Let me put it this way. Suppose you use the methods of mathematics to determine that 7 x 8 = 56, and you decide to test this by arranging physical things in 7

Let me put it this way. Suppose you use the methods of mathematics to determine that 7 x 8 = 56, and you decide to test this by arranging physical things in 7 rows of 8 and then counting them up. Suppose you try this, and you end up counting some other number, like 55. Why could this be? It could be that 7 x 8 is actually not 56, or it could be that your physical experiment was badly suited to test the principle (or badly executed). For example, perhaps you were putting drops of water in rows, and two of them combined into a single drop while you were putting down the rest of the drops. There are two things to say here. First, if the materials you have are highly volatile and it is incredibly difficult to rule out "things happening that you didn't account for" as the explanation when the results of an experiment appear to disagree with your theory, then it's efforts to advance the theory by empirical investigation are likely to be doubtful and fruitless. Note that if your materials are free human beings and you're observing their responses to various stimuli, then the condition in the preceding sentence applies very thoroughly. Second, even if [the reason the results of a physical experiment appear to disagree with the conclusions of a theory] is that the theory is incorrect, then it is always possible to confirm that the theory is incorrect by theoretical analysis. If you find that arranging physical objects in 3 rows of 4 and counting them up never yields 16, then you can verify by the methods of mathematics that 3 x 4 is not 16. And it is only once you've done this that you actually feel sure that the theory was wrong; until then, you always have the doubt that your experiment was badly designed, executed, or interpreted. (Imagine telling mathematicians that you were going to test one of their theorems by physical experiment. Or announcing that you had disproven a theorem by physical experiment. I think you would be laughed at. At any rate, they would want you to turn your results into a mathematical argument, with no reference to your experiment.) I believe it is with these things in mind that Mises and others say things to the effect that economic theory is not subject to empirical verification or falsification.