Bosanquet, Bernard. “The Relation of Coherence to Immediacy and Specific Purpose.” The Philosophical Review 26, no. 3 (1917): 259. https://doi.org/10.2307/2178546.
IT is a long time since Professor Sabine's article entitled "Professor Bosanquet's Logic and the Concrete Universal “appeared in The Philosophical Review.1 I communicated some observations upon it, noting its very appreciative and courteous tone, to Professor Sabine at the time; but was prevented by stress of work from writing anything on the subject for publication. And I should hardly have hoped that after so long an interval any reference to it would still be of interest to the readers of the Review. But the editor has invited me to recur to the question, and I willingly embrace the opportunity.
After giving a very just and generous account of my logical ideas so at any rate it appeared to me Professor Sabine expressed his intention of comparing the results of the coherence theory, as exhibited by my criticisms of points of view other than my own, with certain facts about the reasoning process as it occurs in ordinary experience (p. 550).
I will try to explain my attitude towards the facts thus alleged, taking them on the whole in the order which Professor Sabine has given them, but beginning with a general answer to a fundamental question which he puts at a later stage.
I. I begin, then, with the first of the two questions relating to the paramount importance which I give to coherence; viz. "on just what sort of evidence does this alleged primacy of coherence rest?" (p. 558 top). The second of these questions I defer.
My answer seems to me simple and definite. The primacy of coherence rests on the principle or general fact of implication, which I take to be the core of inference. I mean by implication the kind of relation which is detected by the insight which tells us that two straight lines cannot enclose a space, or that an equilateral triangle must be equiangular and conversely. The main point of it is that the relations, though bona fide different, are connected within a system which has a pervading nature, such that, if certain propositions within it are true, certain others must also be true; and the same applies to the reality of facts or relations. This principle is not assigned in my Logic the full and distinct importance which I now believe to be its due; but enough is said, I think, to make it clear that this was in effect my fundamental conception of inference. I hope to deal more fully with the question in the near future.
Realists, I believe, as a rule accept this principle in the same sense in which I do, though they maintain a principle of simple givenness beside it. There are others who deny it, in this sense, altogether, who deny, that is to say, that an inference by implication can be anything other than a restatement of the original proposition in other words. If so, of course no knowledge can be attained by such a process.
The evidence for the nature and existence of implication suppose is ideal experiment, or the experience of insight within complex distinctly given. The evidence of its primacy is the failure of logicians to produce any principle, not reducible to this, by which facts are ascertained or propositions supported in reasoning. The verification of it through the whole gamut of inference and affirmation is what I have attempted in my Logic under such names as the principle of coherence and systematic inference.
I. Abruptness or discontinuity, if it means richness of difference within a complex having a common nature, is of course again to our experience. But if it means a failure to be included in the total system of facts or truths which imply and are implied, it simply could not be apprehended; and in face of a world of such discontinuities (the word 'world' so applied involves an Irish bull), intelligence could never have developed.
II. "In general," Professor Sabine asks, "is not the tendency of idealism to discredit one phase of experience as an imperfection of the finite mind and to exalt another as an approximation to absolute truth the lineal descendant of the rationalist's distinction between necessary and contingent truths?" (p. 559). Surely not. The distinction between necessary and contingent truths is maintained in principle so long as givenness is held, as I understand that my critic holds it, to be a ground for acceptance of propositions as truth distinct and apart from implication. With the tracing of implication and the presumption of it through all grades of inference and affirmation, the distinction between necessary and contingent truths is finally and in principle set aside, as it is by post-Kantian philosophy in general.
2. As against the theory of coherence, viz., that facts or propositions can only be real or true by being implied in other facts or propositions within a systematic whole, Professor Sabine adduces two considerations on behalf of realism; viz.,
(I) that in all reasoning there is an immediate given and (ii) that every problem has its definite solution (and therefore, as I understand him, is not bound to exhibit coherence with the whole ad infinitum).
I. The immediate given.
(a) Professor Sabine writes (p. 552) as if our criticism of the immediate given had reference only to the supposed sensational in obsolete empirical psychology. It does not present itself to him as to me, in the light of the principal moral and result for logic which is drawn from the empirical sciences themselves. Against the time-honored prejudice that every inference must ultimately have an immediate premise, I advanced many years ago an argument based on analysis of the essential method to be seen in science itself.1 I pointed out that the scientific establishment of the simplest datum such as might be supposed immediate (e.g., the measurement of actual base line for a trigonometric survey) involved reliance on whole systems of the sciences, on the systematic implications of which the simplest observation depends. I pointed out for instance that optical instruments and instruments of accurate measurement are reasoning machines far more truly than Jevons ‘logical instrument, for they include in their structure all kinds of principles and deductions which enter, logically speaking, into the data they furnish. How is one assured that the image observed through a high power of the compound microscope is a correct magnification of the object under the lens? It is impossible to compare them. Your certainty is deductive, and by a pretty complex deduction, if only you knew what it was. It is the very nature of the sciences themselves, and no whim of the idealist philosopher, which suggests that they are not to be thought of as building on immediate premises as foundations,2but rather as something analogous to a solar system determined by reciprocal implications.
(+) But of course, something is in some sense given. The content of our real world comes from sense-perception. But what is it that is given and in what sense? What is given is simply the whole varying world of experience, a mixture of scientific certainty, sense-perception, error, imagination, and dream. This is our prima facie reality, which furnishes our problem. It is within this system, by insight directed to the implications which it contains, that minor systems are discriminated according to the values which their structure reveals, from the demonstration to the dream.
(+) It is this last consideration which enables us to deal with a really formidable difficulty. If comprehensiveness and consistency in a system of content is all that is needed to establish its truth as affirmed of reality, does it not follow that you can create reality at will, in conflict with any of your experience however broad and self-consistent, by imagining, as is alleged to be practicable in every case, a world yet broader and yet more self-consistent, which would by the hypothesis present the character of reality in a higher degree than the world commonly held to be the actual real? And is not this a reductio ad absurdum of the belief that reality is discriminated not by immediate givenness or apprehension, but by the completeness and consistency of the system of implications which it presents?
The first point to be observed here is that, for reasoning, facts can only appear in propositions; therefore givenness, or the nature of an apprehension, can only be expressed in the data of reasoning by further propositions than that which is to be supported by the appeal to such a character; and ultimately, if we insist on trying to push the argument home, by propositions ad infinitum supporting in turn the propositions which support the primary one. Thus, you cannot produce for inspection in an apprehension the immediate certainty which it claims. What you can do and the only thing you can do, is to describe and further describe, one after the other, the conditions which you believe to have excluded error, as in an experiment; and further, to argue in support of any of these, if their actuality is impeached. Thus, you establish wider and wider areas of implication which must stand or fall together, and the result is that classification of realities according to their relative truth and existence which makes up our total world.
Thus, then, the graduated nature of the whole of experience asserts itself, and there is no presumption at worst that you could weight the scale by imagination against the system
supplied by apprehension more than in the opposite direction. At worst the supposed worlds of imagination would cancel. But really it is more than that. Imagination as it departs from reality becomes narrow and thin; and the great source of content is experience, which would always assert itself as superior overall to or in imagination, even in comprehensiveness and consistency of content.
Mr. Russell himself, who has insisted on the objection in question, has also pointed out1 with one of his flashes of insight, that our scrutiny of our world of knowledge, though skeptically regarding every detail, is not skeptical as regards the whole. And therefore, I add, it is the object of thought to find within the whole such propositions as, if this whole is presupposed, must imply each other. And then, because we cannot doubt the whole which indeed is all we have of reality, we are forced to accept the implications which it imposes and to reject those with which these conflict.
Thus, we learn to discriminate, and reject the imaginary and fictitious by their discrepancy, both of general character and of particular detail, with the main bulk and volume of our experience. And it is most important in this connection that the imaginary fictitious and erroneous do, as the objection contemplates, occupy a very considerable part perhaps the larger part of the world which presents itself to us as a whole. But, contrary to the consequence threatened, we find no difficulty of principle in discerning them, and assigning to them a classification within reality in accordance with their defects. And this is really done by the character of their content, and its refusal to harmonize with the whole. The veracity of witnesses, for example, is determined overall by comparison of the content of their testimony with that of our knowledge in general; and even direct evidence of their character comes in merely as propositions which affirm conditions implying the truth of the purport of their testimony or the reverse.
A striking and typical example on a small scale of the real process of ascertaining truth and authenticity among a mass of various data may be found in the account recently given by a great scholar of his enquiry, pursued through many years, into the ideas and doctrines of Plato. "In this case," he writes, “the phenomena to be 'saved' (i. e., the given whole whose reality is to be ordered) are the writings of Plato himself and the statements of Aristotle and others who knew him, and the only proof or disproof the hypothesis admits of is its efficacy in accounting for them." The point of the instance, for my purpose, is that the authenticity of the several 'phenomena, ‘corresponding to the givenness of data in experience, has itself to be judged on the ground of their content, and in fact hardly any of the most important documents has escaped the impeachment of its authenticity by one good scholar or another. So that the precise problem, that of discriminating the fictitious from the real or actual on the ground not of givenness but of coherence, is here exhibited as it solves itself. The comprehensive and coherent real, within a world of mixed experience, becomes in fact a standard by which itself and the imaginary can be tested. On the other hand, there is no error so gross and palpable that it may not be found to be vouched for as given in immediate experience.
II. "This seemingly obvious fact that every problem does have its solution is the fundamental fact on which the realist rests his case against the coherence theory" (p. 553); i.e., as I understand it: if nothing short of the whole is harmonious, how come to an end of any problem?
But here I suppose we should ask with the pragmatist, who not only assumes the solution of problems but accounts for it, “to how many places of decimals is your final solution?" to which he replies, "to as many as you need ad hoc." Of course, you may have relatively distinct problems in science, but surely their finality is only provisional, and relative to subjective purpose. For example, I never heard of a science giving up business on the ground that its whole problem was exhausted. I cannot think that the realist who believes in science is true to his principles if he advances against coherence either the immediate premises or the final solution. I admit, indeed, that the presumptive implication, such as that of causation, which I believe to be in the background of a judgment like 'All men are mortal,' and indeed of all enumerative inductions and ultimately of all serious assertions whatever this I admit to be a favorite of my own not generally recognized. But even for it I am not without hope.
3. I pass to the objections from pragmatism. The specific problem or peculiar 'tension' is advanced from the pragmatist side (556-7) to show that the guidance of thought cannot lie in ‘a systematic nisus towards the removal of contradiction,' but must consist of a special need or purpose, which produces the existence of the problem and sets the standard of its success.
Against this I rely again on the spirit of science; and I further say that a good deal of fallacy is apt to attach to the expression ‘the purpose of the argument.' It seems plain that we must distinguish the private or subjective purpose, need, or tension which expresses itself through the demand for a conclusion of a certain type, and what I may call the logical purpose, which is one with the property or relation in terms of which a conclusion is desired. The former is a private motive productive of action, but when once expressed in the property which covers the desired conclusion cannot possibly have anything further to do with the argument. The latter determines the whole structure of the argument but is not peculiar to any single motive or tension, and, when once determined, cannot be influenced in its inferential development by any such consideration.
I will give a very simple example of the relation between them as I see it; for I do not at all feel sure that Professor Sabine and I have in mind the same kind of case.
I may have any of a hundred or a thousand motives for desiring to have the exact distance measured between the site proposed for my house and the high road across the common. Shall I be out of range of the noise of motors? Shall I have room to drive a golf ball in that direction? The law forbids me to fire a gun within so many yards of a high road. How much clear space have I before me that can never be built on? These are private purposes, tensions set up by my needs, which furnish motives for having the problem solved. But for all of them theological purpose is one and the same, the qualification of reality by a distance to be measured between certain points. Now compare my private motives with my logical purpose in relation to the attitude of the surveyor who makes the ordnance map. The logical purpose is taken up as it stands into the purpose which dominates his work. The private motive wholly disappears and becomes irrelevant.1 But more; the impersonal aim and duty of the mapmaker takes shape in solving not merely the problem of my logical purpose, but the same problem, as a single thing, as between every point and every other point over a whole area; and in solving all these problems at one blow embodies and satisfies a true scientific curiosity the problem of a precise measurement of the district as a whole. Now of course it is an easy joke to insist on all the private motives which may conspire in causing the map to be made, and to point out how the mapmaker himself may be working for his salary and not from a scientific interest. But his work obviously can represent, and on the whole and forever does represent the scientific curiosity which is the mind's demand to have a certain system of truth, excluding innumerable contradictions, cleared up and determined once for all. Anybody can take his profit from it where he finds it. But in knowledge the true and widest reaching motive is as Mr. Russell has well said the motive of scientific curiosity. It is the impulse of the mind to know, by which all private motives and unique tensions are superseded; while all the logical purposes marshal themselves in the order of scientific method and not of chance desires.
If indeed all that is meant by insisting on the unique problem were that in science you must always be at work upon some determinate question, that would of course be true. But the suggestio falsi as I see it is that you are always at some given point which is a limited point; and is so because of the limitations of the special or unique interest which at the moment guides your attention to it. Whereas in the first place it need not be a limited point in the sense in which a special occasion of interest is limited. It may be a fundamental issue affecting a wide range of knowledge and including within its myriads of such specific occasions as arouse at a moment a particular interest. If, further, it is a somewhat specific problem, it need not be taken up because of a specific interest in that problem or a situation leading up to it, but merely because, in the satisfaction of a much more general interest through a much more general enquiry, that particular problem stands now to be dealt with in the order demanded by progressive method.
4. On the pragmatist's objection to epistemology, and his charge that "idealist logic is mainly concerned with the impossible problem of the relation of thought as such to reality as such" I can now refer to my paper on "Realism and Metaphysic “in this Journal for January 1917, especially p. note 2. I may further quote from Mr. Bradley's Essays, p. 117: "And thus, if we are asked for the relation of truth to reality, we must reply that in the end there is no such relation, since in the end there are no separate terms." I am indeed not sure what the charge in question is, unless it were that of making it a problem how you can get from thought to reality, a point of view absolutely abhorrent to and denounced by every student who sympathizes with post-Kantian thought.
5. I now take up the second question on pp. 558-9, the old question of a timeless reality.
I begin by objecting to the expression that change, and creation are to be 'superseded' by metaphysic the word recurs on the following page. That change is to be subordinated within a whole which as such does not change is in my judgment true; that its existence is to be conjured away and replaced by something else has not been suggested. Coherence, further, is a term which I use in speaking of truth, and as applicable to what we know of reality through dealing with it as a system of terms and relations. One would hardly employ the term in speaking of a great work of art; it would fall short. Still less would one employ it in speaking of the ultimate experience. This experience, as Professor Sabine points out below, is not for us the completion of knowledge in its own shape as knowledge. It is something beyond which we believe that a convergence of analogies and the resolution of many discrepancies enables us to divine.
It is easy to explain why new reality, as a bona fide addition to the universe, seems to me a contradiction in terms. A phase of being, which has outside it a past and future, can obviously not itself be a universe. Whatever a universe is, it is surely the whole of reality; and that which has some reality beyond it, is certainly not the whole, but is something that falls within it, and, as demanding completion, might be set down with tolerable certainty as contradictory within itself. If reality is unending in real time, at any moment there is no universe. If it is finite in time, the universe embraces its whole decursus vitae. In any case that which has addition made to it cannot be the universe.
"Even so fundamental an experience as moral development has to be superseded" (p. 560). The note of climax here is suggestive. Moral development is a good example of the kind of experience, which is subordinate in the universe, I should have said, because it is a development which is thoroughly dependent and self-contradictory and incapable of self-existence. It lives only within the social and religious experiences of which it is a subordinate aspect; a purely moral being could not exist. He would have no realized world of goodness but would be cut sharply in two between the 'is' which ought not to be, and the 'ought to be' which is not. Anything more fitted to be a subordinate aspect of a fact such as the social or religious faith and will a solid unity which is the reality that makes it possible, could hardly be conceived.
It is plain that in any case change is here a subordinate aspect of an inclusive experience, and considering the contradictions involved in a past which is really past and gone, phenomena like this point the way to an ultimate subordination of change; though we cannot be put in possession of the actual experience in which its subordination and its reality are reconciled.
6. The topic of selective attention. "The clearness and efficiency of consciousness depend upon it" (p. 561). So do system and relevance themselves (p. 562). "The fact that our interests are limited (I am represented as saying) shows the contingence of our minds, but singularly enough, the interests that we have show our participation in reality" (p. 561). "So far from the elimination of selection being a sign of relevance in experience, as Professor Bosanquet seems to believe, it is quite the reverse" (ibid.).
I am surprised at this attitude, which I should have thought was obsolete since Kant's Critique of Judgment. Either aesthetic or organic experience offers a point of view from which the absolute fusion of whole and part, of end and means, is inevitable. Even Teleology, Professor Burnet points out, has not really to do with Telos as an external end, but with Telos, 'complete.' The sharp contrast of means and ends always betrays the standpoint of mechanism. To refer to a state of semi-coma as the only case in which selection is absent is surely to caricature the totality of feeling, and to misconceive the analogy, which is drawn, by help of experiences like that of life and beauty, from the mere feeling which is below relational knowledge to a unity which is above it. Once more we find the problem ignored of divining the character of a reality transcending the experience which a finite being can expect to possess.
(+) I do not altogether apprehend the argument to be drawn from my reference to our limited interests. If our interests exhibit our participation in reality, how should their limits not display our limitations? Is it implied that the fact of our having specific interests at all betrays a limitation in our selection? But surely the more we are absorbed and the less we negatively select, the greater is our participation and the less the contingency of our minds. We may consider for example the religious consciousness where the finite being tends to merge himself wholly in what he worships.
(+) "The question is whether apart from selection, emphasis, teleological interrelation, there is any way of conceiving a whole"(p. 562). "Hence the pragmatist's preference for specific purposes" (ibid.).
I hope that the answer to this objection has been furnished by the reference above to a whole such as forms the basis of implication, and also to the even higher type of wholes which are familiar in life and beauty. A nexus of relations such as we have in a Euclidean construction is not a purposive whole, but it is a complex on the determinate pervading nature of which complications can be based.
(+) The further answer which I should make to the last-named objection is set out in effect by Professor Sabine himself, in his account (pp. 562-3) of the development which I have found necessary in the analysis of systematic inference, from real teleology to the idea of individuality. It is true that I now believe that the development of the theory of the concrete universal requires the absorption of purposes "in a system so complete that the specification of any part as an end is impossible"(p. 563).
Or rather, the purpose qua purpose is negligible, as we saw above. It is the nexus of relations held together by the distinctly apprehended whole which determines the implications of one in another and vice versa. And in such a nexus, in proportion to the completeness of its interconnection, no part can be idle; and if it presents on one side the aspect of a purpose say, like the total embodied will of a society it must be a purpose such that every so-called means is a modification of the end, and every feature of the end imparts a character to something which might also be called a means. You may find an inferior nexus, an incoherent one and an inferior purpose, which governs only some elements of its means, while others fall apart from it. But these are far below the logical or metaphysical ideal.
7. The ultimate breakdown of the coherence theory is said to be admitted on my side (p. 564).
That is to say, I accept the position that it deals only with the conditions of truth and does not profess to put us in possession of the ultimate experience.
"But is there not," Professor Sabine exclaims (ibid.), "an ultimate improbability in this representation of truth as an ultimate effort to do something which it never can quite do?"
I am afraid I feel strongly the opposite. I think that in all modern philosophy there is no theory more convincingly felicitous than this. We are always, I take it, to remember, as the first condition of the problem, that we are treating of a finite being, and cannot possibly claim for him an experience of an ultimate character. Starting from this point, it is a good deal, I think, to be able to show, that the form in which reality reveals itself in his knowledge in the first place is precisely suited to his finite method of thinking by terms and relations, and in fact consists in pursuing it; and secondly, that it recognizes throughout and indicates by its very spirit and endeavor a completion toward which it declares itself to aspire, but which, in its proper perfection, finite thought self-evidently can never attain. I should have said that these two sides of finite knowledge are most happily confronted and reconciled by the theory in question.
8. "Is Professor Bosanquet's admission of limitation (in truth as compared with reality) so entirely different from the dualism of thought and its occasions which is set down to the discredit of pragmatism?" (p. 564).
Well, is it not? What I admit is a modification in passing from one type of experience to a type kindred with it but more complete, the aspiration towards which is decisively indicated by all features of the former, and by all considerations which bear on the relative character of the lives or spheres to which they respectively belong. What I find in pragmatism is a needless break within our finite knowledge, with an attempt, which seems to me to look something like a parti pris, to deny the pervading unity and constructive operation of the disinterested and methodical curiosity which I believe to be an inherent aspect of man's spirit and to be akin to its noblest features.
I cannot break off without expressing once more my sense of Professor Sabine's courtesy and appreciativeness. That he should deal at all with my book, in such a tone as that in which he did deal with it, was much more than I had a right to expect. It is plain that we apprehend some matters differently, and I hope that the present paper may at least contribute to our seeing more precisely how and why.