Our chapter in Representations in Scientific Practice Revisited discusses the development of authoritative collections of brain scans known as “brain atlases”, focusing in particular on how such scans are constituted as authoritative visual objects. Three dimensions are identified: first, brain scans are parts of suites of networked technologies rather than stand-alone outputs; second, they are specified by means of a “database logic” that makes particular neurological features visible within a register of possibilities; and third, they serve as interfaces that open up a range of possibilities rather than stand in as fixed representations. By tracing how the very concept of the
authoritative image has been transformed, the chapter shows how visual knowing takes shape in
research practices and situates it in the digital and networked settings of contemporary science.