The Killing of Witches

Author(s): Lyle B. Steadman

Source: Oceania , Dec., 1985, Vol. 56, No. 2 (Dec., 1985), pp. 106-123
Published by: Wiley on behalf of Oceania Publications, University of Sydney

Stable URL: https://www.jstor.org/stable/40330856

REFERENCES
Linked references are available on JSTOR for this article:
https://www.jstor.org/stable/40330856?seq=1&cid=pdf-
reference#references_tab_contents
You may need to log in to JSTOR to access the linked references.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

and Wiley are collaborating with JSTOR to digitize, preserve and extend access to
Oceania

This content downloaded from

154.59.124.59 on Sun, 06 Nov 2022 14:15:23 UTC
All use subject to https://about.jstor.org/terms
 THE KILLING OF WITCHES1

Lyle B. Steadman

Arizona State University

The term witch has come to have two very different meanings. On the one
hand, it is used to identify an individual who practises black magic - a
sorcerer, hence the term witchcraft. On the other hand, it refers to
individuals who are accused of being supernaturally evil. While the first
usage refers to an individual distinguished by his behaviour- sorcery - the
second refers to an individual whose actual behaviour is irrelevant; this
second witch is distinguished only by the behaviour of the accuser. In this
paper, the use of 'witch' is restricted to this second meaning, an individual
accused of being supernaturally - that is, non-demonstrably - evil. The
term 'sorcerer' is appropriate for the person who actually practises
sorcery.
The confusion generated by these two meanings may stem from the fact
that the accusation of being supernaturally evil often includes the accu-
sation of sorcery. The practice of sorcery is not only said to be destructive
but also is usually done in secret, and hence difficult to disprove. Thus, the
accusations of sorcery, as well as those of incest and cannibalism, are
appropriate to the assertion that a person is a witch. The important point
to note with a witch, in contrast to a sorcerer, is that it is the accusation of
evilness, not the technique, which is distinctive. We shall see that such an
accusation is a metaphor whose implied simile is that the accused is 'like
evil': he or she only represents evil.
Consistent with this definition, Patterson, in her survey of witchcraft in
Melanesia, distinguishes a witch's actions as being unobservable, in
contrast to those of a sorcerer (1974:140). Likewise, Mary Douglas notes
in her volume on witchcraft, 'most of the contributors have used "witch-
craft" to refer exclusively to internal psychic power to harm, and
"sorcery" for bewitching by means of external symbols, whether by spells,
charms, or potions' (1970:xxxvi). Internal psychic power, of course, is
unobservable: its identifiable reality exists in the claim alone.
Malinowski (1929:47) makes the same observation:
[A] sorcerer actually knows the magic of his trade; when called upon
he will utter it over the proper substances; will go out at night to
waylay his victim or visit him in his hut; and in certain cases, I
suspect, may even administer poison. The witch, on the other hand,
however much she may be believed to play the part of ayoyoya, does
not- needless to say- really fly or abstract the insides of people, and
she knows no spells or rites, since this type of female magic lives
merely in legend and fiction.
And finally, Middleton writes, 'witchcraft may be distinguished from
sorcery, although Lugbara often use the word oleu for both. Witches affect
their victims by mystical means, but sorcerers use material medicines'. He
notes that, as in Europe, 'they both are grouped together, as being evil and anti-
social people' (1960:245, 247).
The most important effect of talk about witches, as here defined, is not
the description of what witches are alleged to do or how to identify them, or

106

This content downloaded from

154.59.124.59 on Sun, 06 Nov 2022 14:15:23 UTC
All use subject to https://about.jstor.org/terms
 L.B.Steadman

even its use of witchcraft in accusing individuals of unprovable evil behav-

iour. The most striking social consequence of a witchcraft accusation occurs
when it is used to justify the execution of innocent individuals - innocent at
least of the supernatural crimes of which they are charged. In the attempt to
identify common elements in the killing of witches, I shall compare the killings
which occurred in late Medieval Europe with those occurring still in a society
in Papua New Guinea, the Hewa.

Europe

In Europe during the 16th and 17th centuries perhaps as many as

million people were killed as witches (Cohn 1970:12). Many slay
occurred along the area divided by the Rhine river, which has long been
important European boundary. During the Roman occupation, it mar
major segment of the northern frontier of the Roman empire. Today the
still coincides with the basic division between the Latin and German
languages, and more or less divides Catholicism and Protestantism. Durin
two witch-killing centuries the Protestant Reformation and Catholic Cou
Reformation raged along this interface. The Protestant reformers,Lu
Osiander, Calvin, and Zwingli, and the humanist Erasmus, all lived near
Rhine. The historian Erik Midelfort (1972) studied court records of mor
1,200 cases of witch-killings which occurred during the 16th and
centuries along a segment of the Rhine in southwestern Germany
carefully collected and described data will be central to the analysis pres
here.

In virtually all of Midelfort's cases, the threat of religious revolut

evident. The first extensive witch-hunt in the German Southwest occurre
1 562 in the community of Wiesensteig. Its religious classification- Prot
or Catholic- was so uncertain that it has been a point of controversy am
historians (1972:90). Because of 'popular pressures for reform', cer
reformers in 1555 were invited to present lectures. These lecturers 'spli
factions, arguing among themselves the merits of Luther, Osiander
Zwingli' (1972:88). This religious competition eventually culminated i
rejection of Catholicism in favour of Protestant authories. Seven years lat
1562, 63 witches were killed, followed five years later by the town's retu
Catholicism with accompanying authority changes. Fifteen years later at
25 more witches were burned by the community (1972:90).
Another example of the relationship between religious strife and w
killing is Offenburg, a city which had remained formally Catholic durin
Reformation despite Protestant sympathies among the citizenry. 'To
against infection of the populace with noxious notions ... the town . . . [cl
its gates on Sundays ... to prevent citizens from walking the short w
Weingarten,where services were Protestant' (1972:127). Eighty-six w
were killed by this community during this threat of Protestantism.
Trevor-Roper focuses on this social and religious turmoil in Europ
communities during the two witch-killing centuries, writing, 'thus, if we
the revival of the witch craze in the 1 560's in its context, we see that it is n
product either of Protestantism or of Catholicism, but of both: or rather of
conflict' (1968:139-140).
While some writers have attempted to attribute witch-killings to

107

This content downloaded from

154.59.124.59 on Sun, 06 Nov 2022 14:15:23 UTC
All use subject to https://about.jstor.org/terms
The Killing of Witches

hysteria or fanatical behaviour, this would be a mistake, for there is

considerable evidence indicating that witch-killings were approved and
rationally defended as necessary in these revolutionary times. In fact, as
Trevor-Roper notes, 'the more learned a man in the traditional scholarhip of
the time, the more likely he was to support the witch-[killers]'( 1968: 154).
A good illustration of this rationality is the argument between two lawyers in
the city of Schwaebisch Gmuend. The distinguished and respected lawyer and
scholar, Dr Leonhard Kager, in a strong criticism against the killing of witches,
noted that people were being killed without evidence, solely on the basis of
accusations and confessions made under extreme torture. The argument made
by a lawyer against Kager is summarized by Midelfort: '1) Witchcraft is
primarily a spiritual crime and need not involve harm to require the death
penalty. 2) Witchcraft is a special crime; therefore, the ordinary rules of law do
not apply. 3) Judges must be concerned with the safety of society and should
therefore be ready to sacrifice individual rights' (1972:120). This counter-
argument was accepted by the community, as evidenced by the large number of
witches killed a few years later. We shall have more to say about the 'safety of
society'.
The facts presented by Midelfort make clear that European witch-killers
were well aware that the kind of evidence used to justify the execution of
alleged witches was never substantial. For example, the Jesuit Fredrich von
Spee, 'with years of experience as confessor to witches about to be executed . .
asserted that he had never seen one who really had done the things she
confessed' (Midelfort 1972:28). Bodin, the 'medieval Aristotle', explained
that 'proof of such evil is so obscure and difficult that not one out of a million
witches could be accused or punished if regular legal procedure were followed'
(Midelfort 1972:19).
The feature distinctive of the European witch, the feature used publicly to
justify her death, was her relationship to the Devil. Midelfort cites Rossel
Robbins: 'the main crime [of the witch] was knowingly invoking the devil and
opposing the Church; quite secondary was causing disease or death'
(1972:16). 'Witches, [Luther]declared, should be burnt, even if they do no
harm, merely for making a pact with the Devil' (Trevor-Roper 1968:137).
The Devil, let us keep in mind, is the Christian symbol used to identify anti-
Christian behaviour - behaviour seen as threatening one's own Christian
community. During the Reformation and Counter-Reformation such anti-
Christian behaviour was seen by individuals as directly threatening their own
particular Christian organization. Pope Innocent VIII was obviously con-
cerned with such challenges to the Church when he issued what turned out to
be an exceptionally important bull against witchcraft. In 1484, near the
beginning of the two witch-killing centuries, he 'sorrowfully reported that he
had heard of widespread witchcraft in Upper (i.e., southern) Germany . . .
Unmindful of their own salvation and deviating from the Catholic faith many
persons of both sexes were supposedly abusing themselves with evil spirits'
(Midelfort 1972:20; my italics).
Witches not only cannot be shown to be guilty of the crimes with which they
have been charged but also they do not seem to have been guilty of any
particularly threatening or antisocial behaviour. Indeed, witches tend to be
chosen from among the least threatening people. Vulnerability in their lack of
social support appears to have been the only quality distinctive of the
European witch. Midelfort discovered that 82 percent of the 1,200 slain

108

This content downloaded from

154.59.124.59 on Sun, 06 Nov 2022 14:15:23 UTC
All use subject to https://about.jstor.org/terms
 L.B.Steadman

witches were female. He noted also that within this category, 'widows and
spinsters were most commonly accused of witchcraft, far out of proportion to
their numbers in society. Certainly it was against them that many hunts were
initially directed; . . . persons without families were . . . unprotected'
(1972:185).
Several other points should be noted in regard to European witch-killing.
Virtually all witches killed in Europe were killed by members of their own
community: 'witch trials [in Europe] were essentially a social internal matter'
(Trevor-Roper 1968:160; see also Mair 1969:9-10, Douglas 1967:78).
When witch-killings occurred they were always supported by many people
in the community, including prominent members - magistrates, priests, min-
isters, and other highranking officials. However, the influentual people alone
were not the cause of witch-killing. Trevor-Roper emphasizes this feature:
4[No] ruler has ever carried out a policy of wholesale expulsion or destruction
without the cooperation of society . . . Great massacres may be commanded by
tyrants but they are imposed by peoples. Without ... the people, social
persecution cannot be organized. Without the people, it cannot be conceived'
Í1968: 114-15).
Indeed, at times authorities were virtually forced to burn witches. Midelfort
describes'the helpless position of the magistrates' when 'the wrath of [the]
populace, unwilling to abide by the rule of law [demanded that witches be
killed]. Witch-killing was not mass madness to be sure, but it most certainly
was not the madness or delusion of a small class of bureaucrats either'
(1972:191).
Midelfort cites other witch trials which broke out, 'not because the
magistrate felt compelled to find a reason, a scapegoat, but because popular
pressures demanded one* (1972:190). At other times preachers vigorously
cautioned magistrates not to 'believe all the denunciations of the vulgar rabble'
(1972:39. In Offenburg, 'local pressure for witch hunting . . . seems to have
come almost entirely from the guilds, against the moderating, cautious attitude
of the councir (1972:127).
Those killed as witches in Europe, then, do not appear to be guilty of any
crime or antisocial behaviour, but rather tended only to be non-threatening,
vulnerable individuals. They were not strangers or enemies but were known by
the accusers, and their killing enjoyed wide community support.
An important question is why European witch-killing finally ended.
Midelfort argues that people became suspicious of the machinery involved in
identifying witches. But why would they not have become suspicious a century
or two earlier? Some historians, Trevor-Roper, for example, have argued that
the reason for the end of witch-killing was that the stereotype of the witch broke
down; but why, then, was it not simply replaced by another? The hypothesis of
witch-killing I propose is that such killings occur only when there is a
significant threat to the relationships of those supporting the killings. It is of
great significance that European witch killing ceased when the Reformation
and Counter Reformation ended. After two centuries of intense religious strife
within individual European communities, especially along the Rhine, the
killings finally ceased when the authority structures of the communities
became relatively stable once again. This is a suggestive correlation.

The Hewa

Witch-killing of course is not confined to Europe. Reports of such kill

are common in anthropological literature, but virtually all such studies w

109

This content downloaded from

154.59.124.59 on Sun, 06 Nov 2022 14:15:23 UTC
All use subject to https://about.jstor.org/terms
The Killing of Witches

made after the killings had been banned by European colonial administrators.
During the years 1 966 to 1 969, 1 conducted a 22 month field study among the
Hewa of Papua New Guinea, a society virtually uninfluenced by Europeans.
At the time of the study, individuals were being killed as witches.
The Hewa live in one of the most inaccessable areas of the Western
Highlands of Papua New Guinea,occupying a mountainous region about half
a mile above sea level. They practise slash-and-burn agriculture and live in
widely dispersed single family houses. Population density averages only about
three or four people per square mile.
The remoteness of the Hewa from one another, as well as from neighbouring
societies, is a major factor which has kept them isolated from European
influence. When I first entered the area in 1966, the only sign of European
contact was an occasional steel axe, obtained through trade with their
neighbours. Up to the time I finally left in 1969, there still were no
administrative officers or missionaries in the area. Thus, in contrast to most
other peoples of Papua New Guinea, the Hewa are virtually independent
politically. This remarkable lack of administrative, missionary, or other
outside influence means that they, at least when I lived with them, can
accurately be described as a genuinely aboriginal society (Steadman 1971,
1975; Steadman and Merbs 1982).
There are three main types of residential groups among the Hewa: the
household, containing an average of seven people; a group I call the
neighbourhood, containing two to four dispersed households; and, finally, a
cluster of three or four neighbourhood groups, usually referred to by a clan
name, which is identified by cooperation in roofing and flooring one another's
new houses. This largest group contains about forty to sixty people who are
dispersed over a large territory and is somewhat amorphous because of the
movement of individuals and sometimes whole households between such
groups. These moves are sometimes related to conflicts which individuals have
had in their own group, but they are also a result of an interest in moving to live
with or near relatives. As a consequence of the Hewa prohibition on
marriage between many identified relatives, including all members of the same
patrilineal clan, marriages tend to be distant, leading to the wide dispersal of
relatives.
During the period I stayed with the Hewa, I was able to gather fairly detailed
information on eighty separate killings. These reflected a killing rate of almost
one percent of the population per year, which may be one of the higher known
rates of killing in the world. In about half these killings, the victim was killed as
a witch, and most of these were female; the males were usually killed in the
retaliatory fighting which sometimes followed a witch-killing.
Each witch-killing is preceded by accusations of witchcraft which may be
made any time by almost any adult, but a killing neither immediately nor
necessarily follows. In cases in which a person lies dying, and it has been
'determined' that his viscera were eaten by a witch (although it is never
actually checked by autopsy), he is encouraged to name the witch responsible
for his death. If he fails to identify the witch, she may be identified by
divination.
I found no case where either close relatives or people closely associated with
the alleged witch agreed with the charge; they invariably and contemptuously
dismissed it as a lie. Witchcraft accusations are seen as threats not only to
the witch but also to those associated with the alleged witch, and therefore
represent a test of their power and courage. Three choices are open to those
associated with an accused witch: 1) they may challenge the accuser by
they dont focus on the nature of evilness of witchcraft itself -- they dont check if death is caused by witchcraft

no

This content downloaded from

154.59.124.59 on Sun, 06 Nov 2022 14:15:23 UTC
All use subject to https://about.jstor.org/terms
 L. B. Steadman

fighting him; 2) they may move away from the accuser's area; or 3) they may
ignore the accusation. Should the choice to fight the accuseds) be made it is
normally done with a staff or cane, not a bow and arrow, the usual killing
weapon. If the group challenging the accusation displays strength, the accu-
sation will usually not lead to a witch-killing. On the other hand, when the
accusation is ignored, there is always danger, for the accused is living under a
kind of sentence of death which may be executed at any time. The group which
does not react aggressively to an accusation is seen by everyone as weak and
vulnerable.
The precipitous moment occurs usually at a funeral, when a decision is
made to kill the person said to have 'eaten' the person being mourned. An
influential man, usually of the deceased's neighbourhood group, may then
attempt to form a band of about eight or ten males to go and kill the accused
witch. These males are tied- politically, economically, and by kinship- to a
number of people most of whom have given at least tacid support to the killing.
Thus, the killing is never automatic: the decision is always influenced by the
adults at the funeral; and the killing party never leaves on its mission without
general support from those attending. The men travel through the forest at
night and, just before dawn, attack the accused with arrows while she sleeps.
When she tries to escape, her attackers shoot arrows into her fleeing body until
she falls, each member of the killing group striving, usually successfully, to
strike her with at least one arrow. I could not find one adult male who had not
participated in a witch-killing. On the other hand, I recorded no instance of a
woman ever killing anyone.
Yet accusations did not always lead to killings. In a number of cases in
which someone would suggest the killing of a particular person as a witch, the
talk would be met with either indifference or an outright refusal to accompany
the killing party. Much indifference or several refusals invariably would end
the talk of the killing. Prior accusations were almost always made months
before the funeral or the killing, and not necessarily by the group doing the
killing. Also, criticism by adult females may stop the killing. In some instances,
females have stopped the killers by leaving the funeral early, gathering on the
track leading toward the person to be killed, and shaming the killing party when
they came along by lifting their skirts and exposing their genitals.
The Hewa claim that the distinctive behaviour of ap isai, a witch, is that she is
a cannibal, that she somehow opens up a living person, eats his viscera, and
then closes him without leaving a trace. Her victim does not die immediately
but gradually, sometimes taking weeks. So far as I could determine, no alleged,
still living victim of a witch has ever been confident that his viscera were
actually eaten. Being eaten is often determined by a diviner, not by the actual
experiences of the victim, and is usually established as the victim lies dying, in
an almost comatose state. Thus, the determination of having been eaten ispost
hoc. Everyone knows that they themselves have never actually seen a witch
performing her 'cannibalistic' behavior, and also, so far as I could determine,
no one has ever been caught in the act of actually eating someone's viscera or
indeed any part of a human being. When I asked men who had killed witches
for the evidence used to justify their accusations, they were always uncertain.
Indeed, when discussing this problem in detail, several men pointed out that,
after they had killed a person as a witch, they had wondered whether the
accusation was true and how they could be certain of it.
Further, no one ever seems to be interested in identifying the opening in a
person through which his viscera is said to have been eaten or even whether the

ill

This content downloaded from

154.59.124.59 on Sun, 06 Nov 2022 14:15:23 UTC
All use subject to https://about.jstor.org/terms
The Killing of Witches

viscera are missing. Although everyone knows that people may lie about such
matters, no one ever seeks to determine the truth by evidence. Indeed, the
actual behaviour of a witch is irrelevant to the charges made. What is clear is
that people are concerned with the consequences of the killing, such as the
likelihood of retaliation, past grievances with the group, whether it might
eventually have to be settled by a 'blood' payment of pigs, and so on.
The evidence I gathered indicates that a Hewa witch, like her European
counterpart, is generally as innocent of wrongdoing as anyone in the society.
Anthropologists have spent much effort in attempting to identify the antisocial
behaviour of those accused ofbeing witches (e.g., Mair 1969: 22-23, 102), but
I found none. The witch, as far as any Hewa was certain, could not have done
the supernatural things of which she was accused, and if she had indeed
committed crimes or misdeeds, she would have been dealt with appropriately.
For example, when a married woman runs off with a lover, she is likely to be
pursued by her brothers and her husband and, if caught, killed. She is not called
a witch.
Thus, the features common to both European and Hewa witch-killing and
perhaps witch-killing generally, appear to be the following: the accusation of
being a witch is, by definition, one that cannot be either proven or disproven;
and it is an accusation of supernatural behaviour. The only characteristic
distinctive to both the European and Hewa individuals killed as witches is their
vulnerability. The witches are individually known by the accusers, and their
killing has wide support.

Alternative Explanations

Most explanations of witch-killing assume that the killers believe

even though the accuracy of such assumptions cannot be verified.
witches are usually said to provide a scapegoat for frustrated
ridden individuals. The term scapecoat usually refers to a person
to bear the blame for others Its meaning is derived from an ancie
ceremony which involved 'a goat upon whc:3 head were symbolica
the sins of the people, after which he was suffered to escape into th
(Webster's Dictionary); hence, 'escaped goaf (see Leviticus 16
William Howells gives a scapegoat explanation:
[It] seems to be a fact that witchcraft works in several ways a
valve through which may escape the accumulations of anxiety, i
envy, and neurotic tensions that arise in all human groups, and
particularly painful in a small, closed, primitive society from w
is no escape (1962: 109- 110).
Trevor-Roper also uses such an explanation. He argues that witch
Europe must be seen as the social consequence of what he calls the
war' between Protestantism and Catholicism. In the climate of fear accom-
panying this ideological struggle, the people sought a scapegoat, the witch
(1968: 114, 140).
Robert Levine (1973) offers yet another scapegoat explanation for witch-
killing in England. He argues that when poor-houses were built, people could
justifiably refuse beggars; but, because the old system had emphasized sharing,
they still felt guilty when refusing them. The person feeling guilty projected his
guilt on to the beggar by calling him a witch; the beggar was then tried and
executed, making the accuser feel betterby allaying his guilt ( 1 973:254-270).
This explanation is defective on two important counts. First, it is not

112

This content downloaded from

154.59.124.59 on Sun, 06 Nov 2022 14:15:23 UTC
All use subject to https://about.jstor.org/terms
 L. B. Steadman

convincing that a person would feel more guilt in refusing a beggar than in
having that innocent beggar killed. Secondly, the accuser is not the person
doing the killing; rather it is the magistrates, supported (or provoked) by the
community. Thus, such killing cannot be explained by the alleged guilt of the
accuser, even if it could be identified. If the community members did not want
to kill such innocent people, they would simply ignore the accusation, perhaps
even ridicule the accuser. Accusations are always socially cheaper than
killings.
The explanation for European witch-killings proposed by Marvin Harris is
similar. He examines data presented by Midelfort and others and argues that
the 'principal result of the witch-hunt system . . . was that the poor came to
believe that they were being victimized by witches and devils instead of princes
and popes' (1974:237). In those economically troubled times, the rich
convinced the poor that their economic troubles lay with imaginary witches.
The anxiety and frustration felt by the poor, and caused by the greed of princes
and popes, was then dissipated by slaughtering innocent people as scapegoats:
'You could actually see the authorities doing something to make life a little
more secure; you could actually hear the witches scream as they went down to
hell' (1974:238).
On the basis of almost two years in the Hewa, however, I found that witch-
killing has just the opposite effect: it invariably created anxiety and fear. When
innocent people are slaughtered, other innocents are vulnerable to being
accused and killed. Indeed, after arguing that witch-killing seemed to make life
a little more secure, Harris more accurately characterizes the situation when
he writes that 'the witchcraft mania . . . made everyone fearful, heightened
everyone's insecurity' ( 1 974:249). Indeed, I shall argue that the aim of witch-
killers is to achieve exactly that: to cause anxiety and fear and thereby to
intimidate a category of people.
Harris's main argument in regard to European witch-killing is that it was
used by the rich to exploit the poor: 'Witchcraft beliefs' he tells us, 'shifted
responsibility for the crisis of late medieval society from both Church and
to imaginary demons in human form' (pp. 237-238). Witchcraft thus 'drew the
poor further and further away from confronting the ecclesiastical and
secular establishment with demands for the redistribution of wealth and
the levelling of rank' (1974:239-240).
This hypothesis overlooks an important question: How did the poor come to
accept a belief which apparently went so strongly against their true interests?
How were the 'popes and princes' (the clergy and nobility) able to convince the
poor to take actions likely to perpetuate their misery rather than end it? If, as
Harris writes, the real devils were the corrupt clergy and the rapacious
nobility, how could they successfully delude the poor for 200 years? He notes
that there was widespread discontent among the poor which led to high fre-
quency of military-messianic movements during this period, sometimes
including pitched battles ( 1 974:239). Thus, how was it possible for the clergy
and nobility to convince the lower classes in such militant times that their
troubles lay with imaginary witches, when it was evident that similarly poor
and dispossessed people were mobilizing and following prophets who were
attempting to cope with their economic and social problems in a practical way?
Earlier Harris stated, 'the witch craze can't be explained in terms of the
consciousness of the people who participated in it' (1974:235-236). And
elsewhere he rejects such explanations when he discusses the Jewish military-
messianic revolts against their Roman oppressors: 'Classes, races and nations

113

This content downloaded from

154.59.124.59 on Sun, 06 Nov 2022 14:15:23 UTC
All use subject to https://about.jstor.org/terms
The Killing of Witches

usually [take desperate measures against great odds] . . . not because they are
duped by irrational ideologies, but because the alternatives are abhorrent
enough to make even great risks worthwhile' (1974:174).
Indeed, the stated aim of Harris' book is to explain 'consciousness' or
ideology- not to use consciousness itself as an explanation. But it is just
consciousness among the poor which Harris uses to explain the witch-killings,
despite the disclaimer twelve pages later that 'Consciousness is adapted to
practical and mundane conditions. These conditions cannot be imagined into
or out of existence^ 1974:25 3). Harris, thus, offers an explanation which he
elsewhere rejects.

The Hypothesis

To deal with the killing of people said to be witches, we must seek t

account not only for why the killings are justified by a supernatural accusatio
but also why innocent people are killed in the first place. Public punishmen
has two important effects: it not only punishes the offender but als
communicates what is likely to happen to future offenders. The second effec
should be the most important: while a crime cannot be undone, its futu
incidence can indeed be influenced. One killing does not eliminate anoth
nor does punishment of a thief eliminate the robbery. But public punishm
of a crime increases the likelihood that individuals in the future will n
commit that crime. Thus, punishment communicates to potential offender
including the one punished, the cost of the crime. A publicly backed killin
then, not only eliminates the person killed and hence any threat he poses bu
also aimed at intimidating those who may behave in a similar way in the futu
But why are individuals killed as witches? Because the reality of the alleg
crime is irrelevant, a person so killed is not killed for the behaviour of which sh
is accused. The punishment of witches cannot be aimed at intimidatin
witches.
Hewa witch killings invariably reflect grievances based on past hostilities.
These hostilities include killings, adultery or even a pig ruining another
person's garden.Members of neighbouring roofing and flooring groups usually
pose the greatest threat, and it is in these groups that witches are 'identified'
and killed. One feature of a Hewa group is that the most extreme form of
violent competition -killing- almost never occurs within it. Other than a few
females killed by their brothers and husbands for adultery, I have no records of
persons being killed by members of their own group- certainly not their
household or neighbourhood and,as far as I can determine, their roofing and
flooring group.
The reason for the lack of violent competition within groups seems to be
related to the behaviour of influential members. The continuous influence of
parents generally reduces competition between their children and promotes
their cooperation; siblings are raised to cooperate with one another. In the
neighbourhood groups the adult males, especially household heads, tend to
suppress violence within them and encourage cooperation. Thus, fathers and
mothers diminish competitive and hostile behaviour between their offspring;
household heads, between their household members; and the leaders of
neighbourhoods and roofing and flooring groups, within their respective
categories. Although there is a competition between individuals who come into
frequent contact, it tends to be sharply limited by the presence of such
influential individuals. And the behaviour of these influential individuals, in

114

This content downloaded from

154.59.124.59 on Sun, 06 Nov 2022 14:15:23 UTC
All use subject to https://about.jstor.org/terms
 L. B. Steadman

turn, has been influenced by their ancestors.

On the other hand, although there is not less contact between members of
different roofing and flooring groups, there is a much greater likelihood of
violence when competition does occur, and almost all witch-killings occur
between them. The leaders of these different groups are not usually in co-
operative relationship with one another. Cooperation is not regularly en-
couraged between members of different groups other than with particular
kinsmen. If someone steals the wife of a member of another such group, he
generally will not be strongly criticized by his own group. This is in sharp
contrast to taking the wife of someone within one's own group. Further, if a
man's pig ruins the garden of a man in another roofing and flooring group, the
owner of the pig will rarely be encouraged to give restitution.
The primary threats, then, are influential men in different, but nearby,
roofing and flooring groups - men prominent in making decisions, some of
which include killing someone. Yet it is not these men who are killed as
witches, but weak and vulnerable individuals, usually women, who reside with
them.
By killing anyone, the witch-hunters announce to everyone their willingness
to dispose of people who threaten them and their social relationships. The
Hewa data indicate that the communication of witch-killing is aimed at
intimidating certain people associated with the witch, individuals who indeed
threaten the witch-killers and those related (through kinship and locality) to
them. The consequences of such a threat are not always predictable and
sometimes include retaliation against, and the killing of, the witch-killers
themselves. It can even lead, years later and after a relative change in power, to
retaliation in the form of a new witch-killing. Often, however, because such a
killing is discussed and calculated long beforehand, it is not revenged and does
tend to have the consequences of intimidation. Everyone in the area realizes
that the killing group is strong, is unified, and is willing both to use physical
violence and to risk the lives of its members to challenge other groups. The
major consequences of intimidation are difficult to demonstrate conclu-
sively, but my data indicate that its major impact is in influencing future
conflicts. When future competition occurs over such things as females, pigs,
fruit trees and other resources, there is readiness among those who have
been intimidated to give in to the interests of the intimidators. They are
much more likely to yield the object of competition and even to move to
different areas when such competition threatens.
Why are innocent, non-threatening individuals killed, rather than those
posing the real challenge? Those killed as witches are politically insignif-
icant - females or weak old men - and do not have close ties to strong men who
would be likely to risk being killed in their defence. Killing strong males, or
people who will be defended by them, is extremely dangerous. Hewa adult
males almost always have their bows and arrows, their main weapon, at
hand, even when they sleep. Thus, while killing anyone is the most dan-
gerous activity among the Hewa, the easiest, safest people to kill are those
whose death is the least likely to be revenged, those with the fewest and
weakest ties to influential men.
But why is a witchcraft idiom used to justify the killings? Why do they not
simply state that they are killing a person as an enemy, or as a member of a
category they fear and wish to intimidate? Why do they not simply say that a
known weak, vulnerable individual of the category of people they wish to
intimidate should be killed because she is easier to kill than anyone else in that

115

This content downloaded from

154.59.124.59 on Sun, 06 Nov 2022 14:15:23 UTC
All use subject to https://about.jstor.org/terms
The Killing of Witches

category? There may be several reasons.

The first one is proposed by Craig Palmer (personal communication): a
witchcraft justification is used when there is a possibility of future cooperation
between the killers and those associated with the victim. This consequence
should be important where future cooperation can be anticipated, as within a
community (for example, in Europe), or where current opponents, because of
shifting alliances, may become allies.The witch justification can later be cited
by those associated with the victim as an excusable motivation, however
mistaken. Thus, a witchcraft idiom facilitates the re-establishment of relations.
In contrast, in the case of feud, which also often involves the killing of
vulnerable innocent individuals, the implicit justification is usually the same as
that stated explicitly: the victim is killed simply as a member of the threatening
category. Future cooperation between the opposing sides is not usually
anticipated; they speak explicitly of each other as implacable, enduring
enemies. A witchcraft accusation is irrelevant.
The second reason is that the killing of an innocent individual known to the
community requires wide support, which is often difficult to achieve, since the
Hewa are killing innocent people whom they know and who may even be
distant relatives of some of the killers or their supporters. The movement of
families and the dispersal of people by marriage leads to people having
recognized relatives living sometimes at great distances. Some individuals in
the killing groups will almost certainly be related to the suggested victim in
some way and, therefore, may not be equally happy with the choice: they may
even suggest others whom they feel would be more appropriate. The selection
would be open to discussion, argument, and criticism by members of the group
associated with the killers. Such a debate would make the selection difficult
and would probably involve controversy, which would in turn impede any
proposed action.
Because a supernatural assertion is not verifiable, it is not disprovable; it is
not challengeable by evidence. The supernatural idiom forces the listeners
either to accept or to reject the claim. If one rejects the claim, one rejects the
speaker's influence and hence, to some extent, his relationship. If his social
relationship with the speaker is valuable- if the speaker is a 'parent' or leader,
for example- he is not likely to reject the assertion and what it implies. If the
speaker is not valuable to him, he may reject the request and indeed may
ridicule him. Hence, the successful use of a supernatural assertion in a request
for support is likely only when it is based on a prior relationship.
A related point is that when a person is finally killed as a witch, accusations
against her have circulated for some time, thereby permitting people to think
about the killing and its likely social consequences. And perhaps most to the
point, a supernatural idiom allows people to demonstrate purely their
willingness or unwillingness to endorse a particlar enterprise, unfettered by the
truth of the accusation.
Thus, the Hewa appear to support the killing of innocent people because the
victims are recognized as being associated with a category of people they seek
to intimidate. Their victims are killed as symbols which communicate a
willingness to engage in violence to protect their own social relationships. The
fundamental consequence of such a communication is that it increases the
likelihood of future conflicts being resolved in their favour.
In Europe, the argument against the lawyer Kager in Schwaebisch Gmuend
rationally emphasized the importance of witch-killing to the community: since
the killing of witches would save their social relationships (the 'society'), the

116

This content downloaded from

154.59.124.59 on Sun, 06 Nov 2022 14:15:23 UTC
All use subject to https://about.jstor.org/terms
 L. B. Steadman

guilt or innocence of the victim was irrelevant.

In a similar way, Benedict Carpzov, a Lutheran scholar, in 1635 . . .
admitted that torture was capable of grave abuse and had led to thousands
of false confessions throughout Europe. But he concluded that ... it
should still be used, even on those who seemed innocent ... He
maintained that even those who merely believed that they had been at the
[witches'] sabbat should be executed, for the belief implied the will . . .
[He] procured the death of 20,000 persons. [Trevor-Roper 1968:151.
Clearly, many individuals were willing io forgo the requirement of evidence
and ignore the normal protection of innocent people, when the public perceived
a great internal danger to their community. Threatened by such danger,
individuals tolerated and even encouraged accusations against vulnerable
individuals, and then had them killed.
Thus, in the sixteenth and seventeenth centuries, members of individual
European communities were willing to kill innocent, known individuals in an
attempt to prevent religious revolution. By destroying innocent individuals
who were alleged to be associated with the Devil, the symbol of the behaviour
threatening each Christian community, they attempted to intimidate those who
posed the true threat, those likely to support religious revolutionaries.
The individuals who actually posed the threat would be neither easily
identified, nor in most cases, guilty of a crime. Often, their observed behav-
iour would be difficult to identify as clearly threatening. Further, those who
might support or bring about a religious revolution probably already had
some influence, and killing or punishing them without good evidence might
be strongly resisted. Thus, non-threatening, vulnerable individuals without
social influence were selected to represent the potentially threatening mem-
bers of the community and then killed. The acceptance by the members of
the community of the supernatural justification demonstrated a willingness
to co-operate in their destruction, uninfluenced by relevant evidence. Vul-
nerable individuals were selected because they were the easiest to dispatch
Such killings of innocent, known individuals, however, should occur only in
special circumstances. Because of the problems associated with foresaking
rules of evidence within a community, witch- killings should be supported only
when something very important to many individuals is endangered. That
something, I suggest, may be the social relationships of the killers and their
supporters.

Salem

If the conditions elucidated in this paper account for European and Hewa
witch-killing, this phenomenon should occur only where there is a serious threa
to social relationships. Therefore, I now examine briefly the witch-hunt of
Salem Village, Massachusetts, in 1 692 and, despite enormous differences, the
McCarthy period in the U.S.A. in the 195O's, also described at times as a
witch-hunt.

There is no witch-hunt more fascinating to Americans than that of Salem

Village, occurring near the end of the two European witch-killing centuries.
Twenty persons were killed in one and a half years. Explanations offered often
focus on the delusions of the children who were the accusers (e.g., Starkey
1 969), and more recently an attempt has been made to explain such delusions
by what the children may have eaten (Caporael 1976). It should be clear,

117

This content downloaded from

154.59.124.59 on Sun, 06 Nov 2022 14:15:23 UTC
All use subject to https://about.jstor.org/terms
 The Killing of Witches

however, that the killing of witches in Salem cannot be explained by hysterical

fantasies of children. Explanations proposed must account for the behaviour of
the adult members of the community, including the magistrates and preachers,
who supported the killings: it was the adults, not the children, who did the
killing.
It is generally known that the Massachusetts Bay Colony was founded in
1630 by a group of religious dissenters for the expressed purpose of religious
freedom. What is not so well known is that within less than ten years these same
people were prosecuting a woman and her supporters for simply emphasizing
their own different religious views: in those few years the Puritan colony had
changed from an assembly of people who used a religious idiom to defy an
authority structure to a group who used such talk to protect one. These
Puritans had become an established community with a powerful authority
structure composed mainly of clergy, who created laws by interpreting the
Bible, and magistrates, who punished violations of these laws.
In 1636 a newcomer to the colony, Mrs Hutchinson, did no more than state
what was then current Puritan doctrine in England. She pointed out that as
each person was individually responsible to God, the clergy had no right to act
as God's intermediaries and interpret His messages: the interpretation of God's
messages, she argued, should be done only by individuals themselves. Mrs
Hutchinson was charged with a political crime, sedition rather than heresy,
tried (the governor John Winthop was both the prosecutor and the judge) and,
together with some of her supporters banished from the colony (Erikson
1966:71-107). The court charged her with having 'seduced and led into
dangerous errors many of the people here in New England' and, interestingly,
compared her to 'others in Germany' (witches?) who 'make some sudden
eruption upon those that differ from them in judgment' (Massachusetts
Records, 1, pp. 211-213, cited by Erikson 1966:92). Note the connection
between European witches and threats to the community.
This reaction against political-religious dissent appears to have effectively
muted further challenges for some time; for it was not until 1 659, twenty years
later, that the colony was again significantly threatened in this way. In this
year, some thirty years before the Salem witch killings, four Quakers were
hanged and many others imprisoned and beaten for allegedly doing no more
than wearing their hats before authorities (including magistrates) and
addressing the authorities by the familiar pronouns thee and thou. When
Charles II prohibited both corporal and capital punishment of Quakers (p.
1 24), the attempt by the colony to cope with the threat posed by them came to
an end. Fifty years later, this incident still aroused the famous witch hunter
Cotton Mather to write of the Quaker's actions as an 'intolerable contempt of
authority' (Erikson 1966:133). In this case, then, the community, relying on
actual evidence, failed to intimidate the category of people seen as threatening
it.

Thirty years later Salem Village supported and encouraged witchcraft

accusations and used them to justify the destruction of twenty innocent
individuals and the incarceration of many others. Partly as a result of their
earlier failure to intimidate dissenters legally, I suggest, this Puritan com-
munity was willing to charge hundreds of innocent people with collaboration
with the Devil in order to intimidate those who might indeed display 'an
intolerable contempt of authority'. For evidence of political and religious
dissent in Salem Village at this time one need only consider the series of events
which surrounded the witch-killings (Boyer and Nissenbaum 1 974:xvii-xviii):

118

This content downloaded from

154.59.124.59 on Sun, 06 Nov 2022 14:15:23 UTC
All use subject to https://about.jstor.org/terms
 L. B Steadman

1626 Founding of the town Salem.

1672 'Salem Farms', a region of Salem Town, becomes the
separate parish of Salem Village; James Bayley hired as its
first preacher.
1679 Bayley resigns amidst criticism by some Salem Villagers.
1680 George Burroughs hired as the new Village preacher.
1683 Burroughs leaves Salem Village.
1684 Deodat Lawson hired to succeed Burroughs as preacher.
1686-7 Futile effort to ordain Lawson and form a Village Church.
1688 Deodat Lawson leaves the Village; Samuel Parris arrives.
1689 April: Governor Edmund Andros overthrown in a coup at
Boston.
November: Formation of the Salem Village church and
ordination of Samuel Parris as its minister.
1691 October: Opponents of Parris win control of Salem Village
parish Committee.
1692 January to May: Witchcraft afflictions, accusations, arrests.
June to September: Witchcraft trials and executions.
1693 Parris' supporters and his opponents jockey for position.
1694 March: the pro-Parris group regains control of the parish
Committee.
1695 April: an ecclesiastical council meeting at Salem Village
under the leadership of the Reverend Increase Mather hints
that Parris should resign; eighty four of Parris' Village
opponents petition the council members to take a stronger
stand.
May: The council members recommend more forcibly that
Parris resign; 105 of Parris' Village backers sign a petition
on his behalf.
June: The Salem Village church endorses Parris, who agrees
to stay on.
1696 July: Resignation of Samuel Parris.
Thus, there is evidence of political and religious competition within Salem
Village, particularly for the religious leadership of the community. Such
evidence of an internal political-religious threat is consistent with our
hypothesis that witch-killing is an attempt to intimidate a category ol
people associated or claimed to be associated with the alleged witch (in
this case being Devil-like or anti-Christian), who actually pose a threat to
the social relationships constituting the community. The individuals killed
as witches, as in Europe and among the Hewa, were innocent and non-
threatening: most of the Salem twenty were vulnerable women.
McCarthy

The McCarthy period in American politics, from 1 950 to 1 954, while of

called a witch-hunt, was obviously quite different from the witch-hunts
sixteenth and seventeenth century Europe and Salem, as well as among th
Hewa, because no one was killed. Nevertheless, an analysis of this period h
distinctive attraction: many Americans living today have a detailed k
ledge of it gained by living through it.
The reason for the term 'witch-hunt' in this context, I suggest, is
individuals were accused and, although not killed, were damaged on the b
of their irrelevant or non-identifiable evidence. What were the conditions
which led to the willingness to abandon, if only temporarily, the normal legal
requirements for evidence?

119

This content downloaded from

154.59.124.59 on Sun, 06 Nov 2022 14:15:23 UTC
All use subject to https://about.jstor.org/terms
The killing of Witches

Virtually all adult Americans knew that in the five years or so before 1950
many Eastern European countries had been drastically affected by successful
Communist revolutions, which altered fundamentally their authority struc-
tures. Similar revolutions appeared imminent in Italy, France, Greece,
Finland, Iran, Indo-China and the Philippines.
Not even the United States seemed immune from such a revolution: it was
clear that in the labour movement, in the arts, and even in government, there
were Communists or Communist supporters. In response to this recognized
threat, President Truman, shortly after his inauguration, required more than
two million government employees to submit to loyalty investigations.
In 1945, the editor of Amerasia had been discovered holding thousands of
classified documents passed on to him by a State Department employee
(Feuerlicht 1972:45). In January 1949, eleven leaders of the Communist
Party went on trial under the Smith Act, and, in October, 1949, were found
guilty of conspiring to overthrow the government. Accused in 1949, Judith
Coplón, a political analyst in the Justice Department, was found guilty in 1 950
of passing secret documents to a Russian spy. In August, 1949, the State
Department announced that 4the heart of China is in Communist hands'. In
January, 1950, Alger Hiss was finally convicted of passing State Department
secrets to Communists (Feuerlicht 1972:5 1-53). Two days later, in response
to Hiss' conviction, Representative Richard Nixon asserted, 'this conspiracy
would have come to light long since had there not been a definite . . . effort on
the part of certain high officials in two Administrations to keep the public from
knowing the facts' (Feuerlicht 1972:53).
Ten days later Klaus Fuchs was arrested in England and accused of spying.
He admitted having passed information to Russian agents at various times
between 1 942 and 1 949. This case eventually led to the arrest and execution of
two Americans, Ethel and Julius Rosenberg, as 'atomic spies' (Feuerlicht
1972:54).
The fear of communism, then, was not simply a result of superstition or
irrationality: there was tangible evidence indicating an internal Communist
threat to the United States. In 1950, eighteen days after Hiss' conviction,
Joseph McCarthy told a group in Wheeling, West Virginia:
'I have here in my hand a list of two hundred and five that were known to
the Secretary of State as being members of the Communist Party and
who, nevertheless, are still working and shJping the policy in the State
Department'. These words . . . were printed in newspapers throughout the
nation (Lu thin 1973:8).
McCarthy almost immediately became the spokesman for a substantial
number of Americans. By 1954, according to Gallup Poll, 'Fifty percent . . .
were favourable to the Senator against only 29 percent unfavourable' ( Varney
1973:170). In this same year McCarthy was described by one journalist as
'the articulate voice of the American people' (Varney 1973:182).
From 1950 to 1954 hundreds of people were accused and dozens of
reputations destroyed. The accusations began with 'middle management'
State Department employees and worked through scholars, movie writers and
actors, librarians, high school teachers, and many others. McCarthy never
formally accused business executives, Senators, Congressmen, Cabinet
members, or the President. He 'was often careful to choose victims who were
either timid or vulnerable or both' (Rovere 1973:261).
In view of the almost instantaneous popularity of McCarthy, we must
assume that many Americans were ready to accept non- verifiable accusations

120

This content downloaded from

154.59.124.59 on Sun, 06 Nov 2022 14:15:23 UTC
All use subject to https://about.jstor.org/terms
 L. B. Steadman

of Communist activities.People were singled out as Communists on the basis

of such bizarre evidence as: being cited in a bibliography in a study of Siberian
folkways; having a library shelf containing detective stories written by a pro-
Communist writer; a boy friend's invitation to spend a weekend with him; or
promoting a dentist in the Army (Rovere 1973:39-40). Such 'evidence' of
communism was used to destroy the careers of many people because, in the
words of McCarthy, 'there is only one issue for the farmer, the laborer, and the
businessman - the issue of Communism in government' (Rovere 1973:41).
The identifiable effect of such witch-trials was the widespread intimidation of
potential Communist supporters, and their intimidating influence is felt still
today.
Conclusion
In this paper I have attempted to account for the killing or harming of
said to be witches. Let me state again the proposition. The essenc
witchcraft does not lie in the alleged practices of witches, but in the u
assertion alone to justify injuring or destroying innocent individuals. A w
an individual accused of being supernaturally evil. Because the accuracy
supernatural claim cannot be verified by the senses, the acceptanc
witchcraft accusation communicates support for the accuser's criticism o
intended damage to, the accused.
A fundamental problem in the analysis of witch-killing or damaging
account of why harm is done to individuals who are identifiably innoc
the crime of which they are charged. Because the killing of such indiv
is otten disruptive, promoting tear and threatening anyone equa
nocent, such killing should be resorted to and encouraged only in
tional circumstances. The hypothesis I am proposing is that
circumstances necessarily involve a situation in which individual
their social relationships are threatened by people in their midst
individual chosen as a witch is selected because he or she is, or w
recognized as being, associated with the category of individuals wh
the actual threat. Witchcraft accusations, therefore, communica
willingness to use violence against the threatening individuals: kill
strong and influential (or those close to them) is likely to lead to susta
violence and social disruption, whereas killing the vulnerable is les
to do this.
Further, the killing of a witch requires co-operation, for the person k
known by the accusers and, on the basis of evidence, is innocent. Any
course, can attempt to justify the killing or harming of someone by claimi
person was a witch, but such a claim is easily dismissed as self-serving,
killer dealt with simply as a killer. Consequently, a witchcraft justific
almost always widely accepted before an innocent, known indivi
harmed as a witch. Such a requirement thus has the effect of 'testing th
it permits individuals to voice either their objection to, or support for the
A prior accusation thereby allows the killers and their supporters to ant
the social consequences of a killing. Such a procedure helps to prev
possible destruction of existing social relationships, whose prote
am arguing, appears to be the aim of the killing.
That is, one identifiable and memorable effect of witch-killing, like
killing, is that it can intimidate those associated with the victim
identification and memory of this effect by those involved in witch-killing
inevitably influence any decision they make in regard to repeatin
activity. It is the attempt to intimidate a category of threatening

121

This content downloaded from

154.59.124.59 on Sun, 06 Nov 2022 14:15:23 UTC
All use subject to https://about.jstor.org/terms
The Killing of Witches

associated with the witch, I am suggesting, which constitutes the primary

aim of witch-killing. Witch-killing, therefore, should not occur in the absence of
this threat, and the identification of such absence would constitute a disproof of
this hypothesis. The hypothesis would be disproved also by showing either that
witch-killing does not often intimidate those associated with the witch, or that
the witch is not identified with the threatening category.
For the Hewa, the threatening category associated with the witch is a
neighbouring, hostile group. This threat is often reduced by the show of force
communicated by the witch-killing.
For European witch-killing, the major threat to the communities was an
internal religious revolution which threatened the existing authority structure.
This threat lasted two centuries, during which time individuals agreed to kill
innocent individuals, and justified such killing not by evidence, but by
supernatural assertion of anti-Christian (Devil-like) behaviour - the
behaviour which most threatened Christian communities.
Although the Massachusetts Bay Colony was itself the product of religious
strife in England, a major concern of the colony was threats to its own
political, religious hierarchy: first, in 1 640, posed by Mrs Hutchinson and her
followers, secondly, by the Quakers, twenty years later; and, finally, the
dissension as seen in Salem Village, culminating in the witch-killings.
For McCarthy's anti-Communist 'witch hunt' the threat was Communist
revolution, and Americans were willing to damage members of their own
country and community on the basis oP supernatural accusations' alone. The
primary aim of the hunt was to intimidate potential revolutionaries and
sympathizers.

Notes

1 The Australian National University generously supported my research in Papua N

Guinea. For assistance in preparation of this article, I am indebted to Melissa Johnson, E
Carson, James Eder, Brian Foster, and Donald E. Brown. Miss Johnson in partic
contributed greatly to both its structure and logic.

References

Boyer, P. and Nissenbaum, S. 1974. Salem Possessed. Cambridge, Massachusetts,

Harvard University Press.
Caporael, Linda R. 1976. Ergotism: The Satan Loosed in Salem? Science 192:21-26.
Cohn, Norman 1970. The Myth of Satan and His Human Servants. In Mary Douglas (ed.
Witchcraft, Confessions and Accusations. London, Tavistock Publications.
DOUGLAS, Mary, 1967. The Meaning of Myth, in E. Leach (ed.),77ie Structural Study o
Myth and Totemism. London, Tavistock Publications.
DOUGLAS, Mary. 1970. Introduction in Mary Douglas (ed.), Witchcraft, Confessions an
Accusations. London, Tavistock Publications.
ÜRIKSON, Kai 1966. Waywara Puritans: A Study in the Sociology of Deviance. New Yor
Wiley.
Feuerlicht, Roberta 1972. Joe McCarthy and McCarthy ism. New York, McGraw-Hill.
Harris, Marvin 1974. Cows, Pigs, Wars and Witches: the Riddles of Culture. New York,
Random House.
Howells, William 1962. The Heathens: Primitive Man and His Religion. Garden City, N.Y.
Doubleday and Co.

122

This content downloaded from

154.59.124.59 on Sun, 06 Nov 2022 14:15:23 UTC
All use subject to https://about.jstor.org/terms
 L. B. Steadman

Levine, Robert 1973. Culture, Behavior and Personality. Chicago, Aldine Publishing Co.
Luthin, Reinhard 1973. McCarthy as Demagogue. In Earl Latham (ed.). The Meaning of
McCarthyism. Boston, D. C. Heath & Co.
Mair, Lucy 1969. Witchcraft. New York, McGraw-Hill.
Malinowski, Brownislaw 1929. The Sexual Life of Savages. New York, Harcourt Brace and
World, Inc.
Middleton, John 1960. Lugbara Religion.London, Oxford University Press.
Midelfort, H. C. Erik 1972. Witch Hunting in Southwestern Germany 15Ó2-1084: me
Social and Intellectual Foundations. Palo Alto, Stanford University Press.
Patterson, Mary 1974. Sorcery and Witchcraft in Melanesia. Oceania XLV(2).
Rovere, Richard 1973. McCarthy: An Unfavourable Summary. In Earl Latham (ed.), The
Meaning of McCarthyism. Boston, D. C. Heath & Co.
Starkey, Marion L. 1 969. The Devil in Massachusetts. New York, Anchor Books, Doubleday
&Co.

Steadman, Lyle B. 1971. Neighbours and Killers: Residence and Dominance amon
Hewa of New Guinea.VhD. Dissertation; Anthropology Department, Australian Na
University Canberra.
Steadman, Lyle B. 1975. Cannibal Witches among the Hewa. Oceania 46:1 14-121.
Steadman, Lyle B. and Charles Merbs 1982. Kuru and Cannibalism? American An
ologist 84:611-627.
TREVOR-RoPER,Hugh, R. 1968. The Crisis of the Seventeenth Century: Religio
Reformation and Social Change. New York, Harper and Row.
Varney, Harold 1973. McCarthy: A Favourable Summary. In Earl Latham (ed
Meaning of McCarthyism. Boston, D. C. Heath & Co.

123

This content downloaded from

154.59.124.59 on Sun, 06 Nov 2022 14:15:23 UTC
All use subject to https://about.jstor.org/terms