Ishii: Acting with things

This work is licensed under the Creative Commons | © Miho Ishii. Attribution-NonCommercial-NoDerivs 3.0 Unported. ISSN 2049-1115 (Online)

Acting with things:

Self-poiesis, actuality, and contingency in the formation of divine worlds

Miho Ishii, Kyoto University

The aim of this study is to investigate how divine worlds can be created, vitalized, and lived by people. Focusing not on cognition and operating through things but on bodily action with things, this paper examines the actuality of these actions, which occur prior to the cognitive articulation of the event and create novel experiences of the world. It reconsiders Alfred Gell’s theory of idolatry through the ideas of Bin Kimura and Hideo Kawamoto. Exploring the making of spirits in Ghana and spirit possession rituals in South India, this paper presents a fresh view of the formation of divine worlds as the actualization of virtual, vital relations between persons and things that emerge only through their contingent coactions.

Keywords: actuality, self-poiesis, coherence, divine worlds, Ghana, India

Action is to move a thing and to be moved by it. We see a thing through our actions; the thing specifies me and I specify the thing. This is action-intuition…. From the standpoint of action-intuition, the self is born of the world of things. To be born of things is to see them truly.

Bin Kimura, Kimura Bin chosaku-syu, 2001


This article investigates how magical, divine, or sacred worlds (hereafter referred to as “divine worlds”) can be created, vitalized, and lived by people. Focusing on the contingent coactions of persons and things, this paper examines the actuality of these actions, which occur prior to the cognitive articulation of the event and create a reality that emerges as a person’s lived experience. Exploring the coactions of persons and things in the formation of divine worlds, this paper sheds new light on Alfred Gell’s (1998) argument on idolatry.

From the modernist point of view, magical-religious phenomena—for example a divine statue manifesting its power or a ritual affecting people’s fates—may be considered as belonging to the realm of peculiar religious beliefs. Here I argue that the formation of divine worlds does not necessarily depend on people’s beliefs in religious powers or in magical things. Rather, divine worlds are created through concrete relations and actions among persons, things, spirits, and deities that take tangible as well as intangible forms. A close investigation of the formation of divine worlds enables us also to see the formation of our “everyday” world from a fresh viewpoint.

To pursue this investigation, it is crucial to consider the work that “things” do as actors, which manifest their power in relation to humans. This is the approach associated with the recent debate on the “ontological turn,” which has been referred to as a “quiet revolution in anthropology” (Henare et al. 2007: 7–8), a revolution to which Gell has contributed as well.1 Gell elucidates the process in which artifacts such as idols and divine statues manifest their agency through a causal nexus of persons and things, which is not merely symbolic, but has real and practical effects. His argument exemplifies a sensibility common among “ontologists,” one that urges anthropologists to engage things encountered in the field as they present themselves, rather than to assume that they signify or represent something else (ibid.: 2). At the same time it must be noted that Gell attributes to human actors the role of “primary” agents, while he treats idols and other artifacts as merely “secondary” agents whose role is to motivate the abductive inferences made by humans. Thus, things are still regarded as the means or instruments by which humans can interpret the world, expand their abilities or transmit their own agency.2

To understand the process by which divine worlds are formed, however, it will be necessary to question the premise that human intentional and cognitive ability, even as it is manifested through things, is central. Rather, I propose that it is not so much thinking through, but acting with things that is fundamental. As we will see later, the notions of actuality and the “coherent”3 movements of bodies and things become crucial for our analysis. These ideas allow us to examine the power of things not just as secondary agents but as coactors with people in a movement that (re)creates all involved. This leads to a further consideration of not only the active but also the passive aspects of human coactions with things.

I start by introducing two sets of concepts concerning things, bodies, and formations of reality. The first comprises the complementary concepts of “reality” and “actuality,” elaborated by the Japanese psychiatrist and philosopher Bin Kimura (2001b, 2005). The second elucidates the notions of “self-poiesis” and “coherence”: ideas stemming from the theory of autopoiesis uniquely developed by the Japanese philosopher of science Hideo Kawamoto (1995, 2006). Based on these ideas, I examine the emergence and formation of divine worlds as a self-poietic process shaped by the coherent actions of bodies and things. Next, I examine Gell’s theory of idolatry, focusing in particular on his notion of nexus formation, and I reconsider Gell’s arguments from the viewpoint of the actuality and coherent movements of bodies and things.4 This discussion is based on my field data on the making of spirits in Ghana and on spirit possession rituals in South India. I focus, in particular, on the simultaneous activity and passivity of human experience during these events and on the unpredictable and fugitive features of the relations that constitute divine worlds.

From this discussion, I go on to explore the emergence and formation of divine worlds, not as the realization of one possible alternative world, but as the actualization of virtual, vital relations between persons and things that emerge only through their contingent coactions. Finally, from the viewpoint of actuality, I explore the experience of a person in the midst of these actions, one who is aware of, and lives in, the nexus created by humans and things.

Actuality and reality

According to Kimura (2001b, 2005: 10–14), two aspects or operations of the self can be distinguished with reference to the experience of living “in reality”: one is the cognitive operation of perceiving and recognizing the world as a given environment; the other is the bodily action through which reality emerges for a person.5 Kimura explains this dual aspect of experience using the concepts of reality and actuality, drawing on the Latin words res (thing) and actio (action). As a conscious subject, Kimura argues, a person usually recognizes the objects perceived by her as a given reality. At the same time, as a living being, she also relates to the objects/environment prior to her recognition or articulation of them as a reality. This primary, vital connection between a person and the objects/environment remains in a virtual state, inasmuch as it is hidden from her consciousness. Nonetheless, fragments of this connection incessantly flash into consciousness via the medium of the body, or the prepersonal self. This contingent flash of virtuality gives the person the sensation of actuality at any given moment of life. At the moment when virtuality rises and becomes individuated, actuality works as the medium to stimulate the emergence of reality in conscious experience.6 A reality thus, does not exist for a person simply as a given setting, but can emerge and be experienced only through the work of actuality in a person’s vital relations with things. Here, the work of the prepersonal bodily self is considered to be more fundamental than the cognitive operation of consciousness.

Kimura’s argument indicates how essential the embodied interaction with things is for the way in which reality is experienced. In a related vein, Kawamoto (2006) considers the work of actuality and the emergence of reality in terms of autopoiesis theory, contrasting the “poietic” to the cognitive aspect of the self.7 According to Kawamoto, the poietic self is that which produces itself or emerges through bodily action: it relates deeply to the notion of kinaesthesia, or the practical sense of “being able to do,” rather than the cognitive sense of “knowing something.” Kawamoto calls this self-producing work, originating in the coactions of body and things/environment, “self-poiesis” (Kawamoto 2006: 15–50, 146–206).

Kawamoto’s argument thus calls attention to the importance not only of the cognitive, but also the poietic aspect of the self. Similar to Kimura, Kawamoto points to the primary function of bodily actions and also to the “delay” inherent in the cognitive function when a particular reality is about to emerge for a person.

Equally important to this investigation of reality formation is the idea of “coherence” between actions and the environment, originally proposed by the German neurophysiologist Viktor von Weizsäcker (1950) and further developed by Kawamoto. In Kawamoto’s terms (2006: 82–85), coherence refers to the relationship between a certain mode of action and the characteristics of the things or environment specified through the action. It can be observed, for example, in the relation between the active mode of a swimming body and the characteristics of the water in which it is immersed (cf. Deleuze 1994: 165). Also, coherence can be seen in the relation between the mode of making an artifact by hand and the characteristic of the thing specified according to the creation process.

The idea of coherence indicates that when a reality is formed through action, persons, and things move together and each of them specifies, transforms, and creates the others’ actions and qualities. It also suggests that, for a reality to be durable, the relation between the mode of one’s action and the characteristics of the things/environment specified by the action needs to become stabilized—at least to a certain degree—through a process of mutual adjustment.

At the same time, coherence between actors and things/environment is not always a smooth, cooperative relation; rather, it involves crisis (Krise), through which a new relation between them can emerge (Weizsäcker 1950: 170). In brief, coherent movements between actors and things/environment undergo a spiraling process of transition: (1) an encounter between actors and things, (2) the beginning of their coaction and mutual adjustment, (3) the creation and maintenance of a specific order or reality, (4) crisis and the dissolution of coherence, and (5) the emergence of a new relation between the actors and things/environment. Coherence is thus indivisible from crisis and emergence, though the link is usually hidden behind the apparent stability of a reality. As we will see later, this inherent instability of coherent movements is conspicuous in the formation of divine worlds.

Although both Kimura’s and Kawamoto’s arguments—one based in phenomenological psychiatry and the other in systems theory—are relatively unfamiliar to anthropological debates, they are still highly relevant to anthropological inquiry into the ways in which divine worlds are constituted. To understand the formation process of such worlds, it is necessary to consider peoples’ practices not only in terms of symbolic representations of their beliefs or cognition about reality, but as practical coaction and movement with things.

In order to understand the formation of divine worlds, however, it is not sufficient to consider the action of a sole person engaging with a single thing. Divine worlds are created through interactions and connections among various actors including persons, things, and deities/spirits: each of them affects its environment, which consists of other actors, and at the same time each constitutes a part of the environment whose characteristics are specified through the actions of others. The formation of divine worlds thus comes into view as a process in which actors, connecting with each other, create both themselves and their environment through coherent movements. Before considering this matter in more detail, however, it is necessary to think about how we can conceptualize the nexus of persons, things, spirits, and deities in the formation of divine worlds. Alfred Gell’s argument on idolatry, which focuses on the agency of both persons and things, is a useful guide in this exploration.

The nexus and agency of humans and things

In Art and agency (1998), Gell examines the formation of the nexus of humans and artifacts to show the process through which these actors or “social agents” manifest their agency. He explains the formation of the nexus of humans and things by using a set of concepts such as agent/patient, index, and prototype.

According to Gell, social relations exist only in so far as they are manifest through actions. Gell sees performers of social actions as “agents” acting on “patients.” Relations between social agents and patients take four different forms: (1) Indexes: “material entities which motivate abductive inferences and cognitive interpretations,” (2) Artists: “to whom are ascribed, by abduction, causal responsibility for the existence and characteristics of the index,” (3) Recipients: “those in relation to whom, by abduction, indexes are considered to exert agency, or who exert agency via the index,” and (4) Prototypes: “entities held, by abduction, to be represented in the index” (Gell 1998: 27). Below I show how Gell uses this conceptual apparatus in relation to three cases: what Gell calls “volt sorcery,” the Oro cult in ancient Tahiti, and kumari worship in Nepal.

Let us first examine how Gell presents volt sorcery as showing the linked work of agency in a causal nexus. This is a type of sorcery that aims to harm a person by injuring an image representing the victim. In Gell’s analysis, the victim, the image representing the victim, and the sorcerer making the image are described as the “prototype,” “index,” and “artist” respectively. Figure 1 shows the process in which the agency of the prototype is mediated via the index and consequently causes harm to the prototype himself. It is remarkable that the victim eventually becomes the victim of his own agency by a circuitous causal nexus. Regarding this sorcery, Gell points out that the action of making a representational image involves a kind of “binding,” which ties the image of the prototype to the index (1998: 102–3). This case illustrates the formation process of the causal nexus through which both persons and idols transmit and manipulate their agency.

Figure 1
Figure 1: The linked work of agency in volt sorcery (Gell 1998: fig. 7.3/1).

Next, let us turn to the Oro cult in ancient Tahiti, a case showing the distribution and control of exuviae. The divine objects at the centre of this cult are called to’o. Gell considers to’o as representing gods who have anthropomorphic attributes “aniconically” and thus indexing the god as prototype. According to Babadzan (1993), to’o, which are made of billets of wood, were usually tightly wrapped and protected from view, and only during a ritual called pa’iatua (the wrapping of the gods) were they revealed to chiefs and high priests. Gell refers to this ritual as the decortication of the divine exuviae.

Among the wrapping materials of the to’o, red feathers—themselves the exuviae of birds—were regarded as most important since they relate to heaven and life. In this ritual, new nonsacred feathers passed from the lesser priests to the higher priests of the primary to’o, and the feathers which had previously been in contact with the primary to’o were passed back to the inferior to’o. Accordingly, the old feathers, as the god’s divine exuviae, were distributed among the people as the currency of political control (1998: 110–13).

Analyzing this ritual, Gell points out the similarity of the Oro ritual to volt sorcery. In both cases, it is by being bound to its index that the prototype is subjected to control by humans such as sorcerers and priests. Even though Oro apparently rules the universe, in practice his power is controlled by humans, both via the indexes that are his exuviae and through the binding of the primary index of his person, the to’o. By binding the to’o, the chiefs and priests make Oro their “passive agent.” Gell thus concludes that at a critical point in the sequence of causes, instrument, and results, human agency is essential (1998: 114).

In both of the cases just examined, the objects representing the prototype manipulate its agency in a specific causal nexus. How, then, do people attribute agency to these objects? To answer this question, Gell discusses what he calls the “externalist” and “internalist” theories of agency-attribution. To put it briefly, in the former, an idol is attributed agency by its behavior, following certain social rules; while in the latter, an inner “mind” is attributed to an idol through human acts such as providing it with eyes, wrapping it, or excavating its body (1998: 126–37). Gell also insists that even a human can become the icon or index of a deity by undergoing the same procedure of endowing agency to things. We see this exemplified in the case of kumaripuja: the worship of the goddess Taleju in the form of a young virgin girl among the Newars of the Kathmandu valley.

According to Allen (1976: 306–7), in the installation ritual of the kumari, the chief priest performs a ritual to remove from the girl’s body of all her previous life experiences in order to enable the spirit of Taleju to enter into her. The girl sits naked in front of the priest who uses a small bundle of plants to purify each of her sensitive body parts. As he removes the impurities of the girl, the spirit of the goddess enters into her, and she is said to gradually become redder and redder. Then the girl is dressed and made up with a kumari hairstyle, jewelry, and a third eye. At the moment she takes her place on the throne and holds the sword of Taleju, the transformation is complete. From then until her abdication, she will be regarded as a kumari.

Based on his theory of agency-attribution, Gell interprets the kumari girl as the “living icon” of Taleju. According to Gell, there is little to distinguish the consecration of the kumari from that of any other idol. Purified by the priest, wrapped in the dress of Taleju and painted with the third eye on her forehead, the girl becomes the index that induces the people’s abduction of the existence of Taleju as her prototype (1998: 67–68, 150–53).

“Secondary agents”? Things, artifacts, and idolized persons

Gell’s analysis offers an example of the process through which idols manifest their agency. This occurs in a causal nexus of persons, things, and spirits/deities, and entails the reception of the agency of others. The actions of social agents in the nexus create real and practical effects (Gell 1998: 134–35), while at the same time they (re)create the nexus itself. Gell’s argument offers a useful model with which to analyze the effects of interactions among social agents, including nonhumans, in causal chains of agency. At the same time, it is apparent that Gell attaches great importance to human cognition and intentions, which after all are seen as constituting “primary agency.”

With this point in mind, let us reexamine the position of human cognition and intentions in the three cases just described. In Gell’s diagram showing the linked work of agency in volt sorcery, the sorcerer occupies a curious position compared with the other agents. Specifically, with the exception of the first phase, he always appears as a “voluntary agent.” Also, in the analysis on the Oro ritual, the power of Oro is considered to be controlled by the chiefs and priests via the indexes (Gell 1998: 114). Furthermore, in the case of kumari worship, Gell suggests that the existence of the goddess/prototype is “inferred” by people through the form of the kumari girl, even if the goddess appears in front of them as the living girl. Here, the girl is objectified or “idolized” as the living index of the goddess.

In these cases, Gell emphasizes the central role of human agency and cognition, which control and infer the prototype by manipulating the index in the causal nexus. At the same time, he treats idols, idolized persons, and other artifacts as “secondary” agents, which simply motivate the abductive inferences about prototypes, such as humans and deities, as intentional beings (1998: 20).

The aforementioned notions of actuality and coherence, in terms of the interactive movement of bodies and things enable us to consider the nexus of persons, things, and spirits/deities from a fresh viewpoint. In particular, this approach allows us to examine the power of things, not merely as secondary agents transmitting the power of primary agents, but as coactors with persons in their coherent and self-poietic movements. Additionally, this leads us to consider not only the active but also the passive aspect of the human actions performed with things, which cannot be reduced to cognitive operations, yet plays an essential role in the process of world-formation.8 Thus, it will be necessary to focus not only on the cognitive, but also on the poietic and generative aspects of self in relation to things. In the next section I examine the formation of the nexus of humans, things, and spirits/deities by focusing on two cases: the practice of making interethnic spirits in Eastern Ghana, and a spirit possession ritual in South India.

Making spirits, producing the self

In the late afternoon, in the back yard of the Tigare shrine, the ritual of making and installing of a new suman (spirit, pl. asuman) is being performed by nana Sakye (the Akan priest managing this shrine), the okyeame (the Akan “spokesperson”), several assistants and the Ewe priest from Togo. They put cowries in a half-cut calabash and kola nuts in a pot bound with red and black threads.9 Then, they bind red, white, and black cloth around a tube (perhaps from a bicycle) and paint it blue…. Past eight at night, they continue the ritual by candlelight. Though I cannot see well in the dark, they are binding the skull of a goat with red, black, and white cloth. Offering prayers to spirits, both nana Sakye and the Ewe priest bind various herbs with threads…. They pour libations and chant in Twi (a dialect of Akan language) for the spirits. After chanting a spell to the animal, nana Sakye sacrifices a gander by cutting its throat with a knife and pours its blood on the ritual objects…. The whole ritual seems to be the process of binding and spreading the thing-spirits, powers, and prayers. (August 20–21, 2007, at the Tigare shrine in Obretema village, Suhum District, the Eastern Region of Ghana)

The ritual described above was held in the shrine of a spirit, or suman, in a small village in the Eastern region of the Republic of Ghana. Though the area was originally inhabited by the Akan people, since the end of the nineteenth century the population has become highly multiethnic due to the development of cocoa plantations and the immigration of various ethnic groups, such as the Ewe and the Guan, as migrant farmers and farm workers. From the early to mid twentieth century, various suman shrines were established and functioned as “anti-witchcraft shrines,” where people were accused of witchcraft or exorcized. Subsequently, the main function of the shrines transformed from being associated with witchcraft, closely connected to kinship orders in Akan society, to the treatment against harmful magic related to ethnic conflict (Ishii 2005).

In these shrines, various magical objects are enshrined and worshipped by priests called okomfo. Most of these objects originate from faraway places such as the Volta and Northern regions of Ghana, or even from foreign countries such as Togo, Burkina Faso and cote d’Ivoire. Both the tangible magical object and the invisible spirit are called suman in Akuwapim-Twi, the mother tongue of the Akan people in this region. The relationship between the tangible and intangible forms of the suman is closely intertwined but ambiguous. While the tangible form is regarded by the priests as the temporary dwelling and medium of the intangible form, it is also considered to possess and manifest its own power. At the same time, the latter is said not to appear or dwell in a shrine without the former enshrined there. In this sense, the suman can be regarded as a thing which refuses the dichotomy between things and spirits, or between materiality and power (Henare et al. 2007: 6).

A priest at a suman shrine often performs rituals, including the offering of libations and blood sacrifices for the asuman in their tangible forms. Through these rituals, the priest vitalizes the power of the asuman to enable them to catch more witches (abayifo) or conquer the harmful magic that causes trouble for people. It is supposed by the priests that if the asuman succeed in these tasks, this success will improve the shrine’s reputation and consequently more people suffering from witchcraft or harmful magic will be drawn to the shrine. The rituals for the asuman-as-things are thus indispensable for the priests to activate their power, while they also contact the asuman in their intangible forms by being possessed by them.

The above case shows a part of the ritual, performed by the Akan and Ewe priests, of making the suman and animating its power. Here we find several characteristics similar to those pointed out by Gell in his analysis of volt sorcery and the Oro ritual. Through the operations of the priests the suman in its tangible form is placed in social relationships that link the priests, materials, and invisible spirits. Also, the power of the suman is activated by binding or smearing onto it the “exuviae” of life, such as the blood and bones of animals. In this view, the tangible suman can be seen as the index of the invisible spirit, whereas the priests are the voluntary agents who control the intangible by operating the tangible. The practice of making the suman, however, also includes a feature that cannot be fully explained by Gell’s analytical framework, since it focuses on the active and principal function of human agency.

The suman made in the ritual originated in Togo, the home of the Ewe people; thus, the Ewe priest took the initiative to instruct the Akan priests on the basic procedure of suman-making. For the Akan priests, however, the procedure of making the Togo-origin suman was uncertain and obscure. They partly followed the instructions of their Ewe counterpart, while partly applying their original methods to the production process. In a collaborative and communicative process more based on practical engagement than language, since they could not fully understand each other, they struggled with the materials that would become the suman. In this sense, the ritual practice turned into a series of contingent and experimental coactions between different priests and things.

Based on Kawamoto’s (2006) argument on bodily action and the practice of making artifacts, we can examine this point in more detail. As noted, Kawamoto focuses on the poietic aspect of the self, which is produced through embodied action. In this view, the self cannot be produced on its own. Rather, the mode and characteristics of the self and its bodily actions are specified through their coaction with things/environment. According to Kawamoto, when a person is going to make an artifact, she usually has a certain model or vision to be realized through her acts. Yet, a person may continue acting even in the absence of a model or vision of what she is going to make. Entering an unknown and unforeseen phase of the process, the human maker must simply proceed to act with the materials. She can pursue these acts only in a process of forming an artifact that is also a shaping of herself—and this is nothing but the process of self-poiesis. In this view, an artifact is not always made the object of one’s intentional acts. Rather, the continuation of one’s act entails the (re)formation of the self, which is, at the same time, the creation of the product (Kawamoto 2006: 190).

The practice of making an idol or magical object almost always entails the self-poietic process described above, even if there is a certain model or precedent for the act. In the case of the suman shrine, even if the Akan and Ewe priests had a completely shared vision of, and procedure for the suman-making, there still would have been a certain degree of uncertainty and unpredictability in the production process. This uncertainty is significantly related to the fact that the process entails encounters between several bodies and things. In this process, according to a certain vision or procedure, a person can partially envisage others’ actions and adjust his acts to them. Still, a person cannot fully predict or control others’ acts—or, indeed, his own—because those acts depend on these contingent interactions. Under a certain set of ritual rules or procedures, the contingency or uncertainty involved in the process may be delimited, yet it is always an integral part of the entire process. Because of the contingency entailed in encounters between several bodies and things, the making process cannot simply be determined as a chain of cause and effect; rather, its indeterminate and virtual feature is guaranteed and even amplified through the interaction. Thus, the suman is created not as the realization of a possible model, but rather as the contingent actualization of its virtual power through the momentum of the various actors’ encounters.10

In contrast with the general prerequisite for the production of industrial manufacturing, contingency, uncontrollability, and unpredictability are indispensable features in the making of the suman.11 To make the suman and animate its power is, in other words, to make it something beyond human control and prediction. Whether the procedure of suman-making is appropriate or not can be recognized by the people, not in the suman’s finished form, but rather in its effects subsequent to its enshrinement with other objects. In short, the suman’s ability or efficacy can be judged only by its work, for instance, capturing more witches, treating harmful magic, and as a result, bringing more clients to the shrine. Indeed, if the suman is powerful enough, it works as a hub of social relations linking various people across ethnic and regional boundaries, who all come to seek its power (Allman and Parker 2005).

We see then, that the power of the suman is manifest and recognized only through the effect of connecting various actors and expanding social relations. It was first made as a link between the priests, materials, and invisible spirits, and later came to include more people, places, and thing-spirits. And it bears emphasizing that this process of forming and expanding social relations cannot be fully controlled even by the priests. What demonstrates the power of the asuman-as-things that exceeds the power of human beings is thus nothing but its uncontrollable and unpredictable characteristics.12

These observations lead to an understanding of the making of idols that differs from the one proposed by Gell. For the practice of making an idol does not necessarily entail an act of controlling the prototype by manipulating the idol as its index, that is, acting as the voluntary agent or originator. Rather, it entails the experience of becoming the patient or “medium” of contingent coactions that relate one’s body to the bodies of others and to things.13 Furthermore, by making and worshipping the idol, a person becomes involved in the expanding social relations between persons and thing-spirits. Based on Kawamoto’s argument, we can thus consider the act of making an idol/suman through one’s bodily actions with things as instantiating a process of self-poiesis.

Passion and possession

Past eleven in the morning, after the long ritual process, the procession of the deities Balavandi and Pilichaamundi,14 patrons, musicians, and ritual workers carrying ritual paraphernalia are marching back to the shrine. Both sides of the street are overflowing with people wishing to have a look at the deities. Though the procession usually enters into the shrine premises from its north gate, this time their entrance is interrupted by two accidents on the way to the shrine. One is that a small brass bell falls to the ground from the hand of Balavandi (incarnated in the impersonator), and the other is that the long handle of borgode (a small umbrella made of silver) snaps in two. Balavandi is enraged by these inauspicious accidents, and for a while the procession comes to a complete halt. Several workers belonging to the Madivaara caste begin to shiver hard while clinging to the broken handle of the borgode. The man of the Safaliya caste holding the mask of Pilichaamundi also starts to shake violently. Finally, in a state of noise and frenzy due to possession, the procession enters into the shrine precinct. The Safaliya man, opening his eyes wide, shakes his stiffened body back and forth, and nearly fainting, staggers into the precinct supported by several workers. Balavandi follows the Safaliya and is about to fall on his back at the main entrance of the precinct when he is caught in the arms of the patrons. Finally, Pilichaamundi, wearing an ani (a big halo-like structure made of coconut-leaves) supported by six workers and accompanied by divine paraphernalia such as a sattige (parasol) and palanquins carried by the ritual workers, slowly enters the precinct. (March 13, 2009, at a buuta shrine in Perar village, Karnataka State of India)

The case described above is a part of the yearly ritual (neema) held at a buuta shrine in Perar village, located in Mangalore Taluk, Karnataka state, India. Buutas are deities and spirits worshiped widely in the coastal areas of Karnataka. They are generally considered deities, such as apotheosized local heroes or heroines who met tragic deaths and the spirits of wild animals such as tigers, serpents, and wild boars dwelling in the midst of deep forests, in dark marshy places, or on steep hilltops (Gowda 2005).

The buuta ritual mainly consists of spirit possession, oracles, and interactions between the devotees and the buutas incarnated in impersonators belonging to the Nalike, Parava, and Pambada castes (all designated scheduled castes). In the rituals held in the buuta shrine, these impersonators play various roles as buutas. They dance in spectacular costumes, sing oral epics, narrate oracles, and hold a court (vaakpiripuni) to judge problems among the devotees. In addition to these impersonators, priests belonging to the Bant and Billava castes are possessed by the deities. Among the devotees at the buuta shrine, the most important role is played by the patrons of the shrine. Most of them are the landlords of local manors called guttus who belong to the Bant caste. Praying for divine protection for their families and for the whole village, they offer local products such as paddy, areca nuts, and coconuts to the buutas. In return they receive blessings from the deities (Ishii 2010).

Similar to the suman shrine in Ghana, the buuta shrine has various ritual objects enshrined both as the dwellings of the invisible deities and as sacred, powerful objects themselves. Among other things they include divine statues, sacred masks, swords, parasols, and palanquins. Most are made of precious metals such as silver and gold, and they are usually kept inside a treasure storehouse (bandaarada kottya) and brought to the shrine only during the neema. Adding to these treasured ritual objects, other paraphernalia such as the garment of the deity to be worn by the impersonator, musical instruments, flowers, incense, and various prasadam (offerings) are also regarded as indispensable for summoning the deities into the bodies of impersonators.

The above event happened on the last day of the neema, when the Pambada impersonator and Bant priest possessed by the deities, along with the patrons, the ritual workers, and the devotees, were on their way to the shrine after the vaakpiripuni held at a sacred place called bantakamba. Let us examine this case in more detail, focusing on the relation between the people and the sacred objects used in the ritual.

As the example shows, the ritual objects strongly affect the people who come into direct contact with them during the ritual. For instance, the impersonator is possessed by the deity at the moment he touches the sacred sword or bell of the deity. Also, the ritual workers often fall into fits of violent convulsion when they touch the deity’s palanquin or parasols, and this occurs even if they are neither impersonators nor medium-priests but merely the carriers of these ritual objects. In the above case, the borgode exercised violent effects on its carriers when it suddenly broke on the way to the shrine.

As in the production of suman in the Ghanaian shrine, here we are also witness to people who cannot fully control or manipulate the ritual objects. Rather than exhibiting the superiority of human agency in manipulating sacred objects, the ethnographic vignette shows divine power overwhelming passionate human affect. At the same time, from the viewpoint of Kawamoto’s autopoiesis theory, the experience of a person affected by contact with ritual objects can be understood in terms of self-poiesis: the kinesthetic feeling of “being able to be possessed,” which indicates the active and at the same time passive state of the self as medium. Through the coactions of persons and things, each draws a certain mode of action from the others and at the same time each (re)creates itself.

However, in the case of spirit possession—contrary to our ordinary experience of a reality stabilizing through the regulation of the coactions of our bodies and things/environment—the coactions of bodies and things rarely converge into a stable state. Instead, these movements are experienced only temporarily as highly contingent and unstable events and they often suddenly break off due to the exhaustion or collapse of the possessed. The same instability may be observed in ritual practices such as hook hanging, walking over burning coals, or swinging in a swing with sharp nails on its seat (see Gell 1980: 219). In these cases of possession occurring through the contact between and coactions of bodies and ritual objects, it is crucial for the people involved that the acts be impossible in daily experience—not only dangerous but often intolerable, even for the mediums. It is precisely the transient and dangerous characteristics of the action that make the medium’s self-poietic experience—of “being able to be possessed”—most prominent and extraordinary; it cannot easily become a part of the given reality, but rather actualizes the virtual and vital link between persons and things-divinities. Thus, the contingent and unpredictable features of the actions are crucial for the people involved in spirit possession because they open a novel phase of experience.

Let us consider this from the viewpoint of the form and extemporaneousness of the performance in the buuta ritual. Spirit possession in a formal ritual is often highly stylized, and the buuta ritual is no exception. In the buuta ritual, the impersonators train themselves to perform in accordance with a sophisticated bodily schema. Above all, however, their skill is expressed in the ability to act in harmony with others’ actions. Similar to the suman priests who anticipate the actions of others in order to adjust their own acts to them, the performer dances, walks, speaks, and breathes in relation to the actions of others, which, aside from humans, include sacred objects and paraphernalia. These actions are inevitably impromptu to a degree, since they emerge through the encounters between several bodies and things. Thus, the contingency involved in the ritual performance makes the ritual not merely the realization or representation of one predictable/possible alternative reality, but the momentary actualization of the virtual state of the world.

This characterization opens a way to understanding the medium’s transformative experience as the actual sensation of becoming the deity itself. Concerning this point, let us reconsider Gell’s analysis of the installation ritual of the kumari. According to Gell, the kumari girl consecrated by the priest is considered as the index or “living icon” of the goddess. Gell analyzes her acts as a performance that induces people’s abductive inferences about the existence of Taleju, the prototype (Gell 1998: 150–53). However, if we consider this ritual not as the installation of an icon of Taleju but as a sort of spirit possession, an alternative understanding becomes possible.

According to Allen (1976), the girl becomes redder as the spirit of Taleju enters into her. As noted, the complete transformation of the girl into goddess occurs when she holds the sword of Taleju. Similar to the contact between the medium and ritual objects in the buuta ritual, the contact of the girl with sacred objects such as the goddess’s sword or bundle of pure plants brings about her bodily alteration, and this enables the girl to transform into the goddess. If we refer to Kimura’s argument on reality and actuality, we can consider the physical contact with sacred things as actualizing the “becoming-goddess” of the girl, even before she grasps and articulates the situation as reality. Also, in the ritual, it is supposed that not only the girl but also the participants such as the priests and devotees experience the physical alteration of the girl as an actual event occurring here and now. This actuality enables them to comprehend the reality of the girl being the goddess.

If we accept that the kumari is the goddess herself, then it follows that it is not adequate to interpret her simply as an index through which a divine prototype is inferred. The reality of the girl as the goddess and vice versa is not based primarily on people’s cognitive interpretations, but rather on the mutual interactions between the girl, things, and the participants. The acts of the girl and participants here can be examined in terms of performativity, rather than mere performance, indicating the poietic function of one’s acts in creating a reality and a self (cf. Butler 1993). In other words, the girl is not simply representing the role of the goddess by play-acting, but is incessantly becoming the goddess through her actions and other peoples’ reactions. In other words, through the ritual process in which the transformation of the girl into the goddess occurs as an actual event, the reality of the divine world, where the girl is “nothing but goddess,” emerges. At the same time, this indicates that this reality involves a sort of inherent instability or vulnerability and that it may easily fade away if not animated or recreated by the work of actuality.15

The case of suman-making revealed that the suman is a thing that cannot be fully controlled, even by its makers. In the same way, the encounters between people and sacred objects in the buuta ritual entail the contingency and unpredictability of their coactions. In both these cases, the uncontrollability or danger of the encounter and the coaction of the persons and sacred objects demonstrates the power of things beyond human intentionality. At the same time, it renders their encounters not as part of a reality already given, but rather, as the actualization of virtual and vital relations between persons, things, and the world. Thus, we see that divine worlds can only emerge through the momentum of such encounters, as creations of the coactions of persons and things through which they actualize their new powers and phases.


The aim of this paper has been to investigate how divine worlds can be created, vitalized, and lived by people. To pursue this investigation, I have considered the work and action of things not merely as implements or symbols, but as coactors with humans. To examine the relationship between humans and things, I have focused not on cognition and operating through things but on bodily action with things, coactions that provoke the emergence of a lived, changeable reality.

Through this analysis, I have also aimed to reconsider Gell’s argument on idolatry. Gell persuasively explains the process through which an artifact/idol interacts with persons as social agents in a certain causal nexus. Analyzing the chained agency of humans and idols, Gell presents a sophisticated theoretical model with which to understand the formation of divine worlds as the everyday social reality of people. At the same time, Gell assumes the primary function of human agency in the nexus, while regarding idols as only the secondary agents used by people to transmit or expand their agency in their efforts to recognize or control the world. Idols or divine objects are seen as mere instruments for humans, and the fact that their power far exceeds human intention is not sufficiently taken into account.

In contrast, this paper has investigated human-thing relations in the formation of divine worlds from the viewpoint of bodily coaction with things, occurring prior to cognitive articulations of events and generating novel experiences of, or postures toward, both oneself and the world. I have argued that the contingency and uncertainty entailed in the coaction of persons and things, which performatively creates a nexus among them, is essential for the formation of divine worlds. Through coherent and contingent action with things, a person becomes simultaneously an intentional agent and a patient or “medium” of others’ actions. In this process, the actuality of the vital and virtual relationship of the person with things is evoked as the momentary event emerging here and now. Divine worlds can emerge from such contingent coactions of persons and things as the ever-unsettled actualization of virtuality, and hence can be distinguished from the everyday world as given, articulated reality.

The contingency and instability entailed in encounters and coactions of persons and things is crucial to the emergence of divine worlds, not as the mere realization of a causal chain of agency, but as the one-time-only manifestation of virtual relations lived by persons and things. Through the emergence of divine worlds in ritual practices such as spirit possession, the making of idols, hook hanging, and sacrifices, peoples’ ordinary “settled” relations with things are disrupted, and in turn, the actuality of their virtual and vital relations are evoked. The emergence of divine worlds thus alienates one’s everyday perception of reality, and in doing so, enables a person to catch a glimpse of the fundamental instability or vulnerability hidden in the everyday world.


Allen, Michael R. 1976. “Kumari or ‘virgin’ worship in Kathmandu valley.” Contributions to Indian Sociology 10 (2): 293–316.

Allman, Jean, and John Parker. 2005. Tongnaab: The history of a West African god. Bloomington and Indianapolis, IN: Indiana University Press.

Babadzan, Alain. 1993. Les Dépouilles des dieux: Essai sur la religion tahitienne à l’époque de la découverte. Paris: Maison des Sciences de l’Homme.

Butler, Judith. 1993. Bodies that matter: On the discursive limits of “sex”. New York: Routledge.

Deleuze, Gilles. 1988. Bergsonism. Translated by Hugh Tomlinson and Barbara Habberjam. New York: Zone Books.

———. 1994. Difference and repetition. Translated by Paul Patton. New York: Columbia University Press.

Derrida, Jacques. 1982. Margins of philosophy. Translated by Alan Bass. Chicago: The University of Chicago Press.

Gad, Christopher, Casper B. Jensen, and Brit R. Winthereik. 2012. “Practical ontology: Worlds in STS and anthropology.” Paper presented at Osaka University, May 7, 2012.

Gell, Alfred. 1980. “The gods at play: Vertigo and possession in Muria religion,” Man N.S. 15 (2): 219–48.

———. 1998. Art and agency: An anthropological theory. Oxford: Oxford University Press.

———. 2006. The art of anthropology: Essays and diagrams. Edited by Erich Hirsch. Oxford: Berg.

Gowda, Chinnappa K. 2005. The mask and the message. Mangalagangothri: Madipu Prakashana.

Haraway, Donna J. 1991. Simians, cyborgs, and women: The reinvention of nature. New York: Routledge.

Henare, Amiria, Martin Holbraad, and Sari Wastell. 2007. “Introduction: Thinking through things.” In Thinking through things: Theorising artefacts ethnographically, edited by Amiria Henare, Martin Holbraad, and Sari Wastell, 1–31. New York: Routledge.

Ishii, Miho. 2005. “From wombs to farmland: The transformation of suman shrines in southern Ghana.” Journal of Religion in Africa 35 (3): 266–95.

———. 2010. “Sinrei tono koukan (Exchange with the spirits).” Bunkajinrruigaku 75 (1): 1–26.

Kawamoto, Hideo. 1995. Otopoiesis: daisansedai sistemu (Autopoiesis: The third generation system). Tokyo: Seidosha.

———. 2006. Sisutemu gensho-gaku: otopoiesis no daiyon-ryouiki (System-phenomenology: The fourth stage of autopoiesis). Tokyo: Sin-yosya.

Kimura, Bin. 1992. Ecrits de psychopathologie phenomenologique. Edited by Yves Pelicier. Translated by Joel Bouderlique. Paris: Presses Universitaires de France.

———. 2001a. “Cogito and I: A bio-logical approach.” Philosophy, Psychiatry & Psychology 8 (4): 331–36.

———. 2001b. Kimura Bin chosaku-syu (The collected works of Bin Kimura) 7. Tokyo: Koubun-do.

———. 2005. Kankei to shite no jiko (The self as the relation). Tokyo: Misuzushobo.

Latour, Bruno. 1993. We have never been modern. Translated by Catherine Porter. Cambridge: Harvard University Press.

———. 1999a. Pandora’s hope: Essays on the reality of science studies. Cambridge: Harvard University Press.

———. 1999b. “Factures/fractures: From the concept of network to the concept of attachment.” Res 36: 20–31.

Leach, James. 2007. “Differentiation and encompassment: A critique of Alfred Gell’s theory of the abduction of creativity.” In Thinking through things: Theorising artefacts ethnographically, edited by Amiria Henare, Martin Holbraad, and Sari Wastell, 167–88. New York: Routledge.

Maturana, Humberto R., and Varela, Francisco J. 1980. Autopoiesis and cognition: The realization of the living. Dordrecht: D. Reidel Publishing Company.

Phillips, James. 2001. “Kimura Bin on schizophrenia.” Philosophy, Psychiatry & Psychology 8 (4): 343–46.

Strathern, Marilyn. 1988. The gender of the gift: Problems with women and problems with society in Melanesia. Berkeley: University of California Press.

———. 1990. “Artefacts of history: Events and the interpretation of images.” In Culture and history in the Pacific, edited by Jukka Siikala, 25–44. Helsinki: The Finnish Anthropological Society.

Viveiros de Castro, Eduardo. 2012. Cosmological perspectivism in Amazonia and elsewhere. Four lectures given in the Department of Social Anthropology, Cambridge University, February-March 1998. Hau Masterclass Series Vol. 1. Available online: http://www.haujournal.org/index.php/masterclass/index.

Wagner, Roy. 2001. An anthropology of the subject: Holographic worldview in New Guinea and its meaning and significance for the world of anthropology. Berkeley: University of California Press.

Weizsäcker, Viktor von. 1950. Der Gestaltkreis: Theorie der Einheit von Wahrnehmen und Bewegen. Stuttgart: Georg Thieme Verlag.

Agir avec les choses : auto-poïésis, actualité, et contingence dans la formation des mondes divins

Résumé : Le but de cet article est d’étudier comment les mondes divins peuvent être créés, vivifiés, et vécus par les gens. Mettant l’accent non sur la cognition et les modes d’action à travers les choses, mais sur l’action corporelle avec les choses, cet article examine la réalité de ces actions qui précèdent l’articulation cognitive de l’événement et créent de nouvelles expériences du monde. La théorie d’Alfred Gell concernant l’idolâtrie est reconsidérée par le biais des apports de Ben Kimura et de Hideo Kawamoto. L’article explore ensuite la fabrication des esprits au Ghana, ainsi que les rituels de possession en Inde du Sud, et présente une nouvelle perspective sur la formation des mondes divins comme actualisation de relations virtuelles mais vitales entre personnes et choses qui émergent seulement de leurs coactions contingentes.

Miho Ishii is Associate Professor of Social Anthropology at Kyoto University. She has conducted fieldwork in Tanzania, Ghana, and India. Among her publications are “From wombs to farmland: The transformation of suman shrines in southern Ghana” (Journal of Religion in Africa, 2005) and “From passion to compassion: Healing rituals and gender in an independent church in southern Ghana” (Japanese Review of Cultural Anthropology, 2008).


1. Other important figures include Marilyn Strathern (1988, 1990), Bruno Latour (1993), Roy Wagner (2001), and Eduardo Viveiros de Castro (2012). Henare et al. (2007: 7–8) argue in favor of refraining anthropological questions in ontological, rather than epistemological terms.

2. A related criticism is made by Henare et al. (2007: 17–18) and Leach (2007). Yet, Gad, Jensen, and Winthereik argue that a similar tendency is still manifest in the volume edited by Henare et al. when they write: “Aside from its introductory manifesto, however, the volume…suggest[s] that thinking is still conceived to be primary, while things, if they are presented, are offered in the form of passive media, through which thinking unfolds” (2012: 9, italics original).

3. “Coherence” (Kohärenz) is one of the key concepts in Viktor von Weizsäcker’s thought (1950). The term indicates the chiasmic relation between an actor and its environment, and also its transformations: the characteristics of an environment are specified and transformed by a certain mode of action, and at the same time, the mode of the act is specified by the characteristics of the environment that emerges in the action (Kawamoto 2006: 82–83).

4. In this paper I will mainly deal with Gell’s argument in “The distributed person” in his Art and agency (1998).

5. Kimura describes these distinct but intertwined operations of the self as the distinction between “simultaneous reflection” (the noetic self) and “secondary reflection” (the noematic self). In the latter, the self is the noematic object of reflection; it is the image of myself after the act, the ‘I’ that has done something and can now be captured in reflection. While the former is immediate and fugitive in a manner that escapes articulation, it—at the same time—very readily flows into a secondary reflection that exists only as a mediated, articulated, noematic self. Kimura develops his notion of the noetic self of simultaneous reflection in terms of the Japanese concepts of jikaku (auto-perception) and koi-teki chokkan (action-intuition), taken from the Zen philosopher Nishida Kitaro (Phillips 2001: 343–344; Kimura 1992, 2001a, 2001b).

6. Kimura creatively applies Bergson/Deleuze’s ideas of the contrastive relationship of the real/possible and actual/virtual. Deleuze writes: “[T]he possible is that which is “realized” (or is not realized). Now the process of realization is subject to two essential rules, one of resemblance and another of limitation…. The virtual, on the other hand, does not have to be realized but rather actualized; and the rules of actualization are not those of resemblance and limitation, but those of difference or divergence and of creation” (Deleuze 1988: 96–97).

7. Autopoiesis theory is a new phase of system theory, first introduced into the field of theoretical biology by Maturana and Varela (1980). Kawamoto has contributed to developing the theory by linking it to German natural philosophy and Deleuzian concepts such as intensity and difference. The most novel point of this theory is its conceptualization of the self-poietic system and its borders: the self is considered to emerge through its movements, creating the borders of self and non-self. Hence, emergence is considered to be fundamental to the operation of the system (Kawamoto 1995: 7–11).

8. These aspects are closely related to the passivity of people in relation to things, which Gell briefly discusses regarding the relation between an artist and his artifacts (Gell 1998: 28–30; cf. 1980, 2006: 181).

9. “Nana” is the Akanese title for a person of honorable status. Tigare is one of the most popular spirits worshipped in this area.

10. Kimura (2001b: 307–308) explains the work of partial contingency and the actualization of virtuality through the example of a board game. Kimura argues that it is the virtual momentum actualized in each move of the players that makes the game not just the realization of possible moves but the actual experience lived by the players.

11. Of course the encounters among industrial or technological products and humans also entail contingency and unpredictability. The idea of the coherent actions of a person and things can be further developed in terms of the cyborg or cybernetic organism (Haraway 1991). Here one’s bodily self is always already acting with things as a cyborg, entailing the contingency emerging through the encounters of the body and artifact.

12. This corresponds to Strathern’s analysis of the encounters between Melanesians and Europeans. Strathern writes: “[T]he Melanesian would understand encounters in terms of their effects. It is the effect which is created, and effects (images) are produced through the presentation of artefacts…. They construct further artefacts…to see what the further effects will be. And the revelation will always come as a surprise” (Strathern 1990: 37).

13. This experience of a person as a medium can be considered as that of living in the middle voice. In his argument on differance, Derrida (1982: 9) expresses the defining feature of the middle voice as “an operation that cannot be conceived either as passion or as the action of a subject on an object, or on the basis of the categories of agent or patient.” See also Latour (1999a: 281–283, 1999b).

14. These are the names of the deities incarnated in an impersonator and priest respectively.

15. As I discuss at the end of this paper, this also can be the case, to a certain degree, in our everyday experience or awareness of reality and the self (Weizsäcker 1950: 172–173).