Noting Witchcraft

It’s impossible – I think – to work on modern European witchcraft without engaging with the challenges Jeanne Favret-Saada laid down in a series of publications from the 1970s onwards. Favret-Saada’s drew on her fieldwork in western France in a series of articles and three books.

I won’t do her ideas the disservice of trying to fully summarize them here.

Instead, this post is about form. Favret-Saada’s three books on French witchcraft took radically different forms. The most recent – Désorceler (2009) – is fairly traditional, in that it gathers essays many of which were published elsewhere. The previous book she published on the same subject – Corps pour corps (1981) – was completely different, a fieldwork diary, edited with her longtime collaborator, the psychoanalyst Josée Contreras. And her first, and best-known, witchcraft book is like neither of its sequels. An iconoclastic assault on the objectivity of ethnography and a call to arms for a generation of fieldworkers, I was introduced to Les Mots, la mort, les sorts (1977) by a mentor who described it as ‘very weird’.

Like all strangeness, the weirdness of Les Mots fades from attention if you reread it several times.

But I can remember its centre of gravity, its scandalous transgression: Favret-Saada declared that the only way to learn about witchcraft was to do it.

Or, to be more fair to her, to become an unwitcher and take on the responsibility for curing bewitched families of their ailments. Much of what makes the book so challenging in form stems from this core argument. The book is both autobiography and theory, a blend of personal revelations and pre-post-(?)-structuralist accounts of the discourse of witchcraft.

My own arguments about witchcraft in France from 1790-1940 draw heavily on Favret-Saada’s ideas. But I’ve also spent many years thinking about where we part ways.

There are many ways that I could describe this which come down to the basic divergence of sources and methods. Favret-Saada’s work was quintessentially contemporary, drawing on press reports at the time but grounded in her fieldwork. When she did cast her arguments backwards in time, it was to the early modern witch trials that she turned.

There is something of a missing piece in between, as if witchcraft disappeared completely from public view in the eighteenth and nineteenth centuries.

I can’t replicate her inside knowledge, her ability to ask her informants questions, to see a conflict over harmful magic evolve.

But rereading her book recently I was also struck by her how claims about the subjectivity of ethnography are in constant tension with a series of normative statements she makes about witchcraft: unwitchers don’t accuse close family members (142, n.17), they are always social equals (195, n.6), etc.

And what drew my attention to this was a formal characteristic of Les Mots: the footnotes.

Very few of these notes engage in detail with other research and ideas, or provide evidence to substantiate the claims in the main text. Many are digressions, small, spin-off arguments of exactly the kind to provoke groans from every academic editor under the sun.

But it is in the footnotes that Favret-Saada tends to bury the most categorical statements she makes about witchcraft, including the claims about family relations and the social status of witches, as well as a claim about the rarity of fraud prosecutions (66-7, n.6).


Formally, this is provocative. Her text becomes the story of a personal journey to understand sorcery, and these notes speak in a different voice: authority. Where the text explains her own evolving positions, the notes describe ‘this is how witchcraft always is’.

I am not convinced that the normative model of witchcraft the notes sketch out is the right way to think about harmful magic. Among the hundreds of criminal trials involving witchcraft from across France in the long nineteenth century, of course there were some where suspected witches were intimate family members. There were many more where the witch and his victim were not social equals. And if Favret-Saada’s claims about how many fraud cases made it to the newspapers hold true for the earlier period, there would have been at least 600,000 witchcraft frauds between 1790-1940, more than 599,000 of which left no trace.

It’s possible… Is it likely, though?

A partisan for Favret-Saada might point out two things: first, her arguments rarely claimed to be valid for that earlier period. This is fair, but I have to point out that what the earlier cases allow is for some testing of the empirical claims she made based on her fieldwork. I have my doubts that by the 1970s accusations against immediate family members had completely disappeared, given that they were still reported 40 years before.

But the second objection is tougher to resolve.

Favret-Saada’s view was that cases that reached public knowledge were aberrations. Any case involving physical violence, she argued, was one where the logic of witchcraft had broken down. Similarly, the reason that just 1/1000 fraud cases ever reached the news, she suggested, was because most unwitchers were very careful in how they went about their work. Those who were prosecuted were really just a decoy, the most extravagant and foolish examples who provided cover for the underground trade in magic (see 61-2). So Favret-Saada might argue that where she studied typical witchcraft disputes, the ones that I can find as a historian are only those that are aberrant. ‘Typical’ disputes are hidden in the archives.

This problem – sometimes known in criminology as the ‘dark number’ – is not easily dismissed. There is plenty of evidence from the historic trial records that unwitchers, the bewitched, and whole communities managed to keep some disputes secret, sometimes keeping them out of court and the newspapers for decades. Perhaps the most typical cases are underrepresented in the historical record.

Or perhaps there is a more fundamental, theoretical divergence at stake.

I think Favret-Saada tends to see witchcraft beliefs as a coherent ideology. Although she is at pains to point out that it is not one shared by the whole rural population, and is only comprehensible from within a dispute, the authority in her footnotes describes a coherent belief system with a logical structure: social life is warfare; kill or be killed; gain power or lose vitality. Appealing in its simplicity, I am not convinced this conceptualisation of ‘witchcraft’ reflects the ambivalences that Favret-Saada herself highlighted in talk of harmful magic.

In witchcraft, nothing is certain.

Aberrant cases were the norm. Or perhaps more accurately: witchcraft is the subversion of all norms. Not a mirror of social conventions, but the chaos of a world where all that is certain is in doubt.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s