Composition and Simples

Core Idea

The Special Composition Question asks: under what conditions do some objects compose a further object? Mereological universalism answers 'always' — any collection of objects, however scattered, composes something (your left shoe and the Eiffel Tower compose an object). Mereological nihilism answers 'never' — strictly speaking, there are only mereological simples (partless entities) arranged in various ways, and what we call tables and chairs are just simples arranged table-wise or chair-wise. Van Inwagen's moderate answer is that composition occurs only when the activity of some objects constitutes a life — so organisms exist as genuine composites but artifacts do not. The question of simples is equally vexing: must simples be point-sized, or can they be extended? Could the material world bottom out in 'atomless gunk,' with every part having further proper parts?

How It's Best Learned

Read van Inwagen's Material Beings chapters 2-4 for the Special Composition Question and his life-based answer. Then read Sider's 'Against Parthood' for a deflationary approach. Test each view against ordinary judgments: does your theory say tables exist? Should it?

Common Misconceptions

• Mereological nihilism does not deny that there are arrangements of matter where we see chairs; it denies that those arrangements constitute a single composite object over and above the simples.
• Universalism does not treat every arbitrary collection as a natural kind — it simply says the mereological sum exists, not that it is interesting or unified.

Explainer

From mereology basics you know the fundamental framework: objects can stand in part-whole relations, and we can ask formal questions about overlap, proper parts, and mereological sums. Now the question becomes substantive rather than formal: when, if ever, do some objects actually compose a further object? This is the Special Composition Question, and the answers philosophers have given range from the radical to the commonsensical.

Mereological universalism says: always. Whenever you have any collection of objects whatsoever — your left thumb, the Eiffel Tower, a photon in the Andromeda galaxy — they compose a mereological sum. There is an object that has exactly these things as its parts. This view has the advantage of formal elegance: you never have to worry about when composition "kicks in" because it never doesn't. But it multiplies entities dramatically. On universalism, the material world contains an astronomical number of objects that no one would ordinarily recognize — arbitrary scattered sums with no causal unity, no natural boundary, no interesting properties over and above those of the parts. Most people feel this proves too much: the world seems stuffed with metaphysical junk. The universalist response is that this "junk" doesn't actually burden us — it's ontologically harmless, positing no causal powers beyond what the parts already have.

Mereological nihilism takes the opposite extreme: composition never occurs. What we call tables, chairs, mountains, and people are not genuine single objects — they are simples arranged in various ways (table-wise, chair-wise, mountain-wise, person-wise). The only things that truly exist are mereological simples: objects with no proper parts. When you see a chair, you are really perceiving a vast arrangement of simples whose collective properties produce the appearance of a unified object. Nihilism has its own appeal: it is ontologically parsimonious, avoids puzzles about identity and persistence of composites, and sidesteps questions like "does a table survive losing a leg?" (There is no table, only simples.) The cost is revisionary: ordinary talk about chairs and people is strictly false, or at best a convenient fiction.

Peter van Inwagen's moderate view sits between these extremes. His answer to the Special Composition Question: composition occurs if and only if the activity of the parts constitutes a life. Organisms are genuine composite objects — a living body is not merely simples arranged organism-wise, but a real thing whose parts are integrated by biological processes. Artifacts like tables and ships are not genuine composites. Van Inwagen thus preserves persons (crucial for his broader project about personal identity) while rejecting the clutter of arbitrary composites. The view is philosophically motivated but strikes many as ad hoc: why should biology be the cut-off? Why not chemistry (molecules as composites) or physics (atoms)?

The question of simples is equally puzzling. Simples are supposed to be the bottom level — things with no proper parts. But what could a partless thing be like? If it is physically extended, can't we ask about its left and right halves? Physics suggests no fundamental partless particles, since quarks may be structured and even particles have quantum fields. The disturbing possibility is atomless gunk: matter all the way down, where every part has further proper parts and there is no bottom level of simples. If gunk is metaphysically possible, then the entire framework of "build complex objects from simples" is not necessarily how the world is structured. The Special Composition Question and the question of simples are thus deeply entangled: what composition is depends partly on what the fundamental constituents are.