Eric B. Dent, Ph.D., Assistant Professor, The George Washington University.

Sampling of recent previous publications.

ABSTRACT

One of the frustrations of working in the exciting area of "complexity science in organizations" is that there is no commonly accepted definition of what this term means (White, et al, 1997). Definitions have been offered, such as "complexity is a watchword for a new way of thinking about the collective behavior of many basic but interacting units... complexity is the study of the behavior of macroscopic collections of such units that are endowed with the potential to evolve in time" (Coveney and Highfield, 1995, p. 7). Although this definition is very descriptive it still seems general and unfocused. The purpose of this paper is to offer a simple definition for complexity science and to demonstrate the shift in worldview necessary for complexity science to become as second nature to people as traditional science now is.

Simply put, complexity science is an approach to research, study, and perspective which makes the philosophical assumptions of the emerging worldview (EWV) (these include holism, perspectival observation, mutual causation, relationship as unit of analysis, and others - see Table 1.). Classical science, as practiced in the 20^th century, for the most part, makes the philosophical assumptions which will be labelled here the traditional worldview (TWV) (which include underlying assumptions of reductionism, objective observation, linear causation, entity as unit of analysis, and others).

This TWV, which has allowed people to make significant achievements in many fields, is no longer serving as a reliable guide. Several brief examples illustrate the dysfunctional nature of TWV assumptions applied inappropriately.

Rent control laws which were intended to maintain a stock of low-cost housing have resulted in a shortage of low-cost housing.

The demise of the Saturday Evening Post and the Curtis Publishing Company has been attributed to "management essentially look[ing] for short and direct cause and effect linkages" (Jacobs and Jaques, 1987, p. 34). Computer simulations have suggested that this company could have been saved if a strategy which incorporated complex, indirect linkages had been employed.

"The largest building in the world, the space vehicle preparation shed at Cape Kennedy, generates its own weather, including clouds and rains. Designed to protect space rockets from the elements, it pelts them with storms of its own" (Gall, 1977, p. 20).

Sick people go to the hospital to be made well. Twenty percent of all patients, however, acquire illness in the hospital as a result of their diagnostic procedures and treatments prescribed (Illich, 1977, p. 23).

The rise of complexity science has paralleled an increase in dissatisfaction with the TWV. Capra (1982, p. 15) labels this dissatisfaction a crisis of perception and says it occurs when people hold to a mental model which no longer achieves their standards of accuracy. Other writers have called this same phenomenon a period of dislocation (Ackoff, 1981) or a time when we are between "stories" (Schwartz and Ogilvy, 1979). We do not yet know exactly what the new story will be. It is easier to see where we have been than where we are going. Consequently, the problems and dilemmas which have arisen are easier to critique than the specific details of a new worldview are to provide. Examples of these difficulties are TWV assumptions that work within a range of conditions, but beyond that range they no longer work.

Some of the underlying assumptions of the shift in worldview are becoming clearer. A difficulty in capturing the TWV and EWV underlying assumptions, though, is that the worldviews cannot be simply stated. One can use simple metaphors like the clock and the waterfall, but these do not capture the full essence of the worldviews. Table 1 contains a list of a number of differences in underlying assumption gathered from a variety of sources.
 

Table 1.
Emerging and Traditional Worldview Descriptors

Most readers of this study have been taught in a learning paradigm so that they are more comfortable with the information presented in the form of Table 1 (Vaill 1996). However, Figure 1, which still has limitations, is a more accurate visual representation of the differences in TWV and EWV underlying assumptions, for reasons discussed below. Including all of the information in Table 1 in Figure 1 would overwhelm the visual representation, so only the three constructs that best differentiate worldview (Dent, 1997) are presented. For clarity of understanding, the word construct is used to denote a phenomenon such as causality. The word assumption is used to indicate a selection within a construct. So, for the construct causality, the two assumptions labeled are mutual and linear.

Taking the Traditional Worldview "Out of Range"

It is important to note that theorists are not suggesting that the traditional underlying assumptions are wrong. In fact, many of them seem to be useful in localized settings. For example, Prigogine and Stengers (1984) see determinism and indeterminism not as irreconcilable opposites but "each playing its role as a partner in destiny" (p. xxiii). Between bifurcation points, determinism is operative. At a bifurcation point, however, indeterminism takes over. Consequently, indeterminism (which doesn't dismiss localized determinism) and the other emerging assumptions seem to be more useful abstract concepts. They reflect reality more accurately in a larger number of instances. Capra (1982) nicely captures the distinction.

Ken Wilber (1995) uses the term fractured worldview to describe the part-right, part-wrong feature of the TWV.

Problems also arise when people assume the TWV is accurate in all settings. Although it is inappropriate, and potentially inaccurate, researchers frequently use linear regression on non-linear phenomena, calculus on discontinuous functions, or chi square when data points are interdependent (Dent, 1994). Priesmeyer (1992, p. 30) has speculated that traditional statistical methods remain useful for systems that are nearly stable. Classic problem solving techniques make perfect sense when reductionism can be assumed. If a single problem can be solved completely independently of everything else in the system and its environment, problem solving is an ideal strategy. However, when interdependencies are present, problem solving becomes less effective.

The comprehension and control model of management makes perfect sense in a relatively stable environment. However, the Relaxation Time Principle has shown that "system stability is possible only if the system's relaxation time is shorter than the mean time between disturbances" (Clemson, 1984, p. 213). In other words, if an organization experiences changes more rapidly than it can comprehend and control them, then it is not possible to keep the system stable. A similar example is provided by Karl Weick (1985, p. 110). He describes the decision-making style of the TWV as rational. Rational decision making is effective in organizations that are in environments which change slowly, have few social groups, and have centralized authority that works reasonably well. Weick observes that these conditions are now relatively rare in organizations.

Consequently, some aspects of the EWV are simply "enlargements" of the TWV. McKelvey (1999) notes that "since the [EWV] does not require axiomatic reduction, it tolerates multiple models. Thus, "truth" is not defined in terms of reduction to a single model. ... That they also have different theoretical explanations is not considered a failure. Each is an isolated, idealized physical system representing different aspects of real-world phenomena (p. 19). Perhaps the most useful mental model for thinking about the TWV and EWV is that of a polarity (Johnson, 1992). Polarities are sets of opposites which cannot function well independently. The two sides of a polarity are interdependent, so one side cannot be "right" or the "solution" at the expense of the other. Johnson contends that, "many of the current trends in business and industry are polarities to manage, not problems to solve" (p. xi). An example of a polarity in worldview is, rather than replacing yang dominance with yin dominance, the EWV includes a balance of yin and yang, not subordinating the yang. Likewise, the example provided earlier suggests that indeterminism and determinism form a polarity. The question of behavior emerging from the bottom up or being imposed from the top down form a polarity.

Each side of the pole has upsides and downsides. A "figure 8" pattern often develops between the upsides and downsides of the two assumptions. People often identify the downside pole as the "problem" and therefore want to abandon it. The upside of the opposite pole is seen as the "solution." When one pole has been emphasized for too long, the result is the downside of both poles (p. 156). In terms of polarities, the shift called for in this paper is from a focus on a single pole (the TWV) to a focus on both poles (the EWV). A graphical representation of a polarity is depicted in Figure 2.

Although a juxtaposition listing such as Table 1 earlier may create the implication, because of instances of synthesis and polarities, these differences should not be pictured as a continuum with the TWV at one end and the EWV at the other. It is more accurate to say that there is a complementarity in the items. In some cases, one is an enlargement of the other, in some they are primarily distinct, and in others there is some overlap.

It is also important to recognize that a breakdown in the TWV does not automatically mean the ascendance of the EWV. A manager, for example, could be totally frustrated by hierarchical structure but not know with what to replace it. And, if one gives up a belief in survival of the fittest, she does not necessarily embrace adaptive self-organization. In this case, there are other alternative concepts about structure.

Organizational Phenomena based on EWV Assumptions

In The Fifth Discipline, Peter Senge (1990) includes mental modeling as one of the five disciplines. He suggests that people must be able to surface mental models by sharing the assumptions they make in a situation. This task is not trivial. Most mental models are so deeply imbedded that people do not even realize they are simply models; we believe they are reality. Mental models are critically important. How we see things determines much of what we see. Consequently, a change in worldview from TWV to EWV would result in major changes in how organizational activity occurs. In this section we will include three examples, one for each of the underlying assumptions which best differentiate worldview: mutual causality in strategic planning, holism in mess management, and perspectival observation in performance appraisal.

Mutual Causality in Strategic Planning

Organizations often assume linear causality. For example, a housing organization which institutes rent control expects the direct result to be low-cost housing. Such officials have not realized the feedback from such a policy. Such feedback consists of developers who will refuse to build additional housing units subject to rent control, landlords who are forced to allow properties to deteriorate because of below-market compensation, and apartment dwellers who may refuse to move to a location with better job opportunities because of the desirability of such low-cost rent. As in this example, when organizations unrealistically assume linear causality, their policies often bring about exactly the effects they were trying to guard against (Begun, 1994, p. 330).

An organization that fully comprehends the effects of mutual causality will engage in strategic planning in a way completely different from traditional approaches. Mark Michaels (1994) has pointed out that the strategic planning process as typically implemented "involves predictions about future events, predictions which the dynamic of sensitivity to initial conditions - the butterfly effect - prove unreliable" (p. 17). Karl Weick advocates "real-time" (or just-in-time) strategic planning. Weick argues that acting should precede planning because by acting we take part in constructing the environment. The environment is not "out there" separate from us. We can help to create the environment. Weick contends that "we create the environment through our own strong intentions" (Weick, 1995). The Spanish have a phrase which nicely captures this connotation. "Compaûero, no hay camino. Se hace camino al andar." A suitable translation is, "my friend, there is no road. You make the road as you walk."

Michaels incorporates the idea of feedback by noting that a strategic plan should be a statement of purpose "of the company's moral response to its broadly defined responsibilities, not an amoral plan for exploiting commercial opportunity" (1994, p. 17). This perspective honors the multiple sources of interconnections that develop over the lifetime of an organization. Weick and Michaels place much more of an emphasis on the present than traditional strategic planners do. Michaels even highlights the importance of the past. His three step process of strategic planning is (1) creating a shared past; (2) defining the present; and (3) steering into the change.

This view of strategic planning is very different from the traditional process which includes developing a vision, a mission, identifying stakeholders, and doing a SWOT (strengths, weaknesses, opportunities, and threats) analysis. This type of analysis assumes that the environment presents opportunities and threats, not that the organization is an active player in creating opportunities and/or threats. Priesmeyer (1992) adds that the traditional strategic planning model is inaccurately simplistic because it "suggests that one can understand the state of the system by assessing current conditions, when in fact an understanding of evolving conditions is important" (p. 195).

Holism in Mess Management

Perhaps the most radical example of holism in practice in organizations is Ackoff's call for an end to problem solving. Ackoff contends that many of the problem solving approaches used in organizations are not effective. His argument is similar to that of Senge's designer role for leaders. Senge (1990) believes that problems should be prevented by proper design. Ackoff would not argue with that, but would add that when anomalies do occur, they should be managed as part of the regular course of things, rather than having a task force convened, or an employee assigned to work on a particular problem.

According to Ackoff, "this whole way of thinking encourages us to focus attention upon bits and pieces of our organizations and thereby leads us to adopt policies and carry out actions that as often as not make the original situation worse" (Clemson, 1984, p. 171). It is rare in organizations that a problem can be isolated so that a fix can be implemented without also altering something else in the organization. Ackoff advocates "mess management," his term for the continuous balancing and navigating of complex, interrelated messes, rather than problems, that most people in organizations face.

Ackoff lists several problems with problem-solving. For example, in many cases the complexity of the problem exceeds the problem-solving expertise of lower-level employees often assigned to "tiger" teams, task forces, or other-named ad hoc problem solving groups. Also, assigning a task force to study a problem and recommend a solution assumes that while the task force is spending time working on the problem, the problem is not changing (Ackoff, 1981, p. 4-5). Anyone who has worked in an organization has the experience of a tiger team coming up with a recommended solution which is not ultimately implemented. Ackoff would suggest that the primary reason is that the tiger team did not take into account the whole - the complete set of interdependent relationships within a given executive=s purview. Mess management requires the executive, who has the responsibility for handling all of these interdependencies, to manage any problems that arise within his natural, normal processes.

Perspectival Observation in Performance Appraisal

In our research, we have discovered that the English language contains a number of rich expressions which convey an appreciation for perspectival observation. These include: where you stand depends on where you sit, beauty is in the eye of the beholder, everyone looks at the world through his own glasses, the glass is half-empty or half-full, a self-fulfilling prophecy, it's the blind man and the elephant, the Rashomon phenomenon, the umpire training school joke about "the pitch ain't nothin' until I call it," and that there are two sides to every story. Such a broad set of common expressions would lead one to believe that perspectival observation is a widely held assumption in society. Paradoxically, our research and experience within organizations has been the opposite. We have interviewed individuals, for example, who were perfectly willing to accept perspectival observation about what happened at an extended family Thanksgiving dinner, but who would insist that at worked there is only one true story of what really happened.

The prevailing view of performance appraisals in organizations is based upon the assumption of objective observation. This dominant TWV is expressed in statements such as, "performance measurement is typically the source of many problems in appraisal because it is seen as subjective" (Cummings and Worley, 1993, p. 403). Subjectivity is assumed to be problematic. An entire industry, led by the Hay Group, is devoted to instituting objective performance appraisal systems into organizations. In summary, the two forms of objective performance appraisal predominant in organizations today site the objectivity either within the manager alone, or in quantifiable metrics such as, number of lines of computer code written, number of academic papers published, or projects completed on schedule and within budget.

Those who assume perspectival observation contend that performance appraisal cannot be objective. For example, for only the simplest of jobs can individuals be given performance objectives that are completely within their control. If the workplace is interdependent, employees are often independently held accountable for the functioning of interdependencies which are operative in the completion of their work. Deming (1986) and the quality experts question objective performance appraisal from another perspective. They argue that it is impossible to define a subset of performance measures which can encompass the full set of behaviors an organization wants from its employees. Empirical research suggests that managers are not capable of reliably evaluating performance over time (Atwater and Yammarino, 1992).

A technique for incorporating the assumption of perspectival observation into the performance appraisal process has recently come in to vogue. This technique has been labelled 360-degree job evaluation, multi-rater performance appraisal, team-based pay, and others. It is based upon the assumption that no single person or collection of metrics can best reflect an employee=s performance. Many organizations are using this technique only as a way of providing feedback to an employee. The employee=s subordinates, however, do not have a say in his or her salary increase or promotion (Antonioni, 1996). Many other organizations do use subordinate appraisals to determine a manager=s raises and promotions (McEvoy, 1987). Motorola bases 20 percent of an employee=s pay on input obtained from peers. They intend to increase this percentage to 50, and contend that peer review for pay has been a major factor in a productivity boost of 126 percent over seven years (Swoboda, 1994). These multi-rater techniques suggest that "reality" is best articulated as a collection of a number of different view points.

Summary

An individual=s worldview may be a major determinant in her success as a practicing manager. Complexity science opens up a whole new vista of perspectives, approaches, and techniques because it is based on a set of underlying assumptions which differ from classical science. Managers need to adjust their mental models to ones that are more useful in accomplishing work. People have been operating with mental models which have not allowed them to achieve the results they have desired. We as inquirers are changing as observing systems. Just as the telescope and microscope revolutionized the way people constructed reality, the computer is having a similar effect today. These tools of intervention are our new sensory organs. Our reality changes as our ability to detect phenomena changes.

While the nearly exclusive emphasis of measurement and quantification has resulted in phenomenal knowledge in the past several centuries, we may be near the peak of the mountain represented by the natural phenomena that can be explained by separate and distant inquiry. We may have passed the peak for organizational phenomena. Research is being conducted to determine whether or not worldview distinguishes between successful and less successful managers. Many management education programs need to be changed to teach more holistic, perspectival, and mutually causal mindsets. Although changing mental models is often difficult, such flexibility is necessary in the demanding, global marketplace of today.

Bibliography