On the malleability of people and computers: a focused approach to office ethnography

Wherever we turn there is talk about the “impact” of new technology upon society. Some of this talk is optimistic — indeed wildly so. Thus Christopher Evans, in his book The Mighty Micro, suggests that the “computer revolution” will lead to the decline of communism, the transformation and rise of the third world, the increasing dominance of the home, the substantial modification of crudely exploitative capitalism and the probable ending of war — a catalogue of benefits which would suggest that the end of human suffering is at hand. On a less global scale, others talk of the arrival of the microcomputer as if it heralds the dawn of a new era of emancipation for secretaries and businesses alike — the boring routines of typing will be replaced by skilled and rewarding tasks while the efficiency of business will be multiplied many times over¹.

However, the optimists are counterbalanced by the pessimists. There are those — for instance Herbert Marcuse — who hold technology to be inherently totalitarian². More moderately, there are many who believe that the impact of new office technologies is disadvantageous if not downright disastrous, especially for those who find themselves at the bottom end of the social scale. Unemployment, or so it is suggested, will increase as tasks are automated and skills become outdated, while the conditions of work of those who still have jobs will deteriorate. In short, the alternative view is that the arrival of microcomputers and word-processors in the office simply marks the latest episode in a continuing struggle by those in positions of power to exploit those who are under them³.

Though we are skeptical whether the arrival of the microcomputer will lead to the ending of war, or indeed, whether new technologies are inherently dominatory, there is doubtless something to be said in favor of the views of both the optimists and the pessimists. In many offices the more routine tasks have been automated and the process of retraining has improved the productivity of the enterprise together with the life-chances of all concerned. In others it has led to unemployment and increased surveillance and the replacement of skilled personnel by poorly paid unskilled employees. Our problem with the literature on the “impact” of technology does not lie in the fact that consequences, sometimes good and sometimes bad, have followed from the introduction of new technologies in different offices. It lies, rather, in the vocabulary that is used to describe these consequences. This vocabulary is Newtonian in character. Microcomputers are treated as being like projectiles which arrive from outer space (or perhaps from the technical agents of the ruling class) and subsequently have an “impact” upon social arrangements. It is as if the social world were best seen as a set of craters, the passive target of bits and pieces that are lobbed at it. This is surely wrong.

Though there are many things wrong with such a vocabulary the most basic is the way in which technology is treated as an object — usually a unitary and relatively unproblematic object — which differs in kind from the social structure to which it relates. Analysis is predicated on the assumption that technical objects inhabit one domain while social relations exist in another. Sometimes one influences the other — indeed, in the more sophisticated versions of the theory they are seen as having reciprocal relations — but whether or not they interact they are seen as being different in kind. The first major purpose of this paper is thus to argue against this assumption and to propose, by contrast, that if we want to understand the way in which office technologies interact with social relations then we need to find a unified and holistic vocabulary for talking about devices-and-social-relations. We should, in other words, study the office as a system of interrelated parts and transactions. Any such system has its own set of more or less stable dynamics — it may, to shift metaphors, be treated as a network of components and interactions between those components. Furthermore, and this is a crucial point, the components, and the interactions between the components, are heterogeneous in character. Some of them have to do with power, authority or duty — that is they are predominantly social in character — while others have to do with circuits and circuit boards — that is, they are primarily technical in character. In between, however, there are many which are “sociotechnical” in character — they run together or depend upon elements of both the social and the technical: the operation of a word-processing program would be a case in point⁴.

Understanding the “impact” of new technologies therefore requires us to describe the important components and interactions that go to make up the dynamics of an office whether these be social, technical, or (as is very often the case) sociotechnical in character. We have, therefore, to minimize the conventional boundaries between the social and the technical (and for that matter, the economic, the political and the scientific) and map out the heterogeneous and interrelated commitments that keep the office or enterprise on the road if we want to understand how and why office information systems are adopted, assimilated, adapted or rejected. In short, we need to avoid the “impact” metaphor altogether, and treat the new technologies as sets of real or potential interrelations that may (or may not) be married with the other sociotechnical relations that make up the average office.

One way of handling the complexity and heterogeneity of office systems is to adopt an ethnographic approach and to consider the empirical detail of the interacting processes that may be detected in the office⁵. If this is to be profitable, however, it must be systematically undertaken. We will, as E.C. Hughes put it, “fall far short of our intellectual duty if we simply record events unsystematically on a sea of change” (Hughes :1971:55). In other words, the most useful ethnographies will be those that rest upon agreement about what should be examined and how it should be analysed. In the present paper we cannot consider our theoretical approach in any detail. In general, however, we share Hughes' view that the study of what he called “going concerns” offers a plausible basis for organizing ethnography. Hughes wrote that going concerns are identifiable enterprises, institutions, organizations, associations or families that are subjected to contingencies that arise from their environmental context. Thus he pictured each educational institution as being “in the market for students, staff, ministers, whatever kinds of help and personnel it wants” (Hughes :1971:64). Such institutions might, he noted “occur in many forms, and may be in any stage of having, getting, or losing moral, social, legal, or simply customer approval” (Hughes :1971:54). They operate, he noted, in a variety of markets and need, if they are to survive, to satisfy a variety of demands.

Consider what happens if we think of a university department — our chosen example of an office — as a going concern and look at the demands that are placed upon it, and which it places upon its environment. The first thing to note is that it draws students, staff, financial and various forms of material support from that environment. In return it feeds various kinds of products — instruction, assessments, graduates, degrees, research papers, salaries and the like — back to its environment. One way of looking at the department is thus as a productive enterprise, or perhaps more precisely, as complex set of heterogeneous transformational chains which converts what it takes from the environment into what it gives back. Little can be said a priori about these transformational chains. Rather their mode of operation, their assessed efficiency, and the alternatives available are all matters for empirical investigation⁶.

The transformation of resources into accomplishments within the department is not the end of the story. Transformations that lie outside the department are (as a thought experiment will quickly show) just as crucial to its character as a going concern as those that occur within its walls. Thus, instruction may not be directly converted into financial support, but failure to instruct will be quite quickly transformed into failure to pay. Thus one way of looking at the productive output of the department, albeit one that is rather unusual, is to treat it as a set of resources that will (or so it is hoped) operate on outside individuals and institutions in order to transform these into or cause them to generate new resources for the department.

This analysis of the relationship between a department and its environment may be sharpened up by noting a further point. This is that much, if not all, of its activities are routine and indeed cyclic in character. It is important to see that it is precisely this routine character of activities — the fact that they are repeated on a regular basis — that makes it possible to talk of the university department as a going concern⁷. The cycle of transformations, starting with resources which are converted in various complex ways within the department to outputs which are in turn, periodically transformed into new resources — a cycle which we will from time to time call the “internal economy” of a going concern — for all intents and purposes constitutes the department. In the absence of routine it would not be a university department, indeed it would not be an identifiable going concern at all.

In what follows we use this holistic but informal “systems” vocabulary to describe the process by which the department in question decided that it needed a microcomputer, the considerations that led to the selection of a particular machine and the sequence of events that followed its introduction⁸. In the course of this description we move, indifferently, between the technical, the social, and the sociotechnical. Our aim is to consider the conditions under which the more technical components of offices such as microcomputers turn out to be more or less malleable when they are juxtaposed with the more social components of those systems. It is, in short, to consider the circumstances under which technologies are molded by the social or, alternatively, the social is molded by the technological — or, as is perhaps the more common case, the sociotechnical is molded by the sociotechnical. We contend, though this cannot be demonstrated in a single case-study, that what it is that is malleable, and what it is that is not, is always a contingent, empirical matter.