The claim that scientific explanation is deductive has been attacked on both systematic and historical grounds. This paper briefly defends the claim against the systematic attack. Essential to this defence is a distinction between perfect and imperfect explanation. This distinction is then used to illuminate the differences and similarities between Aristotelian (anthropomorphic) explanations of certain facts and those of classical mechanics. In particular, it is argued that when one attempts to fit classical mechanics into the Aristotelian framework the latter becomes (...) structurally incoherent. It is suggested that this, together with the fact that classical mechanics embodied the first piece of perfect knowledge, accounts, in part at least, for the historical fact of the rapid demise of the Aristotelian patterns as the new science developed. On the basis of this discussion, the inadequacies of the attack on the deductive model on historical grounds by Toulmin come to be seen. (shrink)