In this paper I analyze manipulation by distinguishing the descriptive question, i.e., what criteria we can elaborate for deciding when a form of influence can be described as manipulation from the evaluative question, i.e., why and when we can hold manipulation to be morally unacceptable. To address the first question, I review the main theories of manipulation and argue that norm based views effectively capture the commonality of different types of manipulation. With respect to the evaluative question, I argue that a relational view succeeds in finding why manipulation is prima facie morally unacceptable, while other normative considerations based on consequences, rights, and character traits can help identify specific situations in which manipulation can be tolerated.
