A community of critics? Thoughts on new knowledge*
The Huxley Memorial Lecture, given to the Royal Anthropological Institute, 8 December 2004.
Department of Social Anthropology, Free School Lane, Cambridge CB2 3RF, UK.
Social anthropology is used to terrains shifting under its feet. Things observed from afar suddenly become near, and the knowledge economy is an example. This article considers the place of anthropology as a discipline in a world where creativity becomes an adjunct of productivity, interdisciplinary collaborations become a paradigm for innovation, and everyone is valued for their expertise. How to lead a critical life emerges as a new kind of problem.
For some years I carried round an issue of the American Anthropological Association's General Anthropology. It had a startling message I did not know how to think about. This was in a study of the way classes discuss controversial issues (Trosset 1999), and the startling message was that tolerance is an obstacle to discussion. The author was commenting on the general assumption that people need to feel safe in order to discuss freely. A typical situation is the open microphone passed around for everyone to speak, with all secure in the knowledge that no one will be challenged for their beliefs, some positive value has to be seen in each view, and when they talk people will express views that belong to them. Commitment to diversity goes hand in hand with a high value placed on comfort. If this is tolerance, how can it be an obstacle?
There are echoes here of what has been called a ‘recognition space’ in the interaction of Australian law and Aboriginal law,1 an arena where Euro-Australian law is prepared to recognize Aboriginal claims to land. This recognition space is not about supporting substantive indigenous understandings; it simply acknowledges the admissibility of such claims in the Australian courts. Nothing larger is ceded. So it does not endorse the reasoning or rationale of ‘Aboriginal law’, but only the fact of these claims. It is not, however, because of its narrowness that Weiner (2002), the anthropologist reporting on Australian ‘recognition spaces’, is uneasy. What is the source of his unease?
Recognition space, Weiner argues, implies the notion of culture as an object occupying a specific (conceptual) domain. ‘Recognition’ becomes an after-thought to what already exists, a means of communication across a divide already there. Yet ontologically one cannot distinguish between a difference that emerges within a culture and a difference that emerges between two cultures. ‘Why not start’, Weiner says, ‘with “one world”, wherein peoples, languages and more-or-less well understood “laws” contingently and praxically exist, and posit as our subject matter the differentiating activity that emerges from it – and results in such categories as “indigenous” and “non-indigenous”?’ (Weiner 2002: 3, my emphasis).
As to the problem of tolerance, the unease here is because everyone talks and nothing is argued. The classroom defracts opposed interests into so many personal ‘points of view’. While you can hold a viewpoint and relay it, you cannot argue from one. In order to argue, you need to have detached yourself from – divided yourself off from – competing positions that you might otherwise (in some other life) have occupied. We might say that division is the essence of argument. Tolerance embodied in the roving microphone is no tolerance at all if there is nothing to cede.
Different as these cases are, in neither does the substance of people's positions have to be conveyed.2 Substance seems all in the very different kind of encounter that Hirsch (2004: 184) draws from Galison's notion of ‘trading zones’:
What happens when an H-bomb designer, a logician, an aerodynamical engineer, and a statistician sit down together? Whatever else they do … they do not found a League of Nations with simultaneous translators (or their scientific equivalents) perched over the assemblage in metaphorical glass booths. No: they work out an intermediate language, a pidgin, that serves a local, mediating capacity (Galison 1996: 14).
This image of practical people getting on with their work, ‘rolling up their shirtsleeves’ (to keep with the 1950s imagery), and being inventive about their means of collaboration is all too compelling. It is one of spontaneity, not contrivance. People's activities are governed by the problem they have to hand; they do not stand on ceremony but get down to business.
Galison presents simultaneous translation and trade pidgin as contrasting ways of managing distinct identities. The former creates value in the speakers (or what they represent), whether through techno-legal instruments of recognition or through an ethos of tolerance; the latter finds value in the objects of transaction, including language, defined by their usefulness or by their ability to communicate. However, I want to suggest that this second, down to earth, transactional paradigm is quite as problematic as roving microphones and technical spaces for recognition. We should be quite as uneasy.
No one would want to diminish the value of tolerance or deny the legal accomplishment of ‘recognition’; the issue is the way in which their management suppresses or displaces division. The same has become true of the once honest trade of collaboration. Trading zones should be micro-managed! The last couple of decades has seen, not least on the part of the UK government, increasingly explicit moves to create arenas for spontaneous synergy and to generate innovation out of boundary-crossing, not least in the name of knowledge. Creativity is seen to lie in the ability to combine elements from many sources to the benefit of business and education alike (Leach 2004: 153). Trading zones, now signs of the ‘creative’, are seen as not robust enough to be left alone and must be actively encouraged.
Social anthropology seems particularly well placed to assess some of the implications of such micro-management, and I contrast a managerial model of knowledge creation with a research model. Here anthropology exists within a field of disciplines, and this is where Huxley will come in. I hope the reader unfamiliar with UK institutions will find enough information to satisfy ethnographic appetite.
Management models and research models
On the face of it, the so-called ‘knowledge economy’ appears made for research. Evidence-based policy-making goes hand in hand with a Euro-American understanding of the world as full of uncertainties (Barnett 2000; Power 2004; Strathern 2005). These uncertainties are not just political or economic but epistemological: we do not know enough – more research is needed. Researchers themselves know that the more they apply knowledge, the more problems proliferate, especially through the engine of technology that turns scientific knowledge into useful products. They become suddenly uncertain about usefulness or about (say) the ethical or social consequences ahead.
Research itself is an engine for turning diverse uncertainties into epistemological ones, for seeing gaps in information, and for creating the very premises of doubt on which knowledge-seeking rests. When knowledge takes the form of information (the ‘knowledge economy’), Euro-Americans produce a sense that it can be quantified, whether in terms of sufficiency (how much is needed) or through multiplying different compartments of it. Each new context creates new reasons to doubt that we know what we are doing, and new scope for research.
Consider the contemporary relationship between research that comes out of universities and ‘society’. It is scientific research in particular that attracts questions, partly because of the massive nature of public funding and partly because of its potential for far-reaching practical consequences. Together these create a new arena (agora, see Nowotny, Scott & Gibbons 2001: 203) concerned with rendering productive the flow and consumption of ‘knowledge’. Questions about ethical or social implications are raised not to understand these processes for their own sake but to seek normative precepts by which to act or give advice. One of the most important axioms of the knowledge economy is that action must be predicated on information, and from this can be extrapolated the more general principle that knowledge is worthwhile to the extent that it can be used. Information must thus be made useful if it is to inform policy. This is where ‘management’ comes in.
The knowledge economy that is made for research is also made for management!
The public sector … represents the biggest single resource for the creation of value-added information content and services. And yet, knowledge management … often falls by the wayside, displaced by other more ‘urgent’ legislative demands. However, implementing a KM [knowledge management] strategy now will actually enable you to meet these legislative demands faster and more effectively … Underpinning all of these challenges is the biggest challenge of them all: culture and change management. You need to find a successful way of re-engineering and distributing knowledge to break through old barriers and reach a new plateau of knowledge sharing. But how can you instigate such cultural change and where do you begin? Our speakers have been involved in key projects designed to seek out knowledge, manage and organise a knowledge-sharing culture and improve cost and time-effectiveness (from conference on knowledge management for the public sector, Ark Group 2004).
‘Knowledge management’ is a burgeoning academic discipline (e.g. Morey, Maybury & Thuraisingham 2002; Newell, Robertson, Scarborough & Swan 2002). When it involves knowledge that members of an organization share, distribute, and indeed create about themselves as an organization we can say that managers are like researchers, producing primary information in the field.3 They filter it, knowing what should be thrown away and what kept according to their own disciplinary protocols.
At the same time, knowledge managers may be dealing with information initially produced elsewhere, as technical know-how often is. Here managers have no option but to take what is already created and to ‘control’ its distribution and consumption.4 Indeed the need for management is generated when people become conscious of specific kinds of information belonging to specific domains that are produced ‘off-stage’. We have a name for information produced off-stage when it appears as an already existing specialism – expertise. Expert knowledge comes already filtered. Of course, other people's expertise may be endlessly questioned, side-stepped, and compared, but it cannot be rectified. One can choose what to take from it, but a non-expert is precisely one who cannot judge in terms of the quality of the information itself what can be thrown away and what should be kept.
Now a large part of the research process consists in throwing away things – blind alleys, dead devices, strained conclusions – and the researcher is constantly experimenting with how arguments do or do not fit. It is important to keep an open mind in order for new combinations to form against the background of epistemological doubt that drives research in the first place. The researcher seeks certainty, confident of finding new sources of uncertainty and sure that one research programme will lead to another. But to arrive at either, one may have to explore avenues that prove worthless, to abandon projects, and to discard theses. Only some ideas get taken on. The dead end comes not from failure to manage information, but from being unable to combine ideas, materials, analyses, and so forth, so as to produce credible outcomes (Hirsch 2004). Much of the activity of research lies in distinguishing between powerful and weak combinations in the light of everything else that is known. The researcher turns into a manager, however, when boundaries of expertise are crossed and research has to be presented to those who do not share that ‘everything else’.
So there is a difference between the goals of research and those of knowledge management. Classically, management seeks to reduce uncertainty.5 Creativity lies in ensuring the best outcomes for all; in management terms everyone could get an A. Since everyone can theoretically get an A, mismanagement is inevitably blamed for failure. The manager is then forced to become a researcher who determines from the evidence what, according to the relevant management model, seems inefficient or superfluous and thus discardable.
What a back-to-front mishmash the RAE (Research Assessment Exercise) turns out to be! This government-led audit of research activity in UK universities that determines the future allocation of public funds encourages academics to be managers of their relations with the outside world when they should be thinking about research. Scholarly reasons for publishing are subordinated to strategy and competition and, in the early RAEs, to the message ‘publish as much as you can, and above all do not throw anything away’. It encourages something closer to research on what should probably be managed insofar as it appraises an institution's output with reference not to its diverse contributions to the distribution and consumption of knowledge but in terms of its relation to the RAE's own paradigms of excellence.
Management is something we all do all the time – the routines of life would be impossible without it – but these days managerial practice is explicitly a specialism (even if ‘managers’ are becoming less visible, Fournier & Munro 2004). It is embedded in techniques such as those of the roving microphone and the recognition space, both of which contain (control) situations that might otherwise be explosive. Management's approach to an uncertain future is to try to secure a future free of the dangers of the present. By contrast with knowledge in a research environment, where every failure yields more information, here we come up against the perception, all too real in business and enterprise, that failure constitutes a risk to the organization.
Management cannot ‘fail’, but it can be good or bad, and organizations can fail. To the risk of (organizational) failure, risk management would seem to be the answer. However, as Power (2004) suggests in his second broadside – the first was the precursor The audit society (1999) – these days the challenging issue is how to contain the growth of the ‘risk management of everything’. Over the past decade in the UK, risk management became a ubiquitous model for administrators. It threatens, however, to immobilize institutions and expertise alike. For what is at stake is not only trust (O’Neill 2002) but, in Power's words, ‘the cultivation of honest opinion and professional truth’: ‘The risk management of everything reflects the efforts of organisational agents … to offload and re-individualise their own personal risk. The result is a potentially catastrophic downward spiral in which expert judgement shrinks to an empty form of defendable compliance’ (Power 2004: 42).
Power says, in terms that could almost describe the research model of knowledge creation, that we need a new political and managerial discourse of uncertainty. A new politics of uncertainty would allow professional competence to flourish and expert judgement to be trusted while, in legitimizing failure, enabling a proportionality of response to erroneous decisions (2004: 63). At present, however, ‘the risk management of everything poses major risks to a society in which the most pressing and most unpredictable problems cannot be solved without the effective marshalling of expert knowledge and judgement’ (2004: 42).
In short, a management model of knowledge creation – disseminating materials, guiding their consumption, and producing information strategic for organizational success – is inappropriate as a total response to risk. In Power's view, far from managing expertise, it threatens to displace expertise altogether. If what is needed is the effective marshalling of expert knowledge and judgement, we have come by another route to our pidgin-speaking party of professionals rolling up their sleeves and getting down to problem-solving. In the case I want to develop, however, management has got there already, and it instantiates the risk it poses – that of pre-emption.
Interdisciplinarity: origin or outcome?
It should be clear that I do not oppose management to research in any totalizing way. I certainly do not suggest that the research model axiomatically creates free spirits and creative thinkers. Each is a point in a particularly Euro-American oscillation between the condition of knowing through investigation (research) and the condition of asking what is to be done with that knowledge (management).6 Each gestures to a particular kind of work. What invites us to look at them apart is how each gets built into institutions and technologies, becomes the ruling paradigm for certain kinds of activities, and provides those activities with their rationale.
What university people have learned to call the research university is also the managed university. Here I take on board Brenneis's challenge (2004: 580; see Lederman 2004) to engage ethnographically with the institutional contexts in which anthropologists principally work. The university is expected to house not just institutional managers but also research managers.7 This is encapsulated in ‘new’ ways of conceiving the organization of knowledge itself. The practice of cross-disciplinary engagement, canonically taken as ‘interdisciplinarity’, is the case I take up. By contrast with multi-disciplinary encounters, which align different voices and provide simultaneous translation, there is with interdisciplinarity a promise of a pidgin, an epistemic transfer, affecting the very knowledge base on which one works (Mansilla & Gardner 2003; Rhoten 2003).
I refer to interdisciplinarity in the abstract because its most powerful grip lies in the very idea of it.8 It combines in itself two sets of values that Euro-Americans, and especially those of the British sort, find compelling. On the one hand is all the creativity contained in the idea of crossing boundaries with the innovative possibilities of making connections. On the other hand are the ‘shirtsleeves’, the logic of marshalling of experts to talk with one another to solve problems, and the practical sense of addressing issues that cannot be handled by one approach alone. It is an unbeatable combination. Needless to say there is also a great pull to imagine that disciplinary boundaries can be transcended altogether, as in the coinage ‘transdisciplinarity’ (Gibbons et al. 1994; Nowotny, Scott & Gibbons 2001).
For, rather as existing ‘cultures’ get in the way of development, existing disciplines get in the way of interdisciplinarity. The Advisory Council for Science and Technology Policy (in The Hague) bluntly asserted that among ‘the bottlenecks’ hindering the growth of cross-disciplinary research is ‘[c]ultural differences and differences of approach between disciplines’ (AWT 2003: 6). Whereas the knowledge manager would manage cultural change to get reality in line with the vision, in the current situation the researcher must sustain those other presences (the disciplines) without which the interpolated nature of the new activity would not be visible. What in the ordinary process of scholarship and research helps keep the practice of interdisciplinarity separate from its parent disciplines is the hybrid seen as the outcome of transactions.9
Yet why must interdisciplinarity be seen to be on everyone's agenda these days? Why its new visibility? Why is it explicitly taken up by all the UK Research Councils (see DTI 2001a; 2001b)? Why does it seem like a new rationale for re-grouping departments or re-conceptualizing teaching? Re-grouping, especially in the natural sciences, is routine, but today re-grouping is not enough; interdisciplinary effort must also be stimulated. ‘To initiate and support cross-disciplinary and multi-disciplinary initiatives’ is a stated objective of the Research Councils UK (DTI 2001b: 1.33 (vi)).
To what microscopic degree? The Advisory Council for Science and Technology Policy suggests that the Netherlands government intervene by funding more interaction, building up networks, developing comprehensive multi-area research questions, and identifying thematic areas where multidisciplinarity will flourish. It adds suggestions as to how to manage this through secondment and selective funding. But these are rather more than suggestions. Under the heading ‘Several alternatives for developing more comprehensive research questions’, we find ‘Incorporate compulsory mechanisms into the funding conditions that will ensure that comprehensive research questions are actually formulated and maintained’ (AWT 2003: 29, emphasis mine). The situation will be managed through administrative measures (policy workers should be trained ‘to develop comprehensive research questions’, AWT 2003: 29) and the scope of inquiry will be driven by the desire for the ensuing research to be multidisciplinary. ‘Themes’ direct the researcher to problem-orientated types of questions that then become ‘comprehensive’ by virtue of the many disciplines brought to bear.10 It is then possible to monitor whether questions really are comprehensive.
‘[B]y now of course the popular maxim is that “all good social science is interdisciplinary” ’ (Cohen-Cole 2003: 149).11 This comes from a survey of social science carried out in the States; the date is 1959. Another moment of micro-management: we have been here before! An important standard for evaluating the quality of a piece of a work being the extent of its interdisciplinarity,12 in the 1950s government and private organizations alike pushed social scientists to become more interdisciplinary. ‘Publications that … championed the maintenance of disciplinary integrity were swamped by the calls for interdisciplinary collaboration, integration, and unification’ (Cohen-Cole 2003: 150).13 If this was a moment when social science was legitimating itself, interdisciplinary explanations of society displayed the useful knowledge that social science could bring to bear on pressing problems in the post-war world. Above all there was a belief in the link between the integration of society and the integration of knowledge. ‘Those who shaped their research with an interdisciplinary “problem focus” and had broad theoretical questions common to all the social sciences were avant-garde; “clinging” to standardized fields was “being traditionalist” ’ (Cohen-Cole 2003: 180). It was no accident that Galison's depiction of the disciplines trading with one another (‘the shirtsleeves’) evokes this period. The new knowledge had to be managed.
Before this period, interdisciplinarity was unlikely to have been labelled as such, so could not be the sign of integrative potential that it became. However, managing integration across disciplines has its antecedents. We might even think of T.H. Huxley in such a light (Beer 1996: 204).14 In the 1860s Huxley wrote of the influence that Darwin's Origin of species had had on him. He prefaced his exposition of Darwin's work by remarking that it was not an easy book to read, more like ‘an “intellectual pemmican”– a mass of facts crushed and pounded into shape, rather than held together by the ordinary medium of an obvious logical bond’ (1906a : 301). That difficulty gives him opportunity, he said, to occupy the humble ‘office of an interpreter’ (1906a : 301), a role facilitated by his own mix of interests.
Interconnections are the subject of Huxley's work. His own combination of expertises – zoology, botany, general biology, and the burgeoning anthropology of the time – was not trumpeted as interdisciplinary, but he was trumpeting the way in which everything organic and inorganic was related (1906b ). That organic matter itself returns at the end of life to the inorganic ‘is true of every living form. From the lowest plant to the highest animal – including man himself … the difference between the highest and the lowest being simply in the complexity of the developmental changes, the variety of structural forms, and the diversity of the physiological functions’ (1906b : 162).
For Darwin and Huxley, the fact that everything can be related also posits a universe in which everything can be known (Beer 1996: 304). This drive took a particular form in the search for evidence of undivided human ancestry;15 the notion of ancestry was already pared down to a lineal connection (Beer 1983). The search was for one distant progenitor, ‘the common parent’. (Two would, after all, spoil the genealogy, constituting a problem neatly solved by ideas about intrabreeding, self-reproducing species which produce two parents of one kind.) Darwin, Beer writes, does away with the sexual pair as an initiating origin: ‘[T]he originary parental dyad is figured as the one, sexually undifferentiated – and irretrievable: “the single progenitor” ’ (1996: 29). These assumptions posed research questions.
However, the anthropologists who were writing in the 1870s or thereabouts then had a problem to manage – the question of affinity in marriage. The facts of human procreation, including the differentiation of male from female parent, posed a problem as to what the originary form of human organization might be (Stocking 1987: 168).16 Doctrines of primitive promiscuity and (later) exogamy were part of a sequence of solutions ranging from the postulate that sexual differentiation made no difference to the homogeneous nature of social organization to assertions that it served the perpetuation of unified groups.
It must be obvious that I have ‘managed’ this account to produce a particular outcome. None the less, mapping a contrast – between undivided pasts that look to multiple futures and undivided futures that reconcile multiple pasts – onto the two models of knowledge creation enables me to cast interdisciplinarity in a particular light. It allows me to specify the different representations of growth implied, and how the models are intertwined.
For interdisciplinarity has it both ways. It can offer diversity as at once foundational to an enterprise and as innovative exploration. To hope future combinations will unify what were once distinctly diverse or divided origins – multiple disciplines – speaks of a management model of knowledge creation.17 A specifiable outcome is sought. To hope to the contrary that new combinations will divide and proliferate what had once been a union or origin in common – in other words create new disciplinary possibilities – is closer to a research model. Here outcomes are multiple, indeterminate. In the first, diversity is pressed into problem-solving; in the second, problem-solving generates more diversity.
Is that all? If there is a sense of disappointment here, is it that the formula seems to exhaust everything one might wish to do?
To the gift and the Indian gift in its different forms (Laidlaw 2002; Parry 1986) let me add another, the Fijian gift. There is a moment in the course of Fijian gift-giving when the givers’ side ‘subjects itself to the gift-receivers’ evaluation, and quietly hopes that the other side will respond positively’ (Miyazaki 2004: 7). Motionless, the givers’ spokesman holds the object in front of him until a recipient steps forward and takes it. ‘In this moment of hope, the gift givers place in abeyance their own agency, or capacity to create effects in the world’ (Miyazaki 2004: 7). There is hesitation; ritual participants regard gift-giving as risky, always seeing in the act its possible failure (Miyazaki 2004: 100). What in other social contexts may be an evaluation that takes place after the event, or is not visible until a return gift is made, is here brought forward into the act of hand-over.
So what is this sudden leap of mine into a quintessentially anthropological account? Partly, after all these generalities, to engage with the relief of the specific; partly to convey what it feels like being at home in one's discipline. What is quintessentially anthropological is not that it is a Pacific island or that in this account from the mid-1990s the object being handed over, a whale's tooth, has a long cultural history. Rather, it is that the evocation of the Fijian gift puts this gift alongside many others, summoning a lineage of analyses that no one but an anthropologist would rehearse. The ethnographer goes on to do something with the iconography of gift-giving not quite like anything before. Its antecedents in anthropological writings give the Fijian gift its newness.
Once the recipients accept the gift they may immediately deny the importance of gift-giving among people and offer the item to God. At the moment at which the giver's hope is fulfilled by the recipients, it is replaced by a second hope that God's blessing will fall on everyone. It is this replication that Miyazaki (2004) takes as his anthropological problematic. He goes back to Toren's (1988) original insight into the way that ‘Fijian’ and ‘Christian’ ritual work together, not as different forms but (in his words) as versions of a single form unfolding in time (Miyazaki 2004: 99).18 This includes the manner in which people cease to emphasize their own actions and deliberately look to others for their response. Fijian participants experience the fulfilment of their hope as the capacity repeatedly to place their own agency in abeyance: ‘[T]he hope produced in this process surface[s] as the replication of a hope … fulfilled’ (Miyazaki 2004: 106). In short, hope recurs. Miyazaki sees a parallel in the anthropologist's hope of an adequate analysis and the possibility of fresh knowledge.19
So what has this got to do with anything? Well, it has everything to do with creating knowledge, with the difference between management and research models, with origins and outcomes, and with interdiscplinarity and the discipline of social anthropology. The Fijian gift contains an unexpected significance when the anthropologist calls that hesitancy a moment of evaluation.
Evaluation is simultaneously a management tool and a research tool. Risk assessment is all about evaluating, measuring, judging possible outcomes, and is, indeed, second nature to systems of audit and accountability. So evaluation is often taken as a crucial step in the chain of scrutiny that leads from fact-finding to decision-making. However, evaluation is equally central to research processes that choose what shall be kept and what discarded or, in other words, that sort the poor data from the rich. That kind of scrutiny is second nature to the researcher or, more accurately, entails deploying a primary ‘nature’, the researcher's disciplinary identity. Disciplines offer a powerful framework for evaluation – criticism.
Disciplinary criticism has affinities with, but is not the same as, self-evaluation for purposes of better management. On the one hand, I am sure that managers (and researchers acting as managers) are told to be critical of themselves as well as of others in order to improve things. The point is that the desired outcome is already specified in the goals of the organization, and goals work best when everyone agrees on them: ‘[M]anagement activities are inseparable from goal setting’ (Reinhardt 2002: 195). On the other hand, researchers (and managers acting as researchers) criticize retrospectively, in relation to the canons of their discipline. A discipline is a body of data, a set of methods, a field of problematics;20 it is also a bundle of yardsticks, that is, criteria for evaluating products and maintaining standards. Knowing that the canons may be constantly changing and that outcomes are uncertain can be taken as a sign of life as much as the reverse. Disciplines live in the prospect of their own renewal, and often do not care too much the form that will take.21 It follows that here there is no desired outcome – only the hope that there will be one. I can criticize Miyazaki's concept of the gift, but there is no ideal version for which to strive. Rather, the aim of criticism in research is to re-multiply, re-divide, the outcomes of any one particular argument.
Criticism bifurcates; it makes a single account multiple again. Unitary in its argumentative focus, the lineage (of writings on the gift) summoned by putting Mauss, Gregory, Parry, Laidlaw – and Sykes (2005) – alongside Miyazaki speaks to diverse conclusions continuing to fuel debate. More emphatically, disciplines look to disagreement as points of growth. In fact disagreement serves to overcome one of the problems that Miyazaki sees in constructing an anthropological account of the Fijian gift, that is, a problem in the kind of knowledge that the anthropologist might wish to make out of it. He observes the temporal orientation of hope to a future that the anthropologist can only deal with through retrospective description by putting hope back into the past. The possibility of describing hope's prospective momentum is drowned in the knowledge of whether it was fulfilled (Miyazaki 2004: 8). This enacts a kind of foreclosure. By comparison, the researchers’ (academics’, scholars’) hope, that there will be an outcome to their labours, is given impetus through practices of disagreement. In looking to colleagues for criticism, they look for life. For the instruments of self-renewal, the papers they write and the books they generate, allow disagreement to remain unclosed. The disagreement, the opening out to further futures, can be left just as that.
This points to one way in which Power's agenda for uncertainty is already built into disciplinary practice. However, there are some interestingly new difficulties in our path. What makes things difficult these days is precisely the striving after renewal. It is not that anyone outside them is particularly bothered by the fate of disciplines but rather that the goal for replenishment is ‘knowledge’ in communicable form, as information or evidence (as in evidence-based policy) that can be put to use as the driving ingredient of the knowledge economy. The shirtsleeves knowledge pressed into the service of problem-solving is the image to which I keep returning. There is pressure for disciplines to seek a kind of instant renewal by communicating with other disciplines. Here the tool becomes also a sign (Riles 2003). Interdisciplinarity that shows an openness to different fields is also seized on as a sign of a willingness to subordinate disciplinary interests to the finding of common solutions.
It is probably fair to suggest that interdisciplinary collaborations work best not as tools (means) in research, but as representations (signs) of desired ends in knowledge management. These social practices differentiate between two kinds of experts. In a research model they will be using their expertise instrumentally as Galison's scientists imagined themselves. While everything they do has its appropriate disciplinary origin, it will be just bits and pieces that turn out to be tradable and useful. Under a management regime, however, it is much more likely that their expertise will have representational status; they will be called on as though they represented their discipline as a whole. So sitting around a table at a policy or ethics forum, experts from diverse disciplines will speak as representatives of their discipline; with no other anthropologists present, it becomes possible to give ‘an anthropologist’s’ view. Plenty of challengers, perhaps, but exit the critic. A multidisciplinary group can of course collaborate with little interest in the backgrounds of one another's contributions; diversifying points of view is not the same as dealing with disciplinary difference. However, even if an interest in disciplinary difference is the hallmark of interdisciplinary collaboration, where representatives do indeed reflect on the contributions they are making, the critic may still be hard to discern. Why? Perhaps we should return to the power of interdisciplinarity to proliferate outcomes and origins alike.
Now it is not the critic alone who becomes shy in the face of interdisciplinarity; so too does the managerial evaluator. This has been the subject of comment, as I note first.
Something very interesting happens to evaluation. As indicated elsewhere (Strathern 2004a), the sign becomes a yardstick; interdisciplinarity evoked as a measure of innovation obscures attempts to apply evaluation procedures to itself. The result is ‘the lack of available criteria to assess interdisciplinary work on its own terms’ (Mansilla & Gardner 2003: 1). Evaluating interdisciplinary endeavour is not the same as evaluating the degree to which a problem is solved or benefit created. Asked about the outcomes of the collaboration itself, researchers rely on indirect indicators such as publications, numbers of patents, or whether the interdisciplinary team will apply for funding a second time round. ‘Measures that directly address epistemic dimensions of interdisciplinary work (e.g. explanatory power, aesthetic appeal, comprehensiveness) [are] rarer and less well articulated’ (Mansilla & Gardner 2003: 1-2).22 In the management view, then, interdisciplinarity runs into some of the problems that bedevil attempts to justify government support for the creative arts in terms of contribution to public well-being (Selwood 2002). What impact do cultural programmes have? How can evidence of their impact be captured within the frameworks of specific projects? Attempts at qualitative assessment have tended to focus on the directly observable, for example surveying participants’ expressed satisfaction. Asking ‘can “culture” [as in cultural programmes] be shown to have an “impact” [on people's lives]?’ is like asking if interdisciplinarity can demonstrate epistemic effect. How does one show that some difference has been achieved?
This is our management model (of the creation of knowledge). Among the reasons why the demonstration is so difficult is perhaps the way the single outcome, the integrated collaboration, is impossible to measure against its own diverse origins. There is little against which to evaluate the effectiveness of interaction as such when the whole point was that there was no pre-existing relationship, only diverse starting points, and interchange was always a hope for the future.
Can we reverse the sequence and suggest that, in contexts of interdisciplinary research, criticism becomes obscured for the opposite reason? Can we assert that the expectation of multiplying outcomes, of keeping the ‘inter’ in interdisciplinarity, compromises the ability to criticize? Constructing disciplines as entities with singular origins makes criticism possible, but it cannot be criticism of the interdisciplinary effort itself. Instead it seems that interdisciplinary effort would discard those origins. I re-engage the point that openness to different fields indicates willingness to subordinate disciplinary interests, but apply it now to the research model. The single starting point, the agreed-upon canon, can be no measure of the combinations and cross-overs that point to future growth. Impatience is the usual reaction. Disciplines simply ‘get in the way’, as traditional cultures everywhere are held up as impediments.23 Barriers seem obsolete.
However, collaboration and criticism are intertwined in ways more complex than sketched here.
Interdisciplinary collaborations seem to promise innovation and creativity by means other than criticism. Instead of generating disagreement and multiplying future possibilities by informed comment from within, interdisciplinary conversations hold out the hope of new sources of synergy. Hope is still there. Indeed hope regenerates exactly as Miyazaki argues it does for those Fijians whose ancestral land is now the city of Suva. The people of ‘new Suva’ (Suvavou) make – indeed must make – constant return to the issues of land compensation. Repeating their efforts, each new effort requires them putting to one side their knowledge about what happened the previous time. As the ethnographer says: ‘How have Suvavou people kept their hope alive for generation after generation when their knowledge has continued to fail them?’ (2004: 3, my emphasis). It does not matter how successful previous attempts have been, institutional knowledge is constantly disappearing and people return to the same points of departure. I gestured towards this in recalling the post-war momentum for interdisciplinary research and the mix of management and research models to which anthropologists have (unwittingly as managers, eagerly as researchers) returned in almost identical terms today.
Yet let us look a bit more closely at the researcher's faculty for criticism. It is a social faculty. A further element in the Fijian encounter prompts the thought that, in that moment of hesitation as the gift is given but not received, when agency is suspended, could we not say that the giver desires to be divided from the receiver and, thereby, not to be assimilated but to be acknowledged as a separate social person? Now could we not also say that the capacity that the division (re-)enacts is the very capacity to separate ourselves not from an other but in the first place from ourselves? In any event, this is a social faculty before it is anything else.24 Anthropologists divide themselves off from anthropologists and multiply their positions precisely because they have common origins. The same could be said of the whole company of disciplines that make up academia. The community in my title could refer to the collectivity (so to speak) either of one or of many disciplines.25
In a sense, disciplinary divisions could not be more different from divergences in points of view. Criticism exploits these divisions in a very different way from the kind of collaboration that would subordinate them to the task in hand. First, divisions are evidently at work in the way that disciplines propagate; they breed through cleavage, whether by the rational calculation that old categories no longer hold things together or through acrimonious wranglings and rifts. Here division operates as a mode of multiplication that generates new forms. Second, divisions can be the impetus to colonization. Each may see something of value in the other but wish to appropriate it for his or her own agenda: that is, it already no longer belongs to the other person alone, as my rendering of the Fijian gift already takes it away from its location in the ethnographer's account. Third, in the kinds of twenty-first-century exhortations to interdisciplinarity that we have been considering, division becomes a sign of failure to communicate, of failure to create a wider community either incorporating the public or with other disciplines. Here disciplines are accused of failing to cross the divide between esoteric and common knowledge. A frequent rhetorical elision in governmental and other public statements is that between dealing with materials in an interdisciplinary way and being able to communicate to anyone (stakeholders).26
However, the divisions that a critic envisages are none of these. Rather, they are (seen to be) created in the course of interaction itself.27 Recall the recognition space and Weiner's admonition that anthropologists take as their subject matter the differentiating activity that emerges from inhabiting ‘one world’ in which categories such as indigenous and non-indigenous are clearly the outcomes of encounters. If, as Viveiros de Castro (2003: 9) says, the procedures characterizing anthropological investigation are conceptually of the same kind as those to be investigated, then what turns out to differentiate people are the radically distinct problems that they think they have.28 By the same token, the critic would have a different relation to the discourse under scrutiny than the proponent of it; united by an interest in a particular work, critic and proponent are not simply differentiated by problems they conceive but, in their interdependency, are specifically divided from each other by these problems.29
Whether within persons or between them, the impetus to divide ourselves from ourselves is a social one. Those second selves emerge as others and this movement is recapitulated in the yielding of one's own agency to the one who is now othered. This gives a further strand to the critic.
When the Fijian gift-givers’ spokesman fell silent, he placed their agency in the hands of the recipients, who would reveal its effectiveness. So ‘they [Suvavou people] experienced the fulfilment of their hope as the capacity repeatedly to place their own agency in abeyance’ (Miyazaki 2004: 106). Surrender before appropriation; here is hope for engagement. Does something akin to such hope feed into the interactions of critics? Could we then think of the critic as someone whose willing suspension of agency, a division of self from self, allows him or her to be captured by someone else's work? Critics find themselves drawn – precisely by their own interest – into other people's agendas.30 Engagement does not proliferate or multiply; it does not to look to standards and requires no evaluation. To argue with an idea is to be captured by it. In this kind of engagement, one can be captured more than once.
This is where I see hope for interdisciplinary endeavour. In this era of expertise, the very idea of traversing disciplinary boundaries uses institutional terms to speak of possibilities that lie in being captured by another's concerns. For it also makes visible the interest of those who are identifiably ‘other’ to the discipline in hand.31 The anthropologist has been here before too; engaging one's interest in ‘other people’s’ agenda is crucial to the enterprise of fieldwork-based research. In interdisciplinary work I see a replication of the anthropologist's hope in the ethnographic moment (cf. Miyazaki & Riles 2005: 328).32 Each interdisciplinary encounter points to a fresh encounter in a terrain only uncertainly mapped. It is the obviousness of the uncertainty that is important here. The constant shortfall of knowledge that never gets beyond recognition spaces holds out the hope that one can always re-engage.
This is why I have kept with this one account (on Fiji). Whether in order to agree or disagree, the possibility of re-engagement is ipso facto‘critical’. A new problem is conceived. Exactly because anthropologists never exhaust the information they collect, ethnography has the remarkable capacity to outlive the particular uses made of it.
There is one caveat. If we think of future engagement predicated upon the hope for it, our two models of knowledge creation will raise their heads. It is a short step to asking how best to manage it or to build such hope into research protocols. I do not think that we should necessarily do either. Re-engagement needs to be re-engagement, a matter for the future that the present should leave undefined. Yet how can one not plan ahead? How can one not make a virtue out of engagement? One way might be by imagining that one has to protect (cherish) it as though it were knowledge to be protected from itself.33 Presumably knowledge best not acted upon is best put into abeyance. So perhaps the answer is to not treat these observations as knowledge at all. Perhaps rather than specifying the faculty for hope, that is, hope for engagement, we might just want to say that engagement is a faculty. This would make everything very simple. For nothing more is implied beyond each act of engagement, insofar as each contains the possibility of re-engagement without specifying what it would be. The future remains indeterminate, and it is simply the possibility that is (already) enacted in the present; one can meet again, come back to the data, revisit the analysis. We just need to keep that as routine as possible. Relations(hips) secure the promise of re-engagement.
So let me make another return to land claims, Aboriginal this time, and to a clear interdisciplinary moment. A remarkable piece of ethnographic elucidation is to be found in a lawyer's commentary on anthropological advocacy (Edmond 2004). Edmond is concerned to show anthropologists that once they are in court they have ceded interpretative space. This is largely an unwilling or unacknowledged surrender; the point is that the roles of expert witness and advocate are procedural, and not open to anthropologists to define.34 Rather, they will be assigned certain duties of care that the court may in effect lay open to dispute by the parties concerned. For litigants will seek to deploy the kind of evidence that the anthropologist brings in terms of (always disputable) points of law. The anthropologist, of course, may well look upon the court transcript as though it were information collected in an interview that is subject to the canons of scholarly verification; the anthropologist may, in short, treat it as though it were a research finding (after Riles 2003; 2004). Although this is often the basis of their complaints, judges in truth do not distort anthropological knowledge, for it is not placed before the courts as knowledge in which they need to have any informed interest. Edmond calls this the legal colonization of anthropology.35
Yet there is a note of interdisciplinary hope here, couched in the invitation to re-engage proffered after his recognition that anthropology is subject to that legal colonization. Edmond quotes an Australian jurist who claimed ‘the last thing I would wish to encourage in humanist witnesses is obsequiousness towards lawyers … There are good social reasons for treating the legal system's normative and adjudicatory authority with respect, but none for endowing it with intellectual authority’ (2004: 221). Anthropologists have an important obligation publicly to criticize legal processes if they feel that their work is misunderstood or that claimants are treated unfairly: ‘[S]ustained and consolidated criticism of legal rules, procedures and doctrines, as well as findings, may find fertile ground among judges and other attentive publics’ (Edmond 2004: 221, emphasis omitted). The law courts are not, however, the place for it; a public space of sorts (the agora) is.
In conclusion, the lawyer asks how anthropologists might respond to the terms of their colonization. One response would be to lay out what the claim to intellectual authority would look like. The first step might well be showing that the issues at stake have been argued over, even fought over, by anthropologists. While there are huge areas of common ground and shared assumptions in their discipline, anthropologists’ engagement with one another in profound disagreement is a crucial part of their work. (More pointedly, agreement and disagreement must go hand in hand for either to have any intellectual purchase.) Although anthropologists have always done this, these days perhaps call for them to stop being defensive about it and make of it a virtue. For they may best validate the role of public critic by being known as critics of themselves. So, sustained criticism, yes. Consolidated criticism, I’m not so sure. However, if anthropologists know themselves as a community of anything, a community of critics is as good a rubric as any.
The focus on divisions was prompted by a session convened by Bill Maurer at the 2003 American Anthropological Association meetings in Chicago, and a discussion with Eduardo Viveiros de Castro at the 2003 Association of Social Anthropologists Decennial in Manchester. For a fruitful interchange, I must thank Rena Lederman, whose book Anthropology among the disciplines is in preparation. Hirokazu Miyazaki and Annelise Riles's influence will be apparent, but here I should thank them for their intellectual hospitality. Christina Toren's perspective has, as usual, been invaluable, as have Eric Hirsch's and James Weiner's comments.
Among the benefits from interdisciplinary conversations at the University of Cambridge, I mention those at CRASSH (Centre for Research in the Arts, Social Sciences, and Humanities) and the inspiration of Ludmilla Jordanova. Jeanette Edwards, Sarah Franklin, James Leach, and Andrea Stöckl have opened up a dialogue about expertises that I do no more than note. Finally, this is a contribution to an ongoing project, ‘Interdisciplinarity and Society’, undertaken with Andrew Barry and Georgina Born (ESRC grant RES-151-25-00042) as part of the ESRC Science in Society Programme led by Steve Rayner.
Weiner refers to Noel Pearson's ‘The concept of native title at common law’, from the Australian Humanities Review (1997), as the source of the phrase. He suggests that the anthropologist's cultural account is in effect a recognition space: accounts of those portions of a culture for which a fair translation can be achieved in another language.
Thus Australian lawyers do not have to concern themselves with the substance of Aboriginal law; they just give Aboriginal lawyers the space with which to deal with what only they know.
This is what Wheatley (2002: 2) argues when she suggests that KM must learn from the organizational failures of the past. It would be a mistake to imagine knowledge as an industrial product; on the contrary, knowledge must be managed as a person is, for (we are told) it is created by persons in engagement with the world and thus with one another. ‘Knowledge creation’ is the phrase used.
‘How does knowledge flow within this organization? … How do you develop the right culture for knowledge sharing?’ (Ives, Torrey & Gordon 2002: 109). Du Gay comments that ‘contemporary managerial discourse approximates to a system of charismatic authority’ (2000: 70).
This is true above all when managers introduce into training regimes deliberate devices to unsettle people so that they learn to deal with moving targets (Martin 1992) and thrive on chaos (du Gay 2000: 71).
In the latter, knowledge is seen as integral to the capacity to act, as a form of information ready to apply to decisions and actions (Ives, Torrey & Gordon 2002: 101).
Today researchers should be their own managers (Fournier & Munro 2004) or dedicated managers are deployed to stimulate research activity where there might have been none before (Strathern 2004b). Behind these developments is the politics of the ‘new public management’ and ‘entrepreneurial governance’ of the last two decades (du Gay 2000: 5).
Inter(or multi)disciplinarity means specific things to specific disciplines across the arts, humanities, and social sciences, not to speak of the natural sciences or the clinical domains (Latimer in press). Rabinow (2003: 5) contrasts the way the funding and facilities regimes necessary to natural science can act as centralized ‘policing’ mechanisms, by contrast with ‘interpretative communities’ among the humanities, where different types of argument can prevail without sanction.
Parties to transactions distinguish themselves from each other before obtaining what of each other's resources they value. Alternative models include (1) knowledge-sharing, as appears in the KM literature, and (2) learning from other disciplines not in order to produce a hybrid but to enhance one's sense of one's own.
‘The Council therefore advocates an approach that facilitates the development of integrated questions. The parties involved should seek each other out at an early stage and formulate the key question together’ (AWT 2003: 28).
I am most grateful to Jamie Cohen-Cole (Princeton University) for permission to cite from his 2003 Ph.D. dissertation. The points in this paragraph come from an unpublished draft.
The term had such cachet that Ford Foundation officials ‘used it in evaluating grant proposals as they would have used other positive descriptors such as “well conceived” ’ (Cohen-Cole 2003: 167, emphasis removed).
Cohen-Cole (2003) archly notes that interaction was always discussed in the positive terms of cross-fertilization, never cross-sterilization.
Osborne (2004) gives modern-day interpreters a special place in his typology of ‘mediators’, a role he regards as diagnostic of the knowledge economy.
The ‘connections’ (missing link) they pursued entailed the investigator trying to make connections between apparently isolated (unique) pieces of data. In the lecture from which I have been quoting, Huxley makes the case for ontogeny, arguing that every living thing begins its existence with the same ‘primitive form’, the egg, which signals the true ‘unity of organization’ of the animal kingdom (1906b : 166).
Some came to believe that everything lay in finding the origin of a single custom (marriage) (Stocking 1987: 167). Similar divergent starting points continue to fuel analysis. McKinnon (2001: 288, 298) describes how Morgan sought to distinguish a clear line of succession to the nuclear family from the promiscuous relations he imaged for primeval man, while Lévi-Strauss took the family as the unit that had to differentiate itself through exchange and create multiple possibilities for human organization.
‘Undivided outcomes’ means not that different voices in a team are merged but that orientation to a joint project (‘problem-solving’) takes precedence. Clearly there are ‘communal’ and other perceptions of collective ownership, in the case of academic knowledge often linked to ideas about the public domain, that fly in the face of the notions described here.
He follows Toren's argument that the Fijian church is not simply a ‘local’ rendering of Christianity but institutes ritual processes through which Fijian chiefship (‘the land’) is also made visible. Toren explores parallels in Fijian meals, including the ways in which chiefs dispense kava, and the ubiquitous imagery of ‘the last supper’.
It is important that he cites earlier discussions on the abeyance of agency, anthropological antecedents including Battaglia (ambiguation), Brenneis (indirection), Marcus (rhetorical manipulation), Herzfeld (structural nostalgia), and Keane (avoidance) (Miyazaki 2004: 105).
Devons and Gluckman's formula remains to be bettered: ‘Different disciplines may study the same events, and even some of the same regularities in those events, but they look for different kinds of interdependencies between the regularities, i.e., for different kinds of relations’ (1964: 160).
‘Anthropology is no longer a singular discipline, if it ever was, but rather a multiplicity of practices engaged in a wide variety of social contexts’ (Moore 1996: 1); diversity belongs to the future, and to the ‘whole series of new questions’ that can be anticipated.
Interdisciplinary work has been defined as purposeful means to a cognitive or practical goal (understanding, solving a problem), with the stipulation that ‘disciplinary lenses be integrated in mutually informative networks of relationships rather than simply juxtaposed’ (Mansilla & Gardner 2003: 2; cf. Strathern 2004b).
Including the ‘culture’ of organizations, which gets in the way of new practices of knowledge-sharing: ‘This cultural issue is … the main obstacle to implementing knowledge management’ (Ives, Torrey & Gordon 2002: 99).
A social faculty with several specific histories, one being the splitting of the self that Hoskin (1995) unfolds in terms of the particular modern (Euro-American) development of the self-examining self lying at the heart first of perspectivalism and then of regimes of accountability. ‘Intersubjectivity’ (e.g. Toren 1999: 2) also has its own, somewhat divergent, history.
Community is not a word I use lightly. An extreme example is the collectivity of authors in Closed systems and open minds (see Devons & Gluckman 1964), who happen to take as their common theme the limits of disciplinary naïvety. The fact that the contributors to this volume were associated with one department (the then Department of Social Anthropology and Sociology at Manchester, its intellectual links with economics – Devons was an economist – and political science prompting the theme), and presented many of the papers together at the 1957 Association of Social Anthropologists meetings in Edinburgh, does not in itself create a community. I see that community in the unremitting criticism with which they approached one another's work, each argument in turn being thrashed out at seminars, and then held out to public view. The conclusion evaluates the chapters one by one and in both positive and negative vein, at times mercilessly detailing errors and difficulties for scrutiny. Communities are not about being nice.
Interdisciplinarity is often taken as the first move towards general intelligibility (Strathern 2004a: 68-86). In a managerialist regime, academics must not be seen as having an ultimate purpose different from the public's at large, no more than (in the new wisdom) ‘science’ should separate itself from ‘society’. The overall onus is to communicate. The public will bring different viewpoints, but the communication can entertain a common purpose. So at the heart of the appeal to overcome division we find the re-creation of a division between the public (lay persons) and whatever it is (from experts) that they are supposed to assimilate. This includes a cynical parallel division between what is produced for colleagues and what is produced for the media.
Not that there is any conclusion that could not be pressed into the service of management (here, see Newell, Robertson, Scarborough & Swan 2002: 119-24 passim on ‘managing communities of practice’, where ‘the community approach sees it [the management of knowledge work] as the product of social interaction’.
This evokes Huxley's arguments (1906b ), later elaborated in terms of comparative anatomy, that the ‘same’ creatures inhabit ‘different’ bodies at different locations, with the forces bearing down on them having different effects.
In the light of the earlier discussion about managerial approaches to problems, I note that Viveiros de Castro (2003) identifies as a common orientation in anthropology the idea that there are generic problems in the world to which ‘cultures’ are specific solutions. By contrast, he invites us to imagine a generic equivalence in problem-solving capacity being applied to distinct and specific problems.
I read Munro's (2004) account of deferral between submission (as of a research project) and evaluation (an organization giving permission for research to proceed) as an unwilling surrender. A more deliberative submission, not caring to know, emerges from Konrad's (2003) account of predictive genetic testing; people suspended between diagnosis and the manifestation of symptoms have to deal with how much they and their relatives ‘know’.
‘Anthropologists’ success in multidisciplinary [interdisciplinary also implied] settings may have positive ramifications for the continued vigour of anthropology as a discipline’ (Gooberman-Hill 2003: 32).
They come to a conclusion very similar to mine, although theirs is couched in an argument about expert knowledge rather than interdisciplinarity.
Following Konrad, this is closer to ‘not caring’ to know than asserting the right not to know. As Miyazaki says, ‘[H]ope cannot be argued for or explained; it can only be replicated’ (2004: 110).
That does not mean to say they will not be disputed by lawyers (Edmond himself gives an alternative view of the expert as a reasonable professional as against the expert as advocate).
His argument applies at least in the context of native title and heritage legislation in Australia, where ‘a hegemonic, though loosely bound, (legal) system [is] coming into contact with a less powerful and even more loosely affiliated set of knowledges and practices (anthropology) … Through its attempts to interpret and understand societies and their cultural practices, professional anthropology is now itself subject to appropriation and legal colonization’ (Edmond 2004: 220).
L’anthropologie sociale a l’habitude que le terrain se dérobe sous ses pieds. Les objets que l’on observe de loin se rapprochent brusquement, comme l’illustre l’exemple de l’économie des connaissances. L’auteur considère la place de l’anthropologie comme discipline dans un monde où la créativité devient une annexe à la productivité, où les collaborations interdisciplinaires sont un passage obligé vers l’innovation et où l’individu est évalué en fonction de ses compétences. On voit se poser un nouveau genre de problème : celui de cultiver l’esprit critique.
Marilyn Strathern is William Wyse Professor of Social Anthropology at the University of Cambridge.