The Fundamentality of Composite Things

We can use grounding to define fundamental entities:

F1. A thing x is fundamental iff the existence and intrinsic properties of x are not wholly grounded by the existence and intrinsic properties of any other thing or things.

Some have suggested that we add the qualification: not wholly grounded by the existence and intrinsic properties of any other thing or things, except for parts of x.

However, this would be a mistake. An Aristotelian substance is not grounded by the existence and nature of its parts: this is what distinguishes substances from mere heaps.

Let’s say that a part x is an essential part of some whole y iff it is essential to x that it be a part of y. Should we allow a fundamental thing to depend on its essential parts? This seems less objectionable than dependence on non-essential parts. An essential part is in a sense an extension of the whole, and so making the whole depend on it wouldn’t seem to compromise its fundamentality.

Is there an alternative way to secure the fundamentality of a composite whole? That is, can there be composite fundamental things that have non-essential parts? It seems reasonable to suppose that the existence and intrinsic character of a whole is a function (supervenes upon) the existence and intrinsic character of its proper parts. If it seemed that that were not the case, it seems reasonable to suppose that we have somehow missed some of the whole’s parts in our inventory. In some sense, the whole cannot be more than the sum of its parts.

But the necessary correlation between the existence and character of the whole and the existence and character of the parts must have an explanation in terms of grounding. Either the existence and character of the whole is grounded by the existence and character of its parts, or the character and existence of the parts is grounded by the existence and character of the whole, or both are grounded by some third fact. The only plausible candidate in this case for a common ground would be the character of the whole’s substantial form.

But, as a matter of fact, the existence and character of the parts are not wholly grounded by the whole substance or by its substantial form. So, if the parts are not essential parts, the composite whole cannot be fundamental. So, the only way for a composite thing to be fundamental is for its parts to be essential parts of it. We could call this Aristotle’s Principle, given its importance in his scheme:

Aristotle’s Principle. All the material parts of a fundamental composite thing are essentially parts of it.[1]

Is it really necessary that a fundamental thing have only essential parts? Wouldn’t it be sufficient if at least one of its parts were essentially a part? What is crucial, I think, is the fundamental thing not be the mereological sum of a class of things that aren’t essentially parts of it. In other words, a fundamental thing is such as to have essentially at least one part that has only parts that are essentially parts of the fundamental thing. For simplicity’s sake, I’ll ignore this partial relaxation.

F2. A thing x is fundamental iff x is essentially such that the existence and intrinsic properties of x are not wholly grounded by the existence and intrinsic properties of any other thing or things, except for essential material parts of x.

What about accidents of a substance? It does seem that the intrinsic character of a substance depends on (is grounded by) its accidents. This will be fine, so long as we assume that each accident is essentially an accident of its actual substance.

In addition, as we shall see, Aristotelians are typically committed to another category of fundamental things: boundaries, like surfaces. These are neither parts nor accidents of substances.

F3. A thing x is fundamental iff x is essentially such that the existence and intrinsic properties of x are not wholly grounded by the existence and intrinsic properties of any other thing or things, except for x’s essential material parts, boundaries, and accidents.

By this definition, are accidents, boundaries, and proper material parts of substances fundamental entities? I think so. This seems pretty clear in the case of accidents. Socrates’ musicality is not grounded by Socrates’ intrinsic properties, since it is that accident of musicality that explains why Socrates is musical.

What about proper material parts of substances? Is the intrinsic character of a proper part of a substance grounded by the intrinsic character of the whole substance? Here too I would say that the existence and character of Socrates’ right thumb is not wholly grounded by the right-thumbedness of Socrates, but the other way around.

Let’s stop and ask again whether it is really necessary for the parts of a fundamental thing to be essential parts, as Aristotle’s Principle. Couldn’t there be other ways of securing the fundamentality of a whole? Let’s consider some possible alternatives—four in fact.

1. Cooperation toward a common goal

We could allow a fundamental thing to have non-essential parts, so long as the parts collectively cooperate toward a simple common goal or end. This would involve including many artifacts as fundamental entities, so long as the artifact’s parts do in fact function together effectively.

The problem is that artifacts seem clearly non-fundamental, precisely because their parts are so completely independent of the whole.

2. Internalization of a common goal

The parts of an artifact have not in any sense internalized the common goal to which they are ordered. Their ordering to that goal depends entirely on the extrinsic agency of their maker. What if we added an internalization condition, by which I mean that the parts somehow acquire fundamental causal powers that are ordered to a common end? An example of this would be the members of a social group who have internalized the group’s internal goods (in MacIntyre’s sense) as their own personal goals. This provides groups with a greater degree of unity than is enjoyed by the parts of an artifact.

The members of a group are intrinsically related to the whole group, but they are related accidentally and not essentially. The members of the group could have existed even if the group had never existed. Their essences make no reference to the group. This is why members are not essential parts of the group.

This still isn’t sufficient, since the fact that the powers of the members of the group cohere as they do could be merely contingent, even coincidental. Suppose a group spontaneously forms to pick up litter from a park. The members of the group cooperate toward a common end and do so because of internalized powers and intentions. However, this isn’t sufficient to fuse them into a fundamental entity. They are just a plurality of entities that happen to be cooperating. They may have acquired new collective powers and dispositions, but collective powers need not be powers of a collective. Even if the powers involved are fundamental powers, their bearers could be the members, taken both individually and collectively.

3. Internalization of ends relating to the group as such

We can close this gap by insisting that the members of the group act for the sake of the group itself—for its prosperity, survival, etc. Now each member is intrinsically related, not just to the other members, but to the group as such. There is now no grounding basis that is group-free, by which the existence and character of the group could be grounded. It seems that the group would be a fundamental entity, even though it has non-essential parts (the members).

This moves a little too fast. We’ve certainly established that in this case some new fundamental entity is in play. However, I think that the fundamental entity is not the group, in the sense of the composite thing that has the members as parts, but something like the particular practice of being-this-group. Take the Houston Astros. The individual members of the team have irreducibly Astros-directed intentions, desires, powers, dispositions, and so on. They want the Astros to make the playoffs. They aim to restore the reputation of the Astros for fair play, etc. So, there is something Astros-related that is not wholly grounded by individualistic facts about the members. But suppose we posit a shared or collective accident of being-the-Astros, an accident that inheres in each of the members. This accident is a fundamental thing and has only essential parts—e.g., essentially Astros-related actions, desires, etc. We could suppose that the team is a derived entity, not fundamental. The existence of the team depends on the existence and character of the players, including this shared accident.

4. Adding novel, emergent powers to the whole (Merricks, Objects and Persons)

Fundamental causal powers can be borne only by fundamental entities. So, it should be sufficient for a composite thing to be fundamental for it to have (as a whole) fundamental causal powers. Perhaps what the Astros team was lacking was any fundamental causal power at the team level. And couldn’t a composite thing have fundamental causal powers while also having non-essential material parts?

Are these active or passive powers?

If the powers are active, then there doesn’t seem to be any reason to posit a fundamental whole as their bearer. Instead, we could simply think of the powers as borne collectively by the non-essential parts (a point made by Meehl and Sellars in 1956 and reaffirmed later by Sydney Shoemaker).  These powers could be novel or emergent simply in the sense of not being manifested except in situations in which many parts are brought together in the right way.

What about passive powers? Passive powers might be able to demand a single bearer. Take sensation, for example. Isn’t that a passive power of the whole organism, not grounded by the powers or dispositions of the organism’s parts?

So, yes, in cases like sensation, there must be a fundamental thing in addition to the material parts. However, we have to ask: what (if anything) makes it the case that this additional is composed of those material parts? Why isn’t the emergent entity some disjoint, immaterial thing that interacts with the material parts without comprising them? It is not enough for this new entity to be causally (i.e., by way of efficient causation) dependent on the material parts. A whole must be metaphysically, constitutively dependent on the parts. If any of those parts are non-essential parts, then the whole is disqualified from being fundamental, and so not fit to be the bearer of a fundamental passive power.

---

[1] What about prime matter? I don’t want to count prime matter as a material part of the substance. A material part is a part that is enmattered, and prime matter is not itself enmattered. Since prime matter has no intrinsic character of its own, it poses no threat to the fundamentality of the substance.