人类理解研究

Verse 2 When we ask, "What is the nature of all our inferences about actual things?" And the proper answer seems to be that they are based on the relation of cause and effect. If we then ask, "What is the basis of all our inferences and conclusions about that relationship?" then we can again reply in one sentence that it lies in experience. But if we continue to indulge our penchant for careful inquiry and ask, "What is the basis of all conclusions drawn from experience?" Then this contains a new question, and one that is perhaps even more difficult to solve and to explain. The two propositions below are not in any way the same. One proposition says: "I have seen that such an object always has such a consequence accompanying it"; the other proposition says: "I have foreseen that other objects similar in appearance will also have similar results accompanying them." I also admit that the latter proposition can be correctly deduced from the former, and I know that in fact it is always so deduced. But if you insist that this inference follows from a string of inferences, then I would like you to indicate that string of inferences. This connection between these propositions is not intuitive, and a medium is needed here to enable one to introduce this conclusion-if it is introduced by means of inference and argument. What exactly this medium is, I admit, that is beyond my knowledge. Therefore, if people claim that it exists and that it is the source of all conclusions in terms of practical things, then they must be responsible for indicating this medium. All inferences can be divided into two categories, one inference is demon A stra-tive, which is concerned with the relation of ideas, and the other inference is contingent, which is concerned with actual facts or existence. In our present case, it seems clear that there is no demonstra-tive argument; for natural paths are changeable, and an object, even if it seems to be the same as the one we have experienced, can give rise to different or opposite results; there is no contradiction in these things. So if there are arguments that make us trust past experience and use it as a criterion for our future judgment, they must be only contingent, and must concern only actual things and real beings - according to our classification above. But if our interpretation of such inferences is solid and satisfactory, we must see that these arguments, as described here, do not exist. We have already said that all arguments for actual existence are based on causality; that all our knowledge of this relation comes from experience; and that all our empirical conclusions are based on the assumption that "the future must be compatible with the past". Therefore, if we try to prove this last hypothesis by arguments of cogency, or of real existence, we are clearly going back and forth, and taking for granted what is being argued.