f LOGIC. 409 LOGIC. merely of the norms or standards to which our tliiuking must conform if its results are to be valid. Taking for granted then that we can have knowledge which is really valid, the question which logic as treated here asks and tries to answer is, According to what standards must the thinking processes work in order that our knowl- edge, so far as it is gained by thought, may be valid knowledge? In other words, What are the laws of valid thinking? But this question cannot be answered until we have first ascertained the method by which these laws of valid thought niav be discovered. There are two theories on this subject, one which maintains that these laws are known a priori (q.v.), the other that they are known a posteriori. Thus we have a school of rationalistic logicians, and a school of experien- tial logicians. Among the most pronounced ad- vocates of rationalistic logic stand Wolll' (q.v.) and Kant. The latter says of 'general logic' (practically logic as here treated) that "as pure logic, it has no empirical principles. . . . Logic is a demonstrative science, and whatever it con- tains must be certain entirely a priori" (Wat- son's translation). Others, such as Eiehl, maintain that these standards of correct thinking are discovered by experience. We think, and some of our thoughts prove valid, i.e. they lead us to results which are confirmed by all experi- ence; others of our thoughts are not valid, i.e. subsequent experience does not confirm them. Now by examination, so it is claimed, we dis- cover that valid thinking is characterized by certain features; it is all reducible to certain types. These types are the norms or laws of thought. Thus logical laws are on a par with pliysical laws. Physical laws are statements of the way in which physical bodies act. Logical laws are statements of the way in which correct thinking proceeds. It is only experience, this school asserts, which can reveal to us these laws. It is impossible in this article to go thoroughly into this question at issue between experiential and rationalistic logicians. All that can be said is that as a matter of fact Aristotle's logic was merely a statement, empirically grounded, of the processes of correct thinking: and all logical advance since Aristotle's day, notably Mill's addi- tion of scientific induction to logic, has been secured bj' a careful study of the actual processes of what is generally regarded as correct thinking. That is, the science of logic in fact originated and has grown on experiential foundations. There is no a priori necessity that we should think logi- cally; otherwise there would be no illogical think- ing going on in the world. Neither is there any way of showing that a certain mode of thinking is incorrect except by showing that it actually leads to error. A teacher in logic who w'ishes to show the fallaciousness of a certain style of reasoning does not appeal to any inner con- sciousness or a priori intuition of his pupil. He shows in the concrete, i.e. by example, that this style of reasoning issues in mistaken conclusions. For example, he shows that it is not correct to say that because all JI is P and all JI is S. there- fore all S is P. How does he prove it? By pointing out that the correctness of the formula would involve the correctness of the argument that all ducks are fowls and all ducks are ani- mals, therefore all animals are fowls. Again, it is not correct to say that because B comes after A, it is the effect of A. Why? Because such a formula would justify the conclusion that "the cock, with lively din, scatters the rear of darK- ness thin." Experience, however, shows that the cock and his din do not have any j>erceptible intluence in dispelling darkness. Hence the for- mula is incorrect, and reasoning according to that formula is fallacious. But there are other formulas which express the processes of correct thinking. These formulas are obtained by com- paring actual trains of thought which lead to valid conclusions, and by discovering their poinU of identity. Now all thinking is judging and all thought is judgment (q.v.). A judgment can always be analyzed into subject and predicate, which in'the act of predication' are recognized as standing to each other in the relation of object of thought and qualification of that object. The subject of a judgment is either a percept or a concept (q.v.). Of the judgments that appear earli- est in thought the subjects are without doubt percepts. It is the business of psychology to in- vestigate the origin of percepts. Logic accepts these percepts as given facts any question about whose origin is of no pertinence to its inquiry, inasmuch as it is merely interested in the manner in which these percepts are manipulated by thought; or, in other words, in the various changes which percepts undergo and in the vari- ous relations into which they enter, when they become elements in a thought process. Xor are even all these changes undergone by jwrcepts of direct bearing upon the problems of logic. For instance, before any judgment can be made upon a percept, attention (q.v.) is necessary. This state of attention is a very difTicull state for the p.syehologist to deal with ; but all the perplexities are of no immediate interest to the logician as logician. He accepts the results pro- duced by attention without bothering himself with questions as to the nature and conditions of attention. Only those changes of percepts which have a direct bearing upon the validity of judgments are treated in logic. These changes are all summed up in the words analysis (q.v.) and generalization, to which may be added com- parison (q.v.), inasmuch as neither analysis nor generalization occurs without comparison. In- deed, analysis, comparison, and generalization are really different aspects of the same process. It is at first probably only as one percept is com- pared with another that any analysis can be made of either of them; and this analysis is pos- sible only as some attribute is discovered to be common to them. And yet this discovery of a common attribute is ipso facto an analysis of the percepts. It is usual to add that before a per- cept can figure in judgment, the common attri- bute predicated of it must be associated with some word or phrase. Tliat is. many logicians claim that language is a necessary precondition of all judgment. Thus Sigwart says that '"it is essential that this" predicate idea "should be taken from those of our ideas which are already luiown to us and named by words which are understood. It must be an idea which has been already received into consciousness, which is con- nected with a word enabling us to retain and reprodvice it, and which is distinguished from all other ideas. . . T cannot say 'this is blue, this is red.' unless I am already familiar with the ideas blue and red, and can reproduce them with the
