Page:Mind (Old Series) Volume 12.djvu/651

 638 NOTES. classify objects in a manner likely to prove of general use without a com- petent knowledge of these objects and their properties. Certain logicians may possess this knowledge, but if so it is not by their logical researches that they have acquired it. Other logicians who have confined their attention more especially to their own science do not possess it, and I do not believe that general directions as to Classification given by a man who has no special knowledge of particular objects are likely to be of much use to the man who possesses the information which his adviser lacks. Mr. Towry raises an important issue by his fourth objection. "Are there," he asks, "in nature, classes clearly marked off from each other, classes that ought to be sought for by us ?" I recognise fully the import- ance of the inquiry, but as a logician, and a logician only, how am I to answer it ? Does the law of gravitation hold good in the solar system only, or does it extend to the region of the fixed stars ? is likewise a very im- portant inquiry ; but is it one that a logician can be reasonably expected to answer ? And in like manner the question, Are there Natural Kinds or not ? is in my opinion clearly one which the physicist, not the logician, is called upon to answer. Then as to the answer, physicists are not agreed. Darwinism is now in the ascendant, but it cannot be said to have been universally accepted. According to this doctrine there are no such things as Natural Kinds separated from each other by impassable barriers; and whenever the line of demarcation between what I may call two adjacent kinds appears to be impassable, it is only because the intermediate members have perished in the struggle for existence. This, at least, is the current doctrine as regards the organic world. As regards the inorganic world, the doctrine of distinct chemical elements separated from each other by impassable barriers (at least so far as the simple elements are concerned) is still the current one ; but many persons are prepared to accept Mr. Lockyer's theory, that the supposed simple chemical elements are all allo- tropic forms of hydrogen. Mill would probably have treated coal, plum- bago and diamond as different Natural Kinds, but they are different forms of carbon, passing into each other under known physical conditions. He would probably have also treated heat, electricity and motion as distinct Natural Kinds, each possessing its own laws, but they can all be converted into each other by known processes. At all events, if Mill would not have treated these things as distinct Natural Kinds, he would have rested his refusal to treat them as such on purely physical grounds. Physically, it may be true that if a number of objects agree in certain qualities, we can predict their agreement in certain other qualities ; and the physicist may also believe with confidence that this agreement extends beyoncl what he has as yet discovered and that new points of agreement at present unknown will be discovered hereafter. But what right (as Mr. Towry very properly asks) has the logician to assume that any two objects agree in more respects than those in which they are known to agree 1 It is not for him to anticipate physical discoveries, and discoveries which it is quite possible may never be made. I do not concur with everything that Mr. Towry lays down in con- nexion with this subject; but I concur with him (if, indeed, he is disposed to go that length) in thinking that the doctrine of Natural Kinds, whether true or false, is entirely out of the province of Logic, and also in thinking that the doctrine in question has not been substantiated on satisfactory physical grounds. To avoid misconception, however, I add, that as judgments or proposi- tions usually contain assertions about classes, the logician is bound to explain briefly what classes are. But the problem of Classification is not to explain how men in fact classify objects, but how they can classify them