We describe data materialization as a practice-based artistic research approach that translates data into physical art objects, emphasizing conceptual connection, interaction, and engagement over legibility. Drawing on an indicative review of recent data art practices, we propose data materialization as a distinct sub-genre of data art, informed by but separate from data visualization and data physicalization. We further propose a series of theoretical principles for this practice, and illustrate them through a series of original artworks. The key principles of data materialization are: 1) data are used as raw material, 2) data have a conceptual connection to the physical material, 3) data are intentionally abstracted and non-decodable, and 4) objects can be two- or three-dimensional and embrace analog processes. We frame data materialization as a mode of epistemic translation, a practice for turning abstract data into physical objects that express emotion and meaning rather than literal values. This paper suggests ways in which data materialization may transform data into shapes, forms, and installations that invite reflection, interpretation, and meaning-making.
Starrett et al. (Mon,) studied this question.