1 64 CAUSALITY AND CONTINGENCY CH.
to cause, then clearly we restore entirely the meta physical character of the condition as not only necessary but sufficient.
There is another compromise between metaphysics and empiricism — one which has played an important c g The com_ part in the history of philosophy from the promise of beginning of the modern era, — the con- occasionahsm. cept o£ occasional causes, which we owe mainly to Geulincx (1627—69) and to Malebranche (1638—1715). These philosophers sought by it to cut the Gordian knot of psychophysical causality, in the Cartesian doctrine of the two substances, soul and body. Occasionalism denied that physical movement can be the efficient cause of ideas, or ideas the efficient cause of physical movement. The parallelism between them was explained as an agreement between soul and body depending on God. It was analogous to that which we may see between two clocks constructed by the same artificer, an agreement brought about, not through the action of one on another, but through their common dependence on the clockmaker's skill.1 1 " Imagine two clocks or watches which agree perfectly. Now, this may take place in three <ways. The first consists in a mutual influence; the second is to have a skilful workman attached to them who regulates them and keeps them always in accord; the third is to construct these two clocks with so much art and accuracy as to assure their future harmony. Put now the soul and body in place of these two clocks; their accordance may be brought about by one of these three ways. The way of influence is that of common philosophy, but as we can not conceive of material particles which may pass from one of these substances into the other this view must be abandoned. The way of the continual assistance of the creator is that of the system of occasional causes; but I hold that this is to make a deus ex machina intervene in a natural and ordinary matter, in which, according to reason, he ought not to co operate except in the way in which he does in all other natural things. Thus there remains only my hypothesis: that is, the way of harmony. From the beginning God had made each of these two substances of such a nature that merely by following its own peculiar laws, received with its being it, THE TWO CLOCKS 165 THE TWO CLOCKS 165 The occasional cause, when we reflect on it, is not a cause at all, except in so far as we transcend it and pass from it to God, in whom is the real principle of its causality, through the relation which it always implies between movement and sensation. And when nevertheless accords with the other, just as if there were a mutual influence or as if God always put his hand thereto in addition to his general co operation " (Philosophical Works of Leibniz, G. M. Duncan's transla- lation, chap. xv.). (Compare also the Troisieme Eclaircissement and the Systeme nouveau, Erdmann, p. 127.) We may remark that the comparison of the two clocks is not Leibniz's own invention, for we find it being commonly cited by the Cartesians as a scholastic illustration (cu. Descartes, Passions de I'dme, i, 5, 6, and L. Stein in Archiv fur Geschichte der Philosophie, i. 59). We may remark, too, that Leibniz's distinction between occasionalism and his system of the pre-established harmony has no great speculative importance 5 for it is easy to see that to dispense with the work of God from the different moments of the process of reality after it has been set going does not eliminate the speculative difficulty of the miraculous character of God's extrinsic inter vention. Without this intervention causality remains just as unintelligible as the harmony, which is already affirmed in occasionalism, and which Leibniz cannot help extending to his pluralism.
Geulincx, also (Ethica, i. sect. ii. § 2), explains the agreement of the two substances, soul and body, as that of two clocks: " Idque absque ulla causalitate qua alterum hoc in altero causat, sed propter meram dependentiam, qua utrumque ab eadem arte et simili industria constitutum est." The body therefore does not think nor make think (" haec nostra corpora non cogitant, licet nobis occasionem praebeant cogitandi ''). But bodies not only do not think, they do not act, they do not move of themselves, for the only mover is God. This most important doctrine was taught by Geulincx in his Metaphysica (published in 1691): " Sunt quidam modi cogitandi in me, qui a me non dependent, quos ego ipse in me non excito; excitantur igitur in me ab aliquo alio (impossibile enim est ut a nihilo mihi obveniant). At alius, quicumque sit, conscius esse debet hujus negotii; facit enim, et impossibile est, ut is faciat, qui nescit quomodo fiat. Est hoc principium evidentissimum per se, sed per accidens et propter praejudicia mea et ante coeptas opiniones redditum est nonnihil obscurius; jamdudum enim persuasum habeo, res aliquas, quas brutas esse et omni cogitatione destitutas agnoscebam, aliquid operari et agere. Existimavi v. gr. ignem, quod ad ejus praesentiam sensum in me caloris produceretur, calefacere; et hoc calefacere sic interpretabar, ac si esset calorem facere. Similiter solem illuminare, juxta similem inter- pretationem, lumen efficere, lapides cadere, ut interpretabar, se ipsos praeci- pites dare, et motum ilium efficere, quo deorsum ruant; ignem tamen, 1 66 CAUSALITY AND CONTINGENCY CH.
Leibniz extended occasionalism, giving it a profounder meaning and making its anthropomorphic bond between the physical and psychical substances the type of the universal relations of all substances or monads, it became in the system of the pre-established harmony the concept of the reciprocal unrelatedness of the monads in their common dependence on God. But through God the occasional cause necessarily conditions its correlative term, although, on the other hand, such condition may be non-sufficient, and for this reason occasionalism can retain a certain metaphysical value, and need not end once and for all in empiricism. That is, the occasion and the occasionate in themselves, the one in opposition to the other, are a mere contingent concomitance, actually like the succession to which solem, lapidesque brutos esse, sine sensu, sine cognitione haec omnia operari existimabam. Sed cum intellectual intendo in evidentiam hujus principii: Quod nescis quomodo fiat, id non fads, non possum non videre, me falsum fuisse, et mirari mihi subit, cum satis clare agnoscam, me id non facere, quod nescio quomodo fiat, cur de aliis aliquibus rebus aliam persuasionem habeam. Et qui mihi dico, me calorem non facere, me lumen et motum in praeceps non efficere, quia nescio quomodo fiant, cur non similiter, igni, soli, lapidi idem illud improperem, cum persuasum habeam ea nescire quo modo effectus fiant, et omni cognitione destitui? " (Opera philosophica, edition Land, ii. 150). It is remarkable that in this passage the negation of efficient causality (operari et agere] is connected with the empiricist opposition between subject and object affirmed in the principle indicated by Geulincx and so nearly resembling the principle of Vico: Verum etfactum con<vertuntur. This in its turn is closely connected with a sceptical theory of the knowledge of nature, analogous, as I have already pointed out, to that of Hume.
That occasionalism and the pre-established harmony both arise from the need of maintaining the unity of the manifold is evident in the proposition which is one of the earliest accounts Leibniz has given of the doctrine (in 1677, in a note to a letter of Eckhard): " Harmonia est unitas in multitudine ut si vibrationes duorum pendulorum inter se ad quintum quemlibet ictum consentiant " (Philosophische Schriften, edition Gerhardt, i. 232). For the genesis and the ancient and medieval precursors of occasionalism consult Zeller, Kleine Schriften, i. p. 316 n., and two writings of Stein, loc. cit. i. 53 OCCASION AND OCCASIONATE 167 the empiricist reduces causality. But the occasion is an occasion in so far as we do not think of it only in itself and in respect of the occasionate, but in relation to God who makes one term the occasion, the other the occasionate. And when this system of " occasion = God = occasionate " is constituted, then the reciprocal relation of the two extreme terms participates in the necessity of the relation between God and the occasion, no less than between God and the occasionate. The relation is that of a necessary but non-sufficient condition. Whence the occasion becomes a necessary and non-sufficient condition of the occasionate, and inversely this of that, even obliging us to think of each of the two terms, the one either as the occasion of the other or as occasioned by the other.
So that the characteristic of occasionalism is to unfold the relation of necessary and non-sufficient § 9. Either condition by duplicating it, in so far as metaphysics or between the occasion or the occasionate empiricism. ^e conditioning is reciprocal. Through this duplication the relative contingency of the effect in regard to the cause, in its character of necessary and non-sufficient condition, can be turned into the relative contingency of the cause in regard to the effect. And therefore the empiricism of the occasionalists is more accentuated than that to which we have called atten tion in the system of the simply necessary and non- sufficient condition, since with the double contingency the multiplicity appears actually loosened from every chain of metaphysical unity.
I say appears, because the so-called duplication, if on the one hand it duplicates and confirms the con tingency, on the other it duplicates and strengthens 1 68 CAUSALITY AND CONTINGENCY CH.
the necessity of the cause in regard to the effect. The body necessarily supposes God who creates the soul, and the soul necessarily supposes God who creates the body. And in the system of the monad- ology every monad supposes God, the creator of all the monads, and therefore supposes all the other monads. In such reciprocity of conditioning between occasion and occasionate, the relative necessity of the cause in regard to the effect becomes reciprocal relative necessity, or rather it becomes necessity which excludes all contingence and therefore all empirical multiplicity.
Between the unity, then, of metaphysics and the > multiplicity of empiricism all attempts to fix a relation of condition and conditionate, as a relation which mediates between unity and multiplicity, are destined to fail.
Setting aside the possibility of stopping at an inter mediate point between the metaphysics of efficient if causa^ty and the empiricism of causality contradiction intended as simple contingent concomi- of metaphysical tance, is it perhaps possible to stop at the concept of metaphysical causality or at its extreme opposite, that of empirical causality?
It is obvious that the concept of metaphysical causality as necessary and sufficient condition is absurd. The concept of condition implies the duality of condition and conditionate, it implies therefore the possibility of conceiving each of the two terms without the other, yet this possibility is negated by the concept of metaphysical causality, which is an a -priori relation, • and implies the unity and identity of the two terms. To use the word causality, therefore, in the meta physical meaning, if we would give an exact account of what we have in mind, is to mean what has no meaning.
xi ATOMISM AT THE CROSS ROADS 169 We are left with empirical causality. Let us not insist now that, whilst all causality implies a relation, empiricism excludes every relation by § ii. Atomism y.
as the basis of postulating a multiple reality or things empirical unrelated. Let us even admit the hypo thesis that there may be such a manifold and that causality may take place in it. Let us simply inquire whether on the basis of pure atomism it may nevertheless be possible to maintain the concept of causality as plain empirical causality. Atomism is always rinding itself at the cross roads. It has either to maintain rigidly the original and absolute multi plicity of the unrelated, and in that case it must give up the attempt to explain the phenomenon which it has resolved into the unrelated atoms; or it has to explain the phenomenon by making it fulfil effectively for the atoms the purpose for which they are destined, by the principle of the reality given in experience. Now to do this it must endow the atoms with a property which renders possible a change in their primitive state, that is, in their state of unrelatedness and absolute multiplicity, in order to bring about their meeting and clash. Movement (the effect of weight) as a property of the atoms is already a negation of their absolute unrelatedness, because we can only speak of movement in terms of the relation of one thing to another, and movement itself, as Epicurus remarked, must be different in the different atoms (through the differences which the new relations and correlation imply) if by means of movement the atoms are to be aggregated and so generate the phenomenal things. For if all the atoms move in the same manner, and in the same direction, and with the same velocity, it is clear that their meeting 170 CAUSALITY AND CONTINGENCY CH.
is for ever impossible, and the atomic hypothesis is useless.
Atomism, therefore, has always of necessity been mechanism, one of the most coherent logical forms § 12. Mechan- of the conception of reality, conceived as a ism. presupposition of mind. Mechanism in resolving all reality into matter and force (atoms and movement) starts with the postulate that nothing of this reality can be lost and nothing can be added to it. Qualitatively and quantitatively, therefore, being is immutable, and all change is no more than an alteration of the disposition in the distribution of the elements of the whole. The intelligibility of the new is a perfect mathematical equation of the new with its antecedents. The sum of matter and force at the moment n is equal to the sum of matter and force at the moment n - i, and also to that at the moment n + i.
Whether it resolves force into matter with the old materialism, which saw in movement the external g The manifestation of the intrinsic property of epistemology matter; or whether with the chemists and of mechanism. physidsts to-day, who think they have got away from materialism because they no longer speak about matter, it resolves matter into force or energy; mechanism, apart from any imaginative representation of the atom and of movement, consists in the conception of absolutely manifold being, the result of elementary units. These units can be variously added up, but always give the same result, so that the possibility of a novelty which is not merely apparent, and of a creation which is really new existence, is absolutely excluded. In its particular application it is clear that in mechanism a relation of condition to conditionate is only thinkable as empirical causality. If a ball struck by another ball THE MECHANISM OF SCIENCE 171 moves, mere empiricism must be limited to the dis covery that the ball after having been struck had moved, without supposing any other relation between the antecedent stroke and the movement which followed it. And this, indeed, is the assumption of empiricism when it insists on what it would have us understand is the true character of causality. But empiricism, when from the particular it passes to the universal and has to make its own metaphysics of reality, in order to enable it to explain the particular itself according to its own scheme of intelligibility, and thereby make credible the mechanism according to which there cannot be a movement which had not a previous movement to account for it, cannot observe the temporal contiguity of the movement of the ball with the stroke received by it without thinking that the movement of the struck ball is one and the same with the movement of the striking ball, the one communicat ing to the other just as much as it loses itself. And lo and behold, the duality of the facts of experience is resolved into a unique fact, whereby in a whole the new is equalized with the old. And when empirical causality wants to affirm concomitance between pheno mena, and, in general, multiplicity without unity, what it comes to is that by empiricizing the causality it attains to mechanism, or rather to the crudest form of metaphysical monism it is possible to conceive.
Against the mechanism necessarily prevalent in modern science since Descartes, Galileo and Bacon, there has arisen in the latter part of the philosophy of last century in France a philosophy, the contingency leading concept of which has been termed contingency^ ramous for its vigorous vindi cation of freedom. Modern science, following Bacon, J 172 CAUSALITY AND CONTINGENCY CH.
has pronounced itself empirical, and even if, following Galileo and Descartes, it has been mathematical, it has always been conceived with the logic of empiricism. Reality for it is a presupposition of thought, and self- identical in its already perfect realization. The new philosophy of contingency, conscious of the freedom of the mind in its various manifestations, has opposed to the concept of a reality always self-identical that of a reality always diverse from itself.1 Contingency is, in fact, an attempt to conceive freedom by denying the unity or identity in which mechanistic empiricism ends, without, on the other hand, abandoning the concept of conditioned reality, that is, of that multiple reality which is empirically given.
In order to understand the starting-point of the philosophy of contingency, let us begin by quoting the first page of Boutroux's thesis on The principle of the Contingency of the Laws of Nature, philosophy of " By what sign do we recognize that a tmgency. ^^g js necessary? What is the criterion of necessity? If we try to define the concept of an absolute necessity we are led to eliminate from it every relation which subordinates the existence of one thing to that of another as to its condition. Accordingly, absolute necessity excludes all synthetic multiplicity, all possibility of things or of laws. There is no place, then, in which we could look for it if it reigns in the given world, for that is essentially a multiplicity of things depending more or less one on another. The problem we have to deal with is really this: By what sign do we recognize relative necessity, 1 There are several indications of this doctrine in Lachelier, but it was first definitely formulated by Emile Boutroux in De La contingence des lots de la nature, published in 1874, republished in 1895, and in many subsequent editions.
that is to say, the existence of a necessary relation between two things? The most perfect type of necessary connexion is the syllogism, in which a particular proposition is proved as the consequence of a general proposition, because it is contained in it, and so was implicitly affirmed at the moment the general proposition itself was affirmed. The syllogism, in fact, is only the demonstration of an analytical relation existing between the genus and the species, the whole and the part. So that where there is an analytical relation, there is a necessary connexion. But this connexion, in itself, is purely formal. If the general proposition is contingent, the particular pro position which is deduced from it is, at least as such, equally and necessarily contingent. We cannot reach, by the syllogism, the demonstration of a real necessity unless all the conclusions are attached to a major premise necessary in itself. Is this operation compatible with the conditions of analysis? From the analytical standpoint the only proposition which is entirely necessary in itself is that which has for its formula A = A. Every pro position in which the attribute differs from the subject, and this is the case even when one of the terms results from the decomposition of the other, leaves a synthetic relation subsisting as the obverse of the analytic re lation. Can the syllogism reduce synthetically analytic propositions to purely analytic propositions? " Starting from this principle it is not difficult to argue that the necessity arising from absolute identity is not to be found in any proposition and tingencyor is not in the syllogism. So that if necessity. mechanics is mathematically conceivable, physics is no longer simple mechanics, and biology is 174 CAUSALITY AND CONTINGENCY CH.
not physics, and neither is biology psychology, nor psychology sociology; and in short, whenever science with its mechanical interpretations is forced to bring a new order of phenomena into line with another, it lets the difference between the one order and the other escape. Therefore, while remaining within the limits of simple experience, the world cannot but appear a hierarchy of different worlds each of which has something irreducible to what is found in the antecedent. The world, then, is not necessary if necessity mean necessary relation, and if necessary relation mean identity.
To begin with being. In its greatest universality and abstractness, can we say that it is necessary? Can we deduce the existence of being analytically from its possibility, just as from the premises of a syllogism we deduce the conclusions? " In one sense no doubt there is no more in ' being ' than in ' the possible,' since whatever is was possible before it existed. The possible is the matter whose being is fact. But being when thus reduced to the possible remains purely ideal, and to obtain real being we must admit a new element. In themselves, indeed, all the possibles make an equal claim to being, and in this meaning there is no reason why one possible should be realized rather than another. No fact is possible without its contrary being equally possible. If then the possible is given over to itself, everything will be eternally floating between being and non-being, nothing will pass from potentiality to actuality. So far, then, from the possible containing being, it is being which contains the possible and something besides: the realization of one contrary in preference to the other, actuality properly so called. Being is the synthesis of these 9 xi FREEDOM WITHIN NATURE 175 two terms, and the synthesis is irreducible." 1 And this is the contingency of being. If being is contin gent, everything is radically contingent, inasmuch as it is being. And if from the abstractness of being we rise gradually to the greater concreteness of the reality presented in experience, we see the range of necessity becoming ever more restricted, that of contingency growing ever larger, and thereby mak ing ever wider way for that freedom, which in the mechanical and mathematical conception of the world is absurd.
It is evident then that the philosophy of contingency is an empiricism incomparably more empirical than the naturalistic and positive mechanism empiricism and of the ordinary empiricism. Should it mechanism of succeed in making freedom, or a possi- mgent. f^-j^y of freedorrl) spring up within nature, which for empiricism alone is real, we should be able to say that it had conquered empiricism with its own weapons. Contingency, in fact, does no more than affirm the reality of the differences or rather of the multiplicity of the real. In it #, b, <:, d^ do indeed constitute a system, but a is not $, nor £, nor d. To make each term the conditionate of the preceding term and the condition of the following term is not to make the preceding term originate the following term because between the one and the other there is no equivalence such as there is between a and a.
Suppose there were such equivalence and that b = a^ as mechanism requires, the relation between b and a would then be necessary, as necessary as the relation of a with <2, and representable in a purely analytical judgment. But if we suppose only a given, 176 CAUSALITY AND CONTINGENCY CH.
by making abstraction from the multiplicity and from every external relation, then it is only a in itself which is absolutely necessary. So if b is not identical with a, it is different; and in so far as it is different and irreducible the juxtaposition of a and b can never give rise to an analytical relation. (Strictly, it could not give rise to any relation whatever, because, as we know, relation is already identity.) It cannot give place to it because a = a and b = b. No term is contingent in regard to another (that is, re latively not necessarily) except on condition that it is absolutely necessary in regard to itself. And all the terms one by one are only contingent relatively to points of view external to their definite and particular essence, whereas, considered absolutely, they are com pletely necessary. But the necessity which clothes them is only that which mechanism affirms, except that, instead of being monochrome, it is many-coloured like a harlequin's dress.
Being" is not deducible from "possible." The proposition is self-evident, — but why? Because being is being, and the possible is possible, antithesis6 shorn of the realization of itself which between con- is purely the exclusion of its contrary, tingencyand gut jf behind the realized being which is not the possible, we know not how to think any other than this possible, toto caelo different, it is clear that being is thinkable only in so far as it is thought as immediate; not as realizing itself but as a reality already established. As such it is self-identical, immutable, in such wise that self-identical must not even mean identical with itself, since even identity is a relation of self with self. And with one term only which can have no other confronting it, even though IS FREEDOM MADE POSSIBLE? 177