Version originale : In the discussion this far, three kinds of "truth" have been involved : - Religious truth, or other "dogmatic" truths, i.e. truth asserted by personal opinion or authoritative statement, having little or no support by evidence, nor any chance to ever have. This truth is object of belief or opinion, but there's little point in arguing on it. I respect those having such beliefs, but I cannot assign any knowledge value to this category of proposition, and tend to avoid it. The Gaia theory I criticised in #232* is a good example.
* #232 est la référence d'un autre mail. - Mathematical truth, i.e. truth inferrable by logic and mathematical proof. This should better be called consistency, and, although undisputable, has no relation to the real world. Proving that a theorem "holds" means only that the thesis is consistent with the axioms of the system. Choosing different axioms carries to different thorems. Which axioms fit best our real world is outside of the scope of pure math. A good example is euclidean vs. non-euclidean geometry. For validating axioms, and thus get a theory of the real world, see 3).
- Scientific truth, i.e. truth as attainable via the scientific method. In this case, perhaps a word like "likelihood" should be preferred. This kind of "truth" has to be supported by carefully analysed evidence, and only theories capable to make predictions are considered. Anyway, no theory is ever considered the ultimate one, and it is only valid within its range of "correctly" described and predicted phenomena. This is the only kind of knowledge we can reliably share, verify and build upon (it is probably not by chance that our ethics is not terribly different from the one of, say, ancient Romans, while our physics definitely is). When it comes to making assertions on the world that need to be "correct within some known range of uncertitude", I take in no account any other source of information.

Traduction : |