Disgust as a moral guide

Why does something feeling repugnant so readily pass for that thing being wrong?

When something revolts us, the revulsion does not arrive labelled as a feeling. It arrives as a report about the world: that is foul. Not "I am reacting", but "it is like that". And in moral argument the slide is quick and almost invisible — from foul to forbidden, from repugnant to wrong.

Some serious people have defended the slide. Leon Kass argued in 1997 that repugnance at practices like cloning is "the emotional expression of deep wisdom, beyond reason's power fully to articulate". So the question is not whether disgust feels authoritative — it plainly does — but whether the thing it is reporting is the thing we take it to be reporting.

Start with what disgust is for, since that is the only way to know what it is competent at. Paul Rozin's long research programme locates its core in food rejection: the paradigm elicitors are decay, faeces, vomit, bodily fluids, corpses, and animals associated with filth. The characteristic expression closes the mouth and nose and gives the gape of imminent rejection. The plausible reading, and it is a reading rather than an established evolutionary fact, is that this is a system for keeping pathogens and parasites out of the body.

Now notice its two strange rules. Contagion: once a cockroach has touched the juice, the juice stays ruined, even after the roach is removed and even when you know sterilisation was complete. And similarity: a perfectly clean object shaped like faeces is hard to bring to the mouth. Rozin and Nemeroff called these the laws of sympathetic magic, and they matter here because they are not about actual risk. They are heuristics that were cheap and roughly right for microbes, and they persist unchanged when the facts have been checked and found reassuring.

So we have a signal that presents its verdict as a property of the object, that cannot be argued down by evidence, and whose triggering conditions are physical resemblances rather than actual danger. Ask what happens when a system with those properties is pointed at moral questions.

Two errors follow directly, and they run in opposite directions.

The false positives are the ones we can list. Disgust fires vigorously at acts that harm nobody — Jonathan Haidt's much-used vignettes about consensual sibling incest with contraception, or a family eating their dog after it dies accidentally. People condemn them, are shown that every harm-based reason they offer does not apply, and hold the condemnation anyway. Haidt named that "moral dumbfounding", and its significance is structural: it shows the verdict was not produced by the reasons, because removing the reasons leaves it standing. Worse, because the system keys on unfamiliar bodies, deformity, and deviations from a bodily norm, its history of misfires is not random. Martha Nussbaum's argument in Hiding from Humanity (2004) is that disgust has been reliably recruited to mark out castes, minorities, disabled people and outgroups as polluting — which is not an incidental abuse of the emotion but an expression of what it is built to do.

The false negatives are quieter and worse. Ask what disgust has to say about tax fraud, about a supply chain that ruins a distant river, about a decision that shortens thousands of lives by a statistical margin. Nothing at all. These leave no bodily trace, so the alarm never sounds — and a moral instrument that is silent about the largest harms of an industrial society is failing on precisely the cases that need it most.

Here is where accuracy matters more than a good line. There is a well-known laboratory claim that incidental disgust — a foul smell, a dirty desk — makes people judge unrelated acts more harshly. It is often cited as proof that the emotion causes the judgement. The honest state of the evidence is that this effect is weak and contested: a prominent 2008 demonstration failed to replicate in a well-powered 2014 attempt, and a 2015 meta-analysis by Landy and Goodwin found the amplification effect small and consistent with publication bias. The stronger evidence for disgust's moral role is not the priming work; it is the dumbfounding pattern and the historical record of what disgust has been used to mark.

Three refinements. First, "disgust is a bad moral guide" does not mean it is uninformative. It tells you reliably about the presence of a contamination cue, and that is real information — it is simply information about pathogens and boundaries, not about permissibility. Second, the argument is not that emotion has no place in ethics. Indignation and sympathy track harm and unfairness in a way disgust does not; Nussbaum's proposal is a swap, not an amputation.

Third, and most important, the misconception to head off: this reasoning does not show that anything disgusting is permitted. Cruelty is often disgusting and wrong, and the disgust is not what makes it wrong. The correct conclusion is narrower and more usable — when the only thing supporting a moral verdict is that the act feels foul, and no account of who is wronged survives scrutiny, the verdict has no independent support left.

• Whether indignation and sympathy are better-calibrated moral emotions, or merely more flattering ones.
• Kass's case for repugnance at its strongest, applied to biotechnology, and what a reply would need to concede.
• How reliably disgust-based norms have tracked group boundaries across different societies.

Every term the collection defines is gathered in the glossary.