Page:An analysis of religious belief (1877).djvu/683

 *rience cannot be put in place of necessity as a test of truth. In fact, every argument drawn from the past fallibility of the test of necessity might be retorted with tenfold force against the test of experience. Observation has constantly misled mankind, and thousands of alleged facts, accepted upon imagined experience, have been disproved by more accurate examination. Observation and reasoning combined (as they often are) are exposed to the double danger of false premises and false inferences from true premises; while the addition of an element of testimony (a circumstance common in scientific inquiries) exposes every conclusion to a threefold possibility of error. Human beings are no more exempt from the possibility of mistaken science than from that of hasty metaphysics. But as, in matters of physical research, we do not discredit the use of our eyes because their perceptions are sometimes inaccurate, so in matters of metaphysical inquiry we need not discredit the use of our minds because their apparent intuitions are now and then fallacious. In the one case, as in the other, the proper course is not to cast contempt upon the only instruments of discovery we have, but to apply those instruments again and again, omitting no precaution that may serve to correct an observation and to test an argument. But when we have done our utmost to attain whatever certainty the nature of the subject permits, we cannot reasonably turn round upon ourselves and say: "True, my eyes assure me of this fact, but human eyes have erred so often that I cannot accept their verdict;" or, "No doubt my mind forces this conclusion upon me as a necessity of thought, but so many assumed necessities have turned out not to be necessary at all that I must refuse to listen to my mind:" for this is not really the caution of science, but the rashness of philosophic theory. For we can have no higher conviction than that arising in a necessity of thought. Nothing can surpass the certainty of this. Grant that we may yet be wrong: we can never know it, and we can have no reason to think it. To oppose to a necessary belief such a train of reasoning as this:

Necessary beliefs (so-called) have often proved false:

This is a necessary belief (so-called):