This paper supports the contention that the general rules introduced by Hume in the Treatise on Human Nature (THN 1.3.15) are a selection mechanism for inductive inferences, which rejects two sources of inefficiency : (i) from emotional origin, which would reduce the uneasiness coming from a possible failure in the uniformity of nature ; (ii) from cognitive origin, which would tolerate the possible overflow of the imagination on judgment. A growing consensus in recent decades, which distinguishes between two kinds of (...) rules — extensive and corrective — is at the basis of this device. Whereas the extensive rules allow us to go beyond a singular experience and derive a wider range of inferences, the corrective rules, whose command opposes the philosopher to the vulgar, control and rectify the effects of extensive rules alone, so as to eliminate emotional and cognitive inefficiencies, and to make inferences that, borrowing the expression to Peirce, we will designate as abductive. (shrink)