In praise of Malcolm Gladwell

While Malcolm Gladwell bashing season is still in full swing and before the mob burns the Gladwell effigy, I want to record a few thoughts that I feel are under-appreciated (or under-emphasised) by some of Gladwell’s critics. (For those not up to speed, Chabris opens, Gladwell responds, Chabris has another go, Gelman comments, twitter goes nuts, and if you google, you’ll find plenty of other contributions).

First, I am not a fan of, for want of a better way of describing it, the conformist strain that underlies some of the critiques. “There’s our way or the highway.” There is a role in science for speculation, for advocacy, for taking a position and seeing how far you can push it (even if you don’t believe it). Gladwell’s stories make you think about issues in ways you haven’t before. It’s through confronting and dismantling those stories that you get to understand the question better. Of those who have read Outliers, who hasn’t thought more about the mix of training and talent behind elite performers. Daniel Kahneman’s Thinking, Fast and Slow was a richer read (for me) having previously read Blink and thought about the two systems that Kahneman dissects (in fact, Blink triggered a path of reading that continues to this day). I largely agree with the “stone soup” section of Gelman’s post, which captures this point better than I do here.

Having said this, there are times where I would prefer more evidence to underlie Gladwell’s claims. In fact, I often wish that about whatever I read. I recently wished the same thing about Dan Ariely’s The Upside of Irrationality. But Gladwell does draw on a lot of peer-reviewed science (just not always the papers preferred by his critics). The 10,000 hour rule did not come from nowhere. Of course, the flip-side of this demand is a world there every word is vetted, every claim pinned to a peer-reviewed paper, and no-one wants to read.

I suspect part of my discomfort with the critiques of Gladwell is that so much of what is published as peer-reviewed science won’t stand the test of time (particularly in economics, psychology and other human focused fields). It won’t (or can’t) be replicated. It relies on tortured statistics. Those who don’t want to believe it won’t, or they’ll stick to their preferred piece of the literature while ignoring the rest. And among that mess, we pick on Gladwell for being too neat or not logical enough or making general claims without evidence (and why aren’t more of the criticisms of Gladwell being directed at the authors of the studies he is referring to and not Gladwell himself).

In that context, another part of my discomfort comes from the expectation that despite my “superior logic”, Gladwell might well be right on some points where I (and others) disagree with him. A small degree of humility does not go astray.

Gladwell is also not as neat in his story telling as he is often accused. You can see why he laments that people only read the first half of Blink, when the latter chapters are filled with examples of where snap judgments fail. If anything, parts of Blink seem inconsistent and by the end of the book, un-reconciled. His recent story on whether drugs may be a fair way to overcome genetic differences in sporting ability is a great example of his ability to raise a murky question and let it linger in the air (in my opinion, Gladwell is a better essayist than book author).

Finally, although Gladwell doesn’t seem keen to engage in a back and forth debate, when he does, he doesn’t do too badly. Or, as I’d rather put it, his opponents don’t deliver the slam dunks you might expect. Andrew Gelman suggested that “Gladwell’s credibility has been weakened over the years by fights with bigshots such as Steven Pinker”. But if you read the exchange between Pinker and Gladwell, Pinker’s major sources were blogs and petitions. Maybe the bloggers and petitioners were right, as Pinker suggested (I think they were). But it’s hardly the dismemberment that someone of Pinker’s stature might be expected to deliver. Of course, over the next couple of years, some of those critiques strengthened. Thanks to Malcolm Gladwell, people thought about it more and nailed their arguments down.

Having said the above, I disagree with a lot of Gladwell’s arguments, including many of the central themes of Outliers. I am with Steven Pinker on the lonely ice floe of I.Q. fundamentalism (although threads started by Gladwell played a small part in convincing me that traits such as time preference and conscientiousness are also important). But I’d like to hear less criticism of Malcolm Gladwell as a package and more of “That’s interesting, but I think it’s wrong. Here’s why.” Because its when people grapple with Gladwell’s ideas that his true value has been realised. Those areas touched by Gladwell are better for it.

*If you want more Malcolm Gladwell love, try this piece by Ian Leslie.

3 comments

  1. A few thoughts in Gladwell’s defence.

    1) I do mostly finish his books.

    2) Even though they are not perfectly robust in their arguments, I think I put each book down a bit smarter than I was when I started it. That’s something.

    3) An example of a perfectly robust, academic-strength book would be Kahneman’s Thinking Fast and Slow. But it took him something like ten years to write it to that standard. Now would we all be better off had he written three slightly sloppier books in that time, and published them earlier? Not sure.

    4) Stories are a very effective form of data compression/storage – and most of us remember them long after charts are forgotten.

    5) You’re right. His real forte is as an essayist. But all sorts of financial incentives – not least the speaker circuit – require you to write a book – to issue albums, not just singles.

    1. On 3 and 4, the parts of Kahneman’s book that I keep quoting to others are the stories he uses to illustrate his points.

      Also, I tend to think that the optimal scenario is a mix of varying robustness across authors, mediums and time. Speculation leads to research, stronger analysis and ultimately more robust outputs. I’m glad Kahneman took the time to write a solid book and I’m not sure Gladwell should slow down his current rate of production.

      1. There is, as you say, an interesting relationship between theory and data. If you only have the first, you risk confirmation bias. But, in the social sciences, if you purely rely on the latter, your theories may be constrained by what you currently measure, or what data happens to be easy to collect.

        I always love scandalising academics by pointing out that, in many cases, the business world provides a much better template for progress than the academic world. My argument is that, to be rewarded, academics have to be absolutely, incontrovertibly right, and need to explain their reasoning perfectly: in business all you need is to be less stupid than your competitors are – and you don’t even need to know why what you are doing works. In most real-world instances, this stochastic approach works much better.

Comments welcome

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s