This paper critically examines one of the most influential arguments against compatibilism-"the Manipulation Argument" (henceforth MA), which has been vigorously defended by R. Kane (1996 ), D. Pereboom (2001 ), and A. Mele (2006 . MA claims that agents in a deterministic world are, with respect to moral responsibility, relevantly similar to agents whose actions, decisions, or processes of acquiring their character were covertly manipulated by other agents. It will be argued that MA fails to

more »

... ute compatibilism. The argument, if it succeeds, is important because it can apply to any MA; that is, the argument does not depend on the specific description of manipulation cases or on specific ways of supporting each premise of MA. Structure of the Manipulation Argument My aim in this section is (i) to present the structure of the Manipulation Argument (MA) as an argument against compatibilism, (ii) to explain the relation between MA and source incompatibilism by extending MA to an argument for source incompatibilism, and (iii) to state my basic assumption about the methodology of MA. MA is a template for a kind of argument that aims to show that compatibilism is false by means of "manipulation" cases. The basic structure underlying MA can be stated as follows. 2 Responses to MA have taken two main forms: hard-line reply and soft-line reply (this terminology is in McKenna (2008) ). Hard-line compatibilists (