Across the ethnographic record, witchcraft accusations recur with remarkable structural consistency. From the Azande of Central Africa to Salem Village, from Navajo communities to the highland Bocage of Normandy, societies separated by continents and centuries converge on strikingly similar patterns of who accuses whom, and under what conditions. This convergence demands theoretical attention.

The anthropological reframing pioneered by Evans-Pritchard, Marwick, and Douglas dislodged witchcraft from the domain of exotic irrationality and relocated it within the analysis of social structure. Accusations, on this view, are not epistemic errors but social diagnostics—they map with uncanny precision onto the fault lines of a given community. Where the seams of a social system strain, accusations cluster.

This essay develops three interlocking claims. First, that accusation patterns systematically illuminate structural tensions inherent to particular kinship and residence arrangements. Second, that witchcraft belief operates as a multifunctional institution—simultaneously enforcing normative behavior, providing causal explanation for misfortune, and channeling conflict toward resolution or expulsion. Third, that historical epidemics of witch-hunting arise under specifiable material and structural conditions, exhibiting recognizable cross-cultural regularities from sixteenth-century Europe to twentieth-century Africa.

The pioneering work of Max Marwick among the Cewa of Zambia established a methodological principle that has held up across decades of comparative research: witchcraft accusations follow the contours of structural ambiguity. Where role obligations are clearly defined and dispute-resolution mechanisms exist, accusations are rare. Where relationships are ambiguous, competitive, or lack institutional adjudication, accusations proliferate.

Consider the matrilineal Cewa: accusations concentrate among matrilineally related males contesting succession to headship, precisely the relationship where jural authority and personal ambition collide without clear resolution. Among the patrilineal Nyakyusa, by contrast, tensions cluster among co-wives and age-mates. The pattern is not random—it is a structural signature.

Alan Macfarlane's analysis of Essex trial records identified an analogous logic in early modern England: accusations typically flowed from wealthier households against poorer female neighbors who had recently been refused charity. The tension point was the transition from communal obligation to emergent market individualism—an ambiguous moral zone where old norms of neighborly reciprocity clashed with new economic realities.

The comparative principle is generalizable: accusations target ambiguous role partners—those with whom one has ongoing obligation but no clean exit, and with whom formal grievance procedures are unavailable or socially costly. In-laws, co-wives, matrilateral kin in patrilineal systems, dependent poor in stratifying economies. The witch is rarely a stranger.

This yields a diagnostic tool: given a society's kinship structure, residence patterns, and inheritance rules, one can predict with reasonable accuracy where its witchcraft accusations will cluster. The ethnographic archive functions here as a controlled experiment in social structure.

Takeaway

Accusations are cartographic—they trace the exact locations where a social system generates obligation without adequate mechanisms for resolving it. Where you cannot easily fight or flee, you suspect.

Functionalist analysis, though rightly critiqued for its teleological excesses, retains explanatory purchase in the domain of witchcraft studies. Beliefs about witches perform identifiable social work, and this work clusters around three interrelated functions: normative enforcement, causal explanation of misfortune, and the enabling of expulsion or realignment.

The enforcement function operates through what Monica Wilson termed the ethical dimension of witchcraft belief. Among the Nyakyusa, the paradigmatic witch is one who eats alone, hoards, refuses to share beer, or displays inappropriate ambition. To be suspected requires only that one violate the community's densely elaborated norms of generosity. Belief in witchcraft thus constitutes a decentralized surveillance system that produces conformity without requiring formal sanction.

The explanatory function addresses what Evans-Pritchard famously called the why me, why now question. Empirical causation may explain that a granary collapsed because termites weakened its supports, but only witchcraft explains why this granary collapsed on this person at this moment. Witchcraft belief supplies the personalistic causation that impersonal mechanisms leave unaccounted for, integrating misfortune into a moral cosmos.

The expulsion function is perhaps the most consequential. Accusation provides culturally legitimate grounds for severing relationships that structural rules otherwise render permanent. In lineage-based societies with no institution of divorce or disinheritance, accusation is the escape valve. The witch is expelled, killed, or induced to flee—and a stuck relationship becomes unstuck.

These functions are not always compatible. Enforcement stabilizes; expulsion fragments. The tension between these outcomes helps explain why witchcraft institutions oscillate between low-level chronic accusation and explosive collective persecution.

Takeaway

Witchcraft belief is not a failed science but a functional institution—it enforces cooperation, explains suffering, and dissolves relationships that structure alone cannot terminate.

Chronic, low-level accusation is one thing; epidemic witch-hunting is another. Comparative analysis of the early modern European hunts, the twentieth-century African witch-finding movements, and periodic outbreaks in New Guinea and Amazonia reveals a recurring conjunction of conditions that transforms endemic suspicion into collective panic.

First, subsistence stress. The European hunts intensified during the Little Ice Age; the Tanzanian Sukuma witch-killings of the 1980s coincided with structural adjustment and collapsing agricultural returns; the Bangwa hunts in Cameroon accompanied cash-crop volatility. Material precarity multiplies misfortunes requiring explanation and depletes the surplus that sustained normative generosity.

Second, institutional restructuring. Epidemics erupt when traditional authorities lose legitimacy but their replacements have not consolidated. Early modern Europe was reformulating the relationship between ecclesiastical, secular, and communal jurisdictions. Postcolonial African societies faced parallel disruptions as chieftaincy, colonial administration, and postcolonial state authority produced overlapping and contested claims.

Third, ideological entrepreneurship. Waves require specialists—witch-finders, inquisitors, prophetic movements—who supply the diagnostic technology and moral framing. The Malleus Maleficarum and the Bamucapi movement of 1930s Central Africa perform analogous functions: they codify detection procedures, standardize the witch's profile, and legitimate escalation.

The pattern suggests that witch-hunting epidemics are not atavistic eruptions of ancient irrationality but predictable responses to specific structural conjunctures. They emerge when subsistence pressure, authority vacuum, and mobilizing ideology align—a formula that renders them, disturbingly, a recurring possibility rather than a vanquished past.

Takeaway

Witch-hunts are structurally recurrent, not historically peculiar. When material stress meets authority breakdown and finds a diagnostic ideology, any society can enter the pattern.

The comparative study of witchcraft accusation yields a general theory of how societies process the tensions their own structures generate. Accusations are not eruptions of irrationality but structured social processes with predictable distributions, definite functions, and specifiable conditions of intensification.

This framework resists both the romanticization of exotic belief and its condescending dismissal. The Azande witch and the Salem witch inhabit the same analytical space as the contemporary conspiracy target and the scapegoated minority—all are products of structural strain seeking a personalized locus.

The comparative anthropologist's task is not merely to catalog these patterns but to specify the conditions under which they arise, persist, and escalate. In doing so, we develop tools for recognizing the accusatory dynamic wherever it appears—including in societies that no longer name their witches as such.