I have read in philosophical literature that in the old days, many philosophers used to equate necessity with always existing. Nowadays, it seems that something eternal can be seen as contingent (I.e. could have not existed) and something that doesn’t always exist can be seen as necessary. For example, if one were to map out all possible worlds (I.e. states of affairs), an eternal being may exist in some of them, but not all. Which position is correct? The correctness of this position seems to bring out some important consequences that are hard to make sense of. Let me illustrate this with an example.

Suppose we take the example of fine tuning. Suppose, further, that this fine tuning occurs at some point in the universe that is not at the initial starting point of the universe. In other words, there is a time at which they do not exist, and then suddenly exist. If this doesn’t cohere with the actual evidence of the universe, replace fine tuning with some other potentially contingent event. The specifics of the example are not relevant.

Now, let us suppose that a theist says that these constants are improbable. Hence, He posits that a God tuned these constants. When pushed on asking to explain God’s existence Himself, the theist points to His eternity. If He exists at all times, it makes no sense to ask for an explanation.

Now, if a theist decides to just assert God as an explanation, without evidence, an atheist may choose the following route. He may simply assert that the constants being fine tuned were necessary. In other words, without explanation, the atheist just says the constants being fine tuned just had to happen.

Now, something seems off here. After all, how can you just assert that something is necessary without offering a how. But that is not the point. It is merely a strategy to point out that the theist does the same thing. After all, he is asserting, without explanation, a necessary Being that fine tunes these constants. Why not just remove one step of the explanatory chain and just assert that the constants were necessary? This seems to create two scenarios with the same explanatory power but gives the naturalist an advantage in that his version of events are simpler.

Now, this may seem fine if we go by some of the more modern versions of necessity. But this doesn’t seem to work if necessity can only be applied to eternal beings. Intuitively, constants being fine tuned by themselves seem to beg for more of an explanation than god himself. If god is eternal, what would a cause of Him look like? On the other hand, a cause of constants being fine tuned can atleast be imagined.

  • Nothing hangs on what people choose to call "metaphysically necessary", another thing will just go by another name. "Whatever is going to seem right is right. And that only means that here we can't talk about `right'", Wittgenstein.
    – Conifold
    Nov 13, 2023 at 13:57
  • @Conifold I suppose my question is whether a thing that begins to exist can be posited as necessary just as much as a thing that always exists. Judging from your quote, you are saying that there is no criterion of correctness here?
    – user62907
    Nov 13, 2023 at 14:08
  • There are no constraints on what "metaphysically necessary" must express in the post (according to you, at least), let alone criteria. There is only: if this then this, if that then that. And so what, where is the rub? When Kripke was formulating what became the modern notion, he was aiming to match colloquial and scientific uses of modal concepts. If you want limits you need constraints. Otherwise, it is up to linguistic preference and there we can't talk about `right'.
    – Conifold
    Nov 13, 2023 at 19:38

2 Answers 2


If necessity is existence in all possible worlds, then if existence claims can one and all be denied, or asserted, without a priori contradiction, then anything can be said to be necessary that can be asserted to exist in all possible worlds. (We must be careful with how we word this or else we will be saying that anything is possibly necessary, which by the usual lights of modal logic would mean a collapse of all possible things into actual necessity.)

If necessity is having counterparts in all possible worlds, then the limits of necessity will be the limits of the counterpart relation, such as they are (if there are any as such).

If modality can be nontrivially iterated, perhaps logical necessity is mere necessity, while metaphysical necessity is necessary necessity; or vice versa; or there is some other iteration of such concepts that might fit to all our intuitions about this topic. It does seem as though, if X is actually in some possible world, then it is necessary that X is possible, assuming unrestricted describability or definability conditions for possible worlds. But the equation "possibly possible = necessarily possible" might not be trivially given, either!


At one point in his first Critique, Kant writes:

Take, for example, the proposition, "The world exists either through blind chance, or through internal necessity, or through an external cause." Each of these propositions embraces a part of the sphere of our possible cognition as to the existence of a world; all of them taken together, the whole sphere. To take the cognition out of one of these spheres, is equivalent to placing it in one of the others; and, on the other hand, to place it in one sphere is equivalent to taking it out of the rest. There is, therefore, in a disjunctive judgement a certain community of cognitions, which consists in this, that they mutually exclude each other, yet thereby determine, as a whole, the true cognition, inasmuch as, taken together, they make up the complete content of a particular given cognition.

More or less immediately thereafter, he discusses necessity itself specifically:

The apodeictical proposition cogitates the assertorical as determined by these very laws of the understanding, consequently as affirming a priori, and in this manner it expresses logical necessity.

Later, when he "categorifies" necessity, he says:

That whose coherence with the real is determined according to universal conditions of experience is (exists) necessary.

But so Kant will say that we cannot use this more substantive conception of necessity outside of experience, or rather that if we try to use it so, we will end up tangled in dialectical illusion, a false sense of being able to interpret the meaning of reality as a determinate whole, whereas our closest empirical representation to such an infinite total is instead an indefinite one, such as does not license us to think through the modality of the entire world to any of the disjuncts he mentions regarding the "origin" of that entirety. In other words, we can entertain the disjunction {blind chance / internal necessity / external necessity} but can never decide which disjunct is true at all save it be for the sake of moral representation (in which case we may hope for the possibility that the world has an external, and morally good, cause, but we do not have to believe, much less know, that there actually is a necessary first cause as such).

  • I think we agree here that if existence claims can be asserted without apriori contradiction, then anything can be. Do you, however, think there is any sort of theoretical advantage for a theorist who asserts the necessary existence of a thing that always exists vs. someone who does this for a thing that begins to exist? As an example, suppose a theist says that today’s lottery winner was divinely fated by necessity. In other words, god is necessary and His will for the lottery win is necessary. If an atheist now says the lottery win was necessary by itself, does he bear more of a cost?
    – user62907
    Nov 13, 2023 at 15:44
  • @thinkingman I would think that if someone says that a thing-that-began-to-exist is necessary, they are talking about relative, not absolute, necessity, i.e. it is necessary if something else is given beforehand, like the necessity of the past say. As for theoretical advantages, I don't tend to play those kinds of language games so I don't have much to say on that score. Nov 13, 2023 at 15:51
  • A thing that begins to exist could be proposed to be necessary even if the things in the past were not, no? For example, God could necessarily fate something to happen next Tuesday, regardless of what happens before next Tuesday. If so, the question is can a naturalist simply assert the same kind of fatalism for that particular event as well, perhaps through some hidden law or just as a brute thing. Of course, this may seem like an assertion without evidence, but maybe one can say that the postulation of a god fating this particular thing is also an assertion without evidence?
    – user62907
    Nov 13, 2023 at 16:00
  • @thinkingman I have added some Kantian considerations to my answer, as a rejoinder to your question in the preceding comment. Nov 13, 2023 at 16:53
  • 1
    Very interesting edit, thanks a lot. @Kristian Berry. I for one think that apriori, any causal reality that proposes less contingencies is better. Aposteriori though, our world seems to exhibit pure indeterminism, which would equate to pure chance in Kantian terms I believe. But even if this weren’t true, and there was some reason to think that everything is necessary, there seems to be reason to prefer our current world as necessary or even some eternal naturalistic cause as necessary over god being necessary given the latter’s additional complexity.
    – user62907
    Nov 13, 2023 at 17:03

Well, in the spirit of Dummett, metaphysical necessity (see the SEP for more info) is simply a logical necessity applied to first principles. If one sees philosophical discourse through a Tarskian meta/object distinction, then one is constrained by one's current metalinguistic description of the world.

Let us follow Quine that we possess a web of belief that always presupposes our current philosophical discourse which like Neurath's boat prefigures and describes our current philosophical discourse. In such a case, one's metaphysical constraints are the sumtotal of one's various metaphysical theories. For instance, Quine himself was constrained by his naturalized epistemology so that his idiolect of philosophy would not accommodate pseudoscientific thinking. Such a constraint wouldn't impact an astrologer, however, because of an astrologer's less empirical approach to knowledge. Quine was also stingy with his ontological commitments, and so that process of stinginess would be presumed as a constraint imposed by his first principles.

The obvious truth to come from this is that Quine wasn't foundational in his approach, and therefore we can see that the constraints of foundationalism are inapplicable to him. Every philosopher, and indeed thinker, therefore has a unique set of constraints imposed by their metaphysical idiolect.

You must log in to answer this question.