Posted on

Jonas Werner – “Extended Dispositionalism and Determinism”

Picture of a tall glass tumbler, meant to represent the glass disposition to break.
“Nr. 1598” (2001) Peter Dreher © Courtesy of the Peter Dreher Foundation

In this post, Jonas Werner discusses the article he recently published in Ergo. The full-length version of Jonas’ article can be found here.

You should handle glasses with care because they are fragile and could break. Dispositions, like fragility or inflammability, go hand in hand with possibilities, like the possibility that the glass breaks or the paper ignites.

Modal dispositionalists take this observation as the starting point for their theory of metaphysical modality. Every possibility, so they claim, is underwritten by a disposition. Some of these dispositions are possessed to a very small degree, some are iterated (dispositions to acquire certain dispositions), and some were lost and are only a thing of the past.

At the heart of the modal dispositionalist’s position  lies the following biconditional:

It is metaphysically possible that p just in case something has, had, or will have an (iterated or non-iterated) disposition to be such that p.

I call the proponent of this biconditional the “classic dispositionalist”.

I argue that, for some p, it is possible that p although nothing has, had, or will have a disposition to be such that p. Some possibilities are only indirectly underwritten by dispositions. 

Why should one be unhappy with classic dispositionalism? Because, when combined with certain plausible assumptions, it quickly leads to a disaster.

The first assumption is that dispositions are always future-directed: nothing can be disposed to change the past.  For the classic dispositionalist, this immediately leads to the result that the first moment in time (if there is one) could not have been different.

The second assumption is that there are immutable truths about the dispositional roles of fundamental physical objects. Plausibly, nothing has the power to change these dispositional roles. For example, nothing can stop electrons from repelling protons. As a result, truths about the fundamental dispositional roles of fundamental objects turn out to be necessary for the classic dispositionalist.

The problem with both assumptions is that they generate too many necessities. To see the force of this worry, assume (something close to) determinism:

A complete description of the state of the universe at the first moment in time, in conjunction with immutable truths about the dispositional roles of physical objects, entails a complete description of every later state of the universe. 

Now, the first assumption gave us that the state of the universe at the first moment in time is necessarily the way it is, while the second assumption gave us that fundamental dispositional roles are necessarily the way they are. Whatever is entailed by necessities is itself a necessity. Hence, we get the result that every state of the world obtains by necessity.

According to this result, for a match that never ignites – because I accidentally dropped in a pond, for example – it is impossible that it ignites. If we were to still subscribe to classic dispositionalism, we would even have to say that it was never flammable. This seems absurd!

Of course, there is some room for manoeuvre for the classic dispositionalist, which I discuss in some detail in the paper. For now, I just wish to mention that the case based on determinism is just an extreme version of the general worry that the classic dispositionalist might be forced to accept necessities that are incompatible with the manifestation of some dispositions.

In the second part of my paper, I propose a variant of dispositionalism that is immune to this problem, which I dub “extended dispositionalism”.

Clearly, the right-to-left part of the biconditional has to be saved. Something having a disposition to be such that p needs to be sufficient for the possibility that p, otherwise the central idea of modal dispositionalism is lost. But the dispositionalist need not say that something having a disposition to be such that p is necessary for it being possible that p.

Extended dispositionalism allows that possibilities are indirectly underwritten by dispositions; it allows that the left-to-right direction of our biconditional fails. This blocks the problem described above, because from the fact that nothing is disposed to be such that p we need not conclude that it is not possible that p.

How can possibilities be indirectly underwritten by dispositions?

In a nutshell, we can take a collection of true propositions to be a candidate for a collection of metaphysical necessities just in case every disposition is such that its manifestation is logically consistent with the conjunction of all propositions in this set.

There will be many such collections. However, some of them might be more plausible candidates for a basis of all metaphysical necessities than others. Maybe there is a maximal collection; maybe a collection is the largest one that avoids arguably objectionable cases of arbitrariness; or maybe it turns out that what’s necessary is indeterminate.

In any case, the method of looking for a collection of necessities that is compatible with the right-to-left direction of our biconditional has it that dispositions keep their role as the source of modality.

Still, we might have possibilities that are not the manifestation of any dispositions. We could, for example, hold that the first state of the universe is not necessary, although nothing ever has, had, or will have a disposition for it to be different.

Want more?

Read the full article at

About the author

Jonas Werner is a postdoctoral fellow at the Massachusetts Institute of Technology. Previously, he was a postdoctoral researcher at the University of Bern. He received his PhD from the University of Hamburg. His research focuses on metaphysics and the philosophy of language.