CHAP. V. SECT. VIII. Hyponature of inferred thetical § 190. The inference of some objective power for the events which we ourselves do not produce is compulsory; but there is great room for doubt and difference as to the nature of that power. Men primarily inferred the existence of innumerable causes. spirits; others have postulated a common substratum, matter; others infer a universal mind; others, again, conclude that there are various powers mutually correlated, which are but forms of a universal Force. As this is not a work upon metaphysics, we do not feel called upon to discuss these conflicting theories. But we may point out that inferences regarding the nature and laws of the unknown cause or causes of phenomena must, from the nature of the case, be hypothetical. And the best that we can do is to adopt that hypothesis which, to our minds, most fully explains the phenomena. It is, therefore, upon a careful study of phenomena and their mutual relations that we should build our hypothetical inferences, taking care that no phenomena be overlooked or placed out of their proper relations. It would take us far beyond our limits to discuss this subject as it should be discussed. Some have held that, because in these matters the absolute certainty of mathematical demonstration or of intuition is unattainable, they ought to be excluded from the sphere of science altogether. This, however, would be an arbitrary limitation of the sphere of science, against which the actual practice of scientific men is a standing protest. Hypotheses fingo, Newton to the contrary notwithstanding, is the practical language of the man of CHAP. V. SECT. VIII. Inferences as to existence and relations of pheno mena. science, even in cases when there appears to be no hope of his hypotheses ever being directly verified. § 191. We now come to the consideration of inferences regarding the existence and relations of phenomenal objects. Illustrations of these inferences have already been given in § 181. . . . We have seen, also, that they may be generally divided into two classes, in the first of which we infer some relation to exist between phenomena which we know; and in the second we infer the existence of a phenomenon not directly known, and holding a certain relation to another phenomenon, or set of phenomena, known to exist. And the question to be decided is, what is the ground of the inferences we draw in each of these cases? Now, it must be observed that, in drawing these inferences, there is great room for error. Thousands of instances might be pointed out in which false relations have been supposed to exist between phenomena. It must be admitted, then, that inferences may be drawn which are not in conformity with the truth of nature. And, therefore, we must consider, not the ground of inference in general, but the ground of correct inference. But in this it is implied that amongst phenomena there exist certain objective relations which may be known by us or not, and if our inferences are not the true expression of these objective relations they are not correct. We observe, for example, that night always precedes day, but if we were to infer from this uniformity of sequence that a causal relation exists between them, our inference would be erroneous, as there are various reasons for our concluding that this is not the true relation. In drawing inferences, therefore, it is necessary to distinguish between those relations of things which are accidental and those which are essential. In order to do this properly, we must have recourse to those laws of induction which Bacon and his followers have formulated. The object of these laws is to enable us to distinguish between accidental and essential co-existences and sequences. Those relations which are of the latter class are thought to have some connection with the cause of the phenomena. And those inferences are thought to be correct which connect together phenomena between which there is objectively a causal relation, or a relation depending upon identity of causation. The relation, for example, between the lowering of the temperature of the atmosphere and the deposition of dew is an essential one, and the inference which connected the one with the other was correct, although, perhaps, it is not proper to call this the causal relation. It is correct, also, to infer that animals which we have discovered to have some of the essential characteristics of birds are oviparous, because we have learned that there is an essential objective connection between these characteristics and that quality which we infer. Thus, it appears that, before we can be satisfied with inferences regarding phenomena, the subjective connection which we form must be recognised as being the counterpart of essential objective relations between things. Inferences are not concerned CHAP. V. SECT. VIII. SECT. IX. CHAP. V. merely with the relations of ideas, but also, and principally, with the objective connections of real things. Contro versy regarding the SECTION IX. THE FORM OF INFERENCE. § 192. Is there any general form according to which the mind acts in the drawing of inferences, syllogism. and, if so, what is that form? This question has given rise to a good deal of controversy, especially with reference to the claims of the syllogism to be considered the universal type of reasoning. It has been asserted, on the one hand, that, if all the mental acts and principles which are involved in the drawing of inferences be explicitly stated in language, the statement will assume the form of a syllogism. It has been held, on the other, that there are many instances of reasoning in which the mental process cannot, without the greatest violence, be put into the form of a syllogism, although it is admitted that many kinds of reasoning do naturally assume that form. Before discussing this question, we must examine the nature of the syllogism. According to the accounts of the syllogism which we find in logical treatises, syllogism is inference which is drawn from the general, from some proposition or judgment which is generally true. There are three terms in every syllogism, two of which, the minor and major terms, are the subject and predicate of the conclusion; while the third, the middle term, is found in each of the premises, and the medium by which the terms of the conclusion are compared together. All these terms are held to be the names of concepts having a greater or less denotation and connotation. And in the syllogism there is a comparison of the denotation or the connotation of the minor or major terms, with the denotation or connotation of the middle term. This comparison has for its object to determine whether the one term contains the other, either wholly or partially—that is, whether the denotation or the connotation of the one term includes, wholly or partially, that of the other. Now, in order to determine the character of the claims of the syllogism to be considered the only type of reasoning, we shall have to consider, (1) whether inference proceeds only from the general, (2) whether concepts are the sole elements of inference, and (3) whether the relation of containing and contained is the only relation regarding which we infer. CHAP. V. SECT. IX. be from par ticulars. § 193. The first of these points has been practically Inferences decided by us already. We have seen that some of may drawn the so-called mathematical axioms are generalised from particulars, and, consequently, that particular inferences are drawn before the generalised axioms are formed. We have seen, also, that in perception an objective cause of our passive states is inferred to exist from the consciousness of some particular sensation uncaused by our conscious effort. Now, the only ground upon which inferences of this kind can be expressed in the form of a syllogism is, that |