Asymmetries in Judgments of Responsibility and Intentional Action

Recent experimental research on the ‘Knobe effect’ suggests, somewhat surprisingly, that there is a bi-directional relation between attributions of intentional action and evaluative considerations. We defend a novel account of this phenomenon that exploits two factors: (i) an intuitive asymmetry in judgments of responsibility (e.g., praise/blame) and (ii) the fact that intentionality commonly connects the evaluative status of actions to the responsibility of actors. We present the results of several new studies that provide empirical evidence in support of this account while disconfirming various currently prominent alternative accounts. We end by discussing some implications of this account for folk psychology. Awareness that an action is intentional plays an important role in evaluations of an actor and her action. This is only natural: if x intentionally acts to bring about a bad outcome, we may form different judgments about x or x’s behavior than if that same outcome is simply an accident or the result of (non-willful) ignorance. In this way, whether or not a given action is intentional matters to us when we assess an action’s or actor’s evaluative status. This relation between judgments of intentionality1 and judgments about the goodness/badness of actions or the responsibility (e.g., praiseworthiness/ blameworthiness) of actors seems to be a relatively straightforward part of folk * Thanks to two anonymous reviewers and [acknowledgments omitted for blind review]. 1 Throughout, we use ‘intentionality’ to denote a particular property of actions, namely, the property of being done intentionally. It is somewhat of an open question whether intentionality requires the specific mental state of intending, as defenders of the so-called ‘simple view’ claim (Adams 1986; McCann 1986, 2005; for recent critical discussion, see Nadelhoffer 2006b). Since our primary focus is the explanation of attributions of intentionality, not intention, we shall set this issue to the side here.