previous
next

In my article “Empiricism” (A.J.P.P., December, 1927)note I presented it as a mark of empiricism that it rejects any doctrine of different kinds or degrees of truth and reality and maintains that there is only one way of being (describable as “being a matter of fact” or simply “being so” or “being the case”). I argued that, in the distinction between empiricism and rationalism (with its division between facts and principles, between actual things and their “grounds” or “explanations”), the question of ways of knowing is a quite secondary matter, though the denial of distinct ways of knowing has still to be recognised as a feature of the empiricist position. It is, in fact, quite illuminating, of the particular question of knowledge as well as of the general question of reality, to present the matter from the side of knowledge and take empiricism as the doctrine that whatever we know we learn — in other words, that to know something is to come into active relations, to enter into “transactions”, with it — a position which at once rules out any rationalist notion of ultimates or principles above the facts, any suggestion of “that whereby” things exist, as something distinct from the things themselves, since, unless we were acquainted with it, had acquired empirical knowledge of it, we could never infer it from what we are acquainted with or assign it any way of operating on objects of our acquaintance.

More broadly, it might be said that we cannot uphold any doctrine of kinds of reality, since to do so we should have to know the distinction or the relation between any two such kinds, and that is something we could not know except as a single situation — which would mean that we knew it as of a single reality, so that the doctrine of distinct kinds of reality would be automatically abandoned. It is in this way that empiricism is seen as a doctrine of what is real as situations, and that therewith goes the denial that anything can be known except as situations, which is to say except as spatio-temporal and except in propositional form. I have argued at the end of the preceding paper, “Relational Arguments” that, important as the distinction is between qualities and relations (impossible as it is to regard, e.g., “A is after B” or “A is beside B” as a description of A), the fact remains that to find a situation is to find a tissue of qualities and relations, a “nest” of situations or propositions. Thus there is no question of separate regions of the existent or the knowable, but the question is always of complex and interrelated states of affairs; and recognition of this makes it possible to have a coherent


  ― 163 ―
view of the acquisition of knowledge of any subject-matter, in a way that cannot be done on a rationalist or “separatist” position.

It is remarkable that Locke, Berkeley and Hume, so widely regarded as the founders of modern empiricism, should take their departure from just such a rationalist doctrine of simple and separate entities — the “ultimates” by reference to which any actual state of affairs is to be explained. It is no less remarkable that these entities should be called “particulars” when, as “whole natures”, as “nothing but” yellow or a shade of yellow, etc., they would appear rather to be “universals”, to be merely descriptive, and to say that their simple nature was of a non-descriptive character would be to say that they were indescribable — and indistinguishable. But, it being on a situational or propositional view of things that such a division, or any belief in simples, would be denied, it is not remarkable that the “British empiricists” are unable to give any coherent account of complexity — of propositions or of relations. (Of course, from an atomistic starting-point, the most that propositions could be would be relations — of juxtaposition, etc.)

The position can perhaps be best illustrated from the work of Berkeley. Of the three he is the most openly committed to “natures”, particularly to the treatment of the “idea” as that whose nature it is to be perceived — the logical outcome of which is the impossibility of distinguishing one idea from another. Yet he is definitely more empiricist than Locke and Hume in taking the truths of mathematics to be as “matter of fact” as those of any other subject, taking them as something that has to be learned just as anything else has, and in standing out the most decisively against the treatment of what we are acquainted with as “standing for” (presenting us with knowledge of) something we are not acquainted with at all. Nevertheless, in spite of all his empiricist strivings, he gets himself into a complete impasse in his attempts to deal with situations.

The failure of Berkeley's attempt to account for “the objects of human knowledge” in terms of atomic “ideas” already emerges in § 1 of the Principles when, having spoken of the ideas furnished to the mind by the various senses (what these are being not itself explicable in terms of simple ideas), he goes on to say that “as several of these [ideas] are observed to accompany each other, they come to be marked by one name, and so to be reputed as one thing. Thus, for example, a certain colour, taste, smell, figure, and consistence having been observed to go together, are accounted one distinct thing, signified by the name apple. Other collections of ideas constitute a stone, a tree, a book and the like sensible things; which, as they are pleasing or disagreeable, excite the passions of love, hatred, joy, grief, and so forth”. Concomitance is only one of the relations which Berkeley “annexes” to his supposedly ultimate objects of knowledge, yet it is clear that to observe such going together is to have as an object (cognitum) a single and at the same time complex situation, and, if an “idea” can be found in such a situation, knowing it is not the same as separating it from all that is not of its nature. If it were, “going together” would at best be another separate object or piece of content,


  ― 164 ―
and there would be no such situation as “the going together of A, B, C and D”; but, if it is not, the knowing of various qualities in various complex situations shows neither the necessity nor the possibility of any such separability (or of any absolute “original”) of any one of them — it brings out no meaning, in fact, in the notion of taking one of them by itself, taking it in any other way than in a complex situation.

This, of course, would be a spatio-temporal situation, for the fact that various contents were near in time to one another (though even that would be unintelligible on a doctrine of distinct atoms of information) would give us no reason for regarding them as making up “one distinct thing” or total content, but spatial togetherness would also be a condition of any such “unification”. Over and above that, however, it should be noted that Berkeley's concomitants are descriptive terms and their concomitance would be a conjunction or addition of descriptions, so that their “unification” would be not merely of a spatial and temporal but of a predicative character. This point is reinforced by the fact that Berkeley's argument gives us no suggestion of any purely “particular” idea but only of classes or kinds of ideas (colours, tastes, smells, etc.) so that such ideas would differ in various ways from others of the same kind and, as thus complex, would not be ultimate units but would themselves be situational or propositional. Thus our recognition of distinct complex things is not accounted for at all by “collections of ideas” (of separate, unitary pieces of content) but is intelligible only as a recognition of complex situations, of situations within situations (in which terms alone “concomitance” can be understood), of interpenetration as well as juxtaposition — in other words, of infinite complexity (with no least and no greatest situation) in place of the “simplicity” which cannot be squared with any complexity or combination.

It may also be noted that Berkeley, in the final sentence quoted, admits situations in which collections of ideas are related, in the way of “exciting”, to mental situations (those of the passions and operations of the mind) — something that is possible only if objects of observation are “on the same level” of reality as mental operations, instead of having the one dependent and the other independent existence. This is to say that the empiricist position in which any coherent account of observation must issue has also to be a realist position, that the rejection of any doctrine of ultimates and derivatives carries with it the rejection of any doctrine of “ideas”, of existence relatively to minds as much as (or as one case of) existence relatively to what is “self-existent”. But of course Berkeley does not recognise this; and, in spite of the soundness of much of his criticism of “representationism”, of any view of an observed content as deputising to our minds for an unobservable content, he himself is not free from representationism, as no one can be who recognises the “idea” (that whose relationship to what knows it is inherent in its own content) in any shape or form. And an essential point here is that, if we find in our minds “a collection of ideas”, what we find (the combination of A, B, C and D) is just as much a matter of fact, something that just is so and not


  ― 165 ―
something that “is dependently so”, as any situation that could be found “existing absolutely”, in minds or out of them.

The untenability of any doctrine of “dependent existence”, then, whether the dependence is supposed to be on the mind which is aware of the “existence” or on anything else, is a particular case of the impossibility in any coherent theory of admitting realities of different kinds or orders. The incoherence of Berkeley's anti-realist position, which is a part of his anti-empiricist position, appears most strikingly in §§ 30 and 31 of the Principles — headed, in Lindsay's text, “Laws of Nature” and “Knowledge of them necessary for the conduct of worldly affairs”. We are told (§ 30) that laws of nature are “the set rules or established methods, wherein the mind we depend on excites in us the ideas of sense”, and that we learn these laws by experience “which teaches us that such and such ideas are attended with such and such other ideas, in the ordinary course of things”. And again it is said (§ 31) that knowledge of such laws “gives us a sort of foresight, which enables us to regulate our actions for the benefit of life. And without this we should be eternally at a loss: we could not know how to act on any thing that might procure us the least pleasure, or remove the least pain of sense. That food nourishes, sleep refreshes, and fire warms us; that to sow in the seed-time is the way to reap in the harvest, and, in general, that to obtain such and such ends, such and such means are conducive, all this we know, not by discovering any necessary connexion between our ideas, but only by the observation of the settled laws of nature, without which we should be all in uncertainty and confusion, and a grown man no more know how to manage himself in the affairs of life than an infant just born” (Berkeley's italics in both the above places).

Now Berkeley is anti-representationist in so far as he maintains that “our ideas” (the objects of our observation or experience) are actual things and not merely ways in which some of the features of an unobservable “matter” are conveyed to us, but in so far as he takes them as relative to our minds he is not treating them as actual and is forced into representationism in order that our knowledge may be taken as in any way confronting us with the actual. As ideas of ours, our successive sensations are dependent on us, but their regularities, as instituted by God for our guidance, are quite independent of us, so that our knowledge of “laws of nature” is knowledge of objective facts. And the only support Berkeley can give for the adhibiting of objectivity to part of what we experience is that our sensations are not under our control (an objective fact of which it seems we are directly aware) whereas our “images” or reproductions of sensations are (another fact, independent of our knowing it, which we can discover); so that Berkeley's position is a hotch-potch of realism and representationism, an unavoidable result so long as “ideas” (entities of a different order of reality from minds or agents) are retained.

Further confusion is disclosed by a more thorough consideration of what can be meant by “the conduct of worldly affairs”, by the regulation


  ― 166 ―
of our actions through attention to “laws of nature” such as that food nourishes and fire warms. Such utilisation of our knowledge involves the securing of results that we desire and the avoidance of others that we object to, and this avoidance can only mean that, on having a sensation that is normally the sign of another sensation, we act so as not to get that second sensation, so as to avoid the experience which the first “signified”. But that is to say that we actually falsify the supposed regularity which we were said to be using as a guide — that, e.g., we get the visual sensation of a fire and then do not get the tactual or cutaneous sensation of being warmed or perhaps burned; in other words, the conduct of life, as an avoidance of certain sensations, would mean a contravening of the “laws of nature”. In fact, however, when we make such avoidances, we do not think there has been any interference with natural regularities; we believe that fire, food, etc., operate as they always did, even if some of their effects are not sensed by us. And this belief is possible only because we distinguish our perceptions or sensations of things (the fact that on certain occasions certain things are perceived by us) from the things themselves, only because we do not identify the effects of fire, e.g., with the “sensations” which ensue upon a “sensation” of fire, only because we distinguish our knowing of a sequence (though this knowing too is an objective fact) from the occurrence of that sequence — only, i.e., because we take a realist view. But if we did not take that realist view, we should have to abandon the recognition of regularities that Berkeley speaks of and thus to deny any possibility of the conduct of life through knowledge of them.

The conduct of life, however, does not involve only that there are ways of working of things we observe, independently of whether we observe them or not; it also involves our working on them — our eliminating certain things and so the effects they have been producing, our instituting certain things, as contrasted with merely seeing signs of their approach. Now on Berkeley's view that our sensations are not under our control and that all we can control is the reproduction of them in imagination, the conduct of life would be restricted to such imaginative exercise while the sequence of sensations proceeded regardless, and we should not be making any use of our knowledge of the laws of such sequence. But, even if it were supposed that we could do so, i.e., that we could take such action as to avoid some sensations and secure others, it is not apparent how such securing and avoidance could make us any less “at a loss” than we should otherwise be. For, on the doctrine of active mind and passive ideas, our sensations would merely confront us, like images on a screen, and it would not make the slightest difference to us whether one or another was there; we could have, say, the sensation of extreme cold, or again that of gentle warmth, or again that of burning (even what is called the burning of our flesh), but which of these was present would not matter to our minds.

Thus, when Berkeley speaks of food nourishing, sleep refreshing and fire warming us, he should mean only that they are signs of other ideas,


  ― 167 ―
of the sort, no doubt, that we call “organic sensations” but having, on the doctrine of ideas, no more intimate relation to the percipient than ideas we do not so describe. But the notion of “the conduct of life” implies that these objects do affect our very selves, that they are not mere data but things that make a difference to our activities. Berkeley to some extent obscures this position by talking of our being affected by pleasure and pain; if this means only our having perceptions of pleasure and pain, this will still not bring us any nearer to the active mind (the perceptions would still make no difference to it), but if, as seems to be the case, pleasure and pain are to be taken as features of mental processes, then their being procured by the securing of some sensations and the avoidance of others can only mean the operation of what is sensed itself on the mental. In any case, unless the sensible has direct effects on us, there can be no conduct of life by means of recognition of its regularities. But, if there is direct interaction between minds and what they perceive, that means that minds themselves come within the range of laws of nature, that they and not just what they contemplate are parts of the natural realm, that there are regular sequences between, e.g., the provision of food and fire and various types of mental occurrence; and without this interconnection, without continuous processes between the mental and the non-mental, there would be no such thing as “affairs of life”, no relevance of “laws of nature” to the sequence (if such there could be) of mental activities.

Even Berkeley's account of the conduct of life, then, is unintelligible except in terms of interaction between minds and things they observe (which, of course, include minds), i.e., except as implying the unintelligibility of any assignment of different sorts of reality to the mental and the non-mental, except as recognising both (besides the relations between them) as equally matter-of-fact, in precisely the same sense situational or propositional — a sense which is realist, recognising independence, the existence of many distinct things in any situation, and empiricist, recognising a single level, on which only situations can and situations always do connect situations. But, while consideration of his inconsistencies leads to the showing up of any attempted separation of mind and nature, Berkeley remains sunk in them just because of his rationalism, his doctrine of natures or elementary entities, the discontinuity between which is set aside by the postulation of impossible leaps, such as he himself has shown Locke's representationism to be. For, in spite of all his efforts to find real connections between minds and what they contemplate, he also, as we have seen, makes the representationist leap; it is something he is forced to do by the doctrine of “ideas”, which must be taken as at once “in” minds and “of” things, leaving us with the insoluble problems which arise on any doctrine of relative existence.

Dualism, disconnections, ambiguities and insoluble problems, can, of course, be brought out just as readily in the theories of Locke and Hume as in that of Berkeley. The point is that on the supposition of elementary ideas it is impossible alike to have any knowledge of them


  ― 168 ―
singly and to have knowledge of any relations among them; we cannot know the absolutely separate but equally we cannot know the togetherness of a number of separates — as argued above, we can know only the togetherness of complex and situated things (things having situations within them and situations around them and themselves being situations). Thus when Locke speaks of our knowledge of “agreement” among ideas, he is (apart from allowing the symmetrical relation “agreement” to do duty for the A proposition which is not, as such, reversible) making possible knowledge of what is in the realm of ideas by recognising the occurrence there of complex states of affairs, but, first, is thereby leaving no room for their supposed elementary constituents and, secondly, is taking all point from the notion of a realm of ideas, since what we know is that A does objectively or as a matter of fact agree with B (assuming that agreement can have a definite meaning here) and there is no conceivable point in saying that that fact further agrees with or “corresponds to” something “in reality” — it is already “in reality” in being a fact.

In line with the general argument against dualism, the distinction between an “inner” and an “outer” reality (between mind and “the external world”, as the phrase goes) is one of which no account can be given, since there will be no reality for that distinction to have. In more detail, to say that anything is external to mind is to say that it is part of a situation which includes mind, and thus to say that it does not constitute a “world” but is just such and such situations related to (in common situations with) other situations; again, to say that A is external to B is to say that B is external to A and thus, just as much as A, exists under conditions of externality (and this may remind us of the fact that what is external to any mind includes minds); and, finally, to note (as has been done above) that what is called “inner” reality contains complex states of affairs is to indicate that it has within itself relations of externality and thus that there can be no “world” (apart from the objections to that notion itself) to distinguish from an “external” one — there will be only variously qualified situations (mental and otherwise) variously disposed spatially towards one another. But if there could be separate kinds of reality, “inner” and “outer”, there could be no relations between them, and when we look closely at supposed relations like “correspondence” or “agreement”, we find nothing that can consistently be meant by them.

We cannot say that X “corresponds to” a fact unless, as I have indicated, X is already a fact, and then how it could be regarded as “standing for” (or being “of”) some other fact is not apparent. We cannot even say that X corresponds to something unspecified without implying our awareness of this whole situation, but there is no suggestion in that that this object of our awareness stands for or corresponds to something else. We cannot say that knowledge of X (something “inner”) “claims to be” knowledge of Y (something “outer”); if this claim were supposed to reside in some resemblance between X and Y (a resemblance which, on the doctrine of “the external world”, we could not know — though, if we could, our knowledge would be directly of X's resembling Y and not of


  ― 169 ―
anything that that could be supposed to “correspond to”), the claim could be made with much more force that knowledge of X is knowledge of Z, another “idea”, the resemblance of which to X we could, on the doctrine of an “inner” reality, be aware of. Or, putting it in terms of “agreement with reality”, we can say that X agrees with Y only if the knowing of X is the very same thing as the knowing of Y (if X and Y are the very same state of affairs); we cannot maintain a dual theory of knowledge and say that knowledge of the situation X is “somehow” at the same time knowledge of a different situation Y. The only solution is to reject the whole doctrine of “inner” and “outer” and to maintain the realist doctrine of a direct knowledge of propositions, of mental as of non-mental states of affairs, which are variously related but none of which can be said to mediate others to our minds.

The force of this propositional or situational realism is commonly concealed by the treatment of the proposition as a tertium quid or mediator, something by which we can assert facts but which is distinct from the facts as well as from us. This, however, is as untenable a view as the doctrine of correspondence, of which indeed it is just a variety. When we assert the proposition “All men are mortal”, what we are asserting is the actual mortality of men, and to call the assertion of the proposition merely a means to the asserting of the fact is to say that we have no way of asserting the fact, just as we have no way of specifying the “reality” with which certain ideas of ours are supposed to “agree” unless those “ideas” (what we know) are the reality. I have spoken in a misleading way on this matter in “Empiricism” A.J.P.P., December 1927, p. 242),note i.e., as if I regarded the proposition as a tertium quid, in saying: “The empiricist, like Socrates, adopts the attitude of considering things in terms of what can be said about them, i.e., in propositions” — when I certainly did not think that there was any other way of considering them, that “things” could be other than propositional (or situational) in their content. I had indeed said earlier (in my discussion paper, “Propositions and Judgments”; Mind, vol. xxxv, N.S., No. 138, p. 239)note that propositions are not “about” anything (i.e., that, as above, the proposition which is commonly said to “assert a fact” just is that fact). I there allowed a loose sense in which a proposition could be said to be “about” its subject; which is connected with my formulation (arrived at very much earlier, but presented in “Hypotheticals”, A.J.P., May, 1952, p.3)note of the “function” of the subject of the proposition as location, as giving a “point of reference” at which we can check the truth of the proposition, in contrast to the predicate's function of description. The fact that any term can have either function (is both locative and descriptive) brings out the fact that the term also is situational in its content, but to be presented with a single state of affairs that we can consider we have to be presented with a proposition; it is a single issue, not a group of issues, and is not “about” any other issue.




  ― 170 ―
The distinction of functions is important (as I have argued in “Hypotheticals”, pp. 3-5)note in the criticism of certain views of the import of the proposition or of possible propositional forms, and also in the consideration of “the false proposition” which has been taken to involve a fatal objection to non-representationist views. The position is, just as in the case of “correspondence”, that if a “false proposition” were simply a certain contemplated content, it would not be false; it would be a state of affairs that we found to occur — and not, as in the similar case of “ideas”, specifically “in our minds”, but just as constituted in a certain way. What could be considered false would be the “claim” of this content to represent or reproduce a different state of affairs, which is to say another proposition; but this is a claim which is not made and could not be made, any more than the “claim” of “ideas” to “correspond to” (or agree with) “reality” can be made. The question, then, is not of the occurrence of a state of affairs with the attribute “falsity” (any more than of a state of affairs characterisable as “not corresponding to reality”); what is meant by the occurrence of a “false proposition” is explained, by reference to the distinction of subject and predicate, as someone's mistaking X for Y (taking X to be Y when it is not) — the question of this threefold relationship not being one that the person who is mistaken intends to raise, and not arising when he is not mistaken, when he is presenting the single situation X is Y.

This predicament is made possible, in the first place, not merely by our knowing X and Y in other propositions but by the fact that they are themselves complex or situational, that the content of a term can be set out in propositions, as when we define A as BC and thus identify its content with that of the proposition BiC (the position is really more complicated — the I is only one of a group of propositions making up the definition); secondly, by the fact that there are different formal possibilities of the “propositional connection” of two terms, that, e.g., knowing situations of the A and of the O forms we can consider whether XaY or XoY is true, without, for the time, being able to decide (though, when we do decide that XaY and thus call XoY “a false proposition”, this, as said above, does not mean that we are recognising a “false situation” but the logical form of “O is false” is A or, as it is commonly put, is “the truth of the contradictory”); and, thirdly, as I have argued in some detail in “Mind as Feeling” (A.J.P.P., June, 1934),note by the fact that knowing is a matter of learning or finding out, that we may be seeking or demanding the occurrence of the predicate Y in the subject X, wishing to introduce, or simply desiring that there should be, this character in a place which we otherwise identify — in particular, by other predicates. No coherent position can be got by dividing the class “propositions” into two species, “the true” and “the false” (the species of “what is” and the species “what is not”) — there is no parallel between this and the division of animals, say, into the male and the female — and the treatment of every proposition as a subject of inquiry and dispute (of seeking and mistaking) is opposed


  ― 171 ―
to the treatment of the proposition as a tertium quid, which would make any discovery, any outcome of discussion or inquiry, impossible.

The rejection of the “tertium quid” view leaves us with a propositional view of reality, which, without fully canvassing the matter here, I have been indicating reasons for identifying with a situational or spatio-temporal view. My general thesis has been that there cannot be separate kinds or realms of reality, and I take the separation of the situational or propositional from the non-situational or non-propositional to be an example of this — though a case might be made out for taking it as the true content of all such separations. I have contended that the rejection of “separatism” of this kind is characteristic of empiricism, even as a doctrine of knowledge, i.e., of the conditions of our becoming acquainted with things, though this is secondary to the characterisation of what we are acquainted with as situational, as “complex and interrelated states of affairs”. Another way of putting the matter is that empiricism is the doctrine of the continuity of things — a continuity which would be broken if we could admit different kinds of reality, notably elemental or ultimate as against secondary or derivative being. Professor Gilbert Ryle who, in an article entitled “Logic and Professor Anderson” (A.J.P., December 1950), criticises my position as set out in a number of A.J.P.P. articles, and particularly endeavours to show the inadequacy of my situational view of things, seems to me to fall conspicuously into errors of discontinuity, of dualism or division in reality.

Ryle takes me to have carried out with some success “deflationist” arguments against “ultimates” and “levels of reality” but considers that in the views I advance, particularly, of mathematics, of good and of implication, I am telling just as “impossible stories” (p. 142) as those I have detected in the work of Plato, Locke and other philosophers. My contention is that in these views (“stories”) I am pursuing the same empiricist line (rejection of divisions in reality) as I do in my “deflating” generally, and that the exposure of such divisions is all that “deflating” can mean. Thus my position on implication (cf. what I have said above in “Relational Arguments”) is that we have to recognise implication among propositions in the same way and by the same acts as we recognise the propositions themselves, i.e., in a continuous situation with them; that the alternative is a complete disconnection between the observed facts and what is supposed to relate them, and thus a quite arbitrary view of what implies what and of what implication is; that only on a view which rejects such divisions, which will not put implication outside the continuum of situations — only on an empiricist view, denying that implication is a special, a “rational”, kind of fact — could it ever be learned that p implies q or that r and s imply t. In the same way, it is a condition of our getting to know mathematical truths, and of their having any relation to, any continuity with, other truths — a condition, e.g., as I mentioned in “Empiricism” of the mathematical having any “application” to the physical — that they should be discovered in the same situations and by the same experiences as those other truths; and, if this were never


  ― 172 ―
so, mathematical truths would have to belong to a different realm of reality from physical truths, and presumably from what is true of ourselves, so that they could never be learned. Similarly on good, a question to which I shall return, I have attacked doctrines of a dualistic kind, and it is not hard to find dualism in the doctrines that Ryle sets against mine.

It is particularly where he touches on categories that Ryle shows the “separatist” or atomistic character of his thinking. He accuses me (quite wrongly) of having such an “exiguous logical alphabet” that I recognise only two categories where Aristotle recognised ten in line with his earlier contention (p. 143) that I have a logical dichotomy “Quality or relation?”, whatever is not a question of one being a question of the other. I think it could easily be maintained that questions of both quality and relation are features of the questions raised under any other category (pretty obviously, e.g., both arise in questions of causality or the continuance of one situation into another); but the vital point is that, while a question of the occurrence of a relation is not a question of the occurrence of a quality, it is only by seeing these different types of question together (as they must be seen in any concrete inquiry) that they can be thoroughly understood. Ryle does not specifically say that he takes questions of quality and questions of relation (and so on with whatever other categories he recognises) to belong to different regions of inquiry; but he gives no sign of seeing that the logician is concerned not with a miscellaneous bunch of types of question which can be raised about this subject or that, but with a group of types of question which have a common ground, which hang together in any inquiry and thus apply to any subject- matter. This, I would say, is the ground of Space and Time (or of being situational) in terms of which the universal application and the interlocking of logical questions appear. It is because questions in all the categories are spatio-temporal, because they all arise within any region or “contour”, to use Alexander's expression, that they are not discontinuous with one another but all form part of a common inquiry (not, of course, an inquiry into everything but inquiry into any specific subject, it being remembered that subjects are not cut off from one another but each of them embraces relations among subjects). Apart from such a common ground, there would be no such thing as logic, no sort of connection between one inquiry and another, and thus no inquiry.

Now just as there are not separate regions of relational facts and qualitative facts but all situations embody both (recognising a situation might indeed be equated with recognising both qualities and relations), so there are no separate regions of the universal and the particular, but any situation exhibits both particularity and generality. This is indicated by the fact that any proposition has both a subject and a predicate or is both locative and descriptive; but here the “convertibility” of terms has to be remembered — there is no such thing as the purely locative or the purely descriptive, there are not two classes of terms, subjects and predicates. Thus, over and above the fact that Ryle's division between universal and particular “statements” would prevent there being any connection


  ― 173 ―
between universality and particularity, we find that he uses “particular” in the confused sense of something “unique” or “individual”, thus making “a particular” something indescribable and not realising that there is no such thing as “a particular” but only “a particular X”, i.e., some subject of which X is a predicate but which can itself be a predicate and thus is a descriptive or “general” term. Thus there is not the antithesis he postulates, and takes me to be possibly conceding (p. 145), between “at least one sort of general proposition” and “a report of [why this?] a particular spatio-temporal situation” (my italics), since this situation also has generality.note

Since I take A, E, I and O propositions to be equally situational and do not admit Ryle's “particulars”, I do not admit that there is any ground in my “Problem of Causality” (A.J.P.P., August, 1938)note for the suggestion that I might have abandoned the identification of propositions and situations, nor do I admit that there is any “affinity” (p. 142) between my “spatio-temporal situations” and “the atomic facts once patronised by Russell and Wittgenstein”. That their “completely elementary propositions assert that named particulars have specified qualities or else stand in specified relation to other named particulars” is nothing to me, since, as I have indicated, I reject “pure particulars” and take any term to be general or have instances; otherwise, it could not be named or have qualities or have relations — thus, there is for me no question of anything atomic or elementary. The representationism bound up with this doctrine of particularity is brought out when Ryle says (pp. 142,3) that “Russell and Wittgenstein did not suppose that all or most or perhaps any of the assertions that we actually make are atomic statements. They are, rather, cheques drawn against these solid coins.” What appears here is that any supposed “solid coin” is merely a postulated “that whereby” cheques are validated, but there is nothing to show that anything validates them or why one rather than another could be said to be validated. Ryle does not commit himself to the Russell-Wittgenstein view he presents, but a similar representationist weakness is seen in his reference (p. 145) to a “report” of a spatio-temporal situation, where the question would be what was the difference between the report and what was reported, just as, when he asks (p. 144) whether “John is not at home” describes the same situation as “John is at the theatre”, we are left with the question of the difference between what is said to describe and what is said to be described. The terminology of “reports” is also used as if it


  ― 174 ―
supported the mere dogmatism of the following (p. 146): “Certainly universal statements are not reports of other-worldly states of affairs; but nor are they reports of this-worldly states of affairs. For we report states of affairs in the idiom of — reports of states of affairs. And this is not the idiom of universal statements.” In each case we say that something is so, and that can be alternatively expressed by saying that something is a state of affairs. I have argued that unless what is asserted in a “statement” of a “particular” spatio-temporal situation is not logically different from, is on precisely the same footing as, a universal “statement”, the two must be taken to be completely disconnected (even though this could not be consistently maintained); and Ryle's reply in terms of “idioms” amounts to no more cogent refutation of my view than that some people do not believe it.

What I have denied, then, and continue to deny, is that universal propositions raise any special problem for our knowledge; they are matters of direct experience, objects of observation, situations, just as particular propositions in the strict sense (the contradictories of universal propositions) are — and the attempt to get round this by any doctrine of “the pure particular”, “the uniquely individual fact”, is a clear failure. The supposed “harshness” of the observational view cannot be escaped from by the substitution of “the verification of hypotheses”, since, apart from direct experience of universal truths, we should not know what the universal form signified (what was being asserted, recognised as actual, when we said that XaY or XeY); moreover, since any proposition can be verified or found to have true consequences (since some propositions recognised as true can always be found to follow from the combination of the hypothesis with other propositions recognised as true, i.e., since there are always propositions which “would be true if the hypothesis were true” and which are true), verification, applying equally to any proposition and its contradictory, can never settle any question (or be proof — cannot, that is to say, bring us any nearer to belief in the truth of a universal (or of any other) proposition.

Again, the setting up of a universal proposition by way of “convention” would (apart from other objections) give us no assurance of anything that was not “by convention”, and any proposition called “axiomatic” would either be a matter of fact or have no bearing on matters of fact. Ryle gives no indication of how my view of mathematics constitutes an “impossible story” (and gives no answer to the arguments in the second part of “Empiricism”), but a passing reference to “analytic propositions” and the use of “modal” expressions make me think that his objections would be couched in terms of the analytic and the necessary. I repeat from “Empiricism” that there are no analytic propositions, that any proposition is an issue or has a contradictory and that it is only by experience that we find how the two terms “combine”, that to say something “could not conceivably be otherwise” is to say we could never find it out — or, more exactly, that no proposition has been advanced at all, though we may be deceived into thinking that one has. With the rejection of “analytic


  ― 175 ―
propositions” would go rejection of the notion of “analysis”, as in “What is the analysis of X?” — either in the meaning of what are X's “ultimate constituents” (which never could be discovered, and any suggested list of which could never add up to a situation but would remain so many “natures”) or in the meaning of what compartment or division of reality does X belong to, with which other things would it fit and from which would it be quite disconnected (with the same problem as before concerning relations or distinctions among such divisions).

The doctrine of “necessary truths”, or, in general, of “modal” distinctions among propositions, is subject to the objections I have advanced in “Hypotheticals” to any attempt to go beyond the four forms. The main relevant consideration is that of contradiction (Is it so or not?), though there are important connected questions of implication. And the position is that any doctrine of kinds of truth (of ways of being so, and not just of being so) is opposed to the recognition of a “straight issue”, of what precisely is or is not so. The attempt is to introduce some qualification into the copula instead of restricting it to the terms, to carry over something of the material of the proposition into the form. But any such “qualification of the copula” is exposed by reference to contradiction; on Berkeley's doctrine of independent or mental reality, as contrasted with the dependent reality of ideas, where we should have to say that some mental process is independently A and some idea is dependently B, the only way in which we could see the difference between the “two copulae” would be by seeing it as a material difference, a difference of concrete character, and we could deny the first assertion only by saying that the process is not independently A, not by saying that it is independently not A. It would then appear that we had been asserting a conjunction of characters, asserting that the process is independent and is A, and, assuming that we could state positively what the character of “independence” is, we should be back at a straight issue with the unambiguous copula of occurrence or being the case.

In the case of “modals” the similar question whether X is or is not necessarily Y will dispose of the special copula “is necessarily” (or “must be”); at the same time, the precise issue will certainly not be a conjunctive one — whether X is Y and is a necessary being or has the characteristic of being necessary — and the same will be the case with possibility. Being necessarily Y is taken to be a special way of being Y, the necessity is taken to “contaminate” the Y-ness and not just to accompany it (though whatever was thought on this point would not prevent there being the straight issue whether X is Y or not) because the question is in fact not one of material but one of form. Before taking up that point in detail, however, we may note the curious position of actuality as a form of modality, viz., that there is no difference between X's being actually Y and X's being Y; that the use of “actually” (like that of “as a matter of fact”) merely “gives emphasis” to the statement that X is Y, by which is meant that it conveys or suggests some additional assertion to the effect “I want you to take particular note of this” — which, however, does not


  ― 176 ―
affect the question of X's being Y. If, then, modal questions were material ones, we could give the emphasis in any case, and say that X's being Y was actually necessary or actually possible, and overthrow “modality” as a special sort of logical question. But if modal questions are formal ones, then the exclusion of any effect of modality on the copula leaves open only the treatment of it as a question of quantity, and this, while it does not affect the “categorical” character of the assertions in question, is in fact the line which is taken by non-qualifiers of the copula.

On this line, the strict form of a statement of “necessity” is that of a universal proposition and that of a statement of “possibility” is a particular proposition, so that the distinction between “X is necessarily Y” and “X is possibly Y” is that between XaY and XiY. So, to contradict “X is necessarily Y” by “X is possibly not Y”, to identify the latter with “X is not necessarily Y”, is simply to contradict XaY by XoY. Accepting this “equivalence”, we often find it convenient in ordinary speech to use “modal” expressions — to ask “Must A be B or need not A be B?”, “Can A be B or cannot A be B?” — without thinking that we have raised anything other than questions of fact, which would be more strictly presented in the four forms. If, on the other hand, the question were of distinct fields of necessity and of possibility (contrasted with the field of the actual), the contrast would be material, and logical relations among the types of assertions would be inexplicable. If we said that the contradictory of “X is necessarily Y” is “X is not necessarily Y”, we could not call this an assertion in the realm of necessity, for what is asserted has not the character of necessity. And if we tried to set up special rules (cutting across the boundaries we had erected) such as that the contradictory of a statement of necessity is a statement of possibility (that “X is necessarily Y” and “X is possibly not Y” are contradictories, and so are “X is possibly Y” and “X is necessarily not Y”), we could know the right way to do it only by getting away from “fields” (or similarly from different “manners” or “ways” of being true) and getting back to a “categorical” issue, an issue with the unambiguous copula which merely indicates that there is an issue, a question whether something is so or not, with no indication or suggestion of what that something is.

But, over and above this general use of “modals” to convey the distinction of quantity (cannot could be said to convey a distinction of both quality and quantity, but that is just can not), there are special uses connected with arguments and with knowledge, though having some affinities with the general use. Sometimes, when we say something is necessarily so or must be so, we mean that we know it is so — which is simply a question of fact. (There may be a suggestion that, as the phrase goes, “there is no possible doubt of it”, but that is never strictly correct, and, suggesting something of a rationalist kind, cannot be said to raise a positive issue.) We also have factual statements of the form that something is possibly so or may be so, meaning that we do not know that it is not so (e.g., the assertion “Plato may have written the Menexenus”, meaning “I do not know that Plato did not write the Menexenus”), and similarly


  ― 177 ―
we have statements to the effect that something is possibly not so. These renderings of “must be”, “may be” statements differ from the previously considered universal-particular renderings in the fact that they introduce terms not in the given formula, but they do present “categorically” types of issue that are raised in apparently modal form.

Again, the question of necessity and possibility may be that of proof or the absence of proof. When we say, All men are mortal and Socrates is a man, hence Socrates must be mortal, our “must be” means that we are saying not merely that Socrates is mortal but that the proposition, Socrates is mortal, is proved (is implied by propositions we believe to be true) — two distinct assertions but both asserting that something is a fact or just is so; though, taking the “hence” (or “therefore”) as indicative of following or being proved, we should have to call the replacement of “is” by “must be” superfluous. Still, the point of this repetition, the thing that is being emphasised, is that proof is relative to premises, that it is by the given data that X is Y is established. Correspondingly, we have the use of “may be” as meaning is not proved not to be, the equation of possible with not disproved, just as of necessary with proved — always with references to premises taken to be true. If we said, Socrates is a man and Socrates is mortal but all men may not be mortal (or it is possible that not all men are mortal), our meaning would be that the proposition, Some men are not mortal, is not disproved by the given information — or, on these premises, All men are mortal is not proved. This is what we mean by saying that All men are mortal is possibly not so or its contradictory is possibly so. But in such cases of not therefore not (possibly), just as in the case of therefore (necessarily), we are concerned with fact or “actuality” and there is no question of special kinds of copula; even though the statements have not always been put in precise logical form, we are clearly not going beyond the ordinary predicative logic, the logic of unqualified fact.

“Possibly not” may also be used in a less definite way than as meaning the non-implication of a proposition (say, All men are mortal) by a given set of propositions we accept; we might mean its non-implication by any information we possess, i.e., merely our inability to prove it. Thus we say “Possibly some men are not mortal” to convey that we are unable to prove that all men are mortal. This comes closer to the first of the special meanings mentioned; where “p is possibly true” (or “p is possible”, though this is not specifying any character of the proposition p itself) means “I do not know that p is not true”. But neither of these meanings (I do not know or I cannot show) is of any importance to inquiry as compared with “I cannot prove from propositions specified” — the fact of non-implication of some propositions by others. The general point remains that none of the uses I have distinguished raise anything but positive questions, questions of fact, though all are distinct from what I called “the general use” — the A or O issue and the E or I issue with the original terms.

Another case in which the question of “necessity” arises is that of imperatives. It is sometimes maintained that an imperative is not an


  ― 178 ―
“indicative” statement in that it cannot be contradicted, but obviously it can. When A says to B “Go away”, B quite often replies “I will not”, where he is clearly rejecting what A has said. Now B replies by making a statement about himself, so that he is taking A's “command” as a statement about himself (B) — thus the conflict of assertions can be presented fairly accurately as “You (B) are departing (or about to depart)”, “I (B) am not departing (or about to depart)”. But what complicates the case, what enables it to be passed off as not a conflict of assertions, is the suggestion of “necessity”; of course, your going is necessary (or is required) can be replied to by the statement that your going is not necessary (or is not required), but this raises the point that “X is necessary” has no meaning (that nothing has “necessity” as one of its own characters or belongs to a class of “necessary things”), that indeed any such formula has to be filled out if there is to be a real issue, that until the question has been answered “necessary for what?” (or necessitated by what?), nothing on which there can be opposing views has been presented. When, however, the statement is completed, when it is said that your going is required for X (that unless you go, there will not be X — or will not be X here), this, firstly, is a straight question of fact and, secondly, does not convey that you must go (the “absolute necessity” of your going), which the imperative form is intended to impress on your mind. Thus we can say that, while the positive content of the imperative form is no more than an asserted fact, it also embodies the logical confusion of maintaining that the assertion cannot be denied, that it is not really an issue but is above question. And the same can be said of the “must” form (you must go, etc.); this is just another way of making out that there is no real issue, that the matter is settled beyond the possibility of question (Cf. “Socrates must be mortal.”)

Of course, the matter of your going may be settled (in the sense that you are going) but this does not settle its “necessity”; and again its “necessity” can be settled only as its necessitation by X, which is not “absolute” (or inherent) necessity. But, however we might expand the statement (your going is necessitated by my desire, or by the promotion of certain things we are agreed to promote — any such expansion giving us something we could argue about), the absence of the expansion that would make it a positive issue conveys an “absolute necessity”, the impossibility of contesting (denying the existence or imminence of) what is “commanded”. There is the suggestion of some reason for the doing of “what is to be done”; you are going (because I tell you), you are going (because it is the right — or required — thing to do), etc.; but, in general, the suggestion is quite vague (you are going — just because!), in order that the pretence may be kept up that there is something (some ground or reason) that lifts “what is to be done” above discussion, something that establishes its necessity or “requisiteness”, when nothing more of a reason has been given than “You are going because you are going”. (It is pretty clear that the same would apply to much that is put forward as “modal” necessity.) Thus, while the real content is only you are going (a


  ― 179 ―
disputable matter), the special “imperative” form introduces logical confusion — inserts a “necessity” that is not in, but is above, the facts and that has to be excised if a definite issue is to emerge. (The issue, “Go”, “No, I won't”, might have been expressed more accurately as “You are going voluntarily”, “I am not going voluntarily “, but this would not affect the above discussion. The important point is that “Your going is imperative” looks like an issue, like something that someone might positively assert, but is not — though if anything could be meant by something's being “absolutely imperative”, this would be a positive issue and would not be a special kind of issue.)

The same question has been very prominent in what has passed as ethical theory, particularly in the conception of the “categorically” imperative, what is “required in itself”, not by or for something else; and here we have a further logical quirk in the notion of “obligation” or of what “ought to be” — of something which is necessary, or must be, and yet need not be. While, then, we have still the objections to dualism, the impossibility of finding any relation or distinction between ordinary states of affairs and what is supposed to “hold” in a different way, we can see what ambiguities, confusions, and, indeed, contradictions arise when such “ethical theorists” try still to maintain a footing in ordinary (historical) reality, to make their “ethics”, their absolute imperatives or obligations, still have some bearing on actual living — be imperative or obligatory on something. It is still true that they cannot do so, cannot cross the barrier between things that “are real in different ways”, cannot find any connection, e.g., between “judgments of value” and “judgments of fact” unless the former are just judgments of fact, the finding of whatever is meant by “value” in certain actual things (itself being an actual thing), and the upshot of their theoretical entanglements is just to encourage the view that there can be no science of ethics, that there is no subject-matter which such a science could consider.

The position is illustrated in what is said by Ryle on p. 141: “Like, for example, some members and followers of the Vienna Circle, Anderson begins by denying that ethical predicates stand for transcendent properties. But while they drew the consequence that ethical predicates do not stand for qualities or relations at all, but are merely emotive or hortatory expressions, Anderson draws the consequence that they stand for empirically ascertainable qualities or relations. Where they said that ethical pronouncements cannot express propositions, since they cannot, in principle, be verified or falsified by observation or experiment, Anderson says that they do express propositions, or describe spatio-temporal situations, and therefore that they are empirically verifiable or falsifiable.” This presentation of the case is vitiated by the use of a representationist terminology to which I at least have never subscribed; what I have said is that what are “pronounced” in ethical matters are propositions, are spatio-temporal situations (granted that those who amalgamate relation and quality or confuse form and matter, as in the notion of “necessity”, have to have these disentangled before the positive


  ― 180 ―
content of their “pronouncements” can appear); and the whole vocabulary of “expression” is simply a means of passing back and forward between what pronounces and what is pronounced. Even what are being called “emotive and hortatory expressions”, even interjections or exclamations, are means of communication, of presenting issues — though certainly in the incomplete form which is a regular time-saving device in social life, and which often occasions misunderstanding; at any rate, where people live in close contact, such “expressions” can convey a great deal not only about a person's emotions (say, of gratification or displeasure) but about the things he is gratified or displeased with. If this were not the case, what are called interjections would be mere noises and not, as they regularly are, part of a body of human communications. But what the “exclamationists” ignore are the facts (a) that, while something of a person's emotions may often be observed in his making a statement, they need not be part of what he is stating at all — he may be, while communicating it in an abbreviated manner, pointing to something that is not a state of himself, (b) that if the only reason for taking it that an objective ethical issue has not been raised is that the speaker's emotions are exhibited, or perhaps just concerned, in the making of his pronouncement, then it would be a reason for rejecting the possibility of raising issues altogether, because a person's emotions are concerned, and are quite often exhibited, in the making of pronouncements in any field at all; e.g., on mathematical issues, on which many people feel very strongly.

But no other reason has been brought out by Ryle in the passage quoted above. It is certainly not apparent how anyone could draw either of the “consequences” that Ryle mentions from the premise that ethical predicates do not “stand for” transcendent properties, and I in particular have not made the inference I am said to have made. I would strenuously deny that we ever start with predicates and then look for subjects for them, whether or not this can be called looking for “what they stand for” (that we ever have, so to speak, a bunch of predicates in search of a subject); what we start with is propositions, which already have subjects as well as predicates. And, that being so, we simply find something to be true and are not in the predicament of having to look for “verifications” of it — for something that would in turn have to be verified (ad infinitum) if propositions needed verification. In fact, what can verify and what can be verified are propositions of the same order (apart from which they could not be connected by the relation “verifying” or any other); and any proposition can have either function, can be checked against other propositions or be directly found, so that there is no question of passing over to “verifiers” from a separate sphere of findings or issues, no question of what in a particular case is a testing proposition being any more empirical than the one that is being tested.

We are not, then, looking for a subject “ethics”; we have ethical beliefs, which can test as much as be tested, which in any case are part of our direct experience. Of course, we can see how sophistical or “propaganda” devices (e.g., the confusions I initially adverted to) have flourished


  ― 181 ―
in the ethical field; but for anyone to say that ethics is “positively all deception” is simply to empty out his past experience. But this is something that no one can thoroughly do; and that is why there is also deception in “anti-ethical” theories, why their proponents are always smuggling in what they profess to have thrown out. Part of the deception, of course, part of the “analytic” myth, is of an egoistic character (what do I find when I set my analysis-machine rumbling?), and this is opposed to the fact that language and inquiry are inherent in social life, are at all stages part of communication — which is the main reason why any doctrine of “the fresh start” falsifies the facts, including procedures of the primitivist himself. It is not in the least the case that pronouncements on “the good life” (or on good activities generally) have as their empirical content emotions of the speaker or that their content need have as part of it anything at all that is going on in him.

We get to know good activities, then, learn their locations and descriptions, by coming in contact with them, by entering into activities which embrace them as well as other people's activities — by there being continuous processes of which all these are parts. And this is the case with our knowledge in general — which makes impossible a recognition of different levels or “types” of reality. Yet Ryle actually presents a theory of types as a way of getting out of logical difficulties; he introduces the fresh obstacle to knowledge of postulating types of questions that can have no relation to one another (new dualisms, new divisions in reality). “When Russell came across the contradiction of the class of all classes that are not members of themselves, and then found the parallel contradictions of ‘the Liar’ etc., he was forced to look for a solution in some sort of theory of Logical Types” (p. 150). And the solution according to Ryle lies in distinguishing “sentences” which do not “express” propositions at all from those which do, the “nonsensical” from the “significant”; i.e., those whose constituents belong to different compartments of reality from those which fit wholly into one box. It is, of course, not difficult to see that words can be strung together without conveying anything (e.g., “I am telling a lie”), but there is in this no suggestion of divisions (or higher or lower regions) in reality. And “the contradiction” depends simply on confused views of the notion of “class”; the class X is just the various things which are X (distributively), what Russell calls “the class as many”, and the question whether this is or is not a member of the class X is the “question” whether or not the X'es are X, i.e., no question at all. To suggest that there is a problem Russell has to introduce the conception of “the class as one” and ask whether it is a member of “the class as many” — but this “class as one” is just the various things which are X taken collectively, or the collection of X'es, and this is not the same thing (it is not a “class” at all), which means that there is no such thing as “the class as one” and the question of what is or is not “a member of itself” simply does not arise.

The details, of course, do not seriously matter; a “real contradiction” is just a confusion; but to say that a contradiction can have a “solution”


  ― 182 ―
is, in any case, to say that there is not a contradiction at all, and no special “logical” conceptions have to be invoked to deal with it. There is nothing here to suggest a distinction between science as concerned with what makes [!] (significant) statements true or else false and philosophy as concerned with what makes [!] statements significant or nonsensical — or to support the further representationism of taking science as talking about “the world”, while philosophy talks about talk about “the world”, when all that the introduction of “the world” means is that science asserts propositions, says that something is the case, uses the copula are — and there is nothing else that talk about talk (about no matter what) can do; it can only talk about states of affairs. There is no basis except different “orders of being” on which we could suppose that it could do anything else. The question of meaninglessness is not a question of uniting different categories, since the whole point about categories, as contrasted with qualities, is that they apply to all material; it is solely in terms of concrete reality that we can speak of things which do not go together — the common case being that of the use of the term XY when XeY (e.g., in XYaZ); and even though XYaZ has no subject or point of reference and so cannot be raising an issue, the consideration of whether it does or not can be an important feature of inquiry, whereas any suggestion of types of inquiry that do not mix (or fields of reality that cannot interpenetrate) can give no stimulus to inquiry. The belief in such divorced realities is only a device for protecting particular views, as, for example, the belief in immortality is protected by the contention that it is “meaningless” to say that a mental process is material (has weight or other such features of material processes) — such a division among regions of reality standing in the way of any coherent theory of mind and body. The division is supported by illogicality (particularly, the relativism of “that whose nature it is to know”) and thus the question of meaninglessness is shifted to that of “logical exclusion” from that of material exclusion, in considering which, in terms of our experience, we can easily see that it does not apply in this case but the mental is bodily. It is only, in short, when we distinguish realms of reality (qualify the copula) that we get involved in logical impasses — not when we deny such distinctions.

Professor Ryle, then, cannot draw a distinction between science and philosophy in the way he proposes; philosophy, when it “talks”, must talk of states of affairs, because there is nothing else to talk of, and if “significance” were not a state of affairs, it would just be — insignificant. I have argued above that recognition of meaninglessness is recognition of concrete fact (“findings of science”), and I would still maintain (from the remarks Ryle quotes, on p. 139, from my review of C. A. Campbell's “Scepticism and Construction”: A.J.P.P., June, 1935) that the philosopher “has true statements to make about the very things any special scientist is examining”, that “what is called ‘method’ is not something different from ‘findings’” (that what we call method, or, more exactly, knowing method, is knowing that any subject we examine has causes, has genera, etc., as a matter of actual fact), and that it is not true that philosophy


  ― 183 ―
has its own province, the province of the ultimate — on which last special point, of course, Ryle is not disagreeing with me. For, if Campbell's view were true, philosophy from its peculiar realm could not impinge in any way on the other realm, and there would be no such thing as the philosophical criticism of science, no appeal that could be made from findings here and now to findings “in the ultimate”. But, in combining the first of the above remarks of mine with the statement that philosophy is science, I spoke in a very loose and misleading way, for there is nothing in that statement to show that science cannot just as readily criticise philosophy as philosophy can criticise science, nothing to show, indeed, what criticism could be other than one scientist's finding out that another was mistaken. I do not think, in spite of my carelessness in expression, that that was ever my real position, or that I had any doubt at the time that it was philosophy that could criticise science and that such criticism was what was properly called “method” — though holding then as now that for the criticism to have any impact, the two sorts of inquiry would have to be directed to the same field.

What would have influenced my apparent assimilation of philosophy to science (not so much besides as incidentally to my repudiation of “the ultimate”) is that each of them is concerned with situational reality, with the spatio-temporal field, with things as they in a single sense are. What distinguishes them is that philosophy is concerned with the forms of situations or occurrences, science with their material; but it is only as forms of such material, as material with such forms, that they can be known. But this means that, if the work of inquiry is to be carried on, it must be at once scientific and philosophic, that if, in particular, the scientist is not philosophic, he will fall into confusions, he will rebuff philosophic criticism — he will lack a theory of categories, of sorts of problem, of “method” — especially he will be carried away by practical interests, by the interest in producing something or implementing a programme instead of in finding something out.

Burnet is dealing with this problem (not always quite clearly) when he presents the account given in the Republic of the work Dialectic does, in showing the defects of the special sciences, as “destroying the hypotheses”, e.g., the hypothesis of three distinct kinds of angles (cf. Greek Philosophy: Thales to Plato, p. 229 — I take this to mean that an angle is a turning, and that, considering a continuous turning round a centre, we shall not be disposed to break up the class of angles into the usual species distinguished in the theory of triangles); and the position is made even plainer (p. 323) when Burnet refers to Aristotle's objection to the assimilation of geometry to arithmetic, the proof of geometrical by arithmetical propositions, on the ground that “the proper hypotheses of each science must be left undisturbed. Aristotle's principle here could be described as that of saving hypotheses instead of destroying (or removing) them; it is one by which realms of reality are separated, and the question in each case is what is “true on these assumptions”, not just what is true — which makes it impossible to examine the assumptions


  ― 184 ―
by reference to findings in the fields they are supposed to mark off, as must be possible if they have consequences in their specific fields.note We can thus connect the opposition between (a) separate fields and restriction of inquiry and (b) unrestricted inquiry into a single field of reality (space and time, interrelated situations with no smallest and no largest) with an opposition between (a) saving hypotheses and removing appearances (e.g., ethical facts) — which is rationalism, and (b) saving appearances and removing hypotheses — which is empiricism. Hypotheses requiring removal are exemplified in axioms of science and elements of things, as well as the general mass of entities supposed to exist or be true in special ways, like normative truths, epiphenomena and sensa.

It is this rationalism, this separatism, this breaking up of reality into sundered sections, that is the mark of the scientist who is not a philosopher, who instead presumes to teach the philosophers (it being mostly scientists who have given the lead to contemporary “philosophic” schools) his own rationalism, his own devotion to ultimates and unquestionables, his own “analysis” (what are the elementary constituents of this?), on the basis of his own practicalism — of what he takes to be absolute ends. It is just because of its concrete concerns that the scientist takes his science to be more capable of criticising philosophy than philosophy is of criticising it, and thus we find a variety of “philosophies” (doctrines of the conditions of inquiry and even of the conditions of social life) which, in accordance with the concrete interests of their authors, lay down the law for technical philosophy, knowing next to nothing of philosophy's history, cling to subjectivism (or the personal outlook) in ignorance of the case either for or against realism or objectivism, and, finally, with their “practical” resources, intimidate the professional philosophers into showing less and less interest in the philosophical tradition and more and more subservience to “science”. Thus with what is esteemed to be the scientific progress of the twentieth century has gone philosophical retrogression. A certain practicalism or instrumentalism has, of course, been characteristic of modern philosophy in general; this is illustrated in the categories employed by Locke which are merely a set of devices by which men can arrange their material and are not exhibited as belonging to it or as having any common ground. This view was considerably corrected by Kant's criticism (followed up in some directions by Hegel's), and the correction


  ― 185 ―
was carried further by Alexander. There can be little doubt, however, that the failure of such criticisms to stick was due to the steadily larger and more influential groups who devoted themselves to inquiry but had little knowledge of the history of critical thought and treated “method” rather in a mechanical than in a logical way.

Professor Ryle speaks of what logic tells me, with the suggestion that I am laying claim to some private communication or even special revelation. What I maintain is that there can be no logic unless it is in the facts, unless their logical characters are found in any facts (or situations) of which we are aware. And what I take myself to be informed of by what might be called my “logical sense” is the continuity of things or their coherence, their “making sense” because they have a common ground, their negating all “breaks in reality”, all doctrines of units or realms — and there seems to me to be no doubt that there is a divergence of view here, that the issue arises anywhere and that to offer a settlement of it is to offer a view of the interlocking of questions or of categories. It would, of course, be maintained by the empiricist logician that no one can offer a consistent “separatist” answer to logical questions, but the primary point is that logical questions arise wherever any questions arise. It is obvious that the history of philosophy offers a very great stimulus to thinking on these questions (as I myself have been greatly stimulated on the general continuity-discontinuity question by the work of Burnet on the pre-Socratics and on Plato's later dialogues, as well as by Alexander's work on “Space-Time”), it is obvious, too, that great stimulus is afforded by the work of separatist thinkers like Descartes and Locke; but the stimulation is possible in either case just because all the time we are confronted by, or rather immersed in, what may roughly be called “the world”, i.e., things which extend around, within and through us. That is something we must always be considering, as part of our consideration of any special situations whatever, and that is how we come to have a “logic” — to see (as I should say), if we are consistent, the errors of dualism, the absence of any breaks in reality, along with special problems of how things intermingle. I have suggested that lack of consistency in these matters is contingent on pursuit of special aims and objects; but the main point is that we cannot have a piecemeal logic, that logic is concerned with the running together of questions of all sorts (“in all the categories”) and that to be confused on this matter is to be hazy in one's “logical (or philosophical) sense”. This I have also described as a sense of form; and it is because form is not additional matter, but is characteristic of any matter that may be in question, that one can speak of logic (or philosophy) as governing or directing science, and not the other way round — just as it is taking it the other way round, making matter do duty for form, making science do duty for philosophy, that has produced the intellectual chaos of the present day.

This sort of “materialism” is well illustrated in Marxism, a leading strain in which is the treatment of social revolution as the common measure of terrestrial events — a role that could be filled only by


  ― 186 ―
something formal. I have myself severely criticised this aspect of Marxian thought, but I was so incompletely emancipated by Marxism or proletarianism at the time of writing “Marxist Philosophy” (A.J.P.P., March, 1935) that I concluded the article by saying that the realist philosopher might find his activity to be part of a producers' movement, so that he would “see his intellectual levelling as an integral part of social levelling”.note The intellectual worker might quite well be said to be engaged in “levelling” in so far as he was denying “higher realities”, but to say that there are not different levels of existence has no connection whatever with being opposed to social privileges and distinctions — there is no reason why seeking to induce people to abolish such distinctions should form part of a common campaign with seeking to persuade them to reject “higher entities”; indeed the social egalitarian is for the most part highly superstitious, though preferring to call his presiding spirits by fashionable names like “progress”.note

It is apparent, of course, that there is in all this a confusion in the meanings of “supporting” and “opposing” as, first, trying to bring about and trying to prevent and, second, trying to prove and trying to disprove. And it seems clear also that such confusion between practical and theoretical questions (connected with which is the sheer superstition of the view that any practice follows from theory, that finding p to be true could in any way require us to do X) or between material and formal questions is one main reason why there could not be co-operation of intellectuals with proletarians, who, in spite of Sorel, do not exhibit the producers' mentality (the condition of co-operation) but are concerned with being better off, with reform, with making society go in the right way — a meliorism and a voluntarism that are opposed to the work of the student of society, to any intellectual grasp of his subject. I have remarked in “The Freudian Revolution” (A.J.P., August, 1953, p. 102)note that the only revolution properly so called is an intellectual revolution, “a revolution in ideas”, not any re-arrangement of externals. This is what the work of the intellectual producer (or the realist or empiricist philosopher, I would say, as in my article) resides in; not “social levelling” or any other “practical” undertaking but simply making discoveries — rejecting conventional or customary associations, like that of “benefiting” with all social activities, or, more broadly, that of theory and practice, and being concerned simply with following an intellectual tradition within an intellectual institution. It is the custody of such traditions and institutions that requires privileges or “charters”, and the work of egalitarianism in breaking these down is one of the major causes of injury to intellectual activity in our present


  ― 187 ―
period. It is true, of course, that social equality is merely a mirage, but devotion to it has still much to contribute to the destruction of culture. Culture, on the other hand, is concerned with the maintaining of boundaries, with opposition to levelling, to the treatment of everyone as of the same standing in any cultural field, and, in the intellectual field particularly, with the search for what is the case as opposed to what can be secured, with the discovery of the laws according to which society works as against ignoring such laws and proposing to make society go according to personal decisions.

It is interesting to note how badly “the ethics of the producers” (G. Sorel, Reflections on Violence, Ch. VIII, esp. § V)note fits the workers of to-day, though it possibly fitted better those of the early century — how opposed current attitudes are to what Sorel took to be the outstanding features of a producers' ethic, the chief “heroic values” of proletarian life; viz., initiative, emulation, care for exactitude and rejection of the notion of “reward”. These “values”, as I say in “Marxist Ethics”, are typical of disinterested activity and may still be found in intellectual institutions, which have not yet been completely eaten up by utilitarianism (interestedness) and egalitarianism, expressed in care for the careers of “individuals” as against the advancement of learning. We may recall here the remark of de Tocqueville that democracies are not interested in questions of form (it is only custodians of traditions that would be so) — their outlook, in other words, is materialistic. The same, in general, may be said of scientists, those, i.e., who are not also philosophers; they also are not concerned with form, or, in substituting for questions of form questions of technical procedures or devices, they are really giving to the material or subject-matter of their science functions it is unable to carry out, they are using it to set aside the exact studies that can be made of formal questions and that can bring out issues which would otherwise be missed.

It is thus that philosophers, as far as they still carry out their own studies and do not, as has happened so much of late, submit to the direction of scientists dialectically untrained, can themselves offer guidance which would remove “hypotheses” or blockages in the lines of scientific study; it is thus that logic, even while asserting the equal reality of all existing things, can claim that it “stands above” the sciences, that it “governs” the various concrete fields of investigation in a way in which science could not govern logic — since it is by common forms, and not by special materials, that investigation can be directed. It is to be understood, of course, that the scientist who sets himself up as a master of method though he has engaged in no systematic study of philosophy will always be found to be borrowing from philosophy, to be using scraps from Locke, Hume and so on, in fixing up his rickety apparatus. But the objections to this procedure, to the light-minded way in which scientists will tell us that we must deny the existence of causality or must take a phenomenalist view of it or must always proceed by “induction” because


  ― 188 ―
we can only start with “particulars”, the objections, more generally, to the substitution of human devices whereby material may be manipulated for forms that are embedded in it — the intellectual mess of all this — can be plainly seen. What can also be seen is the extent to which it has all been furthered by the fashionable philosophy of our day.

previous
next