The Torsåker Witch Trials

Torsåker is a parish in Ångermanland, in northern Sweden, and in 1675 it was the site of the largest single execution for witchcraft in Swedish history. The parish record names a large group, the great majority of them women, beheaded and then burned on one day.

It belongs to a national panic, called in Swedish the Great Noise, which ran through the 1660s and 1670s and had a distinctive engine. The accusations turned on children, who were asked whether they had been carried at night to Blåkulla, a place in Swedish folk belief where witches were said to gather, and who were asked to name who had carried them. Children questioned that way by adults who plainly want an answer will produce one, and each name produced another round of questioning.

At Torsåker the identification is reported to have been done by boys who walked the assembled parish and pointed. Whatever the exact procedure, the result was a mass condemnation from a small community, and the parish never recovered its population balance.

The Swedish panic ended when the central authorities in Stockholm examined the children's testimony properly, found it contradictory, and stopped the commissions. Some of the child witnesses were themselves punished afterwards for false testimony.

The Swedish material is the clearest European demonstration of a general point: witch panics that ran on child testimony were the fastest to grow and the fastest to collapse once anyone examined the testimony carefully.

They grew fast because a child asked a leading question by a frightened adult in authority supplies names, and because a community will believe a child before it believes an accused neighbour. They collapsed because the same testimony, once compared statement against statement, contradicts itself immediately. Blåkulla accounts placed the same people in different places on the same night, and named the dead and the absent.

That is what happened. Investigations ordered from Stockholm in 1676 examined the children's evidence rather than counting it, found it worthless, and the commissions were wound up. The pattern repeats elsewhere with the same shape: the Basque panic of 1609 to 1614 also ran on child testimony and was also ended by an official who tested it rather than tallied it, and Salem's afflicted accusers were mostly young.

The uncomfortable corollary is that nothing about the belief changed in any of these cases. Sweden did not stop executing because it decided witches were impossible. It stopped because a central authority looked closely at the evidence and found it did not stand up, which is a procedural conclusion rather than a philosophical one, and it saved lives immediately.