The book

The author

  • David Shenk is the national bestselling author of five previous books, including The Forgetting ("remarkable" - Los Angeles Times), Data Smog ("indispensable" - New York Times), and The Immortal Game ("superb" - Wall Street Journal). He is a correspondent for, and has contributed to National Geographic, Slate, The New York Times, Gourmet, Harper's, The New Yorker, NPR, and PBS.

    More info here.

    Contact David.

    Follow on twitter.

    Speaking inquiries here.


« Untapped potential | Main | "Gifted and Talented" School Programs, Part 1 »

January 17, 2007



This talk of paradigm shifts reminds me of my favorite poem: Ode by Arthur O'Shaughnessy

David Storrs

The part that especially interested me about this was the assertion that it would take 10,000 hours of practice over the course of a decade or so to achieve mastery. Now, I'm sure those are round numbers, but they have intriguing implications:

1) How granular can the time be? I suspect that doing 1 hour a day for 10,000 days is less effective than doing 10 hours a day for 1,000 days.

2) What is the effect of distribution? If I practice X hours a day and I miss Y days of practice, have I set myself back by XY hours, or by some multiple of that?

3) What is the interference / reinforcement effect of different types of practice? e.g., if one practices music for 6 hours every day, swimming for 6 hours every day, and chess for 6 hours every day, is it possible to achieve mastery of all three in approximately the same time as if one were practicing only one of them for 6 hours per day?

4) Once mastery is achieved, how persistent is it? Human lifespans are increasing; it's something that we as a species devote an enormous fraction of our resources to, so they will probably continue to increase. At some point, lifespans of 500 years or more will be normal. Will it then be normal to find people who are master-level musicians, athletes, chess players, and so on? Or will there come a point where it is simply impossible to maintain all the skills and the older ones fade away as they are used less? Does this vary by skill? If so, why?


David - I didn't mean "fishy" in a sense that the quote or the research was invalid or compromised, but rather that it didn't add up. And I read the link, and it basically says exactly what you quoted.

" High-speed cameras showed that Beckham accelerated the ball to 80 miles per hour, after hitting it about 8 centimeters to the right of its center with the instep of his right foot. The ball spun counterclockwise at about eight revolutions per second and started swerving to the left. The ball rose into the air as if it would soar over the goal's crossbar. Then, it slowed to 40 mph, curved further to the left, and dropped into the top left corner for the goal."

"He discovered that airflow changes when the ball travels below a certain speed. At fast speeds, it experiences turbulent airflow. When it drops below about 23 mph, however, the airflow becomes laminar.

Carré said that when airflow around Beckham's free kick changed from turbulent to laminar, the drag on the ball increased 150 percent in about one second. That caused the ball to slow down suddenly and to drop into the net.

This change from turbulent to laminar airflow, produces the sudden dips of the best free kicks as the ball approaches the goal, Carré said."

So I'm still trying to figure out the slowdown to "laminar" but anyway, I don't think it's all as hard as they make it out to be. I've seen my 15 y.o. daughter bend it pretty good, but it's the drop that is difficult to get right.

David Shenk

I can't defend the research because I haven't seen it directly, but my gut says that this research is basically sound and we shouldn't leap to "fishy." Remember that I'm summarizing an article which summarized two pieces of scientific of research. They left out most of the detail and had to simplify everything for general consumption. If you ignore the 40mph detail, it not only adds up but makes intuitive sense.


There's something fishy about the Beckham research. First, he "accelerates the ball to 80 mph." That seems awfully high - most world-class players can only get it into the 60's, so I'm not sure that 80 is possible, but that's not my main beef. It says that it was going 40 mph when it went into the goal, but research shows that at 23 mph the sudden slowing caused the ball to take a "severe leftward movement." That clearly doesn't add up - it can't be going faster going into the goal than it had when it took that severe leftward movement. Nobody, not even Beckham, can hit a ball that speeds up, slows down, and then speeds up again.


Okay, two things - first, I am someday hoping to write a book on how insight works called 'pop' psychology so, hmmm, sounds like I need to write it soon. It has to do with self-organizing ideas. Second, about the content of this last post - wait, second is actually I lOVE the Tom Wolfe story though there is another explanation for your experience of Tom Wolfe that has to do with the back story of 'Radical Chic' which was that Bernstein almost took him to court for bringing a tape recorder to the party - turns out he hadn't, he has a 'photographic' aural memory. That's what I read a decade ago or so. Anyway, my response to you is this: I think you are misunderstanding the Gould point and the nature of genius.

Genius is a relative term, a historically bound term, it is not timeless. This is what Kuhn's work is all about. At time X you have a paradigm. Everyone and their brother believes it. And then one person sort of 'sees through' the illusion and POPs into a new paradigm. Cut to 50 years later, time Y. Everyone and their brother believes the new paradigm. This has happened around 5 times in my medical lifetime so far - prions, neurogenesis, estrogen replacement being deadly, what have you. Well, it is NOT the everyone and their brothers who are geniuses, even though at time Y they are 'geniuses' compared to the people in time X. The genius is the guy who learned to think Y during the time of X, and whose ideas help everyone else see through their ignorance and understand that they haven't been understanding. The genius brings the future into the present while everyone else is just working out the details of the present in the present.

The assymptote issue pertains not to individuals but to the 'task' at which they are completing. The assymptote is essentialy a constraint on progress imposed by nature. For example, when Fermat's last theorem is solved it is, well, solved. Nothing more to do. Now as a population approaches an assymptote on any task, it takes only very tiny differences between individuals to 'cross' the line and 'win' but that winning is essentially arbitrary and if you re-ran the contest, someone else would win. That's the nature of variation or noise or what have you. But when a population is really, really far away from an assymptote - eg, when nobody even aspired to run a mile in less than 5 minutes, say, or Fermat had just come up with the problem - well THEN variation really has a chance to show real differences. Someone running a 4 minute mile when everyone else is sitting at 5 minutes, and having to wait 15, 20 years for the next human being to break his record; someone solving Fermat's last theorem on day 1; little Gaus busting a move in 1st grade that nobody in his entire city could - that variation is so real - the signal from that individual is so strong - that if you re-ran the experiment it would happen the same way again.

What this tells us is that it is much more likely that the google guys were geniuses, that the beatles were geniuses, that einstein was a genius, than all the thousands who came after them - because they were the ones who shifted the paradigm and that was probably non-arbitrary.

It's your question 2 that is the genius question; question 1 is a herd question.

David Shenk


Onto your Gould/asymptote-of-greatness observation. Very interesting and it seems to obviously contain a lot of truth. It would tend to apply more to athletic/physical endeavors, right? You couldn't say the same thing about painting, scientific brilliance, literary brilliance?

There are two things that come out of this idea for me. One is that it really endorses the notion that greatness is not some innate quality in certain people but an external quality that people can attain and can be attained by masses of people once the path becomes reasonable well-understood and accessible.

The second thought is that this is really critical in dealing with the Michael Jordan/Roger Federer/Tiger Woods question. The question is how did Michael Jordan become Michael Jordan? And if you look at it through the asymptote prism, you see that that is two very distinct questions that need to be addressed together:
- 1. How did Michael Jordan and so many of his colleagues get to be so amazingly great at basketball?
- 2. What makes MJ stand out and how did he attain (or what gave him) that stand-outness?

The third thought is: I also have to wonder if your POP idea isn't a little but in contridiction with the asymptote-of-greatness idea. The POP idea suggests that certain people have huge qualitative leaps past other people, while the asymptote idea suggests that ambitious people in any given class is moving together towards greatness. Is Jordan really POPping above the rest, or is he just moving a little bit ahead in the asymptote?

David Shenk

Add another word to the words-I-will-not-use list: continua.

"POP" I will use, because it's fun. I'm open to POP and discontinuity, but what are we talking about in practical terms -- according to your view of genius, does a genius POP once or many times into new ways of thinking? Do the experts in the lesser class experience any POPs (one or many) to get where they are, and are many of them just one big POP away from being what you consider a genius?

Is a POP basically the same thing as an epiphany? In my own life, I'm pretty convinced my thinking has been enormously influenced by epiphanies that I have every couple of years. I'm not saying these are brilliant thoughts -- but they are profound to me and hugely shape the way I think going forward. For example, in college I once got the opportunity to interview Tom Wolfe, who at the time was my great hero and writing inspiration. He was going to be visiting our campus for a speech and I was writing a preview piece. So we set up a time to talk on the phone. It turned out that I was on a road trip when the appointed time came and I ended up having the most important phone conversation in my life to date from a pay phone in a Connecticut office park -- with, of course, the mandatory burly UPS guy waiting to use the same phone and getting increasingly annoyed and threatening increasing harm to my person.

So this guy is threatening to kick my ass while I'm talking to Tom Wolfe on the phone and I had this list of questions that I'd prepared about how he had approached this scene in Electric Kool-Aid Acid Test or that scene in The Pump House Gang, and my mind is going back and forth between trying to elicit the most interesting response for the interview and trying to glean whatever I could about how Tom Wolfe became Tom Wolfe and how I could become the next Tom Wolfe, and I'm also trying to come up with just the right few words to say to the UPS guy so he doesn't shatter the phone booth glass all over my head and drag me out through the shards. And I'm also hoping that somehow Tom Wolfe is not picking up on the fact that I'm frightened and may be in considerable danger.

So there's a certain surreal quality to this conversation to begin with. But what really made it strange was that, to the extent that I could concentrate on what Tom Wolfe was saying, he was telling me stories I already knew. He was retelling stories from his books. And he wasn't retelling them in a particularly new or different way -- it was very much like hearing an audio version of this or that book chapter. I would ask him a question about some character, and it was like putting the needle on the right place in the LP and there was that story I already knew. (Because I'd read it all pretty carefully by this point.)

I was like, "What the ___? Why is he retelling me stories I already know? And in just the same way he wrote them?"

One instant later:

"OH! I get it. Holy shit -- I get it. He's telling me these stories in precisely this way because it's ALL HE HAS. All this guy has done for his entire adult life is to go spend a few days or weeks or months observing this character or that character and then spend more months or years trying to craft what he's seen into a good story. He works his ass off trying to get these stories just right, and that's his books. He puts everything he has into making every inflection of every sentence just right. OF COURSE he's not going to suddenly tell some hilarious story from behind the scenes of his published life. There is no behind the scenes."

Well, that was huge for me, because up to that time I thought of Tom Wolfe Hero God Writer as being a Magic Man who allowed stories to roll off his fingers. Intellectually, I knew he worked hard but I never really got that he worked hard writing. And in my own writing, I imagined that if I got into the right Tom Wolfeian spirit of things, perhaps I could get brilliant stories to roll off my fingers too.

Now, quite suddenly -- POP -- I understood that I was going to have to work my ass off too. Write and rewrite and write and rewrite and spend my whole life trying to get it just the way I wanted it. Only then could I hope to produce stuff on the level I aspired to.


Wait - I AM suggesting there are three different types of brains (probably a lot more) but I am NOT suggesting they exist on a continuum.

The debate over whether personality types exist on a continuum or as discrete categories is an old one. I personally suspect that while for many measures there is a continuum that runs smoothly from the laggard to the pacesetter (height, weight, IQ, ability to hold a tune, whatever) I also have a more complex view about what happens when various measures combine/interact. I think - I told you this over the phone back in November - that along the self-organizing criticality way of thinking, it is probably true that when things combine in a certain way they 'pop' into a new conformation that is DIScontinuous with previous conformation. This is a major theory of how life began on earth, for example, or rather hypothesis: that when all these RNA precursors were swimming around you had nonlife, and then all of a sudden you get one tiny change or addition and POP you had a self-organizing system. Totally discontinuous with the past. A 'miracle'.

I believe that you will find, if you interview expert groups, people who exist at a very high level, you will hear two themes. First, you will hear that most of them regard most of the others as being essentially in the same class, abilitywise. But if they were honest, you would find that one or two, usually not more, would be repeatedly identified as having capacities/abilities that seemed somehow to exist at a 'POP'above the rest. YOu have heard this a million times so I won't say it again. What I will say is I think the assumption that just because continua exist for individual traits does not mean those traits, en bloc, also create continua.

I also want to say one important Steven Jay Gould point about expertise (though he talked about it mostly in terms of baseball and evolution): room to grow disappears on future generations. When a field is young, eg the violin or whatever, there is a lot of unexplored terrain. That's where 'genius' has the most opportunity to announce itself - because one person can go leaps and bounds beyond his or her peers. When a field has been worked on and worked on by a million people (Eg the 4 minute mile) ability begins to assymptotically (sp?) approach a 'natural' limit, and, simultaneously, the whole population begins to approach this limit en masse. Genius and exceptionality begin to disappear because they have no room to show themselves, until a new environment, unexplored, begins to emerge. Isn't that an interesting idea?

David Shenk

You're suggesting that there are three different types of brains (not rigidly, I know, but along a continuum): disabled, normal and exceptional. Disabled brains have a lower and upper limit on what they can do -- say 0-50th percentile. Normal brains also have lower and upper limits -- say 20-80th percentile. For exceptional, it's 50-100th percentile. Environment (nutrition, parenting, peers, education, etc) decides where a brain will land within its potential.

It's a very reasonable assumption, and I think it's how most people think about this -- sorry to make you one of the crowd.

But I think it is wrong. Your model probably works fine when you're talking about general intelligence (g), which is roughly represented by I.Q. scores. As you probably know, I.Q. scores seem highly heritable, aren't affected much by shared family environment, and correlate well with academic success, socioeconomic advancement and social pathology. (Some of this correlation has been challenged, but for the moment I'm happy to accept it.)

But I.Q. does not correlate with greatness -- highly specialized achievement. Doing well on these analytic tests shows that a brain is finely tuned and efficient, but cannot predict how much intensive effort, desire, and mentorship the person is going to go through in order to develop a brain which performs extraordinarily well in some specialized way. I think the data is pretty solid on this point. (see Ann Hulbert's NYT Magazine piece on Lewis Terman and his "termites.") Absolutely everyone who has studied this stuff agrees that adult greatness requires this intensive effort (the disagreement is on whether the effort is sufficient or requires a genetic advantage as well).

Even kids identified as true prodigies rarely go on to be exceptional as adults -- for a variety of reasons, but again mostly having to do with the enormous difference between A) having a natural affinity for mastering the basics of words, numbers, music, spatial reasoning, etc, and B) the intense effort and desire required to transform that basic mastery into creative envelope-pushing expertise.

Are there limits? Yes -- absolute limits for all humans. No violinist cannot play Paganini's Violin Concerto #1, at 1000 times the tempo it was written. No soccer player can bend a ball at three times the curvature of Beckham. No mathematician can ____ (fill in the blank). But so far, the research suggests that any non-disabled brain is capable of achieving dramatic levels of expertise. And if you look at the extraordinary improvement in whole groups of experts over the last century -- violinists, marathon runners, swimmers, divers, figure skaters, gymnasts -- you see that the only common thread is that they have all improved their training regimens. They don't have more evolved brains, but rather are training better and more intensively than their predecessors.


But just because you can describe it as a back and forth doesn't mean it is one. Epistemology NE ontology. At the cellular level one thing is happening, not two. I don't know it is really confusing but this old paradigm of nature nurture really is dissolving I think. so is mind brain. I really do think that.

Listen I had an idea last night: have you been checking whether your models also can explain mental retardation, eg abilities that are as far from the mean on the 'disability' side as the ability side? do you or do you not think that your explanations should work equally well there? Because I think, for example, parents of autistic kids really have the experience of nonstop care and tutoring and emotional training etc sort of 'hits a wall' at some point and they realize, hey, this kid is never going to be 'normal.' No amount of training or what have you is going to move them into the 50th percentile. Why shouldn't there be a "wall" that normal people hit when going for expertise? Why shouldn't there be a "wall" that exceptional people hit when trying not to be so smart or gifted? I am not sure how this fits into your theory because I think my internal working model of your theory is a bit of a straw man compared to the nuanced way in which you are thinking about it. Talk to you soon,, peter

David Shenk


You're probably right about the software/hardware metaphor. I do have a pretty good understanding of gene expression and I'm absolutely planning on getting to that. It was of course a gross exaggeration to say that that genes don't drive us, but I did that consciously, wanting to set the stage for a new understanding of genetics. My whole project is all a matter of simplifying and generalizing so that non-scientists like me can understand this stuff, and I strongly disagree that we can't describe it as being a back and forth. That's what interactivity is. Just because it's constant doesn't mean that it's not a back and forth.

RNA will certainly be in the book (and the blog), but I hope to keep dualism out of it. I want to keep this as plainspoken as I can, while making sure my articulation of the ideas are basically sound.


Do I get points for having the first four posts on your blog? I hope you don't mind me adopting the gadfly position. It is so much fun. You are a great blogger btw. Anyway after my last post I re-read your last sentence and thought I'd put in one more plug for nonduality, and a small technical correction. Nondual plug: genes don't drive us and we don't drive our genes. Genes and environment are in such constant interaction that there really isn't a possibility of dividing an individual into these two components. Genes and environment are co-constructed mutually interactive and produce emergent phenomena. Maybe you can coin the term 'genvironment' or something? Look at Paul Greengard's work. That's what his Nobel was for. And the technical point is I think you should emphasize the difference between a gene and gene expression. DNA is essentially stable over one's lifetime, but whether and when genes get expressed (turned on, turned off) is what is 'determined' by the environment. I would guess that philosophers of biology, or hell, even biologists themselves, might have trouble identifying where genes end - at their DNA? At their RNA expressions? At the proteins they produce? Do you count transcription factors? Do you count steroids and other molecules that interact with these transcription factors? I think that the closer you look at a 'gene' the more it spreads. But to limit things perhaps you are mostly wanting to talk about gene expression, eg proteins produced by genes, and not DNA? Not sure. Along these lines, have you read WVO Quine's 'Two Dogmas of Empiricism', a short dense non-mathematical article that totally rocked my world when I first read it and had a major influence on the philosophy of science? If not, wait, did I have a conversation with you about him - well, he essentially shows that there is no analytic/synthetic dichotomy because you cannot find where perception ends and cognition begins. That's the idiot's version. The idiot being me. No wait!... I'm just untrained. Anyway the fallout has affected a lot of dualisms and I think most scientists feel that nature-nurture debates are essentially going to disappear. We'll just start describing how things unfold without dividing causes into nature and nurture categories. Everything will interpenetrate. My nondualistic thought is done. Or is that impossible?


David - I'm having so much fun reading your blog. I have never read a blog before let alone posted anything on one of them. One comment, based on ignorance of the literature but some knowledge of brain function: the analogy of hardware and software to the brain seems fukakht, however you spell that word. In the brain, the neurons ARE the software. A neuron sits there and gets EPSPs and IPSPs that either increase or decrease its membrane potential, and when the potential crosses a threshold, the neuron fires. EG, continuous data is perceived, and produces an all-or-nothing response. At the neuronal level, there is no 'program', just a biological threshold for firing. The only change in code that can occur is changing the firing threshold. At the network level, you can change which neurons are connected to which other neurons. This is not the same as running new software, and it is not the same as adding more processors to the hardware. The brain is nondual, with its structures embodying their own code. At least that's my top-of-the-head response. But I'd be wary of using that metaphor until you've researched its validity. My guess is it's one to avoid.

The comments to this entry are closed.