Cause, causer, and the Principle of Causality
Now then:
The cause of a given effect is the true explanation: the one that is the fact. That's what it is, of course; and just defining it this way doesn't get you any closer to finding it.
But it is what science is looking for, obviously; and so Aristotle was right: science is the knowledge of things through causes(1) .
But I am now going to define two terms parallel to our "effect" and "what is affected" which will lead allow us to develop a method that is actually more secure than scientific method (because it is more trivial; but it will do nicely for our very general type of effect):
The cause of a given effect is all and only what is necessary to explain the effect. That is, the cause is another abstraction; it is everything without which the effect would still be a contradiction; but it contains nothing but this. So without a given property or characteristic of the cause, it wouldn't be able to do the job of explaining the effect in question, because in the "cause" as defined abstractly here, all the properties are necessary. And if there is something about the cause that isn't necessary for it to do its job, then this doesn't belong to this abstract sense of "cause" as we are defining it here.
If you consider the effect of the missing keys, then, depending on how you define the effect, the cause will be different--and you won't be able to say in the concrete what it is, though you might be able to know something about it.
If you define the effect as a non-living body's moving from one place to another when it can't move itself, then the cause would be just exactly those properties of anything that would be necessary to account for a non-living body's moving; but it would be just those properties, and anything else about what we would normally call the "cause" are left out (just as the concrete properties of what we normally call the "effect" belong to what is affected, not the effect). That is, suppose a thief stole the keys; the thief as cause of the keys as moving though not living is whatever about him can impart motion to something (the momentum in his hand) and the fact that this got into contact with the keys; everything else about him is not part of the cause of this particular effect, because nothing else is necessary to account for this particular problem.
Obviously, the thief has a lot more characteristics that just these two; but not one of them is relevant to the effect in question. But since the thief would ordinarily be called the "cause" of the motion in question, I want to give the "cause" in this sense a special name to distinguish it from the very precise, technical sense in which I am talking about the cause.
The causer is the concrete object that is "doing the causing"; it is the concrete thing that contains the cause as an abstract aspect of itself.
In other words, I am defining the "causer" as what most people would call the "cause." I am doing so because what we are after is the true explanation of things that don't make sense by themselves and irrelevant facts about the "cause" in the ordinary sense just get in the way: whether the thief had a moustache, whether he had malice aforethought or was brought up badly, whether he had indigestion when he did the thieving, and so on and so on. If you say that the thief is the "cause" of the removal of the keys, then you have all of this excess baggage that has nothing to do with the effect in question, and you cloud up your investigation with a whole lot of irrelevant details that if you aren't careful are going to mislead you.
Furthermore, notice that some aspects of the cause might not be in the causer. The keys couldn't have been moved by the momentum of the thief's hand if they had been made of Jell-O, and just deformed as he grabbed them; so one of the facts necessary for the keys to be moved is whatever it is about them that makes them respond by movement and not being deformed or blowing up when acted on by the momentum. But this particular aspect of the cause is in what is affected, not the causer.
In what follows, it is going to be absolutely vital to understand this abstract sense of "effect" and "cause," since we will be using it in this sense all the time, and not in the ordinary sense; and there are very important differences between the two.
One thing to note here is the relation between this abstract way of considering "cause" and "cause" in the sense of "the true explanation." Very often, in explaining an effect, we are really interested in more than just the abstract cause; we want to know the causer, perhaps, and perhaps circumstances surrounding what is affected and its causer. For instance, John would want to know that he dropped the keys and where, rather than simply that something about the relaxation of his hands plus the gravitation of the earth explains how the keys got away. The fact that he dropped the keys five minutes ago and twenty feet behind where he now is isn't actually necessary to explain their being missing, but concretely it is where and when this "relaxation plus gravity" occurred; and it is of interest to John, because he wants to find the keys again and not simply solve the theoretical problem.
Why, then, would anyone ever bother with "cause" in this very abstract sense? Simply because in many cases, the causer is not observable, either in practice or even in principle. What is the "true explanation" of the interference patterns of light, when there are also photoelectric emissions of electrons? That is, why does light "behave" like a particle and yet a wave? The answer cannot be discovered by finding the causer (the light) and examining it, because "photons" (units of light) are too small to be even in principle observable. Hence, all we can know about light in its relation to this effect is light as cause (in our sense) of the effect: that whatever it is, (a) it has all that is necessary to account for both "behaviors," and (b) that, since we cannot see it "as it is in itself," we can't say anything more about it in relation to this effect without going beyond the only evidence available.
The point is that light, in relation to this and similar effects, is not like the dropping of the keys, where you could in principle see the whole situation, and if you saw it there would be no effect. That is, if John saw the keys drop, he would not wonder what happened to then, but would simply pick them up; there would be no problem. The only way there would be a "problem" (an effect) would be for him to mentally ignore the relaxation of his hands and gravity and consider that without this, the keys would have stayed in his hands. But with the particulate-wavelike behavior of light, there is no possibility of getting into a situation analogous to John's; and so we have to content ourselves with what we know must be the case or the effect would be a contradiction; and we must be careful not to say any more than this, or we are merely saying something that might be true but could never be known to be true (because it couldn't be checked) and we are in the realm of pure speculation, and aren't talking about the cause even in the sense of the "true explanation."
The point, of course, is that this abstract sense of "cause" is not always the most useful in actual investigations, depending on whether or not we can get at the causer and examine it. But (a) it is always, of course, contained in the "true explanation," since the true explanation will obviously have all that is necessary to do the explaining (though it will have more than that), and (b) there are times when nothing but "cause" in this sense will allow us to say anything at all about the true explanation.
Well, but can we say anything at all about the cause in this abstract sense, if we can't actually observe it? Certainly. If it can be shown that without some property or some fact about whatever accounts for the effect, the effect remains a contradiction, then this fact must be true of any causer (or belongs to the cause in our sense). For instance, light cannot simply be a small particle, or it couldn't interfere with itself; so whatever it is, at least this is true of it. That's not much, but it's something; and it may be that more facts like this can be found out by thinking carefully about these "conditions sine qua non" (which is what we are talking about is usually called).
One of the interesting--and perhaps historically rather important--things this abstract way of defining the terms allows us to do is to formulate the Principle of Causality in such a way that it is as obviously true as the Principle of Contradiction:
Every effect has a cause.
This is, as I say, self-evident once the terms are defined as above. An effect without its cause is a contradiction, and there are no such things as contradictions; the effect + its cause is not a contradiction; and so for every effect there is a cause.
Note that this principle does not actually claim that there are such things as effects (that is, that anyone would ever actually get into a situation where the evidence available to him would form a contradiction); but if such a situation happens, then there is a fact which "resolves" the contradictoriness of the situation--either that, or the Principle of Contradiction is false.
So all this really is is an application of the Principle of Contradiction to the special case of finding what on the face of it looks like a contradiction in reality; and all it says is that it isn't really one, and the evidence that it is is misleading you.
This is something else I don't think I need to belabor.
Where the problem has come with the certainty of this Principle is in its various inexact formulations, where "effect" is taken in the kind of rough-and-ready sense that, "Well, everybody knows what an effect is," and various aspects of what is affected get mixed up with it; and where "cause" is understood as what I defined "causer" to be above, and people made great leaps of logic beyond their evidence in attributing to the "cause" all sorts of properties that had no necessary connection to the effect that "it" was accounting for.
For instance, if you formulate the Principle as Kant did, "Every event has a cause," you then have to show what it is about an event as an event that means that it can't stand on its own. Kant said, basically, that since an event begins at a certain time, there must be something "before" it which allows you to experience it as "beginning" when there wasn't any such thing before.
It sounds reasonable; except that he himself noticed that the heat of the fire's "beginning" is really noticed simultaneously with the fire; and so he says the cause has to be "logically" before. Logically before in time? But then what do you do with the "event" of the dawn, which is the light of the sun--which will only appear quite a few minutes later. On the grounds of this notion of "event," the "event" of the sunrise is caused by the dawn preceding it--which is obviously false. It is the (as yet unobserved) sun which is the causer of the light which we call "dawn."(2)
And surreptitiously, Kant was actually using the Principle as I formulated it, because in describing an event as "beginning" and therefore as "beginning-after" which necessitates a "before," he was actually trying to show just what the contradiction is in thinking of an event without a cause. Unfortunately, he hit upon the wrong thing about it.
Of course, this prevents Kant from using causality to explain things that are not temporal sequences, in spite of the fact that he is doing this all the time. What he called the "conditions for the possibility" of something is in fact what I defined as the "cause"; and he is constantly showing that without (for example) the a priori forms of space and time experience is impossible--which is another way of saying, isn't it, that experience is an effect of the structure of our receptivity?
Hume's "refutation" of causality that I mentioned a while back (that we don't observe the "dependence") was intended to explain, as I also mentioned, why we nevertheless tended to think in terms of cause and effect; and what this amounted to is that the apparent contradiction involved in not observing dependence and asserting a necessary connection nonetheless could be resolved by his notion of "customary sequence." It doesn't work, as I pointed out, any more than Kant's refinement of it does. But the point I am making here is that he was using the Principle of Causality as I have stated it to refute the Principle as he misunderstood it because the terms had not been accurately defined.
Every philosopher has actually used the Principle of Causality as stated above, because every philosopher has argued that if his theory is not the case, then something-or-other couldn't be the way it is; or in other words, the something-or-other is "impossible-unless."
The fact that those who try to refute the Principle use the Principle as I have stated it is an indication that the formulation I gave it is the one which is the absolutely certain one. That is, once again we have our criterion for absolute certainty: You can tell that something is absolutely certain, not on the grounds that you have no doubt about it, or even on the grounds that nobody around you has any doubt of it, but on the grounds that any denial of it surreptitiously affirms it.
Hence, without further meandering through the landscape of the historical versions of effect and cause, let this suffice for a reason for holding the formulation of the Principle that I gave above, together with the definitions that make it tautologically evident.
NextNotes
1. Actually, what he meant by this phrase isn't exactly that, because the Greek word translated "science" really just means "factual knowledge," and what he meant by "cause" is close to but not the same as what I have defined above. However, when all is said and done, I think Aristotle would agree with what I have said and not regard this quotation as a travesty.
2. Of course, Kant's formulation would also make "the first event" logically impossible, because by definition, as an event it has to have something before it. But why can't you say that something is an event because it has something after it? That is, supposing there to be a first moment in time (the "big bang"), then it has meaning, not because it follows something, but because events follow it.