This article examines accusations of sorcery as a way to understand the perceptions of sorcery among the Mapuche of central-southern Chile during the colonial period. Local communities believed that illnesses and unfortunate events were caused by the actions of sorcerers, known as kalku, and therefore consulted ritual healers and diviners, the machi and dugul, to identify and punish the supposed evildoers. In accusing local members of being kalku, the ritual specialists expressed a precise perception of sorcery and developed clear strategies for counteracting the sorcerers. This article argues that the accusations of sorcery became political and social instruments in the hands of the local authorities and that they shaped the meaning of sorcery and its perception among the Mapuche.
When the Spaniards crossed the Biobío River in central-southern Chile in the mid-sixteenth century, they encountered the semi-sedentary Mapuche. These hunters and horticulturalists, organized in small communities scattered in an area stretching from the Biobío to the Toltén Rivers, strongly resisted the Spanish conquest. After a widespread and organized rebellion, the Mapuche forced the Spaniards to abandon their towns and settlements to the south of the Biobío between 1598 and 1604. The Spanish Empire abandoned its attempt to conquer the southern Chilean territories, and the Mapuche maintained independence and sovereignty south of the Biobío until the Chilean republic conquered the region in 1883. In an attempt to slowly erode Mapuche resistance and introduce Christianity to the region, the Spaniards resorted to the help of the Jesuit missionaries who had to teach the doctrine, administer the sacraments, and extirpate local traditions as a first step toward the conquest.1 However, both the Jesuit missionary enterprise started in 1608 and the Franciscan efforts that began in the second half of the eighteenth century failed to spread the rites and precepts of Christianity in central-southern Chile. Among the major obstacles encountered by the missionaries, along with the widespread polygamy and a general lack of interest for the gospel, there was the belief that human agents, sorcerers known as kalku, caused illnesses and death.
Among the Mapuche, sorcery was surrounded by secretiveness and vagueness. Local communities explained every unfortunate event, from natural disasters to illnesses, as the result of the actions of the kalku. However, no Mapuche ever described himself or herself as a kalku, and the category of sorcerer remained vague and undefined for the entire colonial period. In the attempt to identify and punish the kalku responsible for causing illnesses and death, the kinship group of the deceased consulted local healers and diviners, the machi and dugul, expecting them to disclose the name of the culprit. This article explores the dynamics that surrounded the process of identification of the alleged kalku and focuses on the methods and criteria used by the ritual specialists for the accusation of the offenders. It shows that Mapuche ritual specialists tended to accuse members of specific social and gender groups of sorcery for social and political reasons. Mapuche authorities started to associate social marginality with sorcery, since the fragile and vulnerable were perceived as socially unproductive members. As a result, through the implementation of various forms of punishment, group leaders expelled from the community ostracized and marginal members accused of antisocial behavior. This article reveals for the first time that accusations of sorcery, especially from the beginning of the eighteenth century, became an instrument of social control in the hands of the religious and political authorities.2 Studying the process of accusation and punishment of the alleged kalku represents an effective instrument to explore the meaning of sorcery, its perception, and social effects in Mapuche society. How did the Mapuche perceive sorcery, and who were the kalku among central-southern Chilean Indigenous communities? How did Mapuche authorities identify the supposed sorcerers and defend the social order from their threats?
The first section explains the importance of sorcery in Mapuche society and defines the role of the different ritual specialists that counteracted the kalku. The article then examines the social importance of the accusations of sorcery and the Mapuche perceptions of sorcery.
Kalku, Machi, and Dugul: Causing Illnesses and Counteracting Sorcery in Mapuche Society
To understand the relevance of sorcery in Mapuche society, it is necessary to analyze who caused the illnesses and death and the relationship between the evil kalku and the benign machi and dugul. The Mapuche believed that sorcerers, kalku, manipulated evil spirits named huecubu who penetrated and possessed the bodies of people in the form of thorns, small sticks and arrows, worms, and lizards (Rosales  1877: 169; Febrés 1765: 506; Olivares 1864a: 179).3 Such external elements represented the physical manifestation of the huecubu and caused illnesses. To counteract the actions of the kalku and remove the origin of the illness from the body, it was necessary to resort to the healing abilities of the machi, shamans who cured patients with a ritual called machitun.4 Among the Mapuche, illnesses did not have a natural cause, but, rather, they were the result of the intervention of a human or supernatural agent. Only injuries and wounds caused by weapons were considered natural illnesses. Huecubu and kalku intervention even caused unfortunate events such as falling from a cliff or drowning in a river, and only in the second half of the eighteenth century did old age begin to be considered the natural cause of the death of elderly people (Núñez de Pineda y Bascuñán  2001: 453; Latcham 1924: 440; Bacigalupo 2001a: 95).5
Since the kalku were considered the cause of every illness in Mapuche society, it became essential to identify and punish them in order to expel dangerous members that could potentially continue to harm others. The function of identifying and accusing the kalku was carried out by the machi and, especially from the eighteenth century on, by the dugul, a term translated by the Spaniards as adivinos, diviners. However, the distinction between these three ritual specialists was rather nebulous. Certainly, the tendency of the colonial authors, mainly missionaries, to brand every activity performed by local shamans and healers as hechicería or brujería (sorcery) fostered a certain confusion about the actual ritual roles of these specialists.6 The word hechicería was used by the Spaniards to indicate the activities performed by ritual specialists who, with the help of the devil, harmed others both physically and spiritually (Real Academia Española 1726–39: 4.134). Although colonial authors soon started to report the presence of local healers that treated illnesses with divination and medical plants, and therefore became aware of the difference between evil and benign ritual specialists, they tended to perpetuate the idea of local shamans as devil worshipers because of their adamant resistance toward Christianity (Prieto 2011: 53).7 It is therefore difficult to identify in these sources clear information on the roles and differences between the kalku, machi, and dugul.
The term kalku appeared for the first time in Luis de Valdivia’s 1606 vocabulary translated as hechicero, and continued to be used to indicate malevolent and deceitful sorcerers during the colonial period (Valdivia 1606; Febrés 1765: 436; Havestadt 1777: 616; Gómez de Vidaurre  1889: 320). The kalku could be both male and female, and they were illustrated as mysterious entities living during the day in dark caves with monstrous creatures named ivumche.8 The ivumche were depicted in many different ways, clearly indicating the elusive nature of these supernatural entities. Sometimes described as dead human bodies without intestines ready to be filled by local deities or as men-beasts able to transform into nocturnal birds, the ivumche were evil spirits helping the kalku to foretell the future and to harm other Mapuche (Febrés 1765: 523; Gómez de Vidaurre  1889]: 320; Oña  1917: 83–84). Since the kalku caused death, locals feared them and consulted the machi when their relatives fell ill. Machi acted as medics and exorcists, and during the machitun ritual, they expelled the huecubu through the invocation of spirits and divination and tried to cure the patient with their excellent knowledge of surgery, bone setting, and medicinal plants (Bacigalupo 2007: 117). In the final part of the machitun, the machi reached a state of ecstasy and, from the supernatural spirits they contacted, obtained the name and location of the kalku (Leiva 1991: 47).
Although the Mapuche clearly distinguished between sorcery and healing, they still feared the powers of the machi and watched their actions closely. Like the kalku, machi interacted with supernatural spirits and possessed botanical knowledge that they could potentially use to either cure or harm others. This is the reason some machi were thought to be kalku (Foerster 1993: 25). As was common among other Indigenous communities in South America, the Mapuche perceived shamans as ambivalent figures that lived in a delicate balance between good and evil (Clastres 1995: 20–29). As Neil Whitehead and Robin Wright (2004: 4–10) observed, killing and curing were complementary opposites and not antagonistic possibilities in the shamanic experience. This dual ability nourished a perception of ambiguity that was, and still is, inherent in shamanic practice. Therefore, the difference between kalku and machi remained blurred during the entire colonial period.
Among the early Mapuche, the machi were the only specialists that dealt with sorcery and with the individuation of the kalku. However, late seventeenth-century sources reported the appearance of a new ritual specialist: the adivino. Although sources started to name this specialist with the local terms duguthue, dungube, llihua, and the most used, dugul, only in the middle decades of the eighteenth century, earlier authors already reported the presence of a different kind of ritual specialist able to discover sorcerers with divination.9 Since the machi also possessed divinatory abilities and used them both during the machitun and for the resolution of internal and military conflicts, it has been argued that the term dugul simply indicated a ritual specialization of the machi and not a different specialist (Bacigalupo 2001a: 106; Boccara 2007: 166).10 Sources frequently confused the roles of the machi and the dugul and sometimes treated them as identical. The Franciscan Antonio Sors ( 1921: 184) stated in 1780 that to identify those who caused illnesses, the Mapuche “consulted their adivinos or machis,” attributing the standard Spanish term used for the dugul to the machi, therefore equating the two ritual specialists. Nevertheless, the specific circumstances in which the dugul used their powers suggest that, more so than the machi, they ritually intervened on different occasions and therefore solved diverse issues.
Both the dugul and the machi used divination to identify the kalku. However, the dugul intervened only when death had already occurred. In 1717, the Jesuit Pedro Aguilar defined the dugul as “those that are thought to be adivinos, and that are summoned when someone dies in order to discover who caused the death.”11 Therefore, it seems they were not medics, and families paid them to discover the culprit when the healing skills of the machi were unsuccessful (Martínez de Bernabé  2008: 117). The dugul sometimes performed an autopsy as part of the funerary rituals and then accused the supposed kalku in front of the community, while on other occasions the relatives of the deceased consulted them in their caves, where, after entering in contact with supernatural entities and performing ritual dances, they disclosed the name of the sorcerer (Martínez de Bernabé  2008: 122; Ramírez  1994: 89).12
Accusation and Punishment of Sorcerers
Whether the family of a victim of sorcery resorted to the machi or the dugul, in both cases they expected the ritual specialists to disclose the name of the kalku. As stated by George Foster (1976: 778–79), non-Western personalistic medical systems, which explain disease as the active and purposeful action of a human or supernatural agent, are primarily concerned with diagnosis and the identification of the “who” and “why.” Mapuche patients and their respective relatives were more interested in identifying those responsible for the illness or death rather than learning its immediate cause. In fact, local communities already knew the origin of the disease, which was believed to be evil spirits and the actions of the kalku, and therefore mainly looked for their identification. The actions of a sorcerer disrupted the social order by spreading illnesses and death among the community. The social order needed to be restored by the punishment of those responsible. However, the persecution of the kalku caused violent strife among Mapuche communities, since those affected by sorcery looked to avenge their families by killing the kalku and those who protected them.
In early Mapuche society, the majority of the transgressions of the admapu, the set of traditional rules, could be solved with a transaction of goods decided by the genvoye, a powerful local leader who acted as a judge for the resolution of internal conflicts (Valdivia 1606; Vivar  1966: 160).13 The genvoye identified those responsible for a specific offense and decided on the compensation that had to be given to the victim’s family (Rosales  1877: 134). However, the genvoye rapidly disappeared from the sources around the second half of the seventeenth century. With the genvoye, the formal process of investigation and resolution of the offenses also seemed to leave space for a more informal judicial system now managed by the families involved in a conflict.14 Important economic changes, such as the breeding of nonlocal livestock and the growing commerce of ponchos with the Spaniards, along with the gradual occupation of the lands beyond the Andes, led to a redefinition of the sociopolitical structure of Mapuche communities by the end of the seventeenth century. Powerful ulmen, local leaders whose leadership and identity no longer depended on military skills but, rather, on commercial activities and political negotiations, obtained the juridical, spiritual, and military functions previously distributed between different members (Boccara 1999: 440–60). However, the ulmen rarely intervened to pacify internal conflicts, especially when the two parties involved were equally powerful and influential (Gómez de Vidaurre  1889: 325). The victim’s family directly requested compensation from the accused without the intervention of an impartial judge.
From the end of the seventeenth century, transgressions of the admapu triggered the following process. The principle of reciprocity inherent in Mapuche culture forced the relatives of the victim to obtain satisfaction from the family of the accused (Jiménez and Alioto 2011: 49–52). The leader of the family of the victim, known as the gen lladcùn, was responsible for contacting the leader of the family of the accused, the gen huerin, and for negotiating a solution that could avoid further conflict between the two families (Febrés 1765: 494). The gen lladcùn usually asked for the payment of a certain quantity of goods. If the gen huerin accepted, a peaceful relationship was soon restored. If the culprit was unable to pay the agreed sum, the relatives would have to contribute to its payment to the gen lladcùn (Rosales  1877: 134). However, if the gen lladcùn and the gen huerin could not reach an agreement, the victim’s family would organize a military party and attack the property of the rival group to obtain the payment requested. Such actions could cause the revenge of the gen huerin and trigger internal wars that, potentially, could last many years (Molina 1795: 67; Rosales  1877: 134; González de Nájera  1889: 49; Gómez de Vidaurre  1889: 325).
When ritual specialists deemed sorcery to be the cause of a death, usually the relatives of the deceased did not engage in formal parleys and agreements with the gen huerin, or asked for payment. In fact, deaths caused by sorcery justified the use of extreme violence, often ending with the death of the alleged kalku (Latcham 1924: 500). Although the Jesuit chronicler Diego de Rosales ( 1877: 134) clearly stated that punishment by death only occurred in cases when murderers and adulterers refused to pay the compensation, coeval observers of Mapuche communities proposed a different view. In 1648 Luis Pacheco, the vice-provincial of the Chilean Jesuits, reported that as soon as the machi informed the family of the deceased that the kalku that supposedly poisoned their relative with an invisible arrow (rapun guecubu) had been identified, the gen lladcùn would soon gather some men and slay the presumed sorcerer.15 There was no formal encounter or negotiation with the gen huerin, and those who tried to defend the person indicated by the machi as being a kalku would incur the same fate.
The accounts written during the sixteenth and seventeenth centuries rarely reported clear information on the social status of those accused of sorcery. Although the Jesuit Alonso de Ovalle ( 1969: 347) stated in 1646, “Some machi had a reputation for being sorcerers,” there is no clear evidence of shamans being accused and executed for sorcery in this period. Both the soldier Alonso González de Nájera ( 1889: 49) and the Jesuit Rosales ( 1877: 169) suggested that machi literally accused whom they wanted to. Therefore the accusations of sorcery seemed to be arbitrary and unilaterally decided by the machi. Moreover, the precise account of the healing ritual of the machitun witnessed by Francisco Núñez de Pineda y Bascuñán in 1629 reflects the same reticence of coeval sources. Asked by the relatives of the ill person to reveal the name of the kalku, the machi simply refused to disclose it (Núñez de Pineda y Bascuñán  1863: 159–61).
It was during the eighteenth century that missionaries and soldiers started to report new details on the dynamics activated by sorcery and on the solutions adopted locally. Sources showed that the ritual specialists that were asked to reveal the name of the supposed kalku started to follow precise criteria in the accusations of sorcery.
The Franciscan missionary Ramón Redrado, writing from the Arauco mission in 1775, reported that when consulted by the family of the gen lladcùn about a death supposedly caused by sorcery, the dugul initiated a personal investigation of the deceased. The adivinos started to gather information on the circumstances in which the deceased fell ill and carefully reconstructed his or her personal relationships with other locals. In particular, the dugul asked the members of the community and the relatives about any former animosity, conflicts, and rivalries that the deceased might have had with other Mapuche.16 Since, according to local beliefs, any Mapuche could hire a kalku to harm others with whom they had conflict, it was credible and logical to accuse the enemies of the deceased of being sorcerers (Carvallo y Goyeneche  1876: 138). Therefore the gen lladcùn would look for the enemies of the dead relative and punish them with death (Gómez de Vidaurre  1889: 319).
In an attempt to find explanations for the rationale behind the accusations made by the machi or dugul, some eighteenth-century observers adapted the early colonial leitmotif of the arbitrariness of the decisions taken by the ritual specialists. According to the missionaries, the machi could take advantage of their ritual role and use the accusations of sorcery to exact revenge on their enemies and rivals, who were subsequently killed or expelled from the community (Molina 1795: 106; Olivares 1864a: 46; Gómez de Vidaurre  1889: 325; Maran in Hanisch  1990: 139). If the machi had no enemies, they usually accused the foes of the ill or deceased person, but sometimes they could even avoid giving precise details on the supposed kalku (Molina 1795: 106). However, the gen lladcùn could often sway the decision of the machi and direct the accusation against a long-time enemy of his kinship group. In fact, while the machi was gathering information on the deceased, the gen lladcùn could strongly suggest the name of the supposed kalku and lead the investigation toward the direction indicated by the family (Olivares 1864a: 46). As confirmed by modern anthropological research on the Mapuche, the accusations of the machi tended to corroborate community gossip and especially family interests (Bacigalupo 2005: 332). Since the gen lladcùn always paid a considerable amount of goods for the intervention of the ritual specialists, accusations of sorcery tended to be subjective, biased, and contextual.17
A careful study of the sources shows the emergence of a clear pattern in terms of accusations of sorcery from the beginning of the eighteenth century. As much as ritual specialists tended to accuse their personal enemies or the rivals of the gen lladcùn, kalku started to be identified among the poorest and most marginal members of the community.18 The Franciscan Redrado, in fact, while explaining how the dugul investigated the enemies of the deceased, reported that the supposed kalku regularly tended to be the most fragile and vulnerable.19 Both Jesuit and Franciscan missionaries frequently reported the emergence of a growing tendency to accuse the poor of sorcery. Sources suggest that the enemies of the machi, dugul, or gen lladcùn often coincided with the defenseless or that, in the absence of clear enemies, the responsibility for sorcery fell on the poor (Olivares 1864a: 46; Sors  1921: 186). Although colonial chroniclers viewed such allegedly arbitrary decisions as a lack of judicial justice and used it to perpetuate the idea of a chaotic and superstitious Mapuche society, there were cultural reasons that suggested choosing the supposed kalku among the poor and marginal people.20
Sources used the terms indefensos and desvalidos to identify those impoverished members accused of being kalku. Such words, literally translatable as “defenseless” or “helpless,” did not necessarily convey the concept of material poverty but, rather, a concept of social isolation and marginality (Real Academia Española 1726–39: 3.240, 4.249). In Mapuche society, those who were indefensos and desvalidos were those unable to defend themselves from the threats made by others. However, the impossibility of defending themselves did not reflect a lack of personal military skills but, rather, the absence of an extended kinship group inside the community that could support their members when necessary. The defenseless were isolated and marginal members who did not manage, because of their material poverty, to establish social relations through the instrument of marriage. The political alliances formed through marriage, in fact, allowed local members to extend their social, political, and commercial influence inside and especially outside their community (Latcham 1924: 283–301; Ortelli 1996: 206–7). Such alliances meant that individual members became part of extended and complex groups that could defend their members in the case of conflicts with other families. What was, therefore, the advantage of accusing the poor and outcast members of the community of sorcery? The desvalidos and indefensos, as marginal members with limited familial relations, were an ideal target because they could not offer resistance to the offenses received. They simply had no powerful relatives that could avenge the violent actions of the gen lladcùn.21 In this way, the victim’s family could obtain satisfaction for the death of their relative without triggering internal strife and causing unnecessary deaths, while the community still managed to neutralize the actions of the supposed kalku.
Accusing marginal members of sorcery became a strategy developed by the religious and political authorities to reduce the high level of internal conflicts and limit their destructive demographic effects. Although the available sources are extremely reticent on the demographic evolution of Mapuche communities during the eighteenth century and do not specify the real impact of such internal conflicts, some accounts reported an extremely high child mortality rate at the beginning of the nineteenth century. According to the Franciscan missionary Melchor Martínez, 75 percent of Mapuche children died before reaching the age of nine.22 The high child death toll probably helped foster the creation of demographic regulative instruments in the middle decades of the eighteenth century, in addition to the fact that the conflicts activated by sorcery frequently led to the killing of entire families, women and children included (Olivares 1864a: 46; Maran in Hanisch  1990: 140). However, accusing the outcast members of sorcery was also related to the increasing poverty of Mapuche commoners in the second half of the eighteenth century. Sors ( 1921: 183) stated in 1780 that the Mapuche were “all poor and with very limited properties,” with the exception of a few who possessed livestock. The emphasis on poverty made by the Jesuit Andrés Febrés (1765: 231) in his manual of confession published in 1765, which contrasts with the superficial treatment of this theme in the 1606 manual of Luis de Valdivia (1606: 23), confirms the increasing impact of poverty in eighteenth-century Mapuche society. The presence of larger groups of marginal people therefore explained the new pattern used by ritual specialists for the accusations of sorcery.
Mapuche authorities started to perceive poverty, social marginality, and the lack of extended familial relations as antisocial behaviors. As modern Mapuche associate excessive poverty and excessive wealth with social isolation and sorcery, in a similar way eighteenth-century Indigenous communities started to identify the outcasts as members that challenged the idea of social reciprocity. In fact, their isolation and poverty made them unable to contribute their share of material goods to the rituals and events of the community (Bacigalupo 2001a: 93; 2005: 323). The impossibility of fulfilling the social duties prescribed by the admapu transformed them into ambiguous members that could be easily stigmatized as kalku. Therefore, accusing marginal people of sorcery was a way to restore the social order while removing members who were perceived as socially unproductive. In 1795 the Jesuit Juan Ignacio Molina (1795: 106) effectively summarized the Mapuche attitude toward sorcery, reporting that the machi and dugul “carefully avoided accusing the members of powerful and influential families” of being kalku.
As the accusations of sorcery started to fall on the marginal members of Mapuche communities, local authorities gradually changed the way they treated the supposed kalku. During the first half of the eighteenth century, as in earlier times, death remained the most common punishment for those accused of sorcery, and only in rare cases would the gen lladcùn agree to spare the life of the kalku and request a payment in the form of material goods as compensation.23 The forms of the death punishment could vary. In the region of Valdivia and especially to the south of the Toltén River, the party of the gen lladcùn would slit the throat of the alleged kalku along with his or her family, while in the area of Arauco the sorcerers were usually hanged (Olivares 1864a: 46).24 Other sources reported that sorcerers were burnt alive with their few possessions.25 However, if a kalku allegedly caused the death of an ulmen, the family of the gen lladcùn would capture and interrogate the sorcerer, attempting to obtain a confession and the names of possible accomplices. The kalku was therefore tied to three wooden poles planted in the ground while a fire was lit under his or her body. After obtaining the confession, the relatives of the gen lladcùn usually stabbed the victim in the chest and then tried to identify and punish the other sorcerers involved in the death of the ulmen (Molina 1795: 64; Olivares 1864a: 46; Gómez de Vidaurre  1889: 325–26; Martínez de Bernabé  2008: 116).
Although punishment by death continued to be performed even at the beginning of the nineteenth century, Mapuche communities began to sell the alleged kalku to the Spaniards in exchange for goods in the second half of the eighteenth century. The circulation of men and women, usually Spanish captives, between different Mapuche communities from the end of the seventeenth century is well documented (Boccara 1999: 441). However, the circulation and selling of the supposed kalku to the Spanish forts and towns is less known. Although some seventeenth-century sources reported the practice of selling the relatives during times of famine to the Spaniards of the forts of Toltén and Boroa in exchange for horses, clothes, and weapons, the introduction of the kalku in this commerce was first reported only in 1755.26 In the district of Valdivia, Spanish merchants, traders, and soldiers obtained the supposed kalku in exchange for payment in the form of material goods to the gen lladcùn.27 The Spaniards often found the alleged sorcerer still tied to the wooden poles and proposed to the Mapuche that the kalku be freed. Along with the payment, the family of the gen lladcùn asked the Spaniards to permanently remove the kalku from their communities. The Spanish merchants could then use the kalku as servants or sell them to other residents of the area of Valdivia. The migrant Mapuches were then baptized and forced to serve for ten years in their new master’s house (Martínez de Bernabé  2008: 116). After that period, many of them married locals and embraced Catholicism, but they never returned to their communities, fearing that they would be executed for being considered kalku.28
The political authorities of Valdivia justified this practice by explaining that the Spanish intervention saved the life of the alleged kalku from the punishment of the gen lladcùn and guaranteed their spiritual salvation through baptism.29 However, the real reasons behind this practice had little to do with religious piety. Both the Spaniards and the Mapuche saw clear advantages in the implementation of such commerce. The Spaniards, in need of servants for the dwellers of the remote southern forts, obtained a constant supply of migrants from the Mapuche communities and implemented forced labor. On the other hand, the Mapuche authorities and especially the gen lladcùn saw the possibility of obtaining an economic as well as social advantage in the selling of the supposed kalku. In this way, those who were perceived as undesired and dangerous members of the community, especially the marginal people or the rivals of the ulmen, were expelled in exchange for a payment. Those who were deemed unproductive members unable to contribute wealth to the community, and therefore participate in the dynamics of reciprocity, could now guarantee an economic gain as compensation for their acts of sorcery. Along with the profit, this practice allowed a smooth and permanent expulsion of the antisocial members of the community, reduced internal conflict, and restored the social balance. It is in this context of expulsion and forced migration of the supposed kalku that the Mapuche started to perceive sorcery as a transitory state intimately related to the community where the kalkutun, sorcery, was performed. In fact, the Mapuche believed that when the supposed sorcerers managed to escape from their community and found refuge among the Spaniards, they lost their status of kalku, as if kalkutun could exist only inside Mapuche society (Sors  1921: 186).
The marginal people and the rivals of the Mapuche authorities were not the only categories that were accused of hiding kalku in their ranks. Women, too, during the mid-eighteenth century became a clear target of accusations of sorcery. Especially in the area of Valdivia, wives and mothers became more exposed to the accusations of kalkutun, and local men and husbands frequently stabbed and burnt Mapuche women they thought were kalku (Martínez de Bernabé  2008: 132).30 Sources also reported that the vast majority of the sorcerers that were sold to the Spanish merchants were women (173). The association between women and kalkutun became so widespread among southern Chilean Indigenous groups that in 1789 the Jesuit chronicler Felipe Gómez de Vidaurre ( 1889: 320) translated the word kalku as brujas, witches. For the first time in the colonial documentation, the kalku were openly identified with the feminine gender, a clear sign of the association between women and sorcery made by late eighteenth-century Mapuche. According to Bacigalupo (2007: 130), the association between women and kalku became part of Mapuche popular belief in the wake of the numerous proceedings held by the Spaniards against Indigenous women accused of being witches (Casanova 1994: 139–50). Sources are too scant to allow a thorough analysis of the issue. However, it is worth noting that such an association coincided with the growth of the number of female machi in the mid-eighteenth century and the simultaneous decline of male machi (Bacigalupo 1995: 52–55). Although there is no evidence in this period that women who were expelled or burned as kalku were machi, the ambiguous nature of the Mapuche ritual healers certainly contributed to concentrate local accusations of sorcery on women. The association between women and kalku that started in the eighteenth century culminated in modern Mapuche society, in which women are associated with sorcery and are considered the most able sorcerers by virtue of being females (Faron 1962: 1159).
The perception of the kalku as evil entities able to disrupt the social order from within the community and capable of hiding even behind relatives, husbands, and wives extended to children. In 1806, the Franciscan Melchor Martínez reported that the dugul frequently accused children and even infants of being the cause of the death of community members, and, for this reason, they were executed.31 However, Martínez is the first colonial author reporting such cases of supposed child sorcery, suggesting that these kinds of accusations were not widespread among the Mapuche and probably began only at the beginning of the nineteenth century.32 Although the Jesuit missionaries reported during the first half of the eighteenth century the sacrifices of children and infants ordered by the dugul to appease the deities in the aftermath of dramatic natural disasters, such as earthquakes and seaquakes, it does not seem that the infants chosen were considered kalku. However, it is worth noting that the criteria used for the identification of the child that had to be sacrificed resembled those followed in the cases of sorcery. The chosen child had to be the son of an unmarried woman, a choice, once again, justified by the necessity to avoid internal conflict raised by the sacrifice and further motivated by the association between social marginality and sorcery.33 In the absence of a husband and of his kinship group, a single woman was unable to defend herself from the decisions of the Mapuche authorities and could do little to avenge the death of her son.
Sources reported the evolution of the perception of kalkutun from the point of view of the accusers, while the voices of the supposed kalku remained largely unheard. However, there are glimpses in the documentation showing a growing skepticism toward the activities of the ritual specialists among eighteenth-century Mapuche communities. Jesuit authors were the first to report the emergence of a certain local disapproval of the dugul actions. In 1717 Pedro Aguilar, the missionary of Dogll (on the Toltén River) reported that a large number of locals were convinced of the falseness of the adivinos’ divinations, and most of them experienced it firsthand when consulting the dugul about the death of a relative, as they often ended up being accused by him of being kalku. However, Aguilar reported that those skeptical members continued to resort to the dugul, since this was part of the admapu and especially because, if they had decided not to consult the adivino, they would have been accused of being indifferent to the matter or even suspected of causing the death. In fact, the community would believe that those who did not ask for the dugul’s intervention failed to do so because they feared discovery.34
Later religious sources reported that a growing percentage of the Mapuche started to brand the healing rituals and the art of divination as nonsense, and they considered the dugul to be deceivers serving the interests of those who paid them (Falkner  1836: 103; Gómez de Vidaurre  1889: 320). Interestingly, the supposed criticism of Mapuche commoners against the dugul very much resembled the missionaries’ opinions of the ritual specialists. It is possible that such an approach was part of a missionary strategy intended to show the erosion of the main local cultural aspects that contributed so long to the Mapuche resistance against the Spanish conquest. In the context of an extremely difficult introduction of the most basic of Christian principles, missionaries saw the supposed detachment between commoners and their religious authorities as a sign of the success of the missionary endeavor. The considerable decrease in the number of duguls at the beginning of the nineteenth century might suggest that the skepticism reported by religious sources was widespread and therefore caused the decline of the adivinos.35 However, the local population still requested their intervention and walked many miles to consult them. In addition, the majority of the Mapuche who migrated to the Spanish forts after being accused of kalkutun, admitted to the sorcery, not because they committed it but because the dugul said so.36 According to the adivinos, the kalku could, in fact, hypnotize people at night and manipulate them to harm others (Latcham 1924: 502). The dugul frequently convinced the accused members that they acted in an unconscious state possessed by evil entities, and therefore they could not remember the facts. Despite a possible erosion of the dugul powers, these ritual specialists still maintained a crucial role in explaining and counteracting evil actions in Mapuche society.
Sorcery was a pivotal element in Mapuche culture, since it was believed that illnesses and death were caused by the actions of kalku that harmed others with the help of evil spirits. With the exception of violent deaths that occurred during military skirmishes, it was believed that every unfortunate event had the intervention of a human agent. To discover the culprits, the Mapuche resorted to the help of the machi, who, with divinatory abilities and medical remedies, could expel the huecubu that entered the body of the patient. The Mapuche believed that kalku operated from within the community to destroy it with illnesses and death, which caused social disorder. Although early colonial Spanish authors depicted the kalku as sorcerers hidden in their caves during the day that sent the evil huecubu to harm others, the cases reported above demonstrate instead that locals believed that sorcerers were ubiquitous. In the vast majority of cases, the kalku were not distant but rather close; they were embedded in the community, and as integral members they could erode local social bonds attacking their own relatives. In addition, since evil spirits could possess everyone, the presence of sorcerers was unpredictable: they could hide behind wives, husbands, children, or relatives. For this reason, the community consulted the dugul to identify the sorcerers.
The admapu prescribed that offenses and transgressions had to be resolved with a transaction of goods or with a compensation given to the family of the offended. However, sorcery was the most serious offense in Mapuche society, a transgression that seriously endangered the community and that broke every previous social bond. The only solution to acts of sorcery was the death of the culprit. The death of a community member always triggered violent internal strife, since the family of the dead person would have reacted to the offense received. This study of the Mapuche judicial system shows that the accusations of sorcery made by the ritual specialists played a crucial role in the balance of local communities. In particular, the machi and dugul developed a strategy for the individuation of the kalku and showed a clear perception of sorcery. As much as the kalku could, at least in theory, hide behind nearly every Mapuche, only specific groups were accused of practicing sorcery. The enemies of the ill patient were easily stigmatized as kalku. In fact, sorcerers were believed to harm others they had disagreements with, or they could be hired by locals to harm their relatives for economic or political reasons. When it was impossible to identify the rivals of the deceased, the ritual specialists accused their personal enemies or the enemies of the gen lladcùn. Accusations of sorcery became effective instruments in the hands of the ritual and political authorities for the elimination of personal rivals or members judged as detrimental.
The Mapuche never accused members of powerful families of sorcery but, rather, started during the eighteenth century to accuse the poor and marginal members of the community. An increasing percentage of poor commoners who were unable to contribute wealth to the community became associated with sorcery. Social marginality and the lack of extended familial bonds obtained through multiple marriages transformed the poor into antisocial members accused of harming the community. The perception of social marginality and excessive poverty as a clear sign of sorcery was also a consequence of the demographic situation of late eighteenth-century Mapuche communities. An extremely high child mortality rate and the frequent internal conflicts threatened the survival of the Mapuche. Accusing the poor of sorcery was a way to limit conflicts and therefore population decline, since marginal members were helpless against the actions of the gen lladcùn, and they had no relatives that could avenge them. Accusing the outcasts and poor allowed Mapuche authorities to extirpate unproductive people and restore the social order based on the admapu without violent consequences. The association between social marginality and sorcery contributed to the development of a new form of punishment of the kalku: they were sold to Spanish merchants in exchange for goods, a system that guaranteed a smooth and permanent expulsion of the undesired members along with a profit.
In the second half of the eighteenth century, Mapuche women became a target of the accusations of sorcery probably because of the existing association between women and sorcery in the Spanish towns. However, since the poor remained the group most exposed to these accusations, it is probable that the women who were accused, expelled, and sometimes killed for being sorcerers were unmarried and poor. Although late colonial religious sources reported a growing skepticism of Mapuche commoners toward the methods and strategies used by the ritual specialists to fight sorcery, the machi and dugul cultural explanations of the origin and cause of illnesses remained widely accepted.
The study of the accusations of sorcery as a tool of political negotiation within Mapuche society in the period analyzed has been overlooked by the extant literature. Ana Mariella Bacigalupo (2001a, 2007) reached similar conclusions on the modern Mapuche, stating that accusations of sorcery help control the behavior of local members.
Among the multiple sources referring to the relationship between the kalku and the huecubu, see Archivum Romanum Societatis Iesu, Roma (hereafter ARSI), Provincia Chilensis 6, fol. 264v, “Letras Anuas de la Viceprovincia de Chile del año de 1649,” 17 December 1649; and Archivo del Arzobispado de Santiago de Chile, Santiago de Chile (hereafter AAS), Fondo Gobierno, vol. 19, fol. 22, “Miguel Vazquez Vazconcelos al Ilustrísimo Obispo de la Concepción de Chile,” Repocura, 4 July 1717.
A thorough analysis of the machitun healing ritual is beyond the scope of this article. The machitun has been widely studied by both colonial and contemporary authors. For a classic description, see Núñez de Pineda y Bascuñán ( 1863: 159–61). For contemporary anthropological studies, see Arturo Leiva (1991: 43–49) and Bacigalupo 2001b, 2007.
See also Archivo Franciscano de Chillán, Chillán (hereafter AFCH), Asuntos Varios, vol. 3, fol. 252, Ramón Redrado, “Relación de Yndios de las dos jurisdiciones de Chile y de Valdivia y de sus inclinaciones errores, y costumbres,” Arauco, 10 May 1775.
Ivumche is a compound word made up of ivum, small animals, and che, people (Latcham 1924: 538).
Rosales ( 1877: 169) simply referred to these specialists as adivinos. Jerónimo Pietas (in Gay  1846: 487) used the term dungube, while Miguel de Olivares (1864b: 493) slightly changed the word into duguthue before Andrés Febrés (1765: 300) finally introduced the most-used forms llihua and dugul.
For the opposite opinion, see Ricardo Latcham (1924: 536).
AAS, Fondo Gobierno, vol. 19, fol. 42, “Carta de Pedro de Aguilar al Señor Obispo de la Concepción,” misión de Dogll, 28 May 1717.
For a clear description of the activities performed by the dugul for the identification of the kalku, see also AFCH, Asuntos Varios, vol. 3, fol. 252, Ramón Redrado, “Relación de Yndios.”
On the role of the genvoye among the Mapuche, see Guillaume Boccara (1999: 434).
ARSI, Provincia Chilensis 6, fol. 247, Luis Pacheco, “Adjunta a Letras Annuas de la Viceprovincia del Reino de Chile desde el año de mil y seiscientos y cuarenta y siete hasta el presente de 1648,” 1648.
AFCH, Asuntos Varios, vol. 3, fol. 252v, Ramón Redrado, “Relación de Yndios.” Nowadays, the machi follow the same process in their attempt to understand the causes of illnesses (Bacigalupo 2005: 332).
Among contemporary Mapuche, accusations of sorcery continue to be subjective (Bacigalupo 2001a: 92).
This tendency is still evident in modern Mapuche society, where those accused of sorcery often belong to families that are marginal to the sociopolitical hierarchy (Bacigalupo 2005: 323).
AFCH, Asuntos Varios, vol. 3, fol. 252v, Ramón Redrado, “Relación de Yndios.”
Colonial authors frequently described the Mapuche as chaotic communities lacking political centralization, social order, and the Christian God (Boccara 1999: 427–29). For the persistence of the identification of Indigenous populations as brutal savages and uncivilized groups during the Enlightenment, see Weber 2005.
AAS, Fondo Gobierno, vol. 19, fol. 23, “Miguel Vazquez Vazconcelos al Ilustrísimo Obispo.”
Archivo de España de la Compañía de Jesús, Alcalá de Henares (hereafter AESIA), PP. Escritores, Caja Astrain IV, legajo 17, 12, fol. 110, Melchor Martínez, “Memorial,” 15 May 1806.
AAS, Fondo Gobierno, vol. 19, fol. 42, “Carta de Pedro de Aguilar al Señor Obispo.”
AFCH, Asuntos Varios, vol. 3, fol. 252v, Ramón Redrado, “Relación de Yndios.”
AAS, Fondo Gobierno, vol. 19, fol. 23, “Miguel Vazquez Vazconcelos al Ilustrísimo Obispo.”
Biblioteca Nacional de Chile, Santiago de Chile (hereafter BNC), Fondo Manuscritos Medina, vol. 142, fol. 32, Antonio Ramírez de Laguna, “Carta a SM,” 30 June 1652.
BNC, Fondo Manuscritos Medina, vol. 188, fol. 150, “Carta de Ambrosio Saez de Bustamante a SM.”
BNC, Fondo Manuscritos Medina, vol. 188, fol. 150, “Carta de Ambrosio Saez.” The Franciscan Redrado is the only author that mentions such a practice in the region of Arauco. On the forced migration of the Mapuche toward the Spanish forts and towns, see Contreras Cruces 2018.
BNC, Fondo Manuscritos Medina, vol. 188, fol. 150, “Carta de Ambrosio Saez.”
See also AESIA, PP. Escritores, Caja Astrain IV, legajo 17, 12, fol. 93v, Melchor Martínez, “Memorial.”
AESIA, PP. Escritores, Caja Astrain IV, legajo 17, 12, fol. 93v, Melchor Martínez, “Memorial.”
It is worth noting that nineteenth-century Franciscan missionaries were also the first to report cases of child sorcery and execution of child sorcerers among the Arawak-speaking peoples of Eastern Peru (Santos-Granero 2004: 274).
AFCH, Asuntos Varios, vol. 1, fol. 4, Martín de Recabarren, “Carta al Señor Presidente,” 17 November 1738. Cases of children sacrificed to appease the deities in the aftermath of earthquakes continued to be documented during the twentieth century (Foerster 1993: 127; Dillehay 2007: 198).
AAS, Fondo Gobierno, vol. 19, fol. 42, “Carta de Pedro de Aguilar al Señor Obispo.”
AESIA, PP. Escritores, Caja Astrain IV, legajo 17, 12, fol. 93, Melchor Martínez, “Memorial”; AFCH, Asuntos Varios, vol. 3, fol. 252v, Ramón Redrado, “Relación de Yndios.”
AFCH, Asuntos Varios, vol. 3, fol. 252v, Ramón Redrado, “Relación de Yndios.” Latcham (1924: 501) recorded at the beginning of the twentieth century the case he witnessed of a man accused of sorcery that confessed to be a kalku even though he did not remember having committed kalkutun. The reason he gave to the anthropologist was that the machi had no reason to accuse him if he was not the culprit, because the machi was an honorable man.