Abstract
The relationship between Adam Smith's official methodology and his own actual theoretical practice as a social scientist may be grasped only against the background of the Humean project of Moral Newtonianism. The main features in Smith's methodology are (i) the provisional character of explanatory principles, (ii) 'internal' criteria of truth, (iii) the acknowledgement of an imaginative aspect in principles, with the related problem of the relationship between internal truth and external truth understood as mirroring 'real' causes. Smith's Newtonian (as opposed to Cartesian) methodology makes room for progress in social theorizing as it allows for a decentralization of the various fields of Moral Science, contributing to the shaping of political economy. On the other hand, the Cartesian legacy in Smith's Newtonian methodology makes the relationship problematic between phenomena and principles.