Afterword: Brokerage as social practice

Brokers and anthropology

The analytical concept of the broker can be considered a genuine invention of anthropology, although the analytical perspectives expressed with this ideal type were similarly already put forward in older sociology work’s (e.g. Simmel’s 1950) theorem of the triad; cf. Bräuchler et al., introduction to this issue).

As is the case with many inventions, the analytical concept of the broker has been invented several times, namely in both economic and political anthropology. Its ‘natural’ home is in the economic world. Only there does it represent – at least in European languages – an emic term which is also used by social actors (Röschenthaler, this issue); in other areas, it is a metaphor like many analytical concepts are. In economic anthropology, the term broker was thus used early on (Hill, 1966, who developed it using the example of African agriculture). All subsequently derived and expanded uses of the term were also oriented towards this original concept of the economic broker. The paradigmatic situation is that of a bazaar economy, that is an ideal type of economy which functions according to market principles, but in which information is highly unequally distributed: ‘In the bazaar, information is poor, scarce, maldistributed, inefficiently communicated, and intensely valued’ (Geertz, 1978: 29). Brokers explore these information gaps and exploit them to their own advantage in order to bring sellers and buyers together. As no real economy corresponds to the ideal type of a pure market economy, they all contain elements of a bazaar economy and thus hold potential for brokers. Especially in the international context, the function of the economic broker also has an important cultural dimension; here, as Röschenthaler (this issue) shows, brokers are also cultural mediators, not least thanks to their language skills. All other forms of brokerage also – to varying degrees – include such functions of interpreting cultures and mediating between them.

Initially independently of the discussion in economic anthropology, brokers were described early on in political anthropology, but avant la lettre, so to speak. For a long time, the predominant perspective was one of mediation between an ‘inclusive’ and an ‘encapsulated’ society (or, to use another metaphor, between ‘above’ and ‘below’). The empirical background for this was initially the African and Middle Eastern colonial and post-colonial situation with the intercalary position of the colonial or tribal chief who – caught between the demands of the colonial administration and the expectations of his local community – found himself in a ‘predicament’ (Fallers, 1955; Gluckman et al., 1949; Salzman, 1974). Since the 1950s, these considerations have been taken out of the explicitly colonial context and extended to peasant societies and their relations to the state or city-based rule in many regions of the world (Cohen and Comaroff, 1976; Epple, this issue; Geertz, 1960; Mendras, 1976; Wolf, 1956). However, notions of a stratified form of social reality or social contexts of different scales – local community versus the state – were often retained. This hierarchising conception of social reality has been criticised by other authors (such as Kuper, 1970) as excessively deterministic. Instead, the literature has increasingly insisted on brokers having room for manoeuvre and creatively exploiting their intercalary position in their own interests (see Marquéz García, this issue).

These considerations were further developed and condensed by Boissevain (1974) using the example of Malta and, more generally, Mediterranean societies. For Boissevain, brokers act at the interface of different social formations, without there necessarily being a hierarchy between them. Boissevain proposed a definition that is still valid today: brokers control second-order resources, that is the access to holders of first-order resources (land, credit, labour, subsidies, study places, specialised knowledge . . .) – he calls the latter patrons. For Boissevain, the broker is an ‘entrepreneur’ who actively manipulates people and information: ‘A broker. . . controls second-order resources. . . (He/she) is a professional manipulator of people and information who brings about communication for profit’ (Boissevain, 1974: 148). This sets out the minimum analytical conditions for being able to refer to an activity as brokerage (and not as other forms of mediation). The broker has relationships in several directions, which they actively shape (they thus have agency); they pursue their own interests in doing so (they receive a ‘commission’ for his brokering activities), and they are positioned more or less equidistantly to the parties between whom they mediate. It is precisely the latter characteristic that makes him morally questionable for other actors, since the parties between whom they mediate doubt his solidarity and would often prefer to cut out the middleman. The ‘commission’ does not have to be in the form of money; the broker’s fundamental interest is in bridging a communication gap, but only on an ad-hoc basis and not in principle, as that would make him redundant. Finally, Boissevain points out that the broker can also always try to convert his relationship capital, that is, his second-order resources, into first-order resources at an appropriate moment, for example, into a job in an administrative body, a piece of land or a position in politics.

In the late 1990s, this concept of the broker, adopted by Boissevain, was used by anthropologists to analyse the dynamics of development policy. However, the phenomenon of development brokerage itself had already been addressed before, again avant la lettre. In a brilliant study, Gonzalez (1972) analysed an association of businessmen and members of the liberal professions which had the goal of promoting the development of the city of Santiago de los Caballeros in the Dominican Republic. The main activity of this association was to channel funds from USAID and other major donors such as the Ford Foundation or the World Bank to this city. Gonzalez describes the members of the association as brokers at the interface of two worlds. They were closely connected to the political elite of their country through family, social networks and friendship ties as well as business partnerships. When a political opportunity called for it, they did not hesitate to express anti-imperialist and anti-American opinions. At the same time, they all spoke English very well (most of them had studied at US universities). For foreign visitors, they stood out because of their lifestyle (and that of their spouses), which was very American in many respects – their dress, ways of speaking, sporting preferences, forms of sociability. ‘So Americans feel at home with them, and take this feeling of well-being as a sign of the honesty, intelligence, competence and political philosophy of this group’ (Gonzalez, 1972: 198).

These strategies were so successful that, in a sense, the American aid donors became hostages to this group of brokers, all the more so because their Dominican interlocutors very cleverly made it clear that other, non-US, organisations and institutions were also hoping to be allowed to support their association. In other words, we are confronted here with a case of international clientelism in which the distribution of power was at least ambiguous. If we assume that both parties derived some benefit from any relationship between the patrons and clients, it is difficult, as Gonzalez concludes, to specify exactly who was the patron and who was the client in this case. Furthermore, it is not clear what these two terms actually mean, because the patron can become just as dependent on the continuation of the relationship as the client. Similarly, in her study of an indigenous community in the mountains of Ecuador, Lentz (1988) shows how villagers were able to redefine and repurpose development aid projects carried out from outside, a situation that has also been analysed intensively in African contexts (Olivier de Sardan, 1988).

The term development broker was coined by Bierschenk et al. (2002) on the empirical basis of studies in West Africa to refer to social actors who are anchored in a local arena and who channel external resources, in the form of development aid, towards this local arena (cf. Knodel, in this issue). Development brokers thus act at the interface of institutions, policies and projects of development cooperation, on the one hand, and the people ‘targeted’ by them, on the other. Towards the former, they present themselves as spokespersons representing the local population and formulating their ‘needs’. They know which funding lines exist and how they can be rhetorically harmonised with these local needs. In contrast, vis-à-vis the local arenas, they position themselves as actors who have the relevant capital (knowledge, language, contacts) to mobilize development aid. The brokers aim to create a demand for their services and thus make themselves indispensable. They translate the discourses and the actions of the actors into terms that make sense to the partners at the other end of the chain. They play with their double affiliation and thus emphasise their social proximity to local actors in some situations, and focus on their mastery of foreign worlds that are unknown to these local actors in other situations. Development brokers are thus not passive recipients of aid; on the one hand, they actively seek projects for their local communities, while on the other hand they support development institutions with finding suitable target groups. Studies on development brokers (Knodel, this issue) as well as on brokerage in the field of migration (Walther, this issue), humanitarianism (Paulmann, 2016) and political activism (Merry, 2005) often overlap with NGO research, because a significant part of the activities of NGO leaders consists in brokering.

In recent times, the concept of the broker has experienced a certain boom, also outside of anthropology. It seems to fit recent societal developments, which are denoted by terms such as network society, multi-nodal security, multi-level governance, multiplicity of regulatory authorities, polycentric social systems, etc. All of these terms refer to social relations that are power-loaded (like all social situations) but do not have a clear hierarchy. Furthermore, according to Marxist political economists, the notion of the broker fits recent forms of rentier capitalism in which ‘brokers of capitalism’ play an important role (Laurens, 2015).

The empirical productiveness of the brokerage concept

The contributions to this issue show that the concept of brokerage can be applied innovatively in new fields such as migration (Walther), public health (Arambewela-Colley) and political activism (Bräuchler) in addition to economics (Röschenthaler), politics and administration (Marquéz García, Epple) and development (Knodel).

The analytical productiveness of the concept of brokerage lies, on the one hand, in the fact that it brings into empirical focus a group of social actors who are clearly distinguished from the groups between which they mediate. Moreover, it draws attention to the fact that these intermediaries pursue their own interests, in both or multiple directions. Such an analytical perspective thus de-moralises practices that are often, as Arambewela-Colley et al. in this issue show, pursued in highly moralised contexts. Thus, the concept of the broker has a strong alienating effect, as is inherent in all ‘strong’ analytical concepts. For the actors themselves, such a de-moralizing analytical lens on their activities is often difficult to accept. Development actors to whom we presented our reflections on development brokers in West Africa in the early 2000s often refused to relate the term to themselves. On the other hand, this alienation effect, as well as the economistic-activist image of personhood from which it is derived, in turn, creates its own blind spots in empirical research (Walther, this issue).

The concept of the broker as derived from Boissevain provides a checklist, as it were, for empirical research. The term invites anthropologists to explore the social profile and position, biographical development, educational and class background, lifestyle, knowledge and skills (not least linguistic skills) of the actors who are described as brokers as well as how they acquire these skills (often as learning-by-doing), their networks and contacts, interests, strategies and techniques. In doing so, an anthropologist should also be particularly interested in the emic terms for brokering and other forms of mediating as well as in their semiology (Epple, Röschenthaler, this issue; Hill, 1966). In the contributions to this issue, familiar but also new themes from the relevant literature can be found: for example, the entrepreneurial orientation of the broker who usually actively creates his role which distinguishes him from other types of intermediaries; the link between brokers in brokerage chains; the overlapping and multiplicity of roles. Roles also overlap in the sense that economic, political and cultural brokerage functions and capitals are often combined, either simultaneously or over the life course (Marquéz García, this issue). In African studies, the concept of straddling (of different spheres of society) has been used to describe such strategies of ‘big men’ (Médard, 1992). It is therefore often not so easy to distinguish empirically between first- and second-order resources. Going beyond Boissevain, the moral considerations and emotions of brokers are also of interest, as Epple and Walther show in this issue. These emotions and moral dilemmas which are also addressed in some NGO research, arise from the feeling of being caught ‘between a rock and a hard place’ (Igoe and Kelsall, 2005). They thereby recall the ‘predicament’ of the colonial chief (Fallers, 1955). Anthropologists pay particular attention to the semiology and symbolic language which expresses moral claims, demands for loyalty as well as the rejection of these demands or even full-scale rejection of taking on a broker role (e.g. through terms from the realm of the family or close friendship, Epple, this issue).

The term broker overlaps but also competes with other terms used to analyse mediation practices, such as translator – a term which, like broker, can be used in a narrow linguistic as well as in a metaphorical sense (Arambewela-Colley, this issue). Epple proposes two other overlapping and partially competing terms in her contribution: ‘street-level bureaucrats’ and ‘para-ethnologists’. There has been little reflection on the definition of these terms in relation to brokers so far. Street-level bureaucrats are state or para-state representatives who have a direct relationship with citizens and whose central function is to mediate between the claims of the state and those of their clients (Lipsky, 1980; see also Schaffer and Lamb, 1974). They can fulfil this role in very different ways, ranging from being an arm of the state to taking sides on behalf of clients against the state, sometimes in a covert manner (Andreetta, 2019). In this sense, African colonial chiefs were street-level bureaucrats (Marquéz García, this issue), as are the Slovakian NGO actors described by Walther (this issue) in the field of migration. In the Ethiopian situation of internal late colonialism, the street-level-bureaucrats-cum-brokers are not only subject to different political demands from different sides. They have to apply politics formulated by actors higher above, fulfil the expectations of their subordinates and see what their room for manoeuvre is, in a situation of de facto legal pluralism. In addition, the native street-level bureaucrats also belong to two different cultural worlds and maintain two quite different lifestyles, separated locally between the city and their home village – a ‘personal and political stretch between two worlds’ (Epple, this issue), which has also been described elsewhere (Arce and Long, 1993). This tension is not only an opportunity to pursue own interests, it also has an emotional price (see also Walther, this issue).

From another perspective, Epple’s southern Ethiopian bureaucrats are also para-ethnologists, a characteristic they share with many other interface officials (Beek and Bierschenk, 2020; Kolloch, 2021). In situations of high cultural diversity, bureaucrats, whether working directly for the state or in a mediated capacity, for example in an NGO on behalf of the state, develop their own ad hoc cultural theories to explain the behaviour of the clientele with which they deal in their professional everyday life. In this respect, they resemble anthropologists. Epple’s bureaucrats are even the ‘native’ para-ethnologist variant, insofar as these street-level bureaucrats bring knowledge about their ‘own culture’ into administrative contexts.

Towards a narrow concept of brokerage

The contributions to this issue show that the term broker, like all analytical terms, can be used in a narrower as well as in a broader sense. However, an extension of the broker concept, as it characterises part of the recent literature, is not unproblematic. If, for example, the above-mentioned analytical criteria – the broker’s equidistance from the groups between which they mediate and the activity of mediating in multiple directions – are disregarded, the boundaries between representation and mediation become blurred. The analytical advantages of the broker concept – namely the understanding of brokers as actors with their own interests who act at an equal distance to the groups between which they mediate – then tend to be lost. Actors are then referred to as brokers who no longer correspond to the classical model of the term and who should perhaps rather be considered, metaphorically speaking, as trade representatives or salespeople: for example, political activists who give local groups a voice vis-à-vis national and international publics in a language that is understood by these publics. A ‘partisan arbitrator’ (Faist, 2014: 45, cited by Bräuchler, this issue) is not a broker in the strict sense. Or the boundaries become blurred with related, overlapping terms such as translator; vernacularisation, after all, denotes an activity in only one direction. With this extensive use of the term, the concept of broker tends to become conflated with that of the active social actor per se, insofar as every social actor is, to a greater or lesser extent, always a mediator between different areas of social reality.

Of course, one can propose a broker concept that goes beyond Boissevain. But for this to be more than contingent definitional positing, an expanded notion of broker would also have to explain what it contributes to illuminating an empirical situation which an alternative, and possibly overlapping, notion fails to do. If the term does not help us to look at familiar things in a new light, it lacks this analytical added value. And if, for example, all translators are defined as brokers, one of the two terms is basically superfluous. Both sociolinguistics and translation studies established some time ago that linguistic acts can take on a new meaning in many ways – through transfer into another language or through reproduction in another context – without recourse to the broker concept (Snell-Hornby, 1988; Wadensjö, 1998).

Reading the contributions to this issue suggests that instead of conceptualising brokers as social types, we should speak of brokerage as a bundle of social practices or a social role. Every social actor is also a broker in certain situations (see also Bräuchler et al., introduction to this issue): professors who refer their students to further study programmes or employers, people who set their friends up with a date, the concierge who finds a flat for his acquaintance. All these actors practise a form of brokering, but often only incidentally. The normal case is one of multiplicity and overlapping roles. Brokering practices take up a more or less broad space in the overall structure of an actor’s social practices. They can also be more or less formalised; only very few people are full-time brokers with formalised training and a licence, such as real estate agents in Western societies. Only in such extreme cases can one strictly speak of someone ‘being’ a broker – in most cases it would be more accurate to say that someone ‘brokers’, among the many other things they do. While few social actors are fully-fledged brokers, many do brokerage.