In Life Itself and in Essays on Life Itself, Robert Rosen (1991, 2000) argued that machines were, in principle, incapable of modeling the defining feature of living systems, which he claimed to be the existence of closed causal loops. Rosen's argument has been used to support critiques of computational models in ecological psychology. This article shows that Rosen's attack on mechanism is fundamentally misconceived. It is, in fact, of the essence of a mechanical system that it contains closed causal loops. Moreover, Rosen's epistemology is based on a strong form of indirect realism and his arguments, if correct, would call into question some of the fundamental principles of ecological psychology.