Data as Collectively Generated Patterns: Making Sense of Data Ownership

Data ownership is power. Who should hold that power? How should data be owned? The importance of data ownership explains why it has been analogized to other domains where ownership is better understood. Several data-as proposals are on the table: data as oil, as intellectual property, as personhood, as salvage, data as labor, etc. Here I propose another way of thinking about data. Like the others, my view characterizes data in ways that make them accessible to ownership considerations and can be expressed as a data-as view: data as collectively generated patterns. Unlike the alternatives, data as collectively generated patterns does not create any equivalence with another domain where ownership is already well-understood. It reveals how ownership considerations enter, but we must explore afresh how they do. Accordingly, I propose a way for ownership considerations to bear on data once we understand them that way. And if we did understand them that way, the internet should presumably be designed very differently from what we have now. The themes in this paper require more elaboration than what can be done here, so some follow-up work is needed.