Anthropologists have long been interested in the role of technology in culture change. This paper describes one of those instances in which technology is the impetus for behavioral change. The behavioral change then diffuses to other technological contexts where its fit is in question. The type of behavior studied here is one of importance in today's American workplace: the move to a self-managed work teams (SMWT) structure for business organizations. A team is a small group with special skills which is responsible for a specific segment of work located within a larger organization (Katzenbach & Smith, 1993; Orsburn, et al., 1990). The work team structure has been primarily developed for use with workers on the production lines in manufacturing plants. Managers have largely considered the team structure to be a success in this environment and as a result, the structure is now being applied to technical professional work environments, the subject of this study. Technical professionals usually have extensive education in specialized knowledge areas and their work involves processing existing information to produce new knowledge (Mohrman, Cohen, Mohrman Jr., 1995). Our research suggests that in this new context, the self managed work team model may be less useful. This paper will examine how the concepts of culture borrowing, especially the importance of fit, can be used to understand the application of work team concepts in knowledge work settings.
Most companies are searching for ways to maximize the utilization of resources in order to remain competitive. Some of the practices being implemented include Total Quality Management (TQM) and methods that tap into employee involvement and self-regulation. A survey of U.S. Fortune 500 organizations (Lawler, 1995) found that more than 50% have either begun implementation of work teams or are trying pilot teams in parts of the company. Since the 1970's there have been many books and articles extolling the virtues of work teams and outlining steps to successful implementation (Beyerlein, 1996). Almost all of the successful work teams have occurred in manufacturing settings and there has not been much research in the area of technical professional teams. The Center for the Study of Work Teams at the University of North Texas applied and received a three-year National Science Foundation (NSF) grant to study the leadership of technical professional or knowledge work teams. This project involves researchers from five disciplines: anthropology, computer science, engineering, management, and psychology. The data reported here is based on semi-structured interviews conducted with 194 individuals, on 20 different teams, located in five companies.
The discipline of anthropology is represented by Sue Gilly and Ann Jordan who are studying the organizational culture changes occurring due to the reorganization to work teams. This paper describes one of their findings. Manufacturing SMWT are created based upon some essential requirements - each team has a discrete segment of work with clear, measurable goals, they continually improve their processes, cross-training ensures multiple team members can perform all team tasks, and teams gradually take over the day-to-day management of their work. This model for organizational structure may be less successful in knowledge work settings than in manufacturing settings due to the nature of the work and the nature of knowledge worker subculture.
First let us briefly examine a perspective on the history of work teams. Much of the recent literature about the movement to a more involved and self-regulated workforce traces its roots to the work done by researchers from the Tavistock Institute of Human Relations in England beginning around 1950. The early implementation of work teams actually occurred in coal mines, not manufacturing plants. In fact, it is not commonly known that their use was a response to the attempt to apply a manufacturing assembly line organizational structure to mining. When that structure was unsuccessful, coal miners returned to the organizational structure of their past. It was this former mining organizational structure which inspired the SMWT structure. The Tavistock researchers involvement in the coal mining organizational changes and their subsequent publications regarding it influenced the birth of the new SMWT structure.
One of the earliest, and most influential, articles from this research (reprinted in an edited book by Glaser, 1992) was E. L. Trist and K. W. Bamforth's "Some Social and Psychological Consequences of the Longwall Method of Coal-getting" published in the Human Relations journal in 1951. Trist and Bamforth reported on the changes that were taking place in the coal mining industry during the 1950s. When they began their research, the preferred work system in the mines was called the "longwall method" which introduced some automation, occupational roles, and attempted to remove as much of the human element from the process as possible. The expected production improvements from this form of coal extraction were not as significant as expected and management began to believe there were problems with the method.
The researchers from Tavistock began their work by talking to coal miners about the method they had used prior to the longwall method, called the "hand-got" system. The miners used to work in small areas of the mines in interdependent pairs or groups and they were responsible for the complete coal-getting task. This structure allowed miners to have pride in their craft, a multiplicity of skills and a sense of independence. Each group negotiated their own contracts with management. The former structure seemed to eliminate any status differences between workers and the men developed strong relationships with each other that lasted for years.
When the longwall method and mechanization were introduced men were put into different functional groups that performed specific tasks on separate shifts. The groups worked together on long sections of a mine which now required external control. The work began to have many of the characteristics of a small factory. Manufacturing organizations had begun following the guidelines of Industrial Engineering calling for reduction of the human element by robotizing tasks and introducing machines. Trist and Bamforth reported there had been almost no experience with machines in mining so "it was almost inevitable that heavy culture-borrowing of this kind should have taken place" (1992, p. 25) from manufacturing. This new situation set up a "conflict between the demands of the new situation and the resources available from past experience" (1992, p. 10) as the miners tried to deal with the required changes.
The long-wall method proved to have many problems such as establishing good working relationships, communication, absenteeism and injuries due to repetitive tasks, and coordination between different functional shifts. Adjustments were introduced to overcome these problems. The coal miners redesigned their work structure to return to one they were used to - small independent work teams that were again responsible for a section of the mine. This time the workers added trading off doing different tasks to lessen the effects of repetition and add variety. Trist and Bamforth point out in their article that manufacturing is different from coal mining because of the difference in the environment or "background" in which the work occurs. The researchers believed that the background in a factory is constant while in a coal mine the conditions continually change, they are impossible to predict and cannot be altered. Work teams were introduced to promote worker flexibility and input.
A Tavistock monograph was published by Fred Emery in 1959 that outlined what researchers had named "sociotechnical systems" (STS). Organizations are composed of both a social and a technical system. The social system includes the division of labor, how work is coordinated and more subjective issues such as job satisfaction. The technical system includes the tools and techniques required for the conversion process to transfer inputs into outputs. The Tavistock researchers saw the need to emphasize the interrelationship between the social and technical subsystems in any organization (Morgan, 1968). As a result of the coal mine study, the researchers noticed what was to become a key aspect for STS: for greatest effectiveness and a better quality of work-life for employees, the organization needed to jointly optimize its social and technical subsystems.
Soon after the publication of Emery's monograph, a manufacturing facility in Norway began implementing these new concepts. Most of the information about work teams was first spread by the researchers at Tavistock and their literature rather than actual contact. Even though the STS redesign was originally implemented to correct problems occurring from a manufacturing model, Pava (1983) reports that by 1973 700 similar projects had been started in factories in Sweden due to the success in Norway. The early adopters of this new work structure kept this a secret so their competitors would not find out and so researchers would not come to study them.
Today there is more opportunity for diffusion of information through direct contact by attending conferences or talking to friends. A team member we interviewed in our research study verified this when she told us about employee evaluation problems her team was having:
But it's not just Company X. I've heard the same thing concerning others at . . . self-managed team conferences. That seems to be an area of discussion quite frequently, how to make your corporate performance evaluation fit a team environment, or how to do this. Or how to do that.So the benefits of STS and work teams are now being transmitted through conferences and direct interactions with individuals who are on teams.
As anthropologists know, change in the environment is the major source of culture change and this has been the primary motivation for the adoption of work teams. Many organizations today believe the environment is too complex and too uncertain to continue to have control and decision making just reside in the management level of their organizations. A different structure has now become necessary. So the key features of STS, as implemented in manufacturing, include the integration of the technical and social subsystems which usually involves setting up some form of work teams. Experience and research has found that teams should have broad flexible jobs within clearly defined work segments that allow members to develop multiple skills so more individuals can perform a task. Work teams are designed to gradually take on management or supervisory functions allowing the team to become self-organizing therefore reducing the need for elaborate and redundant external controls. Organizational hierarchies can then be flattened over time as teams take over coordination and decision making (Purser & Pasmore, 1992).
As previously mentioned, many organizations are now attempting to implement SMWT in technical professional and service work environments. There have been a few success stories in areas such as customer service, banking, and insurance, but one expert on teams said she wondered if these were really "paper assembly lines" (Linda Moran, personal conversation). Organizations who try to implement the model developed in manufacturing with technical professionals are discovering that some things that worked in a manufacturing environment do not fit into theirs. Based upon our research and that of others who have done work in this area (Mohrman, Cohen, & Mohrman, 1995; Pava, 1983; Purser & Pasmore, 1992) there are three key factors that make the adoption of a manufacturing work teams model difficult or impossible for technical professionals. They include: 1. The nature of knowledge work, 2. The nature of technical professional subcultures, and 3. The degree of preadaptation to teams. First we will examine the nature of knowledge work.
One of the primary requirements from the manufacturing model specifies that teams need to be responsible for specific, discrete segments of work that they can control. The work technical professionals do however is nonroutine, dynamic and uncertain. It involves exceptions to routines and requires judgement and interpretation. Knowledge work involves multiple, concurrent, nonlinear conversion processes. The work requires an emergent process rather than a predefined sequence of steps. It is often difficult to draw clear boundaries around projects and many times different processes are interdependent in ways that are not obvious. In fact, Pava (1983) describes it as "saturated interdependence" where everything depends on everything else.
Much of the work involves learning about the content, processes and the organization that is required to carry out the work. Sometimes the result is the creation of new knowledge (e.g. Research and Development) that has to then be incorporated into other processes (e.g. Manufacturing). This means there is often no way to determine the correctness of the work and the cause and effect of effort is often incomplete or removed from each other in time and space. For example, when a new product is not accepted by the consumer is it due to something the Research and Development department did incorrectly or the way it was manufactured or the way it was marketed, etc. Purser and Pasmore (1992) clarify just how different variances between goals and actual results are for knowledge work:
Often, it is difficult to trace these variances to their source or to catalogue them for future reference and control. It is even difficult to recognize variances after they have occurred, because they become an accepted part of organizational life; delays, misunderstandings, and miscommunications are taken to be par for the course, and consequently are ignored by people when they discuss what can be done to improve the system ( p. 53).In a way knowledge work is nothing but variances and exceptions. So since the work is nonroutine and nonlinear it is difficult to set up discrete work processes and specify measurable goals where the team can have control of the processes, variances and exceptions.
In our research, we found many examples of the nonroutine nature of knowledge worker work. The following quote from a worker who provides a quality assurance function for a manufacturing facility is an example.
Interviewer: "Describe the steps in your work process?" Interviewee: We receive a "contract. . . . it's assigned to the individual . . . that has that . . . skill or capability of administering [it]-- contract's reviewed, you go through, you look for what's required. Uh, the degree of effort . . . shifts . . . according to . . . what type of item it is. . . . If it's an off the shelf, then normally, you're going to do just cursory review . . . until such time that the [plant] notifies you that he has this item ready for your inspection, . . . and the other extreme . . . you'll probably be doing a lot of in-plant look-see evaluation, . . . the work load is dictated by the requirements of the contract, the work effort is dictated by that. And they'll be degrees of it all through there. Uh, as to how you know when your job's done, when you ship the product, . . . when it satisfactorily passes whatever's required: inspections, tests, whatever, and you actually ship the product to the customer, and the customer receives it and it's satisfactory.So the work process is emergent depending upon the needs of the contract and what happens during the inspection processes. The work is nonlinear and goals may shift over time.
Technical professionals usually experience an intensive and directed acculturation period through the education and training in their specific disciplines (e.g. computer science, engineering, etc.) or areas of expertise (e.g. quality control, purchasing, etc.) (Purser & Pasmore, 1992). These professionals come to value specific styles and approaches to work, such as doing the majority of the actual work alone, and they have come to expect a certain amount of autonomy or lack of close supervision. To ask technical professionals to work together and teach each other about the work they do requires a substantial change in their beliefs and practices. One of our interviewees tried to explain the difference between knowledge work and manufacturing:
Right now we work on our own individual tasks. Occasionally, we'll have two people together working on a task, but that doesn't often happen because our tasks are small. While on the [manufacturing] floors, everybody is trying to build this product. And if you pass something to me and there is a problem with it, we'll need to work it out, and I don't want to pass the problem on to the next person, because it is a conveyer-belt type of thing. So there is more teaming on the floors because they're trying to come up with process improvement.
Individuals who do knowledge work often do not see any other way to work. Setting up SMWT in manufacturing adds to workers' autonomy as they become responsible for entire segments of work while in knowledge work settings autonomy may be reduced. Variety and Accountability
Also technical professionals are already handling many management functions that are not part of traditional manufacturing work. Technical professionals are usually responsible for such tasks as scheduling, planning, problem solving, and interfacing with other areas of the organization. Manufacturing teams take over some management functions to add variety to their work and move more accountability down to the workers. Since most of the work that technical professionals already do is nonroutine, most would probably like to reduce the amount of variety and accountability in their work (Purser & Pasmore, 1992). Incentives
Rewards and compensation structures are usually changed in manufacturing organizations to include rewards at the team level in order to encourage change in workers' behaviors and values. Technical professionals, on the other hand, are not as motivated by extrinsic incentives. They tend to value more intrinsic rewards such as getting to work on particular projects and achieving particular outcomes from their work. They are usually more loyal to their discipline and concerned with how it determines performance standards than the particular organization to which they belong. (Purser & Pasmore, 1992)
According to the manufacturing work team model, team members should cross-train in order to better understand all aspects of the team's functions, to provide more variety in the work, and to set up continual process improvement. As mentioned above, knowledge work usually requires a lengthy period of education and/or training. This almost prohibits workers from different disciplines learning each other's work. An example from our research involves individuals on the same team from three different functional areas. One of the functions has to attend school and be certified for particular types of work, another has to understand manufacturing processes, and the other has had extensive training in legal contracts. So how do these groups cross-train and to what extent is it possible or desirable?
When workers can cross-train then the loss of one of the team members is not as noticeable. Knowledge workers, on the other hand, have specific skills that cannot be easily replaced by other members. The boundaries between the work of managers and technical professionals is not as clear as it is with production workers. Knowledge workers may be needed to do special staff work assignments for upper management. This creates problems for the rest of the team members. A worker who has a certain skill has been recruited by management without consulting the team or their supervisor. The worker told us,
the day I found out it was going to take me a month of isolated time, no one [on the team] was here to talk to, so I wrote a note and put it on the door. . . . Well, the team hit the roof. They called the supervisor, told him, they did not want me on this project, I didn't get their permission, and I was needed over there to do my job. So the supervisor calls me up, and he chews me out, he says I don't think . . . you know what your mission is.
During the training period for a technical specialty most individuals learn a specific way of thinking and they learn to pay attention to only certain aspects of the environment. They also tend to filter the information they do perceive so that it fits with their paradigms of understanding and action. In other words, they have their own language, beliefs, and behaviors and getting individuals from different disciplines to work collaboratively often results in a clash of cultures. One of these "cross-cultural" work teams in our study was together for about three months before they realized that the acronym CAR they were all using had different meanings. One group monitors manufacturing processes and they initiate "Corrective Action Requests" and another group produces a "Contracts Automated Report" and for three months they all confidently referred to the CAR not knowing there were two different meanings. The interviewee said one of the skills he has developed since moving to teams was "learning another language."
But not all knowledge work is the same and some may adapt to teams more easily than others. There are varying degrees of difficulty in implementing culture change and "additive changes, which do not compete with traditional beliefs or practices" are easier, while "the more difficult change to motivate is that which replaces something in an established culture" (Naylor, 1996, p. 154). For example in our research sample, the programmers who have changed to work teams appear to be finding it easier to accept than other professionals who do other kinds of work. Most programming work requires some amount of teamwork to handle the complexity found in software development projects. So moving to work teams for programmers is more an additive change than a replacement and, therefore, easier to accept and fit into the existing system.
As we have now seen, borrowing the model of work teams from manufacturing appears to require some significant adjustments for knowledge workers. Work teams have not been easily implemented in all manufacturing organizations either. In fact, the greatest successes in manufacturing have been achieved in "greenfield" sites where the manufacturing facility is newly formed and the workers and management are brought on or hired to be in teams. There has been less success with "brownfield" sites where the manufacturing facilities try to redesign their organization to include work teams after years of being in a traditional structure (Pava, 1983). In a greenfield site workers have more choice about the culture change and the organizational environment is more supportive. The most effective and less stressful knowledge worker team in our study is in an organization that has been extensively using teams for about five years in various parts of the company. So the organizational culture has begun to incorporate these new values, beliefs and behaviors as part of their learned ways (Jordan, 1994).
Teams, in general, often experience problems of fit with the larger organization especially when they are only formed at the lower levels or only in certain areas while the rest of the organization remains in a hierarchical structure. This lack of fit with the whole organization is most obvious in the teams in our study who were formed without a supervisor. One of these teams has had problems with the organization's Human Resources Department (HR) supporting the team's efforts to take over management functions. One interviewee said
The first discussion you're going to have before you ever get to whatever the issue, especially with corrective counseling, is "Well, should teams be doing corrective counseling?" And, then somebody whips out the policy and procedure book that says [it should be the] supervisor. . . . I mean, you always have to get past that first. . . . Why put in place . . . an entity that's not supported through your HR practices?
Besides HR, other departments and people within the organization experience confusion and frustration when they interact with "teams" rather than a "manager" because the expected protocol has changed. A team members says,
There are people that don't really understand teams . . . they look at a team and they say, "Well, that's good they can do their job like that. But there's no manager in that group, so therefore there's no equivalent to me. If I'm a manager . . . I don't really know which one to go to, so I'm going to go to their Director.
New ways of interacting have to be formulated before a whole team can be incorporated into the system and treated as equivalent to a single manager.
Teams also require more time to work through issues and make decisions than an individual manager would. One team member complained, "We could have done this in an hour. Why do we have to do a teaming thing, you know? We've wasted 4 hours because we're a team, you know, trying to make the decision." Even though there has been much evidence that teams can make better decisions, many team members as well as the rest of the organization are impatient and are not convinced that members need to spend the time learning new ways of working.
The work from Sociotechnical Systems Theory discovered the importance of the two subsystems of an organization, the technical and the social. STS has been significant in its effort to make us aware of the importance of alignment of these subsystems and how interconnected they are since
any suggested change in a core area represents a potential threat to the group. . . . Change in these areas will be much more difficult to achieve than changes in areas not so important to survival. It can also be expected that a change in a core area will occasion consequences to other core areas, for core areas are closely interrelated (Naylor, 1996, p. 21).Purser and Pasmore (1992) have now come to believe that in the case of knowledge work there are not two separate subsystems. Instead "the nonroutine characteristics of knowledge work are such that the boundary between the social and technical systems are not interdependent and correlative, but interpenetrating and highly confluent" (p. 77). So the types of changes that are required for knowledge workers may be even more difficult to achieve than those in manufacturing.
Another reason why teams are resisted by many knowledge workers may be due to the "shop-floor heritage" (Pava, 1983, p. 42) of work teams. In manufacturing, work teams are usually promoted based on one of their most significant benefits, increased worker satisfaction due to a greater sense of autonomy, more variety, and a sense of meaningfulness in the work. Technical professionals, on the other hand, see their autonomy reduced, the majority of knowledge work already includes plenty of variety, and a sense of purpose is more a part of knowledge work than that found in factories. The work team structure seems to be most suitable for stable work, routine tasks, and an equal skill level of all the members of the team (Pava, 1983).
The above factors support the findings from studies of culture borrowing that "material items and technological processes were generally more readily borrowed than abstract things" (Naylor, 1996, p. 55). Work team models include some aspects that fall in the category of technological processes but most are abstract concepts that call for changes in behaviors, status, and values. Traditional STS redesign for manufacturing work involved examining a depersonalized social and technical system in order to increase worker input and autonomy. Purser and Pasmore (1992) have found however, with knowledge work
the finger seems to be pointed more directly at the knowledge worker, as a professional, the knowledge worker is already supposed to know what it takes to think effectively and is supposed to be able to rise above the dynamics created by the organization or situation to get the job done . . . Knowledge workers are highly sensitive to top-down interventions that threaten to impose methods of working upon them. . . . the first rule of survival [for change agents] might be, "Don't ever tell a knowledge worker what to do or how to do it." Knowledge workers need to think things through for themselves (p. 94-95).
If organizations are going to attempt to maximize their technical professional resources they need to understand the evidence presented in this paper and what anthropologists have learned from the study of culture change. The SMWT model which began in the coal mines has been successfully adapted for use in manufacturing plants. It is now being adapted for use in knowledge work environments and there its success is not so assured. The nature of this new environment is substantially different from the manufacturing plant. The move to SMWT requires knowledge workers to give up some of the work attributes they value like autonomy and the ability to work alone. At the same time the incentives of greater variety in their work or external rewards are not appealing to technical professionals. Technical professionals need to have more say in change that affects how they work. They are trained to solve problems so let them explore why change may be necessary and then design optimal work structures to enhance their collaboration. Change then becomes voluntary instead of management directed, a philosophy the STS researchers advocated in the coal mines. Effort will also need to be put into improving group skills and understanding how to improve communication and collaboration between different subcultures within organizations. What has been presented here is yet another example of the role of technology in culture change. The manufacturing technology served as the impetus for the organizational behavior change to SMWT. This change fits well in manufacturing but less well in knowledge work settings. Technical professionals may need to create different organizational structures that fit their particular situations.
Beyerlein, Michael M.
1996 History of Work Teams. In Handbook of Best Practices for Teams, Volume I. G. M.
Parker, ed. Pp. 1-6. Amherst, MA: HRD Press.
Glaser, Rollin (Ed.)
1992 Classic Readings in Self-Managing Teamwork: 20 of the Most Important Articles.
King of Prussia, PA: Organization Design and Development, Inc.
Jordan, Ann T.
1994 Organizational Culture: The Anthropological Approach. In Practicing
Anthropology in Corporate America: Consulting on Organizational Culture. Ann T.
Jordan, ed. NAPA Bulletin, 14. Washington, DC: American Anthropological Association.
Katzenback, Jon R., and Douglas K. Smith
1993 The Wisdom of Teams: Creating the High-Performance Organization. NY:
HarperBusiness.
Lawler, E. E., III
1995 Discussant remarks for the symposium: A retrospective on work teams: Lessons
from research and practice or "back to the future." Annual meeting of the Academy of
Management. Vancouver, B.C.
Mohrman, Susan Albers, Susan G. Cohen, and Allan M. Mohrman Jr.
1995 Designing Team-Based Organizations: New Forms for Knowledge Work. San
Francisco: Jossey-Bass Publishers.
Moran, L. (1996, February). [Personal conversation].
Morgan, Gareth
1986 Images of Organization. Beverly Hills, CA: Sage.
Naylor, Larry L.
1996 Culture and Change: An Introduction. Westport, Conn.: Bergin & Garvey.
Orsburn, Jack D., Linda Moran, Ed Musselwhite, John H. Zenger, and Craig Perrin
1990 Self-Directed Work Teams: The New American Challenge. Homewood, Illinois:
Business One Irwin.
Pava, Calvin H. P.
1983 Managing New Office Technology: An Organizational Strategy. New York: The
Free Press.
Purser, Ronald E., and William A. Pasmore
1992 Organizing for Learning. In Research in Organizational Change and Development,
Vol. 6. W. A. Pasmore and R. W. Woodman, eds. Pp. 37-113. Greenwich, CT: JAI Press.
Trist, E. L., and K. W. Bamforth
1992 Some Social and Psychological Consequences of the Longwall Method of
Coal-Getting. In Classic Readings in Self-Managing Teamwork. Rollin Glaser, ed. Pp.
2-42. King of Prussia, PA: Organization Design and Development, Inc.