In the first of a new series on this blog, I talk to the American writer David Epstein about his book “The Sports Gene: What Makes the Perfect Athlete”. “The Sports Gene” has been seen as a riposte to Malcolm Gladwell’s “Outliers”, which popularised the view that the key to success, in sport as well as in other fields of endeavour, is not native talent but assiduous practice (Gladwell repeatedly invoked the so-called “10,000 hours rule” in this connection).
Did you conceive the book as a salvo in what one might call the “talent wars”?
That’s an interesting question. I think so. There was nothing in my book proposal about these other books, like Geoff Colvin’s Talent Is Overrated or Outliers. But as I began to investigate these questions of nature vs nurture, I realised that the popular writing narrative was all on one side of the issue. And that was where the public discussion was. It didn’t make much sense for me to discuss genes if the prevailing idea was that they didn’t matter at all. So that’s how I got into assessing the “10,000 hours rule” in the first place.
Why do you think the discussion in the wider culture had taken that turn, why had talent become a dirty word? Why did we become much more interested in practice, hard work and doing the “hard yards”, the 10,000 hours?
I think that genetics has suffered from unfair media coverage in the sense that every time something is reported about a gene, it’s as if it’ s destiny, this single gene. That’s not only not usually correct, but not a very nice idea to think about—because we all work hard, and we all want our work to matter (and I think it does matter). So, I think there was a perception problem created by a lot of reporting around genes. Gladwell wrote a very compelling and interesting book [Outliers]. That said, I think that the way we can all get to the best outcome is by finding the best individualised path, instead of assuming we can all do the same kind of practice and end up in the same place.
I wonder, then, whether the title of your book is potentially slightly misleading, since it might be thought to suggest that there are “genes for sport”, as it were, in that uncomplicated sense that you just referred to.
I did go back and forth a lot with my publisher about that. The title of the article that was sort of the seed of the book was “Sports Genes”—plural. They said to me, “Look at Matt Ridley’s The Agile Gene, for example—it’s a metaphorical concept.” When you say someone has “the music gene”, it doesn’t mean there’s one gene for music, it’s a way of saying, metaphorically, that someone inherited some combination of traits that that equal talent. Ultimately I decided to trust them on that. But I definitely think that that’s a valid question and one that I’ve been asking myself.
One of the many interesting things about the book is the insight it gives into the science of what you call “innate physical hardware”. When would you say the science of that really took off? Is that a relatively recent thing?
I think in terms of genes it’s very recent—as recent as the last decade, since the sequencing of the human genome. But in terms of looking at particular hardware, I actually think, in many cases, we’ve actually gone backwards. Part of the problem was that some of the scientific work in the 20th century was looking at things like skeletal proportions and body proportions, and it became associated with some unscrupulous scientific practices—comparing ethnicities and so on. The result is that there is quite a lot of physical hardware literature that is basically ignored now. And genetics seems so cutting-edge. But it’s really easy to do genetics poorly—if you have somebody in your lab, you can get their saliva really easily and see what their gene sequence is without knowing much about the genes you’re studying. People overlook things like measuring Achilles tendon lengths and instead are skipping straight to the genetics instead of focusing on some of the other physical hardware that was studied in the 20th century and then left alone for a long time.
It’s obviously far too crude to say that this this book is the anti-Outliers. Indeed your discussion at the beginning of the book of “occlusion tests” and attempts to measure the perceptual abilities of athletes actually seems to support the 10,000 hours thesis. The argument for repetitive training seems to work well in the case of brain automation and the sort of improvements in refined perceptual abilities that elite athletes can achieve.
That’s right. That is a skill that I would have guessed would have come more from innate fast reflexes; but it turned out to be very much the result of practice. That said, while I think it shows how miraculous training can be, how important practice is and how useless even the most talented athlete is without it in a sport that requires perceptual expertise, I still see it as not necessarily supporting the 10,000 hours argument in the way that’s often made because even in those skills people progress at very different rates. Once they’ve learned the sport-specific skills, hardware becomes increasingly important.
One of the key statistical insights of your book is that large amounts of accumulated practice can’t explain variance in skill. That seems a really central move in the argument.
That shows up in every single study and I thought that people would be sort of surprised to learn this. The 10,000-hour principles in some writers are more strict than others. Gladwell, for example, says that talent matters too. And some of the others think genes really don’t matter. But I thought that readers who are familiar with this stuff would be interested to know that after scanning the entirety of the scientific literature about learning motor skills, I discovered that there are no studies in which practice accounted for all of the variation in performance—in any complex skill. Occasionally, if the skill was very simple, like hitting a button, then practice would account for all the variants. But in any skill that was more complex, it never accounted for all the variants—it never accounted even for most of the variants.
So you are offering us an account of skill which is based on both hardware and software? It seems to me that this is not straightforwardly an anti-10,000 hours book: it’s about splitting the difference between an argument based on training and one based on genetic inheritances.
Yes. It’s interesting you mentioned that because I was worried about being seen as minimising the importance of practice—I think practice is a miracle. For example, I think I could take any human being who has functioning legs and lungs and train them to run a marathon in six months. I don’t think I could train them to run two hours and 20 minutes in the marathon in six months, but I think I could train them to finish. The more I delved into this area of scientific research, the more I realised that there were two extreme ends of the debate: one side that said genes don’t matter at all—only practice matters. And another extreme that said both genes and practice matters. And only one of those sides, to me, is truly extreme, based on the findings.
This is why the notion of “trainability” is so important for you isn’t it? Because, as you say at one point, it seems to “muddle” the notion of innate talent as something that appears strictly prior to training.
That’s right. If you look back at some of the old sports psychology papers where they define talent, it’s usually defined as something like prowess or skill that precedes the opportunity to train. So it just has to be there, before training. But exercise genetics is totally blowing that notion up. It’s showing that in many cases the most important kind of talent is your ability to profit from your one hour of training more rapidly than your peer does. So just as medical genetics shows that no two people respond to a drug the same way, exercise genetics finds that no two people respond to training the same way.
That clearly has quite profound implications for coaching doesn’t it? To what extent is the theory of coaching catching up with the science in this respect?
It varies on a sport-by-sport basis. In the book I talk about the data scientist Jesper Andersen, who does biopsies of elite athletes. He tells them “Look you’re doing the wrong kind of training for your muscle fibre composition.” He’s had great success with that, sometimes telling guys to train less. I think Usain Bolt has figured this out. Here’s a guy whose achievements are a testament to how much time he takes off of training. For a guy with that much fast-twitch muscle fibre, I think the danger is training those away. And I think he’s really figured out the right balance for himself.
David Epstein’s “The Sports Gene: What Makes the Perfect Athlete” is published by Yellow Jersey Press (£16.99)