A guarded defense of Malcolm Gladwell

Jonah Lehrer now has two posts slamming Malcolm Gladwell's Blink. The second post, currently ScienceBlogs' most emailed story, offers the ultimate slam, proclaiming that Gladwell is the "new Freud," a mere "prose stylist" who "wasn't particularly interested in the neurological foundations of his theories."

As I've said before, I agree with much of what Jonah is saying, especially when it comes to Gladwell's over-reliance on anecdote and over-generalizations about experimental results. And I'm certainly impressed, as Jonah is, with Gladwell's insightful analysis of current events and politics, both on his blog and in his New Yorker articles.

But Jonah's key criticism of Blink doesn't ring home for me. Here's how he describes what he sees as Gladwell's key failing.

Gladwell ended up lumping together all sorts of research, from Damasio's Iowa Gambling Task to Ekman's cartography of facial muscles to brain scans of autistic people, that, at least from a neurological perspective, were totally unrelated.

Jonah's problem with Gladwell's method is that Gladwell doesn't parse the data the way Jonah wants him to. Jonah would like to see Gladwell explain all the data he discusses in the context of showing how the mind works. But that's not what Gladwell's doing in Blink: Gladwell's goal is to show how we respond to a particular type of situation, regardless of whether the same neurological mechanisms are involved.

Jonah and his commenters complain about the overuse of anecdotes, and I find them frustrating, too, since like Jonah I'm primarily interested in the science. But there's also no doubt that the anecdotes make the book much more interesting for the lay reader -- and even for me on occasion. For example, the tragic shooting of Amadou Diallo provides an excellent example of how snap judgments can fail us at critical moments.

Jonah and commenter Katherine would point to this anecdote as an example of Gladwell contradicting himself: at the start of the book, Gladwell's arguing that art dealers make better decisions when they make them quickly, and here he's claiming that police officers are better off taking more time before they act.

But is this really a contradiction? In police work, decisions are often made in fractions of seconds. Wouldn't that be rather a different time frame than even a few minutes taken to decide on a $10 million art purchase? Gladwell's overall point isn't simply "trust your instinct," but "manage your instincts," based on a knowledge of the cognitive processes behind them. That said, I think in many cases Gladwell goes too far. Just because viewing a 30-second snippet of teaching correlates with end-of-semester teaching evaluations doesn't mean we should shift to evaluating teachers with 30-second snippets -- it may mean that we need to reassess the entire process of teacher evaluation.

I do object to the underlying thread of Jonah's argument. Not every book on psychology needs to get at the neurological basis for behavior. Like journalists, neuroscientists (and cognitive scientists) are often guilty of overstating their case. On the whole, however, and especially compared to the vast array of dreck out there masquerading as psychology, I think Blink is a worthwhile book for lay readers.