The Structure of an Accident

William Langewiesche, the author of Columbia's Last Flight, talks about the fundamental problems within NASA that led to the space shuttle's demise.

When the Columbia shuttle began to disintegrate during re-entry over California on the morning of February 1, 2003, few people were even aware that the space shuttle had been in flight. Its launch had not attracted near the attention that that of its cousin in catastrophe, the Challenger, had in 1986. While the Challenger mission drew national attention as the first space flight to carry a private citizen (teacher Christa McAuliffe), last winter's Columbia mission was a low-priority, routine "science" mission. But when the Columbia broke apart over East Texas, killing all seven astronauts on board, the haunting images of a NASA disaster once again hit television screens across the country. Suddenly, what had been a low-profile mission was now a high-profile accident.

—Steve Grove

What was your initial reaction when you heard about the Columbia accident?

I had not been aware that the shuttle was even up in the air, and I'm definitely not a close follower of the space program. I had sort of a vaguely pro-space, pro-human space flight feeling. On the day the thing went down, like everybody else I thought, too bad about the astronauts. But basically my first thoughts were technical—and my first reaction was tile. The shuttle had burned up on re-entry, so they must've had a problem with the tile. And it turned out that was NASA's thinking as well, despite better evidence that in fact it was not the tile.

Very quickly I got a call from Cullen Murphy (managing editor of The Atlantic), and we agreed right then and there that I would pursue the story. But we really thought I would wait a year, to let the dust settle.

That obviously changed when you became integrated within the CAIB. How did that happen?

In the first few days after the accident, NASA was talking in a very peculiar way. They were just all over the map—speculating in public, which is something you really don't want to do. It's very easy to misread or to be misread. So it was odd. I called Bernie Loeb, an old friend who's retired from the NTSB (National Transportation Safety Board), and I said, "What's the deal here? What's going on with NASA?" Bernie said he'd never seen anything like it, and he didn't understand what logic NASA was following.

And then I sort of let it drift for a while, wandered off overseas and continued my work, and at another point when I was coming back through Washington briefly, I called up the NTSB and asked them, "What's this commission that's running the investigation, the CAIB? And do you know any of the guys there?"

It turned out that one of the members of the CAIB, a guy named Steve Wallace, was a fan of my writing. He had taken a quote from the piece I wrote about the Egyptian airliner that went down and hung it on his wall. And so there was a friendly reception from the CAIB. They told me that I should go down to the Kennedy Space Center in Orlando to take a look at the wreckage. No reporters had been allowed in to see the wreckage. As I was driving down there, making calls on my cell phone, the doors started slamming shut in my face. It was a really weird reaction—hostile—and strange because I wasn't coming down as a partisan or anything. I was coming down there because the people who controlled the wreckage, the CAIB, had invited me. I wasn't on a crusade of any kind, but the people at Kennedy were extremely defensive and quite hostile to me.

Red lights must have been going off in your head.

It was amazing. I'd actually had no contact with NASA before. I realized that NASA's position here was really unduly defensive, it was kind of paranoid. I realized that they were under attack. Any investigation is a form of attack, and it's never a pleasant experience. But I've seen other entities and organizations being investigated, especially within the U.S., reacting quite differently. When I got to Kennedy Space Center, some of the people I met struck me as apparatchiks. They were like antiques, in a way, they were a throwback to an earlier time. It was very obvious in the way they saw the outside world, in the way they saw their own institution. And I thought of the Soviets. These people live so deep inside an institution here, that they can't see outside of it.

Later on, NASA actually opened its doors to me. After I was integrated at the CAIB, I was meeting people at NASA who were helping the board, and they wanted to explain their point of view to me. And so I went down to the Johnson Space Center in Houston and spent quite a bit of time down there. And I spent quite a bit of time in Washington, talking to the people who were deep within the NASA organization, and so my relationship with NASA changed and improved. They were much less defensive, and in a sense they understood that we're all in this together. Human space flight was a huge national project, with enormous expense, and it had gone very, very wrong. And the farther I got into this, the more I realized just how wrong it had gone, and how deep the flaws were. In a way the interests of the pro and con finally merged. NASA's still very defensive, but nobody wants to see the U.S. failing miserably with its space program. Nobody wants to see the United States executing bad policy badly. Which is what happened.

Linda Ham, a manager in the MMT (Mission Management team), said that the foam strike represented "no safety-of-flight kind of issue." NASA engineers disagreed. But when their requests for further satellite imagery of the wing were sent to Ham's office, a mid-level employee responded with a denial, assuming "more or less correctly, that Linda Ham had decided against … imagery." The engineers ("hardly rebels," you call them) pushed no further. What is it about NASA that discourages dissent?

That's a very complicated question. All large organizations do. NASA is astonishingly bureaucratic. I'm sure that people who are in the business of studying the sociology of large organizations would put it in a certain age group; it's sort of a young adult or an adolescent bureaucracy. It's at a certain stage in its development where the initial energies and camaraderie and ability to communicate and sense of mission—all that's disappeared. And it's not yet in the really mature stage, like the U.S. military is right now, where they've been able to work through a lot of these problems. I've been impressed by the Army, specifically, in my dealings with them in Bosnia and Kosovo, by the way communication does work within that organization. I mean, yeah, the Army is famously stupid, and there's the right way and the wrong way and the Army way and all this stuff… but in fact in the larger organization they've worked through a lot of problems. NASA hasn't done that.

After the accident, it must have been frustrating for those engineers whose concerns about the foam strike weren't given adequate attention. If their requests for satellite imagery had gotten through, the accident might never have happened.

In a large organization, you're always going to find the whistle blowers—they're always there. And they are always blowing whistles. Most of the time what they blow whistles about turns out not to be a problem. After an accident, it's very easy to go back and find the whistle blower. And I was aware of that trap as a writer, in the way I think accident investigators really need to be aware of it. It's important not to engage in cheap "20/20 hindsight." There were larger structural problems at play here. The problem had to do with the people who were not willing to listen to legitimate concerns.

I wonder if the Columbia accident seemed vaguely similar to the ValuJet crash you wrote about back in 1998. You called the ValuJet accident, which killed 110 people in Southern Florida, a "system" accident: an accident "born of the complex organizations with which we manage our dangerous technologies." Do you consider the Columbia accident to be a system accident as well?

Yes, I think it would be classified as a system accident. And the sign of that was the unusual breakdown in communication at the end, where Linda Ham never heard the request for visual imagery, and the engineers heard the denial of the request that was not intended for them, and assumed that it was for them... That part—the missing of the two ships in the fog—that's a real sign of a system accident. I mean, with the complexities of the communication paths within NASA, no one could have anticipated that particular failure route. And that magic ability of a failure to bypass safeguards and to find new routes to inflict catastrophe is one of the key characteristics of a complex system accident.

Of course you guard against that partly by making your system as simple as possible. And certainly NASA's bureaucracy is not at all a simple system. It's an enormously complicated one. Part of its complexity is that so much of it is not even written, it's just understood. As Hal Gehman (the leader of the CAIB) said to me many times, "These are the unwritten rules." And unwritten rules tend to get really, really, complicated.

Is it easy to have a simple bureaucracy manage a complex technology?

No, it's difficult. But I think it's certainly possible to have a much simpler bureaucracy than NASA does now. In the early days of managing the Apollo program, for instance, I would imagine that in terms of communication, the bureaucracy was simpler. It was more collegial, less hierarchical, and you didn't get these parallel hierarchies in those days, when NASA was new and had this real purpose. And now, in these days of thirty years of defensiveness and dishonesty, of dishonest policy and dishonest missions, these unwritten rules have flourished.

The Challenger disaster in 1986 occurred when faulty O-rings let hot gases leak into an external fuel tank, causing a massive explosion. The investigation that followed concluded that flawed communication and intense scheduling pressures—two of the same criticisms made by the investigators of the Columbia disaster—contributed to the accident. More than fifteen years later, it would seem NASA did not make substantial changes. Or did they clean up their act, and then slip back into bad habits?

It's both. NASA made some changes, but not enough, and then it went back on the changes. So yes, the end result of the Challenger investigation was that basically nothing was solved. Whether NASA is capable of solving its own problems, I'm very skeptical. How realistic would it be to expect NASA to clean up its own act from the inside? It's not realistic. Especially because the entire premise of its operation is flawed.

What do you mean?

There are other fundamental problems here. Obviously, if you look at what NASA's been doing, the problems are deeper than a lack of communication. They have to do with a lack of real direction. There have been thirty years of half-baked policy formed by the White House and Congress, to which NASA acquiesced out of the sense of the need to survive. There's been the dishonesty of NASA leaders in presenting their case to the American public and to Congress. And there are NASA's problems with managing its budgets and over-promising things. All this is much more serious than the internal bureaucratic barriers—which are a very serious problem at NASA, but the larger thing is that the whole direction needs to be questioned at this point. And it's not just because a few people were killed.

As a nation, we need to rethink NASA. What are we doing in space? And to what end, and in what time frame, and how much money are we willing to spend to do this? We need to reconsider the definition of payback. What should we expect to see in terms of tangible benefits for this kind expenditure? One answer may very well be that we should expect a 300-year time period for payback—and that's fine. We can still afford, and in fact we must afford, to continue with human space flight—but let's be honest that there may be no immediate payback.

It seemed to me that you were talking about that lack of vision when you wrote that this mission was undertaken primarily to "clear the books."

Yes, and that's a very odd thing to be doing. In a way, you could say we're treading water as a nation in space. But you can't really tread water like this, because it happens in a political environment, and it's destructive. And in fact you start to sink.

You describe a couple of rescue options that NASA might have employed if they had realized the extent of the damage caused by the foam strike. Do you think a rescue mission would have worked?

Yes, I think it's quite clear that sending up another shuttle to rescue them would have worked. They were saying that if a second mission met all of the standards, they would have had a five- or ten-day window for picking them up. Yes, they said it would have been risky, but they were saying that after the fact, when it's easy to say it was risky. It's true that the chances of foam coming off a second shuttle and taking it down were there. But they were small. So the chances were high that a second shuttle would have gone up there and taken a little battering like the rest of the shuttles did, and then picked up the astronauts and come home.

The shuttle seems like a fascinating machine. As a pilot yourself, you had the opportunity to fly a shuttle flight simulator alongside an active shuttle commander, Michael Bloomfield, when you were writing this piece. What was it like?

The machine itself is complicated in that it has many parts that are redundant and overlap and intertwine. It's nothing but a giant, splendid, system machine. The systems themselves aren't necessarily that complicated, but there are a lot of them. Some of them are quite exotic to a guy like me, just a normal pilot. I'm used to the air and to airplanes, but not to space. Other systems are pretty much just like an airplane. So the shuttle simulator was in some ways very familiar and in other ways of course utterly incomprehensible.

As far as the actual handling of the machine—which only occurs once the shuttle is in relatively low altitude and going through Mach 1—it handles just like a highly wing-loaded and somewhat heavy airplane. Which is what it is. The fact that it is a glider means nothing at all, because there's so much speed and weight behind it. From Mach 1 down to touchdown, it's more or less like an airliner with no engines. You fly it perfectly well. Like a fighter, or a business jet. It handles quite beautifully.

I've got a lot of respect for people involved with the shuttle. These guys are great pilots. And some of the flying I saw inside of the shuttle… there was one crew in particular that was flying, while Mike Bloomfield and I were sitting in the back and watching with headsets on. They were going through a couple of re-entries. It was on autopilot, and they were dealing with failures. The commander on that flight, he took it over and kicked off the autopilot. The simulator was sucking in low on final approach, and he had the intelligence, at this very high work-load moment, to realize that something was going wrong here and to throw all the planning away and to basically go by the seat of his pants. It might as well have been on a Piper Cub coming down to a landing. He just left all the instrumentation, kept only the flight guidance on, and did the actual flying himself. It was a beautiful piece of flying. These pilots are extremely bright, and they're just really decent people. And I find it so sad that all of this intelligence and all of this effort and all of this expenditure, and to some extent all of this risk, seem to have no real purpose. I mean, what's the real purpose?

Whether all this effort is worth it, whether we need to have a machine that can land on a runway and has to take off vertically, this can be questioned. Maybe the Russians have had it right all along: you come back in a can and parachute down. Maybe you come back in a rocket ridden in reverse—though that seems unlikely to be very efficient. But who knows? I think it's an open question if we really want wings on a machine that spends most of its time in space, where the wings don't do any good. And of course in this case, the wing is what was hit.

While the CAIB was critical of NASA's organizational structure, its suggested reforms seem pointed toward further developing the shuttle for human space flight. Critics claim human space flight should be scrapped all together. Do you believe NASA should continue to pursue human space flight?

The rationalists are right. I mean, the rationalists say that manned space flight is a publicity stunt and we can do it all better and cheaper with robotics, and get better science out of it. You can't argue with that. I think they're absolutely correct. Nonetheless, I think, personally, that manned space flight needs to be continued and at considerable expense, maybe in fact at a higher budget than now. Now, that doesn't mean that the shuttle or the International Space Station needs to be sustained. What needs to be done, in my opinion, and this is something that's only come to me as I wrote this piece, is that we have to have a period of rethinking, after a public debate. "Okay, we're gonna go ahead with human space flight, and we're gonna spend a ton of money on it. Now, how are we going to do this?"

Right now, we have reasons that are purely altruistic, purely scientific and exploratory. But a new reason will come. The Chinese are going to end up putting a man in space, and they're going to end up putting a person on the moon. And we could very well at the same time find ourselves in some sort of a cold war with the Chinese, a power struggle for prestige, and all of the sudden we're going to have a different reason for putting people into space—for the same reasons we had before, which were basically Cold War-style power politics, prestige. I think the time has come for a twenty-year stand-down to re-think. But then there should be a definite commitment to proceed beyond there.

The CAIB's report gave twenty-nine recommendations, fifteen of which must be fulfilled before NASA can fly again. Yet while the board was critical, Gehman also said in a press conference, "This board comes away from this experience convinced that NASA is an outstanding organization." Do you think the board struck the right balance in its report?

I think that Gehman and the board in general struck the right balance—NASA has failed really badly here. I think that it was a very tough report, and being in the outer offices when it was being delivered made its toughness very clear to me. This is not something that NASA was very eager to receive.

Gehman is a very wise man about bureaucracy. And he's seen the Navy and the U.S. military in general clean itself up a lot since the Vietnam years. He's optimistic about NASA, but he's not blindly optimistic. And I'm not talking about his political statements. I'm talking about what he really thinks. He has volunteered, as I said in the piece, along with some of the other board members, to go back and monitor the situation for NASA, and that would be a help, to have some kind of watchdog who would come back and who really knows the problems. But that won't be anywhere near as effective as having a strong congressional response. Congress must see its way through its thicket of political thinking and pork-barrel initiatives to an honest and sustained critique, a constructive rethinking of human space flight.

There are a couple of people at NASA who come across in your piece as bad managers. Is there anyone you sympathize with at NASA?

To tell you the truth, I sympathize with all these people. I really do. I sympathize with Linda Ham, who was a terrible manager—she was arrogant and personally difficult and all of this, but I sympathize with her. And that sympathy comes from the recognition that we could all screw up. And the recognition that more importantly, it was the system that screwed up. It was the entire structure that they were operating in that was failing them. And they were sort of cogs in that failure.

I think the person I feel the least sympathy for is Sean O'Keefe (NASA's administrator). Nobody at NASA ever stepped up and accepted responsibility for what had happened, even to the point of saying, "I done screwed up, I contributed to this failure." Nobody at NASA, that I know of at least, ever did that. Not Linda Ham, not Ron Dittemore. Well, they would have a reason not to do it—but the guy at the top, he doesn't have a reason. Why didn't O'Keefe apologize? O'Keefe is the guy who should have said, "We screwed up, our organization screwed up. I'm responsible, I cannot expect others beneath me to accept their responsibility, internally or externally, unless I do. And therefore I am accepting responsibility, and therefore I'm probably resigning." That would've caused a domino effect, then other people could have admitted their partial responsibility. Maybe he didn't even have to resign, he just had to say, "We screwed up, and I screwed up as a manager, and I'm largely responsible for this accident."

What do you think of the American public's reaction to the Columbia disaster?

I think that the public isn't outraged by the loss of Columbia. And that's fine. We don't need the public to be outraged by it. I think that outrage is the wrong emotion anyway. The public probably doesn't feel much one way or the other about it. I think that there was a need for a very strong appraisal and accident investigation, which is what we got. But the public is not going to say we shouldn't continue manned space flight because the program was badly managed—they're going to say we shouldn't continue manned space flight because nobody can tell us why we should continue manned space flight. That's the big one.

I think the reason so many people didn't even know the Columbia was in orbit has to do with NASA's lacking a clear vision—it was just, "Oh, NASA's doing another ozone study."

Plus, NASA's so wrapped up in their own PR that people don't have any faith. People are correctly skeptical of the claims NASA's made for years and years, about how wonderful the science is. It's way too "gee whiz" for our times.

What do you predict for the future of human space flight in the United States?

There's reason for skepticism and pessimism, as long as we don't see a series of really strong public hearings on human space flight in Congress through the fall, and I don't think we're going to. Where's the follow up, guys? That's the question. I am really skeptical, as I know many of the board members are, that in the long run this report is going to have a fundamental curing effect. It probably won't. It's going to require more than just one little ol' report from the CAIB.

Changing the subject a bit, I wonder if you could tell me a little about your process as a writer. You obviously gathered an enormous amount of material to write this story—how do you go about putting together feature-length pieces like this one?

I don't even know how to answer that. I gather as much information as I can. While talking to people I'm always thinking, what's useful to me? What's not useful to me? What is the basic structure evolving to be? What gaps remain? If you're working on a piece, you're concentrating so hard on it that the information goes in and it stays in the pea brain. And then when I have this mountain of stuff together, ranging from taped interviews to all kinds of other documents, I sit down and say, "Okay, how do I want to tell this story?" And I draw some sort of a flow chart, which I'm free to abandon within the half-hour, and I usually do. And then I start writing.

The main thing is, I don't ever want to lose the reader. You've got this rare, privileged position of having the reader's attention, and you don't want to abuse it. That's my number one concern—always, always, always: fear, in some ways, of not treating the reader with sufficient care and respect. Now, there are various ways you can screw up: you can bore the reader, you can lecture the reader, you can talk down to the reader, you can be overly plodding, you can be overly flip, you can be overly manipulative. Our readers are sophisticated, if they think the writer is doing some cheap trick, they'll put it down. So I think the main thing I have is this overriding concern for losing the reader.


One of the many people who knew nothing of the mission until its demise was William Langewiesche, a national correspondent for The Atlantic. Soon after the shuttle went down he agreed to pursue the story. The article resulting from his months-long investigation, "Columbia's Last Flight," is The Atlantic's November cover story. Langewiesche, a professional pilot before he began writing for The Atlantic, had covered aviation disasters before. His cover story in the November 2001 edition, "The Crash of EgyptAir 990," won a National Magazine Award for excellence in reporting. In 1998, in "The Lessons of ValuJet 592," he wrote about the system breakdown that led to the 1996 crash of a ValuJet plane into the Everglades. But Langewiesche had never covered space flight before, and admittedly knew little about NASA. This unique perspective gave him the knowledge of an expert behind the eyes of a beginner, and helped Langewiesche observe serious cracks in NASA's foundation.

Langewiesche gained exclusive access to the Columbia Accident Investigation Board (CAIB), the "interagency" group of high-ranking civilian and military officials who investigated the accident. For several months he followed the CAIB as it cut through a NASA bureaucracy that Hal Gehman, the retired Navy Admiral who led the investigation, described as an "incestuous, hierarchical system, with invisible rankings and a very strict informal chain of command."

The CAIB's report, released in late August, showed that the root cause of the Columbia accident had not been an engineering mistake, but a management one. After a piece of foam the size of a catcher's mitt broke off of the shuttle and struck the leading edge of the left wing eighty-two seconds after lift-off, NASA engineers had requested satellite imagery to assess the damage to the shuttle. Had the pictures been taken, they probably would have revealed the dire condition of the wing and prompted a rescue mission to save the astronauts. Instead, the imagery requests didn't make it through the tangle of NASA's bureaucracy, and complacent NASA managers dismissed the safety concerns posed by the foam strike. When the shuttle broke into the earth's atmosphere on February 1, the gaping hole led to the craft's disintegration.

How had NASA's management failed so miserably? The story Langewiesche found involved "ignorance, insularity, with bad luck allowed to go unchecked." Scheduling pressures enacted by new NASA leadership discouraged dissent. Fear for their jobs silenced engineers. A complex bureaucracy muddled the lines of communication. Frustratingly, these were some of the same problems that the Challenger investigation had uncovered back in 1987. Had time bred complacency, or is there something inherently difficult about managing a complex system with a complex bureaucracy?

Critics have since called for the termination of NASA's human space-flight program, claiming there is no need for humans to be pushing buttons up in a shuttle when those buttons can be pushed from the ground. Meanwhile, NASA has been touting a new shuttle-repair technique: an adhesive application process to patch holes in the thermal protection tiles on the shuttle's belly—the adhesive can be applied with a 49-cent foam brush from Wal-Mart. In this interview, Langewiesche—a self-described fan of human space flight who took part in a shuttle simulation for his story—suggests that both of these ideas miss the point. Human space flight should continue, he argues. But only if we are willing to patch the major holes in America's thinking about space flight, too.

I spoke with him by telephone on September 26.

Presented by

Why Principals Matter

Nadia Lopez didn't think anybody cared about her middle school. Then Humans of New York told her story to the Internet—and everything changed.

Join the Discussion

After you comment, click Post. If you’re not already logged in you will be asked to log in or register with Disqus.

Please note that The Atlantic's account system is separate from our commenting system. To log in or register with The Atlantic, use the Sign In button at the top of every page.

blog comments powered by Disqus


A History of Contraception

In the 16th century, men used linen condoms laced shut with ribbons.


'A Music That Has No End'

In Spain, a flamenco guitarist hustles to make a modest living.


What Fifty Shades Left Out

A straightforward guide to BDSM

More in Entertainment

Just In