AIQ speaks to author David Epstein about his new book, Range, which explores how generalists succeed in a specialised world.
7 minute read
In his book Outliers, Malcolm Gladwell popularised research that indicates world-class expertise in any field requires 10,000 hours of dedicated practice. It follows that highly successful people will find their vocation early and ruthlessly specialise.
Or does it? A recent study has cast doubt on the 10,000 hours premise.1 And in his new book, Range: How Generalists Triumph in a Specialised World, author David Epstein argues that breadth, rather than narrow specialisation, is the key to success. Generalists are more adaptable and more likely to notice productive connections. In a fast-changing and devilishly complex world, these are key advantages.
Epstein cites examples from across business, culture and sport. Tiger Woods might have become a world-beating golfer by starting at an early age – but another great, Roger Federer, spent years sampling different sports before bringing what he learned into tennis. Nobel Prize-winning scientists are much more likely to have an artistic hobby than their less-successful peers. Broadway plays that combine different genres are more likely to be hits at the box office.
In this Q&A, Epstein introduces AIQ to his findings and offers some tips on how individuals can find their own range.
What would you say is the key message of Range?
The obvious message is that society has tended to overvalue specialists and undervalue generalists. However, another message is that sometimes the things you can do to cause the most rapid short‑term improvements can actually undermine long‑term development.
What are the methods individuals, companies and policymakers can use to “embrace the potential that lies on the interface of domains and disciplines”, as you put it in the book?
One has to do with the way we hire. I have been to conferences where people talk about trying to automate human resources, because there is so much résumé information online you can find people who have a direct line of experience to whatever you are looking for. But the work of Abbie Griffin, who studies so‑called serial innovators, shows this approach will screen out the potential serial innovators, because these people tend to have “zigzag” paths, where they have worked and have networks across multiple domains.
Serial innovators tend to have 'zigzag' paths
One of the experiences that led me to this project was working with the Pat Tillman Foundation, named after a former professional American football player who left in the middle of his career to join the army. He was killed in Afghanistan. The foundation gives scholarships to military veterans to aid career changes. I was on the selection committee and the first thing I noticed was the résumés often look a little disjointed. But when you start to learn more about the applicants, you understand there is actually one narrative of individual growth, and they have taken left turns in response to things they have learned, opportunities they did not know existed before, or skills they have uncovered. Then it makes a tonne of sense.
Organisations need to understand these journeys of personal growth. That is how you get people with breadth – what I would call range – and serial innovators.
Beyond hiring policies, are there any other lessons organisations can take from Range?
One thing companies can do to expand their internal range is to make their teams porous. Bill Gore founded the company that created Gore‑Tex based on his notion that a company did its most innovative work when in crisis, because suddenly domain boundaries go out the window and everybody starts figuring out what everyone else’s capabilities are, and working together. He wanted to make that process systematic in a way that did not require a crisis, so there was a lot of moving people between teams.
There should be some flow between teams of people to bring in new ideas
Networks that give rise to creative breakthroughs have porous boundaries between teams; networks that do not give rise to breakthroughs are those in which the same people collaborate with the same people, over and over again. That turns out to be the case whether for Broadway plays or scientific research. That is not to say you have to shuffle everyone all the time, but there should be some flow between teams of people to bring in new ideas. In the process, people familiarise themselves with other areas of the business.
What are the risks of overspecialising? Do you have any examples of the negative consequences of an overly narrow focus?
When I was doing investigative reporting on the medical industry, I started noticing the perverse outcomes that resulted from increasing specialisation in medicine.
For example, specialised surgeons have fewer complications – but there is also evidence specialised surgeons are more likely to do procedures on people who do not need it, so it is a double‑edged sword. Take partial meniscus repairs, which may be the most common orthopaedic surgery in the world. Someone has knee pain, comes in to get imaging of their knee; the surgeon finds a little tear in their meniscus – a crescent‑shaped piece of fibre in the knee – and fixes it. This has been going on routinely for years.
Finally, a team in Finland decided to study this on a large scale and used a control group, in which some people had “sham” surgery, meaning they would have an incision in their knee, the surgeons would act as if they were performing surgery, sew them up and send them home. Those people did just as well as, and sometimes better than, the people who had real surgery. It turns out that maybe the most common orthopaedic surgery in the world does not work, and yet specialists continue to do it because it is what they are trained to do. One of the many reasons healthcare costs have gone completely out of control is the epidemic of unnecessary treatment that, in some ways, is an outgrowth of increasing specialisation.
What are the implications of these findings for training and development?
In the US, our education system was built for the industrial economy and came out of Taylorism, which is basically the science of management efficiency. People were trained to have the basic knowledge needed for industrial economy, where they could expect work next year to look like work last year. They could do the same things over and over.
Workers will have to reinvent themselves more frequently
Now we live in a knowledge economy where work next year might not look like work last year. Many people are stuck with a specialised set of skills, unable to adapt. That has caused tremendous social turmoil in a lot of countries that have switched rapidly from an industrialised to a knowledge economy, faster than workers could adjust.
Workers will have to reinvent themselves more frequently, multiple times over their career, in a way they did not have to in the past. We will need to set up systems that support people’s reinvention, unless we want what we have now: which is a lot of people losing manufacturing jobs and unable to find another job.
Will school-level education need to change as well?
The typical way teaching works, and certainly the way I learned maths, is by “using procedures”, where you essentially teach someone the way to execute procedures, algorithms or sometimes tricks.
That works well in getting people to make rapid progress in what they are doing, but the problem is it does not impart the conceptual knowledge that allows what psychologists call “transfer”. We often lose sight of the fact that transfer is ultimately what you want from a lot of education. It is the term psychologists use to mean your ability to take skills and knowledge and apply them to a problem you have not exactly practised before. That is what you ultimately want but it requires you to form broader conceptual models that allow you to bend your knowledge to a new situation.
How could this different type of teaching work?
A study just came out in which a bunch of seventh-grade maths classrooms were randomised to different types of maths learning.2 Some got “blocked practice”, which means the teacher teaches them a type of problem, like problem type A. They practise, practise, practise, then move on to problem B and problem C, and so on. They get really good at executing whatever procedure they have to. The kids rate their learning as high. They feel they are learning a lot, because they are getting better in front of their eyes. They rate their teacher as good, because they are making progress so quickly.
Other classrooms were randomised to what is called “interleaved training”, where, instead of getting A, A, A, B, B, B, they get A, C, B, D. It is as if you have all the problem types in a hat and you draw from it at random. In that situation, the kids get frustrated, rank their learning as lower and rate their teacher worse, because they are not making progress as rapidly. But, instead of learning how to execute procedures, they are learning how to match a strategy to a type of problem.
When test time came around, the students with interleaved practice blew the blocked practice groups away. They were learning the same problems; it was just that these were arranged in a way that made initial progress slower and more frustrating, but which forced the learners to build a conceptual model from matching strategies to problem types, instead of just executing procedures they had memorised.
Can these insights be applied to everyday life?
Once I started learning about this research, I used interleaving any time I possibly could, in anything I wanted to study, and also things like spacing – another so-called “desirable difficulty”. If you want to retain knowledge, study it, wait until you have just about forgotten it, and then study it again.
In a classic study, two groups were taught some Spanish vocabulary: one group got eight hours of practice on one day and then a test; the other group got four hours on one day, then four hours a month later and then a test. The group with eight hours did better on their test. Then, when both groups were brought back eight years later, with no study in the interim, the spaced practice group remembered 250 per cent more, with no study in the intervening time. One of the ways you move knowledge to your long-term memory is by essentially waiting until it has just been buried, and then drag it back up.
Can technology help people form connections and find range?
Absolutely. A Massachusetts Institute of Technology (MIT) study looked at how business professionals use their social media accounts. On Twitter, for example, the rough pattern was that most people used their social media to follow people who were already in their domain or social sphere, or who entertained them.
We should think of social tools as a way to expand our intellectual tendrils
But a smaller number of professionals constantly curated their Twitter networks. They looked for people outside of their domain. They were constantly taking people off and adding others on, cycling through different industries. The study found project proposals from people who used their social media networks in this way to make connections in different domains were systematically rated higher by their bosses. We should think of those social tools as a way to expand our intellectual tendrils, as opposed to just sharing memes.
You begin the book with a comparison between two sporting greats: Tiger Woods, who specialised in golf from the age of three, and Roger Federer, who was more of a generalist. What does this teach us about range?
After I wrote my previous book, The Sports Gene, I was invited to a debate with Malcolm Gladwell at MIT, co-founded by the general manager of [basketball team] the Houston Rockets. Gladwell and I had never met, and he had written about the importance of early specialisation in sports as an insurmountable advantage. I was the science writer for Sports Illustrated at the time, so I said, “Okay, but that’s just hypothesis: I am going to look at the data.” I saw that, in fact, in almost all sports, when scientists track athletes who eventually become elite, they see a so-called sampling period, when they play a wide variety of sports and play in lightly structured or non-structured activities. They learn about their interests and abilities, and delay specialising until later than their peers. I picked Roger Federer because he is representative of what the science says is the norm.