Charlie Munger on The Need for More Multidisciplinary Skills from Professionals: Educational Implications
All the passages below are taken from the book “Poor Charlie’s Almanack” edited by Peter D Kaufman. It is the Expanded Third Edition, tenth printing 2015.
Fiftieth Reunion of Harvard Law School Class of 1948, April 24, 1998
Today I am going to engage in a game reminding us of our old professors: Socratic solitaire. I will ask and briefly answer five questions:
1) Do broadscale professionals need more multidisciplinary skill?
2) Was our education sufficiently multidisciplinary?
3) In elite broadscale soft science, what is the essential nature of practicable best-form multidisciplinary education?
4) In the last fifty years, how far has elite academia progressed toward attainable best-form multidisciplinarity?
5) What educational practices would make progress faster?
We start with the question: Do broadscale professionals need more multidisciplinary skill?
To answer the first question, we must first decide whether more multidisciplinarity will improve professional cognition. And, to decide what will cure bad cognition, it will help to know what causes it. One of Bernard Shaw's characters explained professional defects as follows: "In the last analysis, every profession is a conspiracy against the laity." There is a lot of truth in Shaw's diagnosis, as was early demonstrated when in the sixteenth century, the dominant profession, the clergy burned William Tyndale at the stake for translating the Bible into English.
But Shaw plainly understates the problem in implying that a conscious, self-interested malevolence is the main culprit. More important, there are frequent, terrible effects in professionals from intertwined subconscious mental tendencies, two of which are exceptionally prone to cause trouble:
1) Incentive-caused bias, a natural cognitive drift toward the conclusion that what is good for the professional is good for the client and the wider civilization; and
2) Man-with-a-hammer tendency, with the name taken from the proverb: "To a man with only a hammer, every problem tends to look pretty much like a nail."
One partial cure for man-with-a-hammer tendency is obvious: If a man has a vast set of skills over multiple disciplines, he, by definition, carries multiple tools and, therefore, will limit bad cognitive effects from man-with-a-hammer tendency. Moreover, when he is multidisciplinary enough to absorb from practical psychology the idea that all his life he must fight bad effects from both the tendencies I mentioned, both within himself and from others, he has taken a constructive step on the road to worldly wisdom.
If "A" is narrow professional doctrine and “B” consists of the big, extra-useful concepts from other disciplines, then, clearly the professional possessing "A" plus "B" will usually be better off than the poor possessor of "A" alone. How could it be otherwise? And thus, the only rational excuse for not acquiring more "B" is that it is not practical to do so, given the man's need for "A" and the other urgent demands in his life. I will later try to demonstrate that this excuse for unidisciplinarity, at least for our most gifted people, is usually unsound.
Broadscale problems, by definition, cross many academic disciplines.
My second question is so easy to answer that I won't give it much time. Our education was far too unidisciplinary. Broadscale problems, by definition, cross many academic disciplines. Accordingly, using a unidisciplinary attack on such problems is like playing a bridge hand by counting trumps while ignoring all else. This is "bonkers," sort of like the Mad Hatter's Tea Party. But, nonetheless, too much that is similar remains present in professional practice and, even worse, has long been encouraged in isolated departments of soft science, defined as everything less fundamental than biology.
Even in our youth, some of the best professors were horrified by bad effects from balkanization of academia into insular, turf-protecting enclaves wherein notions were maintained by leaps of faith plus exclusion of nonbelievers. Alfred North Whitehead, for one, long ago sounded an alarm in strong language when he spoke of "the fatal unconnectedness of academic disciplines." And, since then, elite educational institutions, agreeing more and more with Whitehead, have steadily fought unconnectedness by bringing in more multidisciplinarity, causing some awesome plaudits to be won in our time by great unconnectedness fighters at borders of academic disciplines, for instance, Harvard's E. O. Wilson and Caltech's Linus Pauling.
Modern academia now gives more multidiscihlinarity than we received and is plainly right to do so.
The natural third question then becomes: What is now the goal? What is the essential nature of best-form multidisciplinarity in elite education? This question, too, is easy to answer. All we have to do is examine our must successful narrow-scale education, identify essential elements, and scale up those elements to reach the sensible solution.
Yes, I am suggesting today that mighty Harvard would do better if it thought more about pilot training.
To find the best educational narrow-scale model, we have to look not at unthreatened schools of education and the like, too much driven by our two counterproductive psychological tendencies and other bad influences, but, instead, look where incentives for effective education are strongest and results are most closely measured. This leads us to a logical place: the hugely successful education now mandatory for pilots. (Yes, I am suggesting today that mighty Harvard would do better if it thought more about pilot training.) In piloting, as in other profession,, one great hazard is bad effect from man-with-a-hammer tendency. We don't want a pilot, ever, to respond to a hazard as if it was hazard "X" just because his mind contains only a hazard "X" model. And so, for that and other reasons, we train a pilot in a strict six-element system:
1) His formal education is wide enough to cover practically everything useful in piloting.
2) His knowledge of practically everything needed by pilots is not taught just well enough to enable him to pass one test or two; instead, all his knowledge is raised to practice-based fluency, even in handling two or three intertwined hazards at once.
3) Like any good algebraist, he is made to think sometimes in a forward fashion and sometimes in reverse; and so he learns when to concentrate mostly on what he wants to happen and also when to concentrate mostly on avoiding what he does not want to happen.
4) His training time is allocated among subjects so as to minimize damage from his later malfunctions; and so what is most important in his performance gets the most training coverage and is raised to the highest fluency levels.
5) “Checklist” routines are always mandatory for him.
6) Even after original training, he is forced into a special knowledge-maintenance routine: regular use of the aircraft simulator to prevent atrophy through long disuse of skills needed to cope with rare and important problems.
We need for best results to have multidisciplinary coverage of immense amplitude, with all needed skills raised to an ever-maintained practice-based fluency...
The need for this clearly correct six-element system, with its large demands in a narrow-scale field where stakes are high, is rooted in the deep structure of the human mind. Therefore, we must expect that the education we need for broadscale problem solving will keep all these elements but with awesomely expanded coverage for each element. How could it be otherwise?
Thus it follows, as the night the day, that in our most elite broadscale education wherein we are trying to make silk purses out of silk, we need for best results to have multidisciplinary coverage of immense amplitude, with all needed skills raised to an ever-maintained practice-based fluency, including considerable power of synthesis at boundaries between disciplines, with the highest fluency levels being achieved where they are most needed, with forward and reverse thinking techniques being employed in a manner reminding one of inversion in algebra, and with "checklist” routines being a part of the knowledge system. There can be no other way, no easier way, to broadscale worldly wisdom. Thus the task, when first identified in its immense breadth, seems daunting, verging on impossible.
But the task, considered in full context, is far from impossible when we consider three factors:
First, the concept of "all needed skills" lets us recognize that we don't have to raise everyone's skill in celestial mechanics to that of Laplace and also ask everyone to achieve a similar skill level in all other knowledge. Instead, it turns out that the truly big ideas in each discipline, learned only in essence, carry most of the freight. And they are not so numerous, nor are their interactions so complex, that a large and multidisciplinary understanding is impossible for many, given large amounts of talent and time.
Second, in elite education, we have available the large amounts of talent and time that we need. After all, we are educating the top one percent in aptitude, using teachers who, on average, have more aptitude than the students. And we have roughly thirteen long years in which to turn our most promising twelve-year-olds into starting professionals.
Third, thinking by inversion and through use of "checklists" is easily learned---in broadscale life as in piloting.
Moreover, we can believe in the attainability of broad multidisciplinary skill for the same reason the fellow from Arkansas gave for his belief in baptism: "I've seen it done." We all know of individuals, modern Ben Franklins, who have (1) achieved a massive multidisciplinary synthesis with less time in formal education than is now available to our numerous brilliant young and (2) thus become better performers in their own disciplines, not worse, despite diversion of learning time to matter outside the normal coverage of their own disciplines.
Given the time and talent available and examples of successful masters of multiple disciplines, what is shown by our present failure to minimize bad effects from man-with -a-hammer tendency is only that you can't win big in multidisciplinarity in soft-science academia if you are so satisfied with the status quo, or so frightened by the difficulties of change, that you don't try hard enough to win big.
Which brings us to our fourth question: Judged with reference to an optimized feasible multidisciplinary goal, how much has elite soft-science education been corrected after we left?
The answer is that many things have been tried as corrections in the direction of better multidisciplinarity. And, after allowing for some counterproductive results, there has been some considerable improvement, net. But much desirable correction is still undone and lies far ahead.
For instance, soft-science academia has increasingly found it helpful when professors from different disciplines collaborate or when a professor has been credentialed in more than one discipline. But a different sort of correction has usually worked best, namely augmentation, or "take what you wish" practice that encourages any discipline to simply assimilate whatever it chooses from other disciplines. Perhaps it has worked best because it bypassed academic squabbles mulled in the tradition and territoriality that had caused the unidisciplinary folly for which correction was now sought.
Economics, in turn, took in from a biologist the "tragedy of the commons" model, thus correctly finding a wicked "invisible foot" in coexistence with Adam Smith's angelic "invisible hand,"
In any event, through increased use of "take what you wish," many soft-science disciplines reduced folly from man-with-a-hammer tendency. For instance, led by our classmate Roger Fisher, the law schools brought in negotiation, drawing on other disciplines. Over three million copies of Roger's wise and ethical negotiation book have now been sold, and his life's achievement may well be the best, ever, from our whole class. The law schools also brought in a lot of sound and useful economics, even some good game theory to enlighten antitrust law by better explaining how competition really works.
Economics, in turn, took in from a biologist the "tragedy of the commons" model, thus correctly finding a wicked "invisible foot" in coexistence with Adam Smith's angelic "invisible hand." These days, there is even some "behavioral economics," wisely seeking aid from psychology.
However, an extremely permissive practice like "take what you wish" was not destined to have one-hundred percent-admirable results in soft science. Indeed, in some of its worst outcomes, it helped changes like (1) assimilation of Freudianism in some literature departments; (2) importation into many places of extremist political ideologies of the left or right that had, for their possessors, made regain of objectivity almost as unlikely as regain of virginity; and (3) importation into many law and business schools of hard-form, efficient-market theory by misguided would-be experts in corporate finance, one of whom kept explaining Berkshire Hathaway's investing success by adding standard deviations of luck until, at six standard deviations, he encountered enough derision to force a change in explanations.
Moreover, even when it avoided such lunacies, "take what you wish" had some serious defects. For instance, takings from more fundamental disciplines were often done without attribution, sometimes under new names, with little attention given to rank in a fundamentalness order for absorbed concepts. Such practices (1) act like a lousy filing system that must impair successful use and synthesis of absorbed knowledge and (2) do not maximize in soft science the equivalent of Linus Pauling's systematic mining of physics to improve chemistry. There must be a better way.
This brings us, finally, to our last question: In elite soft science, what practices would hasten our progress toward optimized disciplinarity? Here again, there are some easy answers:
First, many more courses should be mandatory, not optional. And this, in turn, requires that the people who decide what is mandatory must possess large, multidisciplinary knowledge maintained in fluency. This conclusion is as obvious in the training of the would-be broadscale problem solver as it is in the training of the would-be pilot. For instance, both psychology mastery and accounting mastery should be required as outcomes in legal education. Yet, in many elite places, even today, there are no such requirements. Often, such is the narrowness of mind of the program designers that they neither see what is needed and missing nor are able to fix deficiencies.
Second, there should be much more problem-solving practice that crosses several disciplines, including practice that mimics the function of the aircraft simulator in preventing loss of skills through disuse. Let me give an example, roughly remembered, of this sort of teaching by a very wise but untypical Harvard Business School professor many decades ago.
This professor gave a test involving two unworldly old ladies who had just inherited a New England shoe factory making branded shoes and beset with serious business problems described in great detail. The professor then gave the students ample time to answer with written advice to the old ladies. In response to the answers, the professor next gave every student an undesirable grade except for one student who was graded at the top by a wide margin. What was the winning answer? It was very short and roughly as follows: "This business field and this particular business, in its particular location, present crucial problems that are so difficult that unworldly old ladies can not wisely try to solve them through hired help. Given the difficulties and unavoidable agency costs, the old ladies should promptly sell the shoe factory, probably to the competitor who would enjoy the greatest marginal-utility advantage." Thus, the winning answer relied not on what the students had most recently been taught in business school but, instead, on more fundamental concepts, like agency costs and marginal utility, lifted from undergraduate psychology and economics.
Ah, my fellow members of the Harvard Law Class of 1948: If only we had been
much more often tested like that, just think of what more we might have accomplished!
Incidentally, many elite private schools now wisely use such multidisciplinary methods in seventh grade science while, at the same time, many graduate schools have not yet seen the same light. This is one more sad example of Whitehead's "fatal unconnectedness" in education.
Third, most soft-science professional schools should increase use of the best business periodicals, like the Wall Street Journal, Forbes, Fortune, etc. Such periodicals are now quite good and perform the function of the aircraft simulator if used to prompt practice in relating events to multidisciplinary causes, often intertwined. And sometimes the periodicals even introduce new models for causes instead of merely refreshing old knowledge. Also, it is not just slightly sound to have the student practice in school what he must practice lifelong after formal education is over if he is going to maximize his good judgment. I know no person in business, respected for verified good judgment, whose wisdom-maintenance system does not include use of such periodicals. Why should academia be different?
Such periodicals are now quite good and perform the function of the aircraft simulator if used to prompt practice in relating events to multidisciplinary causes, often intertwined.
Fourth, in filling scarce academic vacancies, professors of super strong, passionate, political ideology, whether on the left or right, should usually be avoided. So also for students. Best-form multidisciplinarity requires an objectivity such passionate people have lost, and a difficult synthesis is not likely to be achieved by minds in ideological fetters. In our day, some Harvard Law professors could and did point to a wonderful example of just such ideology-based folly. This, of course, was the law school at Yale, which was then viewed by many at Harvard as trying to improve legal education by importing a particular political ideology as a dominant factor.
Fifth, soft science should more intensely imitate the fundamental organizing ethos of hard science (defined as the "fundamental four-discipline combination" of maths, physics, chemistry and engineering). This ethos deserves more imitation.
After all hard sciences has, by a wide margin, the best record for both (1) avoiding unidisciplinary folly and (2) making user-friendly a big patch of multidisciplinary domain, with frequent, good results like those of physicist Richard Feynman when he so quickly found in cold O-rings the cause of our greatest space shuttle disaster. And previous extensions of the ethos into softer fare have worked well. For instance, biology, starting 150 years ago with a descriptive mess not much related to deep theory, has gradually absorbed the fundamental organizing ethos with marvelous results as new generations have come to use better thinking methods, containing models that answer the question: why? And there is no clear reason why the ethos of hard science can't also help in disciplines far less fundamental than biology. Here, as I interpret it, is this fundamental organizing ethos I am talking about:
1) You must both rank and use disciplines in order of fundamentalness.
2) You must, like it or not, master to tested fluency and routinely use the truly essential parts of all four constituents of the fundamental four-discipline combination, with particularly intense attention given to disciplines more fundamental than your own.
3) You may never practice either cross-disciplinary absorption without attribution or departure from a "principle of economy" that forbids explaining in any other way anything readily explainable from more fundamental material in your own or any other discipline.
4) But when the step (3) approach doesn't produce much new and useful insight, you should hypothesize and test to establishment new principles, ordinarily by using methods similar to those that created successful old principles. But you may not use any new principle inconsistent with an old one unless you can now prove that the old principle is not true.
There is an old two-part rule that often works wonders in business, science, and elsewhere: (1) Take a simple, basic idea and (2) take it very seriously.
You will note that, compared with much current practice in soft science, the fundamental organizing ethos of hard science is more severe. This reminds one of pilot training, and this outcome is not a coincidence. Reality is talking to anyone who will listen. Like pilot training, the ethos of hard science does not say "take what you wish" but "learn it all to fluency, like it or not." And rational organization of multidisciplinary knowledge is forced by making mandatory (1) full attribution for cross-disciplinary takings and (2) mandatory preference for the most fundamental explanation.
This simple idea may appear too obvious to be useful, but there is an old two-part rule that often works wonders in business, science, and elsewhere: ( I) Take a simple, basic idea and (2) take it very seriously. And, as some evidence for the value of taking very seriously the fundamental organizing ethos, I offer the example of my own life.
To this day, I have never taken any course, anywhere, in chemistry, economics, psychology, or business.
I came to Harvard Law School very poorly educated, with desultory work habits and no college degree. I was admitted over the objection of Warren Abner Seavey through the intervention of family friend Roscoe Pound. I had taken one silly course in biology in high school, briefly learning, mostly by rote, an obviously incomplete theory of evolution, portions of the anatomy of the paramecium and frog, plus a ridiculous concept of "protoplasm" that has since disappeared. To this day, I have never taken any course, anywhere, in chemistry, economics, psychology, or business. But I early took elementary physics and math and paid enough attention to somehow assimilate the fundamental organizing ethos of hard science, which I thereafter pushed further and further into softer and softer fare as my organizing guide and filing system in a search for whatever multidisciplinary worldly wisdom it would be easy to get.
Thus, my life became a sort of accidental educational experiment with respect to feasibility and utility of a very gross academic extension of the fundamental organizing ethos by a man who also learned well what his own discipline had to teach.
What I found, in my extended attempts to complete by informal means my stunted education, was that, plugging along with only ordinary will but with the fundamental organizing ethos as my guide, my ability to serve everything I loved was enhanced far beyond my deserts. Large gains came in places that seemed unlikely as I started out, sometimes making me like the only one without a blindfold in a high-stake game of "pin the tail on the donkey." For instance, I was productively led into psychology, where I had no plans to go, creating large advantages that deserve a story on another day.
Today, I have no more stories. I have finished my talk by answering my own questions as best I could in a brief time. What is most interesting to me in my answers is that, while everything I have said is non original and has long been obvious to the point of banality to many sound and well-educated minds, all the evils I decry remain grossly overpresent in the best of our soft-science educational domains wherein virtually every professor has a too unidisciplinary habit of mind, even while a better model exists just across the aisle in his own university. To me, this ridiculous outcome implies that the soft-science departments tolerate perverse incentives. Wrong incentives are a major cause because, as Dr. Johnson so wisely observed, truth is hard to assimilate in any mind when opposed by interest. And, if institutional incentives cause the problem, then a remedy is feasible---because incentives can be changed.
It is neither inevitable nor advantageous for soft-science educational domains to tolerate as much unidisciplindry wrong headedness as they now do.
I have tried to demonstrate today, and indeed by the example of my life, that it is neither inevitable nor advantageous for soft-science educational domains, to tolerate as much unidisciplinary wrongheadedness as they now do. If I could somewhat fix it, there is an ethos, also from Dr. Johnson, that is applicable. Please remember the word Dr. Johnson used to describe maintenance of academic ignorance that is removable through diligence. To Dr. Johnson, such conduct was "treachery."
As Dr. Johnson so wisely observed, truth is hard to assimilate in any mind when opposed by interest. And, if institutional incentives cause the problem, then a remedy is feasible because incentives can be changed.
And if duty will not move improvement, advantage is also available. There will be immense worldly rewards, for law schools and other academic domains as for Charlie Munger, in a more multidisciplinary approach to many problems, common or uncommon. And more fun as well as more accomplishment. The happier mental realm I recommend is one from which no one willingly returns. A return would be like cutting off one's hands.
Talk Five Revisited
As I review Talk Five in 2006, I would not change a word. And I continue to believe that my ideas are important. In my attitude I may be displaying too much similarity to my long-dead relative, Reverend Theodore Munger, former chaplain of Yale.
Theodore published a collection of his sermons, laying out proper conduct with a strong, ex cathedra tone. Then, late in life, he published a final edition, reporting in his foreword that he had made no changes at all and was now producing the new edition only because the extreme popularity of his sermons had caused excessive wear in the original printing plates.
“Read over your compositions, and when you meet a passage which you think is particularly fine, strikes it out” Samuel Johnson