Acknowledgements – I would like to thank Stefan Dienst and Cindy Boyer, who provided me with invaluable help in clarifying and explaining several questions related to the Madiha language. I am grateful to Philippe Erikson, Luiz Costa and David Jabin for their insightful advice regarding previous versions of this article and for being revitalizing interlocutors. Philippe Erikson is also thanked for his kind help in suggesting the evocative formula that opens the title of the article. Finally, I am thankful to Chloe Nahum-Claudel and Olivier Allard, whose careful advice and comments helped me to improve this text and clarify its content.
- 1 The Madiha are known in the existing literature under the name of “Kulina” or “Culina.” I use the (...)
- 2 I will analyze the term hinede in detail below. Note that, while the plural of hinede is hinededen (...)
1This article provides an ethnography of pet keeping among the San Bernardo Madiha people.1 Like many other Amazonian societies (Erikson 1987, 2000), the Madiha capture animals in the wild and raise them in their villages as pets. In what follows, I will account for the particular ways in which the human-pet relationship unfolds among the Madiha people, from the moment animals are introduced into the domestic space to the moment they leave it. I emphasize two series of seemingly contradictory facts. First, I will demonstrate the intrinsic fragility of the process of pet keeping due to its constantly being subject to disruptions of various kinds, and due to the animal’s growth into adulthood (in the rare cases where it happens) sanctioning the dissolution of the pet keeping relation, the latter being conceptually circumscribed by the act of “causing to grow” (this notion is expressed by a series of terms; see below). Second, I will show that pets play an important role in the socialization of children by transforming them into “masters/owners” (hinede),2 hence placing them in the socially valued position of being responsible for others. In short, pet keeping relations are precarious and reversible, but also socially significant and, as we shall see, inherently related to the concept of “growth.”
- 3 I thank David Jabin for encouraging me to reflect on the importance of this contrast.
2In my perspective, the combination of these facts has two main interrelated implications. On the one hand, it suggests that what is most attractive to the Madiha in the activity of pet keeping is less the perspective of producing a fully grown adult pet than the desire to constantly (re)initiate a process of taming. In other words, the process of taming, or familiarizing, pets in and of itself, and regardless of its final outcome, seems to be what is most valued. This explains the almost total absence of adult pets in the face of the constant presence of newly captured younglings, which otherwise remains inexplicable.3 On the other hand, and consequently, the combination of the facts mentioned above adds complexity to the adage that conceives of the taming of wild animals into pets as a form of adoption. Indeed, existing studies on pet keeping practices in Amazonia have stressed the extent to which adoptive filiation constitutes the relational scheme whereby pets are inserted into the human social sphere, thereby equating pets to adoptive children (Descola 1994; Erikson 1987, 2000; Fausto 1999, 2008, 2012a). This analogy is evidenced, for instance, by the fact that in many Amazonian languages (including Madiha) the term for “pet” and for “adopted child” is the same (Erikson 1987, 2000). The fact that, as mentioned, pet keeping among the Madiha is intrinsically fragile invites further reflection on what kind of adoption pet keeping would be, and suggests that pet keeping relations can be modeled after relations of adoptive filiation without necessarily aiming to create such relations. Otherwise, it would be impossible to account for the ease and nonchalance with which, as we will see, ties with pets are undone, broken and destroyed.
3In the following pages, I aim to develop and sustain, through a detailed ethnography, the main points outlined in this brief introduction. The first part of the article discusses the spectrum of animals kept as pets by the Madiha. I observe that, whereas the Madiha consider many species to be tamable, the actual taming most often concerns only a handful of such species. Next, I show how pet ownership articulates with pet feeding and with child making. I argue that pet ownership and feeding are not by necessity superimposable acts, and that pet keeping is a technique of person making. Later, I go on to discuss the extent to which the techniques of familiarization of pets vary depending on the animal in question. I then stress the extent to which pet keeping is a practice tied to the notion of “growth” and, as such, an essentially temporary and reversible relation. I provide further evidence of the temporariness and reversibility of this relation through an account of how the Madiha relate to their chickens and dogs, underlying the differences and similarities between such animals and the younglings captured in the forest. Finally, I propose a comparison between pet keeping and child adoption, in order to emphasize the similarities and differences between the two. In addition to drawing on my own ethnography, I build on, and enter into dialogue with, the work of anthropologists who have dealt with pet keeping in the region, in order to situate the Madiha case in relation to other Amazonian societies.
- 4 The dénichage (literally translated as the finding or unearthing) of hatchlings is carried out eit (...)
4Pets-to-be are usually captured by Madiha men during trips into the forest. Besides being the infants of adult specimens killed during hunting (in the case of mammals) or being randomly found while walking in the forest (in the case of tortoises, for instance), pets-to-be are also actively sought out. Such is the case for parrots, especially tui parakeets, whose nests, once located, become the object of expeditions of dénichage, with the purpose of appropriating the hatchlings.4 Whatever the type of animal, the essential condition for it to be considered appropriate for taming is that it has to be in its infancy (see Costa 2017, p. 30). Specimens considered too big are referred to as bani (meat/prey) and killed as game.
- 5 Up until the 1930s (when they abandoned the interior of the forest to settle along the Purus River (...)
- 6 Trips of this kind are frequent among young Madiha and do not seem to be motivated by any desire o (...)
5The Madiha term used to refer to animals captured in the forest and raised as pets is meze. In its verbal form, meze inanahari can be translated as “to bring up/raise an animal or a child other than one’s offspring” (see Boyer and Boyer 2018, p. 85; Dienst 2014, p. 96). Indeed, meze is a term covering a wide semantic sphere which includes not only pets but also adopted children (who are not uncommon in Madiha communities in Peru) and captive children (now absent but probably numerous in the past).5 Meze is also the term used to refer to those adolescent boys and girls who spend time (ranging from a few weeks to several months) in villages other than their own before returning to their parents.6 Thus, generally, meze designates children and animals, as well as adolescents, who are fed and nurtured, albeit sometimes only for a limited period, by a figure other than their birth parents. Occasionally, the term is also employed to refer to adult humans. Indeed, a few months after my arrival in San Bernardo for field research in 2018, I realized that the Madiha referred to me not by my name, but with the expression Anakha meze, Ana being the mother of my Madiha host family (the person who gave me food) and -kha the alienable possessive suffix (Dienst 2014, p. 215-217). Throughout this article, I will use the terms “pets” and “non-human meze” interchangeably.
- 7 Santa Julia can be reached in four to six days by canoe from San Bernardo. Kinship ties link some (...)
- 8 The pet vocatives I collected in San Bernardo correspond to those collected by Dienst in Santa Jul (...)
- 9 In contrast to other Amazonian peoples (Costa 2017, p. 31, n. 8; Erikson 1988, p. 26), the Madiha (...)
- 10 In this sense, the Madiha are an exception to the norm of many Amazonian societies, which usually (...)
6The spectrum of species considered tamable by the Madiha seems in principle to be very broad, although certain species are tamed far more often than others. This emerges clearly from the fieldwork carried out by linguist Stefan Dienst between 2002 and 2007 in the Brazilian Madiha village of Santa Julia (state of Acre).7 Dienst reports the existence of more than thirty Madiha “pet vocatives,” namely special appellations used by several societies in southwestern Amazonia to call pets (Costa 2017, p. 39-40; Dienst 2014, p. 275; Dienst and Fleck 2009; Erikson 1988, p. 28-29).8 Such terms differ from those used to refer to specimens of the same species that are not pets. Pet vocatives do not substitute referential nouns, insofar as pets continue to be referred to by the species name, nor are they proper nouns, as all animals of the same species are called by the same pet vocative (Dienst and Fleck 2009, p. 210). Despite having collected numerous pet vocatives, Dienst remarks that, in Santa Julia, pet keeping was “commonly done only with monkeys, parrots and tortoises” (Dienst 2014, p. 275). Similarly, missionary Cindy Boyer, who lived for thirty years in San Bernardo, explains that she has seen mammal pets such as deer, coati, tapir and white-lipped peccary only occasionally, while tortoises and parakeets have always been extremely common (Boyer, com. pers. 2020). Boyer’s description closely resembles what I witnessed in the same village during my fieldwork: whereas tortoises9 and parakeets were kept in practically every household (often in more than one specimen), other species, such as monkeys and other land mammals, were less common. Yet other species were kept exceptionally, and were explicitly considered unusual pets. Among the latter, it is worth mentioning a pet vulture, an owl, two iguanas, a star-fingered toad and a shitari fish, all of which I have seen kept as pets. These observations suggest that, although they habitually end up taming only a handful of species, the Madiha are ideally keen on attempting to tame any species of all classes of vertebrates that they find (probably as long as they are not poisonous: I never heard any accounts of snakes or poisonous frogs being kept as pets).10
- 11 San Bernardo has been located on the same site since 1958 (Adams 1976). Consequently, game animals (...)
- 12 It is to be assumed that the chances of hitting an infant monkey clinging to its mother are higher (...)
7Several factors may explain why, although many species are considered tamable, only a few are commonly raised. The fact that in Santa Julia many pet vocatives were known only by adults leads Dienst to suggest that in the past the Madiha used to tame a wider range of species than they do today (Dienst 2014, p. 275). This may be true in San Bernardo as well, if only because the San Bernardo Madiha nowadays rely more on fishing than on hunting for their daily subsistence, and this reduces their chances of encountering land animals.11 Moreover, it is likely that the introduction of firearms has made it harder to capture infant animals since lead pellets in the cartridge disperse radially, hitting everything in the wider vicinity of the prey.12 On the other hand, the desire to raise only certain specific animals should also be taken into account. Some San Bernardo people, for instance, refuse to keep pet monkeys on the basis that they soil everything by defecating indiscriminately and that they are very demanding, problems that tortoises and parakeets pose to a lower extent. Simultaneously, I agree with Dienst that certain species might have always been kept as pets more commonly than others (2014, p. 275), if only for the simple fact that they were more easily found.
8Apart from animals captured in the wild, San Bernardo Madiha families commonly keep chickens, dogs and cats. As I will show, these animals, which I will refer to as “domesticated animals,” occupy a somewhat different position in the eyes of the Madiha, insofar as they are only occasionally treated and referred to as pets (meze).
- 13 This dynamic is well documented. For a few ethnographic cases, see Cormier (2003, p. 114), Costa ( (...)
9Captured animals are never kept by their captors. They are invariably given. More precisely, animals only seem to be caught as long as there is somebody to give them to. This is a typical dynamic among Amazonian groups: men capture animals and give them to women or children who raise, feed and look after them (Erikson 1987, p. 106).13 Among the Madiha of San Bernardo, married men typically give captured animals to their wives or their children (both male and female), whereas unmarried men typically give captured animals to their mothers, their grandmothers, their siblings or young children. The relationship between the captor-giver and the receiver follows the lines of pre-existing close kinship relations and of co-residency (in the cases I have witnessed, captured animals were never given to somebody from a different household). It is worth mentioning that in San Bernardo, people recall with precision the captors of each pet in the house and of all of the pets they have had in the past.
- 14 I refer the reader to Bittencourt (2021) for a comparative analysis of pet designations among the (...)
10One who receives an animal is called the hinede of that animal, and the animal in question starts to be referred to as ponikha/pokha meze (“her/his meze”). Alternatively, pets can be referred to as ponikha/pokha [name of species] bedi (“her/his small [name of species]”; see Costa 2017, p. 31) or, when talking about “pets” as a category, as bani bedi (“small game”), which, just like the Kanamari bara o’pu, is an expression “used to refer either to the young of wild animals or to pets” (ibid.).14 Literally translatable as “the one who causes” and less literally as “owner” (Boyer and Boyer 2018, p. 76; Dienst 2014, p. 68), hinede is one of the terms by which the Madiha designate the widespread Amazonian category of the “master/owner.” First theorized by Carlos Fausto (2008, 2012a), the “master/owner” is a figure who establishes a relationship of “control and/or protection, engendering and/or possession” over a reciprocal term (Fausto 2012a, p. 30). Ambivalent and asymmetrical, in that it casts a relation that is simultaneously of control and care, dependency and nurture, the relation between a “master/owner” and its reciprocal term structures different spheres of the social life of Amazonian peoples, so much as to be considered a key “cosmological operator” in the region, namely a fundamental modality through which Amazonian peoples establish social bonds (Fausto 2008, 2012a).
- 15 Hinede does not exclusively apply to humans, but also to non-human entities in relation to objects (...)
- 16 Birth parents are designated with the terms abi (father) and ami (mother) (Boyer and Boyer 2018, p (...)
- 17 The Summer Institute of Linguistics (SIL) is a US Christian organization dedicated to the translat (...)
- 18 The San Bernardo Madiha often refer to SIL missionaries Cindy and James Boyer with the expression (...)
- 19 Shamanism is most often a man’s affair. However, it should be mentioned that women shamans are inc (...)
- 20 Nowadays, only men are village chief. However, I was told that in the past women chiefs were not u (...)
11In the Madiha language, the terminology for asymmetrical relations of this kind is quite complex, and there are several terms to designate the figures who occupy the positions of generation, possession, nurture and control. Hinede designates owners of pets, objects, garden products, shamanic spirits and immaterial things such as songs or events;15 tamine designates a village chief; mezede designates adoptive parents,16 bosses and, in its translation by SIL missionaries,17 God. Kakawade differs from mezede in that its meaning does not include a “raising” element: it can be translated as the “one who guards/looks after” (Boyer and Boyer 2018, p. 16) and it designates bosses and SIL missionaries,18 but not adoptive parents. It is the other term by which God is translated into the Madiha language by SIL missionaries (ibid.). Finally, nanapide, which I translate as “the one who causes another to grow” (see ibid., p. 90), designates those who feed children or pets, but does not necessarily overlap with birth parents (ami or abi), with the mezede of an adopted child or with the hinede of a pet. This is because, as I will show briefly, “owning” and feeding are not inseparable acts for the Madiha. Hence, unlike many Amazonian societies (Fausto 2008, 2012a), but like the Warao of the Orinoco Delta (Allard 2010, p. 115) or the Yuqui of the Bolivian Chapare (Jabin 2016, p. 450-461, 470), the Madiha have more than one term for “master/owner.” Yet, this abundant lexicon for the figure of the “master/owner” is not matched by as many reciprocal terms. Indeed, pets and adopted/captive children are the only entities designated by a specific term, meze, whereas all other figures encompassed by a “master/owner” are simply designated by possessive pronouns and not by filiation terms. For instance, a shaman’s auxiliary spirit is called pokha dori (“his dori,”19 with dori being the name of shamanic “spirits”), a chief’s villagers pokha madiha (“his people”),20 and someone’s objects ponikha/pokha zepetahi (“her/his things”).
- 21 Just as among the Kanamari (Costa 2017, p. 40), pets are remembered by their owners even years aft (...)
12If we focus on the cases of pet keeping and child adoption, we see that while both pets and adopted/captive children are designated as meze, pet owners and adoptive parents are labeled hinede and mezede respectively, where mezede can be translated, in this context, as “the one who brings up/raises.” Tia mezede (“the one who brings you up/raises you”) was indeed the formula that the Madiha would use, when speaking to me, to refer to Ana—the mother of my host family, namely the person who cooked for me and gave me food. Yet, more than once, when people referred to me as Anakha meze (“the meze of Ana”), I heard children giggle and say meze naraa napiharai (“he is a meze but he does not grow”), revealing that a meze is considered to be an entity that is made, and meant, to grow. It was Ana who would always say of her older sister’s husband, who hunted for her and gave her food after she lost her father as a young girl, powahine onapiharo (“I grew because of him”). Indeed, a mezede can also be called nanapide, a term that, as mentioned, can be translated as “the one who causes another to grow.” The practice of raising pets is also circumscribed by the action of “to cause to grow.” In evoking the many pets she had raised, Ana used to say homo onanapiharo, zazio onanapiharo, zowihi onanapiharo … (“I caused a spider monkey to grow, I caused a howler monkey to grow, I caused a brown capuchin monkey to grow …”) and so on.21 Moreover, another typical expression for raising pets is bani bedi mezede (“to bring up/raise small game”). It follows that human and non-human meze, namely adopted/captive children and pets, are in an analogous position insofar as they are both subject to an encompassing figure who “brings them up/raises them” or “makes them grow” through acts of nurturing and, more precisely, feeding (Brightman, Fausto and Grotti 2016; Costa 2017; Fausto and Costa 2013). In other words, the status of a meze seems to be ideally conceived of as being linked to the process of growth.
13Yet, while in child adoption the mezede, namely the encompassing figure (the adoptive parent), is, as far as I know, always the one who causes the meze to grow (i.e., the nanapide, namely the feeder), in pet keeping the hinede is not necessarily the feeder. In other words, the fact of being the master/owner, or hinede, of a pet is not necessarily tied to the act of “causing it to grow” through acts of feeding. As I mentioned, captured animals are given to very young children and sometimes even to infants. Due to their young age, such children-hinede are most of the time unable to look after their pets, which, consequentially, often die tragically. In one case, a small pet armadillo was killed by mistake by its child-hinede who was playing with a machete. In another case, two small iguanas were eaten by chickens as their child-hinede was playing with them under the floor of the raised house, and was unable to protect them from the birds’ attacks. I doubt that children-hinede are actually expected to nurture and feed their pets. Yet adults constantly remind them to do so: children-hinede are encouraged to carry their pets around, to watch over them, to put them to sleep and to call for somebody to feed them. In other words, they are pushed to act as “true” hinede even though they still lack the means to do so. Feeders of pets owned by children discarded my allusions to their being the actual hinede of the pet they were feeding, firmly replying that it was this or that child who owned the pet. Of course, the instance of the children-hinede should not overshadow the many cases in which the feeder is openly identified as the hinede. Nevertheless, it indicates that, in Madiha pet keeping, being a feeder does not always overlap with being an owner, since owning a pet does not necessarily imply a direct feeding relation. This in turn suggests that one can be a hinede simply by virtue of having been given an animal, prior to performing any specific action (such as feeding) towards the animal in question. The link between ownership and feeding is more mediated in this case than it is, for instance, among the Kanamari, for whom feeding necessarily implies owning (Costa 2016, 2017). This invites us to contemplate the existence of cases where the feeding-fed relationship is not precisely predicated upon the idiom of mastery/ownership (see Fausto and Costa 2013, p. 160).
14The question remains as to why the Madiha insist, by giving captured animals to children, in making hinede who cannot truly act as such. In other words, why do the Madiha try to put children into a dynamic of control that identifies them with an encompassing role (albeit a “fake” or “simulated” one) that they are nonetheless incapable (or only partially capable) of performing? This may seem like a question not worth asking, insofar as similar dynamics of giving pets to children are frequent even in less “exotic” contexts. However, the universality, or presumed universality, of a social practice does not inform us about the local meanings attributed to it. Regarding the Madiha case, I think the answer to this question is probably to be found in the dynamics of “magnification” of the owner typical of some Amazonian societies, according to which being a master/owner is a condition “necessary for personal growth” and “personal autonomy” (Costa 2017, p. 70, 59-74, 123-125; Fausto 2008, p. 342-345). Indeed, the case of children masters/owners is not unique to the Madiha. A quick look at the ethnographic record reveals that children are given (to different extents) the role of “masters/owners” in different regions of Amazonia (Calheiros 2014, p. 62-63; Costa 2016, p. 92, 104, n. 16; 2017, p. 123-127; Pétesch 1993; Walker 2013, p. 40-49). This seems to be done to encourage the transition from dependent child to fully independent adult, as if being a master/owner would instill in children the qualities needed to be a person. Indeed, being a “master/owner” (namely occupying one of the positions described above in relation to someone or something) is a quality inherent to every Madiha person, who, with the exception of old people and young children, is always entangled in asymmetrical relationships with other people (Zuppi 2021, p. 106-113). For the Madiha, to be a master/owner is “to exist socially” (Pétesch 1993, p. 90). Whereas old age is the time when one becomes more and more dependent on others and gradually loses the status of “master/owner,” childhood is the time when mastery/ownership bonds must be created to progressively enable the child to stop occupying only the dependent position (see Costa 2017, p. 123, 223). In this sense, giving a child the role of hinede to a pet would mean to push him or her to occupy, even if only lexically, the position of generation and control of the “master/owner” necessary to be a proper person. To be clear, this would not be merely because, as Cormier notices (2003, p. 116), in having to look after a pet, little girls become familiar with the tasks of motherhood and little boys become familiar with the behaviors of wild animals they will have to hunt. Rather, I argue that pet ownership and person making are connected because of the “relational responsibility” (Moïsseeff 2019) towards other persons, animals or things that being an owner entails.
15The case of children-hinede supports the idea that the pet is less important than the relationships it allows to create. Children-hinede who accidentally cause the death of their pets are neither reprimanded nor scolded and, often, are readily given a newly captured animal, likely to incur the same fate as its predecessor. Furthermore, the death of children-hinede’s pets is met with no surprise. What seems to be really at stake here is less the well-being, development or assimilation of the pet than the desire to create a person through the attribution of a pet. From this point of view, one of the terms of the relationship (the meze) is instrumental to the other (the hinede). We will now turn to taming and familiarization techniques.
16Pets are treated differently depending on the species. The following analysis is limited to the pets I most frequently saw during my fieldwork: tortoises, tui parakeets and monkeys.
17The introduction of tortoises into the house does not usually arouse great interest. Tortoises are kept in buckets or tied to supports with a string threaded through a hole made on their shell. Tortoises do not receive much attention from their owners and they are not fed regularly, although they are sometimes washed by immersion in a bucket full of water. If they do not flee, get eaten by a dog or accidentally get stepped on and killed, tortoises may grow large and survive many years in captivity. Those who grow accustomed to the house are left untied and free to roam around.
- 22 Bananas prepared for tui parakeets (and parrots more generally) are roasted directly on the embers (...)
18By contrast, tui parakeets are better cared for. They are fed with premasticated food (roasted bananas) which they peck directly from their feeder’s mouth. This operation is generally carried out by women, although it is not uncommon to see young male owners of tui parakeets feeding their pets in this way. The preparation of roasted bananas for tui parakeets differs from their preparation for human consumption.22 This adds nuance to the widely affirmed ethnographic datum, applicable to the Madiha case as well (see below), that pets are fed on the same food eaten by humans (Erikson 1987, 1988). This case shows that while the type of food might be the same, it is not necessarily cooked in the same way.
19At night, tui parakeets are put to sleep in woven baskets, but during the day they are left free to roam around the house. As they grow, their wings are periodically clipped. Yet, despite being relatively well looked-after, tui parakeets rarely survive long. They are prey to dogs and cats, they can be accidentally killed by humans or they might venture outside of the house where they are lodged and get lost. Nevertheless, one rarely enters a house where there are not a few tui parakeets chirping and walking around.
- 23 I do not know whether such spitting is also carried out with other mammal pets.
- 24 Judging from my field notes, the verb wati towanaharo is intransitive. This suggests that the capa (...)
20The most elaborate treatment is reserved for infant monkeys. I experienced this directly the day I was given a pet howler monkey by one of the teenage hunters in my host family. The first familiarization act performed on a monkey newly introduced into the house is to spit (pithode) in its mouth.23 In the cases I have observed, this was always done by an elderly woman. The Madiha assured me that spitting is done for several reasons. It serves to make the new pet monkey meek and docile (tohonehine). Spitting also causes the monkey to eat (hipanihine). Ana told me that the pet night monkey she was raising at the time of my fieldwork refused to eat until its mouth had been spat in by Ana’s mother-in-law. Finally, spitting is performed so that they monkey “comes to love/care” (wati towanahihine).24 When a new infant monkey is introduced into the house, it is said to “suffer” (bodi koma tai) and to be “missing its mother” (imenitowi wati wana tai), but as soon as it starts accepting food, the monkey is said to “come to love/care” (wati towanahari). Hence, spitting creates a consubstantiality (in the literal sense of placing human’s saliva in the monkey’s mouth) that engenders a process of familiarization in which being meek and docile, accepting food and “coming to love/care” are inseparable and mutually related expressions.
- 25 In contrast, among the Kanamari, pet vocatives begin to be used when the familiarization process i (...)
21Familiarization of monkeys is also achieved through the use of pet vocatives. I was often encouraged to call my pet howler monkey by its pet vocative, repeatedly and several times a day, so that the animal would grow accustomed to it and come towards me when called.25 The same was done by owners of other pets, which (with the exception of tortoises) are called by their respective pet vocatives when addressed by their owners and by others.
- 26 The “thief” was the husband of the woman who had spat into the monkey’s mouth. This man, himself a (...)
22This said, it should be emphasized that the mere introduction of a newly captured infant monkey into the house, the act of handing it to a woman or a child or the act of spitting into its mouth in no way ensures that the animal in question will be further familiarized. The most common way for a newly captured monkey’s familiarization to be cut short is theft. In San Bernardo, the first few hours after a pet is introduced into the house are the most crucial in this respect: the newly acquired pet should not be left alone or it could be taken and eaten by hungry neighbors. This is not a rare occurrence. During my fieldwork a pet spider monkey, into whose mouth an elder woman had already spat, was stolen and eaten by a consanguine neighbor of another domestic unit while all the members of the household were attending the football match that takes place every day in the late afternoon in San Bernardo.26 The same happened to a pet black caiman that was stolen during the night from my hosts’ house. While these thefts no longer seem to occur once the pet has passed its first night in the house, they nonetheless indicate that the initiation of a familiarization process is neither an obstacle-less endeavor nor a matter of course.
- 27 This is true for all pets kept by the Madiha and is an attitude typical of Amazonian people, who r (...)
- 28 I have never seen women directly suckling pet monkeys, as occurs, for instance, among the Guajá (C (...)
23Once such dangers are averted, pet monkeys are looked after a great deal and any kind of mistreatment, beating or neglectful behavior inflicted on them is met with disapproval.27 Pet monkeys are washed, carried around by their owners (or by their feeders if the owner is too young), constantly surveilled, placed in baskets to sleep, kept warm with old clothes and fed regularly. During the first days after a pet monkey’s arrival in particular, their behavior is on everyone’s lips. When they are not carried around on one’s head, pet monkeys are tied to supports in the house. As they become familiarized with the domestic space, they are left free to roam and climb around the house. Initially they are spoon-fed with powdered milk or, in its absence, with human milk collected in a cup.28 As they grow, they are gradually fed on bananas, banana porridge and papayas. All of the pet monkeys I saw were infants, so I do not know if they are fed on other types of food as they grow.
- 29 Similarly, pet harpy eagles among the Huaorani are not fed on human food and are kept outside the (...)
- 30 Carachama fish constitutes the basis of most meals in San Bernardo during the dry season.
- 31 It should not be inferred, from this expression, that the species in question are considered immor (...)
24While it is generally true that pets undergo a change of diet once captured, in that they are fed on human food (mostly manioc, banana porridge and bananas) (see Costa 2017, p. 33; Erikson 1988, p. 30-31; Goulard 2009, p. 215), there are also instances when no attempt is made to change their diet. The above-mentioned pet vulture was only fed with the intestines of fish and game, namely those parts that the Madiha do not eat. At the same time, it should be said that the pet vulture in question was, in a manner of speaking, only partially a pet as it was not kept inside the house (coresidency being key in defining what a pet is; see below) but on the surrounding patio.29 At any rate, changes in diet are not unidirectional in Madiha pet keeping: the feeders (who are sometimes but not always the owners) of certain pets have to observe dietary restrictions themselves. Failure to observe such restrictions will cause the pet to become ill with worms (shomi) and die. I was made aware of this when raising my pet howler monkey. I was told that if I ate tortoise meat or carachama fish,30 my pet howler monkey would contract worms and die. I collected a few stories about strong and healthy non-human meze that died of worms because their feeders ate certain foods. For instance, a woman owner and feeder of a deer meze had, I was told, caused it to die because she had eaten a type of fish called washa (Blindado siluro). In another case, a woman who was feeding her son’s agouti meze ate deer meat, thus causing the death of the pet. According to these stories, certain species of non-human meze are tied to specific edible species which its feeder should not eat. On the other hand, there are also species of non-human meze, such as the spider monkey, the brown capuchin monkey and the tortoise, that are considered watiera kiri nai (“that do not die”).31 The feeders of these species face no dietary restrictions.
- 32 Just as among the Kanamari (Costa 2017, p. 97-110), Madiha couvade proscriptions aimed at guarante (...)
25Consequently, the well-being of certain pets is tied to the eating behavior of their feeders in ways that evoke the couvade, namely those perinatal practices and proscriptions, widespread in Amazonia, put in place by parents to guarantee their newborns’ well-being (Rival 1998; Vilaça 2002) as well as their own (Costa 2017). Indeed, many couvade practices among the Madiha consist of avoiding eating certain specific foods (especially certain male animals and large fish) to protect the newborn from a fatal disease called ephe tokai, conceptualized as a beetle that takes up residency in the child’s belly if parents do not respect the dietary restrictions (Pollock 1996, p. 327-329, 338, n. 6).32 It is particular of Madiha couvade restrictions that they only apply to the parents of birth children and not to adoptive parents who are raising a human meze. My interlocutors made it clear to me that adoptive parents to an adopted child can eat all kinds of food without any risk to the child. However, the adopted child will get ill if the birth parents (even if they live miles away) do not comply with the proscriptions. In this sense, those pets whose rearing involves dietary restrictions (i.e., not all pets), while being lexically likened to adopted/captive children, are practically likened to birth children. In other words, the context of food proscription reveals how familiarization of certain pets involves embedding them in a process of fictive consaguinization in a way that does not apply to adoptive/captive children, for whom no couvade practices have to be observed by the adoptive parents. Such processes of consaguinization of non-human meze is articulated, it seems, around feeding relations. In fact, in addition to warning me not to eat certain types of food to protect my pet howler monkey from worms, my hosts also clarified that, had I decided to allow others to feed it as well, I could have eaten all types of food without posing any risk to my pet. If I remained the sole feeder to the monkey, however, I would have certainly caused it to die had I eaten the proscribed food.
- 33 In his book, Costa adds, however, that “Although I never saw anyone respect this injunction, it em (...)
26Note, as an additional observation, that what we can label “couvade for pets” is a more widespread phenomenon in Amazonia than has been appreciated so far. For instance, we know that among the Kanamari, “during the first days in which the pet is fed, the woman responsible for its well-being (or, according to some, everyone in the household) must refrain from eating game, particularly meat from the same species being reared […]” (Costa 2016, p. 86).33 Harner (1972, p. 63) writes that, among the Jívaro, now known as Aént Chicham, “When puppies are born, one of the owner’s wives observes a kind of couvade, in which she lies abed with the bitch to protect the litter from supernatural harm.” For his part, Walker (2013, p. 174) reports that among the Urarina of Loreto (Peru), a pet would die if fed by a pregnant woman. These ethnographic examples, however isolated and allusive, should encourage further investigation of a phenomenon (that of couvade practices for pets) that remains understudied.
27The fact that a non-human meze is an entity that “is made to grow” hides a bitter side of the pet keeping relation. Indeed, if to keep a pet means to cause it to grow, it follows that pet keeping ends when the pet stops growing (Costa 2017, p. 32-33, 55-56; Vander Velden 2012, p. 185-188). This is not merely a matter of wordplay. Among the Madiha of San Bernardo, it seems tacitly taken for granted that adult pets have to be gotten rid of, one way or another. Adult pets are, apparently, of no interest.
- 34 Karia is the term by which the Madiha designate non-indigenous Peruvian and Brazilian people. Acco (...)
- 35 This anecdote is reminiscent of another, reported by Cormier, about an adult pet peccary that the (...)
- 36 Costa also reports the case of an adult wooly monkey whose “age was so advanced for a pet and [who (...)
- 37 Cormier goes on to say that “parents never reprimanded children for this behavior and on one occas (...)
28While talking about the pet night monkey she was raising at the time, Ana once said that she would release it in the forest once it came of age. This is what happened to a pet mealy parrot (kowero), which was released by its owner Pedro when it reached adulthood. The pet parrot in question used to fly back to the village and perch on the roof of its old house every day before sunset. Despite still being called Pedrokha kowero (“the kowero of Pedro”), the parrot was generally ignored and nobody fed it. Similarly, the pet vulture mentioned earlier simply flew away one day but still came back near the house occasionally. No attempt at recapturing it was ever made, and it was not fed. Similar cases of release (or abandonment?) of adult pets are also accounted for among the Tukuna (Nimuendajú 1952, p. 24), the Guajá (Cormier 2003, p. 126) and the Matsigenka (Shepard 2002, p. 110). In most cases, when they are imei (big), pets (especially mammals) are sold to the karia34 or to other Madiha in the community, in the knowledge that they will be killed and eaten. Sometimes, pets are eaten by consanguine kin of the pet owner, but never by the coresidents of the owner, let alone by the owner themselves. Indeed, the refusal to eat one’s own pets is a well-documented generic fact among Amazonian Amerindian societies (Erikson 1988, p. 30; 2000, p. 20-21). The possibility is nonetheless evoked, but without ever being put into practice. The large pet tortoise kept in a house that I visited often was on occasion regarded as a kind of mobile dinner by the members of the household, who would sometimes comment bani tohabotenai (“it had almost become meat/prey”). Upon further inquiry, I was told that eventually they would kill it and eat it. Of course, this eventuality never materialized: when I returned to the field in 2022, after an absence of more than two years, I saw the tortoise in question alive and well and exactly where I had left it. Although possibly only ironic, the comments about eating the tortoise meze reveal an inherent discomfort in keeping a pet that is no longer growing.35 In short, it seems that among the Madiha the process of pet keeping is epitomized by the notion of “causing to grow” to an extent that its accomplishment (i.e., having caused an animal to grow) inexorably sanctions the disintegration of the bond between owner and pet. I would like to stress that in numerous Amazonian societies adult pets are gotten rid of. Much like the Madiha, the Kanamari sell and exchange their grown-up pets, though in the latter case adult pets may also be killed in secret (Costa 2017, p. 41, 55-56).36 Remaining in the Juruá-Purus interfluvial region, we know that the Arawan-speaking Paumari release their pets after rearing them to adulthood (Bonilla 2016, p. 129, n. 20; 2022, p. 219) and that the Suruwaha (also Arawan-speaking) kill and eat all of their pets once they reach adulthood, with the notable exception of capuchin monkeys (Huber 2012, p. 207). Further north in Brazil, Helena Valero recalls that grown-up Yanomami peccary pets were killed “secretly” by people, other than their owners, who could no longer stand their aggressiveness (Biocca 1968, p. 221-222). Another striking example concerns the Guajá of eastern Amazonia, among whom, as Cormier informs us, older pet monkeys are “no longer considered hamɨmɨrə” (a term she translates as “my child”) and are targeted by young boys practicing their archery skills (Cormier 2003, p. 117).37 Though, to the best of my knowledge, he does not report cases of pet “disposal” similar to those just listed, Vander Velden has clearly pointed out that among the Karitiana of Rondônia there is a considerable difference in the behavior towards infant pets, treated with love and care, and adult pets, who are generally mistreated (Vander Velden 2012, p. 185-190). In short, all of the authors cited here seem to emphasize that the process of keeping a pet is temporally limited and that the relation between an owner and a pet is never meant to last. As I have shown, among the Madiha this state of affairs is, to a large extent, encapsulated by the notion of “growing.”
- 38 See Calil Zarur (1991, p. 37), Erikson (1988, p. 30), Goulard (2009, p. 218), Nimuendajú (1939, p. (...)
29In any case, among the Madiha, but also elsewhere in Amazonia, most pets die before reaching adulthood (see Cormier 2003, p. 125; Costa 2017, p. 40). It is therefore reasonable to assume that the problem of what to do with an adult pet rarely arises. In San Bernardo, some pets that die are buried. While this is the case with pet mammals and with dogs, for example, I have never seen a burial for a pet bird.38 A wooden cross is constructed and placed on the burial site. At times, the owner’s name is inscribed on the cross, followed by the pet vocative of the deceased pet, or, in the case of dogs (which in San Bernardo are not called by a pet vocative), by the dog’s personal name.
- 39 Simulating a hunt seems to be a common way to put pets and domestic animals to death in indigenous (...)
- 40 The day following these events, the young hunter who had given me the howler monkey as a gift went (...)
30Reaching adulthood, however, is not the only endpoint for the non-human meze. Being a pet is a condition that can end even before the pet reaches maturity. Indeed, pet keeping is a fragile relationship not only because it is tied to growth, but also because it is continually subject to possible disruptions: pets can, as we have seen, be stolen or accidentally killed, or they can flee. Similar disruptions in pet keeping relations can easily be found in the ethnographic literature of Amazonia. For instance, the Kalapalo only raise birds, monkeys and tortoises as pets (Basso 1977, p. 102). However, this does not prevent them from introducing into the villages infants of other species (such as fawns, caimans and peccaries) that will not be referred to as itologu (the local term for “pet”) “except in jest,” and which will quickly die “of maltreatment or lack of food” (Basso 1973, p. 20; 1977, p. 102). Costa writes that among the Kanamari not all pets end up being “loved” by their owners, and that some pets can become “ownerless” and “ignored by those who originally fed them” (2017, p. 39-41). Among the Madiha, pets may also cease to be pets should they find themselves ownerless. Once, tired of spending my days cleaning up after my pet howler monkey, I declared that I wanted to give it away. I naively believed that my pet would easily find another owner. Yet the only interested people were some hungry neighbors who wanted to eat it. What happened next was that a group of boys grabbed the howler monkey, put it on a small tree by the house, waited for it to climb up and hide in the foliage and then, armed with bows and arrows, shot it to death.39 All of the participants of this “hunt” were amused, excited and laughing loudly. So many arrows pierced the howler monkey’s little body that the meat was considered too spoiled to be eaten, and it was eventually thrown in the bushes.40
31I presume that the sudden change of attitudes towards the howler monkey, which shifted from being a pet (towards which all violent behavior was condemned) to being prey, is explained by the fact that I, namely the hinede, had given it up. In other words, it is the fact of having a hinede that constitutes a non-human meze, which, it seems, can exist as such only in a relation of dependency on its owner (see Costa 2017, p. 64). It follows that a pet that suddenly finds itself ownerless ceases to be a pet, despite still being an infant and having a history of familiarization behind it (the howler monkey in question had had its mouth spat in, was fed, was said to “love/care,” was called by the pet vocative and lived inside the house). The statement that, in Amazonia, “everything that lacks an owner also lacks care and protection” (Costa and Fausto 2019, p. 204) finds its ultimate substantiation in the story of my pet howler monkey. Being a non-human meze, therefore, is a temporary, precarious and reversible condition. This aspect will be clearer from analysis of the treatment reserved for certain domesticated animals.
32As previously mentioned, practically all families in San Bernardo raise chickens and have at least one dog. Cats are also very common, but I will not include them in the following analysis. A brief look at the ways chickens and dogs are treated can contribute to a better understanding of how pets are conceptualized among the Madiha.
- 41 Only on a couple of occasions have I seen people eating their chickens or their chickens’ eggs. Ma (...)
- 42 Indeed, whenever somebody goes to Puerto Esperanza, those who remain in the village fantasize abou (...)
33The main difference between pets and chickens is that the latter are only fed exceptionally. Those who occasionally feed their chickens usually do so on leftovers gone bad due to excessive heat, namely on food that humans would not eat, not on food aptly prepared for them. During the day, chickens are free to scratch around the house in search of food, while at night they are kept in the henhouse, which is always located near the house of their owners. Hence, chickens do not share the domestic space of their owners, and are usually shooed away whenever they venture inside the house. Indeed, on the rare occasions on which they are fed, this always happens outside of the house (see Costa 2017, p. 83). Every single chicken has a hinede amongst the components of the household. Chickens are owned by adults, teenagers and children alike. Children exercise the same rights over their chickens as adults and are free to sell them if they wish. The Madiha generally refrain from eating their chickens and their chickens’ eggs (Pollock 1985a, p. 27). Even when there is a lack of food in the house (not an uncommon occurrence at the time of my fieldwork), the idea of eating one of the many chickens scrambling around the house, or their eggs, apparently (and much to my frustration) did not seem to cross anybody’s mind. Not eaten (except in extraordinary cases),41 chickens are instead sold to the karia who stop over in the village during their trips up and down the Purus or are brought to Puerto Esperanza to be sold there. Usually, the money earned from the sale of the live chicken is immediately reinvested in the purchase of a frozen chicken from one of the shops of the little town. This shows that the reason for refusing to eat chickens is not a matter of taste. To be sure, when in town the Madiha attach great importance to eating chickens pina karia (“like the karia”) and, when they finally get to eat chicken meat, they do so with theatrical voraciousness, commenting with satisfaction on its exquisite flavor.42 In sum, while the Madiha love eating chicken, they simply do not eat the chickens that they own. The principle, emphasized by Costa for the Kanamari, but extendable to other Amazonian societies (Erikson 1987, 2000; Vander Velden 2019a), that one does not feed on the animals one feeds (Costa 2017, p. 53) applies to the Madiha case with a slight variation: one does not feed on the animals one feeds and/or owns. Indeed, not being eaten by their owners (except in extraordinary cases) is a feature that chickens share with pets. However, chickens are normally not considered meze. When I asked my hosts to clarify this point, they answered that chickens are not meze because powara hipamanahari (“they eat on their own”) or because oza kahimanahari (“they have their own house”; namely the henhouse). They also suggested, however, that chickens can be considered meze if they are fed by humans and if they are kept inside the house as co-residents. Such answers emphasize the importance of being fed and of coresidency in defining what a meze is (see Costa 2017, p. 30-34; Erikson 1988, p. 31).
34In one case, some Madiha visiting from downriver had given an infant chicken to Ana. This chick was not put in the henhouse. Ana kept it in the house, fed it every day with banana porridge, manioc and other human food, put it to sleep in a woven basket at night, carried it with her on trips to Puerto Esperanza and referred to it as okha takara meze (“my chicken meze”). Likewise, everybody in the household referred to the chick as Anakha meze (“the meze of Ana”). Once the chick became an adult, however, its status changed: it was no longer allowed inside the house, was no longer fed and people stopped referring to it as Anakha meze and merely called it Anakha takara (“the chicken of Ana”), like any other chicken she owned. Put simply, as soon as the chick reached adulthood and stopped growing, it was no longer considered a meze, but a chicken like any other.
35The same is true of dogs. The Madiha treat puppies with a great deal of care: they spoon-feed them with powdered milk, take them on trips whenever they travel and sometimes even put them to sleep by swinging them in hammocks as they would with a human child. These puppies are referred to as meze. When they reach maturity, however, dogs begin to be fed more rarely, are treated with far less care and attention and often cease to be allowed into the house.
- 43 Interestingly, Dienst indicates that “in Deni and in the Sivakodeni dialect of Western Jamamadi [a (...)
- 44 Dogs are likened to other species of domesticated animals known to the Madiha by at least one elem (...)
- 45 In San Bernardo, dogs are often given names that correspond to villages or cities, such as “Manuel (...)
36The case of dogs, however, is more ambiguous than that of chickens. The Madiha word for “dog” is ethe. As a verb, ethe inanahari is a synonym of meze inanahari, and it means “to bring up/raise a child or an animal other than one’s offspring” (see Boyer and Boyer 2018, p. 46; Dienst 2014, p. 96).43 Thus, lexically, dogs should be the quintessential pets. In practice, though, this is only true when they are puppies, and the treatment they receive as adults departs significantly from what their designation (ethe) might suggest.44 The same dynamic with respect to dogs, well cared for as puppies and neglected as adults, is typical of many Amazonian societies (Büll 2018, p. 62; Vander Velden 2012, p. 188; 2019b). Indeed, adult dogs in San Bernardo generally look miserable and emaciated (Zuppi 2022). What we gather is that, while for captured wild animals the condition of being owned is inseparable from the condition of being a meze (as the anecdote of the pet howler monkey indicates), this is not the case for domesticated animals, which can still have a hinede while no longer being considered meze. At any rate, even when they are considered as meze, I have never observed chickens or dogs being addressed with pet vocatives, despite the fact that the San Bernardo Madiha know the pet vocatives for both species: dogs are called by their personal names45 while chickens (in the rare case that they are fed) are called by the onomatopoeic repetition of sounds such as tio-tio-tio-tio. Lastly, domesticated animals such as chickens and dogs are unique in comparison to pets in yet another respect: adult men can own them. In fact, I only recall a few cases where dogs were owned by women; most dogs are owned by men and, at times, are used for hunting. Hence, women feed and raise dogs as puppy meze that are not theirs, but which are recognized as being the property of men. We see, therefore, that while domesticated animals can potentially be treated as meze, they do not seem to be conceived in exactly the same ways as wild animal meze.
- 46 I remain unsure as to how common this naming practice was in the past. I have also heard of adopte (...)
37With all the above in mind, we can now turn to the relation between pet keeping and child adoption, and try to draw out the similarities and differences between them. Firstly, we know by now that pets and adopted children are lexically likened, as both are meze. However, even if they are lexically equated, their position with respect to those who bear their reciprocal terms (hinede and mezede) differs: whereas a pet can be fed by a figure other than its owner, an adopted child is always fed by his or her adoptive parent. This is because in Madiha pet keeping, owning and feeding do not automatically overlap but can be disjointed (see Costa 2016, 2017; Fausto and Costa 2013). This being said, pets and adopted children are fed with the same type of human food (bar exceptional pets like the vulture), just as they are both constantly monitored and their freedom of movement is restricted (small children among the Madiha are never left alone, be they adopted or not). On the other hand, just as pets are called with pet vocatives, adopted children were in the past called with proper nouns that underlined their status as meze. Such nouns were Eze for males and Meze for females. Nowadays, human meze are given normal proper nouns, just like all other Madiha children, but it is not rare to meet adults named Eze or Meze.46 It is tempting to consider these proper nouns as the equivalent of pet vocatives for human meze. While in these respects pets are likened to adopted children, they are dissociated from them with regard to the observation of dietary restrictions that must be observed to prevent them from dying. Even if minimal, in the double sense that they are not observed for all pet species and when they are observed only concern the one who feeds the animal, couvade practices for pets liken them to children by birth. This is because, as I have described, adoptive parents do not have to observe any of the dietary proscriptions that birth parents observe. In sum, the analogies and differences between pets and adopted children depend on the criterion taken under consideration.
- 47 To say that adopted children are familiarized forever does not mean that they are familiarized com (...)
38The fundamental and irremediable difference between the two, however, lies in the fact that, sooner or later, pets are, according to an expression aptly coined by Erikson, “de-familiarized” (Erikson, com. pers. 2022): their meze status is perpetually at risk and eventually revoked. This contrasts with the position of human meze. Whereas their status can, at times, be transitory, as in the case of those adolescent boys and girls who temporarily reside with their kin in a village that is not their own, children adopted as infants are often familiarized for ever. Unlike other societies in the Juruá-Purus region, such as the Jarawara (Maizza 2014, p. 498-500), the Kanamari (Costa 2017, p. 128-135) and the Paumari (Bonilla 2022, p. 281), among whom the adoption of children is conceived of as a form of temporary fosterage, in that adopted children often return to their birth parents, among the Madiha of San Bernardo the adoption of children is often definitive. This means that as they grow, adopted children continue to call their adoptive parents abi (father) and ami (mother), to hunt and cook for them and to work in their gardens; they often continue living with their adoptive parents or build a new house nearby, and often marry according to the kinship terminology they acquired through adoption. In other words, ties between adopted children and adoptive parents, in San Bernardo, are not by necessity severed.47 To be clear, cases where adopted children have cut all ties with the adoptive family are not unheard of (Boyer, com. pers. 2020), but they are not the norm. Thus, whereas non-human meze are irretrievably de-familiarized, human meze practically never are. This last observation brings us back to the question of the fragility of pet keeping relations and opens the way to some recapitulative and conclusive reflections.
39The Madiha of San Bernardo persistently capture young animals in the forest and bring them to their village. Yet adult captured animals in San Bernardo are virtually non-existent. This is because the familiarization process is not only full of obstacles and fragile, but temporally limited as, when successful, it causes its own disintegration. As we have seen, familiarization is difficult to initiate, since pets run the risk of being stolen during their first hours in a household. Once initiated, a whole series of obstacles stand in its way: pets can be killed by accident, get lost, die or be intentionally killed upon finding themselves ownerless. Indeed, most pets never even get close to reaching maturity. Those that do are abandoned or sold or, in the case of chickens and dogs, begin to receive a completely different treatment. It would almost seem that the growth of a pet is a side effect of familiarization: because a pet’s existence is bounded by growth, a pet that no longer grows ceases to have a reason to exist. Yet, despite this, significant time and attention are invested in the process of familiarizing a pet. As I have shown, this investment varies according to the species: not all meze receive the same treatment. Monkeys (and mammals in general, one can assume) enjoy more elaborate treatments and care than tortoises and parrots. That said, sharing the domestic space with humans, being fed with human food and being called by pet vocatives (with the exception of tortoises and those pets that do not have a pet vocative) are generally crucial elements in the taming of all pets. To these can be added (in varying ways according to the species) the observation of dietary proscriptions that can be called “couvade for pets.”
40Thus, a contrast exists between the fragility of the familiarization process, namely the easiness with which it can be undone or interrupted, and the substantial investment in its implementation in terms of time, care and (even if I have not explored this aspect in the article) emotions, given that people remember all of the pets they have had in the past. Such contrast, I argue, suggests that what is important in the process of pet keeping is the process itself, regardless of its final outcome. Or rather, its final outcome is the process itself. Hence, the absence of adult pets in San Bernardo should not be conceived of as evidence of the failure of the familiarization process, but rather as proof that the importance of familiarizing pets lies elsewhere.
41I have chosen to analyze the role of pet keeping in child making to emphasize my point that the importance of familiarizing pets lies more in the process itself than in the fabrication of an adult pet as an outcome of that process. Giving a child the role of hinede to a non-human meze signifies enabling that child to abandon the position of total dependant and to gradually gain that of person, in a social context where a fully social person is always a “master/owner” (in the broad sense) of someone or something. Hence, giving captured animals to children can be read as a way of inducing a change of social status in those children without acting directly on them, but rather by acting “on the relationships [they] entertain” (Bonnemère 2018, p. 12). In this sense, pet owning and person making are tightly linked: making a pet is (in some cases) simultaneous to making a person. Furthermore, one can speculate that pet keeping consolidates relations of intimacy. I have emphasized the importance of the bond between the captor/giver and the receiver of a pet-to-be: captured animals are always given to close kin (or, more rarely, to coresidents). Hence, being put in a position (that of hinede) where one has to look after and take care of a non-human meze only occurs between people who are involved in daily relations of caring and looking after each other. In this sense, to be a hinede of a non-human meze can be considered an index of being implicated in intimate relations of care. The links between pet keeping and kinship, however, do not seem to go any further.
42The eventual unmaking of pets, their de-familiarization, which bears obvious analogies with the “reenimization” of Tupinambá war captives (Fausto 1999; 2012b, p. 256-260), who were referred to as xe-reimbab (“pet”) (Viveiros de Castro 1992, p. 280), shows that permanent adoption is not the purpose of pet keeping relations, although it may serve as its model. Put differently, the case of pet keeping among the Madiha stresses that feeding and/or owning do not automatically result in familiarization or in kinship. Indeed, the same has been noted with regard to pet keeping among the Kanamari and the Piro (Yine). In the former case, Costa points out that there are obvious limitations to the familiarization of pets, which, unlike children, can never really become commensals since “kinship with animals can only go so far” (Costa 2017, p. 127). In the latter case, Gow pointed out that the relationship between humans and pets is a “relationship between […] beings of different ontological conditions, for the same type of relationship between humans generates kinship relationships” (Gow 2001, p. 69). Yet pet keeping is but one in a series of cases in which feeding and/or owning do not “generate kinship.” For instance, Bolivian Chapare Yuqui masters feed their slaves to maintain them as “other,” not to familiarize them, according to a principle that has been called “heterotrophy” (Jabin 2016, p. 467-482). Similarly, the Trio of Suriname “domesticated” the Akuriyo, claim ownership over them, but maintain them in the sphere of affinity, a state of affairs that has led anthropologists to speak of a “domestication without assimilation” (Grotti and Brightman 2016, p. 66). In sum, it seems increasingly demonstrated that relations in which nurture does not necessarily lead to familiarization or kinship are not that unusual in the region. Hence, given that “kinship with animals can only go so far,” and given the ephemeral nature of pet keeping relations among the Madiha, we may ask in conclusion to what extent they should be considered under the prism of adoption.