Saturday, November 26, 2011

All of Malcolm Gladwell in One Sentence

By a series of the very sort of random accidents that lead experts to make faulty inferences, it happens that I've been listening to and reading a lot of Malcolm Gladwell lately. In search of books on CD for the long drive back and forth to my friend's cabin in North Carolina, I've fetched up at my local library, plying the non-fiction CD section and coming away with Blink, The Tipping Point, and Outliers. As an avid reader of The New Yorker, I've encountered individual essays by Gladwell on subjects ranging from Mammograms to criminal profiling. By virtue of having a mother who loves to read, I've received in the mail a collection of Gladwell's previously published pieces, What the Dog Saw. "What a person trained to analyze such patterns would conclude," we may imagine Mr. Gladwell saying about this, "is that Mr. O'Gorman is manifesting a clear intent to read or listen to all of my work, for some specific purpose or agenda." And then, if we've listened to the author via audiobook, we perhaps imagine him then taking a long, not quite stentorian breath, before adding, "But does it really tell us that? Couldn't the fact that O'Gorman has lately read and listened to a lot of Gladwell bespeak a series of outcomes driven by pure chance?"

This then is Malcolm Gladwell's life's thesis: People in the modern labor market have very specialized, intensely technocratic roles, involving the accumulation, collation, and interpretation of highly complex and noisy data. They are thence so inundated with all this data that they are prone to paradoxically flat-footed conclusions -- specifically, to conclusions that are increasingly disconnected from the messy reality they endeavor to explain. In The Tipping Point, marketing experts invest countless fortunes hoping to foreordain explosive bursts in consumer demand for products, but miss the qualitative (and therefore uninteresting-to-them) role of the hipster opinion leaders on the ground. In Blink, a Pentagon with effectively unlimited resources develops a war game -- in which flip charts and payoff matrices figure prominently -- only to be humiliated by an impish, seat-of-the-pants gunslinger who was hired to pretend to be the other side. In What the Dog Saw, oncologists pore over mammograms searching for tell-tale signs of breast cancer, then misdiagnose the patients in their charge because the flood of data overwhelms their intuitions. This is Gladwell's theology, periodically re-dressed to sell more books; this is all of Malcolm Gladwell, in one sentence: Expertise is overrated.

As a professional academician in a technical field, I am obliged to feel a personal investment in pushing back against this thesis; to pretend that this is not personal would be dishonest. But, okay, just how convincing is Mr. Gladwell's case? Is the world as unknowably heuristic as he envisions it to be, or, ironically, does the very argument he endeavors to build in tome after weighty tome suffer from exactly the flaw he keeps noticing in the thinking of all those supposed experts?

Perhaps the best place to start such an inquiry is with Gladwell's essay about the state of mammography. For it is here as in few other places through his bibliography, the author removes the gloves of intellectual coyness and strikes his target-experts a mighty blow from the hammer of his world view:

Calcium deposits, apparently, are the canary-off-the-perch inside a woman's breast. When a mammogram reveals these deposits, the job of interpreting the picture falls to a highly trained practitioner tasked with determining whether the screening process moves on to the next, more invasive level. The trouble, apparently, is that many types of calcium deposits in breasts are benign, others are sometimes benign, others aren't benign but may be so slow to do damage that noticing them would raise a false alarm, and still others are genuinely malignant but have a nasty tendency to hide behind certain types of breast tissue.

Gladwell's conclusion -- bizarrely -- is that the medical field is suffering from a flood of information, and that this flood is overwhelming its capacity to intuitively do its job.

"Would taking a better picture solve the problem? Not really, because the problem is that we don't know for sure what we're seeing, and as pictures become better we have put ourselves in a position where we see more and more things that we don't know how to interpret. ...The picture promises certainty, and it cannot deliver on that promise."

There are several levels on which this conclusion may be considered, and I'd be ignoring my responsibilities to my own convictions and my own profession if I didn't pay at least a passing service to the issue of Gladwell's own lack of requisite training. He speaks to many qualified persons, to be sure -- The New Yorker doesn't run strops, after all -- but all the while his personal education is as a historian.

He has through a combination of imposing raw intellect, abiding curiosity, and virtuoso craftsmanship, established himself as a renowned writer on highly technical subjects -- bringing to each of them a highly technical, razor-edged belief that all of us are drowning in highly technical data -- but none of this completely nullifies the fact that his own educational background excused him from the grim and gritty process of developing the informed intuitions of a technically educated person, with which that technically educated person might just do a slightly better job of sorting the expert opinions he obtains. (A trained statistician, for example, would recognize that richness of data is very often beneficial, and not deleterious, because it serves to rule in the scientific learning process by ruling other, previously accepted theories, out. Noisy data doesn't make it harder to draw conclusions; it narrows the field of available conclusions and thus makes them far more accurate. Take it from a trained statistician.)

But setting aside the issue of qualification, as we must if the matter is to be considered in anything like good faith, what of Gladwell's conclusion itself? Have we got this right? Because the current state of the mammogram is that it is fraught with contradicting signals about calcium deposits, this therefore means that all future improvements in the technology would only make the problem worse? Malcolm, really?

Not to put too fine a point on it, but in trained intellectual discourse, Gladwell has here committed what's referred to as the fallacy of the predetermined outcome. The fact that previous improvements in mammography have only clouded the issue and reduced the rate of successful diagnosis by adding noise to the data (something the author himself interestingly never quite says, by the way), it therefore follows that any subsequent improvements in mammography would have the same result, further exacerbating the current problem. Our intrepid reporter's too-much-information hammer has left him only able to see a clear cut case of not-quite-information, as just another nail: If mammography is resulting in misdiagnosis of breast cancer, it must be because mammography is overwhelming the practitioner with messy data, and that better mammography will only make that problem worse. kuh-WHACK. The good guys lost the Pentagon's expensive war game because they had too much information. kuh-WHACK. The Getty Museum almost bought a counterfeit statue because they had too much information. kuh-WHACK. A company that sells skateboarding apparel confuses the success of Hush Puppies for a general plan for selling hip shoes, by accumulating too much information. kuh-WHACK, kuh-WHACK, kuh-WHACK, kuh-WHACK.

The irony in all of this, perhaps easily overlooked, is that the error Gladwell commits in the structure of these arguments is the very one he implicitly criticizes each time he endeavors to make it. An unabashed expert and unimpeachably brilliant mind, whose knowledge of the world is nonetheless inescapably limited, has difficulty holding perspective on that limitation for long enough to come to the cool-headed conclusion that he might be in over his head. If our author had been alive during the time of Pythagoras, would Gladwell have considered a sun-centered model of the universe necessary to clean up the bizarre tracks of the planets across the sky, or would all the bizarre behaviors of the planets be just another sighing, head-shaking example of all these puffed-up people's insistence on confusing themselves with too much, too technical information?

True, the modern world fetishes information at the expense of common sense. Absolutely it does. And anyone who knows me personally knows that I am every bit as exercised about the problem as Gladwell, at least when it comes to our foolhardy embrace of pointlessly over-improved consumer technologies. In re-acquiring five or six hundred PC's every year or so, for no better reason than because the new PC's are new, the college where I work places itself at the top of the exhibit list in my own, abiding agreement with Gladwell's notion that the modern world often fetishes information at the expense of common sense.

But you know, it also often does the opposite. Increasingly, in fact, it does nothing in-between. Gladwell is right that the CIA should've been more circumspect about the photos it was amassing of the Iraqi countryside in late 2001 and early 2002, but he'd be wrong if he were to argue, as others have, that a snowstorm in May is conclusive evidence that the technocrats are wrong on global climate change, or that the current recession is proof that we should dismantle the technocratic stabilization infrastructures of the United States Federal Reserve. And if this commentary seems a bit over-revved at this point, then perhaps it's time to circle this back to the personal stake that I feel in my heart each time Gladwell over-extends what is otherwise a not entirely invalid point about our information-fetishing world.

It's not just that his thesis so brusquely rules out the possibility that the exact opposite problem might be true in each of his examples -- that mammography might actually get better with an improved picture, and not worse -- it's that in making these observations from a position of limited expertise of his own, Gladwell is contributing to an overall cheapening of the value of expertise. And worse, he does so at precisely the time in world history when that erosion of value is poised to do the greatest damage. As a (supposedly) tenured professor who still stands in front of classes full of students, I'll freely admit that most of the intensity of my reaction to Gladwell is driven by the personal sensation of being one of the world's genuinely endangered species. And being made to feel even more so by the writings of someone as clever and highly educated and intellectually curious as Malcolm Gladwell is hardly doing anything for my peace of mind about the future of intellectual thought in this world -- technical or otherwise.

Dave O'Gorman
("The Key Grip")
Gainesville, Florida


It's jaxun! said...

I just read your review of the Konica Minolta 1600W on the B&H website, posted 2/6/2011, linked below.

I wanted to contact you directly, or at least remark on your review, because...

I was just wondering, did you win with that entry? Win the "The Bulwer-Lytton Fiction Contest", that is. It was such a compelling bit of writing that I was jarred out of my hum drum existence, and in my search for a paragraph, realized that I HAD TO KNOW. I had to know if you actually, consciously INTENDED to pen the single longest, most rambling review I have ever read. Come to think of it, since I wasn't able to get all the way through it (because I was overwhelmed with the compulsion to find you), I am not even sure what your conclusion was about the product.

Was it your attempt to replicate the infamous "Amazon Mountain-Three-Wolf-Shirt-Review" phenom, in a way that only the hippest of printer shoppers would recognize?

If so... bravo! If not... bravo!

Well played, sir. Well played.

The Key Grip said...

I was about to win the award for longest, most rambling review -- but at the last second they gave it to an as-yet-unnamed fellow blogger, in recognition of a comment (s)he left on someone else's blog, about a printer review. The details haven't been released yet but I can give you the number for the public relations office if you're really curious.