92. There is a fact in the history of philosophy which proves with the greatest evidence the truth of what I have just said. This fact is the system of occasional causes maintained by eminent philosophers. If a body, they say, strike another body at rest, it will communicate to it its motion; but this communication does not imply a true causality, but that the motion of the impinging body is a mere occasion of the motion of the body impinged. Here then a thing is conceived as a necessary condition of the existence of another, and yet it is denied that there is between them the relation of causality. In thinking of the two phenomena we cannot invert the order, and conceive the motion of the body impinged as the condition of the motion of the impinging body, yet we can deny the relation of causality between the condition and the conditioned. Therefore the idea of causality represents something besides the necessary order of things among themselves.
93. This brings us to a new phasis of the question. Is the relation of causality faithfully represented in the conditional proposition: if A exists, B will exist? The connection expressed by this proposition is not the relation of causality. If the fruit-tree N flourishes in a certain country, M will flourish. A constant experience proves it. The conditional proposition in this case does not express the relation of causality of the flourishing of N with respect to the flourishing of M; yet the proposition is true. One phenomenon may be the sign of the immediate approach of another, without being its cause.
94. Conditional propositions, in which the existence of one object is affirmed as the condition of the existence of another, express a connection; but this may not be a connection of the objects with each other, but with a third. If a gentleman's servant goes to a place, and then another servant of the same gentleman goes to the same place, the cause of the going of the second may not be the going of the first, but simply that their master wished them to go one after the other. The crops in one field indicate the state of the crops of another field, and this indication may be expressed by a conditional proposition. Why so? Is it on account of the causality of the crops in one field in relation to those in another? Certainly not; but because the circumstances of the climate and the soil produce a sufficiently fixed order between them to verify the conditional proposition, without the intervention of the idea of the causality of one in relation to the other.
95. There are many cases in which the relation between the condition is necessary, and yet the condition neither is, nor can be, the cause of the conditioned. We are here treating of efficient cause, of that which gives being to the thing, and it would often be absurd to attribute this kind of causality to conditions which on the other side are necessarily connected with the conditioned. Take away the pillar on which a body rests, and the body will fall; the connection of the condition with the conditioned, or of the taking away the pillar with the fall of the body is necessary; the proposition in which this connection is expressed is true and necessary in the natural order; and still it cannot be said that the removal of the pillar is the efficient cause of the fall of the body.
96. Even a purely occasional connection is all that is necessary for the truth of the conditional proposition; and no one ever confounds the occasion with the cause. In the present example, the body cannot fall unless the pillar is removed; and it must necessarily fall if it is removed; but the cause of the fall is not in the removal of the pillar, but in the weight of the body, as is evident if we suppose the specific gravity of the body to be equal to that of the fluid in which it is submerged, since in that case, the removal of the pillar is not followed by the fall of the body.
97. Causality cannot express a necessary relation of the condition to the conditioned, unless we deny all free causes. Supposing the idea of causality to be correctly expressed in this proposition: if A exists, B will exist; by substituting God and the world for A and B, it will become: if God exists the world will exist; which would lead us into the error of the necessity of the creation. By substituting man and determinate actions for A and B, we shall have the proposition: if man exists, his determinate actions will exist, which implies necessity, and destroys free will.
98. Here arises the question: would the relation of causality be correctly expressed by a conditional proposition, taken in an inverse sense, or with the effect, as the condition and the cause as the conditioned, (not conditioned in the order of existence, but only as a thing necessarily supposed,) that is, if, instead of saying: if A exists, B will exist, we say: if B exists, A exists? In this case, the proposition may be applied even to the dependence of creatures on God, and in general of all free actions on their causes; for we can say with truth: if the world exists, God exists; if there is a free action, there is a free agent.
99. Although at first sight this seems to explain the relation of causality, this new formula cannot be regarded as correct. For, though it is true in general, that if there is an effect there is a cause, it is also certain that oftentimes one thing supposes another, not as its cause, but as a mere occasion, as a condition sine qua non; which is far from being true causality. Supposing the body supported by the pillar to be so placed that it cannot fall unless the pillar is removed, we might form the conditional proposition: if the body has fallen, the pillar has been taken away; the proposition is true, although the removal of the pillar is not the efficient cause of the fall of the body.
100. God could have so created the world that creatures would have no true action of causality upon one another, and yet have so arranged them that the phenomena would correspond with each other in the same manner as they now do. This is the opinion of defenders of the doctrine of occasional causes, and to this is reduced the pre-established harmony of Leibnitz, according to which all the monads constituting the universe are like so many clocks, which, though independent of one another, agree with admirable exactness. On this hypothesis we might form infinite conditional propositions expressing the correspondence of the phenomena without the idea of causality entering into any of them.