but one based on economy; "relativity," however, would characterize it also. 

There seems to me no question but that the representatives of the view we have been criticizing did in fact take the economy principle as their guiding light, whether explicitly or implicitly; I have therefore already assumed above that on the relativity view there are purposive grounds which determine the selection of protocol statements, and I asked; Can we admit this? 

I now answer this question in the negative. It is in fact not economic purposiveness but quite other characteristics which distinguish the genuine basic statements. 

The procedure for choosing these statements would be called economic if it consisted say in conforming to the opinions (or "protocol statements") of the majority of investigators. Now it is of course the case that we do not doubt the existence of a fact, for example a fact of geography or history, or even of a natural law, when we find that in the relevant contexts its existence is very frequently reported. It does not occur to us in those cases to wish to investigate the matter ourselves. We acquiesce in what is universally acknowledged. But this is explained by the fact that we have precise knowledge of the manner in which such factual statements tend to be made, and that this manner wins our confidence; it is not that it agrees with the view of the majority. Quite the contrary, it could only arrive at universal acceptance because everyone feels the same confidence. Whether and to what extent we hold a statement to be corrigible or annul able depends solely on its origin, and (apart from very special cases) not at all upon whether maintaining it requires the correction of very many other statements and perhaps a reorganization of the whole system of knowledge. Before one can apply the principle of economy one must know to which statements it is to be applied. And if the principle were the only decisive rule the answer could only be; to all that are asserted with any claim to validity or have ever been so asserted. Indeed, the phrase "with any claim to validity" should be omitted, for how should we distinguish such statements from those which were asserted quite arbitrarily, as jokes or with intent to deceive? This distinction cannot even be formulated without taking into consideration the derivation of the statements. So we find ourselves once more referred to the question of their origin. Without having classified statements according to their origin, any application of the economy principle of agreement would be quite absurd. But once one has examined the statements with respect to their origin it becomes immediately obvious that one has thereby already ordered them