Causal Necessity
Causal Necessity
Are causal laws necessary? Are particular causal relations necessary? It has been supposed not: either they are thoroughly contingent or at most weakly necessary (less so than logical necessity). I will put the case for the necessity view. First, they are clearly not epistemically necessary: it could have turned out that causes have different effects from their actual effects (lightning might have turned out to cause shingles). If necessary, they are metaphysically necessary, not epistemically necessary. They are necessary a posteriori not a priori. So, the question is whether they are like water being H2O or like water being plentiful on Earth. I will dismiss the idea that causal relations are totally contingent with not even a hint of necessity in them; it isn’t just an accident that heating water makes it boil or hitting a nail makes it go in. The question is whether there are two types of metaphysical necessity, strong and weak—are there degrees of metaphysical necessity? This is commonly believed, but not usually defended. What would we think if someone maintained that the necessity of origin, say, is weaker than the necessity of identity? No one has ever claimed that to my knowledge, and with good reason: all the recognized examples of metaphysical necessity are equally necessary, and thought to be so. It never crossed Kripke’s mind that some of his examples of metaphysical necessity are stronger than others; they are all totally necessary. So, why suppose that causal necessity is not similarly total? Is it intuition? Let the present causal condition of the universe be repeated at a later time: isn’t it inevitable that the same effect will be produced? You won’t get some massively different effect, or even a slightly different effect. Of course, there might be an epistemic counterpart to the actual condition that gives rise to a different effect, but that is irrelevant, being a proof only of a lack of epistemic necessity. If the world is in the same state through and through, it will give rise to the same effect, as a matter of necessity.[1] And isn’t it very strange to suppose that metaphysical necessity might vary in degree—that some cases of it are only very weakly necessary? How can necessity be weak? It could hardly be that 2 is necessarily even in the strong sense while 3 is necessarily odd in the weak sense. What sense of “necessary” is that? It’s either necessary or it’s not. Are some propositions strongly contingent and some only weakly so? What does that even mean? Isn’t it simpler and more intuitive to suppose that all necessity is equally strong? How many degrees of necessity might there be—three, a thousand? We certainly don’t talk that way.
Suppose we adopt Shoemaker’s view that properties (kinds) are individuated by their causal powers; then it will be a necessary truth that a given property has the effects it has. For example, the shape square will have characteristic causal powers different from the color red—there is no possible world in which red has the causal powers of square. In general, the causal laws of a kind of thing follow from its nature; or rather, the laws constitute the nature.[2] Kinds and causal powers are necessarily joined. A natural kind has a real essence in its composition and structure, but it also has a real essence its causal powers, these being connected to composition and structure. Hume (and Locke) had essentially the same view: causal necessity is real necessity in the objects, whatever our ideas of it might be (nominal essence). It may be opaque to us, but that doesn’t compromise its robustness as necessity. The force of gravity, for example, necessarily obeys the law of gravity; that force would not exist without that law. It doesn’t just happen to obey the inverse square law. Similarly, the laws of psychology are not adventitiously linked to the natural kinds of psychology: it isn’t an accident that impressions of red produce beliefs about being red instead of beliefs about being blue or square. Just as everything is necessarily self-identical, so everything has the causal profile it does as a matter of necessity. The location of an object isn’t an essential property of it, but its causal profile is. So, we can add causal profile to the list of other metaphysical necessities.
In fact, I think that causal necessity is likely the most fundamental of our modal concepts: we grasp it before to get to logical necessity and Kripke-type necessities. It is tied to perception and our primitive grasp of how things work. Indeed, I suspect that our conception of logical necessity is an outgrowth of our concept of causal necessity (the premises make it the case that the conclusion is true).[3] People have only denied causal necessity because they confuse metaphysical necessity with epistemic necessity. There is just no good reason to deny that causal necessity is genuine honest-to-goodness necessity.
[1] I am putting aside objective randomness, but even here a certain probability of a particular effect will be necessary.
[2] See my Principia Metaphysica.
[3] See my “A New Metaphysics”, “Causal and Logical Relations”, and “A Causal World”.
Why do objects obey causal laws? This question leads to a dilemma. On the one hand, if the laws *cause* objects to obey them, then causal laws cause themselves. But this kind of self-causing seems absurd. On the other hand, if the causal laws play no causal role, then they can’t explain objects’ obedience; they seem instead only to describe, not explain, regularities.
I think that the only coherent response to this dilemma is, as you say, that objects have natures and these natures dictate their behavior and interactions with other objects. But this approach only relocates the explanatory problem: Why does Object O’s nature N in situation S lead to effect E rather than to E1, E2, E3, etc.? We can imagine N in S leading to a good number of different effects, so why does it always lead to E? There is something about N that always leads it in S to E, but we cannot imagine *its* – this something’s – nature. We cannot, for example, imagine what would lead particle A to attract particle B or what would lead particle A to repel particle C. The nature of particle A – what leads it to attract something and repel something else – is itself inconceivable.
Because conceivability is necessary for understanding, I conclude that we can’t understand causal necessity. We can’t understand whatever it is that explains causal relationships and therefore regularities. We are cognitively closed to the foundations of science.
A Humean conclusion: we don’t and can’t know the objective basis of causation, yet we know it is necessary.
Let’s see if I understand this correctly.
A wind turbine requires wind to turn it in order to generate electricity. That is causal necessity. But it’s not sufficient. The motor generator needs to be fully functioning and correctly connected in order for the power generation to happen. .
You do not understand correctly.