Ritchie, David G. “The Relation of Logic to Psychology. II.” The Philosophical Review 6, no. 1 (1897): 1. https://doi.org/10.2307/2175584
RECENT logicians have protested against the old tradition of beginning with an account of terms or concepts and have insisted that the judgment is the primary act of thought. But, in the reasons given for taking judgment first, I do not think a sufficient distinction is generally made between theological and the psychological aspects of the question. That the sentence precedes the word in the historical evolution of language, seems proved from an examination of the beginnings of language among primitive races and among children. This is a fact of undoubted psychological interest, but I do not think it has any direct bearing on the logical question of whether the judgment or the concept is prior; for, let it be said once for all, priority in time is irrelevant in logic. The only priority that concerns us is logical priority. That is logically prior which is logically presupposed in something else; in other words, the logically prior is that on whose truth or on whose existence something else is dependent, but not vice versa. Which of them comes first into any individual's mind, or into the average human mind, is a matter which is of itself a nonlogical moment? But without any irrelevant anthropology psychology, it can be shown on purely logical grounds that the judgment is, in a certain sense, prior to the concept; theological character of concepts cannot be known unless they be considered as terms in an actual or possible judgment. The student of elementary logic is asked (e.g., by Jevons) to describe the logical character of such terms as "metropolis," "book,” “library," "prime minister," etc. It is a puzzling question test to beginners, who are always apt to think that every question must have one and only one correct answer. The same term, i.e., what looks the same when stated in isolation, maybe singular or general, collective, or distributive, according to the context in which it comes. " The Library is in this street, “This book is not in the Library, “It is not in any library.” “Here what we call the same term "library" is singular, collective, general, in succession; and in the last example is either general or collective according as we are thinking of the "any" or the "in."
Aristotle's definition of the term, nay the very word 'term, ‘suggests that the term is the element of a proposition: "The term (terminus = limit, end) is that into which the propositions broken up when we analyze it." The two sides of a sheet of paper have no existence apart from the sheet of paper; but they may certainly be considered separately from it and from one another. Is not a similar abstract procedure permissible in logic? Aristotle has been unduly blamed for adopting in the De Interpretation the concept as his starting point and building up the judgment out of concepts. But we may reasonably suppose that, taking for granted the definition of the Analytics (which was an earlier work), he considered himself at liberty, as in the sciences, to show how to construct a whole in thought out of elements that have only been arrived at by a process of abstraction. It should be observed further that, in the passage in the De Interpretation, his object is to show that the isolated concept is neither true nor false, that only the judgment is the real unit of thought. The very passage in which he is supposed to lapse into an erroneous view of the term .is one in which he is practically asserting the logical priority of the judgment. But here, as elsewhere, the disciples have shown a peculiar facility for overlooking the more important aspect of the master's teaching, and his reputation has suffered in consequence.
In regard to the extension and intention of terms and their relation to one another, it is all important to distinguish theological from the psychological aspects of the question. In considering the theory that the extension and intension of terms vary inversely, we must, first of all, absolutely reject the notion that there can be anything of the nature of a mathematical ratio between these logical aspects. This 'inverse ratio' is only one among many examples of the fatal and delusive fascination which the exactitude of mathematics exercises over the students of other subjects. When we find a logician or an economist using mathematical formulae, we ought to be more than usually on our guard. Mathematical formulae in such matters are more insidious than metaphors. The extension of a term is, at least conceivably or potentially, capable of strict quantitative measurement. The number of individuals or the number of species to which a term is applicable is a quantity in the mathematical sense. But the intention of a term, the number of attributes it includes, is not in this exact sense a quantity at all. How many words we take to express what we mean by a term may in any particular case be estimated quantitatively; but how many they are will depend upon what particular words are used and upon what language a person happens to be using. Where a one-person orgone language uses one word to express an attribute, another person or another language may require two or three.
Extension and intension are not, therefore, strictly commensurable quantities between which we can discover an exact mathematical ratio. Nevertheless it is possible to compare them together ;and, so far as I can see, there is a very good sense in which it can be held that as a matter of logic they tend to vary inversely., the larger extension as a rule goes along with the smaller intention, and vice versa.
It seems to me perfectly irrelevant to object to this, that, while a person may with increasing knowledge of a subject come to know more individual specimens or more species of a genus, his conception of the genus may and should simultaneously increase in richness of content and depth of meaning. Thesis an important psychological fact, and as such should find recognition in any psychological account of the growth of knowledge. A complete 'theory of knowledge' may very well be expected to overlap this portion of genetic psychology. But logic has nothing directly at least, nothing primarily to do with the varying degrees of knowledge of different individuals or with the different stages in the history of an individual mind. For logic 'extension' ought to mean the total applicability of the concept, and ' intension ' the total content or meaning of the concept, if its content were completely known. That is to say, here, as in other cases, logic has to do not with what may happen to be in this or that person's mind, nor even with what, as a matter of fact, is in the mind of the average person, but with an ideal standard of knowledge to which any actual human thought can at best only approximate. It is meaningless to attempt to compare such varying and contingent matters as the number of individual roses, or even the number of species and varieties of rose, that any particular person happens to know of at any moment, with the fulness of the description which he could give at the same moment of the genus Rosa. To use and extend the convenient terminology of Dr. Keynes ‘subjective intension' and 'subjective extension ‘are quantities too fluctuating and indeterminate to admit of comparison, whereas 'objective intension' and 'objective extension ' do conceivably at least admit of comparison. For the purpose of illustration and exposition we must be content to take 'conventional intension' and compare it with the actually known applicability of the term.
‘Conventional intention' '' Dr. Keynes uses for "those attributes which constitute the meaning of a name;'' he does not say 'to whom. ‘I suppose we must understand ' to the average well-informed person of our acquaintance. This use of ‘conventional intention' as a substitute for objective intention which in most cases cannot be completely known, and of the extension known to the average person who is well informed on the subject for the complete 'objective extension,' is perfectly legitimate, and is only one example of that use of convention, which is necessary in every science. Because logic must accept conventions, it does not follow that it must confine itself to a manipulation of arbitrary symbols, and leave alone those fundamental problems of knowledge which we have already seen arise even out of such seemingly abstract formulae as the principle of contradiction. It is only the actually known that we are able to analyze, but we can take the best available knowledge as typical of what knowledge must be, and so seek to discover the general laws to which thinking must conform in order tube knowledge. In dealing, then, with this question of extension and intension, our best procedure is to take some well mapped-out province of knowledge where there is a precise terminology and a clearly arranged system of classification. In such a subject as botany or zoology, it is obvious that the wider class needs a briefer scientific description than the narrower class, which includes all that can be said about the higher with the addition of its own differentiae. That this is so, seems to justify us in regarding the inverse variation as true generally of extension and intension. If we look on the whole universe as a classified system of beings, with the summum genus of being' at the one end of the scale and the various individual existences at the other, then we find our law confirmed ; former being is the emptiest of meaning, and the individual being is the fullest. The singular term has thus an infinite intention and is therefore incapable of complete definition. The question which Mill raised about the connotation of proper names, seems to me to turn entirely on whether we mean by the proper name something different from the singular term. If we do, then it may be true to say that the proper name is denotative but not connotative. But such a distinction between proper names and singular terms seems to me entirely extra logical. It is a matter of grammatical or rhetorical import whether I say, 'this person' or call him 'John Smith.' Logic is only concerned with proper names as appropriated to individual beings and can recognize no distinction between them and singular terms. If the question of extension and intension be cleared of irrelevant psychology and irrelevant grammar, and of inapplicable mathematical precision, it does not seem to present much difficulty.
The problem of logic is analysis in order to discover the conditions of validity. As the logical theory of terms, therefore, should be based on a study of concepts whose applicability and meaning are well understood, so should the logical theory of judgments 1 be based on an analysis of highly developed types of judgment. In the light of such an analysis, it is then profitable to look back on the more rudimentary types, in order to understand their logical significance. In the analysis of a complex type it must not be assumed that one and only one form of analysis is legitimate. Logical analysis being analysis made with the view of testing validity, that form of analysis is to be preferred which is most convenient for that purpose. Now the form of analysis which is most convenient in order to make clear the mutual implication of propositions, and the validity or invalidity of the inferences of which judgments constitute the elements, is not necessarily that form which corresponds most closely to what is actually in the mind of any particular individual or of the average person when uttering the proposition. This last is a psychological problem and should not be confused with logic. A complete theory of knowledge may indeed be expected to contain a genetic account of the evolution of the different species of judgment, and to classify these species according to an evolutionary or genealogical principle, as is done so admirably by Mr. Bosanquet in his Logic ; but for logic the primary business is, I think, to give an analysis applicable potentially to every form of judgment, and such analysis must be based on the characteristics of those judgments where the logical aspects are most prominent to consciousness and can therefore be most clearly apprehended. It is undoubtedly very important to recognize that in every judgment, as actually made by anyone, there is a reference to reality in general, or to some portion of reality, as the ultimate < subject of discourse. This account of judgment is confirmed in an interesting way by the fact that in the most rudimentary of all types of judgment the impersonal perceptive judgment (e.g., 'It is hot'; < It hurts,' etc.) there is nondeterminate subject, but only the indeterminate ' it ' = reality in general. But this recognition of the ' reference to reality ‘as ultimate subject of discourse does not falsify nor exclude the traditional analysis of every judgment into subject and predicate, an analysis which is of course based on a study of those kinds of judgments in which the 'subject' is some clearly determined portion of the real world. Furthermore, the recognition that every term as actually used in a judgment must have both a meaning and some objective reference, however slight and indirect, allows us to analyze every judgment according to either extension or intension, or to treat the subject as primarily extensive (quantitative), and the predicate as primarily intensive(qualitative). The last of these modes of analysis may be preferred, because it corresponds best to the ordinary form of language, and to what is most usually in our minds when we say something (predicate a characteristic, i.e., a quality) of something (i.e., of all or some part of a thing or class of things) But the interpretation of both subject and predicate in terms of extension has the convenience that it exhibits most clearly the possibilities of transition from one proposition directly to another, and the implications of combinations of propositions. The continuity which is the essence of all inference can be most easily exhibited by interpreting the 'middle term' in mediate inference extensively. The extensive interpretation of propositions does, of course, make possible the treatment of judgments as equations, and so seems to threaten logic with absorption in algebra. But the logical objection to the quantification of the predicate, which is presupposed in the equational theory, is not that such equational judgments (all men = some animals, etc.) are not very often in our minds; this would be a purely psychological argument. The real logical objection is that proposition with a definitely quantified predicate is always a complex verbal form which expresses two judgments and no-one. Thus, all equilateral triangles are all equiangular triangles 'wraps up into one formula, two propositions which require separate geometrical proof (Euclid, I, 5, 6). Now the business of logic is to analyze complex mental processes into single judgments, and therefore these complex equational sentences do not represent the elements with which we have to deal.
The chief defects of the traditional formal logic seem to me to lie partly in its too exclusive predilection for the extensive interpretation of the judgment, but still more in the absence of distinction between the singular and the universal proposition, and, above all, in the absence of distinction between the mere collective judgment and the true universal. Very different types of judgment are all classed together as A and E propositions. All the books on this shelf are bound in calf ' is an adjudge of a different type from 'The angles of a triangle are together equal to two right angles.' The ignoring of this distinction is the chief thing which has exposed the Aristotelian logic to attack in modern times. Mill's thesis that the Aristotelian syllogism is by its very profession a petition princeliest upon a narrow ' class 'interpretation of the dictum demonic et de null, that is most certainly not justified by Aristotle’s own language, which simply expresses the principle of continuity ("what may be predicated of the predicate may be predicated of the subject"), and on the assumption that every universal proposition is simply a collective judgment. Now certainly if, 'All M is P,' merely means 'A is P, ''B is P,'' C is P,' and 'D is P,' and if we then go on to say,' B is one of this group (M), therefore it is P,' we have made no advance, but, as Mill points out, are simply reading off our memoranda. Where, however, the two premises are both singular, and where (if anywhere) one is a true universal (i.e., necessary), Mill's arguments are inapplicable. That excellent tale of Thackeray's about the too confidential abbe" (it is quoted by Mr. Bosanquet in his Essentials of Logic, pp. 140, 141) seems to me alone sufficient to refute Mill's criticism of the syllogism. "An old abbe", talking among a party of intimate friends, happened to say, ' A priest has strange experiences; why, ladies, my first penitent was a murderer.' Upon this, the principal nobleman of the neighborhood enters the room. Ah, Abbe, here you are, do you know, ladies? I was the Abbe's first penitent, and I promise you my confession astonished him!' ' The company, having the two premises given them from different quarters (and of course they might have been given at any interval of time and through many different channels), are at once able to form a conclusion which is certainly new ' to them. There is no suspicion of petitio principii here. The syllogism arises only from the combination of the premises; but the combination of the premises is the conclusion.
Mill expressly denies the existence of any true universal; all judgments professing to be necessary are, according to him, simply incomplete collective judgments, which we assert as if they were complete. The only necessity he allows is a psychological necessity: a tendency in our minds to expect a repetition of similar experiences. Mill's argument has undoubtedly been made easier for him by the absence of any distinction in the traditional logic between the true universal and the mere collective judgment ; but the main determinant of his whole treatment of the subject of inference has been his assumption that he is dealing with a psychological problem, and that there is no logical problem distinct therefrom. The very question" whether the syllogistic process is or is not a process of inference "shows that he thinks of the syllogism as the consciously recognized and formulated inference. We need only translate Mill’s question into Aristotelian Greek to see its irrelevance as applied to Aristotle's own analysis of inference. 'Syllogism ‘to Aristotle simply means 'inference,' i.e., out of a combination of data arriving at something new in the only sense, of course, in which we can ever know anything 'new'; for we can never learn anything absolutely discontinuous with our existing knowledge. Still less could be said to ' infer 'what has no connection with anything else. But how far we are conscious of the form of our inference is a matter for psychology: whether we formulate it in words is a matter of grammar or rhetoric. Logical analysis applies equally to fully conscious and half-conscious inferences, to fully formulated and half-formulated inferences ; though of course, as already said, our knowledge of the logical forms of inferences is best arrived at by a study of the most fully conscious and clearly expressed specimens we can obtain.
Mill holds that all inference is ultimately from particular to particular. Now if it were true that, as a matter of psychology, we had first one particular case in our minds and then passed at once to the thought of another particular case, this would not prove that, as a matter of logic, inference was possible from particular to particular. Mill speaks of the village matron recommending her neighbor to try the medicine that cured her own child, without uttering any formal universal proposition, or without consciously formulating any universal judgment. But if she is asked why, she must enunciate the major premise of her argument. She must either commit herself to the statement that the drug is a panacea, or she must expressly recognize the similarity of the two cases. But to recognize similarity is, as a matter of logic, to arrive at a 'middle term distributed, undistributed, or approximately distributed:' All such (or some such or most such) cases are cured by this remedy. This is such a case.' Mill himself uses the words" on the recollection and authority of what she accounts the similar case of her Lucy."
Mr. Hobhouse, in his chivalrous attempt to defend Mill against the fierce onslaught of Mr. Bradley, lays stress on this statement of Mill's ; and he seems even inclined to follow Milling making likeness an ultimate category, though he admits that where there is likeness there is generally identity in difference. 1As an argument that there is not always such identity, Mr. Hobhouse asks : " What is the identity and what is the difference between blue and green ?"This question does not seem very difficult to answer : blue (in the widest application of the name) is the identity which links together the most purple of blues and the most yellowish of greens, when we see them in the spectrum. Mr. Hobhouse's chapter on “Resemblance and Identity" seems to me to offer one of the many cases in which a more precise distinction between psychology and logic is needed. "Likeness," he says, "does not in the least bit cease to be real because it is analyzed. “That may be, but it is with the analysis that logic has to doom, Hobhouse seems to think both likeness and identity ‘given' to immediate apprehension. Whether that is so or not is a question for psychology. Logically, identity is the prior, because there can be (in thought) identity without difference, though it is a mere abstraction, whereas we cannot think of ' likeness ' without implying both identity and difference.
Mill's treatment of likeness as an ultimate category rests upon the psychological atomism which forms the basis of his whole theory of knowledge. Mr. Hobhouse is an indignant tamer. Bradley's supposing that when Mill talks of inference from particular to particular, he means 'particular images.' It is quite true that Mill does not mention them in the passage which Hobhouse quotes from the Logic; but we know sufficiently well from other sources notably from his Examination of Hamilton that Mill accepts that theory of knowledge which was most clearly (and with fullest consciousness of its issues) expounded by Hume. Mill's whole argument in the Logic about the nature of mathematical judgments would be without meaning, unless we suppose that by 'particulars ' he means ultimately particular images of particular sense-impressions.
Once admit that, as a matter of logic, likeness may be analyzed into identity in difference, then, if it is admitted that inference is only justified by similarity, it is admitted that inference implies identity and therefore that we cannot logically pass from particular to particular except through a universal. We may not think of formulating the universal principle, the major premise, of our inference till we are met by the question why; and in proportion as we are untrained in abstract thinking or in the habit of scientific expression, we may find it difficult to do so; but the validity of our inference, nevertheless, depends on the truth of the universal principle, whether it be consciously apprehended or not.
Now if it be once admitted that logically no transition from particular to particular is possible except through a universal, this suggests that perhaps the psychological theory which holds that such transition takes place as a matter of fact, may also need revision. It would imply a break in the continuity of our mental life, a break which we should not scientifically be prepared to accept without very distinct proof, if no trace of the identity (the universal element) which comes out clearly in the higher and more fully conscious stage of logical inference could be found in the lower and less explicit stages of association and perception. And modern psychology, though it started from the empirical standpoint of Hume, seems to be coming to recognize that, in Mr. Bradley's phrase, “Association marries only Universals."
It may be considered misleading or inconvenient, as a matter of descriptive psychology, to speak of perception as being an unconscious or subconscious inference ; but it is important as a matter of logic to recognize that the validity of perceptive judgments can be shown to depend on the same principles as those which determine the validity of conscious logical processes. If, for instance, looking at a distant mountain side, I say' I see snow,' this perceptive judgment (which I might quite as well have expressed in the inferential form 'That must be snow') is an inference of a probable kind. It may be analyzed as an Aristotelian enthymeme: ' Snow is white, glistening. (a premise due to past experience lying latent in the mind). This presentation is white, glistening, etc. Therefore, this is snow.' This is an enthymeme in the second figure, an enthymeme of the weakest kind. But as the points of identity become more numerous, the middle term approximates to distribution, and so the major premise approaches the stage at which it admits of simple conversion. 'All that has this particular combination of marks is snow.' And then the inference passes into the first figure.
Nothing, it may be remarked in passing, shows more forcibly the degradation to which Aristotle's logic has been subjected than the perversion of the meaning of 'enthymeme' in the traditional formal logic. To define an enthymeme as a syllogism with a suppressed premise or conclusion, and solemnly to distinguish enthymemes of the first, second, or third order according as one or other of the three propositions is suppressed, all this is, in logic, as absolutely irrelevant and unscientific as if, in zoology, we were to recognize a distinct species of quadruped when one or more of the legs is not seen, and then subdivide the species according as a fore leg or a hind leg, a left leg or a right leg, were at the moment out of sight. How I choose to express my argument, is a matter of rhetoric. If I wish to produce conviction, it may be expedient to conceal my weakest premise or to leave my hearers to make for themselves a conclusion which I only suggest. But such tricks of the platform furnish no special and peculiar types of inference for the science of logic. Aristotle's enthymeme " from signs (or symptoms)" is, on the other hand, a really important contribution to the logic of probable (as distinct from demonstrative) inference, far more important than his " inductive syllogism from all the particulars." The diagnosis of the physician (Aristotle's own illustrations are medical), the circumstantial evidence of the law courts, and, as we have just seen, our ordinary recognitions imperception are affirmative syllogisms in the second figure, which gain in probability as they approach the stage at which the major premise can be converted, and the syllogism becomes of the first figure. Even in the first figure such enthymemes, in Aristotle’s view, fall short of the scientific syllogism, because our middle term is a sign, or a combination of signs, and not a cause or ground. In the ' scientific' syllogism the ratio cognoscendi is the ratio essendi.
Mill's inductive methods are a valuable contribution to theological study of the manner in which, in ordinary life and in the sciences, we test the guesses that we make about the causes of events; but none of them are ‘inductive’ in the sense of being arguments which do not proceed logically from universal to particular. The "method of residues" is a professionally adductive method, and involves the assumption of an axiom, the truth of which is most easily recognized in its purely mathematical form. The other methods are deductive applications of the principle of causation, as Mill himself acknowledges, though he attempts to derive the belief in universal causation and in the uniformity of nature from our experience of particular cases of causation and of particular uniformities of sequence, an argument which turns on the same confusion of psychology with logic as that on which his attack on the syllogism depends. As a matter of mental development, we understand particular cases before we understand the principle involved in them; but the universal principle, thought may be apprehended and formulated later, is logically prior. Our conviction of the universal may come later, but the truth of the particular instance is dependent on the truth of the universal principle. The question of the logical presuppositions of inferences about causation is, however, too large for treatment towards the end of a long discussion. I can only very briefly indicate what seem to me the main points for consideration,
(I) In the sciences and in ordinary life we make abstractions according to our convenience. We isolate certain phenomena as ' causes ' for special consideration, taking for granted the other elements in the total reality. In his exposition of the inductive methods, Mill is obliged to desert his attempt at a philosophical conception of cause as the sum total of conditions, and to adopt the popular use of the term.
(2) A logical analysis of what causation implies, compels us to go beyond the artificial distinction of antecedent and consequent, and to regard the assignment of causes as only one particular aspect of that fitting of particulars into their place in a system which constitutes explanation.
(3) This underlying assumption of the system is identical with the principle of contradiction (or inconceivability of the opposite). In passing from 'formal Logic ' to the logic of probable matter, in passing from mathematics to the sciences of observation and experiment, we do not come across a new set of a priori principles disconnected with our previous canons of inference. Our thinking is determined by the same principle of totality or coherent system (or however we describe it) throughout, though in passing from the more abstract to the more concrete sphere, we pass to a region in which we own certain knowledge is more limited just because it is less abstract. The sphere of the contingent is simply the sphere where it is more difficult for us in intricate material to see the necessity: and the Principle of Sufficient Reason is identical with the Principle of Contradiction.
A due consideration of the difference between the logical question of validity, and the psychological question of the temporal evolution of knowledge, seems to me to vindicate the syllogistic analysis of Aristotle from another charge of incompleteness which is made even by those who recognize the necessity of a universal element in our transition from particular to particular. Such inferences as < A > B; B > C; -. A > C are supposed to be incapable of reduction to syllogistic form. But the psychological fact that it is easier to see the principle., of a fortiori, in a concrete or in a brief symbolic form than when fully expressed in abstract language is no proof that the inference is logically possible except in virtue of the truth of the abstract general principle. The general principle here and in all similar cases (most A are B; most A are C; A is to the north of B, B is to the east of C, etc.) is a principle of quantity or a necessity of spatial relations ; and it is to confuse logic with mathematics, if we set up axioms of quantity and axioms about space as if they were parallel to the dictum de omni etude null. Every science has its own axioms, which may be arbitrary conventions, or derived from other sciences, or capable of proof per impossible (by inconceivability of the opposite) ;but the axioms of quantity or space are no more themselves principles of logic than are the Acts of Parliament which form the major premises of judicial and administrative inferences.
Finally, to guard against misunderstanding, it may be well to point out that the 'Intuitionist' who appeals to the evidence of consciousness or the consensus humane generis in support of his immediate or necessary truths falls into precisely the same confusion of psychology (or anthropology) with logic as his' Sensationalist ' opponent. A priori principles, if we call them so, are not known ' prior to 'experience; they are not ' immediate,' in the sense of being straight away, without any trouble, by anybody and everybody. They are a prior only in the sense of not being dependent upon experience for their validity; they are ' immediate 'only in the sense of not being deducible through a middle term from other logically prior principles. They cannot be 'proved' except by a ‘transcendental proof' i.e., per impossible, by showing that the denial of them makes knowledge impossible and involves us in contradiction. Nothing has more hindered the understanding and acceptance of the idealist theory of knowledge, than the persistent error of treating the logical argument for the validity and necessity of the laws of thought, as if it were an appeal to the average individual's incapacity to analyze some of the facts of his consciousness.
In the attempt to deal with my problem, I have been obliged to sketch in brief outline a good many parts of logic. If I have not altogether failed to make my points clear, I think I have done something incidentally towards vindicating the essential value of Aristotelian logical analysis. I have also tried to show that formal logic is not so barren of philosophical interest as is often supposed, but, if studied seriously, leads us inevitably into problems of epistemology and metaphysics. But we are left with this seemingly paradoxical conclusion, that although psychology ought to be kept out of logic, it cannot be kept out of a complete epistemology to which logic leads up; and, on the other hand, logic ought not to be kept out of psychology. This conclusion is paradoxical only if we have been making the false assumption that logic and psychology are parallel sciences, or that logic is simply a branch or application of psychology. Psychology is, or professes to be, one of the special sciences, like physiology; and yet, as the science of the knowing mind, it occupies a unique 'central' position. So far as psychology is a special science, logic is related to it as it is to any other of the special sciences. But it is difficult for psychology to become one of the special sciences of nature or to remain merely one of them; logic and epistemology claim part of its province for their own, and seek to turn it into a 'philosophy as distinct from a special science of mind.