Review:
Logic of Leadership Research: A Reflective Review of
‘Geeks & Geezers’ by Bennis and Thomas
Geeks & Geezers: How Era, Values, and
Defining Moments Shape Leaders. Book by Warren G. Bennis and Robert
J. Thomas. Published by Harvard Business School Press, Boston,
Massachusetts, 2002, 224 pp., ISBN: 1 57851 582 3, (Amazon Price in
February 2005: $17.79).
Suggested Citation: Dash, D. P. (2005). Logic of leadership research: A reflective review of ‘Geeks & Geezers’ by Bennis and Thomas. Journal of Research Practice, 1(1), Article R1. Retrieved [Date of Access], from http://jrp.icaap.org/index.php/jrp/article/view/2/4
Leadership has turned out to be one of those topics, which persistently slip out of scientific hands. The field of leadership research should be of interest to researchers in various other disciplines, because it serves as a source of examples of many common difficulties faced by researchers in general. These relate to difficulties in defining a research task, specifying quality criteria, choosing methods, ensuring that the research programme remains progressive (the criterion is from Lakatos, see Science and Pseudoscience, 2004; Worrall & Currie, 1978), etc.
The book by well-known leadership researchers, Bennis and Thomas, gives us an occasion to critically appreciate the practice of leadership research so far and assess the book’s potential contribution. This will be done by first outlining the developments in leadership research since the 1930s. It will be shown that although the book’s focus is interestingly different, it does not go so far as to reframe the logic of research in the field.
Here is an abridged account of the developments in leadership research since its initial days. As it can be seen from the following account, the field has steadily changed its focus in search of the right topic.
Although leadership is commonly defined as the ability to set goals in risky and uncertain situations, and the ability to influence a group of people to achieve those goals, it remains by and large an elusive phenomenon. Daily life experiences have persuaded people to look for the secrets of effective leadership. Presented here is a brief sketch of the pathways taken by leadership research since the 1930s (for a more detailed review, see House & Aditya, 1997; for book-length treatments, Northouse, 1997; Yukl, 2002).
The initial approach to leadership research has been to look for personal characteristics that mark out leaders from non-leaders, producing the so-called trait theories. Such theories proliferated, suggesting different combinations of personality characteristics, skills, and competencies such as self-confidence, need for achievement, interpersonal skills, ability to learn, etc., as key to leadership. One review (by J. G. Geier) in 1967 already identified as many as 80 traits reported in 20 different studies, without of course any trait clearly emerging as universal (Robbins, 2003, p. 314).
Eventually, a number of difficulties with this approach came to light. Many of the traits were found to be unstable, i.e., an individual’s traits changed over long durations of time. No one-to-one correspondence could be seen between traits and actions; leaders’ actions depended on the situations they were in and the roles they were playing in it. The focus on traits triggered the question of whether leaders were born or made. Trait theories continued to grow, as they do even today, e.g., theories that focus on emotional intelligence or multiple intelligences. However, difficulties like the above led some leadership researchers to change the focal topic of leadership research.
The next group of theories to appear in the leadership field were the behavioural theories. The focus of these theories was on finding patterns in what effective leaders do (rather than what traits or skills they have) that produce subordinates’ satisfaction as well as high performance. Notions like task-oriented, person-oriented, development-oriented behaviour, etc., marked these theories. However, once again, no pattern of behaviour could be found to be universally associated with high satisfaction and high performance of subordinates. The effects of the context within which a leader operates, the specific role the leader plays, and the disposition of the subordinates were all found to be important in guiding leader behaviour.
The shift of focus from traits to behaviour happened under a persistent demand from people in organisations and communities whose expectations were not sufficiently fulfilled by theories aimed merely at explaining effective leadership. They were apparently more interested in improving the effectiveness of the leadership available to them. Researchers assumed that understanding leader behaviour might help improve leadership through behavioural training. Although the behavioural theories came with this promise, the experience with behavioural training in leadership was rather mixed. It was not clear, if any particular behavioural patterns could be recommended to a specific leader.
Gradually, leadership researchers began to doubt the very possibility of a general theory of leadership, or any general prescription, which could be used to improve the quality of leadership in diverse settings. The multiplicity of factors influencing leadership success, and their complex relationships, suggested a different focus for research in this field. This was the beginning of the so-called contingency theories. As the name suggests, these theories also specified the boundary conditions within which they would apply. Some of these theories still spoke of traits or behaviours, but not in the same generic manner as the earlier theories. For example, a contingency theory called Cognitive Resource Theory (proposed by F. E. Fielder and J. E. Garcia in 1987, discussed in House & Aditya, 1997) seeks to specify the interaction between some characteristics of the leader (as a person), namely intelligence and experience, and some characteristics of the situation, especially the level of stress experienced by the leader and followers. The theory suggests that, in situations of high stress, performance is positively correlated with the leader’s experience and negatively correlated with the leader’s intelligence. The correlations are the reverse in situations of low stress. This may be seen as a modified type of trait theory, where the relationship between performance and leader characteristics is conditional to the stress level prevailing in the situation. This theory found favour with consultants and trainers because it offered possible ways of improving performance, e.g., by implementing a stress-reduction programme.
Similarly, the other contingency theories, e.g., Path-Goal Theory, Leadership Substitutes Theory, etc., sought to explain leader effectiveness under specific situations, such as level of members’ maturity, type of decision-making task, level of control possible, nature of work organisation, etc. Many of these theories led to the development of training programmes to improve leadership effectiveness. However, the more the theories were applied, the more were the doubts concerning their validity. There were practical limitations to testing the validity of such theoretical claims--laboratory-based testing was not realistic and field-based testing had to rely on self-reports, which were not always dependable. The models that translated the theoretical propositions into practical prescriptions, introduced significant uncertainties of their own. Some of the theories emerged virtually untestable because of the very large number of instances required to be tested to cover all the combinations implied by these theories (sometimes running into a million and a half combinations).
In response to the above situation, different kinds of innovation emerged in the leadership field. Some researchers changed the unit of analysis by shifting focus from the leader per se to the quality of the relationship between superiors and subordinates (the so-called leader-member exchange or LMX). The broader aim now was to study how these relationships affect collective outcomes. It was found that the quality of LMX might be a result of various social, organisational, and behavioural processes, which in some cases can even produce dysfunctional consequences, such as discrimination against some members. Moreover, testing of theoretical claims became quite difficult due to the requirement of controlling for a large number of intervening processes.
Subsequently, a class of leadership theories has emerged (the so-called new leadership theories) to focus on extraordinarily outstanding accomplishments alone (thus narrowing down the scope of leadership research). These have sought to find correlations between leader’s charisma and effectiveness. Eventually, a number of boundary conditions have entered into the discussions (e.g., continuity of leader-follower interaction, environmental uncertainty, socio-cultural context, loss of charisma due to routinisation of interactions, etc.). Arriving at any generalisations has remained as elusive as before.
The above outline indicates that, despite repeated change of focus, no focal topic concerning leadership could be found around which general propositions about leadership could be developed, that could be applied and tested in a wide range of practical domains. This is what was referred to as “slipping out of scientific hands” at the beginning of this review. This state of affairs calls for an interpretation. One is presented below.
A careful examination of the above shifts in the focus of leadership research reveals that the field has steadily revised its notion of what to study in order to continue research on leadership. However, there is scant evidence of any innovative reformulation of how to study leadership in order to continue the research. The how to study aspect has continued to follow some received wisdom about rigorous research (remarkably reflecting the so-called Bacon-Galileo-Newton programme in natural philosophy; see Reany, 1995). At the level of tools and methods, it has remained by and large wedded to the notions of dependent and independent variables, intervening and moderating variables, scales of measurement, surrogate measures, correlations, etc. In other words, the research practice in this field has been governed by the so-called language of variables (De Zeeuw, 2001), producing what might be called an intellectual monoculture that discourages any modification to that research language.
Two developments can be pointed out to underline the unplanned effects of this apparently unswerving commitment to the language of variables in leadership research: (i) problem of comparison and (ii) problem of use.
The problem of comparison arises out of over-abundance of variables and boundary conditions, especially when there is precious little by way of shared concepts and definitions. Leadership theories have been prolific in proposing variables and their associations, notwithstanding that the term leadership itself has been difficult to pin down--one survey has yielded 221 definitions of the term (Rost, 1991). Additionally, there is much diversity concerning the differences between leadership and management. As a result, variables used in one piece of research are not easily compared with those used in another. Due to the boundary conditions associated with some of the theories, the number of combinations to be compared becomes so large as to be practically unmanageable. Consequently, testability and accumulation of results have both become rather precarious in the field. It is difficult to spot a consistently progressive research programme in this area.
The problem of use relates to the obsession of the field with explanations. Explanations, even when they are reliable, may not be helpful in bringing about practical improvements. To illustrate the point, explaining earthquakes does not help in improving the lives of their victims. Consequently, the everyday world of leadership has gone on without paying serious attention to what the leadership researchers have been doing. As noted by House and Aditya (1997), no mention of any leadership research was made in the issues of Time magazine (in 1988 and 1993) that published cover stories addressing the need for leadership in the US political system. The same inattention to the research literature is persistent in various public discussions on leadership, almost everywhere.
The rigour vs. relevance debate (e.g., Argyris, 1980, 1992, Chapter 21: “Some unintended consequences of rigorous research”) may be applied to leadership research. However, seeing rigour and relevance as an either-or opposition can be misleading. Moreover, when rigour of a research enterprise appears to preclude its relevance, the time may be ripe for a change in the research script in that area (Dash, 2002; Vahl, 1999). In other words, it may be time to reconsider the rules or conventions that determine the forms of inputs and outputs of the research process.
This cursory review of leadership research indicates that the production of relevant leadership knowledge at the beginning of the twenty-first century would require significant choices to be made with respect to the boundaries that define the methodological conventions of leadership research. The present review uses this idea to pose the following question to the contemporary literature on leadership: To what extent does the new literature help us reformulate what constitutes relevant leadership knowledge, and how that kind of knowledge is to be extended systematically? Upon using this question as a probe, the book by Bennis and Thomas under consideration here, appears rather ambivalent, although it might serve as an inspiration for the new things to come in this area.
Geeks and Geezers seems similar to the bulk of leadership books already in circulation, yet it is different from them. It seems similar because it starts with a list of successful leaders and looks for patterns which connect (the expression is from Bateson, 1979) their individual achievements, in order to conclude about a core set of skills and competencies that hold the key to successful leadership. However, it is also different from this genre of leadership books because it poses a slightly unusual question. Of course, as the student of leadership knows, the usual questions to ask would be like these: What qualities or actions make a leader successful? What type of leadership approach should be adopted when? The question Bennis and Thomas start with is the following: “why some people are able to extract wisdom from experience, however harsh, and others are not” (p. 2, all page numbers henceforth pertain to Bennis & Thomas, 2002, the book under review).
Their focus is not so much on the attributes observable in the sampled individuals, but on the special type of interaction between the individuals and their environments that might have taken place in the past. To unearth that, and especially to compare whether different leadership orientations would emerge in different environments, the authors took two samples--one of young leaders (less than 35 years of age, the so-called geeks) and another of old leaders (aged between 70 and 80 years, the so-called geezers). Broadly, the following kinds of question were asked to all of them: What were the defining moments in your life? How did you get from here to there? How do you define success? How did you define it at age 30? What makes you happy? What role has failure played in your life? (p.9)
It turns out that the world in which the geezers grew up, the “era of limits,” was very different from the world in which the geeks did, the “era of options,” with respect to the guiding values, shared concerns, prevalent images of success, and business practices. The sheer strides made by global business and communication technology provided a different backdrop to the geeks. “Geezers at roughly age 30 were striving to put instability behind them, while geeks were impatient to shake things up” (p. 84).
The authors discover some recurring patterns in the self-reports from their respondents. Almost all of them spoke of their past as involving at least one intense transformational experience, which was potentially debilitating. These were termed the “crucibles of leadership” (p. 87). But, as the self-reports go, the respondent emerged stronger from the ordeal by creating a new meaning, making a new connection, finding a new voice, etc.--in short, emerging as a new self. The respondents had, in all cases, crafted their narratives in which they were the heroes and their lives, “the hero’s journey” (p. 108). The authors generalise the self-reports by hypothesising the existence of something called “adaptive capacity” in those individuals (p. 91).
The book refocuses the reader’s attention on the persistent tussle between a person’s image of oneself and the possibilities (or limits) posed by the changing environment, especially at crucial junctures in one’s life. In doing so, the book encourages readers to recognise the broader social and psychosocial processes that produce leaders. It provokes the reader to consider leadership as the emergent effect of a multi-actor, dynamic, and transformational process.
On the one hand, the book questions the trait theories. “Our study confirmed our belief that traits and other individual factors are given far too much prominence in studies of leadership” (p. 91). On the other hand, the book seems to do exactly what the trait theories try to do. It seeks to identify some key individual-level characteristics that seem common among successful leaders: “To the extent that any single quality determines success, that quality is adaptive capacity” (p. 91). In fact, the book goes on to describe a few other individual-level skills or competencies. It asserts the notion that leaders are a particular kind of human being, having some special abilities. “All our leaders, whatever their age, brought to their crucibles four essential skills or competencies. These are the attributes that allow leaders to grow from their crucibles, instead of being destroyed by them” (p. 121).
This is the ambiguity the reader is left with. Are leaders specially gifted people? Or, are leaders just ordinary people (without any pre-qualification), who happen to have gone through some very special experience? The book seems to postulate an impossibility: It is impossible to know if a person can become a leader, unless the person goes through some kind of ordeal and emerges stronger from it. This can introduce a new twist on leadership research and its systematic extension.
Although the impossibility stated above appears to be a crucial insight, the book has not made it explicit. In fact, the book does not specify how the work might be taken forward in a systematic way. The authors seem to be satisfied with a set of common characteristics that would demarcate leaders from non-leaders--harking back to the traditional inclination of leadership research.
The established logic of research in this area treats individual leaders as cases of some theoretical object having stable properties (call it trait, behaviour pattern, style, gift, or competency). This is not likely to be helpful in furthering the insight one derives from this book.
If we turn our attention from individuals per se to the transformational processes causing the formation of a new self, we have to revise our logic of research in order to produce relevant knowledge about such transformational processes. Such knowledge would not merely stop at specifying properties of successfully transformed leaders, but help people in crucible-like situations to make the appropriate choices in order to emerge from it, successfully transformed.
Somehow, in any field of study, the logic of research tends to be slow to change. It takes some dissatisfaction with a particular logic, for researchers to look for (or invent) new logics and new scripts. It can take time for that kind of dissatisfaction to be sufficiently strong. A climate of self-reflexive research practice is necessary to make the right moves in good time. Leadership research today seems to lack that climate.
The position expounded in the book leaves interesting opportunities for systematic extension of the key insights, although the authors do not pursue that direction. Can adaptive capacity be taught or otherwise developed? Can the so-called crucibles of leadership be re-enacted or simulated, minimising the associated human costs? Can the narrative of the hero’s journey be developed as a method for self-guidance and self-development? The answer to such questions seems to be conditionally positive. To produce meaningful results in this direction, a different logic of research needs to be formulated; different from the Bacon-Galileo-Newton programme in natural philosophy mentioned earlier. That logic may be more like Science Two rather than Science One, as anticipated by Umpleby: “Science Two is a way of developing knowledge for fields that include knowing subjects, just as Science One is a way of developing knowledge for fields that usually do not include knowing subjects” (Umpleby, 2002).
Bennis and Thomas have looked at leadership from its broader social context, so as to inquire into how one emerges as a leader through a constant interplay between one’s own meaning making practice and the challenging context around oneself. Taking a cue from this, we should look at leadership research from a broader context too, so as to inquire into how new forms of leadership research could emerge, which would enhance leadership capacity in human collectives and thus address the current scarcity of this capacity. This might well be the book’s subtle hint to the new researchers in the field.
I sincerely thank the leadership researcher, Devashis Rath, currently at the All India Management Association--Centre for Management Education, India (devashis_r@yahoo.com), for going through an earlier draft of the review and suggesting additional literature references, besides sharing his thoughts on the nature of leadership research.
Argyris, C. (1980). Inner contradictions of rigorous research. New York: Academic Press.
Argyris, C. (1992). On organizational learning. Cambridge, MA: Blackwell.
Bateson, G. (1979). Mind and nature: A necessary unity. London: Wildwood House.
Bennis, W. G., & Thomas, R. J. (2002). Geeks & geezers: How era, values, and defining moments shape leaders. Boston, Massachusetts: Harvard Business School Press.
Dash, D. P. (2002). Improvement of co-ordination: Re-scripting research for multi-agent environments. Systems Research and Behavioral Science (Festschrift Issue on Prof. Gerard de Zeeuw), 19(2), 115-121.
House, R. J., & Aditya, R. N. (1997). The social scientific study of leadership: Quo vadis? Journal of Management, 23(3), 409-473. Retrieved March 19, 2004, from EBSCO electronic database.
Northouse, P. G. (1997). Leadership: Theory and practice. Thousand Oaks, California: Sage.
Reany, P. (1995). Philosophy of science. Arizona Journal of Natural Philosophy, 7, 64-94. Retrieved March 19, 2004, from http://www.ajnpx.com/pdf/AJNP/oct95c.pdf
Robbins, S. P. (2003). Organizational behavior (10th edition). New Delhi: Prentice-Hall of India.
Rost, J. C. (1991). Leadership for the twenty-first century. New York: Praeger.
Science and pseudoscience. (2004). Retrieved August 19, 2004, from http://www.lse.ac.uk/collections/lakatos/scienceAndPseudoscienceTranscript.htm
Umpleby, S. A. (2002). Should knowledge of management be organized as theories or as methods? In R. Trappl (Ed.), Cybernetics and Systems 2002 (pp. 429-427). Vienna: Austrian Society for Cybernetic Studies. Retrieved March 29, 2004, from http://www.gwu.edu/~umpleby/knowledge_of_management.htm
Vahl, M. (1999). A challenge to research: The improvement of communities. Systemica, 12, 443-464.
Worrall, J., & Currie, G. (Eds) (1978). Imre Lakatos: The methodology of scientific research programmes (Philosophical papers, Volume 1). Cambridge: Cambridge University Press.
Yukl, G. (2002). Leadership in organizations (5th edition; 1st edition in 1981). Upper Saddle River, New Jersey: Prentice-Hall.
Zeeuw, G. de (2001). Constructivism: A ‘next’ area of scientific development? Foundations of Science, 6, 77-98.
Published 5 March 2005
Copyright © 2005 Journal of Research Practice and the author