I don't wish to imply that anyone is asking silly questions. Philosophy of science is important, and non-trivial, but not very pretty.
> I really don't see how one could ever be completely certain that a given
> scientific theory is false. As far as I know, for instance, noone has ever
> "falsified" the basic Ptolemaic understanding of the universe. Can't you
> just keep adding epicycles until your theory fits the data? This could get
> pretty clumsy after a while, admittedly, but I don't see how this is
> relevant unless theoretical efficiency and economy get made into criteria of
> truth. Is a basically pragmatist understanding of truth being espoused
> here? Which gets me to my second question:
A theory that becomes a hashwork of exceptions and special rules designed to make it fit the data loses its predictive power. It is impossible to ever falsify it. Since falsification is the criteria I'm using here, that means I would reject the theory. This is the same problem people ought to have with the "God of the Gaps" theory of religion. It's a moving target, which makes it impossible to falsify.
We can reject a set of statements, which we'll call the Ptolemaic theory, by falsification. You can add a new statement to it, but then it isn't the same set of statements anymore. I'm not sure that you can't completely falsify epicycles - I see no way to explain the trajectory of the Mars Pathfinder probe using epicycles.
However, I can accept that in principle we don't know for certain that epicycles don't exist (proving the negative is very hard) however, we know that some specific epicycle based schemes are false, and that some non-epicycle based schemes aren't. We can use some schemes to make accurate predictions, while with epicycle theory we find ourselves constantly needing to add some new axiom every time we find something different. Certainly that kind of repeated falsification, while never proving with certainty that a theory is wrong, makes it difficult to support.
Falsification can't ever tell us absolute positive truth. There are always a large number of different theories that can fit all the data, and things that seemed eternally true (like Newton's laws) prove not to be. We use the theories that have proven the most successful in resisting falsification, but that doesn't mean that tomorrow they won't be falsified.
> These are aesthetic criteria, aren't they? Do we have some kind of
> epistemological basis for saying that truth resides a priori in the general,
> the brief, and the specific, other than that accepting the idea makes life
> easier for us?
Where Popper is mistaken, in my opinion, is failing to look at the process by which theories are proposed and considered, which involves things like aesthetic criteria, sociology and politics.
He does look to some extent at this, by using criteria like generalisability and specificity and Occam's Razor to suggest which theories are the best ones to try, but I do think that's incomplete.
Scott Martens
---
Envoyé par Moncourrier.com Vous aussi obtenez gratuitement votre adresse électronique sur Moncourrier.com