TTT Separating Containing 1

Chapter 6 SEPARATING AND CONTAINING PEOPLE AND THINGS IN MONGOLIA Rebecca Empson ____________________________________________________________________ The concept of ‘fortune’ (xishig) permeates many aspects of Mongolian social life. It motivates practices that involve separating (avax, salgax) a piece from a person, animal or thing at moments of departure or transition, and then containing (xadgalax) it in a different form. In this chapter, I use the concept of fortune to think about things that are either displayed on top of or concealed inside the chest kept in every Mongolian household. Thinking about the Mongolian concept of fortune through these things will allow us to question what kinship is, or looks like, in Mongolia. It reveals that the containment of a part, when separated from a whole, is essential for the maintenance of different kin relations. Furthermore, the invitation to ‘see’ kin relations through these things elicits the realisation of different relations at a single moment in time. When viewed together, they can be said to reveal a person made from each of the parts. Examining the means by which a thing is contained and the access people have to it, as well as its intended audience, will show how kinship is constituted through the separation of people and attention to things. In conclusion, I suggest that in the absence of people, the doing involved in making things visible or invisible makes relations. In this sense, ‘vision’ becomes the tool by which relations are created. Anthropological approaches to Mongolian kinship have tended to emphasise agnatic relations, based on the idea of clans that preserve the ‘shared bone’ of patrilineal ancestors (cf. Vreeland 1954, Levi-Strauss 1969). Moving away from these structural analyses, recent approaches to kinship in anthropology have emphasised a relational approach, whereby people are able to shift between different modes of engagement depending on context. Here, relations themselves exist prior to the position of the person with whom the relation is held and it is people who are able to move between them (Strathern 1988, Viveiros de Castro 2004). In such a way, Strathern (1994) has emphasised that: ‘a performance is always a reduction: a single act created out of composite relations’ (Strathern 1994:248). The aim of this paper is to explore how such an approach, which acknowledges shifting relational perspectives, might inform the way in which relations in Mongolia are created through the separation and containment of things. It does not require any great leap of the imagination to find that people and things are interchangeable in Mongolia in different ways. Objects in Mongolia often command the same respect as would be shown to human beings. This is not because these things ‘stand for’ people. Rather objects, like humans, contain another dimension of the visible world and something of the essence of the person is thought to adhere to their belongings. Chabros (1987) states that in Mongolia: ‘Analogies of form perceived between quite different and unconnected objects may result in the qualities of one object being attributed to the other’ (Chabros 1987:270). Chabros (1987) states: ‘It is difficult to consider the material culture of traditional Mongolia in isolation from non-material or spiritual aspects’ (1987:270). The tendency in Mongolia to perceive analogies, to see one thing as another, to discover relations between visible and hidden things, is the focus of this paper. Mongolians use the very general term ‘thing’ (yum) to refer to made or found things, but things do not always have to be material artefacts. Often, if the ‘thing’ has been made, the intention of the maker or the person who owns or uses it is held to be part of it. If the ‘thing’ has been found, or appears, the way in which it makes itself visible to someone is held to indicate the intention of the thing it stands for. Mongolians distinguish between visible things (xaragdax yum) and invisible things (xaragdaxgüi yum), so that things can also be events and matters. I suggest that things, placed on top of and inside the household chest, together act as a site that absorb aspects of people’s relations and draw attention to relations in the absence of people. These things can be viewed as vessels that remain in place and act as the ‘ideal kin group’ or person, as people necessarily move away from the house and shift different ways of reckoning relations. They allow for the continuation of certain relations that cannot be enacted in shared place. In all cases, the thing is viewed as a piece that has been separated in order to be contained, and is held to be a powerful essence or composite part of a person. Given this similarity, I will proceed to use the term ‘thing’ to talk about actual artefacts as well as pieces of people. It is suggested that the separation of people, or the ability to reflect on social relations through the containment of some part, is essential for the growing of kin relations in Mongolia. The concept of fortune It is important to stress that pieces, which are contained when something has been separated, are not simply visual substitutes, icons or proxies for absent people (cf. Weiner 1985). Instead, relations come into existence through the creation of these things. Mongolians hold that, in order for certain relations to continue, multiply and grow, people, animals, or things have to be separated so that a necessary aspect of them can be contained and a liveable version of a relationship becomes possible. This general concept is grounded in the cosmological idea of xishig or fortune. As an abstract quality, it is difficult to generalise about the principles of fortune. The concept can be understood on various levels, applying to many aspects of social life (cf. Holbraad, this volume). Primarily, xishig refers to the concept of a life-force or animating essence that can be understood through actions that involve attending to a part or portion that fuels a whole. Because it is mobile, the exact place or property of this animating essence is difficult to locate. The uncertain residence or property of fortune means that people take daily precautions so as not to lose it or let it slip away unnoticed to outsiders. For example, when a cow has been sold and is about to be separated from the herd, a woman will silently wipe the inside of her coat across the muzzle of the animal and detach a piece of its tail hair to keep safely contained in the house (cf. Montell 1934:109). Because the fortune that allows cattle to reproduce and prosper may be contained in just one cow, actions that involve keeping back a piece ensures that the animating life-force, essential to the whole herd, does not depart with that single animal. When something is separated, be it an animal, a family member, or some thing, precautions that involve keeping a piece back ensure that the essence, or fortune, is retained to support the whole. Practices that involve ‘separating-yet-containing’ the fortune of animals include: to make sure fortune does not leave when giving away a container of cream or milk, the giver pours the contents of their container into the recipient’s container which is placed on the ground. When the recipient’s container is full, the giver places their own container on the ground and pours back a little of the cream or milk into their own container so that the sacred portion (deej), containing the accumulative fortune of one’s animals, is retained. In response to my enquiries about the Mongolian idea of fortune during fieldwork with Buryat nomadic herders in Northeast Mongolia, people referred to practices, or described scenarios, that would illustrate an aspect of this concept. It seemed as if fortune could only really be understood if thought about through actions that involved attending to things. Frustrated at my inability to grasp the fluid nature of fortune, one friend finally presented the following scenario: Rebecca, if you want to understand what fortune means, imagine a brick building. If you take just one brick out, the whole building might fall down. You may try to find a single brick you can take so that the building still remains. Maybe your whole life you cannot find this brick, so instead you take precautions. You decide not to take a whole brick out but just scrape away a little at a time from different bricks. In this way you ensure you always keep a little back in order to maintain the whole: it may contain the fortune. In tracing the origin of the Mongolian word, Chabros (1992:155) suggests that xishig primarily refers to the idea of ‘portion’. The term is used in ‘The Secret History of the Mongols’ to refer to the selected bodyguards (xishigten) that protected Chinggis Xaan (cf. Lessing 1960:460). ‘The Secret History of the Mongols’ is an account of the ancestry and life of the Mongolian leader Chinggis Xaan, more commonly referred to in English as ‘Ghengis Khan’. Onon (2001) estimates that this account was probably recorded in 1228. Later, Chabros (1992) explains, the word came to be associated with wider notions of favour, good-fortune and benefit. It came to encompass broader meanings, including that of an individual’s share or portion of the vital energy that forms a lineage. Understanding fortune as an energy or vital portion, which animates the kin group and is passed on to each of its individual members, is fitting for my analysis. In this general sense, fortune is seen to increase the kin group’s life-potential as a kind of sacred essence or source, fulfilling a common desire to increase one’s domestic herds and have numerous children. Like Mauss’s (1972:133-149) famous description of Melanesian mana, fortune in the wider Mongolian sense can be used as an adjective, noun and verb. It is divisible yet whole, a force, being, action, quality and state. The idea of fortune can be viewed as a series of fluid notions. For example, while it is possible to exchange an object or animal that may contain fortune for something else, this exchange does not demand a return. In turn, while wealth may be the result of accumulative fortune, fortune does not generate profit in the sense of monetary value. But fortune does not simply dwell on its own. Whether it resides in a herd of animals or a single piece of cloth, it is attached in a relation to a person, or a group of people. Often irreducible to a single meaning, the Mongol term xishig is frequently used to refer to different things when paired with a second term. It can be a rare gift received from a highly esteemed person (buyan xishig), or from one’s animals (malyn xishig), without the giver expecting returns. Such pairing intensifies the good-fortune, vital-essence sense rather than modifying it. A story someone tells you or some crucial information accidentally overheard can be fortunate (amny xishig). A sweet given by a child after they have received blessing (myalaax) or an object that has brought about fortuitous events (xishigtei yum) can also contain fortune. A person or family can have it (xishigtei xün, ail) when their health and work go well and animals and food are plentiful. USING FORTUNE TO THINK ABOUT THINGS We have seen that Mongolians think about fortune in multiple ways. Fortune can change location, disperse, and suddenly emerge again in an interaction. In this section I suggest that the concept of separating in order to contain fortune can be transposed analogically to ideas about kinship, and specifically to the idea that the creation of kinship in Mongolia is achieved through the separation of bodies. Given this similarity, I shall use the idea of a (separated) portion that animates a whole, to think through practices that involve separating yet containing relations through things. This is of great importance for Mongolians because the separation and assemblage of kin members at different seasonal places means that maintaining a connection with a person, place, or thing is essential. Because movement is an inherent aspect of Mongolian kinship, I will show that relations based on affinity, which involve the separation and incorporation of difference, are the necessary, yet invisible, background that support the visibly fore-grounded relations based on consanguinity, containment and sameness. The things that I focus on can be found inside most Mongolian households. I shall argue that the ways in which they are displayed parallels the concealment or display of different ways of reckoning kinship. When viewed together, these things could be seen to provide an ensemble of multiple ways of reckoning kinship in Mongolia. They allow the viewer to apprehend him or herself as an exemplary person made from several severed parts. Drawing on Gell’s (1998) idea that material objects can act as indexes that appear to abduct agency, things kept inside and on top of the chest come to stand for, and act as, instruments of social agency and relatedness. While these things do index certain current relations, viewing or attending to them also initiates the possibility of new relations. These things can thus be viewed as tools or instruments that index certain ideas that produce an effect, such as a motivation or interpretation, on behalf of the recipient who views or uses the thing. For example, we will see that photographic montages are a way for agnatic kin groups to reveal their infinite networks to outsiders. I should make it clear that while I do draw on Gell’s (1998) wider point that objects have an effect (in this paper, kinship is the effect), I do not use his extensive abduction thesis to explain how things appear to have agency. However, it is important to note that some of the things discussed in this chapter do have a kind of agency in themselves. For example, when a person leaves their natal home, pieces that are left behind ensure that they are able to continue these relations, but for the person who remains and attends to these pieces, the objects have a kind of agency in that their containment is seen to increase the life-potential of the kin group. In this sense, the thing can have an effect or a type of agency depending on whose perspective we take. The things which I look at can also be viewed as distributed extensions of a relation or parts of a person. This is not a mystical idea of the person. It is imperative for Mongolians, who are nomadic herders, that people are able to manifest themselves, via things, in different spatio-temporal locations, beyond the confines of a single bodily form. In this way, people are not just where their bodies are, but in many different places simultaneously (cf. Strathern 1994, Gell 1998:21). By exploring the kind of technology that reveals or conceals relations, we will see that objects do not just commemorate past relations, they also initiate relations that make up current kin. UNDERSTANDING KINSHIP THROUGH THINGS Agnatic kinship, based on the idea of shared bone (etsgiin töröl, yasan töröl), permeates much of Mongolian life. Virilocal residence is expected, and property, mostly in the form of animals, is traditionally distributed among sons. The youngest son of a family inherits the bulk of his father’s livestock, including the family hearth that is held to represent the continuity of patrilineal descent. Nevertheless, throughout the year, wooden houses (baishin) as well as Mongolian felt tents (ger), and the people who inhabit them, move over the landscape. As the physical shell of a house reconstitutes itself in different places, the people who inhabit a house also change seasonally. Throughout the year, people move from a house to different locations with different networks of people, in order to attend school, work, hunt, marry, or trade. While summer encampments gather together extended relatives in order to help with the preparation of milk products, winter encampments are often only inhabited by a few family members. During the autumn and spring, children attend school in district centres and younger family members may move away from the household to engage in temporary work, trade, or hunting. Although virilocal residence is expected, and agnatic kinship is held to dominate kin relations in Mongolia, throughout the year different forms of sociality are enacted in different places. People have to move to different locations, activating other types of relations, while still being tied to their agnatic household. This section will focus on the way in which the movement of people and place is managed through different things that are kept in the household chest. Both wooden houses and Mongolian felt tents are prevalent in the countryside, and throughout the year people interchange kinds of residence. For example, a family may live in a wooden house at their summer pasture, but occupy a felt tent at their winter, spring and autumn pasture. When fixed in a particular space, the house (be it a wooden house or a felt tent) is made a container for storing valued possessions, meeting with visitors, sleeping and eating, and for moments when one needs to sit for a long period to fix or sew something. As a one-roomed, open-plan space, there are no personal areas inside a Mongolian house. Instead gender, hierarchy and status define the interior. This adaptability allows for the incorporation of different configurations of kin members, as well as outsiders, at any given moment. For example, an elderly female guest will know in exactly which part of the house to sit as she enters an unknown person’s house. Mongolian kin terms also allow for the incorporation of outsiders. While most terms between kin are fixed and categorise people in terms of hierarchy and gender, some terms are both classificatory and un-gendered. Specific kin terms draw attention to agnatic and non-agnatic kin by differentiating relatives on the father’s side (avga ax / egch), or mother’s side (nagats ax / egch). Male heads of two families, related through the marriage of their children, have classificatory terms (xud), as do female heads of two families related through the marriage of their children (xudgui). Grandparents distinguish their grandchildren as coming from either their daughter (zee), or son (ach, also meaning favour, grace and benefit). See Vreeland (1954), Pao (1964) and Park (1997) for extensive information on kinship terminology in Mongolia. For example the term ‘younger sibling’ (düü), used for people younger than the speaker, does not specify gender or type of relation. Terms such these are not ambiguous for Mongols. It is only because their range of meaning is different from anything we are familiar with in English that they may, at first, seem ambiguous to us. In addition to fixed kinship terms, Mongols need these shifting and flexible terms in order to be able to incorporate people who come in and out of the house whom they sometimes want to treat as kin. In turn, while away from relatives, people are able to establish sibling-like relations with others through the use of such terms. The seasonal movement of people, and the places they inhabit, creates the continual need to relocate both physical and relational boundaries. This can give rise to the feeling that there is no fixed place in which to situate people when trying to define kin relations. For Mongolians, however, the separation and incorporation of people and place is not unsettling or difficult. See Humphrey (1995): ‘The Mongols do not take over any terrain in the vicinity and transform it into something that is their own. Instead, they move within a space and environment where some kind of pastoral life is possible and “in-habit” it’ (1995:135). They overcome this movement by ensuring that certain things remain contained inside the house as people and houses move location. These things act as sites for containing particular aspects of people’s relations in the absence of people. Instead of people constituting a home, in Mongolia, valued things inside the house remain in place and stand for relations that are attached to it. This idea is also extended to the landscape surrounding the house which is marked with stone cairns, sacred trees, buried placentas and tethering posts that invoke a sense of inhabited space in the absence of houses and people. To avoid an ahistorical account, before I begin to explain how people maintain relations through the containment of things, the genealogy of revealing or concealing relations in the house should be placed in brief historical perspective. During the socialist period, Buddhist icons and shamanic implements were prohibited from being placed on view, but statues of Lenin and posters depicting, for example, strong industrious cooperative workers or joyful rosy-cheeked pioneers were openly displayed. Genealogical diagrams (ugiin bichig), going back over seven or eight generations through agnatic lines, were either burnt or hidden in the bottom of chests. Differences, especially of an ethnic or class kind, were thought of as politically polluting and people were forced to use their father’s name as a surname instead of their clan name (cf. Humphrey 1974:477). The Buryats, an ethnic Mongolian group on the northern Mongolian-Russian border with whom I did my fieldwork, migrated to Mongolia from Russian Buryatia in the early 1900s. They experienced extreme forms of political persecution during the socialist period, to such an extent that during mass purges in the 1930s, almost all the male members of the community were killed or taken away. Turning people away from their kin networks, and the distinctions that these create, has concealed the diverse ways in which Mongols actually reckon kin relations and conceive of the person. As a general introduction to ideas about kinship in Mongolia, Mongolian procreation beliefs hold that an infant has to be separated from the spirit world, which makes it vulnerable as a human. After this vulnerable stage, the shared substance of ‘bone’ from the father is emphasised. But while agnatic kinship may be the foundation for many kin relations in Mongolia, these relations coexist with other ways of conceiving kin. Ideas about lay-reincarnations (irgej töröx, daxin töröx), blood relations (ekhiin töröl, tsusan töröl) based on movement, links with one’s birthplace and the deceased, as well as age-sets also form lasting kin or ‘kin-like’ relations. This multiplicity allows for people to be other things, while at the same time being defined by their bones and ethnic identity. The Mongolian concepts of ‘blood’ and ‘bone’ should not be confused with Western essentialist concepts of these terms (a la Gil-White 2001). Through the concept of fortune, this paper proposes that although much Mongolian kinship is ‘given at birth’ through the inheritance of substance, Mongolian kinship is also made in practice throughout life. While classical anthropological accounts (cf. Vreeland 1954, Levi-Strauss 1969) and Mongols alike may hint, through the use of specific terms, at ideas that seem essentialist, in practice we see a multitude of ways of being related. A focus on things kept inside the family chest allows us to explore the ways in which agnatic kinship coexists, and is indeed dependent on, these other modes of kinship, through the way in which these relations are contained in different things that are deliberately displayed or concealed from view. To paraphrase Strathern; these ‘deliberate provocations to vision’ become a way for Mongols to instantiate different networks of relations (cf. Strathern 1994:243). The things on which I focus pivot around two distinct ways of reckoning relatedness. Firstly, the relations that are visible on the chest’s surface, in the distribution of family property, through communal ritual, and in formalised language are based on the idea of shared ‘bone’ from one’s agnatic forefathers. Such relations exemplify an ideology of patrilineal descent and the continuation of relations over generations. They are, however, dependent on a second mode of relatedness, involving the separation and incorporation of people. For such groups to exist, people have to move, establish links with other groups, and incorporate non-kin outsiders. These mobile relations are based on the idea of ‘shared blood’ from one’s mother. This provides an ‘umbilical relation or communication’ (xüin xolboo, tsusan xolboo) that is passed between a woman and her children, and between siblings. Blood relations are given anew each time a person is born and are drawn upon at different periods in a person’s life as people are necessarily separated from each other in different locations. Such relations are not made visible in particular sites. Rather, they are hidden in parts that are kept inside the household chest. I turn now to examine how these relations are maintained, through an analysis of things kept in and around the Mongolian household chest. Analysis of the household chest Relations that are visibly displayed In the northern, rear part of the Mongolian house, in the most honourable section (xoimor), opposite the door as one enters, stands a painted wooden chest (avdar). The chest may be covered in embroideries or painted with inter-locking patterns (xee). The things that I am going to discuss can all be found around or inside this large chest. On the chest’s surface, visible prized possessions that indicate wealth and prestige are displayed. These include objects such as radios, clocks, batteries, perfume and so on. In the centre stands a large mirror. Surrounding this mirror on either side, or attached to the wall above, are two large frames containing a montage of three-quarter-length, portrait-style photographs (jaaztai zurag) of kin members on both the mother’s and father’s side. This montage creates a pile, or layering, of different images over time, as old photographs are concealed behind new ones. It is not just individual portrait-style photographs that are displayed in such frames, groups of people such a family member’s class mates, a pair of brothers preparing to leave for hunting, or a family visit to a historic site, all provide occasions in which the moment can be captured in a photograph and displayed along with more formal portraits. Above the mirror, religious icons and images can be found that comprise a small shrine (Burxan) on which religious books (sudar), pictures of consecrated animals (seterlesen mal) and daily offerings of milk libations (Burxandaa idee tavix) are placed. Above this shrine, on the wall behind the chest, hang large painted portraits of deceased patrilineal relatives (jaaztai taliigaachiin xörög), shrouded in ceremonial silk scarves (xadag). These portraits occupy a high position, comparable to the sacred Buddhist images. They emphasise agnatic dominance in the household (cf. Sneath 2000:224). Around this fixed display, embroideries (xatgamal), sewn by daughters-in-law, are hung, depicting their views on different family relations. Guns, used by men for hunting, are placed at its side. Things kept inside the chest are never revealed to guests and are concealed from general view. Young daughters-in-law, separated from their natal homes, and elderly female household members are in charge of maintaining this very visual display. They feed it with offerings and attend to and change its form. In turn, visitors to a household are expected to respond to it. As one enters a house, after greeting the host, one is expected to go to the chest and, while bowing down towards it, knock one’s head (mörgöx) against its surface three times and turn a prayer wheel or offer some money or sweets to the religious icons, or to a portrait of the host’s deceased relative. In so doing, a visitor pays respect to their host by honouring the fact that they are a part of a wider network of people who respect their elders and the ancestral spirits of the landscape. In addition, because the mirror is at the centre, when attending to or viewing this display a person may catch a glimpse of themselves at the centre of these different imaginings of kinship. The display allows the viewer to respect their host while at the same time to imagine themselves as placed, albeit fleetingly, within this web of relations as a potential part of the network that they are honouring. What kind of relations does the viewer and attendee honour through responding to these things? I suggest that photographic montages of living kin members, displayed in frames above the household chest, can be viewed as a modern take on the traditional Mongolian practice of recording genealogies. While Buryats in Mongolia are interested in maintaining genealogical records, their genealogical diagrams were often burnt during political purges in the 1930s. During the socialist prohibition of recording genealogies, Mongolians embraced the medium of displaying photographs to represent their kin. Unlike anthropological kinship diagrams, Buryat genealogical representations do not define age groups in hierarchy from the top to the bottom of the page, over generations. They depict kin relations in the form of a cluster of male descendents expanding outwards from a single founder in the middle or top half of the page. Given the increasing interest in tracing one’s genealogical background, do these photographic montages provide a new technology for recording genealogical relations based on agnatic kinship? Increasing interest in recording genealogies is, in part, due to the fact that people have to record their clan name on identity cards. For information concerning variations in eighteenth and nineteenth century Buryat genealogies in Russian Buryatia, see Humphrey (1979). Photographic montages mirror some of the compositional forms used to represent relations in traditional Buryat genealogies. The photograph of a patrilineal elder is often placed, with his wife, in the centre of the frame. They are surrounded by their siblings and children, whose images extend outwards towards the periphery of the frame. Such composition mirrors Buryat genealogical diagrams and represents a centric view of kin relations that expands outwards from a patrilineal founder. This circular representation can be found in many other Mongol forms, such as circular offerings at stone cairns (oboo) and to the family hearth (gal golomt). Deceased patrilineal kin members’ portraits are hung above the photographic displays. The photographs can be seen to mimic the style of these portraits and thus enforce the dominant patrilineal ideology that they imply. On closer inspection, however, we see that what links people together in photographic displays differs from the agnatic links that join people together in genealogical diagrams. Photographic montages of kin members reckon relations through both the mother’s and father’s side. They also include photographs of school friends, people from one’s summer pasture, work colleagues, and groups of people at other special events. Information such as ethnic background, class and status can be ‘read’ by viewing the locations in which the images have been taken, the clothes people are wearing, and the type of events celebrated. Viewing the display, visitors are able to infer their host’s relation to other kin members, as well as friends and colleagues. The networks of relations depicted in these montages are an expansion of those found in traditional genealogical descriptions. Alongside relations based on the shared blood and bone from one’s mother and father, other relations are also emphasised. Photographic montages thus replicate Buryat genealogical diagrams in their form, but extend their content to include other types of relations. Bouquet (2001) makes a similar point in relation to photographic displays: ‘ … photographic reproduction [is revealed] as a powerful means of establishing and cutting genealogical relationship[s].’ (2001:110). See also, Bouquet (1996) for the limits of analysing kinship through anthropological kinship diagrams. Like genealogical diagrams, the montages also construct a ‘portrait-chronicle’ (cf. Sontag [1971] 2002) of previous connections, and remind one of kin members who are absent. In so doing, they depart from a single person’s perspective and provide a memory-map in which past and present relations are imagined to exist at once. In Mongolia, having one’s photograph taken involves posing front-on for the camera. To this extent all photographs are reproductions of each other; what makes an image ‘good’ to look at is that it looks at you. Instead of freezing individual characteristics or gestures, we see a replicated pose of motionless groups of people looking at us (cf. Bouquet 2001). As a person is able to quickly glean information from a genealogical record, so too is the viewer able to abstract information about crucial networks and relations from the montage. This view of several groups of people, looking out together from the display, dazzles the viewer with the multiple relations available to the people of that household. In this sense, the groups of people in photographic montages ‘[…] reach out to the consideration of others’ (Humphrey 2002:69, italics in original). Humphrey (2002) has drawn attention to the two-fold way in which, when displayed by their owners, personal possessions reach out to the consideration of others: while they may be displayed to signal social status or vanity, they equally stand for the need for acknowledgement or recognition by other living people (cf. Humphrey 2002:69). By drawing attention to the infinite networks available, the montage confuses the viewer as to who is and who is not kin. For a similar idea, see Gell’s (1998) discussion of Trobriand prow-boards that dazzle exchange partners to surrender their valuables. Viewing the photographic montage in this way, people are not depicted as mobile individuals. Rather people become replicable members of static groups, with potentially infinite links to other groups. Photograph montages, portraits of deceased elders, and shrines that honour the spirits of inhabited places (baigaliin lus savdag) outwardly display relations, with infinite connections, given through agnatic relations in a visible form. Such fore-grounded relations may subside at different periods in a person’s life but they can always re-emerge and be drawn upon again. These images are not about person-to-person relations, but replicate relations between groups. For example, if a man decides to spend the spring building wooden houses and needs help tending his herds, he may turn to his brother who will assist him without expecting anything in return. For the person who attends and adds to the display, there is a sense that, although they move to different seasonal places, the chest’s visible surface remains as a fixed site inside the house that increases over time. We can draw a parallel here with points in the landscape such as stone cairns (oboos). People make piles of offerings to these throughout the year and during communal ceremonies. The comparison is not just analytical: small pieces from these offerings (usually pieces of rice) are retained and stored in a bag (dallagany uut) at the family chest. These places anchor meetings between people in a fixed visible form. This technology is meant to be seen. Its efficacy, as I will show, acts as a shield against those relations that are concealed. Concealed relations in hidden things Things displayed on top of the chest, which are visible as soon as one enters the household, provide a site for preserving agnatic relations and extend these to include relations between wider groups of people. These connections are inherited; one is never fully separated from them. They increase or decrease according to where one chooses to locate oneself. Inside the chest, concealed from general view, are hidden things that have been detached from people at moments of separation and transformation. Such things are individually wrapped in blue ceremonial silk scarves and are carefully placed at the bottom of the chest among winter or summer clothing, as if they have the potential to move but must be contained. The things hidden in the bottom of chests that I focus on comprise of actual parts of people’s bodies, such as pieces of umbilical cords and children’s hair from the first hair-cutting ceremony. They are very rarely handled or exposed for view. I turn now to the way in which these things are produced and concealed in relation to ideas about ‘shared blood’. Like relations based on the idea of ‘shared blood’, things found in the bottom of chests are not passed on over generations. They are the products of the separation and movement of people between groups and are created out of alliance and exchange in one’s own lifetime. The concept of ‘sharing the same blood’ is linked to the movement of women across agnatic kin groups. For further information concerning ideas of separation from people and place, see Stafford (2003), and for information concerning the movement of women in Chinese patrilineal kin groups see Stafford (2000:110-126). Such relations are not contained in any particular visible location or site, such as a stone cairn. Instead, they are realised through a special type of relational communication (xüin xolboo, tsusan xolboo, literally: umbilical communication, blood communication) that allows people to have anticipations, feelings and premonitions about each other (sovin tatax, literally: premonition is pulling at me), even though they may be separated in terms of place. For example, if a person falls ill or suffers some accident while away from home, a mother or sibling will begin to feel pain in their body, have bad dreams and sense that the absent person is in danger. As children grow and leave the home, relations between a mother and her children, and between siblings do not cease. Equally, it is imperative that people who have this type of relation are physically separated from each other. This is because relations based on ‘shared blood’ or ‘umbilical relation / communication’ are considered to be too close to live with. For example, it was suggested to me that when a child is first born, the mother and child merge into and become one another so that the mother’s body becomes child-like (eejiin bie nyalxardag, literally: a mother’s body becomes wet and like an infant) expectantly craving sweets and gifts from visitors and strangers (goridox). The ritual cutting of the child’s first hair creates a necessary distance between a mother and her child, separating their shared physicality. Similarly, twins are considered to have an intense form of umbilical relation. If one twin falls ill or suffers, the other will do so too. In order to lessen the effects of such a relation, twins go through a ritual whereby a piece of red rope, tied between their wrists, is severed. We can begin to see here that it is held to be necessary to detach oneself from the physical intensity of umbilical relations based on shared blood. By creating a physical distance and by giving a part of oneself away, a liveable version of the relation is formed. It is because of this that, when people are physically separated from each other, a part is produced during the act of separation. This part is carefully retained in order to maintain the relation in a separated form. Containing a part of a person’s body at their natal home can be analytically compared to the concept of fortune, and the practice of separating yet containing, in three distinct ways. Firstly, by containing a piece of hair from the child’s first haircutting ceremony, parents separate off and contain a part of their child at a point at which they are introduced to agnatic kin. The haircutting ceremony is seen to mark the point at which a young child, having fully rejected the temptation to maintain contact with difficult un-reincarnated spirits, is secured in the human world. Keeping a part of this hair, which has been severed from the child at a point at which departure from relations with others and acceptance in the kin group is celebrated, children are formally bound to ‘relations of bone’. The hair-cutting ceremony marks the child’s entrance into social life (see Humphrey 1974:479). Secondly, in retaining a piece of the umbilical cord at their natal home, relations between a woman, her children and between siblings, all of whom may later disperse, are maintained. It allows women, who move between groups and never fully belong to their groom’s or father’s agnatic kin, to maintain a partial connection with their natal home. This is exemplified in the term used for a bride as a ‘person with an umbilical cord and an engagement’ (xüitei-süitei xün), indicating that she is someone who is about to move to her groom’s family, but still has continued relations with her natal home. Although people have to be physically separated, due to the intense aspect of these relations, people do draw upon them at different times throughout their life. The umbilical cord is held to stand for a child’s life-power or soul (süns), and the loss of their first hair at the hair-cutting ceremony represents the incorporation of a child into the world of people (cf. Galdanova 1992). See also Humphrey (1973): ‘Buriat women preserve the umbilical cords of the children, since it is thought that the cord is somehow a “line of life” whose magic power would be broken if the cord were to be thrown away. [It is] associated by the Buriat with female descent and thought of as complementary to patrilineal essence.’ (Humphrey 1973:22). For example, a married female friend of mine, after having experienced a series of miscarriages attributed to spirit and human curses, drew on her natal family fortune by making offerings to their sacred tree. When the situation became worse due to intra-household feuds, she was able to return, albeit for a brief period, to her natal home. The umbilical cord is both an actual part of a person achieved through separation and an expression of a relation that, through separation, can reappear: its containment is achieved through departure, but it remains in order to allow for the possibility of return. Finally, by retaining these pieces in her married house, a woman is able to solidify relations with her children even though she lives with her husband’s family and is regarded as belonging to another group. It allows her children, who ultimately belong to her husband’s kin group to maintain a lasting tie of relatedness with their mother, their siblings and their place of birth, regardless of where they happen to be. Women often use these pieces as magical remedies for their children and as an aid for fertility. They both protect a person in a situation of crisis, and facilitate the possibility of future kin. By carefully hoarding a piece of the umbilical cord or pieces of the child’s first hair in the family chest, the mother-child relationship (ekh üriin xolboo) is maintained as a possible relation, regardless of people’s physical location. Through focus on objects in and around the household chest, we see that certain aspects of people’s relations are safely contained and hidden from view, precisely because they are created when that part of a person, determined by gender and birth-order, separates, moves, and changes. See Chabros (1987) who explains that certain things within the family containing magical properties are regarded with the same respect as people. It is equally important not to keep things which have been owned by outsiders. Seen to contain a part of that person, they have the potential to cause pollution. In this way, the contained and hidden parts become the visible manifestation of relations that are concealed from general view and are not displayed openly in daily life or through communal rituals. The concept of revealing and concealing different aspects of people’s relations can, of course, be extended outwards. A case in point would be restrictions placed on the daughter-in-law in terms of language. Uttering personal names in Mongolia draws attention to the hierarchical relationship between the speaker and the person the name refers to. Daughters-in-law are tabooed from uttering the names of senior male affines and must find suitable substitute words to refer to them, even in their absence. For further information concerning name taboos and the suppression of attention, see Humphrey (1993). The practice of separating off a part at moments of transformation is also present at Mongolian rituals of death. Here, the giving away of pieces severs relations with the living in an acceptable. When a person is about to be separated from their body, the dying person gives away their belongings in order to break attachment to people and things and free the soul (cf. Humphrey 2002). Humphrey (2002) states: ‘It is believed that the spirit or “soul” (süns), even after death, remains emotionally attached to one particular object which was much used in a person’s lifetime’ (2002:67). Death can be seen as an extreme form of separation that does not allow for the possibility to return. While people create kinship through the accumulation of things, at death these things have to be disposed of, thereby cutting off further relations. It should be noted that deceased people’s things are not kept in the household chest. Death is, thus, an extreme kind of separation that Mongolians have to manage in relation to ideas about the separation or containment of things. But it is not just at death that ‘… detachment from a person [is] achieved by giving something [of oneself] to them’ (Humphrey 2002:71). We have seen that living people also practice this form of detachment. Movement of people away from the house, in marriage for example, demands that people are separated on the condition that a part, or thing, is kept back. THE VALUE OF SEPARATION Parts, hidden inside the chest, provide a vessel for what are otherwise location-less connections. These vessels are the outcome of relations that are volatile and uncertain, but they anchor people in relations that transcend a person’s physical location and form. Traversing relations that are located in visible sites according to agnatic groups, invisible or hidden things, achieved through separation, allow for people to cross boundaries of agnatic relations. In this way, the parts do not simply contain or stand for relations but actively create and facilitate them. The point is not that these are somehow illicit connections. They are not illicit, but the relation indexed in the thing remains the same as at the point at which it was separated. Instead relations based on separation and departure, are necessary for sustaining the visibly enacted relations based on agnatic kinship. In turn, it is important to note that things contained in the bottom of chests are not about ‘possession’ in the Western sense. Instead, liveable relations come into existence through the creation of these things. Although they are highly valued, they are intrinsically tied to their original producer and cannot be used in exchange for something else (cf. Weiner 1992). They become material parts through movement and, in so doing, create a lasting connection to the person they were once attached to. The thing is, thus, never a full replication or replacement of the person, but a part that is necessarily different from its original form. The idea that things originating from one source are similar, but not the same as each other, could be applied to naming practices in Mongolia. In is not uncommon that children of one household share a name that has a similar part but a different ending, i.e.: Bibish (Not-me) and Terbish (Not-them), or Batchimeg (Bold-decoration) and Battsetseg (Bold-flower), etc. Here, a part of the name is shared, but it is modified each time to be slightly different. They also suggest a difficult connection with a relation that one must be separated from. Instead of viewing these things in terms of a resource, value is given to the exchange in the perspective that they allow. When viewing kin relations through these things, we make what is considered the periphery the centre. Through these parts of persons, difficult relations are maintained as open possibilities. By keeping and retaining a piece, an anticipated return to a possible version of the relationship is created (cf. Weiner 1985:221). Although these things could be viewed as icons, in that they are held to contain some part of a person in their absence, they are not passed on over generations. Instead, we have seen that they preserve the possibility of current relations between living people. Because they cannot be exchanged or substituted for another form, they could be viewed as ‘hyper-personal objects’ (Humphrey 2002). They are comparable to what Humphrey (2002) has termed ‘a refuge thing’ (xorgodson yum) (Humphrey 2002:67), in that they have a hyper-identification with the person they belong to (or have been detached from). The spirit or soul (süld, süns) of a person is believed to be attached to such an object and they have to be looked after with care. When people move away from the household, such a piece (either an umbilical cord or a piece of hair) is retained at the moment of transition, and kept back by the people who remain. The ‘hyper-personal’ aspect of these things means that they have to be detached and separated from the person they are indexing. Once separated, they are carefully cared for by someone else, in part, because they provide vessels that accumulate fortune for the whole family. For example, several umbilical cords in a house are held to attract fertility and the possibility of more children. Instead of merging relations between people into groups and then making them visibly static, as with photographic montages, these things separate bodies and maintain a link with the person they were detached from. They can be seen to act as channels for living person-to-person relations. Having used the concept of fortune to think through things in and around the household chest, certain ideas about Mongolian kinship emerge from our analysis. It has been suggested that relations based on agnatic networks are visibly fore-grounded as immobile centres from which people reach out to different connections with other groups. These group relations are, however, dependent on the separation and incorporation of others. Relations from which one must be separated, in order to support the possible growth of this centre, are concealed from general view. For example, inside the chest a piece of hair is kept that has been detached from a child, ensuring that it transforms from semi-human outsider to agnatic kin member. These hidden pieces intercept and move away from the visible group relations. Through this movement these transformations also support the possible growth of the centre (cf. Empson 2003). By maintaining a part, they facilitate the possibility for relations to continue in a liveable form. Using the idea of ‘separating in order to contain’, the analysis can be extended further to focus on invisible things that emerge and make themselves known in people. Lay-reincarnations, which are common among Buryats in Mongolia, scramble any linear idea of shared substance and bring deceased relatives from either the father’s or mother’s side into the kin group. It is difficult to determine if all children are held to be lay-reincarnations. Sometimes, parents hold that their child is a reincarnation of someone, but do not voice this opinion to others. In turn, parents may fight over who the child actually is a reincarnation of. When they are older, children may inquire about who they were held to have been a reincarnation of and the answer to this is often embedded in who the parent longed to have been reborn in their child. The ethical dilemmas faced by parents when children want to find out about their reincarnated past can be compared to the ethical problems faced in Euro-American kinship thinking, when children want to know about their birth parents’ genetic history (see Strathern 2001). While relations with the deceased are not contained in visible artefacts, they do continue through lay-reincarnations. Here a person’s body becomes the vessel or container for a deceased person that moves, over time, to different bodily containers. It should be mentioned that relations with the deceased do not just make themselves visible though human bodies. For example, it is considered dangerous to bury many family members in the same location because they may start to call on the living to join them. It is, then, important to note that it is not always material pieces that remain when people are separated and move. With lay-reincarnations, people make themselves visible in another bodily form. THE DIFFERENCE BETWEEN THINGS The Mongolian concept of fortune provides a conceptual window through which kinship ideas can be discerned. Things, viewed as parts that are retained to sustain the whole, command our attention to different domains of connectedness between groups or individuals that extend beyond ideas about ‘shared bone’. In this way, a corporal presence is not always necessary for maintaining relations (cf. Telfer 1999). In photographic montages, individual people are dislodged to reveal infinite connections between groups. These connections can be understood as immutable chains that are passed over generations. Parts, hidden inside the chest, provide temporary vessels for future meetings between people. When focusing on these pieces and attending to their containment, they seem to burst out from behind the static groups and draw attention to people’s mobility. Although people may not visibly enact these relations in one location or site, the containment of particular parts, inside the household chest, anticipates alternative meetings between people. In his seminal work on place, Casey (1998:301-308) draws on Deleuze and Guattari’s distinction between ‘smooth’ and ‘striated’ space. Striated space refers to a landscape with distinct points or sites that mark its surface, and can be identified and specified. Rather than residing in place, people move through striated space from one point or location to the next. In contrast, smooth space has indefinite extensions. People move in this space rather than to fixed points along a trajectory. Smooth space is filled with multiplicities. It has a non-limited possibility of localities that resist exact concentration or reproduction. I suggest that the distinction between striated and smooth space can be used to highlight the ideas concerning Mongolian kinship that I have presented so far. With striated space, containment, permanence and renewal over time are paramount. Within the household different sites, such as photograph montages and portraits of deceased patrilineal elders, echo this idea. Like visible places in the landscape, such as oboos which pile high on mountaintops and sacred trees which emerge with fluttering ribbons from dark ravines, people ensure these sites visibly accumulate and grow in order to allow for people to attend to networks that may be drawn upon in various configurations throughout a person’s lifetime. Attending to these places, we find that people merge into groups that are fixed, passed on and contained over generations. In contrast, people also have to negotiate an inevitable series of movements and transitions as they move about in an absolute passage without a fixed centre. Here, we find that people never return in the same way. The things hidden inside the chest provide temporary vessels or places where relations based on the premise of separation and movement may temporarily reside. Like the concept of fortune in Mongolia, these relations have the potential to extend to unlimited places. They are uncertain because they do not guarantee prolonged residence in any fixed place over time. With such varied and unpredictable residence, we find dispersal of these relations is a constantly impending possibility. Exactly the same practice concerns domestic animals. A drawing of a horse or cow is often placed at the altar representing an animal that has been consecrated. This animal is allowed to run freely and will never be ridden or milked (seterlesen mal); it thus provides a container or vessel that accumulates fortune for the whole herd. Inside the chest, however, are pieces of tail hair from individual cattle and horses that have been gathered at points of separation and departure. The varied location of animal fortune and the in-between position of incoming women and young infants allows for their extension to, and partial placement in, any given space as they emerge in varied relations at different moments in time. Given these multiple ways of containing past and present relations, the idea of ‘commemoration’ in the Western sense, seems inadequate. In the West, relations with the deceased are often maintained through commemorative sites and living people need the physical presence of each other for relations. In Mongolia, relations between living people are created via the careful containment of dynamic and variable things, but relations with the deceased are maintained through the site of a person’s body (cf. Empson 2006). I suggest that instead of acting as sources in which people locate the memory of an absent person in a communal place, the things inside the chest work as channels that allow for relations to continue with people who are dispersed and separated. We have seen that these channels allow for spatio-temporal flexibility, whereby people do not need to be confined to relations within particular spatial coordinates (cf. Gell 1998:222). Revealing the whole through vision Although people and animals move across the landscape in Mongolia and there is a sense of unbounded vastness, the household and domestic encampment is desired as a fixed centre that reconstitutes itself in different places. We have seen that the chest acts as this site, gathering together different aspects of people’s relations. In this section I would like to clarify what happens when people actually view and engage with the display. Firstly, I will examine the set of kinship perspectives present in the display. I suggest that they are dependent on each other, but only ever appear independently. Secondly, I show how the two perspectives can, at certain points, simultaneously reveal a whole. This is achieved when: (1) a person observes an exemplary person through their own image in the display and, (2) when an observer views a person looking at himself or herself through the chest’s mirror. Finally, I explore the role of the mirror in Mongolia and examine its capacity to reflect or deflect knowledge. Let us return to the chest, which stands on the ground at the back of the Mongolian house with various objects contained inside and displayed on top of it. The way in which these objects are displayed can be taken as a template according to which Mongols view the person. Photographic montages outwardly depict the person as a replicate member of a whole group with links to other groups. When we switch perspective to the hidden parts that are usually delegated to the periphery, however, relations based on movement and transformation momentarily become a different kind of centre (cf. Wagner 1987). In order to understand how these two perspectives are dependent on each other, I divert briefly to examine the Rubin vase-profile illusion (cf. Arnheim [1954] 2002:223). The Rubin vase-profile illusion, developed in 1915 by the psychologist Edgar Rubin, is an image that most of us are familiar with. In this image, we either see two black profiles facing each other, in front of a white background, or a white vase on a black background. Rubin developed this image to illustrate the dynamic nature of subtle perceptual processes. Because one of the contours of the image is shared with the other, it is difficult to perceive both images simultaneously. Instead, our vision fluctuates between the vase and the profiles. As one image becomes the background, the other becomes the foreground and vice-versa. The ability to see one image, and then the other, but not the two simultaneously is referred to as ‘contour rivalry’ (i.e.: we shift attention between the shape or the contour). The reversal of images that the observer perceives in the vase-profile illusion is due to their individual tendency toward biasing either the shapes or contours, making one interpretation stronger than the other. With regard to the chest, we have seen that things displayed on top of or inside the chest allow us to switch perspectives between different ways of imagining kinship in Mongolia. When first viewed, agnatic relations are fore-grounded on top of the chest and the actual chest, as well as its contents, serve as a physical as well as a relational background for these relations. When we switch perspective to the parts contained inside the chest, however, we see that people have to transform and separate so that agnatic relations can continue. Through the Rubin’s figure-ground reversal, we can see how Mongolians alternate between a set of relational perspectives that are dependent on each other. By this I mean that the two perspectives only make sense in relation to each other. I would like to push the analysis beyond the idea of alternating constituent perspectives and suggest that, at certain moments, the two perspectives can be revealed simultaneously. While each aspect of the display mirrors a different way of tracing relations between kin, together they form an ensemble of the different relations that make a person. In the centre of the display stands a mirror, so that while standing in front of it the viewer sees a reflection of themselves looking back at them. Through this mirror reversal, the display allows us to simultaneously gaze at ourselves, while constructing a figure that stares back at us with our own eyes (see Figure 1.). A similar shift in perspective can be noted in Levin (1988). Levin (1988), drawing on the work of Jean Paris, examines the history of Western painting as a transformation of human vision whereby the seer gradually becomes a part of what is being seen in the painting. In Byzantine mosaics the viewer is an object that the gods look at. But in Renaissance art, due to the gazes in the paintings being cast in different directions, the viewer is afforded the possibility of becoming part of the scene (cf. Reed 1999). The display structures our vision so that a very particular image of ourselves becomes visible. In this sense, looking at the display always involves looking at and making ourselves visible in a particular form (cf. Sobchack 1992:51). The chest structures our vision, gathering together these different aspects of the person into a single image, allowing us to draw out the different relations through our gaze (cf. Strathern 1994). *Insert Figure 1.* Figure 1. The Mongolian Household Chest In viewing the display the parts come together to form a person While viewing each of the separate parts prompts a different reflection of the person, viewing the display in full, and seeing oneself through it, makes all of these relations visible at once. In turn, if we bring in the perspective of an observer who views a person looking at themselves through the display, the chest can be said to reveal a person as constructed out of all the parts. While the viewer’s gaze simultaneously looks at and is looked back at, the observer and the viewer see a figure constructed out of all the parts. Here, the chest is fore-grounded as a figure due to the viewer’s gaze which acts as a necessary background. I am not proposing that the concealed parts are suddenly physically revealed when people look at themselves through the chest. Rather the chest, as a whole, becomes visible as a site that is able to foreground agnatic relations due to the support (or background) of those relations based on movement and transformation. The invisible, hidden parts, generated through separation, support the possibility for the visible parts that constitute agnatic relations. When viewing the display in full, a person is revealed as constituted by all these things through their own gaze. Each of the things that we have focused on make parts of people’s relations visible at different moments in time, but when we view ourselves through the display, the parts appear together to form a site that reveals a person made possible through each of them. At the beginning of this paper I suggested that the creation of kinship in Mongolia is achieved through the separation of bodies. The separating, merging and remaking of bodies, has been an underlying theme throughout this discussion. While much attention has been given to the necessity of separating people through the containment of parts, by viewing the display in full and seeing a person through it, what flashes into view, as one glances at oneself at the centre of this display, is a calling into being of the person who views it. Viveiros de Castro (2004), in a chapter concerning, among other things, Piro concepts of the person, argues that the production of (consanguineal) relatives requires the intervention of (potential affine) non-relatives, and this can only mean the counter-invention of some relatives as non-relatives. Among the Piro, what distinguishes consanguineal relatives from affinal relatives are their bodily differences. If the body is the site of difference, then a difference is required in order to make bodies by means of other bodies (cf. Viveiros de Castro 2004:7). Kinship is thus in a state of reciprocal dependence whereby non-relatives are needed to make relatives and vice-versa. The discussion concerning the two constitutive perspectives resonates with this point. The figure that is revealed through the display also establishes a similar type of perspectival difference. As a person is revealed in the display the chest becomes the site and instrument of this bodily differentiation (cf. Viveiros de Castro 2004:8). In this sense, perception, or vision of self, is dependent on this instant of reciprocal gaze, whereby the detached pieces form a whole that can be looked at, in order to be able to see our self. I am not referring to the separate parts here. I am referring to the image that is produced through the chest as a whole. One result of experiencing this duality of looking at and being looked back at is that the person we encounter, in the context of the mirror, is not a person as an exact replication of our self, but rather a prototype of a person that can only be seen through its separation. For a similar practice see Willerslev (2004:641): ‘Under normal conditions, a person’s body is not presented to him as an object in the world, a thing that he can encounter or straightforwardly observe. Rather, it is an object only from the perspective of another, in the same way as another’s body is an object from the perspective of ego’ (Willerslev 2004: 639). See also Humphrey with Onon (1996): ‘What you see in the mirror is yourself and not-yourself. It is a depersonalised, two-dimensional, image which you do not spontaneously know, but have to recognise.’ (Humphrey with Onon 1996:225). In such a way, the chest transfigures and refracts the visible world, rather than merely duplicating it. It allows the viewer to see a transformed image of himself or herself. The chest thus provides a vehicle for recognising a type of exemplary personhood. It should be noted that I use the term ‘exemplary’ because the perspective that is revealed through the chest is impossible for a living person, who although a mother, daughter-in-law and daughter, cannot visibly enact all these relations at a single moment in time. The figure that is revealed can be said to crystallise the multiple relations necessary for kinship in Mongolia. Furthermore, the figure that is revealed through the chest shatters our own anthropological understandings of ‘blood’ and ‘bone’ as the foundation of kinship in Mongolia, allowing us to see beyond these ideas. It should be noted that there are restrictions on the type of vision that allows us to view a person in this way. The Buryats in Mongolia place limits on who can look at the mirror and when it is possible to look at it. For example, infants should not look in mirrors because they may see their reflected image as that of a stranger and have bad dreams. The soul of an infant is held to be unstable and can easily be ‘dislodged’ finding another bodily container to inhabit. Many things, including fear of strangers, may cause the soul-loss of young infants. In general, people do not look in the mirror at night (or if it is damaged), because the image that is revealed is distorted. When there is a full moon, people cover the mirror in cloth, or turn it around so that it faces the wall, ensuring that the round image of the full moon, which enters through the smoke hole (toono), does not appear in the mirror and illuminate the inside of the house allowing bad spirits to enter. In Mongol culture mirrors are used in two distinct ways. They reveal things that are otherwise not know, or they deflect things. In such a way, the mirror has the double capacity ‘to both gather in and deflect at the same time.’ (Humphrey with Onon 1996:226). In shamanic performance, for example, the mirror is used to deflect, or cast aside, evil spirits. But when used for divinations, vodka is poured on the mirror to reveal and confirm links of causation. A Mongolian friend explained to me that the double capacity of the chest’s mirror means that it is like an eye (nüd-shig). It must be regularly cleaned because the mirror may ‘see’ bad things. These should not linger and accumulate on its surface as the mirror may redirect these back in the house. In its double capacity, the mirror is ‘an instrument not only of containment and absorption but [it is] also [used] for breaking out of the world into another state which reflect[s] the hitherto unseen truth.’ (Humphrey with Onon 1996:226). It should be clear that because mirrors in Mongolia have the capacity to reveal that which is not known, the perspective afforded through the chest is not a perspective that a living person can embody. Not unlike the shaman discussed by Humphrey with Onon (1996:226-7), the image that appears through the chest is one that catches the fragmented aspects of the person and reveals a unity of form to the viewer. I have suggested that when a person looks into the mirror at the centre of the display an exemplary person is revealed. It is important to emphasise, however, that the import of vision does not have to rest solely on the presence of a mirror at the centre of the display. Elkins (1996), in a fascinating book on the nature of seeing, explores relationships established between objects and observers. Elkins (1996) examines the ways in which observes are altered by objects, or fused with them, through acts of seeing. His point is similar to that I have made about the mirror’s capacity to reveal an exemplary person to the viewer. Through looking at the display, the chest and the observer merge into an in-between state, lost in the field of vision. In such a way, vision is not a passive activity involving a discrete viewing subject and a viewed object: seeing alters the thing that is seen and reveals a metamorphosis to the seer (cf. ibid.:11-12). Siding with Elkins’ point, Mongols would argue that it is not just the mirror that reflects or deflects different perspectives; other things also send back our sight. In this way, one could equally make the point about any object in the display. Indeed, the arrangement of things in and around the Mongolian household chest differs between Mongolian groups. For example, among Xalx Mongols, a photographic montage is sometimes placed above the mirror, or the mirror may be in a different location entirely and a single Buddhist image may be displayed on the chest. Among Tuvinian groups, umbilical cords are placed in separate cloth pouches and are sometimes prominently displayed above the chest (C. Humphrey, pers. comm.). In these cases, viewing the display also instantiates kinship in the same manner (through separation, concealing, enclosing and revealing differing aspects), there is just a difference in the details of what people view according to their different ideas about what makes a person. This variation reflects the different ways in which Mongolian groups make kinship differently through similar means. Notwithstanding variations between different Mongolian groups, the household chest needs a corporal presence to be seen as a person made from the different parts. In this sense, vision, situated in another person, is necessary to create this perspective. I should like to point out why it takes a certain type of relation to the thing being viewed, to be able to see the container as an image of our self. That is, I would like to explain why a visitor to a household will not recognise the ‘body’ configured when viewing themselves through the chest’s parts, as something recognisable as themselves. I suggest that in these cases, a visitor may guess and speculate as to the chest’s contents and various parts, but they cannot fully see themselves as constituted through them. This is because the means by which they see an image of themselves does not relate to any aspects that they are familiar with. For visitors, the image that is revealed can only ever be a replication of themselves in a singular form. The concept of ‘reciprocal vision’, that one has to be looked at in order to be able to look (i.e.: that the gaze creates a relation), is also fundamental to the concept of fortune. If we recall the example of the cow presented at the beginning of this paper, something has to be given away in order for it to be kept back to support and increase the whole. In this sense, separation is an ontological precondition for the possibility to retain and increase fortune for growth of the family and its herds. Similarly the separation of people, or the ability to reflect on social relations through the containment of some part, is essential for the growing of kin relations in Mongolia. In turn, a person is only able to see their multiple aspects as an exemplary figure constructed in the display through their own gaze. In this way, we could conclude that Mongolian kinship relies on the separation and transformation of people in order for sameness, or consanguinity, to continue. The necessity of separation, in order for sameness, seems to be a theme that underlies many, if not all, social relations in Mongolia. We have seen how various things, inside and on top of the household chest, are made to stand for relations that are reflected through the gaze and interpretations of others. Through looking at ourselves looking at the display, each of the visible and hidden parts dissolve as separate channels and present themselves as a single form. The things that I have discussed can be seen, in their parts, to reference different kinship perspectives and points of transformation. When viewed together they appear as a whole. Thinking through these things has highlighted relations outside of relations based on agnatic kinship. It has shown that agnatic kinship survives due to these different relations. The visible and invisible aspects of people’s relations are mutually dependent on each other. We have seen that the ways in which these relations are contained in things, mirrors the ways in which these relations are realised in people’s interactions. Conclusion In conclusion, I should like to make some suggestions about why it might be necessary for Mongolians to index relations in things, and why this is tied to the idea that a person has to extend beyond their physical, or spatial location. Firstly, due to the practice of nomadic herding, it is necessary in Mongolia for people to reckon relations with people who are physically absent. When people are dispersed across the landscape, things contained in the household can be seen to allow for the containment of certain relations that cannot be enacted in shared place. We have seen that through the use of photographic montages, far-reaching kin terms, and the spatial layout of the house, Mongolians construct flexible ways in which to incorporate outsiders. These technologies act as potential ways of reckoning kinship that are drawn upon to activate wider networks in the absence of kin. Underneath these visible means, however, we have seen that ways are found to distinguish bodies and create different relations. The theme of retaining relations in the absence of people is something that seems to permeate the need for locating Mongolian kinship in different forms. While people may be necessarily absent, due to the constraints of nomadic herding, historical pressures of migration, and political persecution, relations can be imagined and contained through the construction of different sites. Focus on the display of these relations in households, brings to the fore some of the ways in which objects and people are mutually constituted by processes of objectification (Humphrey 2002:83). These things are the products of relations that, in turn, allow people to make further relations when they are viewed or displayed. The things discussed all transform the typically temporal aspect of people’s kin relations into a contained visual site. With the continual movement of both people and place, I suggest that viewing the household chest as a container or site is necessary for the imagining and creating of Mongolian kinship. ACKNOWLEDGMENTS I am grateful for insightful comments from those at the London School of Economics Anthropology Research Seminar when I presented an earlier version of this paper and from those at the Thinking Through Things workshop at CRASSH, University of Cambridge. Versions of this paper were also presented at St Andrews University and Manchester University Social Anthropology Research Seminars. I am very grateful for comments and suggestions during these seminars, some of which I have not been able to take up in this chapter, but which I hope will be addressed in an extended form. I also thank Martin Holbraad, Caroline Humphrey, Andrew Moutu, Aparecida Vilaça and Rane Willerslev for their important comments and ideas on earlier drafts of this paper. I thank the Press’s two anonymous readers for their helpful suggestions. Naturally, none of these people should be held in anyway accountable for any of my deficiencies, which are wholly my own responsibility. Writing this paper has been made possible due to the British Academy Postdoctoral Fellowship Award (BA PDF 2003 / 145). REFERENCES Arnheim, Rudolf ([1954] 2002) Art and Visual Perception: a psychology of the creative eye, the new version, Berkeley: University of California Press Bouquet, Mary (1996) Family Trees and their Affinities: the visual imperative of the genealogical diagram, in The Journal of the Royal Anthropological Institute Incorporating MAN, March, Vol.2, No.1: 43-66 --------- (2001) Making Kinship With An Old Reproductive Technology, in Relative Values: reconfiguring kinship studies, Sarah Franklin and Susan McKinnon (eds.), Durham and London: Duke University Press Casey, E. S (1998) The Fate of Place: A Historical History, Berkeley, Los Angeles: University of California Press Chabros, Krystyna (1987) The Decorative Art of Mongolia in Relations to Other Aspects of Traditional Mongol Culture, in Zentralasiatische Studien. Des Seminars fur Sprach – und Kulturwissenschaft Aentralasiens der Universitat Bonn, No. 20: 250-281, Germany: Otto Harrassowitz Wiesbaden Press --------- (1992) Beckoning Fortune: A Study of the Mongolian Dalalga Ritual, Germany: Otto Harrassowicz, Wiesbaden Press Elkins, James (1996) The Object Stares Back: on the nature of seeing, New York: Harvest Harcourt, Inc Empson, Rebecca (2003) Integrating Transformations: A Study of Children and Daughters-in-law in a New Approach to Mongolian Kinship, Unpublished PhD Thesis, Department of Social Anthropology, University of Cambridge --------- (2006) No Body, No Memory? The body of the other as a site for recalling and re-inscribing kin in Mongolia, in Ghosts of Memory: Essays on Remembrance and Relatedness (ed. Janet Carsten). Oxford: Blackwell Publishing (forthcoming) Galdanova, G.P (1992) Zakamensky Buryats, Historic and Ethnographic articles. (Second half XIX Century- First half XX), L.P. Potapov and K.M. Gerasimova (eds.), Russia: Russian Academy of Science, Siberian Department, Novosibirsk: ‘Science’ Press Gell, Alfred (1998) Art and Agency: An Anthropological Theory, Oxford: Clarendon Press Gil-White, Francisco J (2001) Are Ethnic Groups Biological ‘Species’ to the Human Brain? Essentialism in Our Cognition of Some Social Categories, in Current Anthropology, August-October 2001, Vol. 42, No. 4: 515-554 Humphrey, Caroline (1973) Some Ritual Techniques in the Bull-Cult of the Buriat- Mongols, The Curl Lecture 1973, in Proceedings of the Royal Anthropological Institute of Great Britain and Ireland, No. 1973 (1973): 15- 28 --------- (1974) Horse Brands of the Mongolians: A System of Signs in a Nomadic Culture, in American Ethnologist, (August 1974), Vol. 1, No. 3: 471-488 --------- (1979) The Uses of Genealogy: a historical study of nomadic and sedentarised Buryat, in Pastoral Production and Society, (ed.) L’équipe écologie et anthropologie des sociétés pastorals. Cambridge and Paris: Cambridge University Press and Maison des Sciences de L’Homme --------- (1995) Chiefly and Shamanist Landscapes in Mongolia, in The Anthropology of Landscape: Perspectives on Place and Space, E. Hirsch and M. O’Hanlon (eds.), Oxford: Oxford University Press --------- (2002) Rituals of Death as a Context for Understanding Personal Property in Socialist Mongolia, in Journal of the Royal Anthropological Institute, Incorporating Man, March, Vol. 8, No. 1: 65-87 Humphrey, Caroline with Onon, Urgunge (1996) Shamans and Elders: experience, knowledge, and power among the Daur Mongols, Oxford: Clarendon Press Lessing, F. et al (1960) Mongolian-English Dictionary, Bloomington: Mongolian Society, Indiana Press Levi-Strauss, Claude (1969) The Elementary Structures of Kinship, Revised Edition Translated from the French by James Harle Bell and John Richard von Sturmer, Rodney Needham (ed.), Boston: Beacon Press Levin, David Michael (1988) The Opening of Vision: nihilism and the postmodern situation, New York and London: Routledge Mauss, Marcel (1972) A General Theory of Magic, Translated from the French by Robert Brain, London and Boston: Routledge Kegan Paul Press Montell, Gösta (1934) Våra Vänner På Stäppen: genom Mongoliet till Torgoterna vid Etsingol (Our Friends on the Steppe: through Mongolia to the Torgots by Etsingol), Stockholm: Lars Hökerbergs Bokförlag Onon, Urgunge (2001) The Secret History of the Mongols: The Life and Times of Chinggis Khan, Onon, Urgunge (ed., trans., introduction), Richmond, Surrey: Curzon Press Pao, Kuo-Yi (1964) Family and Kinship Structure of Khorchin Mongols, in Central Asiatic Journal. International Periodical for the Languages, Literature, History and archaeology of Central Asia, Wiesbaden: Mouton and Co., The Hague and Otto Harrassowitz, Vol.9, No.4: 277-311 Park, Hwan-Young (1997) Kinship in Post-Socialist Mongolia: Its Revival and Reinvention, unpublished PhD thesis, Department of Social Anthropology, University of Cambridge Reed, Adam (1999) Anticipating individuals: modes of vision and their social consequence in a Papua New Guinea prison, Journal of the Royal Anthropological Institute, incorporating MAN, March 1999, Vol. 5, No. 1: 43-56 Sneath, David (2000) Changing Inner Mongolia: Pastoral Mongolian Society and the Chinese State, Oxford: Oxford University Press Sobchack, Vivian (1992) The Address of the Eye: A Phenomenology of Film Experience, Princeton, New Jersey: Princeton University Press Sontag, Susan ([1971] 2002) In Plato’s Cave and The Image-World, in On Photography, London, England: Penguin Books Ltd Stafford, Charles (2000) Separation and Reunion in Modern China, Cambridge: Cambridge University Press Stafford, Charles (2003) Living with Separation in China: anthropological accounts, Charles Stafford (ed.), London and New York: Routledge Curzon Press Strathern, Marilyn ([1988] 1990) The Gender of the Gift: Problems with Women and Problems with Society in Melanesia, Berkeley, London: University of California Press Ltd Strathern, Marilyn (1994) One-Legged Gender, in Visualizing Theory, Selected Essays from V.A.R. 1990-1994, Lucien Taylor (ed.), New York and London: Routledge Press Strathern, Marilyn (2001) Children In An Information Age, Unpublished paper delivered at Brunel University: ‘Children in their places conference’, June 2001 Telfer, Jon (1999) Relationships with No Body? – ‘Adoption’ Photographs, Intuition and Emotion, in Social Analysis, Subjectivities in Material Worlds, Anne Meneley (ed.), Issue 43 (3): 144-158 Viveiros De Castro, Eduardo. B (2004) The Gift and the Given: three nano-essays on kinship and magic, in Genealogy Beyond Kinship: sequence, transmission and essence in ethnography and social theory, Sandra Bamford and James Leach (eds.), Oxford: Berghahn Books (forthcoming) Vreeland, Herbert, Harold 3rd (1954) Mongol Community and Kinship Structure, Human Relations Area Files, New Haven: Behavior Science Monographs Wagner, Roy (1987) Figure-Ground Reversal Among the Barok, in Assemblage of Spirits: Idea and Image in New Ireland, L. Lincoln (ed.), Minneapolis: Minneapolis Institute of Art Weiner, B. Annette (1985) Inalienable Wealth, in American Ethnologist, May, Vol. 12, No. 2: 210-227 Weiner, B. Annette (1992) Inalienable Possessions: the paradox of keeping-while-giving, California: University of California Press Willerslev, Rane (2004) Not Animal, Not Not-Animal: Hunting, Imitation and Empathetic Knowledge Among the Siberian Yukaghirs, in The Journal of the Royal Anthropological Institute (N.S.) 10, 629-652 130 Separating and containing people and things in Mongolia 130 Rebecca Empson Rebecca Empson