Document Type


Publication Date



Professor Judith Jarvis Thomson's provocative article, 'The Decline of Cause,' focuses on the diminishing importance of causation in law and moral philosophy. In this reply, I suggest answers to some of the questions Professor Thomson raises.

Professor Thomson's article revolves around various forms of a classic dilemma: two persons take equal care but, through chance, their actions produce different results. Does the outcome of their actions matter in a moral assessment of those actions? Professor Thomson first sets out what the styles as the Kantian and 'moral sophisticates" position that the outcome of an act does not and should not effect our assessment of the moral quality of the act. Subsequently, however, Thomson rejects the absolute nature of this position. She concludes that, in cases of acts blameworthy in the first instance--i.e., blameworthy on some ground independent of the actual outcome--resulting harm or near harm does and should render our moral assessment more negative than it would be absent such result. Professor Thomson confesses herself to be largely at a loss to explain why we hold these moral intuitions, and thus she does little more than posit--rather than explain--them.

In the first part of my reply to Professor Thomson, I explain why the reasoning that leads her to reject the Kantian position is flawed--why, in fact, her own arguments lead to the conclusion that the Kantian position is correct and that outcome does not matter in our moral assessment of a person's acts. In the second part of my reply, I accept for the sake of argument Professor Thomson's position on Kant and the 'moral sophisticates,' and I address her ultimate intuitive conclusion that outcome does matter, but only in cases of fault in the first instance. I submit that I can offer Professor Thomson reasons for her intuited conclusion, reasons which explain not only why outcome matters in cases of fault in the first instance, but why it matters in all cases--including cases of no fault in the first instance.

Publication Citation

76 Geo. L.J. 151-167 (1987)