Kenneth D. Mackenzie

Kenneth D. Mackenzie (born 1937) is an American organizational theorist, former Professor at the University of Kansas and management consultant. He is known for his work on his early work on the "Theory of Group Structures" and his later work on organizational design

Biography

Mackenzie received his BA in Mathematics with a minor in Physics in 1960 from the University of California, Berkeley, where in 1964 he also obtained his Ph.D. in Business Administration.

After his graduation Mackenzie started his academic career in 1964 at the Carnegie-Mellon University as Assistant Professor of Economics. In 1967 he moved to the Wharton School of the University of Pennsylvania at the University of Pennsylvania. In 1972 he moved to the University of Kansas, where he was appointed Edmund P. Learned Distinguished Professor from January 1972 to January 2006. He further taught at U.C. Berkeley, University of Waterloo, and KU. In 2000 Mackenzie founded the consultancy firm EMAC Assessments, LLC.

Mackenzie has served on numerous editorial boards including Management Science, Organizational studies, International Journal of Organizational Analysis, Journal of Management Inquiry, Human Systems Management, and Engineering Management Research.

He has published 19 books and over 100 articles.

Work

Mackenzie research interests have been in the fields of organization theories, organization design processual models, organizational leadership, multi-level research and the discoverer of the organizational hologram.

Approach to Research

Mackenzie’s research is based on his bed rock belief that there are only three fates of any theory: to be (a) ignored, (b) rejected, and (c) revised. Consequently, it is self-evident and unreasonable for him to expect any of his theories to escape these fates. Therefore, he is not interested in “proving” or justifying them. He knows that they somehow must be wrong. What he chooses to publish is his current best effort but only after finally failing to find grievous fault. The goal is to improve theories rather than to defend them. The best way to do this is to actively search for and highlight hidden flaws and inconsistencies. Finding these flaws provides vital clues for making improvements in the theory. Fixing them paves the path to improvements.

To this end, Mackenzie never begins any research project if he already has a clear understanding of how to solve it. What would be the point? In fact he only works on projects for which he does not know the results. This way he is frequently surprised and grateful to witness new wonders. Surprises refocus thinking and provide information about how one of his theories is wrong. Surprise stimulates the necessity for more improvements. The constant goal is seeking disconfirmations. The motto is “Always Failing Forward!” Obtaining a confirmation adds no new information, does nothing to improve a theory, and uses up scarce resources. Confirming what is already known is thus viewed as an ineffective strategy for theory improvement. But a disconfirmation is a stimulus, a call to action which impels one to stumble forward in order to seek improvements. One succeeds by failing.

Mackenzie is keenly interested in discovering and enhancing explanations of patterned behavior. Thus, he values theory. He is especially interested in the patterns of behavior within groups and organizations.

His approach to research is both deliberate and naïve. There are steps he follows in performing an empirical or conceptual study. However, he tries to ignore any latent preconceived specific theory or a fixed toolkit of methods. To him the phenomena is paramount and not the literature. He always wants a fresh, independent look at a problem whenever possible. The first step is to “wallow” in the phenomena until reaching an intuitive feel for them. (Please note: to wallow in the phenomena is not the same thing as wallowing in the literature about the phenomena). Next is the effort to formulate the research questions, taking care to always to be “anchored” in the phenomena. This formulation provides a narrative and means to begin the development of some explanation for the phenomena. Once an explanation is developed, the question becomes: how can this be refuted? The next step is to refute it. This is the easiest step! This basic research process iterates until one runs out of new ideas for refuting the explanation. Stopping to write up results is the next step. These developments are merged into a wider stream of his past research work and where feasible, the relevant literatures. Because of the importance he places on counterexamples, he eschews the use of mathematical statistics in favor of strong inference. His purpose is to use each research adventure to help erect a less ramshackle theoretical edifice for the group and organizational sciences.

He usually works on many questions simultaneously and moves quickly to reject unpromising avenues of research. He strives to avoid the Type III error of working on the wrong problem. Ideas pop up all over and the trick to give each a chance and then discard the bad ones. In the pungent language of hunters, “do not feed the dog that won’t hunt.”

Mackenzie sometimes works as a management consultant in order to hear of new problems and thus new questions and to fund his research. However, he never accepts a consulting project if he thinks that he already knows the answer. He does not offer recommendations for organizational changes unless the client agrees he knows enough to run the organization. He is always an advocate for the well-being of the organization itself.

Representing Group and Organizational Structures and Processes

As a researcher, a teacher, an author, and an editor, he frequently encounters the difficulty of how to represent fundamental concepts such as leadership, structure, and process. In particular, the representations of structure and group and organizational processes are especially interesting and important because they cut across the group and organizational sciences.

In this regard, he has learned two things worth mentioning produced over a course of many years of research into organizational phenomena: the representation of group and organizational (a) structures and (b) processes. Both concepts are central to discussions of group and organizational processes (GOPs). And neither is represented consistently among scholars and practitioners. The result is confusion and barriers to cumulation. For how can one build on another’s results when the other’s methods and concepts are inconsistent with one’s own?

Representing Group and Organizational Structures

A structure, S_n, of a group or organization of n agents, X_n = (x₁, x₂, …, x_i, …, x_n) can be represented by this equation:

S_n = (X_n; R) (Equation 1)

where the matrix R has n rows and n columns and entries r_ij where row i corresponds to the “sender,” x_i , the column j corresponds to the “receiver,” x_j , and the value of an entry r_ij is a measure of an interaction from x_i to x_j. The value of r_ij can range from a binary relation such as a boss-subordinate to how many thousands of board feet of douglas fir, 4-side finished, eight foot 2 x 4’s were sold by wholesaler i to retailer j during a specified time period.

The x_i ϵ X are engaged Processual Agents such as individuals, groups (e.g. committees, task forces, virtual teams, etc.), and even organizations. The dimension and measures for the entries, r_ij , are chosen for the purposes of an analysis. It is important to realize that in this representation of group and organizational structures, the entries r_ij reflect the actual relationship between its members relevant to the purposes of the study. Please note the asymmetry: r_ij rarely equals r_ji . Also, in some applications, not all of the Processual Agents x_i ϵ X are, in fact, human.

Given this representation of group and organizational structures, the following conclusions have strong empirical support:

1. Structures are more accurately viewed as effects rather than causes of behavior.
2. Different tasks may have different structures.
3. Groups and organizations have multiple structures.
4. These structures are interdependent.
5. Structures can and do change.
6. A structure represents a need satisfying interaction pattern. As needs change, so will the structures to satisfy them as long as there is net benefit for the change.
7. These structural change processes can be modeled and explained.
8. Most group and organizational behavior takes place “outside” of the formal lines of authority. Departures from the organizational chart are normal.
9. Processes and their structures are independent.

Representing Group and Organizational Processes

Any group and organizational process (GOP) can be represented as:

Y = F(C) (Equation 2)

where C is a vector of considerations or steps in a GOP, F is a network illustrating the linkages between each of the considerations or steps, and Y is the set of outcomes of the GOP.

Intuitively, a GOP is a time dependent sequence of behaviors governed by a process framework given by equation (2). The GOP representation of equation (2) is derived from these six ontological axioms:

1. A GOP involves one or more processual agents.
2. A GOP involves two or more linked process elements.
3. A GOP is not random.
4. A GOP is linked to at least one other GOP.
5. A GOP representation requires process resources and involves their characteristics-in-use.
6. A GOP has more than one level.