Walking Back the Cat

The culture of explanation


Upheaval Dome is a craterlike structure in the sere desolation of Utah's Canyonlands National Park. It is a mile wide and 1,500 feet deep, and scientists are still not entirely sure how it came to be. One theory implicates the erosion of a vast underlying salt dome. Another theory invokes an otherworldly collision. Upheaval Dome certainly looks like an impact crater. A sign at its edge offers this scenario: "Some 60 million years ago, a huge meteor streaking through space ... pierced the earth's atmosphere directly above this point." The text goes on, "The meteor hit with little warning."

With little warning? Was the Hubble on the fritz? What about ground-based tracking stations? Where was CNN?

The Park Service's writers, of course, meant nothing more than "suddenly," and the wording they chose came to mind naturally. We inhabit a culture of warning, and have come to expect advance notice of nearly everything. This represents a big psychological shift. Our forebears woke up not knowing if a hurricane was about to strike. They did not anticipate fourth-quarter downturns. The threat posed by arterial plaque never entered their minds.

Today medical-imaging outlets, at shopping malls and elsewhere, offer full-body radiological scans ("the best way to show you care for those you love"); the results, preserved on compact discs, display lifelike color reconstructions of the subjects' bones and organs, with highlighted warning signs of tumors and heart disease. By regulation, restaurant menus in my state now display asterisks alongside items such as clams and sushi and soft-boiled eggs, directing customers to a cautionary footnote: "The Department of Public Health advises that eating raw or undercooked meats, poultry, seafood, shellfish, or eggs may increase your risk of food-borne illness." Last summer the Food and Drug Administration gave its approval to the M2A Swallowable Imaging Capsule, a tiny early-warning spy satellite for the small intestine, which snaps pictures twice a second during the course of its journey. Restaurants may soon provide a bowl of these for departing patrons, next to the mints.

The forward-looking effort to spot trouble on the horizon has a counterpart in the backward-looking effort to figure out what went wrong after trouble has in fact arrived. Intelligence analysts, who often have to second-guess themselves after defections and other failures, have a phrase to describe this kind of diagnostic deconstruction: "walking back the cat." A lot of walking back the cat is being done right now, in the aftermath of the terrorist attacks in New York and Washington. But the basic procedure extends into every aspect of our lives. I won't hazard an estimate as to what percentage of GNP is allocated to walking back the cat; it clearly consumes a growing share. The study of history and cosmology is devoted to it, as is the great bulk of legal practice, much of medicine and market research, about half of geology, theology, and literature, all of psychoanalysis. Walking back the cat is what congressional investigations do. It informs the case-study approach to business-school education and lies behind the idea of the "financial autopsy." It's the whole point of watching videos of yourself at golf or tennis clinics.

An intensifying focus on "What went wrong?" is nowhere more apparent than in the proliferation of "black-box" technologies. The black box that everyone knows about is the flight-data recorder on commercial aircraft, which preserves a second-by-second record of mechanical and electronic performance. One much publicized new application is in stock cars, a consequence of the death of Dale Earnhardt during the Daytona 500 last February. Black boxes are also being installed in some passenger cars, to preserve data on speed, gas-pedal position, brake activity, force of collision, air-bag deployment, and seat-belt use. An experimental black box tested by Toyota goes even further—it has sensors capable of discerning elements of a driver's physical state, such as sobriety and cardiac condition. Rental-car companies are already installing devices that could determine if customers break company rules (for instance, by leaving the country) or drive too fast. Acme Rent-a-Car, in New Haven, fines customers $150 for every instance in which the car's speed exceeds 80 miles an hour for two minutes or longer.

Black boxes are standard equipment on trains and buses. Proposals have been advanced to put them on roller coasters and fishing boats. Although most people don't give it a moment's thought, the "History" folder in a personal computer's Web browser effectively serves as a flight-data recorder for Internet surfing. An item in the satirical newspaper The Onion once presented an interview with a psychologist who specializes in retrieving thought processes from Web browsing.

"Let's look at Christine's time online last night," said Kimble, pointing to an opened history folder on the computer screen. "She starts at the Godiva gourmet chocolates online gift catalog at 7:35 p.m. Then it's on to eDiets.com at 7:43 p.m. and the Nordic Track web site at 7:45. I'm guessing she clicked on a banner ad at eDiets to get there. Then it's on to 'Liposuction FAQs' at 7:52. Next is a page titled 'Sexy Swimwear Sale: Dare To Be Bare.' Then, at 8:06 p.m., it's back to the Godiva chocolates secure-server online order form."

Arguing that the operating room is "the cockpit of the hospital," a professor at London's Imperial College, Dr. Ara Darzi, recently proposed the installation of a black-box system for surgical procedures. Under this scheme a surveillance system would monitor traffic in and out of the operating room and record conversations; surgeons would also wear sensors on their hands to keep a record of every movement. A helpful next step, for the benefit of medical students and other observers, would be to ask the sportscaster John Madden to analyze each stage of the procedure with scrawled diagrams on the replay screen. ("Harbaugh's deep inside, trying for a triple bypass, but he's brought up short. That's an outstanding blockage.")

Nature has built any number of black boxes, which we've barely begun learning to use. Tree rings hold an implicit record of wet years and dry years. Ice-core samples capture a history of pollution and climate. Someday we'll learn how to open the hard drive of stored memory in the human brain. The genome is the most portentous black box of all. Mitochondrial DNA contains a record of genetic mutation going back, in effect, to Eve. Every individual's DNA harbors answers to questions that we haven't even thought to ask, though they all boil down to "What is to blame for yourself?"

Is there a limit to the number of black boxes we're prepared to tolerate? I raise the matter while conceding that many useful applications remain untried. I wish that a black box could be attached to certain waiters and restaurant orders. I wish there was one that could determine when and where I picked up colds, and one that could explain why some people I know always get lost. And I'd like to have a black box that could pinpoint the underlying issue in arguments with family members. ("Oh, so that's what this is about," you would realize after viewing the playback.)

But just as a culture of warning can go too far, so can a culture of explanation. On this question there are two basic camps—those who think that everything is potentially knowable and those who think otherwise. People in the first camp see each advance in knowledge as reducing the total stock of the unknown. They optimistically use the word "yet," as in "Alas, we don't know yet." People in the second camp see every advance in knowledge, counterintuitively, as somehow expanding the realm of the unknown. They believe, further, that there are times when it's helpful not to get to the bottom of something—when the best course is to put an episode under seal and get on with life. They regard an imperfect memory as an adaptive trait, enhancing long-term survival.

I usually find myself in the second camp—the "let sleeping dogs lie" camp. Walking back the cat has its practical benefits, and often it's essential. But there's something unsettling about the process. For one thing, the walk backward can theoretically go on forever. A reductionist pursuit of explanation quickly leads from proximate causes to antecedent factors and then to the tiniest capillaries of contingency. It leads to "recovered memory" and the "Twinkie defense." And, in general, it encourages an unhealthy confidence in human analytical powers.

Yes, from time to time we must rouse the sleeping dog. But when we walk back the cat, let's also remember what killed it.