The Problem of Universals

In brief

Look at a ripe tomato and a fire truck. Both are red. What, exactly, do they share? The naive answer is "redness," a single thing present in both. But that is strange: how can one thing be wholly in two places at once? The problem of universals is the fight over what to say here. A universal is whatever it is that many particular objects can have in common: a property (redness, roundness) or a kind (humanity, gold). Realists hold that universals are real entities. Nominalists hold that they are not, that the world contains only particular things and our habit of grouping them under shared names. Plato and Aristotle staked out the two great realist positions in the fourth century BCE; the medieval schoolmen turned the dispute into a science; and it remains, unresolved, at the root of debates about properties, natural kinds, mathematics, and meaning.

The full treatment

The problem it answers

Language and thought traffic constantly in general terms. We call many different animals "dog," many surfaces "red," many actions "just." This works: the terms cut the world at recognizable joints and let us predict and explain. The question is what in reality makes it work. When we truly say "this rose is red" and "that flag is red," is there some one thing, redness, that both instantiate? Or is "red" merely a label we have learned to apply to a scattered heap of individually red things with nothing literally in common? The problem forces a choice about the basic furniture of the world: does it contain, besides particular objects, a second category of repeatable entities?

Plato: universals exist apart from things

Plato (c. 428 to 348 BCE) gave the boldest answer. Behind the many beautiful things stands a single Form of Beauty; behind the many equal things, Equality itself. These Forms are real, changeless, and perfect, existing in their own right independent of the sensible objects that "participate" in them. A particular is beautiful only by imitating or sharing in the Form. The Forms are not in space or time and are grasped by intellect, not by the senses. This is realism of the strongest kind, often called ante rem realism (universals exist "before the thing"). It explains a great deal: why definitions are possible, why mathematics is exact, why the flux of the visible world can still be known. Plato argues for it in the Phaedo and the Republic, where the Form of the Good stands at the summit of what can be known.

Aristotle: universals exist only in things

Aristotle (384 to 322 BCE), Plato's student, accepted that there are universals but denied they exist separately. For him the universal is real but immanent: humanity exists, but only in individual humans, never floating free. This is in re realism (universals exist "in the thing"). In the Categories and the Metaphysics he attacks the separated Forms as an idle duplication of the world that explains nothing about how things actually change and grow. A living thing has its form as an organizing principle inside it, not in a separate heaven. Aristotle's view preserves the reality of kinds and properties while keeping metaphysics tied to the observable individual. The two founders thus split the realist camp in a way that has never fully closed.

The medieval question and the nominalist answer

The debate was handed to the Latin Middle Ages through a single passage. Porphyry's third-century introduction to logic, the Isagoge, asked whether genera and species exist in reality or only in the mind, and Boethius (c. 477 to 524), translating it, declined to settle the matter. For seven hundred years that open question drove Scholastic philosophy. Peter Abelard (1079 to 1142) developed conceptualism: universals are neither things in the world nor mere words, but concepts the mind forms from real resemblances. Thomas Aquinas (1225 to 1274) defended a moderate Aristotelian realism. The decisive break came with William of Ockham (c. 1287 to 1347), the great nominalist: there are no universals in reality at all, only individual substances and qualities; a universal is just a natural sign in the mind, a general concept, standing for many particulars. Ockham's parsimony, the intellectual habit behind Occam's Razor, told him not to posit repeatable entities the world does not need.

Distinctions that matter

Four positions recur, and it is worth keeping them apart. Platonic (ante rem) realism: universals exist independently of their instances. Aristotelian (in re) realism: universals are real but exist only in their instances. Conceptualism: universals exist only as concepts in minds. Nominalism: universals do not exist in any form; there are only particulars, grouped by names or resemblance. A fifth, modern option, trope theory, agrees with nominalism that there are no shared, repeatable entities but adds that properties themselves are particular: the redness of this tomato is a distinct individual (a "trope") from the redness of that flag, and the two merely resemble each other.

Lineage

The problem is arguably the birth of systematic metaphysics. It runs from Plato and Aristotle through the Neoplatonists and Porphyry, into Boethius, and then across the whole arc of Scholastic thought: Roscelin of Compiègne (c. 1050 to 1125), an early extreme nominalist, Abelard, Aquinas, Duns Scotus (c. 1266 to 1308) with his doctrine of individuating "haecceity" or thisness, and Ockham. Modern philosophy carried it forward: the British empiricists (John Locke and George Berkeley) argued over whether the mind can even form a general idea, while the rationalists leaned Platonist, a tension traced in rationalism and empiricism. In the twentieth century the debate was revived with new precision by W. V. O. Quine and Nelson Goodman on the nominalist side (their 1947 paper "Steps Toward a Constructive Nominalism") and by David Armstrong on the realist side (Universals and Scientific Realism, 1978).

The strongest case for it

Here "it" is realism, since realism is the position that posits something extra and so bears the burden of argument. The case is strong. First, the "one over many": when several things are genuinely the same in some respect, there seems to be one thing they share, and denying it looks like refusing to explain an obvious fact. Second, general truths: "red resembles orange more than it resembles blue" appears to be a truth about redness itself, not about any particular red thing, which suggests redness is something. Third, science: physics quantifies over properties like mass and charge that recur exactly across countless particles, and its laws seem to state relations between these repeatable properties. Armstrong argued that our best scientific theories commit us to universals as the truthmakers for such laws. Fourth, mathematics: numbers and sets behave exactly as Platonic realism predicts, existing outside space and time and known by reason, which is why many working mathematicians are unashamed Platonists.

The strongest case against it

The nominalist reply is old and formidable. Plato himself, in the Parmenides, raised the Third Man objection: if all large things share the Form Largeness, then Largeness and the large things form a new group that is large, requiring a further Form, and so on without end. The separated Forms threaten an infinite regress. Aristotle pressed the charge that Platonic Forms are explanatorily idle, a needless doubling of the world. Against realism generally, nominalists ask how a single universal can be wholly present in many places at once, which seems to violate the ordinary logic of identity. Ockham's razor cuts here: if particulars plus resemblance can do the work, universals are surplus ontology. In the twentieth century Quine urged a "desert landscape" ontology, admitting only what our best theories strictly require, and Goodman rejected abstract objects outright, building a nominalist logic that quantifies only over individuals. Trope theorists (D. C. Williams in 1953, Keith Campbell in Abstract Particulars, 1990) argue they can capture everything realism explains, the sharing, the resemblance, the truthmakers, without any repeatable entity, since a class of resembling tropes does all the same work. The realist's "one over many" is met by the nominalist's charge that the realist has never given a coherent account of how one thing is many.

Where it stands now

The problem is not solved, which is part of what makes it foundational rather than dated. Contemporary analytic metaphysics contains committed realists (Armstrong's immanent universals, and Platonists about numbers and sets), thoroughgoing nominalists (in the Quine and Goodman line, and modern "ostrich nominalists" who deny that the one over many needs any answer at all), and a large and active trope-theoretic middle. The stakes have widened: the same choice reappears in the philosophy of mathematics (do numbers exist?), the debate over natural kinds and scientific laws, and the semantics of predicates. What Plato and Aristotle disagreed about over a tomato and a fire truck turns out to be the same disagreement that divides those who think the world comes carved into real kinds from those who think we do the carving. The terms have been refined for twenty-four centuries; the two intuitions behind them have not gone away.

Test yourself

Point at two red objects and ask yourself, without reaching for the word "red," what they actually have in common. Is it a single thing present in both, a pair of separate but resembling qualities, or nothing at all beyond your own decision to file them together? Notice which answer feels obvious to you. That instinct is your starting position in a debate that has never been won.