Skip to main content

Why specialization can be a downside in our ever-changing world

Why specialization can be a downside in our ever-changing world

/

David Epstein’s new book, Range, explains the benefits of taking our time and learning by doing

Share this story

Graphic by Michele Doying / The Verge

Popular wisdom drives home the importance of a head start and specializing early. Not so fast, advises journalist David Epstein.

After Epstein wrote about the famed 10,000-hour rule in his first book, The Sports Gene, he was invited to debate Malcolm Gladwell, whose book Outliers had brought the rule into the mainstream. To prepare for the debate, Epstein gathered studies that looked at the development of elite athletes and saw that the trend was not early specialization. Rather, in almost every sport there was a “sampling period” where athletes learned about their own abilities and interests. The athletes who delayed specialization were often better than their specialized peers, who plateaued at lower levels.

Epstein filed this information away until he was asked by the Pat Tillman Foundation to give a talk to a group of military veterans. These were people who were changing careers and having doubts that are familiar to many: whether they were doomed to always be behind because they hadn’t stuck to one thing. Epstein became interested in exploring the benefits of being a specialist versus a generalist, and ended up writing Range: Why Generalists Triumph in a Specialized World (Penguin Random House).

The Verge spoke to Epstein about “kind” versus “wicked” environments, the importance of doing instead of planning, and the difference between having range and being a dilettante.

This interview has been lightly edited for clarity.

When did this cult of specialization develop? You mention the 10,000-hour rule, of course, but it was definitely around before then.

David Epstein
David Epstein
Photo: Deb Lindsey

The 10,000-hour rule just gave it more of a common language and intensified it, basically. I really do think that Tiger Woods himself [whom Epstein writes about in the beginning of the book] helped kick this into high gear when he went on television as a two-year-old. It started this explosion of parents thinking, I have to get my kid doing this. There’s a number of prodigies like this. It’s easy to conceptualize giving someone a head start. We give a lot of lip service to “try and fail,” but we don’t actually encourage it when the rubber meets the road.

I also think that sharing videos and movies of prodiges, whether that’s Searching for Bobby Fischer or the Polgár story or the Tiger story — and those things are always in classical music, chess, or golf because those domains are so amenable to that sort of thing — kicked off our natural tendencies to want to give kids a head start. And the 10,000-hour rule codified it and extrapolated it to every other domain, where it doesn’t necessarily belong.

Right. One of the key ideas of your book is that early specialization and lots of deliberate practice does work in certain “kind” environments, but it’s not as useful for succeeding in “wicked” environments where it might be better to be a generalist. Tell me more about this distinction.

Obviously, Tiger Woods went on to be the best golfer in the world. Nothing I wrote was meant to say this didn’t work, but the problem was extrapolating that to everything else that people want to do. Golf is what psychologist Robin Hogarth called a “kind environment” and Hogarth set out the spectrum from kind to wicked environments.  

The most kind environment is one where all the information is totally available, you don’t have to search for it, patterns repeat, the possible situations are constrained so you’ll see the same sort of situations over and over, feedback on everything you do is both immediate and 100 percent accurate, there’s no human behavior involved other than your own. Whereas on the opposite side, with wicked environments, not all information is available when you have to make a decision. Typically you’re dealing with dynamic situations that involve other people and judgments, feedback is not automatic, and when you do have feedback it may be partial and it may be inaccurate.

Most of the things that people want to do are much more toward the wicked end of the spectrum than golf. You don’t know any of the rules, they’re subject to change without notice at any time, and over and over. Early specialization is not the best way to go.

You talk about the importance of a “sampling period.” But how long should this optimal sampling period last? Someone could theoretically just sample forever, but that doesn’t make sense either.

“We learn who we are in practice, not in theory.”

That’s the million-dollar question. Or billion-dollar question. In sports, it looks like athletes who want to be the best start cutting out other things in midteen years, around 15. But it’s not clear to me whether that’s the optimal way to do it or if it’s because they are forced to specialize then. We saw in a study of German soccer players, some of whom went on to play in the World Cup, that they were still doing other sports informally past the age of 22. With Cirque du Soleil, they started having their performers learn the basics of other performers’ disciplines — not because they were going to perform it, but because it helped them be more creative for designing new shows and that cut injury rates by a third. But I don’t think we know exactly what would be optimal.

In addition to discussing the advantages of being a generalist, the book also touches on the disadvantages of specialization and how that can blind us. Do you have a favorite example of this that you like to talk about?

Part of what got me interested in this is that I realized I had committed statistical malpractice in my own master’s degree. I can’t say I “love” talking about it, it’s a little embarrassing. And that research is still published.

The problem is, when I went to grad school, in geological sciences, I was rushed into this very specialized research before I’d even been taught how the tools of thinking and science work. So I was studying very specific information, not knowing what was happening when I was pressing the computer buttons and getting statistically significant results, and publishing them and getting a master’s degree for that.

There’s a replication crisis in science and a huge amount of it is exactly what I was doing: people not thinking about how their statistics work. You can get big enough data now and there are powerful statistical programs, so you don’t have to know how scientific thinking works. I think that’s a huge problem.

At one point in the book, you relay the advice “first act, and then think,” which runs counter to much of what we’re told about the importance of planning. Can you unpack that a bit?

That’s from Herminia Ibarra, an organizational behavioral specialist who studies career transitions. She gave me one of my favorite phrases that’s related to the “act and think” one. It’s “We learn who we are in practice, not in theory.” There’s a lot of research that shows we can take personality quizzes and everything, but our insight into ourselves is restricted. It’s similar to the end-of-history illusion, where we recognize that we’ve changed a lot in the past, but think we will not change much in the future, but we’re wrong at every single stage.

This industry attempts to just give you a quiz or give commencement advice to sit down and introspect and think about what you want to do, and it’s really contradictory to what we know about how people develop and how personality develops over time. We have to actually do stuff and reflect on it. And that’s how we learn about our skills and interest and possibility for the world, as opposed to having a theory of ourselves and assuming. Try stuff and take time to reflect. The best learners have the trait of reflecting on things they’ve done, because they’re learning about who they are.

Another thing that is interesting is the idea of having intellectual range and taking in lots of information. How does that dovetail with all the research on cognitive biases that keep us from believing information that contradicts our beliefs? How do we have intellectual range?

It’s really difficult. Algorithms reinforce that and unless you stop yourself and realize what’s being done to you, you don’t think about it. That’s partly what chapter 10 is about. You look at people who develop good judgment about the world, and people who were really specialized and had a narrow focus actually got worse as they accumulated information because they were better able to fit any story or whatever their views were. One of the main traits of people who had better judgment and were able to avoid falling into their own cognitive biases all the time was a trait called “science curiosity.”

“How do you capture the benefits of range even at the point in your career where you have specialized to a certain degree?”

There were clever studies where people were given statistics to analyze and sometimes it’s just some bland clinical trial of skin cream and other times it’s something very political, like whether gun control reduces deaths. Numerate people often become innumerate when they’re faced with those numbers in that context. It’s not an issue of bright — they could interpret the numbers well when it wasn’t political. The people who bucked the trend were the ones who were highest in science curiosity. Not science knowledge, but science curiosity measured by the fact that when they were faced with information that didn’t agree with their preconceived information, would they follow up and research broadly or would they put that aside and ignore it and leave it there?

So I think we have to very proactively try to step outside the algorithm and do the opposite of what our inclination is. We should see if we can falsify our notions. That was a hallmark of what the people with the best judgment do. It ends up with them widely gathering a large array of sources to try and test their own ideas. You can get so much positive feedback for not doing that as long as you stick to your little corner of the universe.

For those of us who stick in the same field, is it possible to be a generalist and a specialist?

At the end I focused on scientists and scientific research. Scientists, to the outside world, are the epitome of specialization in one sense, and I wanted to make sure that I included people who were viewed that way. Among these people who, compared to the population at large, are very specialized, what does it mean to have range? How do you capture the benefits of range even at the point in your career where you have specialized to a certain degree?

So I looked at people like Andre Geim [who has won both the Nobel Prize and Ig Nobel Prize, which is given to “trivial” research]. I called the guy who started the Ig Nobel and they tell people ahead of time so they can decide if they’d rather not have it and turn it down. But I think Geim was proud of it, basically. He talks about how “it’s psychologically unsettling to change what I do every five years, but that’s how I make my most important discoveries.” He says he doesn’t do research, only “search.” I enjoyed that because we all specialize to one degree or another and the question is how we capture the benefits.

I loved your book and have recommended it to several friends because it speaks to worries that so many people have, like “Am I just lazy because I can’t stick to one thing?” To an extent, the book helps assuage some of those fears, but I couldn’t help wondering: when do you have range and when are you a dilettante?

I don’t think it’s great to give advice like, “Don’t worry about being interested or hard-working at anything.” I like to think of the study of inventors at 3M. They identified generalist and specialist inventors, but there was also a class of inventors who didn’t have that much breadth and who didn’t have that much depth. They didn’t tend to make contributions. They were the dilettantes. They flitted between things to some extent, but didn’t learn about as many different technology classes as the generalists. But also tended not to go deep in any particular technology, so they ended up without an intellectual home, but also without the ability to connect disparate domains in a novel way.

I think that’s symbolic of the difference. You have to give a real effort and be ravenously curious about things because part of the hunt for what economists call “match quality” is diving into things in a way that gives you maximum signal about yourself. And if you’re superficially on the surface, I don’t think you’re getting the signal that helps you find where you are in the world.