The problem with: "Malcolm Gladwell's" "10,000 hour" "rule" 

(or, the K. Anders Ericsson 11,000 hour recommendation)

Malcolm Gladwell is a pop science regurgitator of the TED Talk variety. He is frequently wrong, sometimes shockingly so, like when he ascribes plane crashes to the intricacies of Asian culture (link is a great read, by the way). 

Statistically speaking, you have read one of his books but you probably don’t remember it, having memory-holed it next to Freakonomics. I decided that this style of writing is called “correlationism,” and I refuse to google that term to see if it exists anywhere, because I personally made it up just now. This is a style of writing where a smartypants author tendentiously ties together two subjects, such as [checks Wikipedia]...abortion and crime rates? It brings up interesting ideas with a thin veneer of science, it’s not exactly refutable, and often the data is free. Picking two things out of some free data and inventing a theory is certainly cheaper and easier than conducting an experiment. But as every contrarian knows, correlation does not imply causation.

But as every contrary-contrarian knows, there cannot be causation without correlation. Indeed, events must be related to have a causal relationship. But sometimes, Malcolm, there are coincidences.

A short detour about correlations: in The Handbook of Applied Behavior Analysis (both 1st and 2nd editions – either edition is an excellent read and worth the price; the previous year 2000 version is also good but entirely different; the 1978 Catania version I should review, considering where this sentence is headed), AC Catania writes some truly insightful basic information about reinforcement. Reinforcement must have a causal relationship, not a coincidental one. Catania’s example is that if an abusive parent tries to stop a child from crying by hitting the child, it is likely that crying will temporarily increase; and yet we would not say that crying is reinforced by hitting, and instead would likely (correctly) consider it a respondent behavior. What Catania terms “because of” relationships are part and parcel of an experimental analysis. In the most basic sense, science involves adding and subtracting a stimulus to see whether the presence or absence causes an effect. Taking two sources of data that changed at one point in time is interesting, but it is a weak form of proof. I am suggesting that the authors of Freakonomics run a reversal design experiment with legal access to abortion.

Anders Ericsson was a psychologist who studied human performance. His work was mostly academic and theoretical, stretching back to 1974 (strangely his Wikipedia page is missing many of his publications. Nerds?). Yet mixed in with some standard, indecipherable titles, you get straightforward ones like 1982’s Exceptional memory: Extraordinary feats of memory can be matched or surpassed by people with average memories that have been improved by training. How come this guy didn’t write the TED Talk book? (Foreshadowing)

Malcolm Gladwell’s book Outliers introduces a concept he terms the “10,000 hour rule.” This he mainly attributes to the work of Anders Ericsson, citing Ericsson’s work with violinists. Gladwell does a pretty capable job writing about Ericsson, and he explores a few basic examples of the rule in practice, extrapolating backwards by starting with successful people and speaking to them about their childhood. Strangely, given the successes of Black Americans in sports and music, he picks…hockey and the Beatles (Gladwell is biracial). Gladwell gets a number of things incorrect, or manages to simplify concepts past the point of what the science might support. For example, 10,000 hours as a “rule” is obviously picked because it’s a round number; based on Ericsson’s work one could have easily called it the 11,000 hour rule (and it’s also not really a rule). The “11,000 hour guideline” doesn’t have quite the same ring to it. The fact that “the 10,000 hour rule” has persisted in culture to the point of generification is a testament that Gladwell can turn a phrase.

Other issues are where he takes the uncertainty of science and mashes it into the unequivocal prose of straightforward writing:

You might think that a guy like Anders Ericsson, clearly brilliant but more of an academic, would be well-served by having his ideas chewed-and-spat-out by Gladwell, who is undoubtedly a successful author. But it seems that, after about 8 years of dealing with a regurgitationist using his work to get quite rich, Ericsson was ready to cash in with his “own” (co-written) pop science work: Peak: Secrets from the New Science of Expertise.

Initially I committed to not reading Peak. Why bother when, in point of fact, Ericsson’s “academic” work is…quite accessible. Frankly it’s not hard to see why a person like Gladwell jumped on work like this, from 1993. There is no extreme jargon, or overwhelming statistics, and the graphs are remarkably accessible. In fact, so accessible, one nice round number stands out in particular:

.In The Role of Deliberate Practice in the Acquisition of Expert Performance, he begins with a story of Galton, who in 1869 argued that ability was a pure matter of inheritance. This is now obviously wrong, but Ericsson admits that genetic inheritance plays a part in certain physical characteristics that are beneficial in sports (e.g., height). Next, a multi-million dollar question: is practice enough? No. A few previous researchers have their work turned into interesting anecdotes: Morse code operators could improve their speed – with a monetary bonus; Thorndike found that clerks add more slowly at work than their maximal ability allows. 

So: it’s not just genetics, and not just practice. In fact, it’s deliberate practice, a phrase invented by Ericsson. Deliberate practice includes: 1) a desire to improve (motivation); 2) an understanding of the task; 3) immediate feedback; 4) repeated performance. 

This is essentially the formula for a best-selling science-adjacent book: some anecdotes and interviews, an exhortation that most skills can be mastered, perhaps a personal anecdote or two, and here are the FOUR RULES FOR MASTERY. The title just wrote itself. And to be clear, I’m not accusing Gladwell of plagiarism. In fact, after completing the previous summary of Ericsson’s 1993 work above, I found this published in Slate, which is a remarkably similar summary to what I wrote. I didn’t plagiarize, Gladwell didn’t either, and neither did Slate. We just summarized something that was concisely well-written in the first place.

In any case, what if what Gladwell did was serve as, say, a marketer for a scientific precept? After all, exactly 15 years after Ericsson published his 1993 work, exactly 6 years after publishing a similar book chapter, and quite frankly an entire career of excellent work on the topic, it seemed that most people hadn’t heard of Ericsson or his ideas.

 That’s not some kind of insult to Ericsson; that’s often the nature of being an academic – unless you’re Brene Brown. But Gladwell’s book was quite a success, and “the 10,000 hour rule” quickly entered the public lexicon. 

From my perspective, Gladwell did a (mild) disservice to Ericsson’s ideas. Ericsson himself thought so, and the internet is full of smug people on either side, “debunking” Ericsson or Gladwell.

Debunking Gladwell makes sense, because, well, that’s easy. Debunking Ericsson though, that’s a confusing one. As best I can tell, Slate (and other sites, blogs, etc.) make the case that lots of practice won’t turn you into LeBron. It’s a curious framing device, because Ericsson himself says as much (it also discounts the absurd amount of deliberate practice that LeBron puts into his own performance). The more political side of things seems to leak into the discussion, and that seems to result in two overgeneralizations: either you can achieve anything at all with some hard work, or hard work alone is not enough. These are two reflections on the American dream: in one corner, bootstraps; on the other, privilege. Neither has any relation to Ericsson’s work. If anything, he’s quite specific about the type and amount of practice, the advantages and resources that are necessary, and what it can do for performance. 

Some key points of Ericsson’s work:
1.) 10,000 hours was an (approximate) average for some of the best violinists by the time they were 20

2.) Deliberate practice is designed to be challenging (e.g., you can’t do a lot of shitty practice)
3.) Deliberate practice is very hard, and the best performers can do a maximum of 4-5 hours per day in some fields. This is not true of every skill; some skills cannot be deliberately practiced for 5 hours per day (e.g., Olympic sprints)

4.) To attain mastery in various domains, a person must have access (in most cases) to skilled coaching – this doesn’t always mean wealth, but it does imply that some people have an advantage somehow (e.g., luck, connections, etc.)

Ultimately, scientists infrequently drive the cultural adoption of practices. More commonly that is taken up by someone with something to gain – partisans, or profiteers. There is an ongoing debate in the sciences about whether we are well-served letting scientists handle the messaging. The general public tends to feel that behaviorism is yucky. The general public also tends to feel that life-saving vaccines are yucky. OK, so that’s a little glib and uncharitable, and perhaps why we don’t want scientists disseminating their ideas. But non-experts do sometimes get it wrong, and the appealing stories they tell can communicate the wrong ideas. 

If there’s a middle ground here, it probably makes sense to look at sources – not all the time (you don’t have time for that, you have adult softball league), but for those things that appeal to you. If you read Outliers, or Blink, or Thinking: Fast and Slow, take a look. You might be surprised at what you find.

Postscript: Peak

I mentioned in a previous passage that I found Ericsson’s work so instantly accessible, so readable, that I committed myself to not reading his mass-market manuscript, Peak. Later I felt guilty about that. The book is only about 150 pages long, and I am trying to make a point of researching topics carefully. I should go the extra mile, I think.

As it turns out, I was more or less correct in the first place. Peak is a good read – in the same way that Ericsson’s scholarly literature is a good read. It’s accessible, free from jargon, concise, and illustrated with real-world examples. I’m sure it sold 8 or 9 copies.

The benefit of a mass-market book is that Ericsson can pull in different strands of research; for example, there is evidence in London cab drivers (who have to pass a notoriously difficult navigation test) that parts of their brain grow – and that after retirement, that area shrinks slightly. Frequently challenging the brain or body leads to physical adaptations and eventually what might be termed fluency – performance that is both correct and fast. Everyone knows what physical changes look like, but learning about structural changes to the brain is an interesting and illuminating addition.

Additionally, Ericsson can provide details and background that get left out of his scholarly works. For example, he thinks studying classical instrumentalists (violinists, pianists, etc.) is important because the techniques are well-established and much of the teaching has been refined over the centuries. He never came across a true “prodigy” who never worked hard – every top performer put in more practice than their contemporaries. And people who don’t put in the deliberate practice – well,

What could be done about professions without natural sources of feedback? Ericsson details a solution for radiologists, who in most cases judge x-ray images, but may never receive feedback on their diagnoses. And in fact, much of their training is simply based on what older radiologists have done in the past, whether or not any of it is actually effective (this is a common issue in training – teachers, coaches, BCBAs…). His idea is a training library of images, so that radiologists could practice and receive feedback on cases where the outcome is known. It’s a simple and brilliant idea.

And what about the “10,000 hour rule”? Ericsson himself writes a short sub-chapter covering everything I’ve already written about it. But furthermore, Ericsson details advice he gave to a 74-year-old man trying to become a black belt in karate:

There’s a certain inspirational quality of Peak that reminds me of the behavior analysis training boom of the 1960s – the era of Bob Mager, Aubrey Daniels, Zig Engelmann, and Thomas Gilbert. Ericsson covers times when new training methods led to a shattering of a previous record, and he repeats an idea: nobody knows the limits. What is the fastest 1-mile time possible? How many digits can a human memorize? At points in history, we believe there has been a permanent plateau, only to be shocked as the record is shattered – and later to accept that even high school kids can match the previous plateau.

Towards the end of the book, Ericsson shares how deliberate practice impacted a college-level physics class that was previously receiving standard lectures. The main intervention appeared to be active student responding – something incorporated into Direct Instruction since the 1960s. The promise of that time was that in changing the methods we could change the world. None of us can know the limits, but we have an idea of how to reach them. It’s just a matter of doing it.