← Three short discoveries on WatWat
A choir in blue robes singing at rehearsal inside a stone cathedral
Art

Musical emotion is usually described on two axes, pleasant against unpleasant and calm against stirring. To see whether two are enough, listeners in the United States and in China rated what each of 2,168 music excerpts made them feel, which is a direct comparison across cultures. How many does it actually take?

Tested in peopleSample: 2,168 excerpts, listeners in two countries

Thirteen, and they blur into each other. The categories came out of the ratings rather than from a list the researchers wrote, and they include amusement, anxiety, dreaminess, triumph and a sense of . Specific labels of that kind held up across both cultures better than the broad pleasant and stirring measures did. The feelings behaved as a , not as separate boxes.

Sorting music into happy, sad, calm and stormy is convenient, and the real question is how much it throws away. Two axes can only ever place a piece on a grid, and a listener rarely reports a position on a grid.

Listeners in the United States and in China rated the feelings raised by 2,168 music excerpts, and the categories were extracted from those ratings rather than imposed on them. The cross-cultural design is what makes the result more than a fact about English emotion words.

The analysis found thirteen dimensions, among them amusement, anxiety, dreaminess, triumph and awe. Specific labels survived the crossing between cultures better than the broad measures of valence did, which is the opposite of what a two-axis model predicts.

The second finding matters as much. The feelings did not separate into boxes but formed a continuous space, with no boundary where one becomes the next. Two countries are still not every culture, and the handed to participants sets the edges of any map built this way.

wat. Across 2,168 excerpts, listeners in China and in the United States shared 13 dimensions of feeling with blurred borders between them.

wat wat

Music carries a richer emotional experience than a sorting into happy, sad, calm and stormy can hold, and the extra categories are the interesting part.

How do we know?

Cowen, Fang, Sauter and Keltner ran large-scale rating experiments on 2,168 excerpts in two countries, and derived the dimensions from the ratings with data-driven statistical methods.

Tested in peoplePeer-reviewed studySample: 2,168 excerpts, listeners in two countries

What is still unknown

Two countries do not represent every culture. The word list offered to participants also shapes the map, since a feeling with no available word cannot appear on it.

Learn something?

Cowen et al., PNAS· 2020· DOI 10.1073/pnas.1910704117

Where to next

Why do so few scales get used out of all the possible ones?musicDoes music move you because of the sound, or because you know it?musicWhy would anybody put on music they know will make them sad?music

Want another one worth repeating?

Start with three short discoveries. No sign-up, nothing to pick first.

Surprise me
Browse the topic libraryAbout WatWat, and how sources get checked