Content uploaded by Manvir Singh
Author content
All content in this area was uploaded by Manvir Singh on Mar 13, 2020
Content may be subject to copyright.
This is an accepted manuscript. Please cite as: Singh, M. (forthcoming) Magic, explanations,
and evil: On the origins and design of witches and sorcerers. Current Anthropology.
Magic, explanations, and evil
On the origins and design of witches and sorcerers
Manvir Singh
Department of Human Evolutionary Biology, Harvard University
13 March 2020
Abstract
In nearly every documented society, people believe that some misfortunes are caused by malicious
group mates using magic or supernatural powers. Here I report cross-cultural patterns in these
beliefs and propose a theory to explain them. Using the newly-created Mystical Harm Survey, I
show that several conceptions of malicious mystical practitioners recur around the world,
including sorcerers (who use learned spells), possessors of the evil eye (who transmit injury
through their stares and words), and witches (who possess superpowers, pose existential threats,
and engage in morally abhorrent acts). I argue that these beliefs develop from three cultural
selective processes: a selection for intuitive magic, a selection for plausible explanations of
impactful misfortune, and a selection for demonizing myths that justify mistreatment.
Separately, these selective schemes produce traditions as diverse as shamanism, conspiracy
theories, and campaigns against heretics – but around the world, they jointly give rise to the
odious and feared witch. I use the tripartite theory to explain the forms of beliefs in mystical
harm and outline ten predictions for how shifting conditions should affect those conceptions.
Societally-corrosive beliefs can persist when they are intuitively appealing or they serve some
believers’ agendas.
ON THE ORIGINS AND DESIGN OF WITCHES AND SORCERERS
Page 2
“I fear them more than anything else,” said Don Talayesva
1
about witches. By then, the Hopi
man suspected his grandmother, grandfather, and in-laws of using dark magic against him.
1. Introduction
Beliefs in witches and sorcerers are disturbing and calamitous. Sterility, illness, death, rainstorms,
burnt-down houses, bald spots, attacks from wild animals, lost foot races, lost reindeer races, the
puzzling behavior of a friend or spouse – the enigmatic, the impactful, the bothersome – all can
spark suspicions of neighbors using magic and dark powers; all can precipitate violence. The
suspects are sometimes normal humans, learned in dark magic, but other times, they are rumored
to be odious and other. They devour babies, fornicate with their menstruating mothers, and use
human skulls for sports. They become bats and black panthers, house pythons in their stomachs,
and direct menageries of attendant nightbirds. They plot the destruction of families and then
dance in orgiastic night-fests.
2
Humans in nearly every documented believe that some illnesses and hardships are the
work of envious or malignant group mates. Hutton (2004; 2017) reviewed ethnographies from
three hundred non-European societies and documented pervasive beliefs in sorcerers, witches,
the evil eye, and aggressive shamans. Of the 60 societies in the Probability Sample File of the
Human Relations Area Files – a pseudo-random sample of well-documented human societies –
59 believed in some form of human-induced mystical harm, the only exception being the Kogi of
Colombia
3
(sect. 2). European societies have historically held similar beliefs, embodied in the
Roman strix (Oliphant 1913; Oliphant 1914), the Saxon striga (Cohn 1976), and most famously,
the witches of the Great European Witch Hunt (Cohn 1976), and colonial New England
(Karlsen 1987).
Beliefs about harmful practitioners are profoundly similar across vastly distant societies
(Needham 1978; Kluckhohn 1959). The European witches of the late modern period were said
to eat human flesh, engage in obscene activities, and assemble in conspiratorial, orgiastic
1
The quotation comes from autobiography of Don Talayesva (Talayesva and Simmons 1942:379).
2
The quotes by Don Talayesva (opening) and the Santal guru Kolean Haram (section 3) demonstrate
that these beliefs are disturbing. The destruction mentioned in section 3 demonstrates that they’re
calamitous. Table 2 and section 6.2.1 describe the events that trigger suspicions of mystical harm. Table 3
features examples of animal transformations and attendants. Yamba witches were said to devour children
(Gufler 1999), Apache witches had sex with menstruating family members (Basso 1969), Akan witches
used human skulls for soccer (Debrunner 1961), and Santal witches met naked in nighttime assemblies,
danced, and copulated with their spirit familiars (Archer 1974). Nyakyusa witches had pythons in their
bellies (Wilson 1951).
3
The ethnographic texts included in eHRAF did not describe mystical harm beliefs in two PSF societies:
the Koreans and the Kogi. But researchers elsewhere have reported sorcery beliefs in Korea (Walraven
1980), so their omission seems due to ethnographers underreporting the topic. Meanwhile, Reichel-
Dolmatoff (1997:141; 1976:286) explicitly stressed the absence of beliefs in mystical harm among the
Kogi. Nevertheless, in describing Kogi lineages, he made a vague comment suggesting that people do in
fact believe in mean-spirited, uncanny harm: “Both groups, the Hukúkui as well as the Mitamdú, are
further regarded as vaguely dangerous and endowed with rather evil powers” (Reichel-Dolmatoff
1997:250).
ON THE ORIGINS AND DESIGN OF WITCHES AND SORCERERS
Page 3
nighttime gatherings (Cohn 1976). Similar behaviors were suspected of witches among the
Yamba of Cameroon (Gufler 1999), the Santal of South Asia (Archer 1984), and the Navajo of
the American Southwest (Kluckhohn 1944), among many other societies (Hutton 2017; Mair
1969; see sect. 2). And just as people worldwide believe in sensational and atrocious witches, they
also often suspect that sickness and death are the work of ordinary people secretly practicing dark
magic (e.g., Trobriand Islanders: Malinowski 1922; Tswana: Schapera 1952; Niimíipuu: Walker,
Jr. 1967).
In this paper, I refer to people believed to use magic or supernatural powers to injure
others as practitioners of mystical harm
4
. This term is broad, including, for example, beliefs about
werewolves, abhorrent witches, people whose stares transmit illness, and neighbors who use
voodoo dolls in secret. Magic refers to occult methods with instrumental ends, such as spells,
curses, rites, manipulated objects, and everyday superstitions. Magic can be used
5
to produce
socially-justified ends, such as healing people or succeeding in gambling, as well as less
acceptable objectives, such as inducing illness. I use refer to harmful magic as sorcery. Methods of
sorcery include cursing, stabbing voodoo dolls, and placing charmed poisons in people’s paths.
Sorcerers are people who use magic for malicious ends – that is, people who use sorcery.
Witches, on the other hand, exhibit up to three sets of characteristics: (1) They are existentially
threatening, (2) they have supernatural powers, and (3) they are morally repugnant. Some
practitioners qualify as both sorcerers and witches, such as those believed to both use magic and
engage in activities like graveyard conspiracies and cannibalism. I justify these definitions in
section 2.
The ubiquity of mystical harm beliefs and their striking similarities raise two basic
questions:
1. Why do humans believe in mystical harm?
2. Why do those beliefs take the form that they do?
This paper advances a tripartite theory to answer those questions. I propose that beliefs in
mystical harm, and beliefs about who orchestrates it, are the result of three cultural selective
processes:
1. Selection for intuitive magic. As people try to influence others’ misfortune, they
selectively retain intuitive magic, producing compelling spells and charms for harming
others. This produces intuitive harmful magic, but more relevantly, it convinces
people that sorcery works and that other group members practice it.
2. Selection for plausible explanations of misfortune. People look for explanations for why
things go wrong. When they feel threatened, they suspect distrusted group mates;
when they believe in sorcery, it provides a straightforward explanation for how a
4
I choose the term mystical to refer to harm that is transmitted either through magical means (e.g., spells,
buried poisons, voodoo dolls) or supernatural powers (e.g., transforming into an animal and attacking
someone, inflicting misfortune through an inadvertent envious stare). This usage follows Evans-Pritchard
(1937), who contrasted mystical causation with natural causation, and Needham (1978:26), who defined a
witch as “someone who causes harm to others by mystical means”, corresponding closely with my term
practitioner of mystical harm.
5
Whenever I refer to the effects of magic (e.g., producing illness) or the features of a malicious
practitioner (e.g., flying and eating corpses), I refer to beliefs about those traditions rather than actual
consequences or traits.
ON THE ORIGINS AND DESIGN OF WITCHES AND SORCERERS
Page 4
distrusted rival harmed them from afar. Over time, iteratively searching for plausible
explanations shapes beliefs about sorcerers to become increasingly compelling,
although the same process can produce explanations that do not include sorcery,
including beliefs about werewolves, the evil eye, and conspiratorial governments.
3. Selection for demonizing narratives. Actors bent on eliminating rivals devise
demonizing myths to justify their rivals’ mistreatment. These campaigns often target
and transform malicious practitioners, both because people suspect that malicious
practitioners transmit harm and because individuals accused of mystical harm are
easily demonized and abused.
On their own, these three processes produce beliefs and practices as varied as gambling
superstitions, conspiracy theories, and vitriolic campaigns against heretics, but in societies around
the world, they combine to produce the archetypal, odious image of the witch.
2. Cross-cultural patterns
Researchers struggle over whether beliefs about harmful practitioners are similar across cultures.
Many have emphasized commonalities (e.g., Mair 1969; Kluckhohn 1959), but others have
criticized drawing these comparisons, one scholar concluding that “anthropologists have
committed a possibly grave error in using the same term [witchcraft] for other cultures” (Crick
1973:18).
The most important effort in documenting cross-cultural patterns in these beliefs was
conducted by Hutton (2017; see also Hutton 2004). Hutton reviewed ethnographies in three
hundred extra-European societies and identified five characteristics that malicious magicians
around the world share with the early modern European conception of the witch. Namely, they
tend to (1) cause harm using non-physical, “uncanny” methods, (2) represent internal threats to
their communities, (3) gain their abilities through training or inheritance, (4) have qualities that
incite horror and loathing, and (5) give rise to strategies of resistance, including counterspells and
murderous campaigns. Hutton also reviewed, among other things, similarities in witches’
heinous activities and the social conditions that inspire violence towards suspected malicious
practitioners.
Hutton’s project was ambitious, but he sampled societies opportunistically, risking the
overrepresentation of peculiar beliefs. He also chose not to systematically code traits, such as how
frequently practitioners are believed to kill people or associate with animals. These limitations
prevented him from drawing strong inferences about how these beliefs compare around the
world.
I designed the Mystical Harm Survey (MHS) to systematically capture beliefs about
mystical harm in a representative sample of the world’s societies. The dataset covers the 60
societies of the Probability Sample File of the Human Relations Area Files, a pseudo-random
sample of well-documented cultures that were selected to make inferences about humanity more
generally (see the Supplementary Materials for more details). The full dataset is available at
osf.io/492mj and includes beliefs about 103 malicious practitioners (or practices) from 58
societies. The analyses reported here exclude leaders (e.g., elders, chiefs, senior lineages) and
public magicians (e.g., shamans, priests), because these practitioners are public, institutionalized
classes who advertise and perform their powers rather than simply being conceptions of group
ON THE ORIGINS AND DESIGN OF WITCHES AND SORCERERS
Page 5
mates causing misfortune (including leaders and magicians produces nearly identical results;
compare Supplementary Table 2 with Supplementary Table 4).
6
I used Principal Components Analysis to reduce the 49 raw variables in the Mystical
Harm Survey (e.g., does a practitioner consume flesh? do they cause economic harm?) to two
derived variables (or principal components)
7
, shown in Figure 1 (see Supplementary Materials for
details). This method exposes the axes along which practitioners vary the most and, thus, the
cross-cultural structure of these beliefs. Both of the derived variables are interpretable: The first
dimension represents how witchy malefactors are; the second distinguishes sorcerers, as
classically understood, from the evil eye.
Figure 1. Results of logistic PCA showing practitioners of mystical harm. A single point
represents a belief about a practitioner in a society (such as the Trobriand flying witch or the
Amhara evil eye); the accompany numbers refer to the unique practitioner ID numbers (see
Supplementary Table 1). The points are colored according to the terms used by the
ethnographer(s) who described them. They are scaled according to the number of paragraphs
coded in that society, ranging from 1 paragraph (practitioner 63) to 1,976 (practitioners 1 and 2).
The images refer to the features that characterize a given quadrant: eye = evil eye (unintentional
harm through stares or words); effigy = sorcery (learned magic); owl = witchiness (superhuman
abilities, moral abhorrence, threat).
6
Hereafter, I refer to this restricted dataset as the MHS and to the dataset including leaders and public
magicians as the expanded MHS.
7
There are two reasons to report a two-factor solution. First, a scree plot (Supplementary Figure 1) shows
a dramatic change in slope (or elbow) at the third component; after the second component, the additional
dimensions explain equivalent and smaller proportions of variance. Second, the third component is
uninterpretable (see Supplementary Table 3). The first and second components explain 23.1% and 16.8%
of the total variance, respectively (39.9% in total).
1
2
3
6
7
8
9
10
11
12
13
14
15
16
17
18
20
22
23
26
27
28 29
30 31
32
33
35
36
37
38
39
40
41
42
43
44
45
46
47
49
50
52
53
54
59
60
61
62
63 64
65
67
68
69
70
71
72
74
76
77 78
79
80
81
82
83
84
86
87
88 89
90
91
92
93
94
97
98
99
100
101
103
WITCHINESS (PC1)
SORCERY — EVIL EYE (PC2)
sorcery
cannibal/ghoul evil eye lycanthrope sorcerer witch
ON THE ORIGINS AND DESIGN OF WITCHES AND SORCERERS
Page 6
Practitioners high on the first variable (PC1) are witches.
8
They are believed to kill
people, cause illness, eat human flesh, desecrate corpses, use magic, fly, turn invisible, commit
atrocities at night and in the nude, congregate in secretive meetings, transform into animals or
use them as familiars, and engage in obscenities like incest and nymphomania; shamans and
other magicians are often suspected of being witches (see Supplementary Table 2 for loadings).
Practitioners low on this dimension lack these qualities. Contrary to many writers’ impressions
(e.g., Chaudhuri 2012; Mace et al. 2018; Sanders 1995), I did not find strong evidence that
witches are more frequently women than men.
The second derived variable (PC2) separates everyday sorcerers from the evil eye.
Practitioners low on PC2 use harmful magic, including spells, voodoo dolls, and magical poisons.
They attack their neighbors and family members but sometimes target out-group individuals as
well. Ethnographers often state that anyone can qualify as one of these practitioners, although
men and public magicians are suspected more often. Practitioners high on PC2, in contrast, tend
to possess the evil eye or blasting word: They harm people through their stares and comments,
often inadvertently. Their powers derive from physiological differences, such as special eyes,
rather than from learning specific methods or rites.
A surprising finding is that practitioners high on PC2 also tend to fly and eat human
flesh. But this is less characteristic of the evil eye and more a feature of cannibals, ghouls, and
lycanthropes (humans who transform into animals). In fact, no practitioner labeled “evil eye” by
an ethnographer was said to fly or consume human flesh. Cannibals, ghouls, and lycanthropes
likely appear with the evil eye in Figure 1, because they all tend not to use sorcery (shifting them
high on PC2) and they lack most other witchy qualities (shifting them low on PC1).
In Figure 1, I colored the points according to the ethnographer’s name for that
practitioner. These colors cluster, showing that terms like “sorcerer” or “witch” in fact capture
cross-culturally recurrent beliefs. Sorcerers (blue) are normal humans who use effigies, curses,
and other spells to harm their rivals. Descriptions of sorcerers are very similar to descriptions of
people generally knowing and using dark magic (purple). Possessors of the evil eye (yellow) harm
people with their stares and words, often unintentionally. They do not employ spells, and their
powers tend to be inborn rather than actively procured. Witches (pink) are much more variable
across societies, but they share up to three sets of traits: (1) They are threatening (e.g., they kill
and conspire in secret, nighttime meetings), (2) they are supernaturally powerful (e.g., they fly
and transform into animals), and (3) they are abhorrent (e.g., they consume human flesh and
desecrate corpses) (see Figure 2). This results of the PCA suggest that witchiness is a dimension
rather than a discrete trait – that is, some societies describe practitioners who are more
threatening, supernaturally powerful, and abhorrent than the practitioners described in other
societies.
8
Several variables, all of which appeared very infrequently in the MHS, had unstable loadings that
collapsed when the data from a single region were excluded from the PCA (see Supplementary Materials,
section 2.2 and Supplementary Tables 5 and 6). I have not reported these unstable loadings here, but see
Supplementary Table 2 for the full factor matrix.
ON THE ORIGINS AND DESIGN OF WITCHES AND SORCERERS
Page 7
Figure 2. (A) Witches’ Sabbath (Goya, 1798; ©Museo Lázaro Galdiano, Madrid) and (B) Witches’
Flight (Goya, 1798; ©Photographic Archive Museo Nacional del Prado) depict conceptions of
witches held by many medieval Europeans. The witches are nude and nocturnal; they fly, kill
babies, devour human flesh, associate with nighttime animals, and conspire with evil spirits.
Despite their strangeness and particularity, these traits were not restricted to medieval European
witches. People around the world – including the Tlingit (Pacific Northwest), the Akan (West
Africa), and the Trobriand Islanders (South Pacific) – held similar conceptions of witches.
The analysis helps reconcile a historic debate about the difference between witches and
sorcerers. Evans-Pritchard (1937) drew a strict boundary between the two, specifying that
malicious practitioners are either normal humans who use magic (sorcerers) or different entities
who do not use magic, instead attacking with supernatural powers (witches). He used the
dichotomous scheme to describe Azande beliefs in particular, but other anthropologists applied
the same typology to different ethnographic contexts (e.g., Reynolds 1963; but see Turner 1964).
Figure 1 reveals that Evans-Pritchard’s witch-sorcerer binary does not generalize. Some
heinous, supernaturally powerful practitioners (witches) only attack with supernatural stares and
thoughts, such as those of the Azande (9) and Akan (1), but many are believed to also employ
spells, charms, and other material magic. Some witches, for example, stuffed effigies into the
carcasses of dead puppies (Tlingit: De Laguna 1972:730); others recited spells to fly (Trobriand
Islanders: Malinowski 1922:241) or used horseshoes and keys to conjure evil spirits (Colonial
New England: Karlsen 1987:9). Thus, witches resemble other malicious practitioners, such as
ON THE ORIGINS AND DESIGN OF WITCHES AND SORCERERS
Page 8
sorcerers or possessors of the evil eye, except transformed along a dimension of witchiness, made
more threatening, more abhorrent, and more supernaturally powerful.
3. Existing theories of mystical harm
The most influential theories of mystical harm ascribe a function to these beliefs, often regarding
them as group-level adaptations. Most popular is the theory that these beliefs discourage socially
unacceptable behavior. According to this theory, if people suspect that their irate neighbors will
attack them with evil spells and powers, then people will refrain from upsetting each other, both
to avoid being attacked by mystical harm and to avoid being accused (Whiting 1950; Beattie
1963; Walker, Jr. 1967).
Faulkingham (1971:112) summarized this theory in observations of the Hausa (Niger):
“Sorcery beliefs in Tudù provide people with strong motivations to be gregarious and to avoid
quarrels. One is hesitant to be silent, alone, or bickering, lest he be accused of being a sorcerer.
Further, people are reticent to exacerbate quarrels, for they may become ensorceled.” But he also
recognized that these beliefs entail major costs: “While sorcery beliefs have these social control
functions, I believe that the villagers pay a high psychological price, since hostile emotions are
relentlessly proscribed” (Faulkingham 1971:112).
Other researchers have echoed Faulkingham’s second point, disputing cooperation
theories by noting how sorcery and witchcraft beliefs sow distrust and provoke quarreling
(Gershman 2016; see Hutton 2017:35 and works cited therein). Among the Kapauku Papuans,
most wars in one region (Mapia) started because of presumed sorcery; in another (Kamu),
sorcery accounted “for about thirty per cent of the conflicts” (Pospisil 1958:154). Other examples
of contexts in which sorcery and witchcraft accusations bred violence abound (e.g., Gebusi:
Knauft 2010; Rajputana: Skaria 1997; Yolngu: Warner 1958; Zulus: Bryant 1929). Suspicions of
magical harm can even inspire vitriol among family members, such as when a Klamath woman
slayed “her own mother for the fatal bewitchment of her child” (Stern 1965:21). An
ethnographer quoted the Santal (South Asia) guru Kolean Haram, who summarized the
sociological and psychological stresses of witchcraft beliefs: “The greatest trouble for Santals is
witches. Because of them we are enemies of each other. If there were no witches, how happy we
might have been” (Archer 1984:482).
Other scholars argue that beliefs in mystical harm explain misfortune. Evans-Pritchard
(1937) famously proposed this hypothesis in his report on Azande witchcraft. But the claim that
witchcraft beliefs explain misfortune cannot account for many features of those beliefs. Most
notably, why should people suspect that group mates engineer misfortune through magic or
supernatural powers when they can already blame gods, water demons, and other purported,
invisible harmful forces? Addressing this gap, Boyer (2001) pointed out that we are predisposed
to think about other people harming us. Humans are social animals, he observed, constantly
engaged in reciprocal favors. Thus, he hypothesized, we have evolved psychological mechanisms
that often interpret misfortune either as someone cheating us or as punishment for apparently
cheating others. As people adopt or develop explanations that conform to these expectations,
they produce beliefs in mystically powerful cheaters and cheater-detectors: “People who give
others the evil eye are overreacting cheater-detectors and witches are genuine cheaters” (Boyer
2001:200).
ON THE ORIGINS AND DESIGN OF WITCHES AND SORCERERS
Page 9
I borrow elements of the explanation hypothesis, but Boyer’s formulation suffers from
some of the same flaws as Evans-Pritchard’s: Both leave the content of witchcraft beliefs largely
unexplained, including why people use spells or charms or why witches transform into animals
and mutilate corpses. Boyer’s account also confronts a problematic inconsistency: If people with
the evil eye are “overreacting cheater-detectors”, then why is the evil eye linked so often to envy
(Dundes 1992), rather than feelings of being cheated?
Finally, many researchers connect mystical harm beliefs to sociological events, such as the
envy, inequality, and redistribution associated with social change (Comaroff and Comaroff 1999;
Bohannan 1958), the control of women (Hester 1992; Natrella 2014), and scapegoating (Oster
2004). But these accounts remain atomized and disconnected. They focus on single determinants
(such as rising inequality), most of which only apply in some circumstances, while failing to
describe many of the features of mystical harm beliefs.
I have left out many other explanations for these beliefs, including ones that invoke
repressed sexual impulses (Cohn 1976), distorted perceptions of existing or historic cults (Murray
1921), the inadvertent consumption of ergot fungi (Caporael 1976; Alm 2003), and delusions
resulting from psychiatric illness (Field 1970). These accounts suffer from many of the same
criticisms as those reviewed above. Not only do they fail to explain the content of mystical harm
beliefs, they also leave open the question of how shifting conditions should elicit some beliefs but
not others.
4. Introducing the tripartite theory: Cultural selection
I propose that mystical harm beliefs develop from the interaction of three cultural selective
processes. Cultural selection occurs when people preferentially retain particular practices or beliefs,
such as because they appear to more effectively produce a desired outcome (Blackmore 1999;
Boyd and Richerson 1985; Campbell 1965; Sperber 1996). For example, the cultural selection of
effective killing technology occurs as people adopt and maintain tools that kill animals or
enemies. As people modify their tools and keep the effective versions, they iteratively fashion
technology well-designed for killing, like sleek spears or bows-and-arrows. Notably, cultural
selection occurs whenever people use culturally-transmitted practices for some desired end and
they apply regular criteria to evaluate the effectiveness of those practices. Thus, selection can
produce sleek killing technology, but it can also produce chairs, cheesecake, Disney movies, and
other delights that satisfy desires humans want.
Cultural selective processes are significant for two reasons. First, they produce complex
traditions that no single individual could have devised in a single moment (Henrich 2015). But
just as importantly (although less frequently appreciated), these processes retain those traditions.
A spear, for example, may be used frequently yet remain unchanged for centuries. Although it
does not evolve, people selectively retain it for assassinating game and enemies.
Many scholars assume that cultural selective processes are protracted, involving
generations and many individuals, but they don’t have to be. Yes, selective processes can occur
over many generations: Myths demonizing Jews, for example, evolved over decades as people
throughout Europe borrowed and modified each other’s existing productions (Cohn 1967). But
cultural selection can also produce complex beliefs on very short time-scales with many fewer
ON THE ORIGINS AND DESIGN OF WITCHES AND SORCERERS
Page 10
participants, such as if several people concoct, maintain, and revise heinous myths about a feared
sub-group in the hours or days following a catastrophe.
I propose that mystical harm beliefs develop from three cultural selective schemes that
produce and maintain (1) intuitive techniques of harmful magic, (2) plausible explanations of
misfortune, and (3) myths that demonize a subgroup. The three proposed schemes occur under
different circumstances and frequently act independently of each other, separately producing
superstitions, conspiracy theories, and propaganda. But they also interact and develop each
other’s products, giving rise to beliefs in sorcerers, lycanthropes, evil eye possessors, and
abhorrent witches. In the following sections, I elaborate on each of these selective processes.
5. Magic
Figure 1 shows that people in many societies suspect that their misfortunes are caused by others
using sorcery. Why do people accept that sorcery works and presume that others practice it?
Here, I argue that these convictions develop from a selection for intuitive magic. People adopt
superstitions because of a predisposition to note spurious correlations between cheap actions
(such as wearing special underwear) and important, unpredictable outcomes (such as winning a
football game). As they then select among superstitions, they choose the most compelling ones,
driving the development and maintenance of intuitive magic (see Singh 2018 for an expanded
version of this argument). As a consequence, people accept the efficacy of magic, including
harmful sorcery, and understand that other group mates know it and might practice it.
5.1. The selective retention of intuitive magic
5.1.1. People adopt superstitions (magic) to influence significant outcomes that are important and
unpredictable
Rubbing rocks before giving speeches, wearing special underwear during football matches,
blowing on dice before letting them roll – we regularly use superstitions to nudge uncertainty in
our favor. Humans adopt magic or superstitions, which I define as interventions that have no
causal bearing on their intended outcome, when those outcomes are important (roughly, fitness-
relevant) and occur randomly (Ono 1987; Keinan 2002; Malinowski 1948). Such outcomes
include victory in war, the arrival of rain, recovery from illness, and rivals becoming sick, dying,
or suffering economic losses. That we adopt superstitions to control these outcomes seems a
result of a kind of bet-hedging psychology. When the costs of an intervention are sufficiently
small relative to the potential benefits (like wearing special underwear to win a football match),
and when the outcome seems to occur sometimes after the intervention, individuals benefit on
average from adopting those interventions (McKay and Efferson 2010; Johnson et al. 2013). The
predisposition to adopt superstitions to control uncertainty provides the basis for magical
practices across human societies (Vyse 2014), including, I propose, magic for harming others.
5.1.2. People selectively retain magical interventions that seem the most effective
ON THE ORIGINS AND DESIGN OF WITCHES AND SORCERERS
Page 11
Magic should culturally evolve to become more apparently effective. Humans have intuitions
predisposing us to regard some magical techniques, such as those with more steps and repetition
(Legare and Souza 2012), as more potent than others. As magic-users iteratively innovate and
select these more effective-seeming techniques, they produce intuitive magic. People around the
world share biases about how causality and efficacy work, so this selective process should produce
cross-cultural similarities in magical techniques (e.g., Nemeroff and Rozin 2000; Rozin et al.
1986), discussed below.
5.2. Ethnographic evidence for intuitive magic
At its basis, a selection for intuitive magic demands that people actually attempt to harm each
other using magical means. It also predicts that magic will be effective-seeming and that
common intuitive principles will characterize both harmful magic and other superstitions. Both
claims are supported by the ethnographic record.
5.2.1. People attempt harmful magic
People are notoriously reticent about discussing harmful magic with ethnographers, let alone
admitting to using it (e.g., Ames 1959:264; Nadel 1954:164). Nevertheless, researchers have
successfully documented direct and indirect evidence of people using private sorcery. During his
time with the Azande, Evans-Pritchard discovered two bundles of bad medicine in one of his
huts. One was engineered “to destroy the popularity of the settlement where I lived by killing
some people and making the rest afraid to remain there” (Evans-Pritchard 1937:402). The other
was planted to kill the anthropologist. Richards (1935) examined the magical horns collected in a
Bemba village during a witch-hunting movement in what-is-now Zambia. Although the vast
majority were harmless medicine containers, “11 out 135 horns were admitted by every one to be
undeniably bad destructive magic, that is to say, prepared for the injury of others” (Richards
1935:453). Researchers report other examples such as these (e.g., Anglo-Saxon England:
Crawford 1963; Wogeo: Hogbin 1938:231; Tlingit: Emmons and De Laguna 1991:410),
although people’s admissions of using sorcery and even accounts of other people discovering
evidence are difficult to interpret because of the possibility of deception.
Less contestable evidence of people using sorcery is the frequency with which specialists
sell harmful services and magicians or laypeople performing evil magic to harm out-group
enemies. Specialists sold harmful services in 26 of the 58 societies coded in the expanded MHS,
while in at least 10 of those societies, practitioners used magic and supernatural powers to attack
enemies of rival groups.
5.2.2. Malicious magic is governed by the same intuitive principles as other kinds of magic
The strongest evidence that magic, both harmful and otherwise, develops from a selection for
effective-seeming practices is that all kinds of magic are governed by the deeply intuitive
principles of sympathetic magic.
Sympathetic magic refers to two causal principles – the law of contagion and the law of
similarity (or homeopathy) – which guide magic around the world (Frazer 1920). The law of
ON THE ORIGINS AND DESIGN OF WITCHES AND SORCERERS
Page 12
contagion refers to the implicit belief that “physical contact between [a source object] and [a
target object] results in the transfer of some effect or quality (essence) from the source to the
target” (Nemeroff and Rozin 2000:3). This principle covers contamination or pollution, in which
a negative substance qualitatively changes a target object, as well as notions that acting on a part
(for example, on a lock of hair) can have an effect on the whole (for example, the person who
once owned it). That we wrongly but frequently believe in contagious magic seems in part a
misfiring of psychological mechanisms evolved for noting contamination and illness transmission
and perhaps overinterpreting the lingering effects of objects on each other (Rozin and Nemeroff
2002; Apicella et al. 2018).
In contrast to contagion, the law of similarity or homeopathy refers to the impression
that “things that resemble each other at a superficial level” – like a voodoo doll that resembles a
person – “also share deeper properties” (Nemeroff and Rozin 2000:3) – for example, that acting
on the doll produces effects on the imitated target. It remains unclear why people so habitually
make this association, but as with the law of contagion, it likely reflects misfiring biases in causal
reasoning.
Frazer (1920, Ch. III) famously documented examples of both contagious and similarity-
based magic around the world. Among his many cases of contagious magic, he noted that people
often believe that one can affect a target by magically treating the impressions it leaves, such as
footprints. Footprints feature in malicious magic, like when people tamper with a target’s prints
to induce illness or pain, and in hunting magic, like when pursuers locate the tracks of animals
and doctor them to slow the target (see Table 1). Among his many examples of similarity-based
magic, Frazer (1920) documented the frequent belief that one can influence a target by creating
and manipulating an effigy of it. Table 1 reviews examples of both malicious and non-malicious
magic that uses effigies.
Table 1. Malicious magic is governed by the same intuitive principles of sympathetic causality
that structure other kinds of magic. *Examples documented by Frazer (1920).
Magical method
Examples of malicious magic
(societies with references)
Examples of other magic
(societies with references)
Treating the footprints of a
target, such as to harm a
person (malicious magic) or
aid in the capture or warding
off of animals (other magic)
*Chero
*Maori
Natinixwe (Wallace and Taylor 1950,
pp. 189-90)
Niimíípu (Walker, Jr. 1967, p. 74)
Siwai (Oliver 1955, p. 87)
Tswana (Schapera 1952, p. 45)
Ainu (Munro 1963, p. 113)
Azande (Lagae 1999, pp. 146-47)
Fox (Jones 1939, pp. 23-24)
*Khoikhoi
*Nlaka’pamux
Persians (Massé and Messner 1954,
p. 282)
Manufacturing and treating
an effigy, such as to injure a
target (malicious magic) or
induce birth or drive away
neighbors (other magic)
Ancient Egyptians (Budge 1901, p.
75)
Colonial New England (Karlsen
1987, p. 8)
*Kenyah
*Malay
*Ojibwe
*Basotho
Egyptians (Ammār 1954, p. 89)
*Inuit
*Japanese
*Nisenan
Pomo (Aginsky 1939, pp. 212-13)
ON THE ORIGINS AND DESIGN OF WITCHES AND SORCERERS
Page 13
Sami (Karsten 1955, pp. 43-44)
6. Explanations
The selection of intuitive magic convinces people that malevolent magic is effective and that
others practice it. How does this then transform into beliefs about sorcerers and witches who
cause harm?
In this section, I propose that, under certain circumstances, people’s hypervigilant
tendencies lead them to suspect that group mates engineer inexplicable misfortunes. As they
iteratively consider how those group mates harmed them, people maintain a selection for plausible
explanations of misfortune. When they believe that sorcery is effective, people may suspect and
develop beliefs about sorcerers, although they may consider other means of transmitting harm,
such as animal transformation, the evil eye, and even governmental conspiracies.
6.1. Selection for plausible explanations of misfortune
6.1.1. People suspect distrusted group members in the wake of impactful, negative outcomes
Whether we lose a wallet or observe an epidemic sweeping through our community, we
commonly attribute impactful, hard-to-explain events, especially negative ones, to the wicked
intentions of other humans (Tennen and Affleck 1990). These tendencies seem to have evolved
to vigilantly recognize threat (Raihani and Bell 2018). Our social lives are marked by conflict, so
we benefit from tracing and anticipating when spiteful others harm us, even if it means making
occasional mistaken attributions (see error management: Johnson et al. 2013; McKay and
Efferson 2010).
A growing body of literature, most of it in the psychological sciences, shows that a person
is most likely to suspect other people for causing some misfortune under four conditions:
(1) The person feels threatened (Abalakina-paap et al. 1999; Mirowsky and Ross 1983;
Saalfeld et al. 2018; Mashuri and Zaduqisti 2015);
(2) They are distrustful of others (Abalakina-paap et al. 1999; van Prooijen and Jostmann
2013; Raihani and Bell 2017);
(3) They confront an event that is hard to explain (Rothschild et al. 2012; van Prooijen and
Douglas 2017; van Prooijen and Jostmann 2013);
(4) That event is impactful (van Prooijen and Douglas 2017; van Prooijen and van Dijk
2014; McCauley and Jacques 1979).
These conditions are enlightening for two reasons. First, they provide evidence for
adaptive hypotheses of paranoid thinking. People benefit from identifying mean-spirited rivals
who conspire to harm them, so it’s reasonable that our psychology has evolved to seek out these
individuals when they are most likely to harm us. Second, identifying these conditions generates
predictions for the contexts under which people are most likely to develop beliefs in mystical
harm. If some adaptive psychological machinery provides a psychological foundation for sorcery
and witchcraft, then the conditions that trigger that psychology should in turn breed suspicions
of mystical harm. I discuss these predictions in section 6.2.
ON THE ORIGINS AND DESIGN OF WITCHES AND SORCERERS
Page 14
6.1.2. People selectively retain plausible explanations for how group mates harmed them
Humans constantly seek explanations (Frazier, Gelman, and Wellman 2009; Lombrozo 2006).
When your money-purse goes momentarily missing in a coffee shop and you suspect the wait
staff or your fellow patrons, you automatically consider the various ways by which they might
have accomplished their misdeed. You deem some explanations likelier than others – for
example, that it was stolen once rather than stolen and returned and then stolen again, or that it
was stolen by the grungy crust-punk rather than by the well-to-do suburban family to his left.
The process of inferring an explanation by comparing hypotheses against each other and
selecting the best among them is known as “inference to the best explanation” (Harman 1965).
People suffer many hard-to-explain misfortunes, such as illness, the death of a loved one,
and a burnt-down house. I propose that as they search for explanations for how suspected rivals
engineered those harms, they retain the most plausible explanations. A distrustful person whose
livestock dies, for example, will search for an explanation for how a rival committed the act. They
will consider explanations that they have learned, concoct other stories, and ask knowledgeable
group mates. As other people suffer similar, inexplicable injuries, and as people share their
conclusions and suspicions with each other, communities spin more and more conceivable tales
for how heinous group members abused them from afar. When people believe in the efficacy of
malicious magic (following section 5), it provides a sufficient and parsimonious answer, easily
accounting for invisible, distant harm.
In societies without strong beliefs in magic, this selective process still occurs, although it
converges on different explanations. One explanation is that powerful governments mastermind
misfortune. In his analysis on paranoia in US politics, Hofstadter (1964) noted that people often
attribute their troubles to distrusted governments or the puppeteers controlling them, such as the
Catholics, Free-Masons, and Illuminati. Barkun (2013) showed that these theories evolve.
Milton Cooper, for example, tweaked and synthesized existing theories about the Illuminati, the
CIA, the Kennedy assassination, observations of cattle mutilations, and the AIDs epidemic. His
super-conspiracy theories comprehensively explained both the momentous and the puzzling,
producing an unparalleled appeal. As I am write this, his 1991 book Behold a Pale Horse (Cooper
1991) ranks 2,998th among all books on Amazon.com, besting the highest-selling editions of The
Iliad, War and Peace, and Uncle Tom’s Cabin.
Beliefs about mystical practitioners should evolve like contemporary conspiracy theories.
Over time, they should become more internally consistent and plausible while encompassing a
wider set of inscrutable events.
6.2. Ethnographic evidence for plausible explanations of misfortune
I have argued that beliefs in mystical harm develop to explain how distrusted group mates
attacked a person from afar. At least two basic predictions follow: (1) Beliefs in mystical harm
should track distrust and suspicions of harmful intent, and (2) malicious practitioners should be
suspected of causing calamitous, negative events, especially ones for which people lack alternative
explanations. Meanwhile, that these beliefs develop from a selection for the most plausible
explanations clarifies why malicious practitioners often associate with, and transform into,
animals.
ON THE ORIGINS AND DESIGN OF WITCHES AND SORCERERS
Page 15
6.2.1. Accusations of mystical harm track distrust and suspicions of harmful intent
People who suffer calamity overwhelmingly suspect individuals with a presumed interest in
harming them. When several girls fell into possessed fits in Salem Village in 1692, many of the
girls’ families’ political rivals were suspected of attacking the girls and their allies (Boyer and
Nissenbaum 1974). Among the Azande, “A witch attacks a man when motivated by hatred,
envy, jealousy, and greed… Therefore a Zande in misfortune at once considers who is likely to
hate him” (Evans-Pritchard 1937:100). For the Trobriand Islanders, “the passions of hatred,
envy, and jealousy” are expressed “in the all powerful sorcery of the bwaga’u [sorcerer] and
mulukwausi [witch]” (Malinowski 1922:395). Many ethnographers studying other societies have
made similar comments (e.g., Tlingit: De Laguna 1972:730; Tikopia: Firth 1954:114; Ona:
Gusinde 1971:1102; Tukano: Reichel-Dolmatoff 1971:156-157; Pawnee: Weltfish 1965:337).
People regard envy in particular as a potent, malicious emotion. They not only suspect
that envious individuals want to harm them, but in societies everywhere, people believe that the
emotion itself transmits mystical harm, such as through covetous stares (the evil eye) or jealous
compliments (the blasting word) (Dundes 1992). Beliefs in the harmful effects of envy likely
exist because envy drives malice. Individuals who experience envy are more likely to injure better-
positioned targets (Smith and Kim 2007; Miceli and Castelfranchi 2007) and even derive
pleasure when envied persons suffer (van de Ven et al. 2015; Smith et al. 1996). Thus, a person
who expresses envy betrays a desire to harm, making them a key suspect after things go wrong.
The theory proposed here also predicts that beliefs about witches, sorcerers, and evil eye
possessors should prosper in communities with lower levels of trust compared to those with
higher levels. This explains why mystical harm beliefs increase with conditions that exacerbate
distrust, such as growing inequality and the resulting rise in envy (e.g., Lederman 1981).
9
6.2.2. Mystical harm explains impactful and unexplainable misfortunes
I argued that paranoid tendencies intensify when the impact of a misfortune is high and it is
unexplainable. If beliefs in mystical harm develop from these tendencies, people should fault
malicious practitioners for high-impact and inexplicable injuries.
People overwhelmingly accuse malicious practitioners of causing impactful hardship. Of
the 83 practitioners or practices in the MHS, at least 78% were said to cause illness, 77% death,
30% economic trouble, and 16% catastrophes (such as hailstorms or epidemics). In total, 94%
were reported as producing at least one of those outcomes.
Ethnographic descriptions often focus on the inexplicability of these hardships (e.g.,
Nsenga: Reynolds 1963:19; Kerala Brahmins: Parpola 2000:221). The Navajo attributed illnesses
to witchcraft when they were “mysterious from the Navaho point of view” or “persistent,
9
Analyzing Pew survey data in nineteen sub-Saharan African, Gershman (2016) reported a robust,
negative correlation between the prevalence of mystical harm beliefs and several measures of trust. He
acknowledged that the evidence was correlational yet preferred the interpretation that mystical harm
beliefs erode trust. This is reasonable – people who understand illness and death to be the handiwork of
evil group members should grow more distrustful of them – but the proposed theory also predicts the
opposite direction of causality. As I discussed, people who distrust others should suspect them of causing
unexplainable misfortunes, and sorcery provides a parsimonious explanation.
ON THE ORIGINS AND DESIGN OF WITCHES AND SORCERERS
Page 16
stubbornly refusing to yield to usual Navaho treatment” (Kluckhohn 1944:54). Other strange
circumstances, such as the appearance of unexplained tracks, were taken as further evidence.
When the Tiwi experienced a decrease in mortality from fighting, raids, and neglected wounds,
they attributed the resulting increase in natural deaths to a rise in poison sorcery (Pilling
1958:123).
People attribute random calamities aside from death, disaster, illness, and material loss to
mystical malice. Ten of the 83 practitioners in the MHS were said to produce sterility; 12
influenced love and attraction. Witches in colonial New England were rumored to cause
clumsiness, falling, fires, forgetfulness, barrenness, deformed children, spoiled beer, storms, sleep
paralysis, and unusual behavior in animals (such as a cow wandering off or a sow knocking its
head against a fence) (Karlsen 1987). Table 2 includes every example of harm or misfortune
recorded in the MHS that does not qualify as death, injury, love, sterility, catastrophe, or
economic trouble. Nearly early all are inexplicable and bothersome.
Table 2. Every example of harm or misfortune recorded in the MHS that does not relate to
death, injury, sickness, love, sterility, catastrophe, or economic trouble. Citations appear in the
MHS dataset.
Society (with practitioner* and MHS
practitioner ID)
Harm of misfortune
Akan, obayifo/witch [1]
Accidents (including lorry accidents); bad behavior of
wife; becoming a drunkard; burnt-down house;
cracks in buildings; ill luck; poor performance on
school exams; pregnant men
Amhara, buda/evil eye [3]
Croaking or worsening of singer’s voice
Aymara, laiqa/sorcerer [8]
Accidents; failure in fishing
Azande, aboro mangu/witch [9]
Burnt-down hut; coldness of prince towards subject;
failed magic; ruined performance of witch-doctor;
sulkiness or unresponsiveness of wife
Azande, aira kele ngwa/sorcerer [10]
Outcome of divination (poison oracle)
Azande, irakörinde/possessor of teeth [11]
Broken items, including stools, pots, and bowls
Azande, women’s sexual magic [12]
Bad luck
Chukchee, sorcery [22]
Losing strength while wrestling; slowing down in a
foot- or reindeer-race
Chuuk, souboud/sorcerer [23]
Disturbed growth; falling or tripping during
competition (basketball)
Dogon, yadugonu/witch [27]
Temporary muteness
Highland Scot, buidseachd/witchcraft [40]
Stuck or overturned truck
Hopi, bowaka/witch [42]
Malicious gossip; misbehavior of children
ON THE ORIGINS AND DESIGN OF WITCHES AND SORCERERS
Page 17
Iroquois, witch [47]
Confusion in sports competitions
Lau Fijians, raw eyes [61]
Skin discoloration (i.e., becoming tan)
Lozi, muloi/witch [64]
Inability to perform acrobatics; inability to score
during football
Ojibwa, windigo/cannibal spirit [71]
Overturned canoes
Pawnee, witch [74]
Stopped rain
Santal, sorcery [77]
Deception
Saramaka, sorcery [78]
Boat accidents
Tarahumara, sukurúame/sorcerer [89]
Outcomes of competitions (e.g., races); twins
Tiv, mbatsav/witch [91]
Appearance of baldspots; bad dreams; burnt clothes;
“whatever goes wrong if there is no more convenient
explanation”
Tlingit, land otter sorcery [93]
Disappearance
*The indigenous term for the practitioner or practice with the ethnographer’s term or translation
6.2.3. Animals associated with mystical harm explain impactful misfortune and invisible harm
Those animals associated with malevolent supernatural practitioners provide further evidence
that these beliefs serve as compelling explanations of misfortune. Table 3 displays all of the
animals associated with harmful practitioners recorded in the MHS, separated into those animals
believed to be transformed practitioners and those animals that act as their servants, steeds, or
helpers.
Table 3. Every example in the MHS of practitioners either transforming into animals (including
the practitioner’s soul entering or becoming an animal) or working with animals (including spirit
familiars taking animal form). Citations appear in the MHS dataset.
Animals into which practitioners transform
Society
(with practitioner* and MHS practitioner ID)
Animal
Akan, obayifo/witch [1]
Antelopes, bulls, bushpigs, centipedes, cows, crop
worms, crocodiles, dogs, hyenas, leopards, lions,
lizards, owls, rats, red deer, snakes (including
poisonous ones), squirrels, tsetse fly
Amhara, buda/evil eye [3]
Hyenas
Azande, aboro mangu/witch [9]
Bats
Bahia Brazilians, lobishomem/werewolf [15]
Wolves
Dogon, lycanthrope [28]
Eagles, panthers
ON THE ORIGINS AND DESIGN OF WITCHES AND SORCERERS
Page 18
Eastern Toraja, topokantoe/sorcerer [29]
Snakes
Eastern Toraja, taoe mepongko/werewolf [30]
Buffalo, cats, deer, dogs, pigs, white ants
Garo, lycanthropy [36]
Any beast or reptile, including crocodiles, snakes,
and tigers
Hopi, bowaka/witch [42]
Animals, including coyotes, foxes, lizards, and
wolves
Iroquois, witch [47]
Any animal, including dogs, pigs, turkeys, and
owls
Kapauku, meenoo/cannibal [53]
Dogs, hawks
Lozi, muloi/witch [64]
Hyenas, lions
Mataco, ayīeu/sorcerer [68]
Horses, jaguars, venomous reptiles (including
rattlesnakes)
Santal, tonhi/witch [76]
Bears
Serbs, vještice/witch [79]
Insects, reptiles, sparrows
Tiv, mbatsav/witch [91]
Chicken leopards (?), crocodiles, foxes, leopards,
lions, monkeys, owls, witch cats (?), other birds
(akiki, kpire)
Tlingit, nukwsati/witch [92]
Cranes, geese, owls, porpoises, sea lions
Trobriand Islanders, yoyova/flying witches [94]
Fireflies, flying foxes, nightbirds
Wolof, doma/witch [101]
Ants, cats, donkeys, hyenas, monkeys, owls,
snakes, vultures
Animals associated with practitioners
(e.g., familiars, mounts)
Society
(with practitioner* and MHS practitioner ID)
Animal
Akan, obayifo/witch [1]
Antelopes, bats, chameleons, cocks, crabs, dogs,
eagles, electric fish, goats, horses, house flies,
leopards, lions, lizards, lice, owls, rats, smart
hawks (?), snakes (including black mambas, black
snakes, green mambas, puff adders, pythons,
spitting cobras, thrush striped snakes), soldier
ants, tsetse flies, wasps, weaver birds, wolves
Amhara, buda/evil eye [3]
Hyenas
Aymara, laiqa/sorcerer [8]
Nighthawks, owls
Azande, aboro mangu/witch [9]
Nocturnal birds and animals, including bats,
jackals, and owls
Bahia Brazilians, lobishomem/werewolf [15]
Dogs
Bemba, muloshi/witch [17]
Magical birds, owl-like birds
Blackfoot, medicine [18]
Spiders
ON THE ORIGINS AND DESIGN OF WITCHES AND SORCERERS
Page 19
Chukchee, sorcery [22]
Dogs, reindeer
Eastern Toraja, taoe mepongko/werewolf [30]
Black cats, snakes
Eastern Toraja, taoe meboetoe/werewolf [31]
Black cats
Garo, lycanthropy [36]
Animals that live in the forest, including
elephants, crocodiles, snakes and other reptiles,
and tigers
Hopi, bowaka/witch [42]
Lizards
Lozi, muloi/witch [64]
Jackals, lizards, nightjars, owls, rats, water-snakes
Ojibwa, witchcraft [72]
Snakes, wolverines
Pawnee, witch [74]
Owls
Santal, tonhi/witch [76]
Dogs, tigers
Serbs, vještice/witch [79]
Birds, insects, small reptiles, snakes
Tarahumara, sukurúame/sorcerer [89]
Invisible birds
Tiv, mbatsav/witch [91]
Cats, nightjars, owls, snakes
Tzeltal, witch [100]
Snakes
*The indigenous term for the practitioner or practice with the ethnographer’s term or translation
A cursory glance reveals that many of the animals fall into one of two categories. First are
those creatures responsible for calamities, such as man-killers and crop-destroyers. Snakes, bears,
tigers, wolves, and crocodiles all attack humans, leaving wounded individuals searching for
explanations. Hypervigilant people should immediately suspect their enemies, and ethnographic
descriptions show that this frequently occurs. To the Akan, snakes bring “sudden and most
unpleasant death”, so “anyone who has a narrow escape from a snake comes to ask who sent it
and why” (Field 1970:130). Archer (1984:486) recorded an incident among the Santal of South
Asia when a man was mauled by two bears. He soon consulted a witch finder to learn who was
behind the attack.
Another class of ruinous misfortune is the destruction of crops. The Akan accused
witches of becoming squirrels, rats, crop worms, antelopes, bush pigs, cows, bulls, dogs, and red
deer – but all of those suspicions followed incidents when those animals consumed or destroyed a
person’s harvest (Debrunner 1961).
The second major category includes those animals whose alliance or transformation
explains how dark practitioners commit their wickedness unseen, such as owls, nightjars, flying
foxes, and fireflies. In all of these instances, people seem confident that a group mate harmed
them and, noticing these animals flitting about, find their appearance the missing explanatory
piece for how a distrusted rival harmed them.
Several animals do not fall into the above categories, but their associations with malicious
practitioners still seem to parsimoniously explain puzzling events. The Tlingit believed that
witches could become porpoises and sea lions, but these suspicions occurred when those animals
behaved enigmatically, lacking “the normal fear of human beings displayed by ordinary wild
animals” (de Laguna 1972:731). Thus, an ailing sea lion that remained near people’s houses and
porpoises that swam too close to shore were suspected of being metamorphosed witches.
ON THE ORIGINS AND DESIGN OF WITCHES AND SORCERERS
Page 20
Hyenas were associated with malicious magicians among the Wolof, Amhara, and Lozi,
as well as many cultures not included in the MHS, such as the Kaguru of Tanzania (Beidelman
1975) and Persians in medieval India (Ivanow 1926). This association seems the result of
demonizing narratives feeding back on plausible explanations. If people believe that certain
individuals have superpowers and feast on human flesh (as shown in Figure 1 and discussed in
the next section), they should start to suspect transformation when they witness nocturnal hyenas
digging up corpses.
7. Evil
The above two processes fail to explain the extreme heinousness of witches, such as their
cannibalism and graveyard conspiracies. Here, I propose that these features develop from a
selection for demonizing narratives – specifically, from a selection for those traits that justify the
mistreatment of accused practitioners and even spur other group mates to remove them.
7.1. Selection for demonizing narratives
7.1.1. People promote demonizing narratives when they want to justify mistreatment of a group
The cannibalism, conspiratorial meetings, and existential threat posed by witches are peculiar
commonalities, but they are not unique. Sociologists studying moral panics and elimination
campaigns in Western contexts have documented similar “folk devils”, with target groups
ranging from youth sub-cultures (Cohen 1972) to Jews (Cohn 1967; Cohn 1966). Their
analyses, together with insights from psychological research, reveal why these narratives recur
with such consistency around the world.
Folk demonization usually occurs because one group – hereafter, the Campaigners –
wants to justify the mistreatment of another – hereafter, the Targets (Goode and Ben-Yehuda
2009). Targets can be social groups, such as Jews or heretics, but they can also be those people
who do some behavior, like people who use LSD (Goode 2008).
Campaigners demonize Targets for several, non-exclusive reasons, including (a)
competition, such as when removing Targets opens up resources, (b) existential fear, such as
when Targets are believed to threaten Campaigners, and (c) moral campaigns, such as when
Campaigners want to curb some behavior. The foundations of these motivations can be
legitimate, like if removing victims frees up benefits that the Campaigners can enjoy (e.g., Philip
IV’s motivation to arrest the Knights Templar: Barber 2006), or mistaken, such as when
Campaigners wrongly understand Targets to be threatening (e.g., panics about satanic groups:
Victor 1989).
To mistreat Targets, Campaigners must often gain the approval of other group mates –
hereafter, the Condoners. They can secure this approval by promoting sensational myths that
justify abusing the Targets. People might craft these myths deliberately, as in many propaganda
campaigns (e.g., Desforges 1999), but they can also do so unconsciously. People reflexively
attend to and exaggerate evidence that supports their goals and their claims (Nickerson 1998;
Kunda 1990), a tendency arguably designed to sway others (Mercier and Sperber 2011).
ON THE ORIGINS AND DESIGN OF WITCHES AND SORCERERS
Page 21
As Campaigners refine portrayals of Targets that justify and urge violence, they
selectively retain demonizing narratives. The iterative crafting of heinous myths about Jews
illustrates this process. For example, Cohn (1967) tracked the history of The Rabbi’s Speech, a
fabricated speech by a chief rabbi describing the Jews’ plot to control finance and undermine
Christianity. The speech started as a fictional chapter in an 1868 novel recounting a
conspiratorial meeting between representatives of the twelve tribes of Israel and the Devil. In the
years afterwards, the chapter was borrowed, modified, distributed in pamphlets, and reprinted as
purported fact. In an 1881 version from France, the many speeches had been consolidated into a
single address, the satanic element was absent, and a note was included explaining that the
document came from a forthcoming book by an English diplomat, vouching for its authenticity.
7.1.2. Demonizing narratives develop and are maintained during stressful uncertainty
For demonizing narratives to flourish, Condoners need to believe them. But this is often not the
case because people are armed with cognitive adaptations that recognize and protect against
deception (Sperber et al. 2010). In fact, ethnographers occasionally report people’s skepticism
about the existence or portrayals of evil magicians (e.g., Tswana: Schapera 1952:44).
Condoners should be gullible or credulous in at least two conditions. First, they should
accept information when it comes from influential or trusted sources, such as religious authorities
or the media. Second, and more relevantly, people should become receptive when they need
valuable information, especially during times of unexplainable stress. Research on social learning
and gossip show that uncertainty, especially about important events, motivates individuals to
pursue social information (Boyd and Richerson 1988; Laland 2004; Morgan et al. 2012; Rosnow
1991).
In conclusion, times of unexplainable disaster breed paranoid suspicion while leaving
injured parties intensely credulous. This combination of mistrust and gullibility allows fearful or
exploitative campaigners to invent abominable witches.
7.2. Ethnographic evidence for demonization
7.2.1. Witches are well-designed to induce punitive outrage
In section 2, I showed that witches exhibit many common features, two of the most striking
being (1) their threatening nature and (2) their moral abhorrence, especially their cannibalism
and defilement of human bodies. These behaviors ignite severe punitive ire, encouraging violence
towards those actors.
Depicting a group as an existential threat – organized and secretive yet powerful and
conspiratorial – is effective, because, in short, people want to remove threats. A vast literature
shows that people are more willing to invest in collective action when they feel existentially
threatened (e.g., Johnson and Frickel 2011; Berry 2015; Maher 2010). Meanwhile, researchers
note that people use past harms committed by a group to justify violence and mistreatment
towards it (Sullivan et al. 2012) and people forgive aggressors when reminded of these wrongs
(Wohl and Branscombe 2009). If narratives develop to maximally support and provoke violence
ON THE ORIGINS AND DESIGN OF WITCHES AND SORCERERS
Page 22
towards demonized Targets, Targets should be portrayed as representing as large a threat as is
believable.
Aside from conspiratorially plotting destruction, witches engage in atrocious behaviors,
most frequently cannibalism and corpse desecration, but also acts such as necrophilia (e.g.,
Navajo: Kluckhohn 1944) and incest (e.g., Apache: Basso 1969; Kaguru: Beidelman 1963).
What accounts for their pervasiveness? As readers can attest, these acts trigger an intense,
visceral moral outrage (Haidt, Björklund, and Murphy 2000). For the !Kung, “the two worst
sins, the unthinkable, unspeakable sins, are cannibalism and incest” (Marshall 1962:229), while
among the Comanche, “the very idea that one of them might under stress eat another person was
vigorously repulsed” (Wallace and Hoebel 1952:70). In fact, the repugnance at cannibalism is so
intense that some societies even claim to forbid the consumption of animals that resemble
humans, exemplified in taboos on the Amazon river dolphin and nutria (a large semiaquatic
rodent) among the Warao (Wilbert 1972:69).
One possible reason for our revulsion at acts like cannibalism and necrophilia is that they
indicate that an actor is dangerous and not to be trusted. People may have evolved psychological
mechanisms to select social partners who are predictable and safe. Any individual who even
considers an atrocious behavior, like consuming flesh, having sex with dead bodies, or mutilating
corpses, reveals an underlying preference that makes them perilous social partners (Tetlock 2003;
Hoffman, Yoeli, and Nowak 2015). Our revulsion at these acts may be enhanced by feelings of
disgust, which have been shown to heighten moral judgment (Schnall et al. 2008).
Regardless of why we abhor cannibalism and other obscenities, the broader point is that
those acts invite severe punitive outrage, making them potent for justifying and urging
elimination. Should some other set of behaviors provoke greater outrage, the proposed theory
predicts that witches will do those instead (assuming that people will believe the accusations).
7.2.2. Witches resemble the demonized targets of other moral panics and eradication campaigns
The traits of witches are sensational and atrocious, but they are not unique. Other panics and
campaigns of mistreatment – such as attacks on heretics and dissidents, moral panics during
times of stress, and conspiracy scares – similarly transform targets into witch-like demons. Table
4 lists some examples. Note how frequently these groups supposedly pose existential threats and
violate sacred values.
Table 4. The targets of moral panics and elimination campaigns resemble witches, especially by
posing existential threats and violating sacred values.
Selected groups
Traits ascribed (with references)
Christians,
100s, Roman Empire
Worship a donkey-god or genitals of priest; engage in secretive
meetings, infanticide, child-cannibalism, and nighttime, incestuous
orgies; “threaten the whole world and the universe and its stars with
destruction by fire” (Felix and Rendall 1972, p. 337-41)
ON THE ORIGINS AND DESIGN OF WITCHES AND SORCERERS
Page 23
Knights Templar, early 1300s,
France
Deny Christ; spit, trample, and urinate on the cross; engage in
homosexual practices, including disrobing newcomers and kissing
them; collect in secret meetings at night; are bound by oaths enforced
by death; swear to advance the Order at all costs, lawful or not
(Barber 2006, p. 202-203)
Fraticelli “de opinione” (radical
Christian sect), 1466, Rome
Enjoy nighttime orgies in crypts; sacrifice a small boy, make powder
from his body, and consume it communally in wine during mass
(Cohn 1976, p. 46)
Catholics, mid-1800s, United
States
“The anti-Catholics invented an immense lore about libertine priests,
the confessional as an opportunity for seduction, licentious convents
and monasteries… Infants born of convent liaisons were baptized and
then killed” (Hofstadter 1964, p. 80-81).
Mau Mau rebels, 1950s,
Kenya
Mutilate victims’ corpses; take secretive oaths at night that involve
obscenities like public masturbation and drinking menstrual blood
(Lonsdale 1990, p. 398-400)
Communists, 1965, Indonesia
Murder, torture, and castrate generals; woman’s Communist group
dances naked at night; plot nation-wide purge of anti-Communists
(Henry 2014; Wieringa 2011)
Tutsis, early 1990s, Rwanda
Send women to seduce Hutu and infiltrate positions of power; plot a
war to reestablish control, massacre Hutu, and establish Nilotic
empire across Africa; admire Nazis and engage in cannibalism; elders
kill and pillage and rape girls and women (Desforges 1999, p. 72-83)
8. Discussion
8.1. The origins of sorcerers, lycanthropes, the evil eye, and witches
Table 5 displays the three cultural selective processes hypothesized to be responsible for shaping
beliefs in practitioners of mystical harm. Figure 3 shows how those processes interact to produce
some of the malicious practitioners identified in Figure 1 (sorcerers, the evil eye, lycanthropes,
and witches).
Table 5. The three cultural selective schemes responsible for beliefs in practitioners of mystical
harm.
CULTURAL SELECTIVE
SCHEME
What is being selectively
retained?
CONTEXTS
When should we expect it to
occur?
FEATURES OF BELIEFS IN
MYSTICAL HARM
Which features of mystical harm
beliefs does this process
produce?
ON THE ORIGINS AND DESIGN OF WITCHES AND SORCERERS
Page 24
Intuitive magic (section 5):
Effective-seeming interventions
for harming or killing others
When people want to harm
rivals
That harm can be transmitted
through sympathetic means
(contagion, similarity); that
harmful magic is effective and
that others do it
Plausible explanations (section
6): Explanations for impactful
misfortune
Following unexplainable,
harmful misfortune, especially
when people are distrustful or
persecuted
That impactful and
unexplainable harm is caused by
magic and supernatural powers;
that malicious practitioners are
envious or offended; that they
associate with animals, especially
man-killers and nighttime or
tiny animals
Demonizing narratives (section
7): Narratives that justify and
urge mistreatment of a target
group
When influential individuals
aim to remove a sub-group;
during stressful uncertainty
That malicious practitioners are
threatening (e.g., conspire, kill);
that they violate sacred values
(e.g., eat corpses)
Figure 3. The three selective schemes responsible for beliefs in practitioners of mystical harm.
Practitioners of mystical harm are bolded; examples of other practices and beliefs are unbolded.
The intersection of demonizing narratives and intuitive magic is filled because no beliefs should
exist there – any demonizing narrative in which the target uses magic should also blame the
target for terrible events, shifting them to the center.
shamanism
rain magic
gambling superstitions
some governmental
conspiracies
evil eye
lycanthropes
sorcery
(sorcerers)
Azande witch
Jewish world
conspiracy
(myth)
Knights Templar
(accusations)
Mau Mau rebels
(accusations)
witch who
uses magic
ON THE ORIGINS AND DESIGN OF WITCHES AND SORCERERS
Page 25
According to the theory outlined here, sorcerers are the result of both a selection for
intuitive magic and a selection for plausible explanations. The selection for intuitive magic
produces compelling techniques for controlling uncertain outcomes, including rain magic,
gambling superstitions, and magic aimed at harming others, or sorcery. Once people accept that
this magic is effective and that other people practice it, it becomes a plausible explanation for
misfortune. A person who feels threatened and who confronts unexplainable tragedy will easily
suspect that a rival has ensorcelled them. As people regularly consider how others harm them,
they build plausible portrayals of sorcerers.
Beliefs about werewolves, werebears, weresnakes, and other lycanthropes also develop
from a selection for plausible explanations. Baffled as to why an animal attacked them, a person
suspects a rival of becoming or possessing an animal and stalking them at night. This explanation
becomes more conceivable as the lycanthrope explains other strange events and as conceptions of
the lycanthrope become more plausible. Many societies ascribe transformative powers to other
malicious practitioners (see Table 3), showing that people also suspect existing practitioners after
attacks by wild animals.
Beliefs in the malignant power of stares and words likewise develop to explain
misfortune. As reviewed earlier, people around the world connect jealousy and envy to a desire to
induce harm. Thus, people who stare with envy or express a compliment are suspected of
harboring malice and an intention to harm. A person who suffers a misfortune remembers these
stares and suspects those people of somehow injuring them. In regularly inferring how envious
individuals attacked them, people craft a compelling notion of the evil eye.
Why suspect the evil eye rather than sorcery? There are at least two possibilities. First, an
accused individual may ardently vow not to know sorcery or to have attacked the target (see these
claims among the Azande, both described in text: Evans-Pritchard 1937:119-125; and shown in
film: Singer 1981, minute 21). Alternatively, given beliefs that effective sorcery requires powers
that develop with age, special knowledge, or certain experiences, it may seem unreasonable that a
young or unexperienced group mate effectively ensorcelled the target. In these instances, the idea
that the stare itself harmed the target may provide a more plausible mechanism.
The famous odious, powerful witch, I propose, arises when blamed malicious
practitioners become demonized. People who fear an invisible threat or who have an interest in
mistreating competitors benefit from demonizing the target, transforming them into a heinous,
threatening menace. Thus, witches represent a confluence of two and sometimes all three
cultural selective processes.
In Figure 1, I showed that beliefs about malicious practitioners exist along two
dimensions. The tripartite theory accounts for this structure. All of the practitioners displayed
are plausible explanations of how group mates inflict harm. One dimension (SORCERY-EVIL
EYE) distinguishes those explanations of misfortune that include magic (sorcerers) from those
that do not (evil eye, lycanthrope). The other dimension shows the extent to which different
practitioners have been demonized. In short, all beliefs about harmful practitioners are
explanations; sometimes they use magic, sometimes they’re made evil.
8.2. Ten predictions
ON THE ORIGINS AND DESIGN OF WITCHES AND SORCERERS
Page 26
The proposed theory generates many predictions for how shifting conditions should drive
changes in beliefs about malicious practitioners. I referred to several of these throughout the
paper. Here are ten (the section of the paper is noted when a prediction is discussed in the
paper):
1. People are more likely to believe in sorcerers as sorcery techniques become more
effective-seeming.
2. People are more likely to ascribe injury to mystical harm when they are distrustful of
others, persecuted, or otherwise convinced of harmful intent. (sect. 6.2.1)
3. The emotions attributed to malicious practitioners will be those that most intensely and
frequently motivate aggression. (sect. 6.2.1)
4. People are more likely to attribute injury to mystical harm when they lack alternative
explanations. (sect. 6.2.2)
5. The greater the impact of the misfortune, the more likely people are to attribute it to
mystical harm. (sect. 6.2.2)
6. Practitioners of mystical harm are more likely to become demonized during times of
stressful uncertainty.
7. The traits ascribed to malicious practitioners will become more heinous or sensational as
Condoners become more trustful or reliant on information from Campaigners.
8. Malicious practitioners will become less demonized when there is less disagreement or
resistance about their removal.
9. The traits that constitute demonization will be those that elicit the most punitive outrage,
controlling for believability. (sect. 7.2.1)
10. Malicious practitioners whose actions can more easily explain catastrophe, such as those
who employ killing magic compared to love magic, will be easier to demonize.
8.3. The cultural evolution of harmful beliefs
Social scientists, and especially those who study the origins of religion and belief, debate over
whether cultural traditions evolve to provide group-level benefits (Baumard and Boyer 2013;
Norenzayan et al. 2016). Reviving the analogy of society as an organism, some scholars maintain
that cultural traits develop to ensure the survival and reproduction of the group (Wilson 2002).
These writers argue that traditions that undermine societal success should normally be culled
away, while traditions that enhance group-level success should spread (Boyd and Richerson
2010).
In this paper, I have examined cultural traits with clear social costs: mystical harm beliefs.
As sources of paranoia, distrust, and bloodshed, these beliefs divide societies, breeding contempt
even among close family members. But I have explained them without invoking group-level
benefits. Focusing on people’s (usually automatic) decisions to adopt cultural traditions, I have
shown that beliefs in witches and sorcerers are maximally appealing, providing the most plausible
explanations and justifying hostile aims. Corrosive customs recur as long as they are useful and
cognitively appealing.
Acknowledgments
ON THE ORIGINS AND DESIGN OF WITCHES AND SORCERERS
Page 27
I thank Mia Charifson for research assistance and Steve Worthington at the Harvard Institute
for Quantitative Social Science for statistical help. Nicolas Baumard, Ronald Hutton, Graham
Jones, Ted Slingerland, Dylan Tweed, Max Winkler, Richard Wrangham, two anonymous
reviewers, and members of the Culture, Cognition, and Coevolution Lab at Harvard University
shared comments on earlier drafts of this paper. Luke Glowacki provided detailed feedback on
several versions of this manuscript; this paper and the ideas presented in it are much clearer as a
result of his incisive suggestions. This research was funded by a graduate research fellowship
from the National Science Foundation.
ON THE ORIGINS AND DESIGN OF WITCHES AND SORCERERS
Page 28
References
Abalakina-paap, Marina, Walter G Stephan, Traci Craig, and W Larry Gregory. 1999. “Beliefs
in Conspiracies.” Political Psychology 20 (3): 637–47.
Aginsky, B. W. 1939. “Population Control in the Shanel (Pomo) Tribe.” American Sociological
Review 4 (2): 209–16.
Alm, Torbjørn. 2003. “The Witch Trials of Finnmark, Northern Norway, during the 17th
Century: Evidence for Ergotism as a Contributing Factor.” Economic Botany 57 (3): 403–16.
doi:10.1663/0013-0001(2003)057[0403:TWTOFN]2.0.CO;2.
Ames, David. 1959. “Belief in ‘witches’ among the Rural Wolof of the Gambia.” Africa 29 (3):
263–73.
Ammār, Ḥāmid. 1954. Growing up in an Egyptian Village: Silwa, Province Of Aswan. London:
Routledge and Kegan Paul Ltd.
Apicella, Coren L., Paul Rozin, Justin T. A. Busch, Rachel E. Watson-Jones, and Cristine H.
Legare. 2018. “Evidence from Hunter-Gatherer and Subsistence Agricultural Populations
for the Universality of Contagion Sensitivity.” Evolution and Human Behavior.
papers2://publication/uuid/0545B852-08F3-4979-848D-FFD24E70B4D1.
Archer, William George. 1974. The Hill of Flutes: Life, Love, and Poetry in Tribal India : A
Portrait of the Santals. Pittsburgh: University of Pittsburgh Press.
———. 1984. Tribal Law And Justice: A Report on the Santal. New Delhi: Concept.
Barber, Malcolm. 2006. The Trial of the Templars. 2nd ed. Cambridge, UK: Cambridge
University Press.
Barkun, Michael. 2013. A Culture of Conspiracy: Apocalyptic Visions in Contemporary America. 2nd
Editio. Berkeley and Los Angeles: University of California Press.
Basso, Keith H. 1969. Western Apache Witchcraft. Tuscon: The University of Arizona Press.
Baumard, Nicolas, and Pascal Boyer. 2013. “Explaining Moral Religions.” Trends in Cognitive
Sciences 17 (6). Elsevier Ltd: 272–80. doi:10.1016/j.tics.2013.04.003.
Beattie, John. 1963. “Sorcery in Bunyoro.” In Witchcraft and Sorcery in East Africa, edited by John
Middleton and E. H. Winter, 27–55. London: Routledge & Paul.
Beidelman, T. O. 1963. “Witchcraft in Ukaguru.” In Witchcraft and Sorcery in East Africa, edited
by John Middleton and E. H. Winter, 57–98. London: Routledge & Paul.
———. 1975. “Ambiguous Animals: Two Theriomorphic Metaphors in Kaguru Folklore.”
Africa 45 (2): 183–200.
Berry, Marie. 2015. “From Violence to Mobilization: Women, War, and Threat in Rwanda.”
Mobilization: An International Quarterly 20 (2): 135–56.
Blackmore, Susan. 1999. The Meme Machine. Oxford, UK: Oxford University Press.
Bohannan, Paul. 1958. “Extra-Processual Events in Tiv Political Institutions.” American
Anthropologist 60 (1): 1–12.
Boyd, Robert, and Peter J. Richerson. 1985. Culture and the Evolutionary Process. Chicago:
University of Chicago Press.
———. 1988. “An Evolutionary Model of Social Learning: The Effects of Spatial and Temporal
Variation.” In Social Learning: Psychological and Biological Perspectives, edited by Thomas R.
Zentall and Bennett G. Galef. Hillsdale, NJ: Lawrence Erlbaum Associates.
———. 2010. “Transmission Coupling Mechanisms: Cultural Group Selection.” Philosophical
ON THE ORIGINS AND DESIGN OF WITCHES AND SORCERERS
Page 29
Transactions of the Royal Society of London. Series B, Biological Sciences 365 (1559): 3787–95.
doi:10.1098/rstb.2010.0046.
Boyer, Pascal. 2001. Religion Explained: The Evolutionary Origins of Religious Thought. New
York: Basic Books.
Boyer, Paul, and Stephen Nissenbaum. 1974. Salem Possessed: The Social Origins of Witchcraft.
Cambridge, MA.
Bryant, A. T. 1929. Olden Times in Zululand and Natal. London: Longmans, Green and Co.
Budge, E. A. Wallis. 1901. Egyptian Magic. London: Kegan, Paul, Trech and Trübner & Co.
https://archive.org/details/EgyptianMagic_51.
Campbell, Donald T. 1965. “Variation and Selective Retention in Socio-Cultural Evolution.” In
Social Change in Developing Areas: A Reinterpretation of Evolutionary Theory, edited by
Herbert R. Barringer, George Irving Blanksten, and Raymond Wright Mack, 19–49.
Cambridge, Massachusetts: Schenkman Publishing Company.
Caporael, Linnda R. 1976. “Ergotism: The Satan Loosed in Salem?” Science 192 (4234): 21–26.
Chaudhuri, Soma. 2012. “Women as Easy Scapegoats: Witchcraft Accusations and Women as
Targets in Tea Plantations of India.” Violence Against Women 18 (10): 1213–34.
doi:10.1177/1077801212465155.
Cohen, Stanley. 1972. Folk Devils and Moral Panics: The Creation of the Mods and Rockers.
London: MacGibbon and Kee.
Cohn, Norman. 1966. “The Myth of the Jewish World-Conspiracy: A Case Study in Collective
Psychopathology.” Commentary 41 (6): 35–42.
———. 1967. Warrant for Genocide: The Myth of the Jewish World-Conspiracy and the Protocols of
the Elders of Zion. New York and Evanston: Harper & Row.
———. 1976. Europe’s Inner Demons. Frogmore: Paladin.
Comaroff, Jean, and John L. Comaroff. 1999. “Occult Economies and the Violence of
Abstraction: Notes from the South African Postcolony.” American Ethnologist 26 (2): 279–
303.
Cooper, Milton William. 1991. Behold a Pale Horse. Sedona, AZ: Light Technology Publishing.
https://archive.org/details/WilliamCooperBeholdAPaleHorse1991.
Crawford, Jane. 1963. “Evidences for Witchcraft in Anglo-Saxon England.” Medium Ævum 32
(2): 99–116.
Crick, Malcolm. 1973. “Two Styles in the Study of Witchcraft.” Journal of the Anthropological
Society of Oxford 4: 17–31.
De Laguna, Frederica. 1972. Under Mount Saint Elias: The History and Culture of the Yakutat
Tlingit, Vol. 2. Washington, D.C.: Smithsonian Institution Press.
Debrunner, Hans W. 1961. Witchcraft in Ghana: A Study on the Belief in Destructive Witches and
Its Effect on the Akan Tribes. Accra: Presbyterian Book Depot Ltd.
Desforges, Alison. 1999. Leave None to Tell the Story: Genocide in Rwanda. New York: Human
Rights Watch.
Dundes, Alan, ed. 1992. The Evil Eye: A Casebook. Madison: The University of Wisconsin Press.
Emmons, George Thornton, and Frederica De Laguna. 1991. The Tlingit Indians.
Anthropological Papers of the American Museum of Natural History. Vol. 70. Seattle: University
of Washington Press and the American Museum of Natural History.
http://digitallibrary.amnh.org/handle/2246/253.
Evans-Pritchard, E. E. 1937. Witchcraft, Oracles, and Magic among the Azande. Oxford, UK:
ON THE ORIGINS AND DESIGN OF WITCHES AND SORCERERS
Page 30
Clarendon Press.
Faulkingham, Ralph Harold. 1971. “Political Support in a Hausa Village.” Michigan State
University.
Felix, Minucius, and Gerald H. Rendall. 1972. “Octavius.” In Tertullian, Minucius Felix, edited
by T. R. Glover and Gerald H. Rendall, 303–437. Cambridge: Harvard University Press.
https://ryanfb.github.io/loebolus-data/L250.pdf.
Field, Margaret Joyce. 1970. Search for Security: An Ethno-Psychiatric Study of Rural Ghana. New
York: W. W. Norton & Company.
Firth, Raymond. 1954. “The Sociology of ‘Magic’ in Tikopia.” Sociologus 4 (2): 97–116.
Frazer, James George. 1920. The Golden Bough: A Study in Magic and Religion, Vol. 1 (The Magic
Art and the Evolution of Kings, Vol. I). 2nd editio. London: Macmillan and Co.
https://archive.org/details/TheGoldenBough-Part1-
TheMagicArtAndTheEvolutionOfKingsVol.1.
Frazier, Brandy N., Susan A. Gelman, and Henry M. Wellman. 2009. “Preschoolers’ Search for
Explanatory Information within Adult-Child Conversation.” Child Development 80 (6):
1592–1611.
Galef, Bennett G., Kristina E. Dudley, and Elaine E. Whiskin. 2008. “Social Learning of Food
Preferences in ‘dissatisfied’ and ‘Uncertain’ Norway Rats.” Animal Behaviour 75 (2): 631–37.
doi:10.1016/j.anbehav.2007.06.024.
Gershman, Boris. 2016. “Witchcraft Beliefs and the Erosion of Social Capital: Evidence from
Sub-Saharan Africa and Beyond.” Journal of Development Economics 120. Elsevier B.V.:
182–208. doi:10.1016/j.jdeveco.2015.11.005.
Goode, Erich. 2008. “Moral Panics and Disproportionality: The Case of LSD Use in the
Sixties.” Deviant Behavior 29 (6): 533–43. doi:10.1080/01639620701839377.
Goode, Erich, and Nachman Ben-Yehuda. 2009. Moral Panics: The Social Construction of
Deviance. 2nd ed. Wiley-Blackwell.
Gufler, H. 1999. “Witchcraft Beliefs among the Yamba (Cameroon).” Anthropos 94 (1–3): 181–
98. doi:10.2307/40465701.
Gusinde, Martin. 1971. The Fireland Indians, Vol. 1: The Selk’nam, on the Life and Thought of a
Hunting People of the Great Island of Tierra Del Fuego. New Haven: Human Relations Area
Files.
Haidt, Jonathan, Fredrik Björklund, and Scott Murphy. 2000. “Moral Dumbfounding: When
Intuition Finds No Reason.”
http://faculty.virginia.edu/haidtlab/articles/manuscripts/haidt.bjorklund.working-
paper.when intuition finds no reason.pub603.doc.
Harman, Gilbert H. 1965. “The Inference to the Best Explanation.” The Philosophical Review 74
(1): 88–95.
Henrich, Joseph. 2015. The Secret of Our Success: How Culture Is Driving Human Evolution,
Domesticating Our Species, and Making Us Smarter. Princeton, NJ: Princeton University
Press.
Henry, Adam Hughes. 2014. “Polluting the Waters: A Brief History of Anti-Communist
Propaganda during the Indonesian Massacres.” Genocide Studies International 8 (2): 153–75.
doi:10.3138/gsi.8.2.03.
Hester, Marianne. 1992. Lewd Women and Wicked Witches: A Study of the Dynamics of Male
Domination. New York and London: Routledge.
ON THE ORIGINS AND DESIGN OF WITCHES AND SORCERERS
Page 31
Hoffman, Moshe, Erez Yoeli, and Martin A. Nowak. 2015. “Cooperate without Looking: Why
We Care What People Think and Not Just What They Do.” Proceedings of the National
Academy of Sciences 112 (6): 1727–32. doi:10.1073/pnas.1417904112.
Hofstadter, Richard. 1964. “The Paranoid Style in American Politics.” Harper’s Magazine.
Hogbin, H. Ian. 1938. “Social Reaction to Crime: Law and Morals in the Schouten Islands,
New Guinea.” The Journal of the Royal Anthropological Institute of Great Britain and Ireland
68: 223–62.
Hutton, Ronald. 2004. “Anthropological and Historical Approaches To Witchcraft: Potential
for a New Collaboration?” The Historical Journal 47 (2): 413–34.
doi:10.1017/S0018246X03003558.
———. 2017. The Witch: A History of Fear, from Ancient Times Ot the Present. New Haven and
London: Yale University Press.
Ivanow, W. 1926. “Muhammadan Child-Killing Demons.” Man 26: 195–99.
Johnson, Dominic D. P., Daniel T Blumstein, James H Fowler, and Martie G Haselton. 2013.
“The Evolution of Error: Error Management, Cognitive Constraints, and Adaptive
Decision-Making Biases.” Trends in Ecology & Evolution 28 (8). Elsevier Ltd: 474–81.
doi:10.1016/j.tree.2013.05.014.
Johnson, Erik W., and Scott Frickel. 2011. “Ecological Threat and the Founding of U.S.
National Environmental Movement Organizations, 1962-1998.” Social Problems 58 (3):
305–29. doi:10.1525/sp.2011.58.3.305.306.
Jones, William. 1939. Ethnography of the Fox Indians. Washington: Government Printing Office.
Karlsen, Carol F. 1987. The Devil in the Shape of a Woman: Witchcraft in Colonial New England.
New York: W. W. Norton & Company.
Karsten, Rafael. 1955. The Religion of the Samke: Ancient Beliefs and Cults of the Scandinavian and
Finnish Lapps. Leiden: E.J. Brill. http://ehrafworldcultures.yale.edu/document?id=ep04-
005.
Keinan, G. 2002. “The Effects of Stress and Desire for Control on Superstitious Behavior.”
Personality and Social Psychology Bulletin 28 (1): 102–8. doi:10.1177/0146167202281009.
Kluckhohn, Clyde. 1944. Navaho Witchcraft.
———. 1959. “Recurrent Themes in Myths and Mythmaking.” Daedalus 88 (2): 268–79.
Knauft, Bruce. 2010. The Gebusi: Lives Transformed in a Rainforest World. 2nd Editio. New York:
McGraw Hill.
Kunda, Ziva. 1990. “The Case for Motivated Reasoning.” Psychological Bulletin 108 (3): 480–98.
doi:10.1037/0033-2909.108.3.480.
Lagae, C. R. 1999. The Azande or Niam-Niam: Zande Organizations, Religious and Magical
Beliefs, Family Customs. New Haven: HRAF.
Laland, Kevin N. 2004. “Social Learning Strategies.” Learning & Behaviour 32 (1): 4–14.
Lederman, Rena. 1981. “Sorcery and Social Change in Mendi.” Social Analysis 8: 15–27.
Legare, Cristine H., and André L. Souza. 2012. “Evaluating Ritual Efficacy: Evidence from the
Supernatural.” Cognition 124 (1): 1–15. doi:10.1016/j.cognition.2012.03.004.
Lombrozo, Tania. 2006. “The Structure and Function of Explanations.” Trends in Cognitive
Sciences 10 (10): 464–70. doi:10.1016/j.tics.2006.08.004.
Lonsdale, John. 1990. “Mau Maus of the Mind: Making Mau Mau and Remaking Kenya.” The
Journal of African History 31 (3): 393–421.
Mace, Ruth, Matthew G. Thomas, Jiajia Wu, QiaoQiao He, Ting Ji, and Yi Tao. 2018.
ON THE ORIGINS AND DESIGN OF WITCHES AND SORCERERS
Page 32
“Population Structured by Witchcraft Beliefs.” Nature Human Behaviour 2 (1). Springer US:
39–44. doi:10.1038/s41562-017-0271-6.
Maher, Thomas V. 2010. “Threat, Resistance, and Collective Action: The Cases of Sobibór,
Treblinka, and Auschwitz.” American Sociological Review 75 (2): 252–72.
doi:10.1177/0003122410365305.
Mair, Lucy. 1969. Witchcraft.
Malinowski, Bronisław. 1922. Argonauts of the Western Pacific: An Account of Native Enterprise and
Adventure in the Archipelagoes of Melanesian New Guinea. London: George Routledge &
Sons, Ltd.
———. 1948. “Magic, Science, and Religion.” In Magic, Science and Religion, and Other Essays,
17–92. Garden City, NY: Doubleday Anchor Books.
Marshall, Lorna. 1962. !“!Kung Bushman Religious Beliefs.” Africa 32 (3): 221–52.
Mashuri, Ali, and Esti Zaduqisti. 2015. “The Effect of Intergroup Threat and Social Identity
Salience on the Belief in Conspiracy Theories over Terrorism in Indonesia: Collective
Angst as a Mediator.” International Journal of Psychological Research 8 (1): 24–35.
Massé, Henri, and Charles A. Messner. 1954. Persian Beliefs and Customs. New Haven: Human
Relations Area Files.
McCauley, Clark, and Susan Jacques. 1979. “The Popularity of Conspiracy Theories of
Presidential Assassination: A Bayesian Analysis.” Journal of Personality and Social Psychology
37 (5): 637–44. doi:10.1037/0022-3514.37.5.637.
McKay, Ryan, and Charles Efferson. 2010. “The Subtleties of Error Management.” Evolution
and Human Behavior 31 (5). Elsevier Inc.: 309–19.
doi:10.1016/j.evolhumbehav.2010.04.005.
Mercier, Hugo, and Dan Sperber. 2011. “Why Do Humans Reason? Arguments for an
Argumentative Theory.” Behavioral and Brain Sciences 34 (2): 57-74; discussion 74-111.
doi:10.1017/S0140525X10000968.
Miceli, Maria, and Cristiano Castelfranchi. 2007. “The Envious Mind.” Cognition and Emotion
21 (3): 449–79. doi:10.1080/02699930600814735.
Mirowsky, John, and Catherine E Ross. 1983. “Paranoia and the Structure of Powerlessness.”
American Sociological Review 48 (2): 228–39.
Morgan, Thomas J. H., Luke E. Rendell, Micael Ehn, William Hoppitt, and Kevin N. Laland.
2012. “The Evolutionary Basis of Human Social Learning.” Proceedings of the Royal Society.
Series B, Biological Sciences 279 (1729): 653–62. doi:10.1098/rspb.2011.1172.
Munro, Neil Gordon. 1963. Ainu Creed and Cult. New York: Columbia University Press.
Murray, Margaret Alice. 1921. The Witch Cult in Western Europe. Oxford: Clarendon Press.
Nadel, Siegfried Frederick. 1954. Nupe Religion. London: Routledge & Paul.
Natrella, Kayla Theresa. 2014. “Witchcraft and Women: A Historiography of Witchcraft as
Gender History.” Binghamton Journal of History 15.
https://www.binghamton.edu/history/resources/journal-of-history/k-natrella.pdf.
Needham, Rodney. 1978. Primordial Characters. Charlottesville: University Press of Virginia.
Nemeroff, Carol, and Paul Rozin. 2000. “The Makings of the Magical Mind: The Nature and
Function of Sympathetical Magical Thinking.” In Imagining the Impossible: Magical,
Scientific, and Religious Thinking in Children, 1–34. doi:10.1017/CBO9780511571381.002.
Nickerson, RS. 1998. “Confirmation Bias: A Ubiquitous Phenomenon in Many Guises.” Review
of General Psychology 2 (2): 175–220. doi:10.1037/1089-2680.2.2.175.
ON THE ORIGINS AND DESIGN OF WITCHES AND SORCERERS
Page 33
Norenzayan, Ara, Azim F. Shariff, Will M. Gervais, Aiyana K. Willard, Rita A. McNamara,
Edward Slingerland, and Joseph Henrich. 2016. “The Cultural Evolution of Prosocial
Religions.” Behavioral and Brain Sciences 39: e1. doi:10.1017/S0140525X14001356.
Oliphant, Samuel Grant. 1913. “The Story of the Strix: Ancient.” Transactions and Proceedings of
the American Philological Association 44: 133–49.
———. 1914. “The Story of the Strix: Isidorus and the Glossographers.” Transactions and
Proceedings of the American Philological Association 44: 49–63.
Oliver, Douglas L. 1955. A Solomon Island Society: Kinship And Leadership among the Siuai Of
Bougainville. Harvard University Press.
Ono, Koichi. 1987. “Superstitious Behavior in Humans.” Journal of the Experimental Analysis of
Behavior 47 (3): 261–71. doi:10.1901/jeab.1987.47-261.
Oster, Emily. 2004. “Witchcraft, Weather and Economic Growth in Renaissance Europe.”
Journal of Economic Perspectives 18 (1): 215–28.
Parpola, Marjatta. 2000. Kerala Brahmins in Transition: A Study of a Namputiri Family. Helsinki:
Finnish Oriental Society.
Pilling, Arnold Remington. 1958. “Law and Feud in an Aboriginal Society of North Australia.”
University of California, Berkeley.
Pospisil, Leopold J. 1958. Kapauku Papuans and Their Law. New Haven: Yale University:
Department of Anthropology.
Raihani, Nichola J, and Vaughan Bell. 2017. “Paranoia and the Social Representation of Others:
A Large-Scale Game Theory Approach.” Scientific Reports 7. Springer US: 4544.
doi:10.1038/s41598-017-04805-3.
———. 2018. “An Evolutionary Perspective on Paranoia.” Nature Human Behaviour. Springer
US. doi:10.1038/s41562-018-0495-0.
Reichel-Dolmatoff, Gerardo. 1971. Amazonian Cosmos: The Sexual and Religious Symbolism of the
Tukano Indians. Chicago: University of Chicago Press.
———. 1976. “Training for the Priesthood among the Kogi of Colombia.” In Enculturation in
Latin America: An Anthology, 265–88. Los Angeles: UCLA Latin American Center
Publications.
———. 1997. The Kogi: A Tribe of the Sierra Nevada de Santa Marta, Colombia, Vol. 1. New
Haven: HRAF.
Reynolds, Barrie. 1963. Magic, Divination and Witchcraft among the Barotse of Northern Rhodesia.
London: Chatto and Windus.
Richards, Audrey I. 1935. “A Modern Movement of Witch-Finders.” Africa 8 (4): 448–61.
Rosnow, Ralph L. 1991. “Inside Rumor: A Personal Journey.” American Psychologist 46 (5): 484–
96.
Rothschild, Zachary K, Mark J Landau, Daniel Sullivan, and Lucas A Keefer. 2012. “A Dual-
Motive Model of Scapegoating : Displacing Blame to Reduce Guilt or Increase Control”
102 (6): 1148–63. doi:10.1037/a0027413.
Rozin, Paul, Linda Millman, and Carol Nemeroff. 1986. “Operation of the Laws of Sympathetic
Magic in Disgust and Other Domains.” Journal of Personality and Social Psychology 50 (4):
703–12.
Rozin, Paul, and Carol Nemeroff. 2002. “Sympathetic Magical Thinking: The Contagion and
Similarity ‘Heuristics.’” In Heuristics and Biases: The Psychology of Intuitive Judgment, edited
by Thomas Gilovich, Dale Griffin, and Daniel Kahneman, 201–16. Cambridge, UK:
ON THE ORIGINS AND DESIGN OF WITCHES AND SORCERERS
Page 34
Cambridge University Press. doi:10.1017/CBO9780511808098.013.
Saalfeld, Vanessa, Zeina Ramadan, Vaughan Bell, and Nichola J. Raihani. 2018. “Differences in
Social Rank and Political Affiliation Encourage Paranoid Attributions.”
https://psyarxiv.com/jxkv3/.
Sanders, Andrew. 1995. A Deed without a Name: The Witch in Society and History. Oxford and
Washington DC: Berg Publishers.
Schapera, Isaac. 1952. “Sorcery and Witchcraft in Bechaunaland.” African Affairs 51 (202): 41–
52.
Schnall, Simone, Jonathan Haidt, Gerald L. Clore, and Alexander H. Jordan. 2008. “Disgust as
Embodied Moral Judgment.” Personality and Social Psychology Bulletin 34 (8): 1096–1109.
doi:10.1177/0146167208317771.
Singer, Andre. 1981. Witchcraft among the Azande. England: Royal Anthropological Institute.
Singh, Manvir. 2018. “The Cultural Evolution of Shamanism.” Behavioral and Brain Sciences 41:
e66. doi:10.1017/S0140525X17001893.
Skaria, Ajay. 1997. “Women, Witchcraft and Gratuitous Violence in Colonial Western India.”
Past & Present 155: 109–41.
Smith, Richard H., and Sung Hee Kim. 2007. “Comprehending Envy.” Psychological Bulletin 133
(1): 46–64. doi:10.1037/0033-2909.133.1.46.
Smith, Richard H., Terence J. Turner, Ron Garonzik, Colin W. Leach, Vanessa Urch-Druskat,
and Christine M. Weston. 1996. “Envy and Schadenfreude.” Personality and Social
Psychology Bulletin 22 (2): 158–68. doi:10.1177/0146167296222005.
Sperber, Dan. 1996. Explaining Culture: A Naturalistic Approach. Oxford, UK: Blackwell
Publishers Ltd.
Sperber, Dan, Fabrice Clément, Christophe Heintz, Olivier Mascaro, Hugo Mercier, Gloria
Origgi, and Deirdre Wilson. 2010. “Epistemic Vigilance.” Mind and Language 25 (4): 359–
93. doi:10.1111/j.1468-0017.2010.01394.x.
Stern, Theodore. 1965. The Klamath Tribe: A People and Their Reservation. Seattle: University of
Washington Press.
Sullivan, Daniel, Mark J Landau, Nyla R Branscombe, and Zachary K Rothschild. 2012.
“Competitive Victimhood as a Response to Accusations of Ingroup Harm Doing.” Journal
of Personality and Social Psychology 102 (4): 778–95. doi:10.1037/a0026573.
Talayesva, Don C., and Leo William Simmons. 1942. Sun Chief: The Autobiography of a Hopi
Indian. New Haven: Yale University Press.
Tennen, Howard, and Glenn Affleck. 1990. “Blaming Others for Threatening Events.”
Psychological Bulletin 108 (2): 209–32. doi:10.1037/0033-2909.108.2.209.
Tetlock, Philip E. 2003. “Thinking the Unthinkable: Sacred Values and Taboo Cognitions.”
Trends in Cognitive Sciences 7 (7): 320–24. doi:10.1016/S1364-6613(03)00135-9.
Turner, Victor W. 1964. “Witchcraft and Sorcery: Taxonomy versus Dynamics.” Africa 34 (4):
314–25.
van Bergen, Yfke, Isabelle Coolen, and Kevin N Laland. 2004. “Nine-Spined Sticklebacks
Exploit the Most Reliable Source When Public and Private Information Conflict.”
Proceedings. Biological Sciences / The Royal Society 271 (1542): 957–62.
doi:10.1098/rspb.2004.2684.
van de Ven, Niels, Charles E. Hoogland, Richard H. Smith, Wilco W. van Dijk, Seger M.
Breugelmans, and Marcel Zeelenberg. 2015. “When Envy Leads to Schadenfreude.”
ON THE ORIGINS AND DESIGN OF WITCHES AND SORCERERS
Page 35
Cognition and Emotion 29 (6). Taylor & Francis: 1007–25.
doi:10.1080/02699931.2014.961903.
van Prooijen, Jan-Willem, and Karen M. Douglas. 2017. “Conspiracy Theories as Part of
History: The Role of Societal Crisis Situations.” Memory Studies 10 (3): 323–33.
doi:10.1177/1750698017701615.
van Prooijen, Jan-Willem, and Nils B. Jostmann. 2013. “Belief in Conspiracy Theories: The
Influence of Uncertainty and Perceived Morality.” European Journal of Social Psychology 43
(1): 109–15. doi:10.1002/ejsp.1922.
van Prooijen, Jan-Willem, and Eric van Dijk. 2014. “When Consequence Size Predicts Belief in
Conspiracy Theories: The Moderating Role of Perspective Taking.” Journal of Experimental
Social Psychology 55. Elsevier Inc.: 63–73. doi:10.1016/j.jesp.2014.06.006.
Victor, Jeffrey S. 1989. “A Rumor-Panic about a Dangerous Satanic Cult in Western New
York.” New York Folklore 15 (1–2): 23–49.
Vyse, Stuart. 2014. Believing in Magic: The Psychology of Superstition. Oxford and New York:
Oxford University Press.
Walker, Jr., Deward E. 1967. “Nez Perce Sorcery.” Ethnology 6 (1): 66–96.
Wallace, Ernest, and E. Adamson Hoebel. 1952. The Comanches: Lords of the South Plains.
Norman: University of Oklahoma Press.
Wallace, William J., and Edith S. Taylor. 1950. “Hupa Sorcery.” Southwestern Journal of
Anthropology 6 (2): 188–96.
Walraven, B.C.A. 1980. “The Social Significance of Sorcery and Sorcery Accusations in Korea.”
Asiatische Studien: Zeitschrift Der Schweizerischen Asiengesellschaft 34 (2): 69–90.
Warner, W. Lloyd. 1958. A Black Civilization: A Social Study of an Australian Tribe. Harper &
Brothers.
Weltfish, Gene. 1965. The Lost Universe: With a Closing Chapter on “The Universe Regained.”
New York and London: Basic Books.
Whiting, Beatrice Blyth. 1950. Paiute Sorcery. New York: The Viking Fund Inc.
Wieringa, Saskia Eleonora. 2011. “Sexual Slander and the 1965/66 Mass Killings in Indonesia:
Political and Methodological Considerations.” Journal of Contemporary Asia 41 (4): 544–65.
doi:10.1080/00472336.2011.610613.
Wilbert, Johannes. 1972. Survivors of Eldorado: Four Indian Cultures of South America. New York:
Praeger Publishers.
Wilson, David Sloan. 2002. Darwin’s Cathedral: Evolution, Religion, and the Nature of Society.
Chicago: University of Chicago Press.
Wilson, Monica Hunter. 1951. “Witch Beliefs and Social Structure.” American Journal of
Sociology 56 (4): 307–13.
Wohl, Michael, and Nyla R. Branscombe. 2009. “Group Threat, Collective Angst, and Ingroup
Forgiveness for the War in Iraq.” Political Psychology 30 (2): 193–217. doi:10.1111/j.1467-
9221.2008.00688.x.
S1
Magic, explanations, and evil
On the origins and design of witches and sorcerers
SUPPLEMENTARY MATERIAL
Manvir Singh
Last updated: 13 March 2020
1. The Mystical Harm Survey (MHS)
1.1. Background
The Mystical Harm Survey (MHS) is designed to characterize patterns in beliefs about
practitioners of mystical harm across societies. It covers ethnographic texts from the 60 societies
comprising the Probability Sample File of the electronic Human Relations Area Files, a pseudo-
random sample of well-documented human societies designed to make inferences about
humanity more generally (Human Relations Area Files 1967; Naroll 1967). For each society, I
selected the two ethnographic texts containing the most paragraphs tagged for the code
SORCERY (USE code 754)
1
. Whenever the documents included a total of less than twenty
paragraphs, I included all ethnographic texts with paragraphs tagged for SORCERY until at
least twenty paragraphs were covered or, if that was not possible, until all of the ethnographic
texts tagged for SORCERY in a given culture were included.
Two independent coders read through the tagged paragraphs for each society, identified
the different practitioners of mystical harm discussed, and coded each practitioner for 58
features. Discrepancies between the two resulting datasets were identified and resolved through
discussion to produce a final, merged dataset, available at osf.io/492mj. Supplementary Table 1
displays the societies, ethnographic documents, and practitioner IDs.
1.2. Inclusion criteria
1
According to the electronic Human Area Files, code 754 (SORCERY) includes any reference to
the following: “Ideas of the causation of disease and death through witchcraft and sorcery; actual and
reputed prevalence of sorcery; motives for practicing sorcery; methods (e.g., bone pointing,
manipulation of effigies, exuvial magic, invocation of spirit aids); employment of sorcerers; witches,
wizards, and sorcerers; physical, social and mental characteristics; sources of power; training;
organization; special types of sorcerers (e.g., werewolves and other were-animals, vampires,
individuals with the evil eye); evidence as to the efficacy of sorcery; reactions to sorcerers (e.g., witch
hunts); etc.”
S2
Each row of the dataset corresponds with a conception of a practitioner (or practice) of mystical
harm. The original criteria for a practitioner of mystical harm were as follows (note that this
passage has been edited slightly to make it more readable):
People who are believed to use magical or supernatural powers to attack non-strangers: Individuals,
either in-group or people with whom individuals otherwise frequently interact, believed to harm people
they know through magic (e.g., recited spells, magical poisons, charms) or supernatural powers (e.g.,
becoming a spirit and eating people; transforming into animals and attacking people; harming people
with thoughts or stares). This excludes beliefs about supernatural attackers who are strangers (e.g., beliefs
that individuals from far-away lands transform into bears and hassle travelers). This also excludes
people who only attack out-group members (e.g., shamans who only attack members of other groups).
Each row of the dataset refers to a different practitioner of mystical harm. Some societies have
several such practitioners. For example, an ethnographer might describe one kind of person who can
become an animal at night and a different kind of person who transmits harm through stares. In this
instance, each practitioner is coded as a separate entry in the dataset. In another instance, an
ethnographer may describe a single practitioner in a society, although that practitioner may have many
abilities; for example, they may report that people believe that some individuals can become animals, fly,
and attack with magical spells. In this instance, a single practitioner will appear in the dataset for that
society.
Ethnographers may describe a technique or practice – for example, “black magic” — without
connecting it to a certain practitioner. In these instances, rows in the dataset correspond only to the
practice (when appropriate, variables CLASS01 and CLASS02 will add clarifying information).
Public magicians (e.g., shamans, priests, other magicians) are often said to have malignant,
mystical powers. Whenever this is the case, coding is as follows:
- Whenever an ethnographer reports that a practitioner class includes, but is not restricted to,
public magicians, then public magicians are not coded separately. For example, shamans are not
coded separately if an ethnographer states that people believe that individuals in their group are
witches and that shamans are sometimes accused as witches.
- Whenever an ethnographer reports that public magicians attack non-strangers with mystical
powers and that they differ from other practitioner types on the basis of one or more variables,
then they are coded separately. For example, shamans are coded separately if an ethnographer
states that all witches harm people, but shaman-witches alone can fly and transform into
animals.
Two additional exclusion criteria were added while resolving discrepancies:
Exclude: Spells and curses that are used to enforce contracts or promises
Exclude: Mystical harm that is considered to be “good magic”, such as judiciary magic
1.3. Citations
Any coding decision that reports the presence of some trait includes a citation in the format
refX:Y, where X refers to the number of the document (for example, reference 1, reference 2, and
S3
so on; the title and author of the document appear in the same row) and Y refers to the page
number.
2. Analyses
2.1. Logistic PCA
All statistical analyses were conducted in R (R Core Team 2015). Because the data are binary, I
conducted a logistic PCA, using the logisticPCA function of the logisticPCA package (Landgraf
and Lee 2015). Following Landgraf (2016), I used the cv.lpca function to choose m, an argument
in the logisticPCA function denoting the natural parameters from the saturated model.
I ran the model 10 times, varying k from 1 to 10, and produced a scree plot (see
Supplementary Figure 1). The scree plot showed a break between the second and third
components, justifying a cutoff either at two or three dimensions.
For the reported analyses, I excluded practitioner classes that are exclusively leaders (e.g.,
sheikhs, elders) and public magicians (e.g., shamans, priests) – that is, any practitioners coded 1
for the variable BEHA18. Note that the principal components are very similar regardless of
whether these practitioner classes are included (compare the factor loadings in Supplementary
Table 2 with those in Supplementary Table 4). I removed all free-response variables and
transformed the following categorical variables into binary variables:
TECH10: 1 & 2 --> 1
[“always unintentional” and “sometimes unintentional” coded as “unintentional”]
PROC01: 1 & 2 --> 1
[“biological heredity” and “non-biological heredity” coded as “heredity”]
BEHA01: 1 & 2 --> 1
[“devour flesh” and “devour souls” coded as “cannibalism”]
BEHA14: 1, 2, & 3 --> 1; 4 --> 0
[“harm family members for enjoyment”, “harm family members as obligation”, and “harm family members as
consequence of harm” coded as “harm family members”; “harm family members for other reasons” coded as absence]
BEHA20: 1 & 2 --> 1
[“political leaders” and “household heads, elder lineages, generation leaders” coded as “leaders”]
I also created two new variables from the categorical variable SEX denoting whether the
given practitioner class is mostly or exclusively female (SEX1) or mostly or exclusively male
(SEX2). I binned each practice or practitioner class into a superordinate category (e.g., “evil eye”,
“witch”) based on the ethnographer’s translation (NAME02) and term (NAME03). The binning
decisions are recorded under the variable NAME04 in the dataset.
The first two dimensions resulting from the logistic PCA are interpretable; the third
dimension is not (see Supplementary Tables 2 and 3). I thus reported the first two dimensions
and their corresponding loadings in the main text. I did not report loadings for any factor whose
S4
loading was driven entirely by a single geographic region (see next section and Supplementary
Tables 5 and 6). Figure 1 in the main text shows the 83 practitioners plotted on the two principal
components.
2.2. Testing for geographic dependencies
Some human societies are more similar to others, such as because they share a cultural history or
because beliefs and practices have diffused from one to the other. Comparative analyses that
investigate cross-cultural patterns must ensure that these non-independencies do not influence
results and bias interpretations.
As I mentioned, I tried to minimize dependencies in the dataset by coding the
Probability Sample File of the electronic Human Relations Area Files, a body of high-quality
ethnography covering sixty societies that were chosen to represent human diversity and minimize
similarities that might result from cultural diffusion or shared cultural history. But dependencies
persist in the PSF, so I conducted two additional analyses to test whether geographic patterns in
particular biased the results. First, I conducted an omnibus F-test to evaluate whether regions as
defined by HRAF (e.g., Africa, Asia) have different PC scores on average. I found no evidence
that they do (F14,150 = 1.66, p = 0.070). Second, I re-ran the logistic PCA eight times, in each
instance removing all of the data-points in a single region, and compared the resulting factor
matrices. I found that the eight PCAs produced very similar components, despite using a
substantially reduced dataset in some instances (removing Africa, for example, reduced the
sample from 83 to 54 observations) (Supplementary Tables 5 and 6). Still, this analysis revealed
that several variables, especially those with very few observations, were unstable across analyses,
often with a factor loading collapsing when the observations for a single region were removed. I
have shaded those loadings in Supplementary Tables 2 and 4 and refrained from reporting them
in my description of the components in the main text.
2.3. Testing for the stability of the principal components
There is substantial disagreement about the appropriate sample size for principal component
analysis, but I conducted several analyses to test whether the reported components are stable:
1. I produced jackknife estimates of the proportion of variance explained by the first two
dimensions. The resampling procedure produced means nearly identical to the values
reported above with very low variation (k = 1: mean = 0.231, sd = 0.0026; k =2: mean =
0.400, sd = 0.0030), suggesting that, at the least, small deviations in sampling produce
very similar components.
2. I re-ran the logistic PCA with the full MHS dataset, including leaders and public
magicians. The factor matrix (Supplementary Table 4) is very similar to the factor matrix
for the PCA with the reduced MHS dataset (Supplementary Table 2), suggesting not
only that the sample size of 83 is sufficient to produce stable components but also that
excluding institutionalized classes did not substantially bias the results of the PCA.
3. As I just described, I re-ran the logistic PCA eight times, in each instance removing all of
the data-points in a single region. With the exception of several unstable loadings, which
S5
have been flagged, the resulting dimensions are highly stable and similar to those
produced in the main analysis.
S6
Supplementary Tables
Supplementary Table 1. The sixty societies coded in the Mystical Harm Survey (MHS). The
IDs denote the practices or practitioners coded and refer to the points in Figure 1 in the main
article. Asterisks refer to leaders or public magicians believed to inflict mystical harm.
SOCIETY
(with references)
PRACTITIONER
IDs
Akan (Debrunner 1961; Field 1970)
1, 2
Amhara (Messing 1985; Reminick 1974)
3
Andaman Islanders (Cipriani 1961; Man 1932)
4*
Aranda (Basedow 1925; Spencer and Gillen 1927)
5*, 6, 7
Aymara (Tschopik 1946, 1951)
8
Azande (Evans-Pritchard 1937; Lagae 1999)
9, 10, 11, 12, 13
Bahia Brazilians (Beierle 1999; Hutchinson 1957; Pierson 1967)
14, 15, 16
Bemba (Maxwell 1983; Richards 1935)
17
Blackfoot (Goldfrank 1966; Schultz 1930)
18
Bororo (Baldus and Lillios 1974; Colbacchini and Albisetti 1996)
19*, 20
Central Thai (Hanks 1963; Textor 1973)
21*
Chukchee (Bogoras 1907)
22
Chuuk (Bollig 1967; Mahony 1971)
23, 24*
Copper Inuit (Damas 1996; Jenness 1922; Pryde 1972; Stefánsson 1913)
25*
Dogon (Griaule and Winchell 1986; van Beek 1994)
26, 27, 28
Eastern Toraja (Adriani and Kruijt 1968, 1969)
29, 30, 31
Ganda (Mair 1934; Orley 1970)
32, 33
Garo (Burling 1963; Goswami and Majudmar 1968; Majudmar 1978; Marak
1997; Playfair 1909; Rongmuthu 1960)
34*, 35, 36
Guaraní (Ganson 1994; Schaden and Lewinsóhn 1969)
37
Hausa (Besmer 1983; Cohen 1969; Faulkingham 1971; Greenberg 1946)
38, 39
Highland Scots (Ducey 1956; Geddes 1955; Parman 1990)
40, 41
Hopi (Aberle 1951; Talayesva and Simmons 1942)
42
Iban (Graham 1987; Pilz 1988; Sandin 1967, 1980; Sutlive 1992)
43
Ifugao (Barton 1919; Lambrecht 1955, 1957)
44, 45, 46
Iroquois (Parker 1913; Selden 1966; Wallace 1972)
47, 48*
Kanuri (Cohen 1967; Peshkin 1972)
49, 50, 51*
Kapauku (Pospisil 1958, 1978)
52, 53
Khasi (Godwin-Austen 1872; McCormack 1964; Stegmiller and Knight 1956)
54, 55*
S7
Klamath (Gatschet 1890; Stern 1965)
56*
Kogi
No practitioners
coded
Korea
No practitioners
coded
Kuna (Howe 1986; Marshall 1950; McKim 1947; Nordenskiöld 1930, 1966;
Nordenskiöld and Kantule 1938; Wafer 1934)
57*, 58*, 59
Kurds (Masters 1953)
60
Lau Fijians (Hocart 1929; St. Johnston 1918)
61, 62
Libyan Bedouins (Abu-Lughod 1986)
63
Lozi (Gluckman 1955; Reynolds 1963)
64
Maasai (Merker 1971; Spencer 1988)
65, 66*, 67
Mataco (Alvarsson 1988; Karsten 1932; Métraux 1943, 1959)
68
Mbuti (Turnbull 1965a, 1965b)
69, 70
Ojibwa (Landes 1937; Rogers 1962)
71, 72
Ona (Chapman 1982; Gusinde 1971)
73*
Pawnee (Murie 1914; Weltfish 1965)
74
Saami (Itkonen 1984; Scheffer 1704)
75*
Santal (Archer 1974, 1984)
76, 77
Saramaka (Herskovits 1934; Price 1990)
78
Serbs (Kemp 1935; Pavlovic 1973)
79, 80, 81
Shluh (Berque 1973; Hatt 1974; Hoffman 1967; Montagne 1973)
82
Sinhalese (Leach 1961; MacDougall 1971)
83, 84
Somali (Cerulli 1959; Helander 1988; Lewis 1961, 1963)
85*, 86, 87
Taiwan Hokkien (Ahern 1973, 1978; Diamond 1969; Gallin 1966; Harrell
1974; Saso 1974; Seaman 1981; Wolf and Huang 1980)
88
Tarahumara (Bennett 1935; Kennedy 1978; Merrill 1988)
89
Tikopia (Firth 1939, 1954, 1970)
90
Tiv (Akiga and East 1939; Bohannan and Bohannan 1969)
91
Tlingit (De Laguna 1972; Emmons and De Laguna 1991)
92, 93
Trobriand Islanders (Malinowski 1922; Tambiah 1983)
94, 95*
Tukano (Goldman 1963; Reichel-Dolmatoff 1971)
96*, 97, 98
Tzeltal (Hunt 1962; Nash 1970)
99, 100
Wolof (Ames 1959; Irvine 1973)
101
Yakut (Sieroszewski 1993)
102*
Yanoama (Barker 1967; Chagnon 1968; Early and Peters 1990; Wilbert 1995)
103
S8
Supplementary Table 2. Factor matrix for main PCA; k=2. Loading values that exceed 0.1 are
shaded in blue; those less than -0.1 are shaded in red. Lighter shades of blue or red are used
whenever a loading is unstable (i.e., the loading approaches 0 when re-running the PCA without
data-points in one of the eight world regions; see Supplementary Tables 5 and 6). See section 2
in the main text and section 2.1 in the Supplementary Materials for details.
VARIABLE
DESCRIPTION
LOADINGS
(PC1)
LOADINGS
(PC2)
ABIL01
Fly
0.32516907
0.14302452
ABIL02
Invisibility
0.205909888
-0.009169168
ABIL03
Soul travel
0.056690138
0.033021234
ABIL04
Animal transformation
0.122242946
0.068753986
BEHA01
Cannibalism
0.103315857
0.128334064
BEHA02
Corpse desecration
0.121359352
-0.082833927
BEHA03
Opposite actions
0.281345231
0.265918042
BEHA05
Incest
0.147052238
-0.264364818
BEHA06
Necrophilia
0.33462192
-0.176723353
BEHA07
Nymphomania
0.34353687
0.006144635
BEHA08
Sexual obscenities for transformation
0.086421939
-0.291125273
BEHA09
Nudity
0.102146631
0.08326658
BEHA10
Bad hygiene
0.034940439
0.026754025
BEHA11
Association with excretion
0.039858256
0.017601684
BEHA12
Conspiracy, league, organization
0.138475056
0.019830117
BEHA13
Meet in secret
0.154627629
0.052992401
BEHA14
Harm family members
0.078032123
0.0011897
BEHA15
Nighttime activity
0.249260101
0.08622003
BEHA16
Animal familiars
0.118117498
-0.000942145
BEHA19
Magicians
0.1592915
-0.319589474
BEHA20
Political leaders
0.055161044
-0.064466745
CLASS01
All people capable
-0.146136104
-0.244239998
CLASS02
Unspecified who does harm
-0.250167558
-0.146828521
PHYS01
Physiological differences
0.052933539
0.232937982
PHYS03
Phys. differences enable powers
0.009980626
0.2803895
PHYS04
Possession
0.060299393
0.004828253
PHYS05
Other differences
0.036192844
0.074111038
PROC01
Hereditary
0.07217096
0.071607907
PROC02
Inborn powers
-0.018598541
0.195605675
PROC03
Learn powers
0.105701039
-0.079974362
PROC04
Consume substance to gain powers
0.068185729
0.115230135
PROC05
Kill someone to gain powers
0.083552187
-0.049751624
PROC06
Work with spirit
0.031419489
-0.062361694
S9
PROC07
Self-denial
-0.050628603
-0.162398126
SEX1
Females
0.02358944
0.05692325
SEX2
Males
-0.006154369
-0.130694118
TECH01
Kill
0.270335279
0.001278266
TECH02
Injure/cause illness
0.177113799
-0.018420366
TECH03
Cause sterility
0.051593471
0.013879334
TECH04
Influence love
0.038732201
-0.030082982
TECH05
Cause economic harm
0.027404997
0.00280285
TECH06
Cause catastrophe
0.085194669
-0.01350542
TECH07
Attack out-group members
0.023585796
-0.260215976
TECH08
Cause other harm
0.036128341
0.006527775
TECH10
Unintentional harm
-0.028071652
0.183714785
TECH11
Evil eye/blasting word
-0.049015075
0.116455603
TECH12
Spells, charms, material magic
0.211641972
-0.335998449
TECH13
People pay practitioner
0.08803222
-0.054815857
TECH14
Attack with thoughts
0.068173884
0.021752934
S10
Supplementary Table 3 . Factor matrix for main PCA; k=3. Note that the logistic PCA produces
different principal components depending on the value of k. Loading values that exceed 0.1 are
shaded in blue; those less than -0.1 are shaded in red. Variables with unstable loadings (see
section 2.2 in the Supplementary Materials) are shaded in gray.
VARIABLE
DESCRIPTION
LOADINGS
(PC1)
LOADINGS
(PC2)
LOADINGS
(PC3)
ABIL01
Fly
0.261310321
0.101815617
-0.201846378
ABIL02
Invisibility
0.409703525
-0.09070882
0.180670872
ABIL03
Soul travel
0.071268067
0.030519718
0.04431291
ABIL04
Animal transformation
0.119487455
0.056538577
-0.026263067
BEHA01
Cannibalism
0.153564276
0.152742408
0.051352952
BEHA02
Corpse desecration
0.12080023
-0.1122753
-0.052821421
BEHA03
Opposite actions
0.310724617
0.252884327
0.046453642
BEHA05
Incest
0.080762048
-0.241630038
-0.091354863
BEHA06
Necrophilia
0.181585866
-0.23391196
-0.210422819
BEHA07
Nymphomania
0.115152046
0.024315248
-0.23114178
BEHA08
Sexual obscenities for transformation
0.108406361
-0.195815521
0.116101125
BEHA09
Nudity
0.091572631
0.0873964
-0.096798347
BEHA10
Bad hygiene
0.044302125
0.022824908
0.10176039
BEHA11
Association with excretion
0.044741576
0.001787086
-0.002544706
BEHA12
Conspiracy, league, organization
0.17365596
-0.008798663
-0.204633806
BEHA13
Meet in secret
0.162595738
0.034531167
-0.052372707
BEHA14
Harm family members
0.071465612
-0.00815942
-0.047999928
BEHA15
Nighttime activity
0.281045093
0.080658594
0.109445833
BEHA16
Animal familiars
0.124565927
-0.007817369
0.025785961
BEHA19
Magicians
0.121134907
-0.31346491
-0.043594339
BEHA20
Political leaders
0.052265438
-0.07190035
0.005977858
CLASS01
All people capable
-0.223463755
-0.237884005
0.285994857
CLASS02
Unspecified who does harm
-0.231406022
-0.114906425
-0.155647345
PHYS01
Physiological differences
0.04683958
0.218749501
-0.058742026
PHYS03
Phys. differences enable powers
0.036536654
0.273892194
0.015797165
PHYS04
Possession
0.051172622
0.00496581
-0.053254519
PHYS05
Other differences
0.04498602
0.072057847
0.009876187
PROC01
Hereditary
0.089410421
0.065727736
0.034238935
PROC02
Inborn powers
0.019420081
0.185572523
0.12565599
PROC03
Learn powers
0.228821394
-0.153137777
0.261224069
PROC04
Consume substance to gain powers
0.083781038
0.127130635
-0.023132067
PROC05
Kill someone to gain powers
0.081128883
-0.062608431
-0.001528838
PROC06
Work with spirit
0.021399866
-0.056104371
0.020183767
PROC07
Self-denial
-0.026746541
-0.019501283
0.29936995
S11
SEX1
Females
-0.050390224
0.202456316
-0.370954747
SEX2
Males
-0.016977207
-0.122842641
0.047912838
TECH01
Kill
0.272269589
-0.035747559
0.017591199
TECH02
Injure/cause illness
0.177790256
-0.018873235
-0.07701598
TECH03
Cause sterility
0.045053583
0.015378214
-0.062798915
TECH04
Influence love
0.034943364
-0.035515646
-0.019888074
TECH05
Cause economic harm
0.094707008
-0.020996145
0.453659122
TECH06
Cause catastrophe
0.080095645
-0.019836153
0.001666995
TECH07
Attack out-group members
-0.022017348
-0.319410645
-0.097703912
TECH08
Cause other harm
0.038735278
0.00038124
0.02299223
TECH10
Unintentional harm
0.008195979
0.174330917
0.156955982
TECH11
Evil eye/blasting word
-0.028592515
0.10382641
0.072824517
TECH12
Spells, charms, material magic
0.169514626
-0.339638019
-0.175331041
TECH13
People pay practitioner
0.082383068
-0.059466278
0.012172597
TECH14
Attack with thoughts
0.080965544
0.005891889
-0.000470779
S12
Supplementary Table 4 . Factor matrix for PCA when including leaders and public magicians; k
= 2. The first PC was flipped (factor loadings were multiplied by -1) to make it comparable to
the PCs shown in Supplementary Tables 2 and 3. Loading values that exceed 0.1 are shaded in
blue; those less than -0.1 are shaded in red. Lighter shades of blue or red are used whenever a
loading is likely to be unstable (see section 2.2 in the Supplementary Materials and
Supplementary Tables 5 and 6).
VARIABLE
DESCRIPTION
LOADINGS
(PC1)
LOADINGS
(PC2)
ABIL01
Fly
0.312932832
0.150452661
ABIL02
Invisibility
0.139671194
-0.061132521
ABIL03
Soul travel
0.059654541
0.034065482
ABIL04
Animal transformation
0.112058867
0.060849166
BEHA01
Cannibalism
0.094850787
0.147224247
BEHA02
Corpse desecration
0.131118907
-0.089953239
BEHA03
Opposite actions
0.219594662
0.373463208
BEHA05
Incest
0.167292836
-0.307183159
BEHA06
Necrophilia
0.368757815
-0.211129435
BEHA07
Nymphomania
0.384224961
0.005063995
BEHA08
Sexual obscenities for transformation
0.092059453
-0.356128452
BEHA09
Nudity
0.11628465
0.105273768
BEHA10
Bad hygiene
0.034889771
0.034530096
BEHA11
Association with excretion
0.041458222
0.025661462
BEHA12
Conspiracy, league, organization
0.115818639
0.016503716
BEHA13
Meet in secret
0.134499522
0.059533441
BEHA14
Harm family members
0.082464792
0.001235027
BEHA15
Nighttime activity
0.28572701
0.119801377
BEHA16
Animal familiars
0.125248313
0.003053012
BEHA19
Magicians
0.131307225
-0.271205732
BEHA20
Political leaders
0.040163542
-0.059802318
CLASS01
All people capable
-0.113314878
-0.24511333
CLASS02
Unspecified who does harm
-0.265927537
-0.177547625
PHYS01
Physiological differences
0.046404197
0.139929013
PHYS03
Phys. differences enable powers
0.030134315
0.117287614
PHYS04
Possession
0.053450375
0.011953868
PHYS05
Other differences
0.044383839
0.036624372
PROC01
Hereditary
0.073866425
0.077017831
PROC02
Inborn powers
-0.010528638
0.203671599
PROC03
Learn powers
0.115663249
-0.096759625
PROC04
Consume substance to gain powers
0.073834959
0.133829938
PROC05
Kill someone to gain powers
0.091385668
-0.05587827
S13
PROC06
Work with spirit
0.031828357
-0.043670455
PROC07
Self-denial
0.039799114
-0.176563109
SEX1
Females
0.020079901
0.074343695
SEX2
Males
0.017071117
-0.159396295
TECH01
Kill
0.273974326
0.017403458
TECH02
Injure/cause illness
0.191232124
-0.026061077
TECH03
Cause sterility
0.046118629
0.021921333
TECH04
Influence love
0.03533839
-0.038463531
TECH05
Cause economic harm
0.027617601
0.018021146
TECH06
Cause catastrophe
0.086925475
-0.000474426
TECH07
Attack out-group members
0.021582662
-0.156680262
TECH08
Cause other harm
0.037868411
0.013103683
TECH10
Unintentional harm
-0.034699648
0.202230195
TECH11
Evil eye/blasting word
-0.048879509
0.123233424
TECH12
Spells, charms, material magic
0.176774225
-0.254119465
TECH13
People pay practitioner
0.103091852
-0.08334272
TECH14
Attack with thoughts
0.05328259
0.040583132
S14
Supplementary Table 5 . Factor loadings for PC1 for eight PCAs; all analyses excluded leaders and public magicians. For each PCA,
all of the data for a single region were removed (e.g., the AFRICA column includes factors loadings for the PCA when excluding all
of the data-points in Africa). Whenever necessary, PCs were flipped (factor loadings were multiplied by -1) to make them comparable
to each other and to those shown in Supplementary Tables 2, 3, and 4. Loading values that exceed 0.1 are shaded in blue; those less
than -0.1 are shaded in red. The coloring of the standard deviations corresponds with their values, ranging from white (sd = 0) to
green (sd = 0.15). Factor loadings for PC2 appear in Supplementary Table 6.
VAR
DESCRIPTION
AFRICA
ASIA
EUROPE
MIDDLE
AMER.
MIDDLE
EAST
NORTH
AMER.
OCEAN.
SOUTH
AMER.
MEAN
STD
DEV
ABIL01
Fly
0.277
0.457
0.290
0.325
0.344
0.209
0.296
0.298
0.312
0.071
ABIL02
Invisibility
0.150
0.083
0.391
0.203
0.166
0.092
0.119
0.182
0.173
0.098
ABIL03
Soul travel
0.056
0.051
0.075
0.057
0.051
0.058
0.067
0.050
0.058
0.009
ABIL04
Animal transformation
0.123
0.112
0.109
0.122
0.130
0.117
0.142
0.158
0.127
0.017
BEHA01
Cannibalism
0.089
0.108
0.184
0.104
0.103
0.113
0.082
0.095
0.110
0.032
BEHA02
Corpse desecration
0.125
0.107
0.109
0.121
0.117
0.138
0.213
0.055
0.123
0.044
BEHA03
Opposite actions
0.302
0.329
0.291
0.287
0.282
0.345
0.323
0.000
0.270
0.111
BEHA05
Incest
0.167
0.104
0.160
0.144
0.140
0.120
0.139
0.090
0.133
0.026
BEHA06
Necrophilia
0.359
0.277
0.001
0.332
0.332
0.298
0.365
0.267
0.279
0.118
BEHA07
Nymphomania
0.343
0.273
0.304
0.346
0.356
0.136
0.302
0.269
0.291
0.071
BEHA08
Sexual obscenities for
transformation
0.111
0.045
0.160
0.084
0.089
0.088
0.122
0.000
0.088
0.048
BEHA09
Nudity
0.095
0.098
0.085
0.103
0.102
0.094
0.084
0.269
0.116
0.062
BEHA10
Bad hygiene
0.043
0.026
0.048
0.035
0.034
0.041
0.033
0.011
0.034
0.012
BEHA11
Association with excretion
0.036
0.046
0.089
0.039
0.041
0.037
0.034
0.000
0.040
0.024
BEHA12
Conspiracy, league, organization
0.136
0.129
0.109
0.139
0.143
0.129
0.126
0.343
0.157
0.076
BEHA13
Meet in secret
0.144
0.136
0.152
0.155
0.159
0.139
0.139
0.276
0.162
0.047
BEHA14
Harm family members
0.077
0.085
0.084
0.078
0.080
0.083
0.082
0.042
0.076
0.014
BEHA15
Nighttime activity
0.234
0.253
0.301
0.251
0.232
0.225
0.226
0.260
0.248
0.025
BEHA16
Animal familiars
0.113
0.151
0.128
0.118
0.109
0.119
0.127
0.097
0.120
0.016
BEHA19
Magicians
0.170
0.151
0.115
0.155
0.153
0.177
0.162
0.091
0.147
0.029
BEHA20
Political leaders
0.058
0.050
0.047
0.054
0.053
0.061
0.048
0.066
0.054
0.007
S15
CLASS01
All people capable
-0.118
-0.146
-0.148
-0.154
-0.170
-0.188
-0.160
-0.121
-0.150
0.024
CLASS02
Unspecified who does harm
-0.243
-0.260
-0.224
-0.254
-0.235
-0.217
-0.260
-0.301
-0.249
0.027
PHYS01
Physiological differences
0.033
0.052
0.028
0.055
0.058
0.056
0.083
0.105
0.059
0.025
PHYS03
Phys. differences enable powers
-0.013
-0.014
0.016
0.013
0.023
0.010
0.032
0.041
0.014
0.020
PHYS04
Possession
0.057
0.067
0.065
0.060
0.061
0.054
0.056
0.058
0.060
0.005
PHYS05
Other differences
0.019
0.051
0.052
0.036
0.035
0.041
0.049
0.014
0.037
0.014
PROC01
Hereditary
0.065
0.072
0.078
0.073
0.070
0.086
0.075
0.074
0.074
0.006
PROC02
Inborn powers
-0.040
0.010
-0.011
-0.012
-0.048
-0.008
0.003
-0.015
-0.015
0.020
PROC03
Learn powers
0.113
0.090
0.174
0.104
0.106
0.097
0.106
0.055
0.106
0.033
PROC04
Consume substance to gain
powers
0.061
0.073
0.071
0.069
0.071
0.076
0.076
0.085
0.073
0.007
PROC05
Kill someone to gain powers
0.085
0.077
0.090
0.084
0.085
0.084
0.072
0.161
0.092
0.028
PROC06
Work with spirit
0.038
0.030
0.042
0.030
0.024
0.025
0.031
0.027
0.031
0.006
PROC07
Self-denial
-0.040
0.001
-0.020
-0.053
-0.049
-0.069
-0.059
-0.060
-0.044
0.023
SEX1
Females
0.017
0.022
0.020
0.024
0.035
0.022
0.015
0.035
0.024
0.007
SEX2
Males
-0.003
-0.005
-0.004
-0.008
-0.008
0.000
-0.004
-0.051
-0.010
0.017
TECH01
Kill
0.304
0.275
0.245
0.268
0.268
0.273
0.270
0.273
0.272
0.016
TECH02
Injure/cause illness
0.162
0.163
0.186
0.176
0.198
0.208
0.149
0.221
0.183
0.025
TECH03
Cause sterility
0.052
0.066
0.053
0.052
0.054
0.058
0.045
0.026
0.051
0.011
TECH04
Influence love
0.042
0.042
0.017
0.038
0.040
0.048
0.036
0.035
0.037
0.009
TECH05
Cause economic harm
0.025
0.034
0.037
0.027
0.023
0.029
0.041
-0.007
0.026
0.015
TECH06
Cause catastrophe
0.086
0.096
0.088
0.085
0.087
0.103
0.074
0.065
0.086
0.012
TECH07
Attack out-group members
0.060
0.016
-0.015
0.021
-0.003
0.069
0.025
-0.046
0.016
0.038
TECH08
Cause other harm
0.033
0.044
0.037
0.036
0.033
0.048
0.043
-0.008
0.033
0.017
TECH10
Unintentional harm
-0.035
-0.011
-0.024
-0.021
-0.017
-0.052
-0.006
-0.049
-0.027
0.017
TECH11
Evil eye/blasting word
-0.051
-0.041
-0.037
-0.044
-0.043
-0.042
-0.054
-0.051
-0.046
0.006
TECH12
Spells, charms, material magic
0.247
0.219
0.187
0.208
0.208
0.198
0.224
0.119
0.201
0.038
TECH13
People pay practitioner
0.092
0.078
0.083
0.087
0.085
0.085
0.080
0.076
0.083
0.005
TECH14
Attack with thoughts
0.064
0.073
0.077
0.068
0.066
0.421
0.068
0.011
0.106
0.129
S16
Supplementary Table 6 . Factor loadings for PC2 for eight PCAs; all analyses excluded leaders and public magicians. For each PCA,
all of the data for a single region were removed (e.g., the AFRICA column includes factors loadings for the PCA when excluding all
of the data-points in Africa). Whenever necessary, PCs were flipped (factor loadings were multiplied by -1) to make them comparable
to each other and to those shown in Supplementary Tables 2, 3, and 4. Loading values that exceed 0.1 are shaded in blue; those less
than -0.1 are shaded in red. The coloring of the standard deviations corresponds with their values, ranging from white (sd = 0) to
green (sd = 0.15). Factor loadings for PC1 appear in Supplementary Table 6.
VAR
DESCRIPTION
AFRICA
ASIA
EUROPE
MIDDLE
AMER.
MIDDLE
EAST
NORTH
AMER.
OCEAN.
SOUTH
AMER.
MEAN
STD
DEV
ABIL01
Fly
0.158
0.139
0.201
0.138
0.123
0.117
0.169
0.084
0.141
0.036
ABIL02
Invisibility
0.002
-0.027
-0.117
-0.013
-0.012
-0.002
-0.011
0.141
-0.005
0.070
ABIL03
Soul travel
0.043
0.029
0.006
0.033
0.052
0.031
0.028
0.015
0.030
0.014
ABIL04
Animal transformation
0.079
0.049
0.057
0.067
0.063
0.072
0.087
0.074
0.069
0.012
BEHA01
Cannibalism
0.127
0.132
0.130
0.129
0.169
0.137
0.095
0.084
0.125
0.026
BEHA02
Corpse desecration
-0.065
-0.076
-0.088
-0.085
-0.081
-0.088
-0.135
-0.089
-0.088
0.020
BEHA03
Opposite actions
0.272
0.238
0.238
0.264
0.262
0.218
0.251
0.000
0.218
0.090
BEHA05
Incest
-0.257
-0.339
-0.287
-0.262
-0.258
-0.166
-0.197
-0.045
-0.226
0.090
BEHA06
Necrophilia
-0.157
-0.221
0.000
-0.178
-0.161
-0.141
-0.151
-0.121
-0.141
0.064
BEHA07
Nymphomania
0.020
-0.055
-0.012
0.000
-0.006
-0.009
-0.021
0.130
0.006
0.054
BEHA08
Sexual obscenities for
transformation
-0.307
-0.300
-0.287
-0.294
-0.276
-0.314
-0.316
0.000
-0.262
0.106
BEHA09
Nudity
0.092
0.064
0.081
0.083
0.076
0.080
0.082
0.130
0.086
0.020
BEHA10
Bad hygiene
0.014
0.079
0.005
0.027
0.025
0.022
0.022
-0.015
0.022
0.027
BEHA11
Association with excretion
0.015
0.012
-0.028
0.018
0.012
0.024
0.015
0.000
0.008
0.016
BEHA12
Conspiracy, league, organization
0.030
0.005
0.040
0.018
0.012
0.026
0.039
-0.102
0.008
0.046
BEHA13
Meet in secret
0.059
0.037
0.066
0.051
0.043
0.063
0.074
-0.079
0.039
0.049
BEHA14
Harm family members
0.005
0.000
0.013
0.000
-0.004
0.006
0.004
-0.030
-0.001
0.013
BEHA15
Nighttime activity
0.097
0.103
0.080
0.084
0.077
0.082
0.069
0.006
0.075
0.030
BEHA16
Animal familiars
-0.001
-0.003
-0.006
-0.002
0.006
0.006
-0.024
0.000
-0.003
0.009
BEHA19
Magicians
-0.266
-0.341
-0.260
-0.323
-0.336
-0.320
-0.254
-0.448
-0.319
0.063
BEHA20
Political leaders
-0.057
-0.049
-0.051
-0.065
-0.068
-0.059
-0.060
-0.379
-0.099
0.114
S17
CLASS01
All people capable
-0.278
-0.157
-0.263
-0.246
-0.238
-0.282
-0.262
-0.160
-0.236
0.050
CLASS02
Unspecified who does harm
-0.178
-0.176
-0.114
-0.140
-0.125
-0.131
-0.131
-0.095
-0.136
0.029
PHYS01
Physiological differences
0.238
0.201
0.298
0.235
0.209
0.221
0.261
0.245
0.239
0.031
PHYS03
Phys. differences enable powers
0.272
0.297
0.303
0.280
0.221
0.289
0.323
0.188
0.272
0.045
PHYS04
Possession
0.010
-0.014
0.000
0.005
0.001
0.006
0.011
0.020
0.005
0.010
PHYS05
Other differences
0.129
0.074
0.051
0.075
0.082
0.072
0.047
0.069
0.075
0.025
PROC01
Hereditary
0.075
0.065
0.050
0.072
0.084
0.081
0.062
0.066
0.069
0.011
PROC02
Inborn powers
0.195
0.225
0.175
0.188
0.296
0.178
0.150
0.164
0.196
0.046
PROC03
Learn powers
-0.068
-0.078
-0.155
-0.081
-0.080
-0.073
-0.076
-0.076
-0.086
0.028
PROC04
Consume substance to gain
powers
0.127
0.111
0.085
0.115
0.123
0.115
0.082
0.211
0.121
0.040
PROC05
Kill someone to gain powers
-0.048
-0.052
-0.062
-0.052
-0.056
-0.047
-0.061
-0.116
-0.062
0.023
PROC06
Work with spirit
-0.073
-0.052
-0.083
-0.062
-0.059
-0.061
-0.061
-0.050
-0.063
0.011
PROC07
Self-denial
-0.220
0.000
-0.163
-0.163
-0.154
-0.165
-0.166
-0.112
-0.143
0.065
SEX1
Females
0.057
0.047
0.087
0.057
0.036
0.055
0.067
0.050
0.057
0.015
SEX2
Males
-0.188
-0.113
-0.125
-0.128
-0.122
-0.115
-0.133
-0.122
-0.131
0.024
TECH01
Kill
0.012
-0.007
-0.017
-0.011
0.003
0.002
0.001
-0.031
-0.006
0.014
TECH02
Injure/cause illness
0.013
-0.021
-0.016
-0.028
-0.029
-0.015
-0.076
0.007
-0.021
0.027
TECH03
Cause sterility
0.025
0.013
0.010
0.013
0.009
0.014
0.020
-0.009
0.012
0.010
TECH04
Influence love
-0.024
-0.043
-0.015
-0.030
-0.031
-0.032
-0.017
-0.031
-0.028
0.009
TECH05
Cause economic harm
0.003
0.022
-0.013
0.004
0.006
0.003
0.003
-0.035
-0.001
0.017
TECH06
Cause catastrophe
-0.010
0.013
-0.024
-0.015
-0.017
-0.026
-0.013
-0.058
-0.019
0.020
TECH07
Attack out-group members
-0.180
-0.179
-0.256
-0.263
-0.302
-0.332
-0.268
-0.216
-0.250
0.055
TECH08
Cause other harm
0.011
0.011
0.008
0.008
0.009
0.009
0.021
-0.056
0.003
0.024
TECH10
Unintentional harm
0.174
0.177
0.179
0.175
0.154
0.238
0.215
0.144
0.182
0.031
TECH11
Evil eye/blasting word
0.099
0.118
0.102
0.112
0.106
0.112
0.104
0.183
0.117
0.027
TECH12
Spells, charms, material magic
-0.312
-0.361
-0.322
-0.343
-0.328
-0.333
-0.357
-0.397
-0.344
0.027
TECH13
People pay practitioner
-0.045
-0.062
-0.051
-0.056
-0.053
-0.054
-0.053
-0.081
-0.057
0.011
TECH14
Attack with thoughts
0.030
0.028
0.014
0.021
0.036
-0.003
0.023
-0.045
0.013
0.026
S18
Supplementary Figure
Supplementary Figure 1. Scree plot for main logistic PCA. Gray points show the additional
variance explained by each principal component. Black points show the cumulative variance
explained.
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
12345678910
PRINCIPAL COMPONENT
PROPORTION OF VARIANCE EXPLAINED
S19
References
Aberle, D. F. (1951) The psychosocial analysis of a Hopi life-history. University of California Press.
Abu-Lughod, L. (1986) Veiled sentiments: Honor and poetry in a Bedouin society. University of
California Press.
Adriani, N., & Kruijt, A. C. (1968) The Bare’e-speaking Toradja of central Celebes (the East
Toradja), vol. 1. Human Relations Area Files.
Adriani, N., & Kruijt, A. C. (1969) The Bare’e-speaking Toradja of central Celebes (the East
Toradja), vol. 2. Human Relations Area Files.
Ahern, E. M. (1973) The cult of the dead in a Chinese village. Stanford University Press.
Ahern, E. M. (1978) The power and pollution of Chinese women. In: Studies in Chinese society
ed. A. P. Wolf, pp. 269–290. Stanford University Press.
Akiga, & East, R. (1939) Akiga’s story: The Tiv tribe as seen by one of its members. Oxford
University Press.
Alvarsson, J.-Å. (1988) The Mataco of the Gran Chaco: An ethnographic account of change and
continuity in Mataco socio-economic organization. Academiae Upsaliensis.
Ames, D. (1959) Belief in “witches” among the rural Wolof of the Gambia. Africa 29:263–273.
Archer, W. G. (1974) The hill of flutes: Life, love, and poetry in tribal India : a portrait of the
Santals. University of Pittsburgh Press.
Archer, W. G. (1984) Tribal law And justice: A report on the Santal. Concept.
Baldus, H., & Lillios, I. (1974) The social position of the woman among the Eastern Bororo. Human
Relations Area Files.
Barker, J. (1967) Memoir on the culture of the Waica. Human Relations Area Files.
Barton, R. F. (1919) Ifuago law. University of California Press.
https://archive.org/details/ifugaolawroy00bartrich
Basedow, H. (1925) The Australian aboriginal. F. W. Preece and sons.
Beierle, J. (1999) Culture summary: Bahia Brazilians. HRAF.
Bennett, W. C. (1935) The Tarahumara: An Indian tribe of northern Mexico. The University of
Chicago Press.
Berque, J. (1973) Social structures of the High Atlas. Human Relations Area Files.
Besmer, F. E. (1983) Horses, musicians and gods: The Hausa cult of possession-trance. Bergin &
Garvey.
Bogoras, W. (1907) The Chukchee, part 2. - religion. E. J. Brill Ltd. and G. E. Stechert.
Bohannan, P., & Bohannan, L. (1969) A source book on Tiv religion in 5 volumes. Human
Relations Area Files.
Bollig, L. (1967) The inhabitants of the Truk Islands: religion, life and a short grammar of a
Micronesian people. Human Relations Area Files.
Burling, R. (1963) Rengsanggri: Family and kinship in a Garo village. University of Pennsylvania
Press.
Cerulli, E. (1959) Observations on the Moslem movement in Somaliland. Human Relations Area
Files.
Chagnon, N. A. (1968) Yanomamö: The fierce people. Holt, Rinehart, and Winston.
Chapman, A. (1982) Drama and power in a hunting society: The Selk’nam of Tierra del Fuego.
Cambridge University Press.
S20
Cipriani, L. (1961) Hygiene and medical practices among the Onge (Little Andaman). Anthropos
56:481–500. http://ehrafworldcultures.yale.edu/document?id=az02-006
Cohen, A. (1969) Custom & politics in urban Africa: A study of Hausa migrants in Yoruba towns.
University of California Press.
Cohen, R. (1967) The Kanuri of Bornu. Holt, Rinehart, and Winston.
Colbacchini, A., & Albisetti, C. (1996) The eastern Bororo Orarimogodogue of the eastern plateau of
Mato Grosso. HRAF.
Damas, D. (1996) Culture summary: Copper Inuit. Human Relations Area Files.
De Laguna, F. (1972) Under Mount Saint Elias: The history and culture of the Yakutat Tlingit.
Smithsonian Institution Press.
Debrunner, H. W. (1961) Witchcraft in Ghana: A study on the belief in destructive witches and its
effect on the Akan tribes. Presbyterian Book Depot Ltd.
Diamond, N. (1969) K’un Shen: a Taiwan village. Holt, Rinehart, and Winston.
Ducey, P. R. (1956) Cultural continuity and population change on the Isle of Skye. Columbia
University.
Early, J. D., & Peters, J. F. (1990) The population dynamics of the Mucajai Yanomama. Academic
Press.
Emmons, G. T., & De Laguna, F. (1991) The Tlingit Indians Anthropological Papers of the
American Museum of Natural History Vol. 70. University of Washington Press and the
American Museum of Natural History. http://digitallibrary.amnh.org/handle/2246/253
Evans-Pritchard, E. E. (1937) Witchcraft, oracles, and magic among the Azande. Clarendon Press.
Faulkingham, R. H. (1971) Political support in a Hausa village. Michigan State University.
Field, M. J. (1970) Search for security: An ethno-psychiatric study of rural Ghana. W. W. Norton &
Company.
Firth, R. (1939) Primitive Polynesian economy. George Routledge & Sons.
Firth, R. (1954) The sociology of “magic” in Tikopia. Sociologus 4:97–116.
Firth, R. (1970) Rank and religion in Tikopia: A study in paganism and conversion to Christianity.
Beacon Press.
Gallin, B. (1966) Hsin Hsing, Taiwan: A Chinese village in change. University of California Press.
Ganson, B. A. (1994) Better not take my manioc: Guarani religion, society, and politics in the Jesuit
missions of Paraguay. University of Texas at Austin.
Gatschet, A. S. (1890) The Klamath Indians of southwestern Oregon. Government Printing Office.
https://archive.org/details/klamathindiansof02gatsuoft
Geddes, A. (1955) The Isle of Lewis and Harris: A study in British community. Edinburgh
University Press.
Gluckman, M. (1955) The judicial process among the Barotse of northern Rhodesia. The University
of Manchester.
Godwin-Austen, H. H. (1872) On the stone monuments of the Khasi hill tribes, and of some of
the peculiar rites and customs of the people. The Journal of the Anthropological Institute of
Great Britain and Ireland 1:122–143.
Goldfrank, E. S. (1966) Changing configurations in the social organization of a Blackfoot tribe during
the reserve period (the Blood of Alberta, Canada). University of Washington Press.
Goldman, I. (1963) The Cubeo: Indians of the Northwest Amazon. University of Illinois Press.
Goswami, M. C., & Majudmar, D. N. (1968) A study of social attitudes among the Garo. Man
in India 48:55–70.
S21
Graham, P. (1987) Iban shamanism: An analysis of the ethnographic literature. Department of
Anthropology, Research School of Pacific Studies, the Australian National University.
Greenberg, J. H. (1946) The influence of Islam on a Sudanese religion. J. J. Augustin.
Griaule, M., & Winchell, J. H. (1986) Dogon masks. Human Relations Area Files.
Gusinde, M. (1971) The Fireland Indians, vol. 1: The Selk’nam, on the life and thought of a hunting
people of the Great Island of Tierra del Fuego. Human Relations Area Files.
Hanks, J. R. (1963) Maternity and its ritual in Bang Chan. Cornell University, Department of
Asian Studies, Southeast Asia Program.
Harrell, S. (1974) Belief and unbelief in a Taiwan village. Stanford University.
Hatt, D. G. (1974) Skullcaps and turbans: Domestic authority and public leadership among the Idaw
Tanan of the western High Atlas, Morocco. University of California, Los Angeles.
Helander, B. (1988) The slaughtered camel: Coping with fictitious descent among the Hubeer of
southern Somalia. Uppsala University.
Herskovits, M. J. (1934) Rebel destiny: Among the bush Negroes of Dutch Guiana. Whittlesey
House, McGraw-Hill Book Company.
Hocart, A. M. (1929) Lau Islands, Fiji. Bernice P. Bishop Museum.
Hoffman, B. G. (1967) The structure of traditional Moroccan rural society. Mouton & Co.
Howe, J. (1986) The Kuna gathering: Contemporary village politics in Panama. University of Texas
Press.
Human Relations Area Files (1967) The HRAF quality control sample universe. Behavior Science
Notes 2:81–88.
Hunt, M. (1962) The dynamics of the domestic group in two Tzeltal villages: A contrastive
comparison. University of Chicago.
Hutchinson, H. W. (1957) Village and plantation life in northeastern Brazil. University of
Washington Press.
Irvine, J. T. (1973) Caste and communication in a Wolof village. University of Pennsylvania.
Itkonen, T. I. (1984) The Lapps in Finland up to 1945, vol. 2. Werner Söderström Osakeyhtiö.
ehrafworldcultures.yale.edu/document?id=ep04-017
Jenness, D. (1922) The life of the Copper Eskimos. F. A. Acland.
Karsten, R. (1932) Indian tribes of the Argentine and Bolivian Chaco: Ethnological studies.
Akademische Buchhandlung.
Kemp, P. (1935) Healing ritual: Studies in the technique and tradition of the southern Slavs. School
of the Slavonic and East European Studies, University of London, Faber and Faber
Limited.
Kennedy, J. G. (1978) Tarahumara of the Sierra Madre: Beer, ecology, and social organization.
AHM Publishing.
Lagae, C. R. (1999) The Azande or Niam-Niam: Zande organizations, religious and magical beliefs,
family customs. HRAF.
Lambrecht, F. (1955) The Mayawyaw ritual: VI. Illness and its ritual. Journal of East Asiatic
Studies 4:1–155.
Lambrecht, F. (1957) The Mayawyaw ritual: VII. Hunting and its ritual. Journal of East Asiatic
Studies 6:1–28.
Landes, R. (1937) Ojibwa sociology. Columbia University Press.
Landgraf, A. J. (2016) An introduction to the logisticPCA R package. https://cran.r-
project.org/web/packages/logisticPCA/vignettes/logisticPCA.html. Accessed 11 July 2018
S22
Landgraf, A. J., & Lee, Y. (2015) Dimensionality reduction for binary data through the projection of
natural parameters arXiv preprint arXiv:1510.06112 .
Leach, E. R. (1961) Pul Eliya, a village in Ceylon: A study of land tenure and kinship. Cambridge
University Press.
Lewis, I. M. (1961) A pastoral democracy: A study of pastoralism and politics among the northern
Somali of the Horn of Africa. Oxford University Press.
Lewis, I. M. (1963) Dualism in Somali notions of power. The Journal of the Royal Anthropological
Institute of Great Britain and Ireland 93:109–116.
MacDougall, R. D. (1971) Domestic architecture among the Kandyan Sinhalese. Cornell University.
Mahony, F. J. (1971) A Trukese theory of medicine. University Microfilms.
Mair, L. P. (1934) An African people in the twentieth century. Routledge & Songs.
Majudmar, D. N. (1978) A study of culture change in two Garo villages. Anthropological Survey of
India, Government of India.
Malinowski, B. (1922) Argonauts of the western Pacific: An account of native enterprise and
adventure in the archipelagoes of Melanesian New Guinea. George Routledge & Sons, Ltd.
Man, E. H. (1932) On the aboriginal inhabitants of the Andaman Islands. The Royal
Anthropological Institute of Great Britain and Ireland.
Marak, K. R. (1997) Traditions and modernity in matrilineal tribal society. Inter-India
Publications.
Marshall, D. S. (1950) Cuna folk: A conceptual scheme involving the dynamic factors of culture, as
applied to the Cuna Indians of Darien. Harvard University.
Masters, W. M. (1953) Rowanduz: A Kurdish administrative and mercantile center. University of
Michigan.
Maxwell, K. B. (1983) Bemba myth and ritual: The impact of literacy on an oral culture. P. Lang.
McCormack, A. P. (1964) Khasis. In: Ethnic Groups of mainland Southeast Asia eds. F. M. Lebar,
G. C. Hickey, & J. K. Musgrave, pp. 105–112. Human Relations Area Files.
McKim, F. (1947) San Blas: An account of the Cuna Indians of Panama; The forbidden land:
Reconnaissance of upper Bayano River, R.P., in 1936: Two posthumous works. Etnografiska
Museet.
Merker, M. (1971) The Masai: Ethnographic monograph of an East African Semite people. Human
Relations Area Files.
Merrill, W. L. (1988) Rarámuri souls: Knowledge and social process in northern Mexico.
Smithsonian Institution Press.
Messing, S. D. (1985) Highland plateau Amhara of Ethiopia. Human Relations Area Files.
Métraux, A. (1943) Suicide among the Matako of the Argentine Gran Chaco. América Indígena
3:199–210.
Métraux, A. (1959) Report on the ethnography of the Mataco Indians of the Argentine Gran Chaco.
Human Relations Area Files.
Montagne, R. (1973) The Berbers and the Makhzen in the south of Morocco: Essay on the political
transformation of the sedentary Berbers (the Chleuh group). Human Relations Area Files.
Murie, J. R. (1914) Pawnee Indian societies. Anthropological Papers of the American Museum of
Natural History 9:543–644.
Naroll, R. (1967) The proposed HRAF probability sample. Behavior Science Notes 2:70–80.
doi:10.4135/9781412953948
Nash, J. C. (1970) In the eyes of the ancestors: Belief and behavior in a Mayan community. Yale
S23
University Press.
Nordenskiöld, E. (1930) Picture-writings and other documents by Néle, Charles Slater, Charlie
Nelson and other Cuna Indians. Elanders Boktryckeri Aktiebolag.
Nordenskiöld, E. (1966) Miracle men and diviners among the Cuna Indians. Human Relations
Area Files.
Nordenskiöld, E., & Kantule, R. P. (1938) An historical and ethnological survey of the Cuna
Indians. Göteborg Museum.
Orley, J. H. (1970) Culture and mental illness: A study from Uganda. The Makere Institute of
Social Research and East African Publishing House.
Parker, A. C. (1913) The code of Handsome Lake, the Seneca prophet. University of the State of
New York.
Parman, S. (1990) Scottish crofters: An historical ethnography of a Celtic village. Holt, Rinehart, and
Winston.
Pavlovic, J. M. (1973) Folk life and customs in the Kragujevac region of the Jasenica in Sumdaija.
Human Relations Area Files.
Peshkin, A. (1972) Kanuri schoolchildren: Education and social mobilization in Nigeria. Holt,
Rinehart, and Winston.
Pierson, D. (1967) Negroes in Brazil. Southern Illinois University Press.
Pilz, A. (1988) Manang Jabing Anak Incham: A study of an Iban healer, Sarawak. D. Reimer.
Playfair, A. (1909) The Garos. David Nutt. https://archive.org/details/in.ernet.dli.2015.500276
Pospisil, L. J. (1958) Kapauku Papuans and their law. Yale University: Department of
Anthropology.
Pospisil, L. J. (1978) The Kapauku Papuans of West New Guinea. Holt, Rinehart, and Winston.
Price, R. (1990) Alabi’s world. Johns Hopkins University Press.
Pryde, D. (1972) Nunaga: My land, my country. M. G. Hurtig Ltd.
R Core Team (2015) R: A language and environment for statistical computing. R Foundation
for Statistical Computing. http://www.r-project.org/
Reichel-Dolmatoff, G. (1971) Amazonian cosmos: The sexual and religious symbolism of the Tukano
Indians. University of Chicago Press.
Reminick, R. A. (1974) The evil eye belief among the Amhara of Ethiopia. Ethnology 13:279–
291.
Reynolds, B. (1963) Magic, divination and witchcraft among the Barotse of northern Rhodesia.
Chatto and Windus.
Richards, A. I. (1935) A modern movement of witch-finders. Africa 8:448–461.
Rogers, E. S. (1962) The Round Lake Ojibwa. Ontario Department of Lands and Forests for the
Royal Ontario Museum. https://archive.org/details/roundlakeojibwa00roge
Rongmuthu, D. S. (1960) The folk-tales of the Garos. University of Gauhati, Department of
Publications.
Sandin, B. (1967) The Sea Dayaks of Borneo: Before White Rajah rule. MacMillan and Co.
Sandin, B. (1980) Iban adat and augury. Penerbit Universiti Sains Malaysia for School of
Comparative Sciences.
Saso, M. R. (1974) Orthodoxy and heterodoxy in Taoist ritual. In: Religion and ritual in Chinese
society ed. A. P. Wolf, pp. 325–336. Stanford University Press.
Schaden, E., & Lewinsóhn, L.-P. (1969) Fundamental aspects of Guaraní culture. Human
Relations Area Files.
S24
Scheffer, J. (1704) The history of Lapland: Containing a geographical description, and a natural
history of that country; with an account of the inhabitants, their original, religion, customs, habits,
marriages, conjurations, employments, etc. Tho. Newborough, at the Golden-Ball in St.
Paul’s-Church-Yard and R. Parker under the Royal Exchange.
https://archive.org/details/historyoflapland00sche
Schultz, J. W. (1930) The sun god’s children. Houghton Mifflin.
Seaman, G. (1981) The sexual politics of karmic retribution. In: The anthropology of Taiwanese
society eds. E. M. Ahern & H. Gates, pp. 381–396. Stanford University Press.
Selden, S. W. (1966) The legend, myth and Code of Deganawidah and their significance to Iroquois
cultural history. Indiana University.
Sieroszewski, W. (1993) The Yakut: An experiment in ethnographic research. Human Relations
Area Files.
Spencer, B., & Gillen, F. J. (1927) The Arunta: A study of a Stone Age people. Macmillan and Co.,
Ltd.
Spencer, P. (1988) The Maasai of Matapato: A study of rituals of rebellion. Indiana University
Press.
St. Johnston, T. R. (1918) The Lau Islands (Fiji) and their fairy tales and folk-lore. The Times
Book Co.
Stefánsson, V. (1913) My life with the Eskimo. The Macmillan Co.
Stegmiller, P. F., & Knight, E. (1956) The religious life of the Khasi. Human Relations Area Files.
Stern, T. (1965) The Klamath Tribe: A people and their reservation. University of Washington
Press.
Sutlive, V. H. (1992) Tun Jugah of Sarawak: Colonialism and Iban response. In: . Fajar Bakti.
Talayesva, D. C., & Simmons, L. W. (1942) Sun Chief: The autobiography of a Hopi Indian. Yale
University Press.
Tambiah, S. J. (1983) On flying witches and flying canoes: The coding of male and female
values. In: The Kula: New perspectives on Massim exchange eds. J. W. Leach & E. Leach, pp.
171–200. Cambridge University Press.
Textor, R. B. (1973) Roster of the gods: An ethnography of the supernatural in a Thai village.
Human Relations Area Files.
Tschopik, H. (1946) The Aymara. Smithsonian Institution.
Tschopik, H. (1951) The Aymara of Chucuito, Peru. 1, Magic. Anthropological Papers of the
American Museum of Natural History 44:133–308.
Turnbull, C. M. (1965a) Wayward Servants: The Two Worlds of the African Pygmies. Natural
History Press.
Turnbull, C. M. (1965b) The Mbuti Pygmies: An Ethnographic Survey Anthropological Papers of the
American Museum of Natural History Vol. 50. The American Museum of Natural History.
van Beek, W. E. A. (1994) The innocent sorcerer: Coping with evil in two African societies
(Kapsiki & Dogon). In: Religion in Africa: Experience and expression eds. T. D. Blakely, W.
E. A. van Beek, & D. L. Thompson, pp. 196–228. James Currey.
Wafer, L. (1934) A new voyage and description of the Isthmus of America. Hakluyt Society.
Wallace, A. F. C. (1972) The death and rebirth of the Seneca. Vintage Books.
Weltfish, G. (1965) The lost universe: With a closing chapter on “The universe regained.” Basic
Books.
Wilbert, J. (1995) The Sanema. HRAF.
S25
Wolf, A. P., & Huang, C. (1980) Marriage and adoption in China, 1845-1945. Stanford
University Press.
