I fear we will never know what exactly happened to that plane. One of our sources said, “The Earth is big. The Earth is really big.” The open ocean is really vast, especially when you don’t know where to start looking. See our story from Monday: There’s stuff out there, but it’s not plane wreckage.

How does a plane in the modern era simply vanish? This is the question I posed in the first story I wrote about Malaysia Airlines Flight 370, and it’s still the key question, the critical question and the question that gives us a path to something like an answer. This is a tragedy wrapped in a riddle.

Because I’ve written some stories on this topic in the last three weeks, people stop me and ask where is the plane. I’m sorry to say I don’t know. I have no theory of the case, no hypothesis, no conjecture even. But I’ve put together a framework for thinking about what probably happened.

The rational thinker in this complicated world must sometimes entertain multiple, contradictory thoughts, and remain comfortable with ambiguity, with a range of possibilities in lieu of certainty or conviction. There is a quantum weirdness to this plane story. A lot of scenarios are still in play. I wrote about some of these scenarios a couple of weeks ago.

Even if we can’t know what happened, we can narrow down the range of possibilities, or at least create some hierarchy of probabilities. To do this we rely on that old standby for rational thinking, Occam’s Razor.

There are several categories of explanation, including:

1. Mechanical failure

2. Diversion/sabotage/suicide by lone individual

3. Hijacking as part of larger operation

In the first category you’d put a fire of some kind, perhaps an electrical fire that takes out the transponder and ACARS system, etc., or a sudden decompression, such as a “Payne Stewart event,” in which the occupants of the cockpit are incapacitated, and the plane keeps flying on autopilot until it runs out of fuel. For the first couple of weeks of this case, I found this to be a plausible scenario.

This explanation now feels unsatisfactory. In this case there are really two anomalies at work: First, something went wrong on the plane to keep it from reaching its destination in Beijing, and second, something went wrong with all the systems that normally keep a plane in contact with the rest of the world. The second anomaly has to have the special feature of allowing the plane to keep flying (assuming the satellite data are correct, and the plane stayed in the air for another six hours or so).

A few days ago, we published a story about aviation safety, and discussed whether planes need more technology to enable them to be tracked by satellite. But of course, this plane did have such technology — it just got turned off or disabled somehow. It obviously wasn’t foolproof or tamper-proof. There’s a legitimate question about whether it makes sense to reconfigure the world’s aviation fleet because of what might be the ultimate “one-off” event.

Flying is safer than ever. Here are some of the things that John DeLisi of the U.S. National Transportation Safety Board told me:

“If you go back further than 10 years, many accidents were the result of engine thrust issues, loss of power, the engines not doing their job. But holy smokes, there have been great advances in the design of turbine engines, so that they are highly reliable devices.”

When I asked him what’s more dangerous, takeoff or landing, he said: “Accidents occur during landing more so than during takeoff. But you’re talking about something that’s so statistically rare that you’re in the 8th or 9th decimal place looking for where to tell that there’s a difference.”

In general, if you want to look for a place where an aviation hazard may lurk, look at the cockpit. The common failure mode, he said, “is humans figuring out ways to crash perfectly airworthy airplanes. That’s the area that we really need to be focusing in on.”

To be clear: He was not making a statement about Flight 370. We were speaking about aviation safety in general.

But the broad picture right now of civilian aviation is that it is extraordinarily reliable and really one of the great achievements of civilization. When you get on a plane you can rationally expect to get where you’re going (if, perhaps, hungry and cranky).

It’s possible there’s a failure mode that would cause the pilot and co-pilot to divert back to the nearest airport while losing all ability to communicate — no transponder, no ACARS, no radio — and then that failure mode would knock them out so that the plane flies on its own. Invent your own narrative: There’s no evidence for it, but you can’t rule anything out. Still, the fact that the plane deviated from its flight path, and turned back, just as the plane left Malaysian airspace and was not yet in touch with Vietnamese air traffic controllers, raises suspicion that this was a consciously timed event. The plane was cruising: The safest point in the flight.

Usually in these cases, the simplest explanation is the best one. You look for explanations that don’t require too many moving parts or extraordinary occurrences or coincidences. You want the parsimonious theory. Generically, this is why most conspiracy theories aren’t correct. They’re too elaborate and require too many assumptions. Conspiracy theories are Swiss watches; when trying to solve a mystery, first look for a sundial.

Let me throw in the JFK assassination just to keep things hopping here: As I wrote a few months ago during the 50th anniversary of the assassination, the conspiracy theories are innately elaborate, whereas the Oswald-did-it scenario is simple. Oswald’s motivations remain the subject of much intriguing investigation — see Ron Rosenbaum’s exhaustive discussion in Slate a few months ago of Oswald’s contacts with Cubans in Mexico City (channeling the author Philip Shenon’s new book on the assassination). Leaping lightly over that rabbit hole, let me note that, if Rosenbaum and Shenon are correct, the JFK assassination is not a meaningless, random event, but rather is located very much in the history of the Cold War and the relations between the United States and Cuba (and is specifically a kind of blowback for attempts by the CIA to kill Cuban leader Fidel Castro). But that still doesn’t make it a conspiracy. Oswald is the shooter, and he’s a rogue operator, under this formulation.

So let’s look at Flight 370 scenarios in categories 2 and 3. They have in common the idea that someone with a malevolent or lunatic mind, someone who knows how to fly a plane, has intentionally diverted Flight 370 from its normal flight path and flown it somewhere else (probably far into the Indian Ocean if we are to believe the satellite data). Of these two scenarios, I favor the first on simplicity grounds: It’s more likely to be a rogue agent, a single person at the controls and causing the entire tragedy.

When my sources tell me they think perhaps the bad guys wanted to use the plane for a subsequent operation to drop a bomb on a distant city, I fail to detect the resonant ping of plausibility. It’s unwieldy. There are so many moving parts, and co-conspirators who remain invisible, the whole scenario by its nature unfalsifiable.

So, who was this bad person? Dunno, and pending further information I don’t feel comfortable speculating. Why would he do it? Dunno. You’re into motive. Speculating freely, we might entertain the idea that someone suffered from mental illness, a mania, a bad drug reaction, temporary insanity, some kind of fanaticism. There is no publicly shared evidence that implicates anyone in this case.

But people do strange things. We are a complicated species. There is sickness in this world and depravity. You can make up all kinds of scenarios about cockpit invasions, and foiled hijackings, or semi-inept hijackings, and so on, and we just don’t know how it played out. But the simplest explanation is that someone obtained solo control of the plane, turned off the transponder and the ACARS, stayed silent on the radio and flew the jet to the middle of nowhere (possibly with a change of mind along the way). This appeals to me in part because it is technologically doable without extraordinary measures. It can be done, sad to say.

Every year there are millions of planes taking off, and they almost always land where they are supposed to touch down. Anyone in the cockpit — especially a veteran pilot — gets multiple layers of review by superiors. But aviation is growing steadily around the world, and there are so many flights in which something could, potentially, somehow, inexplicably, bizarrely, go catastrophically wrong. Given enough chances, even the exceedingly improbable event can happen.

But no, I don’t know where the plane is.