ABSTRACT
Aspects of realism are present in Einstein's earliest contributions to physical theory on "the molecular-kinetic theory of heat" and the 1905 account of Brownian motion. Scientific realism is broadly the view that the best theories in the "mature" sciences are true, or at least "approximately true", descriptions of both observable and unobservable aspects of a mind-independent nature. The semantic apparatus of reference and truth characteristic of contemporary scientific realism presupposes both empirical success as well as relative stability of theory. Realism as a "program" pertains to the intended target of description of any physical theory, the real external world as partitioned by physics into systems and subsystems. Einstein's realism lacks the semantic apparatus of truth and reference of scientific realism. Many contemporary philosophers of science view the confirmation of Einstein's molecular hypotheses explaining the mechanism of Brownian motion as vindicating realism over positivism, and there is a sense in which this is correct.
