only in our experience. But long before the half of the tests had been carried out the probability of our being in error would have become infinitely small, and we should accept as virtually proved the existence of a real serpent as the single adequate explanation of our various experiences. In quite the same way we prove beyond a reasonable doubt the existence of other persons. We cannot directly experience the existence of their minds, and we cannot be absolutely certain of them, but the totality of their acts of behaviour form a convergent system quite of the type we have been discussing, and the alternatives that confront us are: (1) the real existence of other minds similar to our minds to the extent that their behaviour is similar to our behaviour; and (2) a miraculous combination of independent or chance factors which perfectly counterfeit the effects of such minds. Between these two alternatives the probabilities in favour of the first are overwhelming.
We have now surveyed the four principal arguments that have been brought forward by sceptics in support of their negative contention that none of the five sources of knowledge is adequate to provide a reasonable basis for human beliefs. Our conclusion is that the historical, the dialectical, the physiological, and the psychological arguments do indeed show that the human mind is unable to attain absolute certainty in any field of inquiry. But between this certainty maintained by the extreme anti-sceptics, or, as Kant called them, the "dogmatists," and the blank indifference of complete doubt maintained by the sceptics, there exists the intermediate realm of the probable. And as our position with respect to the historical and dialectical arguments was the modestly optimistic claim of ignoramus as against the sceptic's despairing ignorabimus, so with respect to the physiological and psychological arguments our claim has been that the main systems of belief, such as those resting on the Uniformity of Nature, and on the reality of a single set of physical objects underlying the varying perceptions of individuals--maintained by common sense and by natural science--possess a very high