1 - Perspectivism

Summary

Chapter 1 introduces the fundamental tenets of perspectivism, its main characteristics and principles and the problems and challenges it poses for archaeology. A brief account of the genesis of perspectivism as a theory is provided, drawing principally on Viveiros de Castro’s writings and comparing it to other ontologies, particularly animism. The key perspectivist characteristics shared by many Amerindian populations are detailed. These include the quality internal to many entities of possessing a human soul, the importance of the body as the distinctive mark of subjects, seeing the world from a human point of view and predation as the model for human relationships. These fundamental cosmological premises derive from a set of underlying metaphysical principles with consequences for social practices, all of which are relevant for thinking about the archaeological record.

In one of the founding texts of post-processual archaeology, Reading the Past (Hodder 1986), Ian Hodder asked what a series of figures nowadays familiar to archaeologists were (Figure 1.1). According to the point of view of the observer, each object can be one thing or another: a box with a side near or far from the observer, a deer or a bird, a bear behind a trunk or a knotted trunk. One might say they are ‘perspectivist objects’. Hodder told us that before we could do anything with them, we had to decide what they were. We had to categorize. These categories, he argued, are formed through a process of perception involving the real world, our theories regarding them and our own social and cultural context (Hodder and Hutson 2003). That is, there must be a point of view and at least one subject that interacts with those figures. However, Hodder (1986) argued that the decision about what they are is neither subjective nor a matter of alternative interpretations: they are in fact different things for different people according to their relative position in the world. That is, what they are comes down to different subjects and their respective points of view of the world. This is what perspectivism is all about.

Figure 1.1 Perspectivist objects

(based on Hodder 1986)

There is of course much more to perspectivism than this practical, simplified introduction. Perspectivism is both a theory of the world and a way of being in the world. As a theory, it is a set of philosophical assumptions about reality; as a way of being, it implies effective practices in that world. Perspectivism, expressed as theory and practice, is based on the anthropology of Amerindians, mostly native groups from the lowlands of South America, but also others worldwide. Perspectivism was developed chiefly by Brazilian anthropologists, among whom the figure of Eduardo Viveiros de Castro (1996a, 1996b, 1998a, 1998b, 2004a, 2010a, 2012a, 2014) stands out as its main proponent and theorist, followed by Tania Stolze Lima (1995, 1996, 1999a, 1999b, 2000). To explain the characteristics and scope of perspectivism, I begin with its meaning in Amerindian ethnography, which will help to reveal gradually its main characteristics as an anthropological theory and the issues and challenges it poses, particularly for archaeology. Subsequently, I describe the various manifestations of perspectivism beyond the lowlands of South America, laying the foundation for a discussion of the specifics of the concept of materiality and the nature of objects in perspectivism in the following chapters.

The Genesis of Perspectivism

The sense that some native groups in South America had a particular way of comprehending animals, spirits and some material objects as animate that differed from traditional animism can be found in many Amazonian ethnographies from the twentieth century (see Reichel-Dolmatoff 1996). Likewise, a variety of ethnographies of native North American groups, such as the Cree, the North Pacific coast cultures (for example, see Hallowell 1960, 1975, on Ojibwa ontology) and the Zuni, pointed towards this particular cosmology. The first concrete insight into perspectivism and the concepts and precepts associated with it relate to an earlier concern of Brazilian anthropologists with the idea of the person (Seeger, Da Matta and Viveiros de Castro 1979; see also Lévi-Strauss, 1963, 1964, 1967). To a great extent, the theory was also anticipated in the fieldwork and dissertation of Viveiros de Castro (1992) among the Araweté, ‘From the enemy’s point of view: Humanity and divinity in an Amazonian society’, in which he wrote about the concept of the person and cannibalism. In the mid-1990s, a number of nearly simultaneous publications on Amazonian groups not only sparked interest in Amazonian peoples in contemporary global anthropology but also highlighted the principles of perspectivism and other ontologies (Århem 1990; Descola 1992, 1996; Gray 1996; Lima 1995, 1996; Rivière 1994, 1995; Viveiros de Castro 1992, 1996a). Philippe Descola’s (1992, 1994, 1996) recouping of traditional anthropological concepts such as animism and totemism was particularly impactful in an international context, where the relationship and limits between nature/culture and human/non-human were being discussed and a new debate on animism was gaining momentum (Ingold 1995, 1998; Bird-David 1999).

Three publications made prominent contributions to the understanding of Amazonian ontologies. With hindsight they are perspectivist in content, while at the time they were seen as describing an animist ontology. The most important was an article written by Kaj Århem (1990), a Swedish ethnographer who worked with the Makuna people in Colombia; the other two were written by Peter Rivière (1994, 1995), a British anthropologist working with Amazonian groups in Suriname. Based on his 1989 fieldwork among the Makuna of Colombia, Kaj Århem (1990) proposed novel ideas about the way the Makuna think about the relationships between nature and culture and animals and humans. His interests were centred on ecological relationships; he wanted to understand the use of forest resources and how this manifested in the Makuna cosmology. Århem (1990, 2001) argued that ecological knowledge that translated into beliefs resulted in more harmonious and balanced environmental relations and management. The Makuna worldview held that humans and other living things were related in spirit and substance by a common human-like soul. Any living being (including plants) could be referred to as a ‘person’, all having a material and an immaterial form, or intangible spiritual essence (Århem 1990: 112). The cosmos was a large community organized under the same principles as human society, with all beings living a human life in their own worlds (Århem 1990: 125). Not only did they live in their own human-like world, but they were in fact human beings, their true form hidden beneath animal bodies and skins. All this implied a mode of interaction with nature completely different from a mode in which human beings define themselves as radically different and superior to other beings. It was a comprehensive system of ideas, values and practices that made up a true philosophy of nature, or ‘ecosophy’, given its ecological emphasis.¹

Århem made a further insightful observation, though he did not develop it: that living things, as humans, had their own point of view or perspective. As humans, he noted, they saw the world as we do, but from the perspective of their own species. Thus, what was water to humans was beer to a deer; what were fish to vultures were worms to people, and so on. For the Makuna, animals are organized into communities and do in their worlds the same things as humans do in theirs. They have their own territories, homes, rituals, customs and objects. In a word, they have ‘culture’. Crucially, it is not just that animals live as humans; they are in fact people. There is a shared identity between humans and other beings based on a shared spiritual essence (soul) and a common mythical origin. Fish, prey and humans are different bodily manifestations of that shared spirituality. This common essence becomes evident when one of these species transforms into another. Fish can become prey animals, prey can transform into fish, and both become people when they enter their homes: ‘they strip off their skins and put them back on when they leave, while people remove their shirts’ (Århem 1990: 121). That is, different kinds of living beings institute different worlds appropriate to their particular needs and characteristics. Despite the disparities in the external forms, a shared spirituality circulates among them: bodily appearances conceal a common internal unity. Yet an interrelated whole exists, a diverse society united by a cosmological order.

The various forms of life are analogues of each other but differ in the social groups each form. Moreover, just as the relations between social groups are reciprocal and mutually dependent, the same applies to the relations between humans and animals. Hunting, for example, is a type of exchange between beings. At a very general level, Makuna order the living world into three broad classes of entities based on relative position in the food chain: ‘eater’, a central ‘ego’ and ‘edible’ (food). The ego can be both an eater and edible. Humans can hunt deer in the same way that humans in turn can be hunted by a jaguar.

As noted, Århem (1990) described a typical feature of the Makuna worldview that he called a ‘perspective quality’. For him, a perspective position is one that sees the world from a specific point of view dependent on the ‘viewer’. The capacity to see the world is shared but always manifests from a particular point of view according to the class of being. Something that appears as one thing to humans is different for other species, though each perspective is equally true and valid per se. For vultures, maggot-infested rotting carcasses are really rivers full of fish; as noted, for humans, what deer drink is water, while for deer it is beer. Shamans are particularly qualified to see these different worlds or, more correctly, to see worlds from the point of view of other species. Humans, as a species, are decentred; their point of view is only one among others. There is no single true or correct representation of the world – there are many.

Though Rivière has not been as influential as Århem, he elucidated concepts that enabled perspectivism to be seen as a shared phenomenon among many lowland groups in South America. Around the same time as Århem’s article was published, Rivière published an article in Portuguese (Rivière 1994) and one in English (Rivière 1995) in which he made a similar argument about the body and interiority, though not about points of view. Rivière identified a defining feature of Amerindian ontology: that the native people of the Amazon live in a highly transformative world where appearances can be deceptive. Using the acronym ‘WYSINWYG’ (What You See Is Not Necessarily What You Get), he focused attention on the difference between external appearance and internal reality. External forms can be put on and taken off, like clothing, and conceal an underlying reality. Although he based his conclusions on Amazonian myths, he nevertheless argued that transformations occurred not only in mythical space-time but also in the everyday world (Rivière 1994: 256). Rivière emphasized transformation and instability in living beings, changeable external appearances (that can be reversed) and the presence of an internal ‘soul’ or spirit. Thus, a human form can conceal an animal interiority and vice versa. Importantly, Rivière intuited a shared communality among living beings as well as a capacity for transformation based on the bodily surface or outer covering. Although he is silent on perspective and worldviews, he makes a brief comment that will prove important for perspectivist theory: that behaviour is a better guide than appearances in everyday life, as ‘it is never safe to believe the evidence of your own eyes. It is better to wait and see what transpires’ (Rivière 1994: 261). As we will see, what living beings do – their affects and behaviours – is what differentiates them as species, not their soul or external bodily form.

Animism and Perspectivism

Amerindian studies and their relation to anthropological theorizing were brought to the fore by the work of Philippe Descola (1992, 1996) among the Achuar from the Ecuadorian Amazon and his reformulation of the concept of animism. His general approach aimed to establish general schemes of praxis arising from the operation of underlying principles of reality constructed by different cultures. Descola (1996) considered the objectifications of nature and otherness as key problems. In addition, Nurit Bird-David’s (1999) reformulation of Tylor’s (1871) traditional concept of animism as a relational epistemology had a significant impact and opened up new horizons of inquiry on the subject. Descola (1996) argued that the principles underlying the construction of social reality should not be sought in the relationship between human beings and their natural environment, but rather in operational schemes originating in cognitive devices shared by all humans. It is through their operation that the social objectification of nature is implemented, resulting in native taxonomies, for example. In such schemes, ‘nature’ is always constructed with reference to the human domain, which is how animism is generally explained. The difference is that Descola noted that these schemes are consequently informed by ideas and practices relating to ‘self’ and ‘otherness’. For the author, there is a homology between the ways in which ‘nature’ and ‘others’ are conceived, thus highlighting continuities and discontinuities between the social and natural domains (Descola 1992: 110–111).

At the time, Descola (1992) proposed totemism and animism as two distinct processes by which humans impose continuity between the domains of nature and the social. Following Lévi-Strauss (1963), he argued that totemism makes use of empirically observable discontinuities between natural species to conceptually organize an order that delimits social units. Totemism expresses a relationship between a social group and a natural category; discontinuities in nature are signs used as metaphors for differences in the social order (Descola 1992: 124). In contrast, the traditional anthropological notion of animism is that it endows natural beings and things with human dispositions, with will and intention or agency. Moreover, natural beings possess their own spiritual principles, making it possible to establish personal relationships with them, including relationships of protection, hostility and exchange. Animate entities have the status of persons with social attributes, including social hierarchies, kinship relations and behavioural norms. Unlike totemism, in which relationships between natural species provide a conceptual ordering for society, animism holds that social categories order the relationships between humans and natural species. For Descola (1992: 114), animist systems are a symmetrical inversion of totemic classifications: whereas totemism expresses a relation between a natural category and social aggregates, animism posits relations between individual entities treated as single persons, whether a natural category, plant or individual animal. Descola’s work has had a significant impact on anthropology, consolidated by the subsequent publication of the more detailed and complete versions of his model some years later, in which he includes analogism and naturalism as two additional modes of relationship and identification between nature and society (Descola 1996, 2006, 2009). Descola’s conceptualization of animism allows for equivalence between humans and animals as persons, as well as the recognition that there are no fixed boundaries between nature and society. It was not hard, consequently, to interpret Århem’s perspective quality and Rivière’s human–animal transformations in terms of Descola’s animism. Though several authors (Alberti and Marshall 2009; Descola 2006; Halbmayer 2012; Harvey 2006; Holbraad and Pedersen 2017; Pedersen 2001; Willerslev 2007), and even Århem (2016) himself more recently, consider perspectivism in a broad sense to be a variant of animism (a ‘companion concept’, as Århem puts it), it nonetheless has some unique characteristics that distance it from animism as a distinct ontological phenomenon (Lima 1996, 1999b, 2000; Viveiros de Castro, 2002a, 2011a).

While for Descola animism is a set of practices in the world and a classificatory system in anthropological theory, for Viveiros de Castro (2009, 2014) perspectivism is a theory about the world and not a classificatory system, though it does involve practices. Even as a set of practices, however, it differs from Descola’s animism, since two entities cannot be human at the same time (Viveiros de Castro 2004a). Relations are not ‘animist’, since relations are between people and other entities under their non-human form rather than between people and other people. As Viveiros de Castro (2014: 69–70, emphasis in original) argues,

Neither animism, which would affirm a substantial or analogic resemblance between animals and humans, nor totemism – which would affirm a formal or homological resemblance between intrahuman and interanimal differences – perspectivism affirms an intensive difference that places human/nonhuman difference within each existent. Each being finds itself separated from itself, and becomes similar to others only through both the double subtractive condition common to them all and a strict complementarity that obtains between any two of them; for if every mode of existent is human for itself, none of them are human to each other such that humanity is reciprocally reflexive (jaguars are humans to other jaguars, peccaries see each other as humans, etc.), even while it can never be mutual (as soon as the jaguar is human, the peccary ceases to be one and vice versa).

The debate over whether perspectivism is just one of several variants of animism remains unresolved.² If we define animism in a broad sense – or, as Århem (2016) and Ingold (2006) suggest, in terms of a ‘new animism’ – encompassing a spectrum of cultural phenomena and characteristics that are associated with the interrelationships between beings and souls or spirits (Harvey 2014), perspectivism could be classified as animism. That is, it would be part of ‘a propensity among indigenous peoples worldwide to anthropomorphize non-human beings and things – i.e., the notion that not only human beings have soul (consciousness, will, intentions) but also animals, plants and a whole host of other objects and phenomena’ (Århem 2016: 4–5). But of course, it is precisely the richness and particularity of perspectivism that is lost by labelling it as and fitting it within such a broad description of animism.

Perspectivism as Cosmology

Perspectivism as a distinct cosmology appeared simultaneously in articles by Tania Stolze Lima (1996) and Viveiros de Castro (1996a) in the same volume of a Brazilian journal published in 1996. Each author contributed in different ways to its definition. This conjunction, as the authors acknowledge, was the result of a dialogue between the two. Lima (1995, 2005) had completed her dissertation among the Yudjá or Juruna,³ a Tupi group, under the direction of Viveiros de Castro at the National Museum of Rio de Janeiro, Brazil, while the latter had been working on questions of otherness, personhood and the body among Amazonian groups for many years (Seeger, Da Matta and Viveiros de Castro 1979; Viveiros de Castro 1979, 1992 1996a). In her 1996 article, Lima developed the idea of ‘point of view’, drawn from the relationship between humans and animals in the cosmology of the Yudjá. An analysis of the relationship with peccaries in hunting helped her to highlight and define the difference between the concept of point of view and animism. In the same article, she elucidated the relationship between the soul, the body and the individual and in the process discarded the concept of cultural relativism as an adequate framework to understand the nature–culture relationship. Based on myths and ethnographic records, Lima (2000) writes, peccaries for the Yudjá are people. And, as such, they have a human spirit and way of life in their own domains but the outward appearance of an animal; that is, they have a perspectival quality that makes them see things from their own world in a human way. However, Lima also found that the Yudjá actually recognize peccaries as animals, not humans; the key point, in fact, being that peccaries consider themselves to be human (Lima 1999b: 113). Humanity is not an intrinsic property of humans, but a characteristic that must be produced, which many entities can do, particularly animals. Animality, Lima (1999b: 115) argues, is a form of other-consciousness, while self-consciousness refers directly to the human. In fact, Lima (2000: 45; original emphasis) found that the concepts of nature and culture have no counterparts in Yudjá cosmology:

humans do not belong to the class of animals, nor do they distinguish themselves precisely from the latter through the possession of culture, language and social life.

These last three functions are primarily related not to humans but – how to put this? – to the living beings that inhabit the different regions of the cosmos, some of whom are defined as having souls and others as being souls. For animals, spirits and humans, having a soul means having awareness of oneself and others, being able to think, being a subject. Whoever thinks or lives in effect behaves like humans: in this sense, animals have an awareness of their own humanity, act in accordance with this, and consider humans properly speaking to be their similar; in turn, the souls of the dead think of themselves as living people.

Here, culture denotes a universal function which is simultaneously defined as thought and sociality (and is, therefore, neither a domain isolated from an exterior reality, nor a distinctive function of humanity in opposition to animality).

Although Yudjá cosmology is at first sight reminiscent of animism – an anthropocentric worldview that takes humanity as a model for other animate entities – the author argues that it is not the same, and particularly not the same as Descola’s version. The human qualities held in common have more to do with differences than with similarities: for the Yudjá, the relation of identity between humanity and animality is a condition for highlighting their differences. Though both kinds of being have culture, they differ in eating habits, artefacts, musical instruments and spirits. Even the environments where each live are different. They are different social groups living in their own worlds with their own particular modes of action and subjective experiences. That is, each class of being has its own point of view. ‘Point of view’, Lima (1999b: 117) explains, ‘implies a particular conception according to which the world only exists for someone.’ Reality depends on the subject; that which exists, exists only for someone. Therefore, there is no objective or external reality or nature (in Western terms) that is interpreted differently by different classes of entities, as relativism would claim. The same referent (thing or event) is seen as irreducibly different – a condition that implicates not only what you see and the nature of material things but also actions such as hunting, war and daily life. Even sounds are experienced differently (flute music to peccaries is a whistle to humans). In short, there is no external world of reference, and humanity is not a single shared essence, but rather there are co-occurring parallel and simultaneous dimensions or worlds (Lima 1999b: 120–121).

Lima (2000) concludes her reflections on Yudjá perspectivist cosmology with a discussion of the nature/culture divide and differences in Western and Yudjá dualisms. For the Yudjá, there are three basic categories of living beings: humans, animals and spirits. These are not separate classes in the usual usage of the term. Each may contain aspects of the others or may even be contained by the other. While humanity is characteristic of animals, it is also characteristic of spirits, while animality and divinity are also characteristics of some humans, and spirits can live as souls or as concrete entities. Moreover, every being can become another. In a word, all three ‘classes’ of entities contain part of each other: human, divine and natural characteristics are found in humans, animals and spirits. Entities may have different quantities of each of these characteristics, making them more or less akin. More importantly, entities are not ordered by opposition, since each contains something of the other two (Lima 1995: 59–60); if they were, Lima (2000: 47) argues, there would be three relations of opposition, human/non-human, animal/non-animal and spirit/non-spirit, which are not defined by their reciprocal opposition but as a triad of oppositions or ‘group of oppositions’.⁴ For example, a collared peccary may actually be a spirit rather than a peccary or an animal. Humanity, moreover, is divided into the Yudjá proper, the forest people and the white people. Each is differentiated from the other humans by different values: the Yudjá are simply human, the forest people are linked to animality (by their eating habits) and the white people are closer to the divine shaman who created humanity (due to their technology and mobility) (Lima 2000: 47). Yudjá cosmology thus implies a potential asymmetry in relationships between entities. Sometimes a human point of view prevails, for example, in the encounter with prey; but at other times, it may be the point of view of a spirit that is foremost. Asymmetry, according to Lima (2000), derives from the capacity of one subject (human, animal or spirit) to impose its point of view on another. This capacity can be reversible and is not an a priori. It is a relational and situated occurrence where the dominance of one entity over the other is established as a result of the encounter. Importantly, the differences between entities are not fixed but unstable and relative. The distance between humans and animals, therefore, is not constant: it depends on which entities are in relation and their species characteristics.⁵ Lima (2000: 48) takes as an example the differential relationships between humans and animals: the differences between Yudjá and white-lipped peccaries, for example, are not the same as those between humans and collared peccaries or jaguars (Lima 2000: 48). A point of view is entirely relational and relative.

Perspectivism as a Native Theory of Reality

Thus far, I have provided a summary of the important texts that help delimit perspectivism as a characteristic phenomenon of Amerindian cosmologies or ontologies. Eduardo Viveiros de Castro has undoubtedly made the most important contribution to the subject and has gone a step further than other authors by proposing perspectivism as a theory of reality or, more accurately, as a native anthropological theory (Viveiros de Castro 1996a, 1998a, 2004a, 2009, 2013, 2014). Perspectivism emerged as a dialogue between Lima and Viveiros de Castro, as noted above. Lima (1996, 1999b, 2005), too, writes of perspectivism as a theory, but her focus was on the Yudjá case and the key concept of ‘point of view’. Although Viveiros de Castro (1996a, 1998a) in his initial writing on the subject introduced perspectivism in relation to a discussion of the categories of ‘nature’ and ‘culture’, his real goal was to establish the basis of this native theory. In what follows, I synthesize Viveiros de Castro’s writings on perspectivism and the latter’s status as theory.

Perspectivism and point of view are terms that had been used in philosophy, particularly by Leibniz and others after him, such as Nietzsche, Whitehead and Deleuze, all of whom influenced Viveiros de Castro. Viveiros de Castro (Viveiros de Castro 2007: 121, 2009) has recognized as particularly important Deleuze, along with his collaborator Guattari (Deleuze and Guattari 1977, 1987), as well as the perspectivist ideas of Tarde, von Uexküll, and Ortega y Gasset (Viveiros de Castro 2007: 121). The clearest connections with perspectivism as a theory are the perspectival ideas of Leibniz in relation to the point of view and the constitution of the subject and world, as well as Deleuze’s analysis and development in The Fold: Leibniz and the Baroque (Deleuze 1989, 1993), together with his lessons (Deleuze 1986).⁶ Viveiros de Castro essentially put into effect Deleuzian philosophy in anthropology. Even so, however, he does not project Western philosophy onto Amerindian ontology and epistemology but instead finds philosophical statements that help to understand and translate them in Western terms. Furthermore, just as Deleuze and Guattari in Anti-Oedipus (1977) intervene in current social issues, Viveiros de Castro’s Anti-Narcissus engages contemporary politics, considering anthropology to be the theory and practice of the permanent decolonization of thought.⁷ One of his questions is: ‘what does anthropology conceptually owe to the people it studies?’ (2010a: 14, my translation). The antecedent authors contributed to the delineation of that ‘conceptual complex’, as Viveiros de Castro (2009) calls it, of perspectivism.

Perspectivism characterizes a set of principles shared among many Amerindian populations beyond local variation, as outlined in the Preface. To reiterate, these principles are the quality of an internal human soul possessed by some entities; the importance of the body as an entity’s distinctive character; an entity’s ability to have a human point of view; and predation as a model of human relatedness or kinship. Viveiros de Castro (2004b: 3–4) summarizes perspectivism as follows:

I use ‘perspectivism’ as a name for a set of ideas and practices found throughout indigenous America and to which I shall refer, for simplicity’s sake, as though it were a ‘cosmology’. This cosmology imagines a universe peopled by different types of subjective agencies, human as well as non-human, each endowed with the same generic type of soul, i.e., the same set of cognitive and volitional capacities. The possession of a similar soul implies the possession of similar concepts, which determine that all subjects see things in the same way; in particular, individuals of the same species see each other (and each other only) as humans see themselves; that is, as beings endowed with human shape and habits, seeing their bodily and behavioural aspects in the form of human culture.

Additional principles are derived from these fundamental cosmological premises, which also imply specific explanations for the way things are and have consequences for social practices (which are important for thinking about the archaeological record in perspectivist terms). These principles are explained in greater detail in what follows.

Subjectivity, the Soul and the Body

Amerindian perspectivism is a conception among some Indigenous peoples according to which the world is populated by different entities, agents or persons (i.e., certain animals, spirits, objects, phenomena of nature, artefacts and plants), all of who consider themselves to be subjects, and who, as such, see the world in the same way as humans. Humanity – and culture – is a condition shared by all entities that are considered to have subjectivity, that is, to have a point of view. ‘Everything that possesses a soul is a subject’, Viveiros de Castro (2004a: 467) writes, ‘and everything that has a soul is capable of having a point of view’. According to many Amerindian myths, all entities – animals, plants, objects, people, meteorological phenomena and geographical accidents – were originally human (persons). However, all except for humans lost that general condition and became what they are today. Nevertheless, they retained an anthropomorphic soul behind their present appearance. This external appearance, which is a kind of shell, is not a disguise that hides an essence but a tool that gives a specific form to the indeterminate universal human spirit (Viveiros de Castro 2007: 101). The skin, or external surface of the body, is like clothing that can be taken off to show the true human entity that lies beneath it. Animals alone in the forest or fish in the river, for example, take off their clothes and live a human life in their human world in a human form. Transformation is a typical widespread phenomenon of body change in the ever unstable and transformational world of Amerindians. Through it, some entities can become others: shamans can become jaguars, animals can become other animals, humans can be transformed by other beings into animals and artefacts can become animals. In short, while the internal form of the animal, the ‘soul’ or ‘spirit’, is an intentionality or subjectivity like human consciousness, the external form also materializes a human bodily scheme, but one that is hidden behind an animal disguise.

Since bodies can change, they are considered unstable forms. They are not conceived as fixed nor as the distinctive substance of a species, since a body can hide another species. Difference lies in the point of view of each body, and this will only be perceived when entering into relationships – through habits, dispositions and behaviours. As such, it is what subjects do with their external bodies as much as how they appear that is important. Entities are specified more by their ways of being, by their habitus and their associated affects and capacities, than by a stable form or by their physical substance. Viveiros de Castro (2004a) summarizes it this way: ‘the point of view creates the subject’. Any entity to which a point of view is attributed is a subject. And affects and behaviours are precisely the clues that help distinguish an animal when it presents itself in human form; it is a way of knowing the other in front of you is a subject.

It is worth clarifying that not all entities in the world have ‘souls’. Animals commonly do, but not all of them. In general, animals with souls are animals that play an important symbolic role for humans or those involved in predator–prey relations. These animals include large forest predators such as jaguars, panthers and boas, birds such as vultures and game animals such as peccaries, deer, wild boar, some monkeys and fish. Other kinds of entities, whether immaterial like spirits or solid like landscape features such as mountains, rocks, water and even some plants (certain trees and cassava, for example) (Descola 2013), may also have a soul and a point of view. Many artefacts can be people, including mundane everyday objects such as hammocks, baskets, pots, tobacco pipes and stools, as well as those used only exceptionally, including musical instruments such as flutes or drums (Hugh-Jones 2009; Lagrou 2009a; Santos-Granero 2009a).

While not all entities in the world have a point of view and are persons, there is no impediment to any of them gaining such. A human essence is not the property of an individual being but the consequence of a relative position. All entities have the potential to be considered human at one time or another according to the relationships they establish and the context of interaction. Moreover, varying degrees of personhood may be associated with different entities. Some objects are animated by their association with a subject, for example. The archetypal human is clearly the complete subject, one with the capacities of self-reflection and intentionality. As such, as Viveiros de Castro (2007: 36, my translation) explains, Amerindians do not claim that animals are persons qua persons: ‘Everyone in their right mind, and that of the Indians is as healthy as ours, “knows” that the animal is an animal and that the person is a person. As Derrida says somewhere, even animals know it. But under certain points of view, in certain contexts, for the Indians it makes perfect sense to say that some animals are human.’ Having a point of view is a matter of relative position, degree and context.

The Point of View

Humanity and soul are thus formally identical qualities for all entities, and the ability to have a point of view is a matter of the soul. The differences in viewpoints, however, do not reside in the soul but originate in the specificity of bodies, since the soul is of the same type for all beings. Each member of a subject species sees itself as human with a human body; but other kinds of entities are seen as animals, spirits or non-humans. Each species, for itself, occupies the place of humanity (i.e., what Western science designates as the species Homo sapiens) and sees itself as occupying that place in relation to the rest of the cosmos, which is non-human. The fact that no other being will be human except their own kind implies that each species perceives otherness differently. Thus, humans see peccaries as animals, but peccaries see themselves as people and see humans as predators or enemies; or what humans see as blood, jaguars see as cassava beer. The entire world changes according to the point of view; but the quality of ‘point of view’ – the way of seeing itself – is always the same because each species considers itself human and therefore always has a human point of view. Moreover, to be human – to have a human point of view – is to have culture. Everything seen is seen through the lens of human culture. Animals see the same elements of culture that we humans see: houses, food, objects and so on. That is, they see a human world. Importantly, as things are seen from each species’ own point of view – its body – the same thing is seen as a different thing from another species’ point of view (blood is beer, or fish is bread, etc.). As Viveiros de Castro (2004b: 3–4) states:

What changes when passing from one species of subject to another is the ‘objective correlative,’ the referent of these concepts: what jaguars see as ‘manioc beer’ (the proper drink of people, jaguar-type or otherwise), humans see as ‘blood’; where we see a muddy salt-lick on a river bank, tapirs see their big ceremonial house, and so on. Such difference of perspective – not a plurality of views of a single world, mind you, but a single view of different worlds – cannot derive from the soul, since the latter is the common original ground of being; such difference is located in the bodily differences between species, for the body and its affections (in Spinoza’s sense: its capacities to affect and be affected by other bodies) is the site and instrument of ontological differentiation and referential disjunction.

Humans generally do not see animals as persons, and, reciprocally, animals do not see humans as such, since from their own points of view, each sees itself in the human corporeal form, as a member of the human species.⁸ One characteristic of these multiple perspectives is their incompatibility. Two species cannot simultaneously see themselves as persons: a human and a jaguar cannot be persons at the same time. It is not possible to see blood as beer without first becoming a jaguar. Only shamans have the capacity to temporarily acquire the point of view of the other and thus see the world from the perspectives of those others – although this implies a prior transformation of the shaman into the other.

Culture is one and the same for all subjects, but nature – the world that is seen – changes according to point of view. While there is only one culture (human), there will be many natures. This is the basis of what Viveiros de Castro calls ‘multinaturalism’ which he opposes to multiculturalism. The latter holds that one nature exists for all humans that are divided by multiple cultures, each of which views that same nature differently. Multinaturalism implies that perspectivism is also a way of knowing and ordering the world, that is, an epistemology. Perspectivism it is an ontology proper – a multiplicity of perspectives on the real, a multinaturalist ontology – and simultaneously an epistemology, a way of knowing those realities. The exchange of points of view is always at stake while interacting with the world, while trying to know and order it. For example, to know an other the shaman must change their point of view. Implicit in this process is an acquired knowledge of the other, as both points of view participate, that of the shaman and that of the other. There is a certain performative mode of perception involved here, in which the action of ‘seeing’ is a constitutive element of the other entities, human or non-human (Viveiros de Castro 2010a: 35), like Rivière’s WYSINWYG. Perspectival knowing is not knowing ‘the world’, then, but knowing many worlds at the same time as there is, ‘one culture, multiple natures – one epistemology, multiple ontologies’ (Viveiros de Castro 2004a: 474).

Multinaturalism may lead one to think that perspectivism assumes the existence of an objective external reality that is apprehended in different ways according to each species, but this is not the case. As Viveiros de Castro (2012a) argues, Amerindians do not imagine that there is one thing in itself that humans see as blood and jaguars as beer. There are no self-identical substances categorized differently, but rather ‘immediate relational multiplicities’ (Viveiros de Castro 2012a: 34), depending on who one enters into relation with. Importantly, then, these multiple worlds are not representations of an objective reality that is out there and that is interpreted (represented) from different points of view. Perspectivism is not a type of cultural relativism; nor are these worldviews. Worlds are not representations, because representations are a property of the mind or spirit, while the point of view is found in the body (Viveiros de Castro 2004a: 474). In addition, since the constitution of subjects – on both sides of an interaction – is a relational and partly performative occurrence in which there are no fixed essences or stable materialities, there is no room for a representation. Things simply are as they are seen by each species of subjects. And, given that all share the same culture, they see the same type of things. In other words, different beings see (interpret) the world in the same way, using the same categories. What changes is what they see (i.e., there is one culture and many worlds). There is no possibility of thinking in terms of representation, therefore, and things do not acquire specific meanings as signifiers. In representational terms, they would be the same representation of different things: the same meaning but multiple referents. Such things are therefore closer to ‘symbols that stand for themselves’ (Wagner 1981; see Viveiros de Castro 2009) – as there are many external natures to take as references, they are at the same time their own symbols and referents. Moreover, things change their character as entities according to the perspectives – the subjects – involved in the relationships. For example, things will be the same for everyone in relations between beings of the same species, but in an interspecific exchange things will be different: beer will be blood, fish will be worms and so on.⁹

Predatory Relationships

Any being can become human to another due to the simple fact that it can transform or show its human side hidden beneath its appearance. This has two important consequences: first, that humanity is relational and to an extent contextual; and second, that that any relationship with the outside world, or ‘nature’, can be a social one, a relationship between subjects. There are nuances to relationships, but the general principle that characterizes them in the Amazon is the model of a predator–prey relationship between subjects. Relationships of affinity and exchange are dimensions of the same (as in cannibalism or war; see Viveiros de Castro 1998a, 2004a). Consequently, jaguar-human sees humans as peccaries (prey); humans see peccaries as prey; peccary-humans see humans as jaguars, and so on. It is a continuous chain of eat or be eaten (‘a trophic chain’) in which entities find themselves – as subjects – always in one of the two relative positions of predator and prey (Viveiros de Castro 2004c, 2007: 96). Even so, these are social relations: both the prey and the predator believe that they are human – or, at least, that the other could potentially be human. These predatory relations – or the logic of predation more appropriately – is a basic and pervasive regime of sociability, which is expressed symbolically – and sometimes materially as well – in other domains. The same logic is found in the capture of enemies as an appropriation of the other and as a form of constitution of the self. Spiritual and physical attributes of the other are incorporated into the self – the soul, names, words, memories, trophies and objects (Lagrou 2007; Vilaça 2002; Viveiros de Castro 2010a).¹⁰ Predatory relations can go from an extreme of de-subjectification – as in the case of hunting, where the ontological status of potentially human prey must be changed by shamans to avoid cannibalism – to that of the necessity to subjectify the other – as in the case of war or enemies, who must become human beings so that they can be incorporated by their captors. Since objects also have a subjective dimension, procedures must be carried out to de-subjectivize them and turn them into mere objects. Common procedures include intentional omissions in the decoration or in the shape of the body when making the object, or deliberately leaving it incomplete, to prevent its subjectivity and intentionality from becoming active (Lagrou 2009a).¹¹ The ability to de-subjectify belongs to shamans who, by taking the place of the other, acquire their point of view.

Since the personality or subjective dimension of non-humans is not immediately apparent, an encounter with an entity that can become a subject can be a dangerous situation. In such enchained predatory relationships, the other entity may become the predator. The human being is not physically consumed, but rather their soul is appropriated, and they are consequently rendered by the predator into one of its own species. This process takes place through an exchange of points of view, where the dominant party in an asymmetrical relationship is the strongest. The other will make the human being adopt its point of view without the human being realizing it. The consequences for the human being are serious illness, the adoption of behaviours belonging to the predatory subject (from the perspective of other human beings) or the realization that they are dead and already live in the world of the other. In the first two cases, the shaman has to intervene and transform themselves into the species of the other to recover the stolen soul. In the third case, the human being realizes in time that the other is not a human but an animal or spirit; they become suspicious, perceiving details in the body or behaviour of the other that are not entirely human, though carried out in a human social context. In fact, reaching this point means that the human being has been caught by the other and is already seeing the world from the perspective of that other.¹² Only shamans can see what is going on or change their body and become another at will. They can also recover the soul of a subject so transformed. Relationships with other entities, whether they have a point of view or not, are always fearful. If humanity is a latent possibility behind any existing thing, then any relationship with the external world – or ‘nature’ – should be conducted with care and caution.

A further important form of relationality between subjects in perspectivism concerns the construction of affinity and identity. Revolving around relations of commensality, it represents another way of appropriating or gaining access to the subjectivity of the other (Fausto 2002, 2007, 2012). In this case, there is no exchange of points of view, but rather transformation in relation to food consumption and ways of eating, with the ingestion of other beings as a form of incorporation. Eating as and with someone is a process of familiarization in which the dispositions of the other are acquired and a shared identity is constructed. That is, eating is a way of forming people of the same species, including relatives. What is eaten matters. The case of hunted animals is particularly interesting as in the process the animal must be turned into an object; it must be de-subjectivized and thus transformed into food. Otherwise, the act would one of predation and not of feeding, with the resultant possibility of identification with the prey as a subject in what would become an act of cannibalism. Someone would be eaten rather than eaten with (or as). One way of converting prey into inert food is to cook it – eating raw meat, as in the case of enemies, results in the acquisition of the abilities of the victim’s soul.¹³ Hence the importance of culinary codes, and cooking fires, and the significance of which species are eaten, since not all animals have subjectivity or possess it to the same degree (Fausto 2007: 503). Predation and commensality are two different but dynamically articulated ways of producing people and socializing in the Amazon.

In this chapter, I have described in some detail the basic principles and general characteristics of Amerindian perspectivism, which are shared with some variation among the various Amazonian groups. Humanity (and culture) is a shared condition of many types of entities, hidden behind a distinctive body that provides subjectivity and a specific perspective on a world in which predation functions as a principle of generalized relationship. In what follows, we will see how perspectivism as an ontology can also be understood as an anthropological theory, on a par with those of the Western academy.