That Causality is Never
a Logically Necessary Relation

When we speak of event P causing event Q, what we mean is that if it weren't for P, there would not have been Q. When I drop a pen, and it falls, I say that, had I not dropped it, it would not have fallen; this is what it would mean if dropping it caused it to fall. Likewise when we say rivers swell due to rain we mean that had it not rained more than usual, the rivers would not have swollen; this is what causality means. It requires us to think in counterfactuals that must always be possible to cogitate; we can mentally imagine it not having rained as much--which is perfectly imaginable--and conclude that were it not for the rain the rivers would not have swollen. This is what we mean when we speak of causes.

The very notion of a cause is that, were it not for the cause, there would not have been the effect. But when the cause is a rule of necessary logic, it is impossible to cogitate the counterfactual "were it not the case". Now, take the example of the rains causing the rivers to swell, where I say had it not rained, they would not have swollen. Would it make any sense to say, "Had it rained and not rained, they would not have swollen"? Does the proposition "rained and not rained" make any coherent sense at all? This is what we are doing when we try to cogitate the counterfactual of: "Were it not for the necessary logic that forced such-and-such a conclusion, such-and-such would not have followed." When we are speaking of some effect that could not have been otherwise due to logical necessity--that is, could not have been otherwise without there being a true contradiction--it is not causality we are speaking of, since it is impossible to imagine a world wherein there is a privation of the cause at all. When we try to cogitate the, "Were it not for P, there would not have been Q" in this sense--speaking of P as an aspect of necessary logic--we are merely uttering gibberish in our counterfactual "were it not for" and this is not a coherent proposition at all. We might as well be saying, "Were there true contradictions, there would not have been Q." The fact is that speaking of "true contradictions" at all is nonsensical gibberish, and not coherent speech.

This is the ontological sense in which necessity can never produce effects as their causes. Since the notion of P causing Q for us is so bound up in us being able to cogitate, "If not for P, there would not have been Q," when we speak of necessary logic "causing" anything, it is impossible to cogitate the absence of the cause. Both cause and effect are equally necessary (if we are to suppose that it would take a true contradiction to avoid both), and it is impossible for their relation to be considered what we mean by "cause" at all. Does it make any sense to say, "If the shortest distance between two points were not a straight line, Euclid would not have taken up geometry"? We may as well be saying, "If Euclid studied arithmetic at 12, and did not study arithmetic at 12, he would not have taken up geometry." The first condition in both these propositions is simple nonsense, impossible because it is contradictory, and therefore does not propose anything at all. Thus, any state of affairs that results from purely logically necessary laws, this is not what is meant by causality at all. In the same way, it is absurd to speak of anything at all "causing" two plus two to equal four.

So much for ontology. But there is also an epistemological sense in which logical necessity is not involved in causality as well. Epistemologically, causality is always an induction and not a strict, necessary consequent. When we speak of knowledge of what would have happened if not for a given P, we inevitably go some distance from the actual, into the hypothetical. I say of a man who dies after being shot in a robbery, "Were it not for the gunshot, he would not have died." But when I consider "the world as it would be without the gunshot" I am altering some actual fact, and going into the possible but not actual. Thus, I can conclude nothing with epistemic necessity about these contingents. Yes, what I mean by "The gunshot caused his death" is that "Were it not for the gunshot, he would not have died." But this must always be qualified; for it is always possible that he would have died anyway, were it not for the gunshot. Were he not robbed, something else may have caused his death. He may have--instead of being robbed--gotten to his car, got into it, and had an accident and died. Once we alter the fabric of the actual facts, this leaves us with a world that is altered from the actual one, and no deductive conclusions can be drawn about this other world. That "Were it not for the gunshot, he would not have died" is always a hypothetical speculation, and not a necessary conclusion. We are dealing with induction here: On no night before has he died, therefore were it not for this event he would not have died. This is never strictly certain; when we speak of causes our conclusions are always inductive. Like I said at the opening, the concept of causality is the sort of thinking where we say, "If P then Q" and reason about it thus: "Not P then not Q," an inference we can only ever know through induction and not necessity, as any beginning logic student will tell you.

But then we have the inverse, where we say, not that without the cause there would not have been the effect; but, given the cause, the effect could not have been otherwise. This is the only sense in which there is necessity in causality, but it is not a logical necessity. If it were a logical necessity, then, given the cause, only a true contradiction would avoid the effect. Give that man shot in the robbery a survival from the bullet; suppose now that he survived. Reason demands that there be some alteration in the physical circumstances of the wound, if we suppose he did survive. Intuitively--and in a strong rational sense--we cannot suppose he could have lived instead of died, with all the physical facts of nature exactly the same between both cases. But here is where we realize the genius of Kant--there is no contradiction in supposing so, any more than there is a contradiction in an uncaused event. Just why do we perceive things in chains, events leading to events leading to events? Not because of analytic logic. Try as I might to find a contradiction in the idea of a lion suddenly appearing on a city street, with absolutely nothing that led to it appearing or caused it to be there, I cannot see that in such an instance I have affirmed what I have denied (which is what a contradiction is). The only conclusion to be drawn is that in our rational apparatus, and probably in nature itself, there is causal thinking and causality, respectively; but we cannot conclude that it is based upon logical necessity. Physical necessity perhaps, or perhaps what Kant would term "rational necessity"--but in terms of analytic logic, causality seems to be superfluous to its fundamental principle that there are no true contradictions. After all, the difference between physical necessity and logical necessity may lie in the very notion of cause--and this is probably the reason laws of nature are given the title physical "necessity" at all. It is hard for us to conceive everything physical being the same in nature, and only one thing being different without any impact on anything else. It is hard for us to imagine the man shot in the robbery living instead of dying with every single atom of the situation the same in both instances, over and above the life or death. In terms of the physical world and our thinking about it, we cannot imagine this; but there is no contradiction in it being so. So causality, while being physically necessary, is not so logically.

In conclusion, we can never pronounce knowledge of causes and their effects in a necessary manner: epistemologically, causality cannot be necessary, according to the inferences we can make about it. But there is also an ontological sense in which causality is never necessary. This comes from the fact that when effects issue from necessary logic, necessary logic is not, properly speaking, standing in a causal relation to them at all. When something is caused, this presupposes that it could have been otherwise; we must be able to cogitate the idea that the cause was never present at all, in order to suppose that were it not for the cause there would not have been the effect. Since it is impossible to cogitate a true contradiction, we must hold that everything that issues from the fact that there are no true contradictions are not caused by that fact, as the term is properly used. The conclusion is that nothing can be caused by there being no true contradictions, any more than it is possible to suppose that, were there true contradictions, any given situation would not have arisen from the fact. It is impossible to cogitate a situation where there are true contradictions to begin with; hence, if P is a necessary truth, it is not merely untrue, but utter nonsense, to do thinking along the lines of: "If not for P, such and such would not have followed."

Hence, both from an epistemic and ontological perspective, causality is never a logically necessary relation.

Angelhaunt.net: Because earth's madness is heaven's sense.