When Does a Heap Stop Being a Heap

The Greek philosopher Eubulides of Miletus posed the puzzle in the 4th century BCE, and 2,400 years later nobody has cleanly solved it. Start with a heap of, say, 10,000 grains of sand. Remove one grain. Still a heap, obviously — one grain can't make the difference. Apply that same harmless rule 9,999 times and you are staring at a single grain on the floor, which the rule insists is also a heap. The conclusion is absurd. The reasoning looks airtight.

The puzzle's name comes from the Greek soros, meaning heap. It is not really about sand. The same structure breaks any vague predicate: a person with no hairs is bald, one extra hair won't change that, repeat enough times and someone with a full head of hair must be bald. Or in reverse: a person 1.50 m tall is short, an extra millimeter won't change that, so a 2 m basketball player is short, too. Vague concepts seem to lose their footing the moment you put pressure on them.

Proposed escapes all hurt. "Epistemicism," defended by Timothy Williamson, says there is a precise sharp boundary — heap above N grains, not below — but humans simply cannot know where it sits. Most people find that worse than the paradox. "Supervaluationism" lets the boundary be many possible boundaries at once. "Fuzzy logic" gives propositions truth values between 0 and 1, so a borderline heap is 0.6 true, which philosophers tend to dislike for reasons that are themselves vague.

The paradox lives because human language and human cognition both run on fuzzy categories — bald, tall, rich, old, dead — that work fine until you ask exactly where they begin. The Sorites is the oldest reminder that the universe is continuous and our words are not.