Abstract
Functionalism is retro, and —in the philosophy of science— retro is en vogue. In this talk, I argue that conceiving of Lewisian functionalism (Lewis, 1970, 1972, 2010) as a programme aiming to explain emergence is to misconstrue it, and misses the point of what makes the framework attractive. While I do this in dialogue with the literature on emergent spacetime in theories of quantum gravity, the argument is applicable to a broad range of subdisciplines in the philosophy of science.
The presentation starts by examining what kind of problems spacetime functionalists, such as Huggett and Wüthrich (2025), want functionalism to solve. In examining the interrelated notions of empirical incoherence and physical salience, I argue that Maudlin (2007) and Huggett and Wüthrich (2025), among others, are implying that physically salient derivations are coherent narratives, in the following sense: if a derivation of a higher-order phenomenon, P, from a more fundamental system, S, is to account for the existence of P—the derivation must be formally correct, and S must be interpretable as beables evolving in local spatiotemporal regions. This notion corresponds to Albert’s (2014) “narratable physics” and Abell’s (2004) view of explanations as narratives, hence the term “coherent narrative”.
I argue that Lewisian functionalism amounts to the denial that there is a story to tell. As is pointed out by Butterfield and Gomes (2020), Lewisian reduction is achieved through a deductive argument concluding in a mandatory identity claim—it does not make any sense, on this view, to ask how S can “give rise to” P, because P just is S, and we must identify them.
To argue, in the face of a derivation, that there are conceptual gaps or hard problems (e.g. Le Bihan, 2021) that must be bridged, amounts to a category error where some concept is distinguished from its physical realisation without proper justification.
In this, I side with Huggett and Wüthrich (2013, 2025) by defending them from critics, like Knox and Wallace (2023), that seemingly conflate Lewis’s views on theoretical definitions on the one hand, with Lewis’s position in the philosophy of mind on the other.
I argue that some criticisms aimed at Huggett and Wüthrich (2013) is grounded in a misunderstanding of why Lewisian functionalism is useful for their project—functionalism guides derivation. Once a derivation is achieved, there is nothing left to explain. Functionalism does not add anything to the derivation. This nullifies criticisms such as those advanced by Linneman (2021), which seem to presuppose that Lewisian functionalism should provide some further ontological explanation—when, really, Lewisian functionalism denies that any such task remains.
I conclude by discussing how Lewisian functionalism denies that there is anything that we must say about how spacetime came to be. Taking seriously Maudlin’s demand that physical derivations must be embeddable in local spacetime regions, on pain of incoherence, the functionalist remains unthreatened by pointing out that a physical derivation often consists of two narratives running in parallel: one (internal) narrates what some system is doing, and another (external) narrates how to derive some result from a theory. Many derivations do not provide internal narratives—state comparisons, gauge fixing, etc. As long as the external narrative is embeddable in spacetime (i.e. as long as we are able work with physical theories), then Lewisian functionalism denies that more is required. Functional coherence is coherence enough.
