| |
|
Picasso Triggerfish I think that Dawkins is mentioned in the book, albeit only in the footnotes. I cannot be sure, because my fiance managed to leave my copy somewhere in Suffolk where, no doubt, it is starting an epidemic as we speak.
More generally the stuff further up the thread about something being 'valid' even if it is not 'scientific', and questions about whether it could be 'tested'.
A.J.Ayer and the Vienese school of philosophy did a lot on the philosophy of language and the test of meaning being whether a statement was verifiable. This had quite a lot of impact on the whole Popper-ian philosophy of science in turn. But there was really one major problem with the 'test meaning by means of verification' - how do you test that statement? And if that statement is meaningless because there is no method of testing it, how can the whole concept work?
Speaking of Popper - I think his philosophy of science might have something to add to this discussion. His view (and this is a MASSIVE simplification, I know, don't shout at me) was that a scientific theory is just that - a theory which can be tested against data. The moment you can prove something logically it goes beyond being a scientific theory and becomes a logical necessity, and in the physical world there aren't many of those around because of the necessity to have a priori assumptions which render the 'necessity' suddenly only theoretical. On the other hand, if you cannot test it against something it may well be a theory, but it doesn't really count as scientific because that term implies a degree of 'testability' (and particularly, objective 'testability').
The way you test a scientific theory is twofold:
(i) Check it logically - make sure it doesn't have mathematical errors and the like; and
(ii) Check it against the physical world - make sure that experimental data conforms with it, and/or that it can predict experimental data. (The same thing theoretically, but the latter is sort of better because it tends to suggest that you have some sort of fundamental understanding of the problem rather than a description of the situation. The wider in application a theory is, the more useful a theory it is, the 'better' a scientific theory it is.)
The big problem with science is that you never prove a theory is the case - you just don't disprove it and, in the process, eliminate situations where you think it won't work and the theory becomes more and more useful. Obviously, there is a stage at which a theory becomes so unlikely to be proved false that it becomes thought of as true in the sense of 'proven absolutely' but all that it ever is really is 'proven on the basis of the evidence'.
{I suspect that many times there is a paradigm shift in science it is a case of persuading scientists to remember this last thing, and accept a new theory as being better. If only they remembered that the old one was not Gospel, things might be easier for the innovator.)
Finally, getting to the point, if during the test the theory doesn't work you have three options:
(i) Discard the theory.
(ii) Check the data
(iii) Amend the theory - possibly so it encompasses the data, or possibly so it excludes the sort of situation where the data is provided.
As to the Tipping Point theory - it probably doesn't come under my understanding of Scientific Theory because it is too broad-brush and therefore testing it would either be almost impossible or would result in so many false results that one would be constantly refining it until it became either unrecognisable or entirely unusable.
Rather, it is what I would call a Social Science Theory - useful, because it gives us a handle on certain complicated structures that are not easily susceptible to testing or analysing, but at the same time not so strong as to be anything more than a loose guide to the future. |
|
|
