Three Cheers for Double Effect


  • This paper resulted from discussions that grew out of a graduate seminar on non-consequentialist ethics that we co-taught at UCSD in 2010. We would like to thank the following seminar attendees for their helpful comments and probing questions: Craig Agule, Amy Berg, Nanhee Byrnes, Alex Marcellesi, Veronica Pear, and our colleague, Dick Arneson. We presented a shorter version of an ancestor of this paper at a stimulating workshop on doing harm at Université Paris 1 Panthéon-Sorbonne in July 2011. We are very grateful to the members of the Equipe NoSoPhi who organized and participated in the workshop: Laurent Jaffro, Catherine Larrère, and Vanessa Nurock. We benefited greatly from the constructive comments and helpful suggestions we received from David Brink, William Fitzpatrick, Matthew Hanser, Derk Pereboom, and an anonymous referee for this journal. Thanks to you all. Finally, we would like to acknowledge our greatest debt, the one we owe our teacher, mentor, and friend, Warren Quinn, whose work and intellect we so admire, and who will never know just how grateful we are for his support and encouragement at a critical time in our lives.


The doctrine of double effect, together with other moral principles that appeal to the intentions of moral agents, has come under attack from many directions in recent years, as have a variety of rationales that have been given in favor of it. In this paper, our aim is to develop, defend, and provide a new theoretical rationale for a secular version of the doctrine. Following Quinn (1989), we distinguish between Harmful Direct Agency and Harmful Indirect Agency. We propose the following version of the doctrine: that in cases in which harm must come to some in order to achieve a good (and is the least costly of possible harms necessary), the agent foresees the harm, and all other things are equal, a stronger case is needed to justify Harmful Direct Agency than to justify Harmful Indirect Agency. We distinguish between two Kantian rationales that might be given for the doctrine, a “dependent right” rationale, defended by Quinn, and an “independent right” rationale, which we defend. We argue that the doctrine and the “independent right” rationale for it are not vulnerable to counterexamples or counterproposals, and conclude by drawing implications for the larger debate over whether agents' intentions are in any way relevant to permissibility and obligation.