In the Natural History of Song project, we have assembled a database of 118 field recordings from hunter-gatherers, pastoralists, and other small-scale societies. Each has been transcribed by a team of musicians with efforts made to reduce a variety of potential biases present in traditional staff notation. For instance, we do not impose any particular meter on the transcription, nor do we segment melodies into what we hear as different phrases — the transcriptions are just an effort to get a rough sketch of the pitches and rhythms present in the singing.
With the raw audio and transcriptions, we are now coding a variety of musical features of each song, with the help of a team of volunteers from a variety of musical backgrounds. Each volunteer listens to each of the 118 tracks, while reading the transcriptions, and answers a series of questions about the song. Some questions are pretty objective (e.g., "is there more than one person singing?") but others are highly subjective and somewhat controversial (e.g., "is there a tonal center?").
The idea is to get input from as many expert listeners as possible, so that we can begin to characterize which features have agreement among many listeners and which do not. This is an interesting question in its own right from the perspective of music perception research, but it will also help us to guide analyses comparing the songs to one another across behavioral contexts (like lullaby vs. love song) and across world regions.
Anyone with training in music theory and/or ethnomusicology is welcome, and we are especially interested in hearing from those whose training is not in traditional Western music and/or those born in non-Western countries, to help diversify our team of listeners. At this time we do not have any compensation available, but volunteers will have first access to the datasets, if they are interested in collaborating with us on follow-up projects. The time commitment is modest; most coders work only a few hours per week for a few weeks, and all work can be done from anywhere in the world with an internet connection. The transcriptions and audio recordings are shared online and the coding is done via a simple and straightforward Google Form.
If you or anyone you know is interested in helping us out, please write me at
sam@xxxxxxxxxxxxxxx. This work is in collaboration with Manvir Singh, Nori Jacoby, Daniel Ketter, Daniel Pickens-Jones, Max Krasnow, Steven Pinker, and Luke Glowacki. It is based at Harvard's Evolutionary Psychology Lab.