falsifiability

Karl Popper’s idea of falsifiability states that in order for a theory to be a scientific one, it need not be verifiable, but falsifiable - meaning that one can never prove a scientific theory, but one can prove it wrong.

Something’s always bothered me about this though. It seems too “unsymmetrical”. By that I mean that for any theory which you prove wrong with only one instance, are you not proving the opposite theory to be true? You’d have to be, wouldn’t you? For example, if I had a theory that all snails carry shells on their backs, and one day I found a snail without a shell, my theory would be proven wrong. But suppose I had a competitor whose theory was that not all snails carry shells on their backs. If the one shell-less snail I found proves my theory false, wouldn’t it prove his theory true by the same token?

I’m guessing that Popper took cases like these into account. By “scientific theory”, he must have meant theories stated in a positive manner - that is, theories stated as what the case is rather than what the case is not. I don’t know Popper’s work that well. Would anybody be able to shed some light on this?

No. The reason is that scientific “theories,” to which that name properly applies, are much more complicated than “all snails carry shells.” That’s a hypothesis, not a theory. Theories don’t exist in simple binary form like that, such that there are only two possible and mutually-exclusive theories, and to disprove one is to prove the other.

An example of a theory is Newton’s theory of gravitation. This was not a simple statement. It held that all bodies which have mass attract one another, as if a force were acting at a distance to pull their centers together, this force being proportional to the product of their masses divided by the square of the distance between their centers of mass.

Now, one could argue that the converse to this is, “all bodies with mass do NOT attract one another,” etc., but this would not be a theory accounting for gravity, it would merely be the statement that, in some part or other, Newton was wrong. And so, a demonstration that Newton’s theory was incorrect (which it was) would not amount to proof of anything except the mere statement that Newton’s theory was incorrect, which is not itself a theory of gravity.

One may identify two theories (at least) competing with Newton’s, one of which predated him and the other came long after him. The earlier theory was Aristotle’s view of all matter as having a proper resting place according to the proponderance of its elemental composition: earth in the lowest place, then water, then air, then fire. Each substance tended to move toward its resting place, which is why solid objects (preponderantly earth) fall when dropped, but smoke (fire and air) rises. The later theory is Einstein’s gravitational field theory, which holds that mass warps space-time, so that bodies in inertial motion in a gravitational field follow normal intertial paths that are curved by the nearby presence of other masses. But neither of these theories follows automatically from the disproving of Newton’s theory.

Does that help?

Actually, yes it does. I seems to have failed to take into account that theories explain phenomena. They aren’t just descriptive statements. So if one explanation fails, you can’t just take the inverse of the theory and use that as a substitute explanation - you would just be describing something that needs an alternate explanation.