In this article, I argue that Williams's sceptical view about the value of economic models expressed in 'The philosophy of economic modelling: a critical survey' [South African Journal of Philosophy, 18(2): 223–246, this issue], and widely shared amongst philosophers of science, is not warrented. Williams's error, I maintain, lies in his failure to adequately distinguish, (a) between theories in general and what he calls 'folk theories', and (b) between the different roles that models play in different sciences. With respect to (...) (b), I suggest that Williams fails to recognize that scientists who lack the ability of physicists to constrain theorizing through rigorous controlled experiments must generally do a good deal of work investigating theories by means of models solely designed to test their formal implications, before further models with potentially direct applicability to the world can even be developed. Williams's central mistake, I argue, lies in confusing models of the first sort with models of the second. (shrink)