a note nobody plays, assembled in the listener
Play two notes together, loudly enough and in tune, and a third note arrives that no one played. The violinist Giuseppe Tartini noticed it in 1714, while playing double-stops — two strings at once — and waited forty years before setting it down in his Trattato di musica in 1754. He called it the terzo suono: the third sound.
Its pitch is not a mystery note. For two tones at frequencies f1 and f2, the most easily heard third tone sits at the difference, f2 − f1. Take a just perfect fifth — say 400 Hz and 600 Hz, a 3:2 ratio — and the difference is 200 Hz, exactly an octave below the lower note. A fainter sum tone can also appear at f1 + f2. Violinists still use the difference tone as a tuning instrument: when the fifth is exactly in tune, the third tone below it sits steady and clear, so the ear can check the interval by listening for a note that is not being bowed.
Headphones help, at a moderate volume. First play the pair and listen underneath it for the low note; then play the 200 Hz tone alone, so you know what you were listening for; then try the pair again. Nothing on this page generates a 200 Hz signal in the first button — your browser sends only the two sines. If you hear the low note anyway, it was made after the sound reached you.
Where is it made? Mostly, in the ear itself. A perfectly linear hearing system would simply add the two incoming waves and pass them on, and no new frequency could appear. The cochlea is not linear — its active amplification of quiet sounds distorts loud ones slightly — and a nonlinear system fed two frequencies produces new components at their sums and differences. The third tone, on this account, is a distortion product the listener manufactures: a sound with no storage state and no source in the room, assembled at the moment of hearing from two harmless parts, the way the onion's tear gas is assembled at the moment of the cut.
Except the story refused to stay that tidy. Tartini himself insisted the third sound was physically there, and for centuries the respectable position was that he was wrong and it was all in the ear. Then a Florence-led team recording a professional violinist on five violins (work reported in 2020–2022) decomposed the sound in the air and found combination tones there too — present for every violin, strongest in the oldest, a Bologna instrument of 1700, whose most prominent combination tone measured about three-quarters larger than a modern mass-produced violin's. So the assembly has two workshops: instruments can make a little of the third tone in the air, and ears make more of it on arrival. Which one Tartini heard, in which proportions, is not a question the record can now split.
Organ builders turned the effect into architecture long before it was settled. A true 32-foot organ pipe is enormous and expensive; a resultant bass instead sounds a 16-foot pipe together with its fifth — the 10⅔-foot rank — and the listener's ear supplies the difference tone an octave below, a floor note no pipe in the building is long enough to make. The deepest note in the church exists, in part, only in the congregation.
An intact pair of tones contains no third tone, in the way an intact onion contains no tear gas: the parts are present, the partitions are the listener, and the thing itself is an event. It is stored nowhere. It happens — in you.
Sources: Combination tone — Wikipedia · PNAS, “Physical basis of two-tone interference in hearing” · New Scientist (2022) on violin combination tones in the air · Masetti et al., characteristics and perceivability of combination tones in violins (PDF)