One Song, Many Free Uses
From public domain to Spotify: understanding real-world access to culture
Livonian is a critically endangered European language related to Estonian, Finnish, and Hungarian. A small community of Livonians are re-learning their ancestral language, and singing together is one of the most effective ways to practice.
For the first time, we made a recording of a Livonian dance song accessible on Spotify, Apple Music, and Amazon. These platforms make the song available in almost every country, and some—like Spotify—offer it for free if you are willing to listen to advertisements.
You could say that the music is free. But since you are not allowed to download or reuse it from Spotify, is it public domain?
One song, many uses
In music, audiovisual works, and translated literature, asking whether something is “public domain” is often misleading. A musical work cannot be experienced without a performance, a recording, and a distribution platform. Each of these layers can have its own rights and conditions.
This is why we designed a different rights model in the Open Music Observatory and its national and regional modules. Instead of asking whether a work is simply free or not, we describe how it is available across different contexts—public and private, free and remunerated.
The Livonian song illustrates this clearly. The underlying composition is in the public domain. Yet its recordings, notations, and specific uses may be subject to different conditions. These conditions often do not pose any monetary burden on your use, but it ask for attribution, respect, or other conditionality. Our database reflects this reality. It shows that the same song is available through multiple access points, each with its own usage policy.
However, one policy does not fit all uses. If your café is licensed for public performance, you can legally play this recording in a commercial setting. If you are working in a non-commercial context, such as research, you may obtain access to the originally digitised .wav file under more permissive conditions.
For the archival version, we create a separate node in our knowledge graph. This is presented as a readable table for human users, and as structured data (TTL, XML, JSON, or other serialisations) for machines.
The data page of Tšitšōrlinki, tšitšōrlinki (also discoverable as Tsitsorlinki, tsitsorlinki without diacritics) shows the different ways you can access and use this public domain folk dance song:
as a sound recording with different usage conditions
as musical notation (handwritten and printed)
as standardised Livonian lyrics
Each of these manifestations has its own access points and usage policies. The advantage of our graph model is that these options are connected and can be explored together.
Is this public domain?
It depends—and this is where public domain databases often fail to give a useful answer.
The song itself—the melody and lyrics—is in the public domain. You can learn it, perform it, and adapt it freely. Our database makes this explicit.
Yet the ways you can access and use the song today depend on the specific manifestation: the recording, the format, and the platform.
