Convicting Witches

To the Editors:

In his review of Richard Godbeer’s The Devil’s Dominion: Magic and Religion in Early New England [NYR, May 28], Edmund S. Morgan writes:

“Witch trials in both England and New England (before Salem) differed markedly from those on the European continent in that the usual outcome was acquittal. In Europe, where witch hunts were much more numerous and protracted, more than 80 percent of those accused, numbering in the thousands, were convicted and executed.”

This is incorrect. Recent research indicates that the rate of executions, on average, across Europe was closer to 50 percent. In France, as Robert Muchembled’s review of 10,000 witch trials indicates, the rate was almost exactly 50 percent. In the Isle of Guernsey it was 46 percent; in Fribourg, 33 percent; in Geneva, 21 percent; in Finland, approximately 16 percent; in Denmark and Hungary, approximately 50 percent; in Scotland, just over 50 percent (54%); in Estonia, 32 percent; in Norway, approximately 25 percent; and, most remarkably, in Spain, out of 4,000 witch trials heard by the Inquisition between 1550 and 1750, there were 11 death sentences. In Italy, the Venetian Inquisition heard 500 witch cases between 1550 and 1650 and condemned no one to death; and in Portugal, between 1536 and 1821, there was only one conviction for witchcraft—in Evora in 1626. (I will be happy to supply fuller details and sources to anyone who desires them.)

Based on these statistics we need to reconsider the nature of the witch craze in Europe and any judgments we might wish to make about the comparison of the Salem and European witch craze.

Steven T. Katz
Department of Near Eastern Studies
Cornell University
Ithaca, New York

Edmund S Morgan replies:

Professor Katz is right. My estimate for the rate of conviction in European witch trials was too high, and I am grateful for this correction. Perhaps it is not meaningful to generalize about “European” convictions at all, not only because the surviving records are so incomplete but also because the estimates based on them vary so widely from place to place. For example, the same scholars who put the rate for Geneva at 21 percent, put it at 90 percent for another part of Switzerland, the pays de Vaud, where there were more than two thousand trials. (William Monter, Ritual, Myth, and Magic in Early Modern Europe (Ohio University Press, 1984), p. 47; Brian P. Levack, The Witch-Hunt in Early Modern Europe (London and New York: Longman, 1987), pp. 19–20.