Abstract

In the debate over the function of images in science, a cornerstone has been set by the work of the historical epistemologists Lorraine Daston and Peter Galison. In this article I criticize their idea of mechanical objectivity and their account of mechanical images as passive representations of reality. By examining the interplay between the indexical, iconic and symbolic modalities in the processes of signification of microscope and X-ray images, I challenge the idea implicit in Daston and Galison's argument that mechanical objectivity relies on self-operating instruments and self-evident images, and call into question the common sense notion of mechanical images' photographic realism.