Studies in Logical Theory · John Dewey
A CRITICAL STUDY OF BOSANQUET'S THEORY OF JUDGMENT[44]
The subject is wholly within the judgment, not in any sense outside of it; but it is at the same time true that the subject of the judgment is reality. The fact that the subjects of all judgments--even those of the most elementary type--bear evident marks of the work done by thought upon them, ceases to be a problem. The subject is essentially a thing constituted by the doubt-inquiry process, and functioning within it. The necessity for an intermediate _real world as it is to me_ between the real world and the knowing process disappears, because the _real world as it is to me_ is the only real world of which the judgment can take account. There is no longer any divorce between the content of the subject and its existence. Reality in his sense of the term--reality as fact--does not fall on the side of _existence_ in distinction from content, but on the side of _function_ in distinction from content.
The predicate of the judgment is that part of the total experience which is taken as doubtful, or tentative. As we have seen, every act of adaptation involves a definite situation to be reacted to (subject) and an indefinite or tentative material with which to react (predicate). We have pointed out that a situation which demands a judgment never appears in consciousness as mere questioned or questionable situation.[52] There is always present, as soon as the doubt arises, some sort of tentative solution. This is the predicate or idea. Just as the fact, or real in the narrower sense, is that which is taken as given in the situation, so the ideal is that which is taken as tentative. Its ideality does not consist in its reference to another order of existence, the objective world of meanings, but in its function within the judgment, the estimate of the whole situation as leading up to the adequate act. Just as we no longer have any need for the mediation of the _real world as known to me_ between subject and reality, so we no longer need _the objective world of meanings_ to bridge the chasm between the predicate and reality. The difficulty of understanding how ideas can be used to build up facts disappears when we regard fact and idea, not as different orders of existence, but as contents marking different phases of a total function.
Ideas, as Bosanquet represented them, proved to be extremely unsatisfactory tools to use in building up a knowledge of reality. In the first place, their value as instruments of thought depends upon their universality. We have already reviewed Bosanquet's difficulties in attempting to explain the universality of ideas. The universality of an idea cannot reside in its mere existence as image. Its existence is purely particular. Its universality must reside in its reference to something outside of itself. But no explanation of how the particular existence--image--could refer to another and fuller content of a different order of existence could be discovered. The fact of reference remained an ultimate mystery. From the new point of view the image gains its universality through its organizing function. It represents an organized habit which may be brought to bear upon the present situation, and which serves, by directing action, to organize and unify experience as a whole. It is only as function that the concept of reference can be made intelligible. Of course, considered as content, the idea is just as particular from this point of view as from any other. We still have to discuss the question as to whether or not the particularity of the idea has a logical value. The fact that it had none in Bosanquet's theory sets a limit to the validity of thought. But if the real test of the validity of a judgment is the act in which it issues, then the existential aspect of the idea must have logical value. The existential aspect of the idea is the "my" side of it. It is as my personal experience that it exists. But it is only as my idea that it has any impulsive power, or can issue in action. Far from being ignored, therefore, the existential aspect is essential to the logical, the determinative, value of an idea.
Ideas, according to the representational theory of knowledge, proved to be a poor medium for knowing reality in still another respect. They are in their very nature contents that have been reduced from the fulness of experience to mere index-signs. Even though their reference to a fuller content in the objective world of meanings presented no problem, still this objective world of meanings is far removed from reality. And yet, in order to know, we must be able to affirm ideas of reality. On the functional theory of ideas, their value does not rest at all upon their representational nature. They are not taken either in their existence or in their meaning as representations of any other content. They are taken as contents which mark a given function, and their value is determined entirely by the adequacy of the function of which they are the conscious expression. Their content may be as meager as you please. It may have been obtained by a long process of reducing and transforming sensory experience, but if it serve to enable its possessor to meet the situation which called it up with the appropriate act, then it has truth and value in the fullest sense. The reduction of the idea to a mere index-sign presents no problem when we realize that it is the tool of a given function, not the sign for a different and fuller content. The idea thus becomes a commendable economy in the thought-process, rather than a reprehensible departure from reality.
We have already upon general considerations criticised the point of view which holds that ideality consists in reference to another content. In arguing that this reference cannot be primarily to reality itself, but rather to an intermediate world of meanings, Bosanquet cites the question and the negative judgment. In the question ideas are not affirmed of reality, and in negation they are definitely denied of reality, hence their reference cannot be to reality. It must therefore be to an objective world of meanings. It may be worth while to point out in passing that, from the functional point of view, the part played by ideas in the question and in negative judgment is the same that it is in affirmation. We have brought out the fact that all judgment arises in a doubt. The earliest stage of judgment is accordingly a question. Whether the process stops at that point, or is carried on to an affirmation or negation, depends upon the particular conditions. The ideas which appear in questions present no other problem than those of affirmation. They are ideas, not by virtue of their reference to another content in the world of meanings, but by virtue of their function, _i. e._, that of constituting that part of the total experience which is taken as doubtful, and hence as in process.
In order to make this point clear with reference to negative judgments, it will be necessary to consider the relation of negative and positive judgments somewhat more in detail. All judgment is in its earliest stages a question, but a question is never _mere_ question. There are always present some suggestions of an answer, which make the process really a disjunctive judgment. A question might be defined as a disjunctive judgment in which one member of the disjunction is expressed and the others implied. If the process goes on to take the form of affirmation or negation, one of the suggested answers is selected. To follow out the illustration of the bread used above, the judgment arises in a doubt as to the nature of the white object perceived, but the doubt never takes the form of a blank question. It at once suggests certain possible solutions drawn from the mass of organized experience at the command of the person judging. At this stage the judgment is disjunctive. In the illustration it would probably take the form: "This is either bread or cake." The further course of the judgment rejects the cake alternative, and selects the bread, and the final outcome of the judgment is formulated in the proposition: "This is bread."
But how did it happen that it did not take the form: "This is not cake"? That proposition is also involved in the outcome, and implied in the judgment made. The answer is that the form taken by the final outcome depends entirely on the direction of interest of the person making the judgment. If his interest happened to lie in obtaining bread, then the outcome would naturally take the form: "This is bread," and his act would consist in eating it. If he happened to want cake, the natural form would be, "This is not cake," and his act would consist in refraining from eating. In other words, the question as to whether a judgment turns out to be negative or positive is a question of whether the stress of interest happens to fall on the selected or on the rejected portions of the original disjunction. Every determination of a subject through a predicate includes both. The selection of one or the other according to interest affects the final formulation of the process, but does not change the relations of its various phases. An idea in a negative judgment is just what it is in a positive judgment. In neither case is it constituted an idea by reference to some other content. So far we have outlined Bosanquet's theory of the judgment; have noted the apparently insoluble problems inherent in his system, and have sketched a radically different theory which offered a possible solution for his difficulties. It now remains to develop the implications of the new theory further by comparing its application to some of the more important problems of logic with that of Bosanquet. In closing we shall have to inquire to what extent the new theory of the judgment with its metaphysical implications has proved more satisfactory than that of Bosanquet.
The special problems to be considered are (1) the relation of judgment to inference; (2) the parts of the judgment and their relationship; (3) the time element in the judgment; and (4) the way in which one judgment can be separated from another. 1. The discussion of the relation between judgment and inference comes up incidentally in Bosanquet's treatment of the distinction between a judgment and a proposition (p. 79). The proposition, he says, is merely the enunciative sentence which represents the act of thought called judgment. With this distinction we should agree. In his discussion of the point, however, he criticises Hegel's doctrine that a judgment is distinguished from a proposition in that a judgment maintains itself against a doubt, while a proposition is a mere temporal affirmation, not implying the presence of a doubt. The ground of his criticism is that judgment must be regarded as operative before the existence of a conscious doubt, and that, while it is true, as Hegel suggests, that judgment and inference begin together, they both begin farther back than the point at which conscious doubt arises. Doubt marks the point at which inference becomes conscious of its ground. Now, it is undoubted that inferences in which the ground is implicit exist at an earlier stage of experience than those in which it is explicit. The former we usually call simple apprehension, and the latter judgment. What Bosanquet wishes to do is to make the term "judgment" cover both the implicit and the explicit activities.
The question at once arises whether such a use of terms is accurate. There is certainly a wide difference between an inference which is conscious of its ground, and one which is not. It is conceivably a distinction of philosophic importance. To slur the difference by applying one name to both accomplishes nothing. It will be remembered that the presence of a conscious doubt is the criterion of judgment adopted in the standpoint from which we have been criticising Bosanquet's theory. We should accordingly make the term "inference" a wider one than the term "judgment." A judgment is an inference which is conscious of its ground. Since fact and idea have been represented as constituted in and through judgment, the question which at once suggests itself is: What, from such a standpoint, is the criterion of fact and idea in the stage of experience previous to the appearance of judgment? The answer is that the question involves the psychological fallacy. There is no such distinction as fact and idea in experience previous to the appearance of judgment. The distinction between fact and idea arises only at the higher level of experience at which inference becomes conscious of its grounds. To ask what they were previous to that is to ask _what_ they were before they _were_--a question which, of course, cannot be answered.
Our reason for not adopting Hegel's distinction between a judgment and a proposition would accordingly not be the same as Bosanquet's. The question has already been touched upon in the distinction between dead and live judgments. What Hegel calls a proposition is really nothing but a dead judgment. His illustration of a temporal affirmation is the sentence: "A carriage is passing the house." That sentence would be a judgment, he says, only in case there were some doubt as to whether or not a carriage was passing. But the question to be answered first is: When would such a "statement" occur in the course of our experience? It is impossible to conceive of any circumstances in which it would naturally occur, unless there were some doubt to be solved either of our own or of another. Perhaps one is expecting a friend, and does not know at first whether it is a carriage or a cart which is passing. Perhaps some one has been startled, and asks: "What is this noise?" What Hegel wishes to call a proposition is, accordingly, nothing but a judgment taken out of its setting.
2. In dealing with the traditional three parts of the judgment--subject, predicate, and copula--Bosanquet disposes of the copula at once, by dividing the judgment into subject and predication. But the two terms "subject" and "predication" are not co-ordinate. Subject, as he uses it, is a static term indicating a _content_. Predication is a dynamic term indicating the act of predicating. It implies something which is predicated of something else, _i. e._, two contents and the act of bringing them into relation. Now, if what we understand by the copula is the _act_ of predicating abstracted from the content which is predicated of another content, then it does not dispose of the copula as a separate factor in judgment to include thing predicated and act of predicating under the single term "predication." The term "predication" might just as reasonably be made to absorb the subject as well, and would then appear--as it really is--synonymous with the term "judgment." But Bosanquet's difficulties with the parts of the judgment are not disposed of even by the reduction to subject and predication. He goes on to say: It is plain that the judgment, however complex, is a single idea. The relations within it are not relations between ideas, but are themselves a part of the idea which is predicated. In other words, the subject must be outside the judgment in order that the content of the judgment may be predicated of it. If not, we fall back into "my idea of the earth goes round my idea of the sun," and this, as we have seen, is never the meaning of "The earth goes round the sun." What we want is, "The real world has in it as a fact what I mean by earth-going-round-sun." (P. 81.)
We have already pointed out the difficulties into which Bosanquet's presupposition as to the nature of reality plunges him. This is but another technical statement of the same problem. If the subject is really outside of judgment, then the entire _content_ of the judgment must fall on the side of predicate, or idea. In the paragraphs that follow, Bosanquet brings out the point that the judgment must nevertheless contain the distinction of subject and predicate, since it is impossible to affirm without introducing a distinction into the _content_ of the affirmation. Yet he considers this distinction to be _merely_ a difference within an identity. It serves to mark off the grammatical subject and predicate, but cannot be the essential distinction of subject and predicate. His solution of the puzzle is really the one for which we have been contending, _i. e._, that "the real world is primarily and emphatically my world," but he still cannot be satisfied with that kind of a real world as ultimate. Behind the subject which presents my world he postulates a real world which is not my world, but which my world represents. It is the relation between this real world and the total content of a judgment which he considers the essential relation of judgment. This leaves him--as we have pointed out--as far as ever from a theory of the relation of thought to reality, and, moreover, with no criterion for the distinction of subject and predicate within the judgment. To say that it is a difference within an identity does not explain how, on a mere basis of content, such a difference is distinguished within an identity or how it assumes the importance it actually has. He vibrates between taking the whole intellectual content as predicate, the reality to be represented as subject (in which case the copula would be the "contact of sense-perception") and a distinction appearing without reasonable ground or bearing _within_ the intellectual content. When subject and predicate are regarded as the contents in which phases of a function appear, this difficulty no longer exists.
3. In discussing the time relations within judgment (p. 85) Bosanquet first disposes of the view which holds that the subject is prior to the predicate in time, and is distinguished from the predicate by its priority. He emphasizes the fact that no content of consciousness can have the significance of a subject, except with reference to something already referred to it as predicate. But while it cannot be true that the parts of the judgment fall outside of one another in time, it is yet evident that in one sense at least the judgment is in time. To make this clear, Bosanquet draws a provisional distinction between the process of arriving at a judgment and the completed judgment. The process of arriving at a judgment is a process of passing from a subject with an indefinite provisional predicate--a sort of disjunctive judgment--to a subject with a defined predicate. This process is evidently in time, but it is as evidently not a transition from subject to predicate. It is, as he says, a modification, _pari passu_, of both subject and predicate. The same distinction, he thinks, must hold of the judgment when completed. But this throws us into a dilemma with reference to the time-factor in judgment. Time either is or is not an essential factor in judgment. If it is not essential, then how explain the evident fact that the judgment as an intellectual process does have duration? If it is essential, then how explain the fact that its parts do not fall outside one another in time? Bosanquet evidently regards the former problem as the easier of the two. His solution is that, while the judgment is an intellectual process in time, still this is a purely external aspect. The essential relation between subject and predicate is not in time, since they are coexistent; therefore time is not an essential element in judgment.
The first point at which we take issue with this treatment of time in relation to judgment is in the distinction between the process of arriving at the judgment and the completed judgment. Bosanquet himself defines judgment as an intellectual act by which an ideal content is referred to reality. Now, at what point does this act begin? Certainly at the point where an ideal content is first applying to reality, and this, as he points out, is at the beginning of the process which he describes as the process of arriving at a judgment. It is nothing to the point that at this stage the predicate is tentative, while later it becomes defined. His process of arriving at the judgment is exactly the process we have been describing as the early stages of any and every judgment. When he talks about the judgment as completed, he has apparently shifted from the dynamic view of judgment implied in his definition to a static view. All he could mean by a completed judgment--in distinction to the total activity of arriving at a judgment--is the new content of which we find ourselves possessed when the total process of predication is complete. But this content is not a judgment at all. It is a new construction of reality which may serve either as subject or as predicate in future judgments.
Now, if we regard the judgment as the total activity by which an ideal content is referred to reality, then must we not regard time as an essential element? Bosanquet answers this question in the negative, because he believes that if time is an essential element, then the parts of the judgment must necessarily fall outside one another in time. But is this necessary? If the essence of judgment is the very modification, _pari passu_, of subject and predicate, then time must be an essential element in it, but it is not at all necessary that its elements should fall outside of one another in time. In other words, the dilemma which Bosanquet points out on p. 87 is not a genuine one. There is no difficulty involved in admitting that the judgment is a transition in time, and still holding that its _parts_ do not fall outside _one another_ in time. His own solution of the problem--_i. e._, that, although judgment is an intellectual process in time, still time is not an essential feature of it, because subject and predicate are coexistent and judgment is a relation between them--involves a desertion of his dynamic view of judgment. He defines judgment, not as a relation between subject and predicate, but as an intellectual _act_.[53]
Take such an every-day judgment of mixed perception and inference as, "He is coming down stairs and going into the street." It is the merest chance whether I break up the process thus, into two judgments as united by a mere conjunction, or, knowing the man's habits, say, when I hear him half way down stairs, "He is going out." In the latter case I summarize a more various set of observations and inferences in a single judgment; but the judgment is as truly single as each of the two which were before separated by a conjunction; for each of them was also a summary of a set of perceptions, which might, had I chosen, have been subdivided into distinct propositions expressing separate judgments; _e. g._, "He has opened his door, and is going toward the staircase, and is half way down, and is in the passage," etc. If I simply say, "He is going out," I am not a whit the less conscious that I judge all these different relations, but I then include them all in a single systematic content "going out." (P. 89.)
But is it a question of merest chance which of these various possibilities is actualized? Is Bosanquet really looking--as he thinks--at the actual life of thought, or is he considering, not what as a matter of fact does take place under a concrete set of circumstances, but what might take place under slightly differing sets of circumstances? If it is true that judgment is a crisis developing through adequate interaction of stimulus and response into a definite situation, beginning with doubt and ending with a solution of the doubt, then it is not true that its limits are purely arbitrary. It begins with the appearance of the problem and its tentative solutions, and ends with the solution of a final response. It does, of course, depend upon momentary interest, but this does not make its limits arbitrary, for the interest is inherent, not external. In the case of Bosanquet's illustration, the question of whether one judgment or half a dozen is made is not a question of merest chance. It depends upon where the interest of the person making the judgment is centered--in other words, upon what is the particular doubt to be solved.
Ideas discussed in this chapter
Text from Project Gutenberg, public domain in the United States.