Witches, Man-Eaters, and Female Slaves

By James R. Burns

The DoSSE project team is working on a sourcebook for late antique slavery, featuring Latin, Greek, Arabic, Hebrew, Syriac, and Persian texts, and more. This post showcases some of the sources we’ll include: extracts from Gregory of Tours’ ‘Histories’ and the Lombard Edict of Rothari dealing with female slaves as witches. It concludes with a few remarks on ‘numinous women’ in the early Middle Ages and John Blair’s Killing the Dead.

In late sixth-century Merovingian Gaul, an enslaved woman was burnt at the stake:

When she was being led off, the miserable woman began to cry out that she had told lies. But her words were to no avail. She was sent to the stake, and incinerated alive by the flames.[1]

This is a scene we may normally associate with the ‘witch-crazes’ of early modern Europe. The slave, indeed, had been accused of witchcraft. The sixth-century bishop-historian Gregory of Tours gives the background:

a certain person came to the Queen [Fredegund] and said, ‘You sit disturbed by [the deaths of] your sons because of a trick of Clovis has been at work. For he passionately desires the daughter of one of your female slaves. He kills your sons through her mother with witchcraft (maleficium). Therefore, I warn you, that you might not hope for better for yourself, since the hope through which you should have reigned is snatched away.’ Then the Queen, greatly frightened to fear, with her fury set ablaze and compounded by her still fresh bereavement, had the girl, upon whom Clovis had cast his eyes, apprehended. She had her gravely beaten; she ordered the hair of her head to be cut, and once it was shorn, she ordered her to be fastened on a stake fixed in the ground before the dwelling of Clovis. And also the mother of the girl was removed and tormented by tortures for a long time; she forced a confession from her, so that she declared that these rumours to be true.[2]

When another of her children died on a later occasion, Fredegund apparently condemned as witches and tortured to death several women from Paris; the status of these women is unspecified, but in the absence of any hints otherwise, they were likely free.[3] Female slaves were not the only people accused of witchcraft. Nonetheless, in the anecdote quoted above, we can see how the intimacies and tensions of the household made female slaves especially vulnerable. Clovis desired the daughter of one of Fredegund’s slaves: female slaves could be sexually exploited by free male members of their owner’s family. When those men had hostilities with their owners, as Clovis did with his step-mother Fredegund, enslaved women might be caught in the middle. Moreover, as a slave in Fredegund’s household, the mother accused of maleficium and burnt to death could have been close contact with, or even helped to bring up, Fredegund’s young sons. Slaves were employed to cook food and serve drinks, so it may also be relevant here that the term maleficium can also refer to poison, which conceptually overlapped with witchcraft in this period. When free children died seemingly without explanation, slaves, as potentially untrustworthy and legally defenceless members of the household, were easy targets for suspicion and blame. We must remember, however, that we are reliant on Gregory of Tours, who did not have a favourable view of Queen Fredegund and was happy to imply that she was violent and vengeful, her actions possibly driven by paranoia and hysteria.

It would be all too easy to use Gregory’s story as proof that the early Middle Ages was riven by cruelty and superstition. But an edict issued by the Langobard king Rothari in 643 shows legislators taking steps to try to protect female slaves from the worst excesses of violence driven by fear of witches:

 Let no one dare to kill another’s female slave (ancilla) or half-slave (haldia) as if she were a witch, which they call mascam; because it should not be believed possible by Christian minds that a woman might be able to devour a man alive from within. If anyone dares to perpetrate such a wicked and unlawful thing, if he has killed a female half-slave, let him compensate for her status with 60 solidi, and, on top of that, let him add 100 solidi for his crime: half to go to the king, the other half to whose half-slave she had been. If however she was a slave, let him compensate for her status, as is dictated above [elsewhere in the Edict], depending on if she was a household or field slave; and, on top of that, for his crime, 60 solidi: half to the king and half to whose female slave she had been. If truly a judge ordered him to perpetrate this evil deed, let the same judge compensate for the above penalty from his very own property.[4]

Thom Gobbitt’s excellent blog makes a number of key points about this strikingly specific law:

• The conventional assumption that if something is banned it probably took place in practice probably applies here. The reference to judges (illicitly) ordering the deaths of suspected witches among female slaves and half-slaves is especially suggestive.
• Nonetheless, the edict explicitly emphasises that this is an unchristian and unlawful act, and makes the culprit liable to pay a fine to the king.
• There is no similar provision in the edict about the killing of free women accused of being witches. This probably reflects the need to issue extra protection for female slaves because of their vulnerability, and the tendency for slaves to be subject to particularly extreme violence, rather than an absence of interest in protecting free women.[5]

The term used for a witch in this law is striga, which several dictionaries as well as Katherine Fischer Drew’s translation render as ‘vampire’, though there is no hint here that the accused slaves here were undead. Quite the opposite. Yet with the vivid reference to the striga eating a man alive from within, the type of witchcraft does seem conceptually different from maleficium. Whereas men, including male slaves, were also sometimes accused of maleficium, the striga was stereotypically female. Daniel Ogden’s recent book on the explains that she ‘was a woman that flew by night, either in a form akin to that of an owl or in the form of a projected soul, in order to penetrate homes by surreptitious means and thereby to devour, maim, blight or steal the new-born babies within them’.[6] So the use of the term striga may indicate that, as in Gregory’s anecdote about Fredegund, the sudden deaths of infants provoked suspicions and violence against female slaves in the early Middle Ages. But the idea of a striga is not a straightforward reflection of class conflict: in Roman sources, free women are also depicted as witches and the babies of slaves are sometimes depicted as their victims.

In addition, perceptions of ‘supernatural’ power among female slaves were not always causes of antagonism with their owners. Gregory of Tours states that an enslaved woman earned freedom from her owners after she correctly predicted future events to their advantage.[7] In this instance, the accusation later made against her by suspicious elites was not witchcraft but demonic possession, though the two often come together. The incident recalls, probably deliberately, a similar story from the Acts of Apostles where a female slave likewise profits her masters through prophecy – except for the fact that, while in the biblical narrative St Paul successfully exorcises a demon from the slave, the Merovingian bishops, in Gregory’s telling, failed to do this.[8] Freed, the woman gained wider renown and associated herself with Fredegund.

It was with these sources on my mind that I started reading John Blair’s new book, Killing the Dead, in which he looks at the global history of people ‘killing’ corpses perceived to be harming others from beyond the grave. Blair argues that, in the early Middle Ages, girls and young women were disproportionately targeted for corpse-killing. He notes that strigae offer a precedent for an idea of female malicious, vampiric power, though strigae were often presented as being old rather than young women. The more immediate early medieval context is a contemporary belief that ‘numinous power resides in women’; those thought to exercise extraordinary power in life might be thought to exercise it in death too.[9] In Blair’s interpretation, people killed corpses by removing their heads, jawbones, and possibly their hearts too, among other methods. Might there be female slaves among some of the burials with these features? Gregory’s account of the future-predicting ancilla offers a precedent for a ‘numinous’ woman being attached to a household as a slave and attracting both support and suspicion.

Enslaved women might have recourse to ‘magic’ for their own advantage. But they were also vulnerable to being demonised as witches, which might lead to them being violently attacked in life – and potentially in death.

References

[1] Gregory of Tours, Ten Books of Histories, 5.39: ‘Quae cum duceretur, reclamare coepit misera, se mendacia protulisse; sed nihil proficientibus verbis, legatam ad stipitem, vivens exuritur flammis.’

[2] ‘Post dies vero aliquot adveniens quidam ait reginae: “Ut urbata de filiis sedeas, dolum id Chlodovechi est operatum. Nam ipsi concupiscens unius ancillarum tuarum filia, maleficiis tuos per matrem eius filios interficit, ideoque moneo, ne speres de te melius, cum tibi spes per quam regnare debueras sit ablata”. Tunc regina timore perterrita, furore succensa, nova orbitate conpuncta, adpraehensam puellam, in qua oculus iniecerat Chlodovechus, graviter verberatam, incidi comam capitis eius iussit ac scisso sode inpositam defigi ante metatum praecipit Chlodovechi. Matrem quoque puellae relegatam et turmentis diu cruciatam, elicuit ab ea professione, quae hos sermones veros esse firmaret.’

[3] Gregory of Tours, Ten Books of Histories, 6.35.

[4] Edict of Rothari, c. 376: ‘Nullus presumat haldiam alienam aut ancillam quasi strigam quem dicunt mascam, occidere; quod christianis mentibus nullatenus credendum est nec possibilem ut mulier hominem vivum intrinsecus possit comedere. Si quis de cetero talem inlecitam et nefandam rem penetrare presumpserit, si haldiam occiderit conponat pro statum eius solidos 60, et insuper addat pro culpa solidos centum, medietatem regi et medietatem cuius haldia fuerit. Si autem ancilla fuerit, conponat pro statum eius, ut supra constitutum est, si ministiriales aut rusticana fuerit; et insuper pro culpas solidos 60, medietatem regi et medietatem cuius ancilla fuerit. Si vero iudex huic opus malum penetrare iusserit, ipse de suo proprio pena suprascripta conponat.’

[5] Thom Gobbitt, ‘Vampires, witches and witchcraft in the Lombard Laws’, thomgobbit: early medieval laws and law-books, https://thomgobbitt.wordpress.com/2016/01/04/vampires-witches-and-witchcraft-in-the-lombard-laws/ (4 January 2016), accessed 28 November 2025.

[6] Daniel Ogden, The strix-witch (Cambridge: Cambridge University Press, 2021), p. 1.

[7] Gregory of Tours, Histories, 7.44.

[8] Robert Wiśniewksi, Christian divination in Late Antiquity (Amsterdam: Amsterdam University Press, 2020), pp. 179–182.

[9] John Blair, Killing the dead: vampire epidemics from Mesopotamia to the New World (Princeton: Princeton University Press, 2025), especially pp. 15, 125–126, 214–215, 231.