Abstract
The distinction between the context ofdiscovery and the context of justificationrestricts philosophy of science to the rationalreconstruction of theories, and characterizesscientific discovery as rare, theoreticalupheavals that defy rational reconstruction. Kuhnian challenges to the two contextsdistinction show that non-rational elementspersist in the justification of theories, butgo no further to provide a positive account ofdiscovery. A gradualist theory of discoverydeveloped in this paper shows, with supportfrom ecological cases, that discoveries areroutinely made in ecology by extending modelsto new domains, or by making additions toearlier models. The logic of discovery isphilosophically accessible once it isappreciated that model truth is presumed, evenif counterfactually, in ecologists' applicationof models. A gradualist view shows thatmodels' heuristic power routinely leads todiscoveries.