The ongoing use of diagrams in anthropology has its roots in the emergence of the discipline itself. Ever since the work of Malinowski and a number of notable predecessors, diagrams (along with maps) have become a customary feature of ethnographic monographs – with some more standardised and familiar than others. A two-dimensional, often schematic, arrangement of lines drawn to show the organisation, appearance, arrangement, mechanisms or interactions within an area or action of analysis, the diagram has appeared in many different forms.This introductory review focuses first on two particular kinds: those used to convey information regarding kinship, and those depicting different forms of exchange.
Critiques and Challenges
Compared with other practices that rely on the visualisation of ideas and data, and which also operate within an interdisciplinary context, diagrams in anthropology have received less critical scrutiny than, for example, cartography and visual research methods. Photography and film – within Visual Anthropology – have become established forms of both presentation, and of method. They also provide objects of analysis in and of themselves. Interrogating the pitfalls and potential of their display via digital media has led to the development of ‘hypermedia anthropology’ (Pink 2006: xi) – enabling novel combinations of the visual, aural and textual. For some, this counteracts a previous rejection of the ‘visual, sensory and applied’ that coincided with social and cultural anthropology establishing itself as ‘a scientific discipline’ – a rejection of the ‘subjectivity of photography and film’ in favour of adopting ‘visual metaphors such as diagrams, grids and maps to synthesise and objectify knowledge’ (Pink 2006: 8; Grimshaw 2001: 67).Framed this way, diagrams lack the sensory transmission that multimedia forms of presentation seek, in part, to address.
Another critique questions the ‘decontextualisinglinearity’ (Ingold 2000: 140) of diagrams. In this light, the kinship diagram, for example, is seen as a chart that ‘can be taken in at a glance’ and ‘scanned indifferently from any point in any direction’, thus presenting ‘the complete network of kinship relations over several generations… as a totality present in simultaneity’ (Bourdieu 1977: 38, at Ingold 2007: 111). For Ingold, such a ‘snapshot’ resembles ‘the sterile austerity of an electrical circuit board’ – a schematic devoid of human inspiration – even adopting the technical convention of drawing a ‘hump’ where unconnected lines cross one another, echoing the circuit drawing of electrical engineers (Barnes 1967: 122; Ingold 2007: 111).
However, given their innate reliance on the visualisation of data, diagrams also appear to have much to offer the development of forms of ethnographic presentation that challenge, or augment, an exclusive reliance on text. Relationships between the two vary greatly, and critical approaches to cartography raise questions that are equally applicable to diagrams. For example, the idea that they conceal as much as, if not far more than, they reveal, and that any sense of accuracy comes at the cost of minimising complexities inherent in the lives and locales of research. Recognising that maps, as representations, are necessarily selective (Turnbull 2000: 101) leads many, among them Monmonier (1991), to emphasise how all maps ‘tell lies’.
This is not simply because the quest for an ‘accurate’ or ‘precise and comprehensive’ representation of reality raises impossible questions regarding what counts as ‘detail’ and ‘information’ on one hand and what constitutes the infinite, remaining ‘particulars’ on the other, but is because in the cartographic world, all is ‘still and silent’ – as opposed to the world of our experience that is ‘suspended in movement’ (Ingold 2000: 242). Crucially, for the types of diagram under consideration here, this cartographic or diagrammatic world threatens to conceptualise social relations as static social facts rather than as ‘dynamic phenomena,’ offering a particularly empty conception of social life (Kertcher 2006) and envisaging these relations without space to query how they persist or diminish over time (Suitor et al 1997). As we shall see, however, questions around how diagrams are used in anthropology are as numerous as the forms they adopt. Maps of places can be used for guiding and informing our interaction with the world. Diagrams of human relations of different kinds tend not to share such an explicit purpose, however. The role of an exchange diagram, for example – what we might decide it is for – depends very much on the ethnographic material that accompanies it, and which generated it in the first place. In what follows, I present various examples drawn from the anthropological to begin exploring some of these issues and relations.
Bourdieu questioned the origins and meanings of familiar ‘graphic representations of kinship,’ recommending a ‘social history of the genealogical tool’ (Bourdieu 1977: 38, 207) – a task which Mary Bouquet addresses by highlighting affinities between ‘European iconographical tradition’ in ’sacred, secular and scientific family trees’ and the ‘conceptual field’ around the anthropological kinship diagram (Bouquet 1996: 45, 59). As elsewhere, these traditions and influences are seen as coalescing and finding form within the work of W. H. R. Rivers and his visualisation of kinship in the genealogical diagram (ibid.).
Rivers is usually credited with developing the ‘genealogical method’ within anthropological inquiry. In his words, this was to involve the means of both ‘obtaining information’ and of ‘demonstrating the truth of this information.’ In this, diagrams were seen as crucial devices in the presentation of ‘facts,’ and as a way to ‘guarantee the accuracy and completeness [of those facts]’ (Rivers 1910: 11). This was a ‘staunchly positivistic approach’ (Stocking 1992: 34) and explicitly sought to establish the emergent discipline of ethnology ‘on a level with other sciences’ by ‘demonstrating the facts of social organisation’ in such a way as to ‘carry conviction to the reader with as much definiteness as is possible in any biological science’ (Rivers 1910: 12). Visual representations thus became ‘an argument for the credibility of the scientists’ inferences’ (Gifford-Gonzalez 1993: 26, at Bouquet 1996: 45). Kinship diagrams were not his invention, however: Morgan’s ‘diagrams of consanguinity’ in his Systems of Consanguinity and Affinity of the Human Family (1871) were also based on historical models of the ‘family tree.’
Rivers argued that the systematic presentation of genealogical ‘facts’ offered a way to get ‘beneath the skin’ of human beings to the relations that people were born into and developed throughout their lifetime, admiring how once people had been identified in a genealogical diagram they ‘became real personages… although I had never seen them’ (Rivers 1968: 105, at Bouquet 1996: 45). This reflects both the ‘concrete method’ of questioning in order to learn personal names and terms ‘known by informants,’ and also the weight given to the ‘abstract system of relations underlying those names’ – an abstract order that was itself ‘reconcretized (visualized)’ in the ‘genealogical diagram’ (Bouquet 1996: 45). Rivers’ diagrams led to the conventionalisation of inverting the ‘tree’ of family trees, placing ‘its roots at the top’ (Bouquet 1995: 42–3; 1996), and thus erasing “the image of the tree as a living, growing entity, branching out along its many boughs and shoots, and [replacing] it with an abstract, dendritic geometry of points and lines, in which every point represents a person, and every line a genealogical connection” (Ingold 2000: 135).
This inversion had lasting effects.The stories that ‘people tell about themselves’ and ‘the information gleaned from them by systematic forensic inquiry’ (Bouquet 1993: 140) that Rivers described continued to influence the ‘systematic collection of genealogical data’ – methods that in 1967 Barnes acknowledged could ‘scarcely be improved’ (Barnes 1967: 106, at Ingold 2007: 110). Bouquet also suggests that visualising kinship in the genealogical diagram reflects “the limits of a specific ideological consciousness, [marking] the conceptual points beyond which that consciousness cannot go, and between which it is condemned to oscillate (Jameson, in Clifford 1988: 223)” (Bouquet 1996: 44). Their presence persists (Bouquet 1996: 44) and genealogical diagrams are established as images for use “on the edge of the text” (Stoller 1994: 96) – each (diagram and text) expanding on the explanatory reach of the other.
Diagram as Method and Delivery
To recognise this is to emphasise how “the diagram is a possibility of fact – it is not the fact itself” (Deleuze 2004: 110). That is, genealogical diagrams are ‘contemporary models for social relations’ (Barnard & Good 1984: 9), portraying the inter-relationships of ‘real or imaginary individuals’ (ibid. p.8). The significance of these diagrams is not established until ‘the nature of those relationships between the individuals portrayed is clarified’ (Bouquet 1996: 45). Malinowski recognised the visual clout and direct efficacy of the reduction of data within visual forms, whilst at the same time elaborating on the kinds of details and observations that are necessary in establishing the relationships portrayed – how to flesh out the bones of the genealogical diagram: “The method of reducing information, if possible, into charts or synoptic tables ought to be extended to the study of practically all aspects of native life. All types of economic transactions may be studied by following up connected, actual cases, and putting them into a synoptic chart. Also, systems of magic, connected series of ceremonies, types of legal acts… a table ought to be drawn up of all the gifts and presents customary in a given society, a table including the sociological, ceremonial, and economic definition of every item… Besides this, of course, the genealogical census of every community, studied more in detail, extensive maps, plans and diagrams, illustrating ownership in garden land, hunting and fishing privileges, etc., serve as the more fundamental documents of ethnographic research” (Malinowski 1922: 11).
Malinowski, along with Radcliffe-Brown, Evans-Pritchard, and Fortes (among others) sought to understand the ‘basis for the orderly functioning’ of small-scale, effectively state-less societies, and kinship was seen as ‘constituting the basis and structure for social continuity’ in these settings (Carsten 2004: 10). Latter work was dominated by ‘avowedly ahistorical studies of African unilineal kinship systems,’ treating the lineage as bounded units, and developing a ‘complex typology’ to describe the functioning of these systems, involving ‘maximal’ and ‘minimal’ ‘lineages’ and ‘sublineages’ (Carsten 2004: 11).
Evans-Pritchard adapted the visual metaphor of the tree to account for such notions of scale in the inter-relationships between Nuer clans and lineages. He also made attempts to represent Nuer descriptions and depictions of these inter-relationships.
In these attempts, Evans-Pritchard explicitly states that it was only the analyst (or ‘we’) who insisted on this visual metaphor, highlighting something of its limitations and biases: “[the Nuer] do not present [lineages] the way we figure them as a series of bifurcations of descent, as a tree of descent, or as a series of triangles of ascent, but as a number of lines running at angles from a common point… they see [the system] as actual relations between groups of kinsmen within local communities rather than as a tree of descent, for the persons after whom the lineages are called do not all proceed from a single individual” (Evans-Pritchard 1940: 202).
During this era of kinship studies in Britain, ‘largely preoccupied with the analysis of descent groups’ (Carsten 2004: 12), such projects in France followed a route influenced by Lévi-Strauss’s The Elementary Structures of Kinship with an emphasis instead on social rules, the generation of exchange, and marriage (ibid.). The once-raging debates between adherents of “alliance” or “descent” theories do not need to be repeated here. For current purposes, I focus on how the established symbolic formulae of kinship diagrams have been adapted for use in different ethnographic works, and how the diagram has been modified to focus analytical attention on different aspects of social life. Kinship diagrams do not always fit the static model they imply: it’s not always the intention for ‘each triangle and circle [to represent] one real man or woman’ since they may be used ‘to be used to represent fictive genealogies of imaginary persons’ (Barnard & Good 1984:7). Even when the correlation between diagram symbols and living individuals is more direct, kinship diagrams have been diversely adapted, and ‘constructed so as to bring out certain structural features’ that the work seeks to draw attention to (Bouquet 1996: 60).
Relationality and Decentering
The different variations on kinship diagrams above share in common a recognisable linearity. Reflections on these conventions question their impact on reinforcing particular notions of relationality, and subsequent effects on ideas around alterity and the individual. In the above examples, the passage of time (in people’s lives) has a directional, generational thrust that can be depicted (across the page) accordingly. For Ingold, this trend reinforces anthropological habits of insisting that ‘the way people in modern Western societies comprehend the passage of history, generations and time’ is ‘essentially linear,’ which casts anything that is not immediately recognisable in an opposing category: “alterity, we are told, is non-linear” (Ingold 2007: 3) – and this, in turn, equates ‘the march of progress’ with the ‘increasing domination of an unruly – and therefore non-linear – nature” (Ingold 2007: 155).
Sahlins, meanwhile, suggests that “the partible ‘dividual’ has become a regular figure of kinship studies as well as a widely distributed icon of the pre-modern subject,” perhaps as a result of anthropologists “staring too long at ego-centred, cum egocentric, kinship diagrams” (Sahlins 2011: 13). As such, we have learned to make the mistake of “rendering the relationships of kinship as the attributes of singular persons” (ibid.). Not only this, but also considering ‘kin persons’ as the only kind of persons who are ‘multiple, divisible, and relationally constructed’ leads to a tendency to overlook the fact that more familiar terms are also relational, among them ‘employees’, ‘clients’, ‘teammates’, ‘classmates ‘, guests’, ‘customers’ and ‘aliens’: “When aspects of the same person, variously salient in different social contexts, they are instances of partibility. But they are not instances of ‘dividuality,’ since they do not entail the incorporation of others in the one person” (Sahlins 2011: 13).
In her description of the Yarralin people’s world view, Rose (2000: 221) describes individuals as shaped by their own personal ‘angle of perception,’ the angle of their matrilineal identity, and their ‘various country angles which tie them into other species and to the workings of the world” (ibid.). The diagram drawn to reflect this resembles Deleuze and Guattari’s ‘rhizome’ (Deleuze and Guattari 1988: 18): “a dense and tangled cluster of interlaced threads or filaments, [where] any point in which can be connected to any other” (Ingold 2000: 134). As we have seen, the tree has become one of the ‘most potent images in the intellectual history of the Western world,’ not only used in diagrammatical form to represent ‘hierarchies of control and schemes of taxonomic division’ but also, and above all, ‘chains of genealogicalconnection’ (ibid.). The rhizome model, by contrast, looks beyond the ‘static and linear, arborescent and dendritic imagery of the genealogical model’ to begin thinking ‘about persons, relationships and land’ in a world in movement, “wherein every part or region enfolds, in its growth, its relations with all the others” (ibid.): “a continually ravelling and unravellingrelational manifold” (Ingold 2000: 140).
Rose’s description of relationships and contexts – based on Yarralin ideas about wisdom, difference and interconnection – includes the influence of (physical and relational) positioning on perception, in a strikingly ‘rhizomatic’ account: “an angle of perception is a boundary, and boundaries are both necessary and arbitrary. Necessity lies in the fact that there are no relationships unless there are parts, and without relationships there is only uniformity and chaos. Arbitrariness lies in the fact that since all parts are ultimately interconnected, the particular boundary drawn at a given point is only one of many possible boundaries. Each line in Figure 24 is both and boundary and a relationship. Each node (A, B, C, etc) is both a context and an angle of vision, another centre… One particular human angle defines our world as it is because it is we who are looking. Perception distorts, but wisdom lies in knowing that distortion is not understanding” (Rose 2000: 222).
Such developments and reflections take us beyond the more recognisable examples of kinship diagrams, especially those focused on lineages and descent. As mentioned above, Lévi-Strauss’s work on kinship shifted focus to the importance of marriage, and of exchange more generally, in ‘establishing and maintaining relations between groups, rather than just individuals’ (Carsten 2004: 14). In this, he developed models for ‘elaborate, long-term exchange[s] involving the transfer of goods, services, and people that cemented relations between groups’ (Carsten 2004: 14) – making extensive use of diagrams.
There is a tension at the heart of anthropological diagrams of exchange concerning attempts to represent movement (spatiotemporal change) and the effects of time passing. Holbraad (2012: 101) asks why a line is ‘appropriate for representing a trajectory [of change]’ and how the inherent continuity of trajectories relates to the ‘momentum’ of movements and action. Is demarcating, visualising and representing the ‘continuity of plotted trajectories’ not a ‘very faint way of expressing momentum’ (ibid.)? He adds that ‘tota simul representations on paper’ have to be ‘economical’ since ‘they do not move in themselves, and hence they cannot really have a momentum,’ but argues that this economy ‘comes at a price’: “For the point about momentum is not only that it renders motion both continuous and directional, but also that it does so as a matter of necessity: momentum describes the inner compulsion of motion. The best way to understand this, I think, is cinematic: imagine panning away from the bird‘s-eye perspective of diagrams, and placing the “camera” at the helm of a moving trajectory, cockpit-style” (Holbraad 2012: 101).
With kinship diagrams, their linearity directed the passage of time and the segregation (or interaction) of generations. Attempts to visualise and represent exchange, however, emphasise the ‘movement’ of transfer – relying on directional arrows to depict action and change, often across both time and space. As the following examples illustrate, on the more abstract level of economic theory, diagrams are apt devices for illustrating modes and relations of trade and transfer – operating at different scales in order to reflect different ‘flows’ (Appadurai 1996) of goods, labour, capital, value, commodities, people and technologies.
Building on diagrams of neoclassical economy (see above), Gudeman (2001: 7) draws the Economy as a complex of ‘practices and relationships’ that are ‘constituted within the two realms of market and community’ and the four ‘value domains’ he terms ‘the base, social relationships, trade, and accumulation’ (Gudeman 2001: 5). In this diagram (below, left), he emphasises the difference between established contemporary theories of value relativism through individual preference and its influence on demand and supply, and his own that proposes a world of ‘inconsistent, or incommensurate, domains of value that are locally specified’ – ‘culture’ is thus “made and remade through contingent categories, such as home and work, body and the other, weekdays and weekends, beauty and efficiency, or friendship and love. Different value arenas make up economy” (Gudeman 2001: 6-7).
The economic diagram format suits cases where the directional flow of abstract goods or entities is depicted in transfer or exchange with similarly abstract (or, rather, generalized) actors. Gregory’s work on gift economies makes extensive use of such diagrams: at the initial level, distinguishing between the single, quantitative exchange relation ‘established between objects’ in commodity transfer, and gift exchange that ‘consists of two transactions [where] the transactors become mutually indebted to each other – the exchange relation is established between the transactors rather than the objects’ (Gregory 1982: 46). The gist of these differences is summarized in two, simple figures (3.1/2, below).
Sites of Exchange
Another broad category of anthropological exchange diagrams attends less to abstract principles or temporality and instead focuses on the specific locales, or sites, of exchange interactions. As such they more closely resemble maps/plans, but often also contain or suggest particular forms of movement and/or interaction.
Gell’s account of the Dhorai market (in Madhya Pradesh, central India) pictures the market as a wheel: different groups of traders are able to sit and transact business in particular areas according to social rank, and the goods they trade in are also ranked, from the most prized (more central) to the less valuable (more peripheral) (Gell 1999: 121).
Malinowski’s (1922: 63) famous map of the Kula ring – an extensive form of exchange carried out across a wide range of islands that form a ‘closed circuit’: “in the direction of the hands of a clock… long necklaces of red shell, called soulava… in the opposite direction… bracelets of white shell called mwali… Each of these articles, as it travels in its own direction on the closed circuit, meets on its way articles of the other class, and is constantly being exchanged for them” (1922: 64). We are told that ‘every movement of the Kula articles’ is ‘fixed and regulated,’ that no one ‘ever keeps any of the articles for any length of time in his possession,’ and that transactions lead to ‘permanent and lifelong’ connections – none of which is visualized around the text (Malinowski 1922: 62). Others (two examples follow) have subsequently revisited the Kula ring.
Hage (1977: 29) describes his diagram as “an undirected graph of the Kula Ring” – it follows Malinowski’s descriptions and plots 18 points (each representing a Kula community) at their ‘approximate relative locations’: “Each point is enclosed by a broken line roughly indicating the territorial extent of the Kula community as an island, a part of an island or a group of islands as in Map V in Argonauts of the Western Pacific. The unbroken lines represent trading relations between these communities” (ibid.) – adopting this form to highlight how trade links may be of any physical distance but ‘may not pass through the territory of another Kula community’ (Hage 1977: 30).
Godelier and Jablonk’s (1998) diagram (above) combines elements of each category outlined above: flows; sites; sequences (interactions); routes. The Baruya had trade links with 12 other tribes; whose territories are located from 1/2 day’s walk to more than 3 days’ walk away’ – journeys were made to exchange ‘bark cloth, bows and arrows, stone adzes or steel axes, feathers, shells, dogs, and pigs’ (ibid.). This account queries standard notions of the operations within cashless economies: “With such a diversity of goods being exchanged, it might be difficult to find just the partner who had on hand the item one wanted. The problem does not arise, however, because salt bars, like currency, can be exchanged for all kinds of subsistence goods (e.g., bark cloth, stone adzes, arrows) and all kinds of luxury goods (e.g. feathers, shells). Salt bars crisscross all these distinctions. There is a known and accepted rate of exchange of salt bars for any given type of item with each other tribe. The partners in this exchange system are not trading in order to make profit, but rather in order to fulfill their needs as individuals and as members of their society. Nobody hoards salt bars, and nobody withholds goods in order to create an artificial scarcity to force a rise in price. This trading system requires regular, permanent, face-to-face relationships with people with whom one will continue to deal for many years. Everyone knows the accepted rates of exchange” (Godelier & Jablonko 1998).
Directionality and Irreversibility
Questioning and expanding on the diverse use of diagrams in anthropology parallels broader concerns within the discipline as a whole – not least how we understand attempts to create ‘a moving picture of a world that doesn’t stand still’ (Clifford 1997). Bourdieu (1990) challenged the structuralist analysis of gift exchange and the idea of ‘some abstracted and synchronic “law of reciprocity,”’ highlighting instead ‘the political judgement of the agents involved as regards the timing of the giving of the initial gift and then of the counter gift’ (Jedrej 2010: 692). This is to question structural analyses that ‘deal with a synchronic virtual reality’ and tends to ‘privilege spatial relations and their analogues in such forms as synoptic tables, diagrams (structures) and figures,’ and is instead to deal with ‘practice,’ which ‘necessarily unfolds in time and has all the properties which synchronic structures cannot take into account, such as directionality and irreversibility’ (ibid.). There are works – such as those on the concept of landscape – that explore and articulate the intersections of time, space and practice (Jedrej 2010: 692). As the above examples suggest, those same intersections urge further examination and exploration through the use of diagrams in anthropology.