The Special Composition Question
24 Aug 2025Praise, condemnation, questions, and comments can be sent via email
Last week I mentioned that one of the questions that metaphysicians ponder is “Under what conditions, if any, does a group of objects form a composite object?” This question is known as the Special Composition Question, and it will be the topic of today’s blog post. The Special Composition Question (henceforth SCQ) was coined by perhaps the greatest U of R metaphysician to date, Peter van Inwagen. Van Inwagen distinguishes the SCQ from the General Composition Question (henceforth GCQ), which asks “What is the nature of composition?”
It may seem that these two questions are both aiming to understand what composition is. But note that we don’t need any rigorous account of what composition is in order to answer the SCQ. Consider by analogy the questions “What is (romantic) love?” and “Under what conditions do people fall in love?” I can’t tell you with any certainty what the nature of (romantic) love is, but I can tell you that it occurs when there is trust, familiarity, fondness, and sexual attraction. In answering the latter question, I haven’t told you anything about what love is, but I have given you the proper conditions for love.
Today the focus will solely be on the SCQ and some solutions to it. I will caution my readers that this discussion will be woefully inadequate. There is a massive philosophical literature on the SCQ, and I have neither the time nor the space here to do it any real justice. What I hope to instill in you, dear reader, is at least a passing familiarity with the broad structure of the debate and the curiosity necessary to further the SCQ further.
A note before we consider solutions. In exploring the solutions below, I’m going to say some strange things, like that tables and chairs don’t exist. I don’t mean that there aren’t objects in the world that we can sit on or that we can eat dinner at. When philosophers consider questions like the SCQ, we’re concerned with what the real, underlying structure of reality is. When I say that tables and chairs don’t exist, I just mean that they are not part of the basic furniture of reality (pun intended). We don’t necessarily have any reason to think that the basic furniture of reality will correspond to our common sense notions of what stuff exists in the world.
Design for Export Furniture, drawing, anonymous, Chinese, 19th century
The first reply to the special composition question we’ll consider is nihilism. The nihilist says that composite objects don’t form under any conditions, the only thing that really exists are physical simples. Physical simples are commonly taken to be the fundamental particles given by the standard model of physics. On this view there is no such thing as chairs or tables or cars or any of the other everyday objects that make up our lives.
One advantage to this view is that it is extremely clean, the basic furniture of the universe is soundly based in scientific fact and there is little vagueness to consider. A disadvantage is that we must do some extra work to make sense of all the every day talk about composite objects that seems perfectly reasonable. A common reply for the nihilist is to make a distinction between speaking strictly and speaking loosely. When I say “there is a chair in my room,” I’m speaking loosely, and using convenient terms even though these terms don’t match the actual structure of the world. If I were speaking strictly, I might instead say “There is a collection of physical simples arranged in such a way that I can sit on them, place clothes on them, roll them across the carpet etc.”
The second reply we’ll consider is universalism. The universalist says that any collection of simples forms a composite object. Thus, for the universalist not only do chairs and tables exist, but so does the composite object made up of my chair and the moon, or the composite object object made up of only the north half of the Eiffel tower. The universalist need not think that these objects are fundamental, though they are committed to these objects being part of the real structure of the world in some metaphysically robust sense.
The universalist position, of course, has the advantage of accounting for all of our common sense objects, like tables and chairs. The disadvantage is that we seem to have overshot the mark, and allowed a whole bunch of seemingly nonsense objects, like the object composed of my chair and the moon, into our ontology. These objects also overlap, so there is both an object that is my chair and an object composed of my chair and the moon that are at least partially co-located. The world, then, becomes much more crowded for the universalist.
The universalist, like the nihilist, also has to deal with problems of language. When I say “there is a chair in my room”, which object am I referring to? There are, according to the universalist, millions of objects in the vicinity of what we might commonsensically consider my chair, differing in composition by as much as an entire armrest or as little as a single electron. One reply the universalist might give is that what occurs is a reference lottery. I am definitely referring to a single object when I say “there is a chair in my room,” and it’s just a matter of chance which definite one of the millions of candidates I actually refer to.
Design for Export Furniture, drawing, anonymous, Chinese, 19th century
Any answer in between universalism and nihilism is a restrictivist response. Restrictivists take the middle position that simples sometimes form composite objects, but not all the time. There are many different ways to be a restrictivist but here we’ll focus on van Inwagen’s response. Van Inwagen, in his book Material Beings, argues that a collection of physical simples forms a composite object only when the composite object is a living thing.
One reason to think van Inwagen’s answer is right is that, for one, humans are composite objects. We should be skeptical of the nihilist’s answer for the reason that I exist and I’m a composite object so composition has to happen under at least some conditions. We also need some principled way to cut down the massive ontology of the universalist. Living things naturally have a certain organization of their composite parts, as well as a reasonably clear boundary of non-existence (death).
Van Inwagen’s account does have some issues. As van Inwagen himself admits, he is committed to the existence of apple trees, which are living things, but not apples, which are merely former parts of a living thing. This is a strange line to draw. Van Inwagen also admits to some amount of vagueness in his account: there are some things for which it is not clear whether they are definitely living or definitely not living. We might reasonably prefer an account of composition that draws more definite lines between composite and non-composite objects.
I generally am inclined towards mereological nihilism, as I think it gives us a nice, relatively clean ontology (the standard model of physics), and I don’t think we have any reason to think that our natural language talk should be perfectly attuned to the fundamental nature of reality. But I also am still learning about the debates, and constantly reformulating my own opinions. In fact, just last spring I presented a paper that argued for a restrictivist account.
If you are interested in exploring further, I recommend this article from the SEP on material constitution that gives a simultaneously higher level and more detailed account of debates around the metaphysics of objects, and this article on ordinary objects that nicely covers the different ways that ontologists might accommodate the existence of the everyday furniture of reality.
Please let me know your thoughts! Thanks for reading!