A Case for Necessitarianism

Let's workshop this short defense of necessitarianism, the view that everything that happens is necessary (necessary to such an extent as to exclude all unrealized possibilities)

A Case for Necessitarianism

1. Introductory Remarks

Necessitarianism, according to which nothing happens unless it necessarily happens, is one of the most radical metaphysical views. It entails that ultimately nothing is up to any dependent creature like us, which seems to rule out our having moral freedom (the sort of freedom required for moral responsibility). It entails that whatever takes place (down to the finest detail: how I tie my shoe this morning or how I plan my day while in bed) is the only thing that could take place—leaving no room, in effect, for unrealized possibilities.

Challenging as it is to our commonsense understanding of agency, responsibility, and possibility, here I offer a defense of Necessitarianism. My proofs, which largely reflect the spirit of Spinoza (a 17^th century philosopher on whom I wrote my PhD dissertation), stem from those developed in further detail in my paper “A Rationalist Defense of Determinism,” which was published in Theoria a few years back.

2. What is Necessitarianism?

Necessitarianism is the most extreme form of Determinism. Determinism is the view that there are conditions jointly sufficient for anything that is the case in a specified domain (such that when those conditions obtain so do as well that for which they are jointly sufficient). Different forms of Determinism differ regarding what these conditions are and what the domain is. The most common form of Determinism discussed in contemporary metaphysics is what we might call “Chronological Determinism”: the view that the past provides the sufficient condition for whatever is the case in the domain of the future—or, more simply put, the view that the past guarantees the future.

Chronological Determinism, although entailing that nothing is ultimately up to us, still leaves room for things to have gone otherwise. How so? Well, look at it this way. Either the past goes back infinitely or not. If not, then (for all that Chronological Determinism says) the first state of reality from which all the other states flow could have been otherwise. And if the past does go back infinitely, then (for all that Chronological Determinism says) the whole infinite series of states, the entire beginningless flow of reality, could have been otherwise.

Necessitarianism, on the contrary, cuts deeper. It is the view that—no matter the domain—there are conditions jointly sufficient for anything that is the case, such that nothing is the case unless it had to be the case. Necessitarianism says, in other words, that there is a sufficient condition for whatever obtains (including that sufficient condition itself), in which case—since what suffices for x guarantees x (see the fourth point in the next section)—whatever obtains is guaranteed to obtain: every feeling, thought, action, process anywhere could in no sense have been otherwise.

Since either there is or there is not an ultimate sufficient condition for whatever is the case, there are two forms of Necessitarianism. According to Conditional Necessitarianism, which is rarely endorsed because of the infinite regress worries it invites, there is no ultimate sufficient condition even though there is a sufficient condition for whatever obtains (ad infinitum). According to Absolute Necessitarianism, which is Necessitarianism in its strongest sense, there is an ultimate sufficient condition for whatever obtains.

I want to give some proofs for Absolute Necessitarianism, a view most famously endorsed by Spinoza. Before I get into the arguments, it is important to introduce some terms and make some clarifications first..

3. Background Definitions and Assumptions

First, there are only three options when it comes to something, A, obtaining. (1) A is other-caused, that is, A is caused by something nonidentical to A, that is, A is excreted (whether in time or not) out of some other being (and so exists ab alio, from an other). (2) A is self-caused, that is, A is caused by A itself, that is, A is excreted (whether in time or not) out of its very own being (and so exists a se, from itself). (3) A is uncaused, that is, A is caused neither by its own being nor by some other being but rather by nothing, that is, A is excreted (whether in time or not) out of nonbeing (and so exists ex nihilo, from nothing).

Second, the uncaused category is necessarily empty, in which case all things are either (1) explained by themselves (self-caused, causa sui) or (2) explained by an other (other-caused, causa aliena). Why is it impossible, in effect, for anything to be uncaused? I defend this controversial view in detail in section 5.1 of my paper “A Rationalist Defense of Determinism” (onlinelibrary.wiley.com/doi/abs/10.1111/theo.12280), but the short answer is: you cannot get something from nothing. To deny that the uncaused category is empty is to say, in effect, the absurd: that something, x, has reality even though reality—reality all-things-considered (and so including x itself)—does not ultimately suffice for x to have reality in the first place (whether the phrase “in the first place” be understood in a temporal sense or not). Being, however, cannot be engendered by nonbeing. Something cannot be generated from power-bereft metaphysical nothing—from metaphysical nothing, remember, in a way that does not mean that it is generated from itself. Ex nihilo, that is to say, nihil fit: from nothing—which does not have any powers or states or dispositions or laws or potencies or fields or information or implications or energy fluctuations or even lacks (if lacks have being)—nothing comes.

Third, when it comes to that which is self-caused there are three main options. According to the first option, that which is self-caused is that which exists before it exists in order to cause itself to exist. This option is a nonstarter. It is absurd, after all, for something to exist before it exists. According to the second option, that which is self-caused is that which causes itself to exist at the very same moment it comes to exist in the first place. This option too is a nonstarter. Nothing, after all, can create itself at some moment despite not existing before it is created. There is only one remaining option, which unlike the other two does not involve any absurdity: that which is self-caused is that which is necessary in virtue of existing by its very definition (that is, in virtue of having an essence that—rather than just happening to be instantiated—all by itself guarantees that it is instantiated). For something to be self-caused in the third sense, in other words, is for it to have an essence that cannot truly be conceived except as existing (since that essence inherently involves existence in the way that the concept “bachelor” inherently involves the predicate “male”). If something is self caused, that something must be self-caused, therefore, in this third sense, in which case it always already exists by the necessity of its own nature.

Fourth, what is sufficient for o is what guarantees o, such that o obtains whenever what is sufficient for o obtains. After all, what is sufficient for o is what fully explains why o does rather than does not obtain and what fully explains why o does rather than does not obtain is what guarantees o. To avoid confusion about the claim that what is sufficient for o guarantees o, consider the following points. (1) The sufficient condition for what occurs is not to be confused with the difference-maker for what occurs. Despite how people sometimes speak, the sufficient condition for the match lighting is not the match being struck. Besides the match being struck there are other conditions that need to obtain: match being dry, strike-surface being gritty, fan not blowing, oxygen being in the room, or whatever. (2) Some people may think that even though there is a sufficient condition for me pressing the shift key on the keyboard, it was not guaranteed that I would press the specific one I did: the one on the right. Such a scenario is no counterexample to the claim that what is sufficient for o guarantees o. To be sure, there was a sufficient condition for pressing the shift key in general, that is, any shift key. In this case, it was guaranteed that some shift key would be pressed. But there was also more specifically a sufficient condition for clicking the shift key on the right instead of the one on the left. The sufficient condition guaranteeing that specific action presumably involves the factors for pressing the shift key in general and also the factors that made the difference for my pressing the one that I did: it is my habit to hit the one on the right; the one on the right was closer to the other key I wanted to press while holding it down; and so on. (3) If q does not guarantee z, then q might do some work to explaining why z obtains. However, if q does not guarantee z, then q does not sufficiently explain why z obtains and so is merely a partial cause, an insufficient condition.

4. Proof A for Absolute Necessitarianism

Consider Proof A for Absolute Necessitarianism—in particular, for the view that whatever is nonultimate, that is, whatever fails to exist by the necessity of its own nature, is sufficiently caused by that which is ultimate. Consider, in other words, a proof for the view that there is an ultimate reality, a self-caused and so absolutely necessary reality, that provides the determining condition for itself and whatever else there may be.

1. The sufficient explanation as to why the totality of absolutely everything is the way that it is (as opposed to some other way) is provided either (1) by nothing, (2) by something beyond the totality, or (3) by something not beyond the totality. In other words, the totality of things is either uncaused, caused by something outside of it, or caused by something not outside of it.

2. The totality is caused by something not outside of it, which means either that the totality as a whole sufficiently explains the totality or that some part of the totality sufficiently explains the totality.

Rationale.—Since nothing is uncaused and since there is nothing outside of the totality of absolutely everything, the first two options listed in premise 1 are ruled out.

3. Some part of the totality sufficiently explains the totality.

Rationale.—The totality as a whole does not sufficiently explain itself—sufficiently explain itself, that is, in a way incompatible with merely some part of the totality sufficiently explaining the totality. For the totality as a whole sufficiently explaining itself would mean that each being—including cups and cars—is the sufficient cause of itself, which is false. (And even if the totality as a whole explains itself, absolute necessitarianism still follows. For whatever has reality would be self-caused in this case.)

4. Nothing beyond the totality sufficiently explains the part that sufficiently explains the totality.

Rationale.—Nothing is beyond the totality since the totality is the totality of absolutely everything.

5. No other part within the totality explains that part, that part that sufficiently explains the totality, in which case that part is self-caused: it sufficiently explains itself.

Rationale.—To say that the mere part that sufficiently explains the totality is itself sufficiently explained by some other mere part is to say the absurd: that the nonidentical effect of the sufficient cause is itself the sufficient cause of that cause. (Even if by some magic it is not absurd to say that the mere part that explains the totality is sufficiently caused by some other part to which it is not identical, then by the transitivity of sufficient causation the mere part in question would be self-caused anyway, which is precisely the point here.)

Therefore, the part that sufficiently causes the totality is self-caused, that is, the part that provides the determining condition for the totality provides the determining condition for itself.

Proof A is straightforward. Since the totality is sufficiently caused by something not outside of it, either the totality as a whole is self-caused or some part of the totality that sufficiently explains the rest of the totality is self-caused. The totality cannot be self-caused as a whole because that would mean that each thing is self-caused, which is false. Hence some part of the totality must exist by the necessity of its own nature. Since that part sufficiently explains the whole (in which case that part is the ultimate cause of everything else), the necessity of that part transfers to everything else.

5. Proof B for Absolute Necessitarianism

      Consider finally some versions of Proof B for Absolute Necessitarianism.

Version 1.—Assume that R is the collection of all things not self-caused. Now, there must be a sufficient reason for R—a reason for each part of R and for why R obtains as opposed to some other collection. Suppose that S provides the sufficient reason for R. Either S is self-caused or S is not self-caused. To suppose that S is not self-caused entails a contradiction. After all, since S would be self-caused were it beyond R, the only hope for S not being self-caused is that S is either R itself or some part of R. But since S provides the sufficient reason for R, S provides the sufficient reason for itself no matter whether S is R itself or some part of R. That which provides the sufficient reason for itself, however, is that which is self-caused. The absolute necessity of S, which it has since it exists by the necessity of its own nature, transfers over to that which it sufficiently explains: R and everything that makes up R.

Version 2.—Assume that R is the collection of all things not self-caused. To explain why the whole collection R obtains, it is not good enough even to explain why all the specific parts obtain if that explanation merely refers to other members of that collection. If the explanation for why each part obtains involves only other parts of R, further explanation is needed to explain why the whole of R obtains. Unless, say, the parts are each individually necessary in virtue of being self-caused or unless, say, just one part is self-caused and serves as the full explanation of the whole (neither of which is being granted here given the definition of R), we must appeal to something beyond those parts in order to explain the whole. Since we must appeal to something beyond the parts of R in order to explain R, S cannot be one of the parts of R. So that leaves only two options: S is outside of R, which means that S is self-caused (being that R is the collection of everything that is not self-caused), or S is R itself, which also means that S is self-caused (because S would be providing the full explanation for itself in providing the full explanation for R). As the full explanation for R, S’s necessity (which is a function of it being self-caused) transfers to R and each member of the collection.

Version 3.—Even if it is merely metaphysically possible that R has a sufficient explanation, which is a much more modest claim than that R definitely has a sufficient explanation, it follows from the reasoning in versions 1 and 2 of Proof B that it is at least metaphysically possible that there is a self-caused explanation, in which case there is a self-caused explanation for R at least in one “possible world.” But if there is a self-caused explanation for R at least in one “possible world,” then there is a self-caused explanation for R in every “possible world.” For what is self-caused is what exists by the necessity of its own nature and so is an absolutely necessary being.

This piece is unpublished

Photo: rediscovering-black-history.blogs.archives.gov/15-0728a/