Modern life is full of potentially terrifying “What if?” possibilities. What if a pharmacist decided to substitute morphine pills or strychnine for the next prescription you pick up? What if a school-bus driver decided to swing the wheel, and plow a full load of children head-on into incoming traffic, or off an overpass?

What if a FedEx or UPS courier decided to deliver a box full of explosives, or anthrax spores, to an office building, rather than business supplies? What if a disturbed student, teacher, or parent walked into a public school and opened fire on everyone in sight?

The last possibility is a reminder that there are risks some societies will define as acceptable. All the rest illustrate the reality that our lives hang by threads that someone else could decide to cut. The ability to inflict harm, whether intentionally or accidentally, rises more or less in pace with the technological complexity and interdependence of modern life.

Every modern city dweller depends for daily well-being and even survival on systems that make up the hard and soft infrastructure of society: water, power, sanitation, public health, and on down the list of services no one notices until something goes wrong. Most are run by people we don’t know, whose competence and good intentions we have no choice but to take for granted. As for people determined to do harm—the pharmacist who wants to poison customers, the bus driver intent on suicide—the only absolute protection would be surveillance and regimentation on a draconian scale. (Want to avoid the risk that any bus driver, ever, could do something rash? Send them all through full FBI criminal-background checks, plus psychological testing, and then staff every bus with both a driver and a co-driver, each to keep an eye on the other. Any school system could do this. None that I’m aware of does, since it would price bus service out of the realm of practicality.)

Thus sane approaches to security have been careful to set the goal of reducing risks, rather than eliminating them. The first is possible, and it naturally leads to discussions of cost, practicality, and the trade-offs between security and liberty. The second is in most cases impossible, and it naturally invites “security theater”-style posturing in fending off threats, and “How could this have happened???” overreaction when something inevitably goes wrong. (For more of The Atlantic’s case against security theater, especially involving the early years of the TSA, see pieces by Jeffrey Goldberg here, here, and here, and by me here and here. )

So we come to the bizarre, frightening, and tragic episode on Friday night in Seattle, in which a ground-staff baggage employee of a regional airline got into an empty twin-engine turboprop, started it up and took off without permission, flew dramatic aerobatic maneuvers over Puget Sound, and then crashed on an island off Tacoma, killing himself in an apparent suicide.

Bizarre, frightening, and tragic this certainly was. But was it a sign of an alarming failure in security practices, as some press accounts immediately asserted? (For instance, from the United Kingdom’s The Telegraph, soon after the event: “It has raised fundamental questions about airline security at America’s major airports after the mechanic was able to board the plane, taxi onto the runway and take off without being stopped. Aviation experts questioned what the authorities would have been able to do if the pilot was determined to fly the plane into a city rather than do loop-the-loops.”)

Maybe this will be the appropriate response when more facts are known. For the moment, as is usually the case with aviation disasters, many of the most important questions about what happened are impossible to answer right away. Here are some of the aviation details, known and still puzzling, and then my hypothesis as to how this could have happened. (I trained for and got my instrument rating at Boeing Field in Seattle in 1999, and flew frequently in Seattle airspace when I lived there in 1999 and 2000.)

  • The specifics: The most useful overall summary I’ve seen is in The Aviationist. It gives details about the plane (a Horizon Air Bombardier Dash 8, with no passengers aboard but capable of carrying more than 70); the route of flight; the response of air traffic control; and the dispatch of two F-15 fighter jets from the Oregon Air National Guard’s base, in Portland, which broke the sound barrier en route toward Seattle and were prepared if necessary to shoot down the errant plane.
  • The real-time drama: A video of the plane’s barrel rolls and other maneuvers, plus the F-15 interception, from John Waldon of KIRO, is here.The recordings of the pilot’s discussions with air traffic control (ATC) are absolutely riveting. A 10-minute summary, featuring the pilot’s loopy-sounding stream-of-consciousness observations in what were his final moments of life, is here. A 25-minute version, which includes the other business the Seattle controllers were doing at the same time, is here. The pilot makes his final comments at around time 19:00 of this longer version. A few minutes later, you hear the controllers telling other waiting airline pilots that the “ground hold” has been lifted and normal operations have resumed. In between, the controllers have learned that the pilot they were talking to has flown his plane into the ground.
  • How did he do it? Part 1: The Dash 8, which most airline passengers would think of as a “commuter” or even a “puddle jumper” aircraft, differs from familiar Boeing or Airbus longer-haul planes in having a built-in staircase. When the cabin door opens, a set of stairs comes out, and you can walk right onto the plane. This is a very basic difference from larger jets. The big Boeing and Airbus planes require a “Jetway” connection with the terminal, which is the normal way that passengers, flight crew, and maintenance staff get on and off, or an external set of stairs. Also, big jets usually require an external tug to pull or push them away from the Jetway and the terminal, before they can taxi to the runway. They cannot just start up and drive away, as the Dash 8 did. Was the Dash 8’s door already open, and the stairs down, so a ground-staff member could just walk on? Did he have to open the door himself? I don’t know. But either way, anyone who has been to a busy airport knows that it’s normal rather than odd to see ground-crew members getting into planes.
  • How did he do it? Part 2: However the pilot started the plane (switches? spare set of keys?), the available ATC recordings suggest he didn’t fool the Seattle controllers into giving him permission to taxi to the runway or take off. He just started taxiing, rolled onto the runway, accelerated, and left. As you can hear from the 25-minute recording, ATC at big, busy airports is an elaborately choreographed set of permissions—to push back from the gate, to taxi to a specific runway, to move onto the runway, to take off. For safety reasons (avoiding collisions on the runway), in this case the Seattle controllers had to tell normal traffic to freeze in position, as the unknown rogue plane barged through.
  • How did he do it? Part 3: In the 10-minute ATC version, you can hear the pilot asking what different dials mean, saying that he knows about airplanes only from flight simulators, and generally acting surprised about where he finds himself. But the video shows him performing maneuvers that usually require careful training—for instance, leveling off the plane after completing a barrel roll. Was this just blind luck? The equivalent of movie scenes of a child at the wheel of a speeding car, accidentally steering it past danger? Was his simulator training more effective than he thought? Did he have more flying background than he let on? At the moment I’ve seen no explanation of this discrepancy.
  • How everyone else did: I challenge anyone to listen to the ATC tapes, either the condensed or (especially) the extended version, and not come away impressed by the calm, humane, sophisticated, utterly unflappable competence of the men and women who talked with the pilot while handling this emergency. My wife, Deb, has written often about the respect she’s gained for controllers by talking with them in our travels over the years. These are public employees, faced with a wholly unprecedented life-and-death challenge, and comporting themselves in a way that does enormous credit to them as individuals and to the system in which they work. In addition to talking to the hijacker pilot, Seattle ATC was talking with the scores of other airline pilots whose flights were affected by the emergency. See if you detect any testiness, confusion, or exasperation in those pilots’ replies.
  • What could have happened next? The two F-15s were in range of the plane within a few minutes of its takeoff. They were awaiting orders to shoot or force it down, if the pilot had turned toward Seattle or Tacoma rather than away from them.

Could this have been much worse? Clearly it could have. If a ground-crew member had been purposeful, better trained, and able to haul explosives or other damaging cargo with him onto the plane, he could have reached downtown Seattle before the F-15s got to him. That’s a plainly possible “What if?” like the strychnine, anthrax, and school-bus examples with which I began. And even if the resulting toll were “merely” comparable to that of a large-scale American gun massacre—it could not have been comparable to the 9/11 attacks, since a commuter plane like this is so much smaller and carries so much less fuel—the psychological impact of airplane-borne threats is always greater than sheer numbers would indicate. (Airline travel in the United States is statistically about the safest thing you can do; many people are nonetheless afraid of flying.)

Could it have been avoided altogether? We’ll see what the details show. As with the Germanwings crash of 2015, in which a depressed 28-year-old airline pilot deliberately flew himself and 149 other people into the ground, the easiest break point would have been someone noticing how disturbed a staff member had become.

As for (even) greater screening and supervision of airport and airline staff members, that is sure to come. One reason North American and European airlines are now so exceptionally safe is that they view each crash, death, or even near miss as the occasion for systematic learning and error reduction, to address the vulnerability that has just been revealed. Because decades of refinement have eliminated so many air-safety problems, the crashes that do occur tend to involve some previously underappreciated point of failure—for instance, the autopilot and crew-management issues that led Air France Flight 447 to plunge into the Atlantic Ocean. (Think of how different this is from car-crash investigations or the stories of mass shootings, where the background factors are so familiar that most people can guess them before the full details come in.)

I hope that, when the facts are in, the response to this odd, sad incident will resemble what the aviation system usually does with its failures, rather than the way the political system typically behaves. That is, I hope it serves as a source of guidance for further threat reduction, rather than as fuel for panic and finger pointing about the modern reality that some “What if?” peril will always remain.

We want to hear what you think about this article. Submit a letter to the editor or write to