Sumários

Panpsychism

18 Março 2026, 17:00 • David Yates

In this topic we will examine issues surrounding structuralist ontologies, starting with the case from quantum physics for a metaphysics of relations. We will then move on to consider Galen Strawson's panpsychist response to the puzzle of structuralism, which is to say that while it is true that physical science only tells us about structure, that does not mean that the world is pure structure, it means that physical science is incomplete. Finally, we will briefly consider the prospects for a Buddhist metaphysics of science in which there is fundamentally only the cosmic whole, and what seem to be distinct entities standing in certain relations to each other are merely abstractions from the one.

This class focuses on Galen Strawson's arguments for panpsychism, specifically: the intrinsic natures argument, stemming from Russell; and the anti-emergence argument. Panpsychism offers a unified account of the natures of fundamental physical particles and of the emergence of consciousness from physical processes. We will examine the most serious problem for panpsychism, the combination problem, and consider how it might be solved.

Reading week 5: Galen Strawson, "Realistic Monism"

Here, Strawson argues that in fact it can't be relations all the way down, but he agrees that physics only tells us about relations. There must therefore be something that physics leaves out, but that something is not unknowable: the fundamental nature of reality is consciousness.

Ontic Structural Realism

11 Março 2026, 17:00 • David Yates

According to ontic structural realism (OSR), there are fundamental non-supervenient relations between things in nature that are not determined by intrinsic properties of the relata. In Lewis' ontology, of course, spatiotemporal relations are non-supervenient, but almost everyone accepts that. In OSR, there are fundamental non-supervenient relations such as being of opposite spin, and many proponents of OSR say that fundamental particles in fact do not have intrinsic natures at all. There are no intrinsic powers or qualities, there are just fundamental nomic relations between particles. The world is a world of pure structure. It is relations all the way down. But is this picture even coherent? Surely, for there to be structure, there must be something that is structured?

Reading week 4: Michael Esfeld, "Quantum Entanglement and a Metaphysics of Relations"

Sections 1-3 are the most important ones. If you are not familiar with quantum physics, you may find some bits of this confusing, but there is also plenty of metaphysics here that does not depend on physics. All Esfeld really does with the quantum physics is argue that it leads to a certain kind of relational metaphysics. I will explain the relevant points in class, but try to read the paper in advance.

Causal Structuralism

4 Março 2026, 17:00 • David Yates

Sydney Shoemaker thinks that first-order properties are intrinsically powerful, so there is no need for N-relations. It is also not possible to recombine properties in any old way: not all patterns of properties in spacetime are possible. Shoemaker agrees with Lewis that causal laws are descriptions (i.e. not genuine, external relations) but thinks they describe the essential natures of properties such as mass and charge. Properties are powers and concrete things, it seems, must ultimately be defined by what they can do. The laws of nature are metaphysically necessary.

Shoemaker's position is known as causal structuralism (sometimes also dispositional essentialism).

In lecture 3 we discussed some issues related to relational essences. There are various problems with these. If you say that the natures of properties are determined by their places in a structure, then it seems you need that structure to somehow govern what properties do on specific occasions. In other words, the relations that fix the essence of a dispositional property like mass must also somehow make it the case that the mass of the Sun attracts the mass of the Earth. But that seems to be exactly the same as the governing role that Armstrong assigns to relations. How do you then embrace a powers ontology without appealing to the governing laws? Furthermore, the very idea of a relational essence seems to presuppose that relations could determine the identities of their relata. But how can the relata stand in a relation "before" having their identities determined? In other words, if you say two things A and B are essentially related by a relation R, then the relational state R(A,B) determines the natures of A and B. But how could it, given that the relational state seems to presuppose A and B for its existence? My own preferred view is a "no-priority" theory in which A, B and R(A,B) are all at the same level of fundamentality, and mutually ontologically dependent. This would be a case of symmetric grounding.

Suppose we say that powers are qualities as well, which have their causal aspects in virtue of their qualitative aspects. In other words, they are very much like sphericality, which enables spheres to roll in virtue of being the mathematically defined quality that it is. The so-called "powerful qualities" theory is also a response to the "filling in space" problem. If we can make sense of powerful qualities, it seems they resolve a lot of problems with powers ontologies! A somewhat different solution is to say what Neil Williams says: powers form a structure at our world based on their essential natures, but that structure does not determine their identities or govern what they do. Each power has the blueprint for the whole structure built into its nature. See section 4.3 of his book. Here is a taster: "The result of packing all this information in to the powers is that each property then contains within it a blueprint for the entire universe (not only as it is, but as it could have been). Each power property has within it organized plans for every possible circumstance in which it might find itself. These plans constitute instructions for what manifestation is to be produced in each circumstance", pp. 88-9.

Reading for week 3: we will discuss Sydney Shoemaker's paper, 'Causality and Properties'. You will find that (and also some of the other papers referenced in the course) in Kim & Sosa, Metaphysics: an Anthology, pp. 253-268.

If you want to read a very short paper that expresses the puzzlement many people feel when they try to imagine that every property is a disposition or powers, take a look at Simon Blackburn's article 'Filling in Space'.

Also very useful is Neil Williams' book, 'The Powers Metaphysic', which covers several of the key topics of this course including powers, structuralism and causation.

DTA account of laws

25 Fevereiro 2026, 17:00 • David Yates

Armstrong thinks there are necessary connections between properties, but that these are contingent necessities that only hold at worlds with the same laws of nature as ours. Properties are powerful, on Armstrong's account, but only because they are related by laws of nature (external N-relations that bind properties together and make it the case that the regularities hold (e.g. whenever you drop something, it falls; like charges repel). This week we will examine the advantages and disadvantages of this view, including Armstrong's solution to the problem of induction and the inference problem.

Inference problem (Bird's version): Armstrong only needs N-relations because he agrees with Lewis that first-order properties like mass and charge are not intrinsically powerful, but it seems that N itself must be intrinsically powerful if it is to make properties like mass and charge do what they do at our world. Why not then just say that mass and charge are intrinsically powerful?

Laws of nature - Humean supervenience

11 Fevereiro 2026, 17:00 • David Yates

We will begin our survey of the metaphysics of science by focusing on the central issue of laws of nature. What is a law of nature exactly? Is a law just a way of codifying the observed regularities, or do laws govern the regularities we observe? That is, which comes first: the patterns we observe in nature or the laws? There has been much debate over this issue, primarily between those who (following David Lewis) endorse a "best-systems account" of laws and those who endorse the "Dretske-Tooley-Armstrong" (DTA) conception of laws. To undetstand this debate we will first introduce "Hume's Dictum", the claim that there no necessary connections between distinct events, since this is the backbone of Lewis' best-systems approach to laws.

On the best-systems approach, the patterns come first: laws of nature are the axioms and theorems of a deductive system that is the best combination of simplicity and strength when it comes to explaining and predicting phenonena. There are no necessary connections between phenomena: nature is a vast Humean mosaic of property-instances distributed throughout spacetime, and the laws of nature are the best deductive system for describing that mosaic.

Reading for week 1: SEP Laws of Nature

For more details on Humean Supervenience, try this paper by Brian Weatherson