A definition that turns back on itself

“Polysyllabic” is easy to classify: it has several syllables, so the word describes itself. “Heterological” is designed to mean the opposite: a word is heterological if it does not describe itself. The trouble begins when we try to apply that definition to the word “heterological.”

Suppose it does describe itself. Then it is not heterological, because heterological words do not describe themselves. But if it does not describe itself, it meets the definition of a heterological word—which means it does describe itself after all. Either answer seems to reverse itself.

Why the loop forms

The puzzle is a form of self-reference: a statement or description is applied to the very thing that expresses it. The definition does not merely classify other words; it invites the word naming the category to be classified by that same category.

That creates a cycle between two claims: “This word describes itself” and “This word does not describe itself.” Each proposed classification triggers the opposite one. The puzzle is not about how many syllables “heterological” has, or whether it contains a literal description of itself. It depends on treating “describes itself” as a precise yes-or-no test and then applying the test to its own label.

The useful distinction

Many words avoid the problem because their meanings concern a property that can be checked directly. “Polysyllabic” can be counted. Heterological, by contrast, refers to whether a word applies to itself, so its meaning points back toward the classification process. The paradox comes from that self-application, not from an ordinary contradiction hidden in the word’s spelling. It is a compact demonstration of how a definition can become unstable when it is allowed to judge itself.