Ergot and Salem: a tempting but weak key

The ergot hypothesis about the Salem witch trials has all the features of a story that is easy to remember. In a wet year the rye is infected by Claviceps purpurea, the inhabitants eat contaminated flour, the alkaloids cause convulsions and disturbances of perception, and the community interprets the symptoms as witchcraft. Biology supplies a material cause where the sources speak of invisible agency. The problem is that an elegant chain does not become true merely because each of its links is separately possible.

Ergot replaces the grains of grasses with dark, elongated resting bodies called sclerotia. It contains alkaloids acting powerfully on the nervous system and blood vessels. Historical grain poisonings could take a convulsive form, with spasms and disturbances of sensation, or a gangrenous form with ischaemia, necrosis of the limbs and burning pain. The composition of the alkaloids, the dose, the raw material and the duration of exposure all change the picture. Ergotism is a real poisoning, not an invention created for Salem.

In 1976 a proposal linking the epidemic of accusations with the consumption of contaminated rye was published in Science. The authors of later analyses examined the weather, the crops, the distribution of farms, the ages of those reporting complaints and the chronology of events. The criticism did not consist of claiming that ergot cannot produce strange symptoms. It asked whether the predictions of the hypothesis fit the full set of data better than competing explanations.

If a shared batch of grain were the main cause, one would expect a pattern of consumption: illness in households drawing on the same supply, symptoms in many family members, and a decline once the source was removed. The Salem sources do not form so clear a map. The "afflicted" girls were not the only people eating local produce, and accusations and testimony developed through a social process over months. The behaviour could appear or intensify in particular public situations, which is hard to reduce to a constant dose of a toxin.

A further problem is the selection of symptoms. When a researcher starts from ergot, they may pick out trembling, pain, a sensation of pricking or unusual visions, and pass over the structure of the testimony, the pressure of interrogation and reactions to particular people. In the documents a word describing a "fit" need not correspond to a modern neurological diagnosis. Witnesses, officials and later chroniclers used the language available in their culture. Retrospective medicine has to retain uncertainty, because there are no biological samples and no independent examination of a patient.

The trials neither began nor ended with the complaints of a few people. They involved a network of local conflicts, religious beliefs, disputes over authority, family reputations, legal procedures and the authorities' acceptance of a particular kind of evidence. An accusation could change the social position of many people, and the testimonies were known to one another. Even if some participants did experience illness, that would not explain why the community selected particular suspects, why the court admitted particular evidence, and why the executions continued.

The toxicological hypothesis is attractive partly because it relieves the history of the uncomfortable fact of agency. Instead of asking how ordinary people sustain a persecution, one can put the blame on a fungus. Biological material does not, however, accuse a neighbour, draw up a warrant or set the rules of a court. Illness may affect behaviour, but institutions interpret behaviour and make decisions. A biological explanation should not remove the social level.

On the other hand, excluding the environment entirely would also be a mistake. Food, infections, malnutrition, sleep and stress affect historical people just as they affect us. A sound analysis asks for evidence along each route. Was susceptible grain grown in the region? Did the weather favour infection? Were sclerotia, unusual flour, similar illness in animals or ischaemic symptoms recorded? Does the timing of exposure match the chronology? The absence of one record does not close the case, but an accumulation of gaps weakens the claim.

Ergot also changes during processing and storage. It is not enough to say the summer was wet. One has to reconstruct the variety of grain, the timing of flowering, the harvest, the sorting, the milling, the mixing of supplies and the dose in the diet. Climate may raise the probability of infection, but it does not prove that a particular group consumed a toxic quantity. In historical analysis every stage should have its own source and its own range of uncertainty.

The popular comparison of ergot with LSD also needs correction. Some alkaloids share a chemical history with the synthesis of lysergic acid diethylamide, but eating sclerotia is not equivalent to a controlled dose of a pure substance. The mixture can cause severe vascular and neurological poisoning, not the colourful "trip" familiar from film. Such a simplification trivialises both the toxicology and the victims of historical ergotism epidemics.

How should this history be presented in a museum or an article? First show the organism and the well-documented effects of poisoning. Then give the date and the arguments of the original hypothesis. Next make room for the criticism: the demographic, chronological and social inconsistencies. Finally explain that historical scholarship judges explanations comparatively, and that the absence of one universal key does not mean giving up on understanding events.

Illustrations of the trials and photographs of ergot are not evidence of the same kind. An engraving was made later and shapes collective memory; a specimen of C. purpurea shows what the pathogen looks like, but it does not come from the flour of the people of Salem. Placing them together is valuable if the caption clearly separates the functions of the images. Without that, a viewer may read proximity as confirmation of a causal link.

Today ergot remains a hazard controlled through choice of material, agriculture, grain cleaning and standards. Do not use sclerotia yourself or prepare "historical experiments". The alkaloids have pharmaceutical uses in purified, dosed preparations, but the natural mixture is not a medicine. A suspicion of contaminated food requires a batch to be withdrawn and the relevant services contacted, not a tasting.

The most useful lesson of Salem is neither "a fungus caused the witch hunt" nor "biology had nothing to do with it". It is this: a resemblance between symptoms creates a hypothesis, not a verdict. A good hypothesis has to predict who will be affected, when and how, and to explain the data better than the alternatives. Ergot opens a fascinating question about environment and history, but the present evidence does not allow it to be made the sole engine of the events of 1692. A responsible account keeps that uncertainty instead of selling it as the solution to a mystery.

A hypothesis, a counter-test and the limits of explanation

In 1976 Linnda Caporael proposed in Science that ergot poisoning could have produced symptoms interpreted as witchcraft at Salem. The idea was testable, but later analyses challenged its central premises. Nicholas Spanos’s 1983 critique disputed the proposed fit with weather, accusers’ ages and symptom patterns, offering instead a framework of sociopolitical conflict and role enactment. Alan Woolf’s 2000 toxicology review concluded that ergotism was unlikely to account for much of the episode. None of this means Claviceps purpurea cannot contaminate rye or that ergotism is imaginary. It means that a biologically possible mechanism is insufficient for a retrospective historical diagnosis. The entry should present the sequence—proposal, attempted falsification and present evidentiary limits—rather than turning an appealing fungal clue into a single-cause solution for collective violence. Possibility, plausibility and demonstrated historical causation are three different claims.