsibles | Pataut, Dubucs, Panza |
|---|
| |
|