Envisioning Power. Eric R. Wolf

of a giant monster such as Leviathan or Behemoth, or else as a machine that grows in capacity and ferocity by accumulating and generating more powers, more entities like itself. Yet it is best understood neither as an anthropomorphic force nor as a giant machine but as an aspect of all relations among people.

      I first encountered this formulation when I heard Norbert Elias lecture in the summer of 1940 at the Alien Detention Center at Huyton near Liverpool in England, where all male Austrian, German, and Italian citizens living within a certain range of London were interned by the British government, as the German armies overran France and an invasion of England seemed imminent. There I not only had my first lesson of sociology but also learned from Elias that “more or less fluctuating balances of power constitute an integral element of all human relations” (1971, 76–77; my translation). Elias likened the shift of power balances to a game: balances may change and produce gains for one set of partners (individuals, groups, or whole societies) and losses for another; a cumulative run of gains may eventually build up monopolies of power yet simultaneously generate efforts to test and disestablish these favored positions. Particular moves in these games can bring on violence and war; but violence and war, too, were to be thought of in relational terms as interdependent phenomena, and not as forms of anomic disorder.

      Thinking of power in relational terms, rather than as a concentrated “power-pack,” has the further advantage that it allows one to see power as an aspect of many kinds of relations. Power works differently in interpersonal relations, in institutional arenas, and on the level of whole societies. I have found it useful to distinguish among four modalities in how power is thus woven into social relations. One is the power of potency or capability that is seen to inhere in an individual. Power in this Nietzschean sense draws attention to how persons enter into a play of power, but it does not address what that play is about. A second kind of power is manifested in interactions and transactions among people and refers to the ability of an ego to impose its will in social action upon an alter (the Weberian view). Left unspecified is the nature of the arena in which these interactions go forward. A third modality is power that controls the contexts in which people exhibit their capabilities and interact with others. This sense calls attention to the instrumentalities through which individuals or groups direct or circumscribe the actions of others within determinate settings. I refer to this mode as tactical or organizational power.

      But there is still a fourth modality of power, which I want to focus on in the present inquiry: structural power. By this I mean the power manifest in relationships that not only operates within settings and domains but also organizes and orchestrates the settings themselves, and that specifies the direction and distribution of energy flows. In Marxian terms, this refers to the power to deploy and allocate social labor. It is also the modality of power addressed by Michel Foucault when he spoke of “governance,” to mean the exercise of “action upon action” (1984, 427–28). These relations of power constitute structural power. Marx addressed the structural relations of power between the class of capitalists and the class of workers, while Foucault was concerned rather with the structural relations that govern “consciousness.” I want to trace out the ways in which relations that command the economy and polity and those that shape ideation interact to render the world understandable and manageable.

      Ideas or systems of ideas do not, of course, float about in incorporeal space; they acquire substance through communication in discourse and performance. We therefore need to attend also to how ideas are communicated, from whom to whom and among whom. The term “communication”—generating, sending, and receiving messages—was in common usage in the 1950s (for example, Ruesch and Bateson 1951), but it yielded pride of place, after a brief reign, to “meaning.” It nevertheless remains a useful term, because it covers both messages expressed through human language and those transmitted nonverbally. Nonverbal communication embraces many modes through which messages can be conveyed. They can be transmitted through human gestures and bodily comportment, and also communicated iconically through displays of objects and representations.

      Both modes of communication provide vehicles to convey ideas, but messages have first to be cast into appropriate cultural and linguistic codes. To speak and understand a language, one needs access to its linguistic codes, so as to identify its phonemes and morphemes as well as the syntax through which these elements are formally combined. Similarly, to take part in a ritual, one needs to have a formal script of required acts, set out in the memory codes of participants or in the written instructions handed out to an expectant audience. Codes arrange the constituent elements of the message in particular ways, in order to convey which notion or notions are to be broadcast to an audience and how it should decode the messages heard. There would be no communication without codes, and to the extent that all social relations involve communication, they must also utilize codes and engage in coding and decoding. Thus, this concept of code and codes is applicable not only to language and formalized behavior such as ritual but to other facets of cultural life as well. One can speak, for example, of dress codes, culinary codes, codes of appropriate comportment, or codes that govern gifts of flowers.

      Yet these codes should not be understood as fixed templates for how social life is to be lived. They vary with the social contexts in which they are deployed, whether these be on the level of household, family, community, region, or on the level of society at large. They also vary according to the domain they address—such as economics, politics, or religion—and according to the social characteristics of the parties to the communication process, including their social origins, gender, age, educational milieu, occupation, and class position. Since these social categorizations involve variabilities in access to power, power equalities or differentials are at work in defining who can address whom, and from what symmetrical or asymmetrical positions. The grid formed by these rankings and positions, in turn, sets up the contexts for how things are said or performed and codifies how they are to be understood.

      Communicative processes must thus strike a balance between adhering to codes and to the formal properties of codes, on the one hand, and fostering variability in application, on the other. Adherence to rules supports intelligibility and coherence; variability permits communication to be fitted to changing circumstances. Yet these operations of replication or variation do not take place in the minds of isolated individuals. The signs and codes employed possess a tangible, public quality, a reality that anyone attempting to communicate must take into account; no one can simply invent a language or a culture individually. The processes of reproducing or modifying communicative traditions are social, carried on by socialized participants with communicative means and skills deployed publicly in social contexts.

      Just as all social arrangements, including those of communication, involve relations of power, so also is that true of ideas. Contrary to the old German revolutionary song that proclaimed that thoughts are “free,” (“die Gedanken sind frei”), ideas and idea-systems are often monopolized by power groups and rendered self-enclosed and self-referential. While ideas are subject to contextual variation, moreover, this variation itself encounters structural limits, since contexts too involve social relationships and thus acquire their structure through plays of power. A key question that emerges is how power operates in these contexts to control potential disruption. More concretely, we need to inquire into how conflicts between tradition and variability in communication are fought out.

      This kind of inquiry shifts attention away from an internal analysis of how codes are arranged, transmitted, or altered to questions about the society in which these messages are sent and received. Linguistics and semiotics investigate the mechanics of communication that lay the groundwork of signification, but they do not yet address what the communicatory act is about, what it asserts or denies about the world beyond the vehicle of discourse or performance itself Communicative acts impute attributes to the world and convey them as propositions to their audiences. It is part of the ethnographer’s task to bring together the different pronouncements thus made, to note their congruence or disjunction, to test them against other things said and done, and to guess at what they might be about. It should also be his or her task to relate these formulations to the social and political projects that underwrite discourse and performance and to assess the relevance of these projects to the contests over power in social relations. These contests involve ideational repertoires; stress on one repertoire over another can affect the outcome of power struggles, opening up opportunities to one