詹姆斯·馬奇（James Gardner March）：如藝術(shù)般的思想
Ideas as Art??by Diane Coutu (October 2006)
Summary.??
?A professor emeritus in management, sociology, political science, and education at Stanford University, March has taught courses in subjects as diverse as organizational psychology, behavioral economics, leadership, rules for killing people, friendship, computer simulation, and statistics. He is perhaps best known for his pioneering contributions to organization and management theory. March’s accomplishments in that field, and in many others, have conferred on him an almost unprecedented reputation as a rigorous scholar and a deep source of wisdom. As University of Chicago professor John Padgett wrote in the journal Contemporary Sociology, “March’s influence, unlike that of any of his peers, is not limited to any possible subset of the social science disciplines; it is pervasive.”
?March approaches thought aesthetically; he cares that ideas have “some form of elegance or grace or surprise.” His poetic sensibility can be felt in the metaphors he has created over the years—the “garbage can theory” of organizational choice, for instance, and the “hot-stove effect” in learning.
?In this edited interview with HBR senior editor Diane Coutu, March shares his thinking on aesthetics, leadership, the role of folly, and the irrelevance of relevance when it comes to the pursuit of ideas. He also comments on the fundamental differences between academic and experiential knowledge, underscoring the need for both.
In these pages, three years ago, consultants Laurence Prusak and Thomas H. Davenport reported the findings of a survey of prominent management writers who identified their own gurus. Although his is an unfamiliar name to most readers of this periodical, James G. March appeared on more lists than any other person except Peter Drucker.
?
March is a writer to whom the experts turn when they want to engage new ideas. He is a polymath whose career over the past five decades has encompassed numerous disciplines. A professor emeritus in management, sociology, political science, and education at Stanford University, he has taught courses on subjects as diverse as organizational psychology, behavioral economics, leadership, rules for killing people, friendship, decision making, models in social science, revolutions, computer simulation, and statistics.
?
March is perhaps best known for his pioneering contributions to organization and management theory. He has coauthored two classic books: Organizations (with Herbert A. Simon) and A Behavioral Theory of the Firm (with Richard M. Cyert). Together with Cyert and Simon, March developed a theory of the firm that incorporates aspects of sociology, psychology, and economics to provide an alternative to neoclassical theories. The underlying idea is that although managers make decisions that are intendedly rational, the rationality is “bounded” by human and organizational limitations. As a result, human behavior is not always what might be predicted when rationality is assumed.
?
In addition to this work, March has been a leading contributor to the study of political institutions, particularly through the books Rediscovering Institutions and Democratic Governance (both coauthored with Johan P. Olsen); the study of leadership, through Leadership and Ambiguity (with Michael D. Cohen) and On Leadership (with Thierry Weil); and the study of decision making, through his books A Primer on Decision Making and The Pursuit of Organizational Intelligence.
?
March’s influence on students of organizations and management and his accomplishments in other areas of social science have conferred on him an almost unprecedented reputation as a rigorous scholar and a deep source of wisdom. In the academic literature, it has become de rigueur to cite his articles. Professor John Padgett of the University of Chicago wrote in the journal Contemporary Sociology that “Jim March is to organization theory what Miles Davis is to jazz…. March’s influence, unlike that of any of his peers, is not limited to any possible subset of the social science disciplines; it is pervasive.”
?
Despite his renown in the social sciences, March has not confined his interests to those fields. Besides his professional and scholarly articles and books, he has made one film (Passion and Discipline: Don Quixote’s Lessons for Leadership), is currently working on another, and has written seven books of poetry. His poetic sensibility can be felt in the metaphors he has created over the years: the “garbage can theory” of organizational choice, the “technology of foolishness,” the role of consultants as “disease carriers,” and the “hot-stove effect.”
?
In a recent interview with HBR senior editor Diane Coutu at his home in Portola Valley, California, March shared his thinking on aesthetics, leadership, the usefulness of scholarship, the role of folly, and the irrelevance of relevance when it comes to the pursuit of ideas. The following is an edited version of that interview.
?
?
How did you come to have such influence on management thinking without being as public a figure as Peter Drucker?
?
I don’t claim to have any notable impact, partly because I question the methodology by which such an impact is assigned to me, and partly because I suspect it is easier to infiltrate the idea consciousness of the business community if you do not worry about attribution, reputation for significance, or direct communication. And even if that is not true, I fear that my own enjoyments come more from playing with ideas than from selling them. I am a scholar; I do what scholars do. I think about things, conduct research studies, and write up my thoughts and research in professional journals. The writing is probably more a way of understanding my thoughts than of communicating them.
?
That sounds more reclusive than I believe I am. I am hardly so perverse as to resist having my writings read by others. The articles I write are accessible to any manager who chooses to read them. I don’t write obscurely—at least not deliberately—and the ideas are not exceptionally arcane. Dull, maybe, but not arcane.
?
You liked to begin your classes at Stanford each year saying, “I am not now, nor have I ever been, relevant.” What did you mean by that?
?
It was a signal to students that it would not be fruitful to ask me about the immediate usefulness of what I had to say. If there is relevance to my ideas, then it is for the people who contemplate the ideas to see, not for the person who produces them. For me, a feature of scholarship that is generally more significant than relevance is the beauty of the ideas. I care that ideas have some form of elegance or grace or surprise—all the things that beauty gives you. The central limit theorem, for example, is one of the more important theorems in classical statistics; it allows you to say things about sampling errors. But for me, the theorem is primarily a thing of extraordinary beauty. I think that anyone who teaches the central limit theorem should try to communicate that aesthetic joy to students.
?
Charlie Lave and I wrote Introduction to Models in the Social Sciences. Among the books I’ve written, it is one of my favorites. It introduces students to the rudiments of four very fundamental models in social science. It treats modeling as an art form. Scholarship will always have an element of aestheticism, because scholars are obliged to advance beauty as well as truth and justice.
?
This seems to be your artistic sensibility. How do you justify such a sensibility in a world where business is so desperately in need of practical solutions?
?
No organization works if the toilets don’t work, but I don’t believe that finding solutions to business problems is my job. If a manager asks an academic consultant what to do and that consultant answers, then the consultant should be fired. No academic has the experience to know the context of a managerial problem well enough to give specific advice about a specific situation. What an academic consultant can do is say some things that, in combination with the manager’s knowledge of the context, may lead to a better solution. It is the combination of academic and experiential knowledge, not the substitution of one for the other, that yields improvement.
?
No organization works if the toilets don’t work, but I don’t believe that finding solutions to business problems is my job.
?
Have you ever consulted to businesses?
?
When I was younger—younger and poorer, I suppose—I used to do some relatively technical consulting about things involving statistics or research methods. I don’t do that any longer. I still occasionally do something I humorously call “consulting” but probably is better seen as getting someone to buy lunch for me. If someone calls me up and says a manager would like to talk to me, I’m inclined to respond that I almost certainly don’t have anything useful to say. But if the answer is that it doesn’t matter, that the manager still wants to have lunch with me, then I am happy to have lunch. I think that it would ordinarily be difficult to discover any practical use for such conversations, but I may occasionally have a way of looking at things that is sufficiently different to help a manager in some marginal way. Usually, managers are sensible enough not to ask me to lunch, and I end up paying for most of my lunches myself.
?
Leadership has become a big concern and a big industry in recent years. What is happening in leadership research?
?
I doubt that “l(fā)eadership” is a useful concept for serious scholarship. The idea of leadership is imposed on our interpretation of history by our human myths, or by the way we think that history is supposed to be described. As a result, the fact that people talk about leaders and attribute importance to them is neither surprising nor informative. Although there is good work on several aspects of asymmetric relations in life, broad assertions about leadership are more characteristic of amateurs than of professionals. Unless and until a link to significant scholarship can be made, the thinking on leadership will produce more articles in popular journals than in professional ones, more homilies and tautologies than powerful ideas. In the meantime, in order for leadership scholarship to generate some good ideas, it needs to build buffers to protect itself from the temptations of immediate relevance.
?
What kinds of questions do you think are important for leaders?
?
In my course on leadership and literature, I ended up with a list of about ten topics—for example, power, dominion, and subordination; ambiguity and coherence; gender and sexuality; the relation between private and public lives. Not a unique list, and hardly a complete one. Each of the topics can draw illumination from social science, but I think they often are more profoundly considered in great literature. One issue, which I used to talk about by looking at George Bernard Shaw’s Saint Joan, is how madness, heresy, and genius are related. We often describe great leaders as having the drive, vision, imagination, and creativity to transform their organizations through daring new ideas. Retrospectively, of course, we sometimes find that such heresies have been the foundation for bold and necessary change, but heresy is usually just crazy. Most daring new ideas are foolish or dangerous and appropriately rejected or ignored. So while it may be true that great geniuses are usually heretics, heretics are rarely great geniuses. If we could identify which heretics would turn out to be geniuses, life would be easier than it is. There is plenty of evidence that we cannot.
?
Most daring new ideas are foolish or dangerous and appropriately rejected or ignored. So while it may be true that great geniuses are usually heretics, heretics are rarely great geniuses.
?
?
In your film on Don Quixote and leadership, you say that if we trust only when trust is warranted, love only when love is returned, and learn only when learning is valuable, then we abandon an essential feature of our humanity. How do we lose part of our humanity?
?
We justify actions by their consequences. But providing consequential justification is only a part of being human. It is an old issue, one with which Kant and Kierkegaard, among many others, struggled. I once taught a course on friendship that reinforced this idea for me. By the end of the course, a conspicuous difference had emerged between some of the students and me. They saw friendship as an exchange relationship: My friend is my friend because he or she is useful to me in one way or another. By contrast, I saw friendship as an arbitrary relationship: If you’re my friend, then there are various obligations that I have toward you, which have nothing to do with your behavior. We also talked about trust in that class. The students would say, “Well, how can you trust people unless they are trustworthy?” So I asked them why they called that trust. It sounded to me like a calculated exchange. For trust to be anything truly meaningful, you have to trust somebody who isn’t trustworthy. Otherwise, it’s just a standard rational transaction. The relationships among leaders and those between leaders and their followers certainly involve elements of simple exchange and reciprocity, but humans are capable of, and often exhibit, more arbitrary sentiments of commitment to one another.
?
You’ve said that scholars and managers do fundamentally different things. Can you elaborate on that?
?
You need both academic knowledge and experiential knowledge, but they are different. The scholar tries to figure out, What’s going on here? What are the underlying processes making the system go where it’s going? What is happening, or what might happen? Scholars talk about ideas that describe the basic mechanisms shaping managerial history—bounded rationality, diffusion of legitimate forms, loose coupling, liability of newness, competency traps, absorptive capacity, and the like. In contrast, experiential knowledge focuses on a particular context at a particular time and on the events of personal experience. It may or may not generalize to broader things and longer time periods; it may or may not flow into a powerful theory; but it provides a lot of understanding of a particular situation. A scholar’s knowledge cannot address a concrete, highly specific context, except crudely. Fundamental academic knowledge becomes more useful in new or changing environments, when managers are faced with the unexpected or the unknown. It provides alternative frames for looking at problems rather than solutions to them.
?
Along with Richard Cyert and Herb Simon, you laid the groundwork for the field of behavioral economics. Do you think you started a revolution?
?
Scholarship is a communal activity. No one comes first; and no one stands alone. Insofar as there has been a behavioral economics “revolution,” it is a revolution started by many. Nowadays, economists and others do talk a fair amount about bounded rationality. But the economists are more inclined to see the limits on rationality as minor perturbations, easily accommodated within some variation of neoclassical economic theory, than as fundamental challenges. I once wrote a paper on the tendency of economists to maintain neoclassical theory by reinterpreting its definitions and constraints. I called the paper “The War Is Over, and the Victors Have Lost,” to note the way economics has tended to become so tautological as to “explain” everything at the cost of abandoning predictive power. At times, economics as a theory threatens to become economics as a faith.
?
You’ve been unusually interdisciplinary in your life’s work. Is there some overarching question that you’ve tried to answer?
?
I don’t think it’s been that grand. It just happens that the disciplines are organized in ways that distribute my rather narrow areas of research focus across standard disciplinary lines. I’ve studied problem solving and decision making, risk taking, information processing, innovation and change, learning, selection, and the creation and revision of rules and identities. In a very loose way, I think, you could say my main focus has been on the cognitive aspects of organizations, as long as you include in the term “cognitive” such things as conflict, bias, rule following, and confusion.
?
Much of your research has focused on learning in one way or another. How is this tied to the hot-stove effect?
?
That term comes from an article by Jerker Denrell and me, but it is stolen from some of Mark Twain’s wisdom. Twain said that if a cat ever jumps on a hot stove, he will never jump on a hot stove again. And that’s good. But he will also never jump on a cold stove again—and that may not be good.
?
The hot-stove effect is a fundamental problem of learning. Learning reduces your likelihood of repeating things that got you in trouble, as you hope it will. But that means you know less about the domains where you’ve done poorly than about the domains where you’ve done well. You might say, “Well, why should that cause problems?” It causes problems whenever your early experience with an alternative is, for whatever reason, not characteristic of what subsequent experience would be. It clearly causes problems in domains where practice makes a difference. For example, you are likely to abandon an approach or a technology prematurely.
?
One form of the hot-stove effect is the competency trap, where learning encourages people to stick to and improve skills they have already honed to a fine degree rather than spend time gaining new ones. Some of my grandchildren say to me, “We’re not very good at mathematics, so we’re not going to take any more mathematics.” I say, “Wait a minute. Mathematics is a practice sport. If you’re not very good at it, you take more of it.” That’s counterintuitive, and it goes against the main logic of experiential learning, not to mention grandchildren’s sentiments about control over their own lives. It has also been demonstrated that the hot-stove effect leads experiential learners to be risk averse. It is possible to limit the hot-stove effect by slowing learning so that you increase the sample of alternatives that have poor results. That obviously has the cost of incurring short run losses and consequently is hard for an adaptive system to do.
?
You’ve written about the importance of a “technology of foolishness.” Could you tell us a little about it?
?
That paper sometimes gets cited—by people who haven’t read it closely—as generic enthusiasm for silliness. Well, maybe it is, but the paper actually focused on a much narrower argument. It had to do with how you make interesting value systems. It seemed to me that one of the important things for any person interested in understanding or improving behavior was to know where preferences come from rather than simply to take them as given.
?
So, for example, I used to ask students to explain the factual anomaly that there are more interesting women than interesting men in the world. They were not allowed to question the fact. The key notion was a developmental one: When a woman is born, she’s usually a girl, and girls are told that because they are girls they can do things for no good reason. They can be unpredictable, inconsistent, illogical. But then a girl goes to school, and she’s told she is an educated person. Because she’s an educated person, a woman must do things consistently, analytically, and so on. So she goes through life doing things for no good reason and then figuring out the reasons, and in the process, she develops a very complicated value system—one that adapts very much to context. It’s such a value system that permitted a woman who was once sitting in a meeting I was chairing to look at the men and say, “As nearly as I can tell, your assumptions are correct. And as nearly as I can tell, your conclusions follow from the assumptions. But your conclusions are wrong.” And she was right. Men, though, are usually boys at birth. They are taught that, as boys, they are straightforward, consistent, and analytic. Then they go to school and are told that they’re straightforward, consistent, and analytic. So men go through life being straightforward, consistent, and analytic—with the goals of a two-year-old. And that’s why men are both less interesting and more predictable than women. They do not combine their analysis with foolishness.
?
How do you encourage people to be foolish?
?
Well, there are some obvious ways. Part of foolishness, or what looks like foolishness, is stealing ideas from a different domain. Someone in economics, for example, may borrow ideas from evolutionary biology, imagining that the ideas might be relevant to evolutionary economics. A scholar who does so will often get the ideas wrong; he may twist and strain them in applying them to his own discipline. But this kind of cross-disciplinary stealing can be very rich and productive. It’s a tricky thing, because foolishness is usually that—foolishness. It can push you to be very creative, but uselessly creative. The chance that someone who knows no physics will be usefully creative in physics must be so close to zero as to be indistinguishable from it. Yet big jumps are likely to come in the form of foolishness that, against long odds, turns out to be valuable. So there’s a nice tension between how much foolishness is good for knowledge and how much knowledge is good for foolishness.
?
Another source of foolishness is coercion. That’s what parents often do. They say, “You’re going to take dance lessons.” And their kid says, “I don’t want to be a dancer.” And the parents say, “I don’t care whether you want to be a dancer. You’re going to take these lessons.” The use of authority is one of the more powerful ways to encourage foolishness. Play is another. Play is disinhibiting. When you play, you are allowed to do things you would not be allowed to do otherwise. However, if you’re not playing and you want to do those same things, you have to justify your behavior. Temporary foolishness gives you experience with a possible new you—but before you can make the change permanent, you have to provide reasons.
?
What role would there be for foolishness in business education?
?
We have some foolishness already, though we dress it up as fairly serious activity. For example, we have students play roles. We have them pretend they are the CEO of IBM, and that’s foolishness. They aren’t, and they can’t be, and they won’t be. But if you are encouraged to think of yourself as somebody else, you start acting the way you imagine such a person ought to act and experimenting with who you might become.
?
On the whole, I think that American management education is so deeply embedded in a rational mystique that pressure toward foolishness often has to become extreme in order to have even a minor effect. At the same time, I don’t think any of us would want to live in a world of foolishness that ignored the fact that one of the major glories of the human estate is the capability to practice intelligent rationality.
?
It’s all a question of balance. Soon after I wrote my paper on the technology of foolishness, I presented it at a conference in Holland. This was around 1971. One of my colleagues from Yugoslavia, now Croatia, came up and said, “That was a great talk, but please, when you come to Yugoslavia, don’t give that talk. We have enough foolishness.” And I think he may have been right.
?
You’re famous for your garbage can theory of organizational choice. Can you sum up the theory for us?
?
The original article on the garbage can theory was written jointly with Michael Cohen and Johan Olsen, so they have to share in whatever fame or shame there is in it. A fair number of people took the organized-anarchy notion that life is ambiguous and said, “The garbage can is really a label for confusion.” That wasn’t quite what we meant.
?
We were operating at two levels. On one level, we were saying that choice is fundamentally ambiguous. There is a lot of uncertainty and confusion that isn’t well represented by standard theories of decision making. Opportunities for choice attract all sorts of unrelated but simultaneously available problems, solutions, goals, interests, and concerns. So a meeting called to discuss parking lots may become a discussion of research plans, sexual harassment, managerial compensation, and advertising policies. Time is scarce for decision makers, though, and what happens depends on how they allocate that time to choice opportunities.
?
On the second level, we tried to describe the way in which organizations deal with flows of problems, solutions, and decision makers in garbage can situations. The central ideas were that a link between a problem and a solution depends heavily on the simultaneity of their “arrivals,” that choices depend on the ways in which decision makers allocate time and energy to choice opportunities, that choice situations can easily become overloaded with problems, and that choices often can be made only after problems (and their sponsors) have moved to other decision arenas and thus typically are not resolved.
?
In our minds, the garbage can process is a very orderly process. It looks a little peculiar from some points of view, but it isn’t terribly complex, and it isn’t terribly jumbled. The good thing, I think, is that our perspective has opened up the possibility for people to say, “That’s a garbage can process”—meaning it’s an understandable process in which things are connected by their simultaneous presence more than by anything else, even though they look all mixed-up.
?
Does it concern you that people sometimes misunderstand your ideas?
?
In a real sense, there is no such thing as “my” ideas. Scholarship and notions of intellectual property are poor bedmates. I have often read things, both by critics and by enthusiasts, that seem to me to be based on a less than precise reading of what I have written; but once you publish something, you lose special access to it. The interpretations of others have as much legitimacy, if they can be defended, as yours do. In the best of all worlds, others will generate interpretations that are more interesting than the ones you had in your mind. In fact, a basic goal in writing is to choose words that can evoke beautiful and useful meanings that were not explicit in your own mind. Some very good writers resist that idea. They want to be their own interpreters. I think that is a mistake. The evocative ambiguities of language are sources of creativity.
?
You’ve said that the numbers of women coming into the workforce have changed the sexual character of organizations but that many people will pine for the simplicity of the old order.
?
Oh, sure. But I’m also committed to the notion that a lot of happiness comes from dealing with complexity. We may regret the passing of a simple life—a simple division of labor between the sexes or a work world without heterosexuality—but we wouldn’t like it if that life were restored to us. I think the problems of wrestling with the issues of gender and sexuality have been, for my generation, very important. They’ve obviously been important to women, and, as a result, they’ve been important to men. Many of the beauties we now see in people have come out of that struggle, and the struggle is by no means over. Nor am I sure you would want to have it over. For example, life might be much less interesting if you actually took sex and sexuality out of management. It would be easier, and it might, in some ways, be less prone to atrocity—but it surely wouldn’t be as much fun.
?
Are there any practitioners you admire?
?
I guess I admire all of them, even the scoundrels. Modern organized life poses problems that are not trivial; and anyone who is prepared to try to function meaningfully in a modern organization has my respect. Dealing with the simultaneous demands for self-respect, autonomy, control, coordination, order, freedom, imagination, discipline, and effectiveness that are essential parts of modern organizations seems to me worthy of respect, even when it is done in a less than perfect way. I know there is some sense in the observation that hierarchies and competition can make monsters out of ordinary, good people. They often do, and I think we have an obligation to recognize that problem. The business firm is one of the few contemporary institutions in which the arbitrary and gratuitous cruelty of the powerful in dealing with the weak is tolerated, even encouraged. Even so, most of the executives I’ve met seem to be trying to do as decent a job as they can. Of contemporary figures, probably the one I know best is John Reed, the longtime CEO of Citibank and recently the supervisor of reforms in the New York Stock Exchange. I admire John. I think he has a sense of what it means to be a human being.
?
I think practicing managers are sometimes less reflective than they might be. The rhetoric of management requires managers to pretend that things are clear, that everything is straightforward. Often they know that managerial life is more ambiguous and contradictory than that, but they can’t say it. They see their role as relieving people of ambiguities and uncertainties. They need some way of speaking the rhetoric of managerial clarity while recognizing the reality of managerial confusion and ambivalence. In a recent paper, I argued that reading poetry helps, but I fear that is a small response to a large problem.
?
You are a poet yourself. Why do you write poetry?
?
I’m not sure why I write poetry. I’m not always sure that it is poetry. It has something to do with an affection for the beauties and grace of life, along with an affection for its efficiency or effectiveness. I think it is the beauty of rationality, as well as its utility, that attracts me to it. It is the beauty of emotions and feelings that makes them compelling for me. Poetry is a way of contemplating and augmenting those beauties, as well as the absurdities of their presence in the dustbins of life. Poetry celebrates the senses; it celebrates the feelings in ways that other things don’t. It’s also a place where you can play with the splendor, sound, and combinatorics of words. And you don’t usually do that in other genres.
?
You began your courses by denying claims to your relevance. How did you end them?
?
It depended on the course; but quite often I would end with a quote from a French writer, étienne Pivert de Senancour. Even in English, it provided a patina of culture to an otherwise midwestern sensibility: “Man is perishable. That may be; but let us perish resisting, and, if nothingness is what awaits us, let us not act in such a way that it would be a just fate.”
?
In the end, you know, we are very minor blips in a cosmic story. Aspirations for importance or significance are the illusions of the ignorant. All our hopes are minor, except to us; but some things matter because we choose to make them matter. What might make a difference to us, I think, is whether in our tiny roles, in our brief time, we inhabit life gently and add more beauty than ugliness.
