Reference:Evolution of socio-technical systems/The historical background (1950-1970)

Origin of the concept

The socio-technical concept arose in conjunction with the first of several field projects undertaken by the Tavistock Institute in the British coal mining industry. The time (1949) was that of the postwar reconstruction of industry in relation to which the Institute had two action research projects^[1]. One was concerned with group relations in depth at all levels (including the managemenUlabor interface) in a single organization - an engineering company in the private sector. The other project focused on the diffusion of innovative work practices and organizational arrangements which did not require major capital expenditure but which gave promise of raising productivity. The former project represented the first comprehensive application in an industrial setting of the socio-clinical ideas concerning groups being developed at the Tavistock. For this purpose a novel action research methodology was introduced. The book describing the project became a classic (Jaques, 1951). Nevertheless, it approached the organization exclusively as a social system. The second project was led, through the circumstances to be described below, to include the technical as well as the social system in the factors to be considered and to postulate that the relations between them should constitute a new field of inquiry.

Coal being then the chief source of power, a lot depended in industrial reconstruction on there being a plentiful and cheap supply of it. But the newly nationalized industry was not doing well. Productivity failed to increase in step with increases in mechanization. Men were leaving the mines in large numbers for more attractive opportunities in the factory world. Among those who remained, absenteeism averaged 20%. Labour disputes were frequent despite improved conditions of employment. Some time earlier the National Coal Board had asked the Institute to make a comparative study of a high producing, high morale mine and a low producing, low morale but otherwise equivalent mine. Despite nationalization, however, our research team was not welcome at the coal face under the auspices of the Board.

There were at the Institute at that time six postgraduate Fellows being trained for industrial field work. Among these, three had a trade union background and one, the late Ken Bamforth, had been a miner. After a year, the Fellows were encouraged to revisit their former industries and make a report on any new perceptions they might have. Bamforth returned with news of an innovation in work practice and organization which had occurred in a new seam in the colliery where he used to work in the South Yorkshire coalfield. The seam, the Haighmoor, had become possible to mine 'shortwall' because of improved roof control. I can recall now the excitement with which I listened to him. No time was lost, in my going up to visit this colliery where, since we were introduced by him, the local management and union readily agreed to our 'researching' their innovation with a view to its diffusion to other mines. The area general manager (who had the oversight of some 20 mines) welcomed the idea. The technical conception of the new scheme was his, though the men, with union support, had proposed the manning arrangements.

The work organization of the new seam was, to us, a novel phenomenon consisting of a set of relatively autonomous groups interchanging roles and shifts and regulating their affairs with a minimum of supervision. Cooperation between task groups was everywhere in evidence; personal commitment was obvious, absenteeism low, accidents infrequent, productivity high. The contrast was large between the atmosphere and arrangements on these faces and those in the conventional areas of the pit, where the negative features characteristic of the industry were glaringly apparent. The men told us that in order to adapt with best advantage to the technical conditions in the new seam, they had evolved a form of work organization based on practices common in un mechanized days when small groups, who took responsibility for the entire cycle, had worked autonomously. These practices had disappeared as the pits became progressively more mechanized in relation to the introduction of 'longwall' working. This had enlarged the scale of operations and led to aggregates of men of considerable size having their jobs broken down into one-man-one-task roles, while coordination and control had been externalized in supervision, which had become coercive. Now they had found a way at a higher level of mechanization of recovering the group cohesion and self-regulation they had lost and of advancing their power to participate in decisions concerning their work arrangements. For this reason, the book which overviewed the Tavistock mining studies was subtitled, 'The Loss, Rediscovery and Transformation of a Work Tradition.' (Trist et aI., 1963). The transformation represented a change of direction in organizational design. For several decades the prevailing direction had been to increase bureaucratization with each increase in scale and level of mechanization. The organizational model that fused Weber's description of bureaucracy with Frederick Taylor's concept of scientific management had become pervasive. The Haighmoor innovation showed that there was an alternative.

Those concerned with it had made an organizational choice (Trist et aI., 1963). They could, with minor modifications, have extended the prevailing mode of working. They chose instead to elaborate a major design alternative. It was not true that the only way of designing work organizations must conform to Tayloristic and bureaucratic principles. There were other ways, which represented a discontinuity with the prevailing mode. The technological imperative could be disobeyed with positive economic as well as human results. As became clearer later, what happened in the Haighmoor seam gave to Bamforth and myself a first glimpse of 'the emergence of a new paradigm of work' (Emery, 1978a) in which the best match would be sought between the requirements of the social and technical systems.

Some of the principles involved were as follows:

1. The work system, which comprised a set of activities that made up a functioning whole, now became the basic unit rather than the single jobs into which it was decomposable.
2. Correspondingly, the work group became central rather than the individual job-holder.
3. Internal regulation of the system by the group was thus rendered possible rather than the external regulation of individuals by supervisors. .
4. A design principle based on the redundancy of functions rather than the redundancy of parts (Emery, 1967) characterized the underlying organizational philosophy which tended to develop multiple skills in the individual and immensely increase the response repertoire of the group.
5. This principle valued the discretionary rather than the prescribed part of work roles (Jaques, 1956).
6. It treated the individual as complementary to the machine rather than as an extension of it (Jordan, 1963).
7. It was variety-increasing for both the individual and the organization rather than variety decreasing in the bureaucratic mode.

Conceptually, the new paradigm entailed a shift in the way work organizations were envisaged. Engineers, following the technological imperative, would design whatever organization the technology seemed to require. This was a rule accepted by all concerned (Davis et aI., 1955). The 'people cost' of proceeding in this way was not considered. Any people cost, it was presumed, could be compensated for first by improving the socioeconomic conditions or employment, next by improving 'human relations'. The movement under this latter title arose during the interwar period when the model of the technocratic bureaucracy was becoming entrenched. It failed to arrest the spread of work alienation after World War II (Baldamus, 1951, 1961; Walker and Guest, 1952). At the Glacier Metal Company where Jaques (1951) carried out his research, it was observed that, despite the progressive personnel policies adopted and the far-reaching changes made in the character of management-labor relations, there was no reduction in 'the split at the bottom of the executive chain.' Nothing had happened to change the structure of jobs. There was no change in the nature of immediate work experience.

The idea of separate approaches to the social and the technical systems of an organization could no longer suffice for one such as myself who had experienced the profound consequences of a change in social-technical relations such as had occurred in the Haighmoor development. Work organizations exist to do work - which involves people using technological artifacts (whether hard or soft) to carry out sets of tasks related to specified overall purposes. Accordingly, a conceptual reframing was proposed in which work organizations were envisaged as socio-technical systems rather than simply as social systems (Trist, 1950a). The social and technical systems were the substantive factors - the people and the equipment. Economic performance and job satisfaction were outcomes, the level of which depended on the goodness of fit between the substantive factors.

The following research tasks emerged in the Tavistock program:

1. The theoretical development of the core concept.
2. Methods for the analytical study of the relations of technologies and organizational forms in different settings.
3. A search for criteria to obtain the best match between the technological and social components.
4. Action research to improve the match.
5. Ways to measure and evaluate outcomes through comparative and longitudinal studies.
6. Ways to diffuse socio-technical improvements.

These tasks could not be carried out in a preplanned sequence. The research team had first to make an extensive reconnaissance of the field to locate relevant opportunities. It then had to become actively linked to them in ways which would sanction their study in a collaborative mode. The idiom of inquiry was action research (Trist, 1976b).

Socio-technical studies needed to be carried out at three broad levels -from micro to macro- each of which is interrelated.

1. Primary work systems. These are the systems which carry out the set of activities involved in an identifiable and bounded subsystem of a whole organization - such as a line department or service unit (c.f. Miller, 1959). They may consist of a single face-to-face group or a number of such groups together with support and specialist personnel and representatives of management plus the relevant equipment and other resources. They have a recognized purpose which unifies the people and the activities.
2. Whole organization systems. At one limit these would be plants or equivalent self-standing workplaces. At the other they would be entire corporations or public agencies. They persist by maintaining a steady state with their environment.
3. Macrosocial systems. These include systems in communities and industrial sectors and' institutions operating at the overall level of a society. They constitute what I have called 'domains' (Trist, 1976a, 1979a). One may regard media as socio-technical systems. McLuhan (1964) has shown that the technical character of different media has far-reaching effects on users. The same applies to architectural forms and the infrastructure of the built-environment. Although these are not organizations, they are socio-technical phenomena. They are media in Heider's (1942) as well as McLuhan's sense.

As the historical process of a society unfolds individuals change their values and expectations concerning work roles. This changes the parameters of organizational design. Conversely, changes in technology bring about changes in values, cognitive structures, lifestyles, habitats and communications which profoundly alter a society and its chances of survival. Socio-technical phenomena are contextual as well as organizational.

Not all social systems are socio-technical. Emery (1959), following Nadel (1951), distinguished between 'operative' and 'regulative' institutions and proposed to restrict the term 'socio-technical' to the former. Regulative organizations are concerned directly with the psychosocial ends of their members and with instilling, maintaining or changing cultural values and norms, the power and the position of interest groups, or the social structure itself. Many such organizations employ technologies as adjuncts and have secondary instrumental systems which are socio-technical. By contrast, organizations which are primarily socio-technical are directly dependent on their material means and resources for their outputs. Their core interface consists of the relations between a nonhuman system and a human system.

There are mixed forms typified by the co-presence of psychosocial and socio-technical ends which may be congruent or conflicting. An example of the latter would be a prison with both an electronic surveillance system and a therapeutic community. Hospitals are inherently socio-technical as well as psychosocial, which accounts for the complexity of some of their dilemmas.

From the beginning the socio-technical concept has developed in terms of systems, since it is concerned with interdependencies. It has also developed in terms of open system theory, since it is concerned with the environment in which an organization has actively to maintain a steady state. Von Bertalanffy's (1950) paper on 'Open Systems in Physics and Biology' became available at the time that the socio-technical concept was being formulated. It influenced both theory building and field projects, compelling attention alike to self-regulation and environmental relations. As regards the special role of technology, Emery (1959) put it as follows:

the technological component, in converting inputs into outputs, plays a major role in determining the self-regulating properties of an enterprise. It functions as one of the major boundary conditions of the social system in mediating between the ends of an enterprise and the external environment. Because of this, the materials, machines and territory that go to making up the technological component are usually defined, in any modem society, as 'belonging' to an enterprise, or are excluded from similar control by other enterprises. They represent as it were, an 'internal environment'. This being the case, it is not possible to define the conditions under which such an open system achieves a steady state unless the mediating boundary conditions are in some way represented amongst 'the system constants' (cf. Von Bertalanffy, 1950). The technological component has been found to play this mediating role and hence it follows that the open system concept, as applied to the enterprise, ought to be referred to the sociotechnical system, not simply to the social system.

Source influences

An interest in social and technical relations arose in my own thinking first at the macrosocial level, next at the whole organization level and thence at the level of primary work systems. This last, however, became the crucial level as regards the initiation of field projects which provided the concrete route through which the broader levels could again be reached.

Mumford (1934) in 'Technics and Civilizations' had introduced me to the idea of linking the two. Anthropology and cultural history suggested that, if the material and symbolic cultures of a society were not connected by any simple principle of linear causality (as some interpreters of Marx have implied), they were nevertheless intertwined in a complex web of mutual causality (Trist, 1950b). In the language of E.A. Singer (1959) they were co-producers of each other. The technological choices made by a society are critical expressions of its world view. As new technologies develop, new societal possibilities mayor may not be taken up. The mode of their elaboration may be constructive or destructive. There are unanticipated consequences. In the period following World War II the information technologies of the second industrial revolution were already beginning to make themselves felt. It seemed not unlikely that there would be as big a cultural shift associated with them as with the energy technologies of the first industrial revolution.

As regards the whole organization level, the first industrial project in which I was involved made it impossible not to look at the relations between technical and social systems. This encounter was with the jute industry in Dundee, Scotland, where in the late thirties I was a member of an interdisciplinary research team studying unemployment. The spinning section of the industry was being 'rationalized', causing not only more unemployment but a de-skilling of the remaining workers, along with an extension of managerial controls. As to alienation, workers in the interview sample would say that they might as well be unemployed, while the appearance of time-study men provoked a bitter reaction in the trade unions. In the changes taking place, the technical and social aspects were interactive. A new socio-technical system emerged - that of a more controlling 'technocratic bureaucracy' with very different properties from the earlier system in terms of which jute spinning had been, and jute weaving still was, organized. Then came World War II. A new military socio-technical system appeared in the form of the German Panzer Divisions, formidably competent in the way they linked men and machines to fit their purposes. The French army had failed to develop an equivalent, despite de Gaulle's proposals.

As the war proceeded, military technology gave increasing scope for, and prominence to, small group formations, recognizing their power to make flexible decisions and to remain cohesive under rapidly changing conditions'. This led to a recasting of the role of junior officers and the kind of relations (more open and more democratic) best maintained between them and their men. In Britain the War Office Selection Boards (to which I was attached) were created to choose officers capable of behaving in this way. The Boards made extensive use of W.R. Bion's (1946) method of leaderless groups, which allowed leadership to emerge and rotate in a variety of group settings. All this opened µp new areas of group dynamics - extended after the war when Bion (1950, 1961) introduced therapy groups at the Tavistock Clinic. A parallel influence was that of Lewin's (1.939, 1951) experiments on group climates and group decision making, together with the beginnings of the National Training Laboratories. These traditions became fused at the Tavistock. Bion focussed on the unconscious factors obstructing the attainment of group purposes and on group creativeness; Lewin on the commitment to action consequent on participation and on the performance superiority of the democratic mode. Both emphasized the capacity of the small group for self-regulation, an aspect of systems theory which received increasing attention as cybernetics developed (Weiner, 1950).

Going against the grain of the fifties

To a number of us at this time, and certainly to me, it seemed that the small self-regulating group held the clue to a very great deal that might be improved in work organizations. Knowledge about it had made considerable advances during and immediately after World War II. Yet experiences in industry in the reconstrUFtion period had shown that socio-technical relations were patterned on the breakdown of work into externally controlled one-man-onejob units and that top-down management hierarchies were being even more rigidly maintained than in the prewar period. The pattern of technocratic bureaucracy was increasing in strength.

Hence the interest of the Haighmoor development, which pointed to the existence of an alternative pattern going in the opposite direction to the prevailing mode. The Divisional Board, however, did not wish attention drawn to it. They feared the power change that would be consequent on allowing groups to become more autonomous at a time when they themselves were intent on intensifying managerial controls in order to accelerate the full mechanization of the mines. They refused to allow the research to continue and balked at Bamforth and myself referring to it in the paper that we published (1951) on conventional longwall working. It would lead, they said, to expectations that could not be fulfilled; for, while autonomous groups might be successful on the Haighmoor shortwalls, they would not be feasible on longwall layouts which represented the prevailing method of mining. Later, this opinion was found to be false, though widely believed. The Divisional Board's reaction suggested that any attempt to reverse the prevailing mode would be met with very serious resistance. To move in the opposite direction meant going against the grain of a macrosocial trend of institution-building in terms of the model of the technocratic bureaucracy, which had yet to reach its peak or disclose its dysfunctionality.

Several major pioneer studies were carried out during the decade. They established a number of research findings of key importance. Their effect on industrial practice was negligible. Neither what happened nor what failed to happen is widely known. These studies are reviewed here to provide a short account of what turned out to be the laterrcy decade of the socio-technical approach.

The continuation of the mining studies

If the Haighmoor development had general meaning, it was reasonable to assume that similar developments would occur elsewhere. In fact, a parallel development in a more advanced form and on a larger scale emerged in another Division of the National Coal Board (East Midlands), where one of the Area Managers, W.V. Sheppard (1949, 1951), was developing a method of continuous mining - a radical innovation designed on what appeared to be socio-technical principles. There were two versions: the semi-mechanized (Wilson and Trist, 1951) and the fully mechanized (Trist, 1953a). The second was delayed because of teething troubles in an ingenious but somewhat underpowered cutter-loader invented by Sheppard. Faces were 100 yards in length, alternating advance with retreat and concentrated in one district so that only one main road needed to be maintained. Autonomous groups of 20-25 conducted all operations on one shift. There were three production shifts every 24 hours instead of one - the other two shifts had been concerned with coal face preparation and equipment shifting which were now done in parallel with coal getting. All members were multi-skilled and were paid the same day wage, which was judged more appropriate for continuous mining than a bonus. Productivity and work satisfaction were unusually and consistently high. A beginning was made in spreading the new system to six pits. Emery (1952), who was over at the Tavistock on sabbatical from Australia, made a study of this process, paying special attention to required changes in the supervisor's role. After area-wide appreciation conferences had been held for managers and under-managers, an Area Training School was designed (Trist, 1953b) to which groups of eight (operators, foremen and mechanics) from each pit scheduled to go over to the new system came for a week (during which they visited ,the original mine). Members of these groups began to meet weekly to compare experiences. A kind of socio-technical development center was created. This model was not picked up again for another twelve years, when something like it emerged both in the Norwegian Industrial Democracy project (emery and Thorsrud, 1976) and the Shell Philosophy project (Hill, 1971). It was a forerunner of 'the deep slice' used by Emery (1976) in his method of Participant Design.

A study of overall area organization was made (Trist, 1953c). The incoming technology in association with autonomous work groups reduced by one the number of management levels underground. Group Centers between collieries and the Area Office were obviously redundant. Divisional Boards between operating Areas and the National Headquarters in London also seemed unnecessary. These superfluous levels of management were based on narrow spans of control which implied detailed supervision of subordinates at all levels rather than the socio-technical concept of boundary management which was congruent with maximizing the degree of self-regulation throughout an entire organizational system. In the course of time, these levels were in fact eliminated. This showed how the socio-technical concept could affect the organization as a whole and reduce the administrative overhead which has become so excessive in large technocratic and bureaucratic organizations.

Having reached the whole organization system level, our research efforts (though on independent funds) were again stopped when a new Divisional Chairman took over. What had happened was seen in an entirely technological perspective - that of the new cutterloader which had been introduced. Since this was judged not as good a bet for further mechanization as another similar machine, the whole project was regarded as not meriting continuation. Besides, granting more autonomy was not popular. The union regionally negotiated special pay for operators of new equipment. This broke up the unity of the face groups, which were further decimated when bonuses were introduced for various classes of workers. Within a year or two, the conventional system reinstated itself.

Sociologically, this setback and the earlier one over the Haighmoor may be seen as examples of what Schon (1971) has called the 'dynamic conservatism' of organizations. Psychologically, at the unconscious level, these setbacks may be seen as stemming from 'envious attacks' on the innovations and the innovators. In psychoanalytic object relations theory (Klein, 1958) a good object, which one cannot bear because it is not one's own, may for that reason be turned into something bad, which then becomes a threat through having one's hostility projected on to it. Creativeness is apt to stir up jealous hatred of this kind and the creators all too often become the targets of destructive spite. I have encountered a number of cases of this in studies of innovation with which I have been associated.

A search of other coalfields produced only one, Durham, where the Divisional Board and the regional organization of the National Union of Mineworkers said they would like to proceed with social research into mining methods. Virtually all extant methods were available in the same low seam in a single area in the older part of the coalfield where customs were uniform and traditions common. Here, the research team found what the conventional wisdom had held to be impossible: the working of the conventional, semimechanized, three-shift longwall cycle by a set of autonomous work groups (locally known as composite). Groups of 40-50 men interchanged the various jobs required while alternating shifts in ways they felt best and evolving an innovative pay system that seemed equitable to them. Output was 25 percent higher with lower costs than on a comparison face similar in every respect (conditions, equipment, personnel) except that of work organization. Accidents, sickness and absenteeism were cut in half (Trist et aI, 1963). Only one man left the composite faces in two years. Over the four-year period of the project, the conversion of an entire colliery with three seams from conventional to composite working was followed in detail. Much was learned about the conditions under which autonomous groups prosper and under which they fail. The potential of selfregulating groups in fully mechanized installations was studied and the research team began to collaborate in the design of sociotechnical systems for the most advanced technology then available. A meticulous study of a single face team was made by Herbst (1962); it explored the mathematical relations between a number of key variables.

A report was submitted to the National Coal Board (Trist and Murray, 1958). The results were not disputed. But the Board's priorities were elsewhere - on the closing of uneconomic pits in the older coalfields and carrying the union with it in implementing the National Power-Loading agreement, deemed critical for full mechanization. It was not willing to encourage anything new that might disturb the delicately balanced situation as the industry contracted in face of the greater use of oil. On the union side, the Durham Miners' Association sent the report to the National Executive. No reply was received at the Tavistock Institute.

Dr. Hugh Murray has since^[1] made an archival study of composite agreements in the various British coalfields. There were quite a few of these in the mid-fifties, but they were regarded simply as wage settlements. There was no understanding that they might have implications for work-organization.

In the late sixties Murray carried out an action-research study of layouts using very advanced technology. He found that the coincidence of specialized work roles and high absentee rates was giving rise to wide-scale disruption of production processes. Men were posted to places in their speciality all over the mine through a 'pit market.' There was little cohesion in work teams. Efforts to introduce multiskilling, which would have afforded the basis for greater team cohesion, met with little success (Murray et aI., 1969).

During the seventies an experimental section based on autonomous groups was tried out in a mine in the American coal industry with its room-and-pillar layouts and very different technology of roof bolting, continuous miners and shuttle cars. Positive results were obtained comparable to those obtained earlier in Britain; not only as regards productivity but as regards safety, which was the reason for union collaboration. Although a second autonomous section was started, an attempt to diffuse this form of work organization to the mine as a whole encountered insuperable difficulties which were not foreseen by members of the Labor-Management Steering Committee or the research team (Trist, Brown and Susman, 1979). This project has been independently evaluated by Goodman (1979).

The difficulties centred on the resentment of those not included in the experiment towards the privileges of those who were. This resentment would not have become acute had not expansion of the mine led to some inexperienced new recruits winning places on the second autonomous section (and hence the top rate) when experienced men withdrew their bids at the last moment in order to stay with a foreman (who then deserted them). There was no infringement of seniority rules, but the issue split the union.

The project shows in great detail how unanticipated and uncontrollable events in the broader as well as the immediate context can influence outcome in the later stages of an action research undertaking. It also shows how the encapsulation of an innovation can prevent its diffusion and the dangers of applying classical experimental research design in the 'moving ground' of a real life field situation - even though this was a condition of receiving initial support at the mine and from the sponsors of the national program of which it was a part.

Studies in other industries

Meanwhile, at the Tavistock, opportunities were sought in other industries. The first to arise was not only in another industry, textiles, but in another culture -India. In 1953 the late A.K. Rice (1958, 1963) paid his first visit to the Calico Mills in Ahmedabad in which an automatic loomshed was converted from conventional to autonomous group working, with results that surpassed expectations. Later, the change was diffused throughout the non-automatic weaving sheds in this very large organization, which employed 9,000 people. Rice did no more than mention through an interpreter the idea of a group of workers becoming responsible for a group of looms. The loomshed employees took up the idea themselves, coming back next day with a scheme which they asked management's permission to implement. Terms regarding a progressive payment scheme were negotiated, and the first trials of the new system began. As with the mines, major initiatives were taken by the workers themselves. The depth of their commitment became apparent later, when the Communist Party of India (orthodox) took offense at the 'Ahmedabad Experiment' and drafted a number of their members from various parts of the country into the city, swollen with refugees from West Pakistan, to agitate against it. Though their families were threatened and attempts were made to set Hindu and Muslim workers against each other, the Calico's employees stood by an innovation which was largely their own creation.

Yet the group method, as it was called, did not spread to other mills as originally expected. I asked Shankalal Banker, the venerable leader of the Ahmedabad Textiles Union, about this when I was in Ahmedabad in 1976. He replied that the other owners did not want to share the po wee Also, as Miller (1975) reports, the non-automatic loomsheds gradually regressed to conventional ways of working. Training was not kept up. New middle managers took over who knew little of what had originally taken place. Senior management became preoccupied with marketing and diversification. The automatic loomsheds, however, have retained the group method and their high level of performance and satisfaction with it.

During the early fifties also, Seymour Melman (1958), who had come over from Columbia to Oxford, made an in-depth study of work practices in the Standard Motor Company at Coventry. This company, which made both tractors and automobiles (and some airplane engines), employed 12,000 workers who, through their unions, largely controlled work arrangements and practices on the shop floor. There were only 70 foremen in the entire organization. Only 16 people were in the personnel department. There were only 8 time-study men. The ratio of administrative to production workers was far lower than in the rest of the industry and had been held steady while it increased elsewhere. At the automobile plant, the workers formed themselves into 15 large, internally differentiated groups varying from 50 to 500, each of which comprised a worker constituency which negotiated its detailed conditions of work and operating rules within a plant-wide union agreement, itself separate from the rest of the industry. The large groups were known as 'gangs'. They controlled upgrading and deployment among eight broad classes of jobs (reduced to these few from a very large number). They negotiated the bonus for the number of products turned out in a given time. These products constituted a major subsystem of the automobile. The bonus was large and induced component groups in the gangs to cooperate. The primary work systems, which contained many component groups, represented a sophisticated adaptation of earlier gang systems (which were disappearing) and constituted a complementary decision system to that of management. The foremen controlled the boundaries of productive activities, not the people.

The company increased its market share during the five years in question beyond that of other automobile companies in Britain, introduced automated equipment at a much earlier date, paid very much higher wages (yet had lower unit costs), remained attractively profitable and increased its assets by a third. In later years (the company was eventually taken over by British Leyland) this pattern of work organization met with severe management opposition.

Too much power was being shared. Yet where the prevailing mode of a highly controlling technocratic bureaucracy has been imposed, there have been substantial increases in administrative costs and huge labor trouble. ^[1]

The Tavistock workers sought to discover how far alternative organizational patterns existed in service industries. An instance was found in a large national retail chain consisting of small shops run by 4-6 employees with shared tasks and all-around skills; the 'manager' was a working charge-hand. (Pollock, 1954). When, however, this organization enlarged its shops and extended its lines of sale, specialized jobs with several different statuses and rewards appeared along with formal control mechanisms.

At roughly the same time, opportunity arose to explore the possibility of an alternative organizational mode in a large teaching hospital. Advances in medical technology had turned the hospital into 'high pressure' center for intensive treatment while reducing the length of patient stay and extending the range of diseases coped with. This had created quite severe problems in nurse training. The work system consisted of a set of tasks broken down into narrow jobs in a closely similar way to that in large-scale industry. An attempt to introduce, in an experimental ward, the concept of a group of nurses becoming responsible for a group of patients met with both medical and administrative resistance, though much was learned about the embodiment in social structure and professional culture of psychological defenses against anxiety (Menzies, 1960). Integrated ward teams have since been developed in Australia by Stoelwinder (1978; Stoelwinder and CLayton, 1978).

As the last years of the postwar period came to a close in the earlyfifties, the mood of the society changed from collaboration, which had fostered local innovation, to competition and an adversarial climate in management-labor relations, which discouraged it. No further instances of an alternative pattern were identified. Nevertheless, the mining, textile, and automotive studies had suggested that continuous production industries which were advancing in automation might develop requirements which could eventually lead in a direction counter to the prevailing mode. Accordingly, analytic sociotechnical studies were instituted in chemical plants and power stations (Murray, 1960; Emery and Marek, 1962). These studies· disclosed a basic change in. the core shop-floor tasks: the worker was now outside the technology, adjusting, interpreting, monitoring, etc; he had become a manager of a work system; he needed conceptual and perceptual skills rather than manipulative and physical skills. He usually worked interdependently with others because his essential task was to keep a complex system in a steady state. The opportunity to go over to an alternative pattern, however, did not seem to be under any 'hot pursuit', though Bell (1956) had pointed to the possibility and Woodward (1958) noted the presence of fewer supervisors in continuous process than in mass production plants.

For a moment it looked as though a major action research opportunity would be forthcoming in Britain. Richard Thomas and Baldwin (RTB), the largest complex in the British Steel industry, were preparing to build the most modern steelworks in Europe. They wanted to break with many constraining precedents in management and work practices that would inhibit full advantage being taken of the most advanced equipment. The Director of Education and Training invited the Tavistock to collaborate with him in evolving a new set of roles and decision rules, indeed a whole organizational structure, that would be a better match to the new technology. The method proposed was a series of participative workshops to be held in the RTB Staff College which would be attended by the different levels and functions of management, foremen, key operators and shop stewards. But there were delays in site construction - the ground proved more marshy than expected - and huge additional expenditures were incurred. The participative workshops were never held. In the end, an organizational structure and the various associated appointees were crash-programmed, and all the old roles and practices were reinstated with negative consequences (as time showed) of a severe kind (Miller and Rice, 1967).

There was a rising interest in socio-technical relations among a number of social scientists concerned with industry in the British setting. In Scotland, Burns and Stalker (1961) observed a new management pattern which they called 'organismic' as contrasted with 'mechanistic,' in more technologically advanced industry. Woodward (1958) related changes in organizational structure to broad types of technology. Fensham and Hooper (1964) showed the increasing mismatch between conventional management and the requirements of a rationalized rayon industry. Such studies, however, were widely interpreted (not necessarily by their authors) as supporting a theory of technological determinism. There could be no organizational choice, as had been suggested by the Tavistock researchers.

In the U.S., attention had been drawn to the counterproductive consequences of extreme job fractionization (Walker and Guest, 1952). But concepts of job enlargement and rotation and later of job enrichment (Herzberg et aI., 1959), though concerned with sociotechnical relations, focused on the individual job rather than the work system. In its orthodox form, job enrichment did not countenance participation but relied on experts brought in by management.

In Continental Europe there were occasional signs of a concern with alternative organizational modes. Westerlund (1952) reported the introduction of small groups on the Stockholm telephone exchange. Indeed, a similar transformation had been carried out in Glasgow by a telecommunications engineer (Smith, 1952). King (1964), from a training approach, had introduced groups with a good deal of scope for self-regulation in small textile firms in Norway. Van Beinum (1963) had completed his studies in the Dutch telecommunications industry. In the U. S., Davis (1957) introduced the concept of job design. This constituted a basic critique of industrial engineering and opened the way for systems change which could involve groups and encourage participation. A working relationship between him and the Tavistock group was established.

An opportunity for stocktaking occurred at an International Conference on Workers' Participation in Management in Vienna (Trist, 1958). Interest centered on co-determination in Germany and the Yugoslav workers councils. The idea of involving workers directly in decisions about what should best be done at their own level seemed strange to those concerned with industrial democracy. Only marginal attention was paid to the idea that an alternative pattern of work organization to that prevailing might be on the horizon; in the end, however, it was not entirely ignored (Clegg, 1960).

Confusion regarding the forms and meaning of industrial democracy has persisted and has still not been entirely cleared up. Four different forms may be distinguished, all of which represent modes of participation and the sharing of power. They are:

1. Interest group democracy, i.e., collective bargaining, through which organized labor gains power to take an independent role vis-a-vis management.
2. Representative democracy whereby those at the lower levels of an organization influence policies decided at higher levels (workers on boards, works councils).
3. Owner democracy, as in employee-owned firms and cooperative establishments where there is participation in the equity.
4. Work-linked democracy, whereby the participation is secured of those directly involved in decisions about how work shall be done at their own level.

These four forms may be found independently or together, in consonance or contradiction and in different degrees in various contemporary industrial societies. The work-linked form has been the last to appear historically and is that with which socio-technical restructuring of work is associated (Trist, 1979c). increasing congruency may be hypothesized among the four factors in the longer run. Table 1-1 summarizes their current relations in selected countries. Organizational democracy would be a preferable term to industrial democracy.

Norway and Sweden exemplify a congruent Scandinavian pattern which Holland and Australia approximate. The larger European countries show no consistency. The U.S. and Canada express a North American form. Yugoslavia is very different with no independent unions.

Conceptual developments

A monograph by Emery (1959), who had returned to the Tavistock, put forward a first generalized model of the dimensions of social and technical systems, showing that, though they were multiple, they were not so numerous that analysis would become unmanageable. Eight were identified on the technical side, including level of mechanization/automation, unit operations, the temporo-spatial scale of the production process, etc^[1]. On the social side, rigorous attention had to be paid to occupational roles and their structure, methods of payment, the supervisory relationship, the work culture, etc. - all of which belong to the 'socio' rather than the 'psyche' group (Jennings, 1947). The psyche group, concerned with interpersonal relations and Bion-type 'basic assumptions' regarding group behavior, however important, was not the starting point. Appropriate structural settings had to be created before desirable social climates and positive interpersonal relations would have the conditions in which to develop.

The original formulation of social and technical relations had been made in terms of obtaining the best match or 'goodness of fit' between the two. In conjunction with the Norwegian Industrial Democracy project, to be referred to in what follows, Emery reformulated the matching process (in terms of the more advanced systems theory that had become available) as the joint optimization of the social and technical systems. The technical and social systems are independent of each other in the sense that the former follows the laws of the natural sciences while the latter follows the laws of the human sciences and is a purposeful system. Yet they are correlative in that one requires the other for the transformation of an input into an output, which comprises the functional task of a work system. Their relationship represents a coupling of dissimilars which can only be jointly optimized. Attempts to optimize for either the technical or social system alone will result in the suboptimization of the socio-technical whole.

In the language of Sommerhoff (1950, 1969), a work system depends on the social and technical components becoming directively correlated to produce a given goal state. They are co-producers of the outcome (Ackoff and Emery, 1972). The distinctive characteristics of each must be respected else their contradictions will intrude and their complementarities will remain unrealized.

This logic was held to underlie job and organizational design. Failure to build it into the primary work system would prevent it from becoming a property of the organization as a whole.

The conceptual advances were 'directively correlated' with the involvement of the Tavistock research team in the action-research opportunities which occurred as the decade of the sixties unfolded. A further round of developments took place in 1965 (Davis, Emery and Herbst, 1965) which are incorporated in the next section. 'On Purposeful Systems' (Ackoff and Emery, 1972) has had far-reaching influence on subsequent work.

The pathfinding role of the Norwegian Industrial Democracy project

The hypothesis was made that no further advances could be expected until changes occurred in 'the extended social field' of forces at the macro-social level. Any happening of this kind would change the opportunities for and meaning of the efforts at the primary work system and whole organization levels. While no one could foretell where and when this might occur, such a happening could be expected from the increasing impact of the new infqrmatlon-based techriologies.

The science-based industries were 'the leading part' of the Western industrial system. They functioned as the principal change-generators and brought about many other changes, directly or indirectly (Emery and Trist, 1973). Western societies were beginning what is often referred to as the second industrial revolution.

The anticipated happening occurred in 1962 in Norway, where little modernization of industry had taken place in comparison with other Scandinavian countries. Economic growth had slowed down; the largest paper and pulp company went bankrupt; Norwegian firms were being taken over by multinationals. In many other respects this very small country began to feel that it had lost control of its own destiny. Its environment had become what Emery and I (1963) have called 'turbulent'.

A sudden demand for workers' control erupted in the left wing of the trade union movement. Neither the Confederation of Employers nor the Confederation of Trade Unions felt they understood what it was about. Having set up an Institute for Social Research at the Technical University of Norway, they asked it to conduct an inquiry into the matter. Given the political pressures, Einar Thorsrud, the Director, who had close contacts with the Tavistock, felt the inquiry would be better undertaken in association with a group outside Norway, which had accumulated relevant experience. Accordingly, he invited the Tavistock to collaborate. Very soon Emery and I became, with Thorsrud, part of a planning committee composed of representatives of the two Confederations. The task was to work out a jointly evolved research design. Involvement of the key stakeholders in each step was a basic principle of the design.

The fIrst inquiry undertaken was into the role of the workers' directors, whose existence was mandated by law in both state-owned enterprises and those in which the state had some capital (former German capital given to Norway by the Allies after World War II). Various members of the board were interviewed, including the workers' directors, the principal members of management and of the trade union organization. It was found that, whether the workers' directors were outstanding performers or not, their presence, though valued as enhancing democratic control, had no effect on the feelings of alienation on the shop floor or on performance (Emery and Thorsrud, 1964, 1969). Accordingly, it was proposed that a complementary approach be tried - that of securing the direct participation of workers in decisions about what was done at their own level. These findings were widely discussed throughout the two Confederations and in the press. A consensus was reached that the mode of direct participation should be tried. The committee chose two sectors of industry which were not doing well and which were of strategic importance for the future of the economy (paper and pulp and metal working). Criteria were established in terms of which plants might be selected to conduct socio-technical field experiments which would serve as demonstration projects. Joint committees within these sectors then chose likely plants, which the research team visited to test their suitability and to secure local participation.

The research team made a study of the culture and history of Norwegian society. Industrialization had been late and more benign than in those European countries (or the U. S.) where industrialization had occurred earlier. Industrial relations were stable at the national level where the two Confederations accepted their complementarity. Norway had not passed through a period during which patterns of deference to authority had become entrenched. Traditions of egalitarianism were deep and had been more continuously maintained than in most western societies. The hypothesis was made that this configuration would be favorable for the development of direct participation in the work place. These favorable conditions were strengthened by the homogeneity of the society and by its small size. Members of key groups knew each other and overlapped. If they decided to move in a new direction, networks existed through which a wide support base could soon come into existence.

These contextual conditions permitted a series of four major sociotechnical field experiments involving work restructuring not only to be launched but in three cases to be sustained (Emery and Thorsrud, 1969, 1976). Yet the hypothesis that widespread diffusion into Norwegian industry would occur from high profile field sites turned out to be wrong. They became encapsulated (Herbst, 1976). The diffusion took place in Sweden at the end of the decade - when the Norwegian results created great interest in the Employers and Trade Union Associations. Thorsrud was invited to visit. By 1973 between 500 and 1,000 work-improvement projects of various kinds, small and large, were going on in many different industries. A new generation of Swedes (better educated and more affluent) refused (by absenteeism and turnover) to do the dullest and most menial jobs. The importation of Southern Europeans created social problems. Something had to be done. Managers and unions took up the Norwegian approach and adapted it to their own purposes.

After that, shifts in the macrosocial field in Scandinavia recentered attention on the representation of workers on boards of management just when, in Germany, some interest appeared in direct participation. A number of laws have been passed in Norway and Sweden whose effects are still being assimilated. In both countries a third of the members of the boards have to be workers' representatives.

The Shell Philosophy project

In Britain a large-scale socio-technical project began by the Tavistock with Shell (U.K.) in 1965 showed the need to develop a new management philosophy to establish values and principles which could be seen by all to guide work redesign, if commitment was to be secured not only from the various levels of management but also from the work force (Hill, 1971). This project led to a whole series of two-and-a-half day, off-site, residential conferences to discuss the original draft philosophy and to amend and ratify it.

These conferences involved all levels of the organization from the Board to the shop floor and the outside trade-union officials as well as the shop stewards in five refineries.

After some four years, the advances brought about by this project were arrested by an exceedingly complex situation within both the company and the industry. The ways in which the clock began to be turned back are described in Hill's (1971) book. The approach, however, was taken up by Shell in other countries - Australia, Holland, and more recently, Canada. It appears to be characteristic of innovative processes that after a certain time particular implementive sites reach their limit. The burden of trail-blazing is then taken up by others where favorable conditions emerge.

Meanwhile, what had happened regarding work restructuring and participation, especially in Sweden, created interest in the United States. Though one or two pioneer socio-technical projects had been under way for some time in the U. S., it was not until 1973 that wider public interest was awakened. Notions of work alienation were popularized by the media and associated with the threat of declining productivity in the face of Japanese and West German competition.

At an international conference held at Arden House in 1972, the term 'quality of working life' (QWL) was introduced by Dr. Louis Davis. Along with 'Work in America' (O'Toole, 1972), which extended consideration to the mental health aspects of the workplace and the work-family interface, this conference has set the tone for further developments. In Bateson's (1972) sense, it repunctuated the field. The two volumes of papers emanating from it (Davis and Cherns, 1975) have become its standard reference work. Since then, socio-technical concepts and methods have become one input into a wider field concerned with changing social values and studying the effects of values on organizations and their individual members. The age of resource scarcity has coincided with increasing recognition that advanced industrial societies are producing conditions which are impoverishing the overall quality of life. The quality of life in the workplace is becoming seen as a critical part of this overall quality. It is now less accepted that boredom and alienation are inherently a part of work-life for the many, or that they must perforce accept authoritarian control in narrow jobs. Examples can be pointed to in almost any industry of alternative forms of socio-technical relations where these negative features do not have to be endured. For individuals and organizations alike, there is a choice.

In the fifties, the societal climate was negative toward socio-technical innovation. Thirty years later, as the eighties begin, the societal climate is becoming positive (Walton, 1979), though in most Western countries the support base remains limited in face of the persisting power of the technocratic and bureaucratic mode. Yet this mode is being experienced as increasingly dysfunctional in the more complex and uncertain conditions of the wider environment. Emergent values are moving in the direction of regarding personal growth as a human right. All who wish it should have the opportunity to cultivate it. The work place constitutes a key setting for this purpose. A Norwegian law of 1976 gives workers the right to demand jobs conforming to the six psychological principles described in the next section of this paper. These are the principles which shaped the original socio-technical experiments of the Norwegian Industrial Democracy project.

Notes