we just posted the Expanded Natural History of Song Discography, a corpus of audio & metadata. 1007 songs in many languages, for behavioral experiments, cross-cultural research, etc preprint: osf.io/preprints/psyarxiv/d2ftg_v1 corpus: zenodo.org/records/14927216 led by @milabertolo.bsky.social 1/ - ThreadSky

mehr.nz • 1 day ago

we just posted the Expanded Natural History of Song Discography, a corpus of audio & metadata. 1007 songs in many languages, for behavioral experiments, cross-cultural research, etc

preprint: https://osf.io/preprints/psyarxiv/d2ftg_v1
corpus: https://zenodo.org/records/14927216

led by @milabertolo.bsky.social

1/

1 / 2

Comments

mehr.nz•1 day ago

the psychology of music is a small field with big problem: like many areas of cogsci, most of the time we study western music, western musicians, western listeners, etc. but music is way more global than that (obviously)

2/

mehr.nz•1 day ago

unlike linguistics which has tons of cross-linguistic resources, music only has a few systematically built corpora, and what's available is often hard to use, doesn't have audio or metadata available, or has metadata that's unreliable, etc. we're trying to help fix that

3/

mehr.nz•1 day ago

6 years ago we published a paper in Science that did cross-cultural analyses of song recordings in a small but carefully built corpus we called 'the natural history of song discography'. it had 118 songs used for dance, lullaby, healing, and love in 86 societies

4/

mehr.nz•1 day ago

that first corpus enabled all sorts of fun research, like cross-cultural data science in the first paper (https://www.science.org/doi/10.1126/science.aax0868) but also, since the corpus is made of audio files, it can be used as a stimulus set for behavioral experiments

5/

mehr.nz•1 day ago

so for example you can sample lullabies from the corpus and play them to babies in the lab to see how they respond. @milabertolo.bsky.social and Connie Bainbridge did so, for example, finding that lullabies are relaxing to babies even when unfamiliar/foreign

https://www.nature.com/articles/s41562-020-00963-z

6/

mehr.nz•1 day ago

or you can play the songs to people all over the world, speaking many languages and in far-flung places (incl smaller-scale societies), to see what people think. Lidya Yurdum, @manvir.bsky.social and others did so — turns out listeners know what they're hearing

https://doi.org/10.1073/pnas.2218593120

7/

92hits.bsky.social•1 day ago

📌

maxprimbs.bsky.social•1 day ago

Very cool!

Comments

Posting Rules

Reply