Knobe versus Machery: Testing the trade-off hypothesis

Recent work by Joshua Knobe has established that people are more likely to describe bad but foreseen side-effects as intentionally performed than good but foreseen side-effects (this is sometimes called the 'Knobe effect' or the 'side-effect effect'. Edouard Machery has proposed a novel explanation for this asymmetry: it results from construing the bad side-effect as a cost that must be incurred to receive a benefit. In this paper, I argue that Machery's 'trade-off hypothesis' is wrong. I do this by reproducing the asymmetry between judgments about good and bad side-effects in cases that cannot plausibly be construed as trade-offs.