Range
David Epstein
This is a book about the power of generalists. It starts with a comparison of Tiger Woods and Roger Federer. Both are generational talents, but they developed very differently. Tiger Woods was playing golf by age 2 and was in tournaments. All he did was golf and he was quite successful (and may not be done). Roger Federer played multiple sports and only began to focus on tennis when he was a teen. Woods father pushed him, but Federer’s parents did not; his mother was a tennis coach and she refused to coach him. Federer was still the world number one in his thirties and is currently (March 2020) ranked fourth at age 38. It is often thought that great skill requires early dedication and examples from music (piano, violin) and chess are often given.
Federer clearly has done fine despite his late start, and the book suggests that he is not an exception but the norm. It seems that the advantage of an early start is enjoyed in what are called “kind” systems. A kind system is highly predictable and limited in its possibilities. Golf requires some complex mechanics, but hits a ball at rest towards a fixed target. There is not opponent to outwit and reaction speed barely comes into the game. In contrast, tennis is a “wicked” system. The ball is moving, the opponent is trying to make it harder for you and reflexes play a critical role – there is strategy. Economic forecasting, scientific discovery and social change are wicked domains where the “rules” are unclear, feedback is slow or patterns of response are rare. Playing classical music or chess are more like golf than tennis. Practice time is a powerful influence on future success in kind systems.
It turns out that across athletics as a whole, Federer followed a path more common for elite athletes – near-elite athletes often follow a path more like Woods. The book provides numerous examples of world champion level athletes who did not specialize until their late teens or early twenties. This observation applies beyond sports. Musicians, scientists and artists often sample many domains before picking one.
In fact, there is something about generalism that works really well, and this is contrary to what we commonly think today. Psychologists have known for a while that “slow learning” is superior to fast learning over the long run – in fact, test performance has little predictive power for what is learned. The suggestion is that our ability to succeed in unpredictable situations depends on varied experience over time – which does not come from hyper-specialized early learning or practice.
For a long time, people beat computers at chess. As computer performance improved,people could no longer beat computers and eventually the reigning world champion player lost decisively to a computer. The computers could replicate what the hyper-specialized chess grandmasters did, which was recognize patterns and play out different patterns. While a superior human player could consider 3-5 moves ahead, a computer could calculate 10 or 100 moves ahead. In fact, it seems that chess players did not succeed based on superior strategy, but superior tactics learned by intensive practice. This was shown when people using a computer began to beat the best computer programs. To make this clearer, a moderately good player with a computer could easily beat a supercomputer. People are good at strategy - which involves a combination of imagination and logic. There are important implications to this observed limitation to bot humans and computers.
Specialized training has been suggested to be extremely pattern based. A chess expert can look at a chess board in mid-game for less than 5 seconds and recreate the game board. But if the same person can’t reproduce the board when they were shown a board with pieces placed at random. Similar results have been shown for musicians presented with “real” and “artificial” music passages. Specialists learn to chunk information as long as it fits a learned pattern, but deprived of suitable context they are ordinary. Similarly, many elite athletes are clumsy outside of their main sport.
Chunking is an everyday practice that we all use. For example, we read rapidly because every language uses chunking to set sentence structure and word order. For a native speaker, interpretation becomes reflexive and second languages a challenge. We remembered phones numbers (before cell phones with address lists) via chunking, and chunking is why phone or credit card numbers tend to be clustered in groups of 2-4 numbers.
Video gaming offers another perspective on human vs. computer capabilities. The game Star Craft is a strategy game where a player needs to manage many “societal development” activities while fighting a war. People lose to the computer regularly, but they can adapt. Studied closely, it is easy to see that humans are weak at almost every activity compared to the computer, but humans can adapt their approach in ways that the computer can’t. It was not until 2019 that an artificial intelligence computer beat a top human player, but after a few losses the human began winning again. The superior human long-term adaptability in the complex game was more important that the specific tactical ability of the computer. The more complex the environment, the greater the human advantage to use broad integration of information, skills, and imagination. A few years ago, there was great hope that AI would permit rapid progress in cancer understanding, but it has been a complete flop. Unlike Jeopardy (a game show where people supply the questions to match offered answers) where the answers are known to start with, cancer research does not even know the right questions yet. Humans are better in such domains.
Interestingly, incentives may have a detrimental effect on our ability to solve such problems. In an experiment, students had to figure out the right order to flip a series of switches; there were 70 possible successful outcomes. A tiny monetary reward was given for each successful outcome and students could play as many times as they liked. Students would find a solution and repeat it to gain more money. Other students were asked to find the general rule that solved all 70 cases, and every student succeeded. Only one student in the reward situation found the general rule. Rewards apparently do not help people explore complex topics. In contrast to simple domains like chess or golf, much of business life is closer to “Martian tennis”. There are racquets and balls, but the rules are unknown and must be learned by playing.
In domains where repetition is common, practices makes a big difference and an early start provides more practice time. But people who occupy such practice-dominant roles can struggle with change. The author cites the example of experienced accountants faced with a change in tax law. Novices were able to understand and properly apply the new law better than the experts – who suffered from “cognitive entrenchment”. The implication is that expertise must be balanced with generalism to be effective in the real world. While anecdotal, it might be relevant that compared to the general scientist, Nobel Prize winners were 20x more likely to be an active participant in the performance arts. Widely recognized scientists are also much more likely to be involved in arts of some sort than average scientists. Dean Keith Simonton, who studies creativity commented that rather than obsessively focus on a narrow topic creative achievers tend to have broad interests. This breadth often supports insights that cannot be attributed to domain-specific expertise alone.* Steve Jobs attributed his design aesthetic to a calligraphy course and Claude Shannon (the fact that you can Google his name is due to his creation of information science) attributed his insight into information to a philosophy course. These parallel interests allow these people to recognize and get out of a rut when a problem is hard.
As we have all been taught, evolution acts slowly. From this, we assume that we are about as smart today as our parents and grandparents. Actually, you were probably smarter than your parents, and your children are probably smarter than you. That is because how smart we are is the result of how stimulating our environment is during development. The US Army measured the IQ of incoming soldiers over decades, and someone observed that an average solider in World War I would have been in the 22% percentile by World War II. Follow up work looked at IQ trends using a test called the Raven’s Progressive Matrices test, which is language and culturally independent. The gain is about 3 points every 10 years. An average adult today would have been in the 98th percentile 100 years ago, Tests based on “school work” did not change, but general intelligence did/is. Even more specifically, what has been improving is the ability to think abstractly. In some places, scores for verbal or mathematical ability have declined while Raven’s scores have increased.
What we might be getting really good at is classification - the ability to separate things into groups based on common features. In a complex world, the ability to recognize groups with common behaviors (chunks) allows us to take useful action (as will be discussed below, the ability to use an analogy is critical and the ability to classify is related to the ability to use analogies). However, it is unclear that education is adapting. Most teaching is dedicated to accumulation of facts and resembles the idea of mastering the kind world of chess and not critical thinking which might be more useful in a complex world. Critical thinking can be measured, but shows no correlation to grades, as measured at top American universities across a wide range of subjects. In general, students understood material in their own major study area, but could not apply that knowledge to another area. General information of broad applicability is presented, but recitation of specific facts and methods is emphasized so overwhelmingly that it is missed by students. The book suggests that higher education should focus 1-2 years just on thinking before introducing content on what to think about. Given the rapid increase in information, the bigger challenge for students will always be in how to learn rather than remember what was once true.
One method of teaching this sort of thinking comes in the form of “Fermi” problems. These are problems that can’t be answered in the time frame allowed, but which can be approximately estimated. The book’s example is determination of the number of piano tuners in New York City. Starting from the city’s population, a series of assumptions gives rise to an approximation. For most purposes, answers of this sort are sufficient. One of the things it is most useful for is “calling bullshit” on “expert” analyses. We are routinely exposed to supposed facts, and the ability to do a back of the envelop estimation enables bad information to be identified and thus be ignored. Over-specialization deprives us of the thinking tools to do this sort of general analysis.
The author devotes almost an entire chapter to the music scene in 17th century Venice. The background is too extensive to recite here, but in essence there were a lot of orphanages. Venice was a wealthy city with a wealthy church that undertook to help these orphans as best they could. They would teach the children and where possible help them become apprentices or provide dowries to girls to marry. To read the description, they were exactly what we hope an orphanage would be. However, there was an interesting aspect of these orphanages from the perspective of breadth. They became musical hotbeds. Some orphanages had acquired instruments as gifts and given children, specifically girls, a chance to learn about them. In most of Europe, instruments were played by men, but the Venetian orphan performers were female. Some of these young women became the most prominent and accomplished performers of the age. Though they developed fame for individual instruments, they were multi-instrumental. Their small ensembles needed to perform a wide variety of music so performers needed to shift instruments between (and sometimes within) pieces. Thanks to their general competence, a number of (now forgotten) instruments were created just for them. It was an age of great musical innovation. The concerto form was developed, major and minor keys were created, and the great Cremona instrument builders (like Guarneri and Stradivari) were at work. Musical performance became a way for orphanages to support themselves. Antonio Vivaldi composed 140 concertos specifically for these performers. Not only were they multi-instrumentalists, but they taught themselves and each other how to play. They would improvise in public on their instruments. Older students taught younger students and training was formally part of the weekly schedule. In fact, some affluent parents tried to get their daughters positions in these ensembles.
Everything about our modern view of specialization and learning seems different from what happened in Venice. Online forums suggest forcing a child to choose an instrument by age seven, to have structured music practice, and trying to convince children under three to pick an instrument. Yo-Yo Ma is a famous prodigy, but before he settled on the cello, he played violin and then piano. Some of the same things happen in athletics. Parents want their children to do what Olympic or professional athletes do, but are not interested in what those athletes did when they were children. Elite athletes, as mentioned earlier, played many games. To be clear, there is a large premium for early intensive practice for performers in a classical setting. A classical performer is working in a kind world where reproduction is the ideal. Musicians in other genres benefit from a range of early experiences. Yo-Yo Ma is well known for seeking other musical genres to play, consistent with his early range of interests. The book provides numerous other examples of quite famous musicians who had a late or erratic start into their musical life and often had no formal training and never learned to read music. One way to frame these stories is that these people across the centuries learned how to learn music through irreproducible specific circumstances. Of course, this is how we all learned to speak, walk and ride a bike. Formal instruction did not matter and we chose what interested us to devote our practice time to. Some children learn to walk faster and some learn to speak sooner because they choose what they are interested in learning themselves.
Learning is a key activity in building range, but some ways of learning are more effective than others. Humans are quite good at learning, but they especially like to learn procedurally. We try to deduce the procedure for doing something and apply it repetitively, rather than deduce the principle. For example, college students were asked to evaluate whether 55 + 93 = 148; they knew to subtract 93 from 148 to get 55. Asked for another way to check the answer, they were stumped. They knew the rule “subtract the number to the left”, but did not generalize to use the other number. This is a failure in connection making. Students learn a procedure (rule), but don’t make the connection that allows them to generalize. The root of this problem is a desire to make learning easier and more efficient. It turns out that this has the opposite effect. We learn best when it is difficult. Teachers who offer hints, make it easier. Removing distractions makes the immediate learning easier, but long term learning harder. Being forced to imagine or discover the answer make subsequent learning better. In fact, learning is even better when the imagined answer is wrong. The more wrong the original “guess” the better the learning. The process of “guessing” the answer is called the “generation effect” and is part of why pre-tests can be so effective. Finally, standard teaching methods will focus on a subject for a while and then move to another topic. Students who had their learning interrupted with other material before a return to the original material had a more durable memory of the material. To repeat, durable learning is more likely when it is made difficult. Confusingly, an immediate test of knowledge favors the focus/easy approach.
Overall, this is called “desirable difficulty” and this difficulty fosters connection between bits of information. A study was conducted at the US Air force Academy. All students must take three math courses and almost everything is standardized. Hundreds of professors taught about 10,000 students over many years. Students were randomized after each course. Students received grades in each class and provided teacher ratings. At the end of the first class, there was a high correlation between grades and teacher ratings. But in subsequent classes, students who initially did very well struggled. Student who had lower grades and gave poor teacher ratings did much better in the subsequent math and engineering courses. Interestingly, higher initial ratings and grades were associated with inexperienced teachers. More experienced teachers made learning “better” by making it “harder” and forced student to learn more deeply and connectedly. This result has been replicated elsewhere.
These methods don’t just apply to intellectual tasks. Piano students were asked to make a specific 15-key left handed jump in 1/5 of a second. They were given 195 practice attempts. Some students used all 195 practice on a 15-step jump, but others did a variety of jumps (8, 12. 15, 19, etc.). The focus- practicers did more poorly than the variety-practicers. It is counter-intuitive that making the worker harder generates better results. It is easy to see short-term progress, but immediate impressions deceive.
One the ways that learning feeds thinking is by building up a stock of analogies. People use analogies to make estimates (budget, probability of future events) and to solve problems (how can adapt the solution to that other problem to this case?). While use of analogies is common, it seems that training people to use analogies has a significant effect on people’s ability to think. In particular, learning about new “classes” of analogies has an effect. These classes (chunks) of analogies are a form of mental model. As is often the case, the experiments were done with university students. There was relatively little difference between students from different major fields with one exception. The university had a major called “integrated science” where the students get minor degrees in the full set of science areas rather than concentration is any one. These students had a large number of analogies to apply to the problems and had significantly better results. These analogies allow a person to imagine the structure of the problems presented in a deeper and more holistic way, which means they had more possible strings to pull while untangling a solution. In another study of how laboratories used analogies, the most productive labs comprised people from a range of specialties, and the least effective comprised people with similar expertise. Diverse groups had diverse analogies to apply, but similarly-specialized groups had a limited set of analogies that weren’t always applicable. If it is hard to find individuals with range, a group with range can often achieve the same thing. In our world, the belief in specialization means that people must take the risk that a diverse early education will be rewarded later in their careers, because it is usually punished initially.
The earlier examples of athletes or musicians that jumped from topic to topic are an example of “match fitting” A person tries one thing, doesn’t like it, and tries another until they find the thing they like. Van Gogh tried general teaching, art dealing, carpentry, religious teaching, and bookstore clerk before finding oil painting at age 34. In a brief period, he became one of the most influential and recognized painters of modern times. The many things he tried were part of finding the right match for his deepest interests. In a modern context, comparison between educational systems that require early specialization and systems that permit/require later selection show that students in the latter case are less likely to make big career changes in the years following graduation. The career changes of early specializers suggests that a period of experimentation helps people try out different fields and then settle with a good fit. This diverse experience also means that they have more mental models and analogies to tap as their career progresses. The book makes the point that we are also deluged with advice about sticking it out, showing grit, or sticking to the plan. This might be good advice in some contexts, but career choice might not be one of those contexts. Similarly, even within a career, one company may be a poor fit for a person while another company is a great fit (but there may be two more companies in between). Switching in this context is a winning strategy.
Epstein considers the study of grit in this context. First, it is worth noting that grit studies were originally done on a highly selected group of people (US Army cadets). With some much in common, minor differences between them may have seemed bigger than they would be with a more diverse group of subjects. Second, people who stuck out the initial study period often dropped out soon after. Many concluded that the system had mistakenly picked these people or that these people we somehow “weak”, but an equally valid conclusion could be that the cadets were sampling this career and rejected it. Knowing when to quit can be a big advantage and that advantage can show up in many ways. Decades ago, companies depended on specialized people with specialized skills, so changing careers was very hard. Today’s companies depend more on generalists or on expertise built on experience, so matching goes on longer. For companies early in the match process, this is a problem (and one they probably can’t solve). For the individual, this is a powerful tool to achieve career fit and happiness.
The book devotes 2-3 chapters to the stories of people who tried numerous careers or took “odd” routes to prominence. These help build the case that success does not always involve early specialization, though the people in these stories did tend to go about their experiments with great intensity. They were not trying things out so much as testing and pivoting away towards a better choice. One such chapter discussed Gunpei Yokoi. Yokoi was hired directly out of university to maintain the card making machines at Nintendo (Nintendo got started as a playing card company). This was before they became a video gaming powerhouse. He was actually the entire maintenance department, and he had a lot of free time. Nintendo had been in business for a long time but was confronted by competition by other forms of gambling. The company diversified in many directions (all failures), when Yokoi was hired. He like to make things and one day he made an “extendable gripper” to allow him to reach for things without really reaching for things. He was playing with this, when the company president passed through. He was called to the president’s office the next day and told to make a few hundred to sell as toys (they sold 1.2 million). They sold very well and he was removed from maintenance and directed to build a toy R&D department. Yokoi already had many hobbies, but his biggest hobby was making things. Nintendo began making toys and did very well, but that had one flop that turned into the center of subsequent success. The toy was called Drive Game and involved a steering wheel, a plastic car and a race track. For the time, this was a complex electronic device that was expensive, fragile, difficult to make and frequently defective. Nintendo had powerful competitors in both the toy and electronic arena, so direct competition seemed unpromising. So, Yokoi adopted what he called “lateral thinking with withered technology” – today it might be called simplification. The idea was to use technology that was very mature to do new jobs rather than create new technology to do an old job. For example, Nintendo made a “Love tester” which made holding hands fun for teenagers. It involved technology that was already many decades old; it became a popular party game. Another example was radio-controlled cars. This is another complex electronic toy, but Nintendo decreased the complexity enormously by only letting the car turn left. Lefty RX was fine for racing on a counterclockwise race and kids soon learned how to use the car to make more complex maneuvers. But the big breakthrough came when Yokoi was riding a train back home. He saw a commuter playing with his hand-held calculator to pass the time. He realized that commuters would like something to pass the time – like a game. Shortly after this he was asked to drive the company president to a meeting (he was the only other employee who knew how to drive a left-hand drive vehicle in the company). During the ride, he mentioned his idea of a hand-held game to the president (who seemed to ignore him). The next week, some executives from Sharp visited him, at the request of the president. He explained the concept of a game device the size of a card that was played with two thumbs. A few innovations later, they had a device that could tell time and play a game. Three versions were launched hoping to sell 100,000. Demand rocketed and the Donkey Kong version alone sold 8 million units two years later. This led to the Nintendo Entertainment System and the Game Boy. Every device was the most technologically backward in the market but they were tough, easy to play and easy to program for. Game designers had an easy time using the simple technology and consumers loved the challenge. It was also deeply insightful about how people see and experience the world. They don’t require everything to be spelled out; they just need the right clues. Though trained as an engineer, Yokoi’s contemporaries did not see him as a great engineer, nor did he. His gift was to think widely about how to use existing things. Though Yokoi died before the Wii came out, it was inspired by his design principles. While some criticized the Wii as too simple, consumers loved it because it made playing easy for them.
One of the best features of this book is the wide array of stories, so the chapter that talked about Yokoi and withered technology also described Andy Ouderkirk who made the world best glitter as an outgrowth of work on reflective films (the films are used in many more ways than glitter). In a way, this is the opposite of withered technology because it involved all kinds of insights into the nature of materials and creation of things that were deemed impossible by experts. They key was that Ouderkirk was not exactly an expert in the field of reflective films, but his expertise was adjacent to it. The book describes numerous studies about inventors that conclude that the most likely breakthroughs come from people who might be described as polymaths. These are individuals with a core area where they have deep knowledge combined with a range of areas where they have moderate knowledge. Difficult problems with high uncertainty might be solved by a specialist, but a diverse team is more likely to solve the problem. But a team with a diverse individual is even more likely to succeed. Organizations have trouble with polymaths. Employee recruitment often seeks specialists and has no idea how to categorize a person with breadth. Even once hired, companies tend to lock people into a category because that seems to work fine for most people. Companies seeking innovation may need to change how they search and use people in projects to foster polymaths.
Narrow expertise has another weakness, overconfidence in certain facts. Expertise creates a kind of blinders that makes it hard to see a problem from other perspectives. People with moderate, but not deep, expertise often outperform experts to predictions in uncertain environments. It is not a question of intelligence, but it is a question of openness to multiple interpretations. A person with broad experience is more likely to consider alternative explanations or possibilities. Teams of such people are even better at making such judgments which use an open-minded approach to analogies, evolving information and probability to drive discussion. To be even clearer, collections of individual expert opinions were not as good as the group predictions arrived at by interaction. Experts see their answers are answers, but non-experts see them as hypotheses to be tested. Experts are more likely to suffer from confirmation bias in their area of expertise (other studies show that the same people outside their area of expertise suffer less rigidity). In this context, breadth involves actively seeking contrary information and then seriously considering its validity.
In one sense, a person seeking to be a polymath dedicates themselves to maintaining the curiosity and energy of an amateur. Today we think of amateurs as people who aren’t very good at something, but the meaning is much closer to “one who adores something”. The chapter dedicated to deliberate amateurism recounts multiple stories of people who changed subjects multiple times, worked because they loved the work, and were devoted to “trying something out”. One such person was Andre Geim who may be distinguished by being one of few people to win both an Ig Noble prize (see links below) and Nobel Prize. Both prizes resulted from what Geim called Friday Night Experiments. The Ignoble prize was won for levitating a frog with a magnetic field. The Nobel Prize was for discovery of graphene, which was first made by ripping a piece of Scotch tape off a block of graphite. When this work was first submitted for publication, one reviewer described it as impossible because something so simple could not possibly work while another considered it trivial. The same mind also create Gecko tape in a similar informal experiment. “A paradox of innovation and mastery is that breakthroughs often occur when you start down a road, but wander off for a ways and pretend as if you have just begun…” (quoted from Sarah Lewis) makes the point that a certain open-mindedness is required to develop range. It is not just openness to new experience and information, but also to the insecurity of ignorance. The people profiled in the book demonstrate “tools for thinking” more than knowledge and most had a mentor who encouraged them to retain a wide view and curiosity.
Another way to look at these people is to see them as occupants of interfaces. A portion of their individual success arises from being able to bring many perspectives to bear on a problem. It seems that groups are much the same. Some groups are characterized by movement of people and ideas from similar fields while others have more diversity in their exchanges. When the same people collaborated constantly, they were efficient but rarely high impact. Groups that experienced more novel interactions were more likely to have bigger impacts. In fact, the real force for impact seems to come when an entire research ecosystem has more novel interactions. In other words, the entire field of knowledge is building more interfaces with more fields of knowledge. It is as if people with many interfaces are better able to import and export good ideas and information from other fields than people with highly defined boundaries. This phenomena was observed in physics, economics and Broadway musical productions.
The idea that got the author started was that early specialization was a sort of life hack that would lead to greater success. By focusing on one thing and practicing in an efficient way, people would be more skilled and more successful. This is true in some parts of life with relatively low uncertainty. But much of life is uncertain and this means that specialization may be the opposite of efficient. Many of the examples from the book are very unusual, cherry-picked to tell the story (as acknowledged by the author), but that is not different from the cherry-picked stories of the benefits of early specialization. The author takes this perspective about individuals and applies it more widely. Society must solve many kinds of complex problems, yet has come to idolize specialists and application. By overly focusing on immediate benefit, society has imposed a sort of early specialization. Too many major breakthroughs come from inefficient what-the-heck experiments that would be eliminated in strongly “efficient” or “focused” environments. The ironic question becomes how do you cultivate inefficiency, dabbling, and following a hunch? People, organizations, and societies that want to succeed in complex environments need the range to apply diverse perspectives to problem solving and the mentality to see this as valuable.
Comment and interpretation
- When organizations have problems to solve, the standard approach is to find an expert. This book makes you wonder what we mean by expert. One view might be someone who has deep knowledge of an area with the ability to apply the mental models of that field in combination with the knowledge to the problem. This book suggests that a different kind of expert is someone who knows how to think about problems, but only has moderate knowledge and experience. It suggests that organizations must think much more carefully about the kind of problem they have, before deciding the kind of expert they want to engage. The odd thing might be that the right kind of expertise is already in the organization but being ignored out of a misunderstanding of what is required.
- The book comments that most professional training uses the concentrated/focus/efficient approach. This means that the perception that such training does not make much difference is probably correct. Students feel like the course is effective because it seems efficient and tests taken during or immediately after the course show good learning, but this knowledge fades fast. If we want professional training to be more effective, it must seem inefficient, unfocused chaotic and disorganized. Years ago, I was in a Cargill training course on communication. The teachers barely taught anything. They just gave us problems and exercises, and left us to work things out. At one point, a classmate noticed that one of them had gone to sleep during a discussion (they were outraged at this obvious disengagement, I now suspect that the instructor was just sitting with closed eyes, which I do too). What I noticed was that I learned a lot from this experience (I did not notice the napping) and that knowledge stuck with me for decades. I was constantly lost and confused in the class, and this reminded me of graduate school.
- I struggled in my undergraduate studies where I was expected to learn (memorize) large volumes of material. I don’t have that kind of memory, and it took me years to figure this out. Somehow high school had not exposed this problem, though retrospectively – there were hints. I came to understand that I am a slow learner. One of the big challenges in my graduate studies was a set of tests called “preliminary exams”. These were custom tests created just for me by the committee overseeing my education (other schools and departments may do this differently). At my school, there were three written and one oral exam. Within the broad scope of the topics selected by the committee and me, anything was in scope. In my case, this included basic biochemistry (excluding lipids, thankfully). I had to know all of the basic compounds, their synthesis pathways, structures, etc. A huge memorization task and thus a big problem to solve. I knew the approximate dates of my exams, so I started one year ahead. I wrote down all of this information on cards and spent one month studying them. Then I ignored the material until 6 months out, when I spent two weeks studying intensely. At 3 months, I spent one week and at one month I spent one day. The exams were not easy, but my problems had nothing to do with memory. The professors challenged my memorization and I could answer. I had no idea that theory would show that this is the right way to build durable memory. I remember much of that material nearly 40 years later, though I had no real use of it for a very long time. At the time, I thought this approach was just a way to work around one of my weaknesses. It seems I luckily found a good general way to learn.
- Americans perceive that the quality of education in the US has fallen over the years. Objectively, this is incorrect. Standardizing testing allows comparisons across years, and the bottom line is that the tests have been getting harder and are asking different kinds of questions. Questions in the past were mostly procedural, but now require both connected thinking and procedural thinking. This type of learning may involve fewer facts, but also enables better generalism. Your children are smarter than you, and they are getting a better education. Our standards have been moving upwards, and perhaps a bit faster than teaching practices have been able to keep up. If your children are struggling – rejoice. It is easy to see how this also fits with an understanding about growth and fixed mindsets in education. If the struggle is what drives learning, then a growth mindset that accepts that struggle makes it easy to accept the struggle.
- I work at a Midwestern US company (Cargill) with a long history. For many years, the typical manager had a particular background and came up through the company in a particular way. In the mid-1990s when MBAs were becoming more common in business, the company hired graduates from highly reputed schools from the east and west coast states. Many such hires left within a few years, in part because they did not feel that the Midwest was a good fit for them. Midwestern graduates slowly displaced them. My brother and sisters finished college and moved to California. After a few years, they moved back because they did not feel like they fit in. Many of the coastally educated MBAs hired in my company, grew up here. Companies seeking diversity must struggle with many layers of fit and even within the broader American culture, there are elements of fit that result from how people grow up.
- What is the connection between the problem of finding fit from the individual’s perspective and the dynamics for an organization seeking a diverse workforce? This is interesting in two different aspects of diversity. First, does good fit between individual and organization inevitably mean that there must be less diversity in thinking styles? Second, if you are attempting to achieve greater diversity in other dimensions (gender, race, geographic origin, religion, etc.), how hard is it to achieve the desired diversity if the diversity itself makes it harder to find fit?
- Which leads to polymaths – people who are individually intrinsically diverse in thinking style. Corporate life can be quite difficult for a person who isn’t “one thing” or who seems to change interests. It is clear from some of the description in the book that polymaths are curious and are not good at predicting what they will be curious about. If an organization wants to create and stick to a plan, polymaths are probably not that compliant. The book also make clear that polymaths make look scattered to others and be described as lacking focus. In the context of seeking adjacencies, I suspect they are actually relatively focused. It is just that they focus on the interface rather than the core, and they see that everything has many interfaces.
- One of my colleagues once commented that “one experiment is worth 100 opinions”. The discussion in the book about the importance of resisting expert opinion reminds me of the most important experiment I ever did. My boss at the time was an excellent organic chemist. I had an idea to use a particular reagent to titrate a particular enzymatic reaction. He said it would not work and explained why. I’m not that good a chemist, and could not understand why it would not work the way I expected. So I did the experiment and it worked as I expected. This opened up a whole area for us to explore, because I did the experiment. I was wrong and he was right plenty of times, but they key was that opinion had to be backed up with good experimentation. It is not always simple to do a good experiment, and experiments can contradict each other. But this is how you really learn. Some companies, some disciplines, and some national cultures believe in expert opinion and knowledge. Expertise works, until it doesn’t. This is one of the reasons that I distrust consultants; it seems impossible to do useful experiments to test their recommendations.
- The concept of foxes and hedgehogs has become fairly well covered in recent years. I am writing this book summary in the middle of the Covid 19 epidemic, and it is quite interesting to see how politicians, health experts, and the media are interacting. The situation is very uncertain and conditions are both similar to and different from past epidemics. There are numbers flying all over of varied quality with limited understanding of their limitations. The last few years as seen an erosion in public faith in science, perhaps because indifference seemed safe. Now that indifference seems unsafe, there are many challenges. How do we look at experts? Politicians that have expressed certainty have been exposed, while politicians that have expressed uncertainty have bene criticized for not being decisive. There are many models being used, but people do not understand the use of such models in making decisions – and fail to recognize that these models are used to predict ranges of conditional outcomes. And the consequences of all of these decisions are quite domain spanning. This is a condition highly suited to generalists in health and economics, but probably not for either health or economics specialists.
- Information on the Ig Nobel Prize can be found at https://www.improbable.com/ig-about/winners/ . Information on Gecko tape can be found at https://web.stanford.edu/group/mota/education/Physics%2087N%20Final%20Projects/Group%20Gamma/gecko.htm . Information on graphene can be found at https://en.wikipedia.org/wiki/Graphene .
- Academic science has many “meta” phenomena, often related to citation. Someone did a study of about 500,000 journal articles and observed that less than 1 in 10 made a novel combination of citations. Scientific authors often cite prior work as the basis for creating a hypothesis or for corroborating a finding. These results suggest that authors mostly looked at the same things as all other authors and made the same connections. There is a huge number of siceintific papers published every year – nobody could possibly read them all; there should be many novel connections to be made. Papers that made such novel connections between prior art were published in less prestigious journals and were slow to be cited further. But it was the papers in this class than tended to go on to have the highest impact and by 15 years were much more likely to be in the top 1% of papers cited. This raises the question of why we are so initially resistant to new information – here in the form of a new combination. Organizations often made a big deal about being unique or differentiated, but a discipline that is entirely about change (research) demonstrates that safe and comfortable thinking is the norm. An organization that wants to be differentiated probably needs to reject things like best practices (conforming to others’ procedures), intensive judgement of ideas based on knowable value, early proof-of-value and intensive planning.
- In this context, it is interesting to consider one of the growing fetishes of innovation – customer validation. It is widely observed that innovation success is low probability. Many innovations fail commercially – at least in the time frames that interest most businesses. One reason often proposed is that the innovations reflect the innovators interests rather than the customers’ needs. This definitely happens. The solution is to look much more closely at what the customer needs using both direct (survey, focus groups, etc.) and indirect means (anthropology, empathy, etc.). But rather than stopping there, rapid prototyping can be used to check that the solution is valuable to the customer before going too far into development, and then pivoting to better match their needs. All of this should lead to better, faster fit of solution to problem. The cost of this may be that bigger, better ideas that are initially hard for the user to understand get rejected in favor of easy-to-understand incremental ideas. It is absolutely fascinating to learn how much of modern telecommunication technology was created before 1970 by people who studied the practical problems of a telephone system and then were imagination-driven in finding solutions (see Idea Factory ). And then to see this essentially repeated at the Palo Alto Research Center (see PARC ). There is a strong negative pressure against “technology push” innovation, and strong positive drive for customer pull. Does this lead to incrementalism? To create a breakthrough, MUST you do something they were not asking for and did not even recognize as a need? And that leads to a question about focus. What percentage of breakthroughs come from dabbling? If the people in a company have no time to dabble and feel that they need approval to follow hunches, is that company doomed to incrementalism?
- Scientists are taught something about past scientists - often to illustrate a principle that we should use going forward. One of the paradoxes of science is that the instrumentation and methods available to scientists even 50 years ago was primitive. Those people had to work so hard to do an experiment compared to people today. The pace of science progress has really accelerated in the last 50 years as better equipment and materials have been created. But, there has not been much fundamental progress by comparison. Scientists in the past were much more dependent on thinking than we are today. Interestingly, they were also much less specialized. Darwin went on his trip around the world as a geologist. Marie Curie won two Nobel Prizes – one in physics and one in chemistry. Newton not only invented calculus, but he discovered fundamental laws in motion, optics, and heat transfer and went on to make contributions in economics. Benjamin Franklin was a diplomat, printer, physicist, and more (Franklin was consider one of the most prominent scientists of his day and his work on electricity was important).
*text in italics was quoted directly from the book
Recent Comments