Airline Safety: A Special Report

it takes a lot of people to put an airplane in the sky--but any one of them can bring it down

August 3, 1979: Grass doesn't grow here anymore. Yesterday it rained. Now the sun is out and there is a smell coming from the black, pear-shaped scar that stretches 100 yards across this green field, a smell of kerosene and ashes, like ancient lamp oil and burned insulation. And another smell, too, a strange incipient searing smell that makes you want to move away and discourages any further investigation.

The field is many acres, surrounded by high barbed wire. Grass, cornflowers, scrub oak and weeds grow wild out here. Except on this one spot. If you look closely, you can see that thousands upon thousands of pieces of white wire are embedded in the rich, black mud that sticks to your shoes as you walk along. The strands are buried deeply, as if by unimaginable force, and when you pull on them, bits of metal come out of the ground. Each length of wire pulled free unearths more and more parts, tiny electronic components, devices, shards of plastic and scraps of aluminum melted into odd shapes. Some of the fragments have identifying numbers on them and some are still painted with yellow-green inhibitor. There are rivets and bolts, nuts and doublers and a few hefty remnants as big as a man's hand. Suddenly, it dawns on you that this is not merely a bald, scarred patch of mud. It is a virtual warehouse of scrap metal, avionics, knobs, switches, dials, parts, evidence....

They say there is other evidence buried here, too, and if you dig long enough, you begin to believe it. Stand on this spot and watch the jets scream past overhead and think about what you have smeared on your hands. The impulse to go on digging just disappears.

"It's spooky," a cop out here tells me, "a very spooky spot." He should know. He has to work here. "Must be weird for you," he adds, taking great care to clean his boots of any mud clinging to them.

On May 25, 1979, about three o'clock in the afternoon, American Airlines flight 191 crashed here. When I arrived, I found the area sealed off by police. I went through a trailer park to get around the blockade. The flight had terminated right inside the Chicago Police Department's Canine Training Center. Just beyond that, the quiet community on West Touhy was laid out on nameless, tree-lined streets with double-wide trailers in rows, and beyond that were the great Standard Oil fuel-storage tanks that could have made the crash even more spectacular if that McDonnell Douglas DC-10 had been able to stay in flight for 41 instead of only 31 seconds.

As I approached the crash site from the interior of the trailer park, passing through dappled sunlight and shade, I turned a corner and came face to face with an enormous piece of the plane. It was a section of fuselage fitted for a cabin door. Apart from the fact that it was separated from the plane, it was undamaged and sat there as it might have sat in a museum, revealing nothing of the magnitude of the explosion that had put it some 150 yards from the rest of the smoking wreckage, and nothing of the damage it would have done had it landed on a trailer instead of on the narrow street.

Beyond the trailers was the entire plane, some 270,000 parts that had once been collected into a whole by 2,000,000 fasteners, bolts, nuts, rivets. The fires were out now, but the story was only just beginning for most of us.

Everyone's had the experience: You want to see it and you're sorry when you do. I couldn't connect with it then, but four of my friends had boarded that plane. I found myself unable to react; that came later, with interest.

Yellow body bags and brightly colored flags dotted the area. There were police and helicopters, cars and equipment. There were black flags, too. The largest piece of the plane was the number-two engine, the one that runs through the vertical fin. Everything else was chips and slices, fragments and hunks, blackened beyond recognition; and when you saw something that might have been an unfortunate tree in the path of destruction, you didn't go near it. Pieces smaller than that you'd have to touch to be sure whether they were made of metal or plastic or something else altogether.

When the police picked me up, I was standing in the street by the clean, gleaming section of fuselage (so pristine it made you want to look for a plaque, MC DONNELL DOUGLAS DC-10 SERIES 10 CABIN DOOR...). I was put into the caged, locked back seat of a police car and driven out of the area to a blond brick building, where they determined that I was a reporter and sent me off to join the other reporters.

It's true what they say about air disasters: You've never seen anything like it, no matter what you've seen. I once covered a disaster in West Virginia. A makeshift dam broke and wiped out 18 miles of coal-mining communities. A lot of people died there, the bodies were everywhere, and the power of that water sluicing through the narrow valley had made the land look like the surface of Mercury, 90-pound railroad track in pretzel shapes. But it was nothing like the crash of flight 191--that looked like a nuclear-reactor meltdown. In West Virginia, I had seen bodies. In this field, I just saw faces.

An aircraft-accident investigator told me all about it, how it is when you're in there, really in there, going down. "You know what happens in many of these cases," he said. "If a person knows that he's going to die, he produces enormous tensions and pressures and literally destroys his heart muscle--not exactly destroys it but causes it to appear as a heart attack. On more than one occasion, we've had that kind of report from a coroner--you know, 'The captain died of a heart attack.' And you say, 'Oh, Christ, but did he know what was happening to him? Does that mean something?' Of course, it means a lot. It means that guy was all tensed up." He laughed sadly. "He knew he was going. All it says to you in essence is that it was no surprise. A guy who gets shot in the back of the head doesn't die of a heart attack. A guy who has control of the airplane, literally trying to keep it airborne, will sometimes die of a heart attack. I've seen them tear control columns right out and break them from the tension in their hands. And I've seen them break their own bones doing so."

I stood with the reporters in the heat at the departure end of runway 32 Right for a few hours, and then we were led on a tour of the site, which consisted of having us stand in the field about 50 yards from the main wreckage and reminding us from time to time not to wander off. By then, I had had enough. I snapped a few pictures--of the emergency helicopters parked behind us, the police cars, the green fields, of the unfathomable, surreal destruction and the reporters, laughing and joking about it (if you'd just happened on the scene unaware, you'd have thought someone was shooting a movie about Vietnam--the high-intensity lights they use for night work standing on their yellow-painted stalks, turned off and gleaming in the sun, as if waiting for the director to call another take). And I photographed Elwood T. Driver in his blue jump suit with the snappy insignia of the National Transportation Safety Board (NTSB). He was leading the investigation and would make much news in the coming days and weeks.

Now, as I stand on this spot again, holding a two-pound piece of the DC-10 that crashed here over two months ago, Driver is a few blocks away at a hotel, chairing the hearings into why and how this happened. When the hearings began, I showed up to sit in the audience with the pilots, who moaned and groaned each time a "professional investigator" or an "expert witness" made another Orwellian statement, another improbable grope for an answer. Even chairman Driver expressed his exasperation with the nonanswers he was getting, as each group participating in the hearings attempted to blame the others for the crash, making no real attempt to generate useful information. "I feel like I am being caught in a game of dominoes," Driver said wearily when Douglas Sharman, a Federal Aviation Administration aerospace engineer, once again refused to answer a question. "One man says, 'I can't answer it, the next one will,' now Mr. Sharman tells me he can't answer and you tell me Mr. Foster will. We are going to run out of people pretty soon." And, of course, they did.

When I arrived on the opening day of hearings, the NTSB public-relations man shoved a ream of paper at me, saying, "Press handout." It fell open on the table in front of me and I read, "Aviation Toxicology Laboratory, Case No. 3206.... Received by: P. Roberts from Dr. Kirkham at 8:30 A.M., June 1, 1979. SAMPLES: One bag of bone, one jar each of skeletal muscle and hair...."

I couldn't read any more. And after a time, I couldn't listen anymore to the tragic comedy of those hearings, so I've just come out to stand on this spot again, to see if what the officials tell me is true ("It's all overgrown now, they've plowed it under, reseeded it. There's nothing to see."). Well, some of it is overgrown and they did plow and rake and reseed it. But there is plenty to see (and feel and smell) and it will be a long, long time before anything grows on this spot again.

•

A year has passed and the questions just get larger: If there are avoidable aircraft accidents, why aren't they avoided? If there are survivable accidents, why are there so few survivors?

If flight 191 were the only crash or if the DC-10 were the only airplane, there would be fewer questions. But the closer you look at airline travel, the more it looks like a game of angels and great good luck, rather than skill and know-how and high technology. An engineer at McDonnell Douglas told me that flying in an airliner was 115 times safer than riding in a car, 28 times safer than walking and three times safer than riding a bicycle. And I tried to tell him that, based on the same statistical manipulations, it was safer to walk the tightrope than fly his planes and it was also safer to repair your roof than to take a bath. "Well," he said, "these statistics aren't meant to be exact measures, they're to help you get some perspective...." The point is, we should be looking at the problems as they happen and before they turn into major air crashes.

The airline industry (meaning not just the airlines themselves but also the airframe manufacturers, engine and component manufacturers and various Government organizations) would have you believe that airline travel is less risky than climbing into bed. Some days it is, some days it isn't. On May 25, 1979, it was 100 percent fatal for the 271 on board. The day before, on the same flight, it was 100 percent safe. Statistics can devil the hell out of you if you let them, but you pay your money and you take your chances, and in this game, undelivered goods are nonreturnable. It is therefore more instructive sometimes to put the numbers aside and look at a few unarguable examples: the actual air crashes.

Start with June 24, 1975, when an Eastern Airlines crew flew a 727 with 124 people on board into a known thunderstorm hazard on the approach to Kennedy International. The aircraft encountered a powerful wind shear called a downburst, was forced down into the approach lights, went out of control and was destroyed. Only 14 people survived at the scene; four died within a few days and one of them held on for nine days before giving up. Wind shear is a meteorological condition in which sudden changes in wind direction and/or velocity occur. An airplane, encountering such a wind change, can suddenly lose a significant amount of flying speed. Without adequate flying speed, the plane will simply return to earth. Downburst is a thunderstorm-related phenomenon that can cause wind shear. The crew of that Eastern flight had been warned about the problem. A pilot coming in ahead of the Eastern 727 said to the tower, "I'm just telling you that there's such a wind shear on the final on that runway you should change it to the northwest." Another Eastern pilot in an L-1011 had abandoned the same approach a few minutes earlier. The 727 captain, however, flew right into it.

Now, consider some amended statistics: that Eastern Airlines flight 66 on June 24, 1975, provided the passengers with an eight percent chance of getting home alive. For one of the most dangerous myths of statistics is that they hold forth the tempting notion that every flight is the same, that each time you board an airliner your chances of survival will be 99.9999 percent sure, which is clearly not the case. When you fly directly into a wind shear, your chances drop--along with your airplane.

Not two months later, a Continental Airlines crew did the same thing at Denver, only it was an outbound jet instead of one trying to land. Continental flight (continued on page 142)Airline Safety(continued from page 138) 426 was safe on the ground; all the captain had to do was wait and the thunderstorm would have gone past. Instead, he flew into the vicious winds that always surround thunderstorms. He was caught in a downburst and his plane was forced into a field off the end of the runway. The plane was destroyed.

Although there are some very sophisticated wind-shear monitors at certain airports now, numerous pilots recognize the fact that even simple wind cones, placed at various points along the approach path and the runway, could allow them to get a much clearer picture of wind conditions. But when there is a market for sophisticated monitoring equipment, the market for wind cones diminishes. And, of course, there is no real guarantee that all pilots would pay attention to the information even if they had it.

Thomas E. Gullett of Continental Airlines was captain of another 727 on June 3, 1977, when he departed Tucson, climbed to 30 feet and slammed through power lines and utility poles. "Before flight 63 started its take-off roll," said the NTSB, "the captain had clues that should have alerted him to the likelihood of a wind-shear encounter." That "should have deterred him from taking off under the conditions, especially since the wind factor was critical to remain within allowable weight limitations."

At the start of the taxi from the gate toward the runway, the flight engineer, who was computing the airplane's weight, said, "Well, we're overgrossed without wind." That meant that the plane was overloaded (by some 900 pounds, according to the NTSB) and might not lift off in the available runway length without a head wind to assist it. The captain went ahead, nevertheless.

But there's more: "All crew members," the investigators said, "were properly certificated, except the flight captain, who had not been route certified." That means he had not flown into or out of Tucson in so long that he should not have been there without a check pilot to oversee his flight. "Furthermore, a check airman, who had occupied a seat in the pilot compartment...did not certify as required by regulation that the captain possessed adequate knowledge of the assigned route."

The litany of incompetence goes on. The runway Gullett used was 7000 feet long, but he started his take-off roll 500 feet beyond the normal starting point, giving himself less space to work with. Captain Gullett did not realize he was starting at what is called the displaced threshold--which means there were 500 feet of runway that could not be used for landing, though he could have used the extra runway for take-off. He didn't. The NTSB observed that Continental's failure to ensure that its captain was route certified did not "lessen the captain's responsibility to have recognized the displaced landing threshold markings on [the runway] which conforms [sic] to the standard marking explained in the Airman's Information Manual, Part 1 [a basic text used by all pilots]. This part contains 'basic fundamentals required to fly in U.S. National Airspace System.'"

And that should help you rethink any assumptions you may have made about airline crews. Don't misunderstand this: The men piloting your ship are not necessarily a trio of full-bird idiots thinking of nothing but getting home to learn how their $100,000 a year is being invested. But they aren't always steel-eyed superheroes, either. I know a pilot who recently received approval to join United. She has 1200 hours accumulated over ten years in single-engine aircraft. She happens to be extremely competent and, of course, won't start out as a captain, but it does give you some notion about qualifications. In other words, put yourself in the pilot's place: You could do it, too. How good would you be?

Would you be like Continental's Captain Gullett? Or would you be like the captain of American Airlines flight 248 on August 5, 1977? He was flying a 707 out of San Francisco with 51 passengers on board. With a ceiling of 100 feet, the weather was bad enough to require that he fly by instrument flight rules (I.F.R.), and he was departing through what they call The Gap, a mountain pass out beyond runway 28. About 75 feet in the air, his left outboard engine exploded so violently that not only did it break away from the aircraft but it forced the left inboard engine back to idle power as well. With only two of his four engines operating normally, both on the right wing (the word unstable doesn't even apply), the pilot flew it away, got the damaged engine operating, turned around and started back to the runway. The tower called him to say that runway 28 was closed (his engine was lying on the far end of it). He called back, "Well, I don't intend to roll down there and hit it." And he landed. No one was even scratched. He got another plane and the 51 people went with him to Chicago. They had just learned one of the secrets of flying (what are the chances of that same group's losing a plane twice in one day?).

As a footnote, the cause of the engine disintegration was the use of a faulty fan. According to a crew member, the large fan at the front of the engine had been "oversped" on a previous flight, which meant it had had to be removed. American Airlines had removed it from another plane, checked it out and pronounced it airworthy. They then put it on one of the engines for flight 248. The hub cracked, releasing numerous fan blades, which exploded outward like shrapnel. However, a principal maintenance inspector for the FAA said the disintegration was the result of "normal wear and tear."

So would you be like that American Airlines captain and save 51 lives? Or would you be like Captain George T. Kunz of National Airlines? On May 8, 1978, he was flying a 727 with 52 passengers and six crew aboard, heading for Pensacola, Florida. His Class I medical certificate had the restriction that he wear glasses while flying. He wasn't wearing them. Notices to Airmen (NOTAMs) are issued by the FAA every two weeks in printed form for flight-planning purposes. It is mandatory for pilots to be familiar with these before making a trip. NOTAMs contain information critical to flying safety (how else would you know if, for example, your destination airport were closed?). The NOTAMs on May eighth said the instrument-landing system was not working on the runway Kunz planned to use; in fact, the runway was closed for repairs. Kunz was unaware of that.

Adding to his problems was the fact that Kunz previously had failed his proficiency test for descending too low on the approach. In his report, the airman who gave Kunz the check ride he failed said, "Kunz was having instrument-scan problems (sometimes referred to as tunnel vision).... The captain was given additional training and flew a recheck successfully."

On the night of May 8, 1978, however, (continued on page 268)Airline Safety(continued from page 142) Kunz once again descended too low and made a perfect landing three nautical miles short of the runway in the waters of Escambia Bay. (Kunz wasn't alone. The copilot failed to give proper altitude call-outs and the flight engineer turned off the device that would have warned them that they were too low.) It makes you wonder. If a pilot fails his regular check ride, what do they do to him? The answer is that they train him until he passes the exam, unless he's really burned out. The philosophy is called train to proficiency. It is no guarantee that the pilot won't make the same dumb mistakes again--nothing in the tests ensures that he will wear his glasses or read the NOTAMs or observe the dangers of had weather.

Some passengers think weather is not a problem for big jetliners, either because they are so large or because radar and navigational aids are so sophisticated that a regular air carrier couldn't possibly have a weather problem. It does not work that way. First, the laws of aerodynamics do not change when you slap an American or United insignia on the tail. All planes are susceptible to weather. Second, many general-aviation aircraft have more up-to-date equipment than that found on airliners. Bendix, for example, makes color radar for airplanes, just like you see on television. You won't see that in most airliners for some time to come. And, finally, airliners do end up in severe weather conditions, even if the airlines are not actively encouraging them to be there. In fairness to all pilots, it should be pointed out that quite often they just don't have the information with which to save their own lives.

On April 4, 1977, a Southern Airways DC-9 flew into a thunderstorm the National Weather Service characterized as one of the most severe in three years in the entire United States. The plane lost both engines and crashed trying to land on a highway. The pilot had no business being there, and no one will ever know exactly what he was thinking when he penetrated that fatal weather situation. There are pressures on pilots to fly under any conditions, though airlines and pilots may well deny it. In this case, in fact, Southern Airways would surely point out that its own operating manual states, "Flights shall not intentionally be conducted through thunderstorms...." Yet airlines make a practice of advertising their punctuality, and a pilot, sitting in the cockpit with two other pilots, watching other planes take off and being aware of the 100 or so passengers behind him, ready to go, may sometimes be forced into a position of rationalizing a need to fly in spite of forbidding weather.

In the Southern crash, the investigators decided that the engines had quit because of the ingestion of enormous quantities of water. An ironic note: Although the airplane's operating manual contained no information about how far the pilot could glide, he could have glided to the airport, some 14 miles away, even with no power. In fact, he flew 32.5 miles before crashing. Unfortunately, he flew in the wrong direction. The NTSB said, "The safety board was unable to determine precisely why the flight crew turned the aircraft about 180 degrees back toward the west-northwest instead of continuing toward Dobbins [the airfield]." Then he had to attempt a landing on a highway. He didn't make it. The FAA tests do not test whether or not the pilot will decide to go the wrong way when he could be flying safely to an airport.

Neither do they guarantee that the cockpit crew won't simply be asleep at the wheel. That was part of the problem on September 25, 1978, when a PSA 727 crew landing at San Diego was just shooting the breeze, not paying enough attention to flying. In strict fairness to the crew, a deadheading pilot, sitting in the jump seat, was running his mouth, distracting everyone. But no one seemed too concerned. It was a beautiful, clear morning and they had abandoned the instrument approach in favor of a visual, meaning they became responsible for keeping clear of other planes--which would have been fine, if they had bothered to look out the window. They didn't, and ran down two men in a Cessna 172, well-qualified pilots who were doing exactly what they should have been doing.

The PSA crew had been warned repeatedly about the small plane. The NTSB described the crash: "According to witnesses, both aircraft were proceeding in an easterly direction before the collision. Flight 182 was descending and overtaking the Cessna, which was climbing in a wing-level attitude. Just before impact, flight 182 banked to the right slightly and the Cessna pitched nose up and collided with the right wing of flight 182. The Cessna broke up immediately and exploded. Segments of fragmented wreckage fell from the right wing and empennage of flight 182." The 727's wing caught fire and the plane began to dive. A photographer snapped a picture of the people looking out the window at the fire, their faces illuminated by the orange glow.

The windshield of a 727 offers very poor visibility--like the slit through which a tank driver views the battlefield. Also, the Cessna is a high-wing aircraft, which makes it impossible to see what's directly above. Nevertheless, what is so maddening about that particular accident is that all the tools needed to avoid it were available and in use at the time. It puts one in mind of a man who jumps from an aircraft wearing a parachute--and simply neglects to pull the rip cord.

A look at some of these gross errors of omission is eye-opening. Because there is more to the story than pilot error. The air-traffic-control (A.T.C.) system is designed, according to the Government, "to promote the safe, orderly and expeditious flow of air traffic." That quote is from the Airman's Information Manual, designed to provide airmen with basic flight information and A.T.C. procedures for use in the National Airspace System.... The manual says, "Radio communications are a critical link in the A.T.C. system. The link can be a strong bond between pilot and controller--or it can be broken with surprising speed and disastrous results."

The accident report after the San Diego crash offers this: "The evidence indicates that there may be a communications gap between pilots and controllers as to the proper use of the A.T.C. system. The A.T.C. controllers are responsible for, and are required to apply, the procedures contained in handbook 7110.65A in their control of traffic. Despite the fact that the successful use of these procedures requires a mutual understanding on the parts of pilots and controllers of the other's responsibilities, pilots are not required to read handbook 7110.65A."

If the book that explains the procedures is not being read by pilots, the controllers might just as well talk to the stewardesses. Then at least the pilots would be free to look out the window. The fact that the NTSB cautiously allows that "the evidence indicates that there may be a communications gap" is equally frustrating to anyone interested in air safety. If a mid-air collision--at the time the largest crash in U.S. history--is not clear evidence that there is a communications gap, what is?

That is especially true if you realize exactly what took place that day. The controller issued a traffic advisory, telling the PSA captain where the Cessna was. There is evidence that the controller gave the wrong position for the Cessna in question. Other planes were in the area. We'll never know what the captain saw, but he acknowledged seeing something, which the controller interpreted to mean that the conflict had been resolved. Subsequent advisories were given to PSA 182, at least one of which may also have been wrong.

The airspace in which PSA 182 was operating contains a restriction that jets be kept above 4000 feet. The controller allowed flight 182 to descend below that (impact occurred between 2000 and 3000 feet). The controller had a "conflict alert system" that, by computer, projects the paths of aircraft and lets the controller know when they might come dangerously close to one another. There is a blaring horn, or klaxon, and a visual signal: The aircraft identifications flash off and on on the radar screen.

In the case of PSA 182, someone must have turned down the volume on the warning horn, because the tape recordings from the control positions did not pick up the sound of any klaxon. As for the blinking "data block," either the controller didn't see it or he ignored it. During legal action resulting from the crash, the attorneys representing families of victims were closing in on that area of investigation. The following control-tower action took place just prior to the collision. The sequence of events is in Greenwich mean time. The quotes are transmissions of the air-traffic controller.

1600:31--"Eleven Golf [the Cessna], traffic at six o'clock at two miles eastbound, PSA jet inbound, has you in sight."

1601:01--"Hey, what's this? I see Spike has some competition in the facility."

1601:28--The conflict-alert system went off, according to records, but no sound was heard on tape.

1601:47--"Eleven Golf, traffic in your vicinity."

And then the crash occurred.

The reference to Spike was explained by the controller himself during a sworn deposition. "Spike is a fellow controller that I work with. He kind of buddies up to the higher-ups in the facility, and we have a new fella that came into the facility that moves right along the same tracks." The attorneys claim that in order to see that "new fella" from his scope, the controller would have to have been turned around in his seat. The controller has denied taking his eyes from his scope.

During the NTSB hearings, this question was put to the controller: "Prior to the 25th of September, had you ever personally given a conflict alert, an alert warning?" That is, had he ever told the pilots of the aircraft involved in the alert?

Answer: "No, I've never given a warning." At the time, the controller had been at that facility five and a half years.

Before the attorneys were allowed to present their evidence in court, however, the FAA lawyer admitted liability, which kept the case from progressing any further. The NTSB report, which insiders have called a whitewash, blithely states, "The conflict-alert procedures in effect at the time of the accident did not require that the controller warn the pilots of the aircraft involved in the conflict situation." One might reasonably ask--even if the controller knew what was going on--what the expensive warning system is for if everyone using it may simply elect not to issue the warnings, as this controller admits doing as long as he's been using it.

While all that was taking place, the copilot was asking, "Are we clear of that Cessna?"

The flight engineer said, "Supposed to be."

The captain said, "I guess," and everyone thought that was real funny. They all laughed.

The pilot in the jump seat said, "I hope." And about 20 seconds later, they found that Cessna.

Both planes were under radar surveillance. The NTSB again: "The capability existed to provide... separation," but "Stage II terminal service does not require that either lateral or vertical separation minima be applied between I.F.R. and participating V.F.R. [visual flight rules] aircraft." In other words, keeping those two planes from running into each other was not required, so it wasn't done. The list of steps that could have been taken is long and frightening, The end of the cockpit voice recording is not long, merely frightening:

Captain: What have we got here?

Copilot: It's bad.

Captain: Huh?

Copilot: We're hit, man, we are hit.

Captain: Tower, we're going down, this is PSA.

Tower: OK, we'll call the equipment.

Unidentified Voice: Whoo!

(Sound of stall warning and expletives)

Captain: This is it, baby.

Unidentified Voice: Bob....

Captain: Brace yourself.

Voice: Hey, baby....

Voice: Ma, I love ya.

(Impact)

In all, 144 people were killed, including seven on the ground. Twenty-two homes were destroyed.

•

I sat and talked with a retired high-level FAA official about those and other crashes, the baffling waste of lives, the sometimes astonishing inattention and blundering. "It's the reason I quit the FAA," he said. "My heart couldn't take it anymore. If I'd thought I could have done any good, I'd have stayed with the agency." But 40 years of trying had done nothing but give him a bad heart.

Most air crashes are a collaborative effort. Big jets are generally well-made machines; pilots and air-traffic controllers are generally competent. So a lot of people have to cooperate to bring down an airliner. In the PSA crash, for example, four men in the 727 cockpit collaborated in not seeing the Cessna, the controllers offered assistance by issuing imprecise or inaccurate advisories and by not seeing or not issuing the conflict alert, the FAA helped by not making the alert system a useful tool and by other quirks of rule making--such as not requiring pilots to read the book that tells how to talk to a controller.

The men in the Cessna may even have helped a little by not asking themselves, "If I am being issued as traffic to a 727, just exactly where is that big old jet now?" Because for the Cessna, a collision with the 727's wake could have been as disastrous as the actual collision with the plane. Everyone lent a hand. As Gerald Sterns, a top aviation attorney, put it, "They must've sat up nights to do this one."

Another joint-effort crash occurred on March 10, 1979, and it is a good illustration. Swift Aire Lines flight 235 from Los Angeles International (LAX) to Santa Maria was a twin-engine prop plane. It went down in Santa Monica Bay because the crew, having lost one engine (no big problem there), inadvertently shut down the sole remaining engine.

How, you may well ask, could anyone possibly do that? The same question was asked by the NTSB, the air carrier, the surviving families, the insurance companies, the aircraft manufacturer, the FAA and a lot of reporters. And, furthermore, having done something so stupid, why didn't the crew just restart the good engine?

"The engine-restart procedures contained in the aircraft operating manual did not contain sufficient information to effect a restart"--even though "there was enough altitude and time available for the crew to get a restart." That's from the NTSB. The manual that does not tell how to restart an engine in flight and the flight crew using it were both certified by the FAA as safe and reliable. The manufacturer wrote it, the FAA read it and said it was fine, the airline bought the plane, learned to use if, trained its crews: the crew accepted it and flew it into the bay.

The FAA and the manufacturer (Nord) can protest all they want, but they cannot escape the fact that they produced a flight manual that was, to put it politely, inadequate. Pilots, too, can protest, but the Swift Aire crew demonstrably did not know how to operate that plane. Airline executives will further protest that Swift Aire Lines is not a scheduled air carrier, it is a commuter airline. But passengers find those distinctions difficult to appreciate when the airline takes their money and flies them into the ocean.

The manual for that plane is not unique. If you lose all engines in a large jet, for example, how far can the plane glide? It's not in the manual. Manufacturers say that it's not important, you're going to go down then and there anyway. They also say that the chances of losing all engines are infinitesimal. That's probably true--until it happens. Maybe if the pilot of that Southern Airways DC-9 had known he could glide 32 miles, he would have gone the 14 miles to the runway instead of turning away from it. A National flight 150 miles out over the Atlantic (with no life rafts, incidentally), on its way from Ft. Lauderdale to Newark, lost all three engines. The engineer made a procedural error in the process of fuel transfer that starved the engines. Luckily, they got a restart. United Airlines had reported this type of problem to its pilots but not to other airlines. And a DC-8 belonging to United just plain ran out of fuel one day near Portland, Oregon. The crew had not been told about a built-in fuel-gauge error, even though United knew it existed. Once again, statistics do not provide much comfort in those circumstances.

Another manual that is--or was--deficient is that of American Airlines for its DC-lOs. In the crash of flight 191, the NTSB says that the captain did exactly what the manual told him to do--and it resulted in the loss of the aircraft. There is a great deal of controversy about whether or not any right-minded pilot would follow the instructions in that manual, and those instructions were recognized--after the crash--to be so badly deficient that the manual was changed. A more detailed discussion of that crash will follow, but the fact remains that the manual contained instructions that could be fatal. Whether or not the captain followed them is a matter of debate. He is not available for comment.

•

Japan Airlines flew a DC-8 out of Anchorage, Alaska, on January 13, 1977. It reached an altitude of 160 feet before stalling and crashing. The captain was drunk. He was observed in a taxicab on the way to the airport exhibiting "mental confusion, dizziness, impaired balance, muscular incoordination, staggering gait and slurred speech." Once behind the wheel of that DC-8 (how he got that far is anyone's guess), he became lost on the surface of the airport and taxied to the wrong runway, where he reported ready for take-off. When the tower finally shepherded him to the correct runway, he flew the plane out to an almost immediate stall and crashed. The other two crew members did not challenge him, though it's difficult to believe that his condition wasn't obvious. And if you're thinking it was just some undisciplined Japanese pilot who didn't know how things are run in this country, the captain's name was Hugh L. Marsh.

United Airlines sent up a DC-8 on December 18, 1977, from San Francisco bound for Chicago, with an intermediate stop in Salt Lake City. While holding for the approach to Salt Lake, it was flown at 7200 feet into a mountain that was 7665 feet high. The results of two toxicological tests, verified by the Armed Forces Institute of Pathology, revealed that "the second officer would have to have had the equivalent of seven to eight ounces of 80-proof alcohol in his body when he left the hotel to report for duty...the degree of impairment, if any, of the second officer's physiological and mental faculties could not be determined." Just for the record, I would suggest that investigators interested in determining the degree of impairment caused by seven or eight ounces of 80-proof alcohol should try knocking back seven or eight shots of bourbon and then seeing if they can leave the bar, let alone the hotel.

Although the official and probable cause of the crash was the failure of the crew and the controller to communicate effectively, the finding is nonetheless interesting, since Federal law specifically states that at least eight hours must elapse between the consumption of any alcoholic beverage and the beginning of a flight. Most airlines have their own rules as well, some extending this to 24 hours. If United Airlines cannot assure its passengers that a flight-deck crew member is sober, who can? Certainly not the FAA. Probably no one. I called United to discuss air safety, but before I could even begin, I was told by spokesman Marc Michaelson: "Frankly, no one here wants to talk to you." That's understandable, given the facts.

There isn't a lot of information on airline-pilot alcoholism, though Dr. J. Robert Dille, the head of the Civil Aeromedical Institute (CAMI), says, "There's no reason to think the national figure doesn't apply to airline pilots." That means six to ten percent of airline pilots could suffer from this disease. "We don't have access to airline pilots," says Dille. But he points out that a large number of formerly alcoholic airline pilots are flying regularly now, having been rehabilitated. They fly under a special exemption that requires regular checkups and careful monitoring. "Better to have them out in the open and flying than go undetected, continue drinking and flying."

I interviewed a senior-level American Airlines captain, who confirmed that. "We've had a number of cases where we've actually had to take over from a guy and move crews up to cover him," he says. "The guy who is the real problem is the one who drinks on the QT and no one detects it for a long time--drinks actually on the airplane sometimes. We've had cases where the flight attendants can't allow the captain to walk past the open containers of vodka and gin, and so on. They watch him carefully when he goes into the washroom. We have terminated people like that. We've lost some."

And while the airlines understandably remain silent on the subject, CAMI and the FAA's Office of Aviation Medicine discuss it openly. "It's treated like any disease," says Dille. One of the problems CAMI faces is that no check exists that will show who is and who isn't alcoholic. "The best test we have today for alcoholism," says Dille, "is that if a person has had two arrests for D.W.I. [driving while intoxicated], there's a good chance he is an alcoholic. This test has proved to be about 73 percent accurate." Since neither CAMI nor the Federal Air Surgeon has access to police records, all alcohol rehabilitation must be done on a voluntary basis.

An article by Barton Pakull, chief psychiatrist for the Office of Aviation Medicine, says that the situation has improved greatly over the past ten years, when no pilot would admit having an alcohol problem for fear of being permanently grounded. Now, he writes, "Over 250 airline pilots have been returned to flight duties in the past three years." Until fairly recently, any alcoholic pilot wishing to return to duty would have to abstain from drinking for roughly two to five years. "It is now possible," writes Pakull, "for a commercial-airline pilot, sponsored by an appropriate medical department, to be considered for an exemption and return to duty within three months after completion of an initial intensive rehabilitation program."

This information is not meant to alarm passengers or exaggerate the problem. Programs such as those mentioned above are a definite step toward improved safety. But as an air-transport pilot says, "People don't seem to realize that pilots are just people. Some of these airline captains don't even like to fly. They've just gotten into it because of the military or the money or some other reason. And it's too late to get out."

One stunt pilot tells me, "You know, people don't realize how many pilots are actually scared of flying. They don't like being in airplanes."

Yet Tom Wolfe, in his best-selling book The Right Stuff, talks of that ineffable quality of certain flying men that is at once thrilling, frightening, admirable and confidence inspiring. Most passengers assume it is with that stuff they are flying when they go anywhere at all, that the right stuff is a prerequisite. That, of course, is why Frank Borman, the former astronaut, makes a perfect head for Eastern--to project that image and further that myth.

And it is a myth. Promoting airline pilots as superhuman is like promoting an image that equates railroad engineers with Wally Dallenbach and Johnny Rutherford. That is not to say that airline pilots are not good pilots. You'll find that if push came to shove, a lot of them could put down a 747 with ten slick tires on a rain-swept skating rink and not kill anybody.

But there's nothing in the rules, nothing mandatory in the training, that guarantees that level of skill. I know two captains who do this: hang a handkerchief over the copilot's instrument panel and make him land the plane. "These guys complain when I do that," one says. "They say, 'We're supposed to use all these instruments for landings.' But the guy who wrote those rules never had to land a big jet with his electrical system gone to hell." These captains are working to breed some of the right stuff in their copilots. Sadly, they are violating the rules to do it.

This may draw a storm of protest from some pilots, but this article is not written for them. This is for the passenger, for whom it may be valuable--perhaps vital--to realize that there is more to flying than chicken Kiev and some 60 Minutes reruns. Most passengers, if they genuinely understood flying, might never get on a plane. In fact, 37 percent of the American population will not get on a plane. Another 11 percent fly only when they have to--"Like, to a funeral," says leading aviation attorney John Kennelly.

"Passengers must be subtly coaxed onto airplanes," says the Aviation Safety Institute of Worthington, Ohio, the only truly independent air-safety organization in the United States today. Why do you suppose the oxygen masks are hidden in the ceiling of the airplane, instead of hung by your seats where you can reach them, as they are by the pilots' seats? No one wants to see them. Why do only the crews have shoulder harnesses? No one wants to think about the possible need for them. And that is at the heart of numerous safety problems. Because until passengers demand a more rational airline industry, they won't get one.

•

"Punchin' and jammin'," they call it, "movin' iron," punching them out and jamming them in. Air-traffic control is an interesting profession, especially if (as one controller told me) "you happen to be strapping 200 tons of aluminum to your fanny."

They have their own language, their own style, and it takes a while just to talk with them. It may take years to drink with them--at a recent gathering of "the brothers," as they like to call themselves, they told the hotel to double the normal liquor order. The hotel management thought that was merely a macho gesture. But the draft beer ran out in the first hour. The liquor ran out in the second. One controller leaned on the bar, waiting for the delivery truck they'd been promised, and told the bartender: "This is really a nice place for a bar. You ought to consider opening one here."

One of the first things you may notice in a big, hot, action-control-tower cab is the light rock-'n'-roll music playing from a radio or a tape deck. It's not turned up loud, but it's there, almost like a reminder ("When everything goes wrong, you have to go on, and do it or die..."). On top of that, every person there is whistling a different tune or humming something, working his position and tapping a pencil or his fingers or his foot (or both feet); the entire cab sounds like a quiet, metallic insect colony backed up by the Atlanta Rhythm Section or the Eagles or Linda Ronstadt. The immediate impression is that something is about to happen, but, of course, you don't know what it is. It makes you feel that if--God forbid--some fool should drop a glass, you'd have six or eight coronaries on your hands. During my recent visit to Opa-Locka, Florida--one of the country's busiest general-aviation airports--an outbound plane reported that it was returning due to rough engine operation and one of the controllers next to the local position had her hand on the red phone, her eyes wide, the white showing above and below the irises. It was a few seconds of the most intense anticipation and tension I have ever seen. Yet the humor--the language--is always there.

"Eastern sebendy-six, you'll be number two behind the condominium," said the controller at Miami International. Condominium is just an affectionate word for the Boeing 747, also known as the hog or the concrete eater. At Los Angeles, when an international 747 leaves, they call it a sand-dune departure--because the plane is so low when it clears the end of the runway that its wake kicks up a sandstorm. For this largest of the world's passenger aircraft, there just aren't any adequate runways. The 747 can gross out at something like 400 tons.

You can sit up in the O'Hare Tower in Chicago and watch them count off the Honolulu bomber, as they call United flight 111, a 747 that is generally well loaded with humans and cargo when it leaves every day for Los Angeles and Hawaii. Runway 32 Left has seven numbered exits that intersect it, and it's not nearly long enough for the Honolulu bomber. "Cleared for the take-off," the controller said and then laughed, turning to the others in the tower. "I say she'll take seven to VR." meaning the point at which the nose-gear lifts will be at the last taxiway, which is just about the end of the runway. The main gear will still be on the ground.

"It was seb'm yesterday, wudd'n it?" another controller asked.

And then they counted in unison, all eyes on the bomber, as it rolled out, faster and faster, eating up the concrete. When it finally lifted its nose wheel at the seventh intersection, they all cheered. They joke a lot, but they know: If the Honolulu bomber ever had to stop, why, good luck. "It could ruin your entire day," a controller told me, and then smiled thinly and reached for the antacid tabs--a family-size box left out on the counter in so many towers.

Controller locutions: Getting drunk is called "going I.F.R." At O'Hare, where small aircraft are not particularly welcome, an entire section is reserved for them where the strips containing identifications and flight plans are stacked. It says AIR TAXI--FLIB CITY. FLIB is an acronym in a world fraught with acronyms. It means fucking little itinerant bastards.

And the instructions the controllers give to pilots are sometimes more inventive than the FAA--their employer-- would like. Alter clearing an Avianca DC-8 (a plane easily recognized in the air by its four thick, black smoke trails) for take-off from Miami, the controller watched him sit and do nothing. He cleared him again and still nothing happened. Finally, he shouted into his mike, "A-vi-an-ka, you gonna show me some smoke, or whut?"

The humor can be macabre, too: A pilot who goes down with his plane is called dead right if it isn't his fault; and the decorative longitudinal stripe around some airplanes is referred to as the water line. (The pilots aren't without their humor, of course. Asked if he could descend to 12,000 feet in the next 20 miles, a British captain at 39,000 feet responded laconically, "Yes, sir, but I'm afraid I can't bring the aircraft with me.")

But this image projected by the professional air-traffic controller and his union (PATCO) is just a little too delicious: Those crazy guys who run the tower, real pros, real crazy. PATCO even put out a manual for its members to show them how to manipulate the press. When the humor falls away, though, when the jokes get old, when they shelve the hype manual and sit down for a serious talk, being around a group of air-traffic controllers can be like hanging out with the P.L.O. on planning day. Saying controllers are militant is like saying Idi Amin had a short temper. These people, once lowly radio operators, have turned modern aviation into a bitter war. Nobody is winning and the real issues lie buried and depend largely on where you go for information. The only participant with no say in the matter is the passenger, who pays controller salaries, pilot salaries and FAA management salaries--and who, for all his contribution to their livelihoods, sometimes gets cut down in the cross fire.

There are many points of view, but the three most apparent are those of PATCO, the pilots and the FAA bosses. To PATCO, pilots are ungrateful prima donnas, chimpanzees in space capsules pushing a few buttons and getting 100 grand a year. The FAA is Big Brother personified, an inhumane, repressive obstacle to safety and harmony.

To the pilots, the PATCO brothers are civil servants, no more, no less. They are sitting safely on the ground. "We're the first to the scene of the accident," pilots are fond of saying. "No controller's ever died falling off his stool. And they don't do such a hot job, either," a pilot may tell you. "Let me take you up on instruments and watch them lose us." And the FAA? It's not concerned with safety. It cares about politics, overcomplicated rule making and profitable air carriers.

The FAA claims its goal is safety--but that it's caught in the middle between the pilots and the controllers. We're just another agency asking for a handout from Congress, goes the FAA line; if we mandated every safety improvement they asked for, sure, it would be a little safer. But it would also cost $500 to go from Miami to Palm Beach, and we'd still have an occasional crash (which is all we have now, anyway).

This internecine war certainly isn't promoting safety. When I went to Miami to visit the air-traffic-control facilities there, my presence caused a lot of problems. A controller had invited me--and the FAA management has strict dibs on reporters. When the FAA learned that I was there, it mounted a massive handholding campaign to make sure I got the official tour, and several controllers were ordered not to talk to me.

This, naturally, made the controllers more eager to talk, which led to several secret late-night meetings with some PATCO members. They said they wanted to alert me to some safety problems, which they did. But their militant attitude was a bit disconcerting ("How'd you like to be up there in a plane when I decide to ram a screw driver through something?" I was asked by a technical man PATCO had invited to the meeting). The phrase criminal negligence was tossed about in reference to FAA inaction in the face of serious equipment faults.

On one day, Carlisle Cook, deputy chief of the Miami Air Route Traffic Control Center (A.R.T.C.C., also called Miami Center), took me on the grand tour of the facility. He showed me the great computer room, an acre of equipment PATCO told me was so outmoded it could not be relied upon. Cook pooh-poohed this with a smile, saying they were replacing it. "Wanna rent some space?" he laughed, indicating that the newer, smaller computers would need only a corner of this room. I was convinced PATCO had exaggerated the problem.

That night I told the PATCO members what Cook had said. "You go back there and ask him what date those computers are being replaced," one technician told me. "And come back here and tell me the brand and model of equipment we're getting."

Cook hithered and thithered a lot when I asked the question, said something about maybe in 1987 they might have something and admitted that they had no idea what kind of equipment they were going to put in. In fact, I asked FAA Administrator Langhorne Bond the same question, and he said that computer companies were even reluctant to bid on the design of these computers-- a design that hasn't even been decided yet. The companies are apparently worried about liability if somebody gets killed. Bond said that by 1985 something might be worked out and by 1990 they hoped to have some of the new systems running. He admitted that revamping the A.T.C. computer system was a monumental project. Avintion Week said in its November 26, 1979, issue, "FAA's schedule does not call for introduction of a new-generation computer system until the late 1980s." So much for Cook's real-estate business.

Considering how ragged the present system has become, it's no wonder some people in the FAA are trying to keep controllers from talking to the press. The equipment does fail--often enough to merit attention. To understand what this means, you have to understand something about how it's all supposed to work.

When you are flying across country, your pilot is in contact with an Air Route Traffic Control Center. A controller (and there may be as many as 120 in one center) watches a scope while talking to the pilot. The plane is displayed on the scope as a data block--numbers and letters describing who he is, his altitude, speed, and so on. The data is generated by a computer. But when the computer fails, the controller is left with only a mark or smear on the screen, indicating that something is there. In such cases, the controller has to get out small plastic chips called shrimp boats and write on them with a grease pencil, reproducing the data he has lost. Then he pushes them around with his fingers. There is nothing inherently dangerous about this method of controlling traffic, if you are trained to do it and if you start out with the airplanes separated for that kind of control and if you don't have too many aircraft to control. But when you change back and forth from one method to the other, and have to do so unexpectedly, you risk losing airplanes. Things can stack up very quickly and get out of hand, as they have numerous times in the past year.

The day I arrived in Florida, two airliners nearly collided over the East Coast just after a computer failure. The Aviation Safety Institute reported the near miss like this: "At about seven P.M. on October 31, 1979, the radar...computer system went out of service at the Washington Center...and lost radar contact with PALM 721 [Air Florida flight 721, a Boeing 737]. A few minutes later...radio contact was established with PALM 721 [and he] quickly stated that he just had to take a sharp turn to avoid collision with a southbound wide-body L-1011 [Delta Airlines flight 1061]." The controller was blamed, but, as always, there is more to it than that.

At Miami Center, FAA officials told me that every controller is assigned a midnight shift. That's a time when the computer is intentionally taken off line for service. Therefore, three FAA officials insisted, the controllers get regular, continuing, frequent practice in the shrimp-boat system of traffic control. All that manure I'd been hearing about how tough and how dangerous things got during a computer failure was (I was assured) just plain wrong. That night, when I reported this to the controllers, they laughed. It turns out they trade off those "mid shifts" to the hard-core handful of controllers who like to work at night. "I haven't worked one in over four years," a top-level controller told me.

There are other equipment failures, too, and no amount of training can help with them. At Miami Tower, logs are kept, and the ones we obtained show that the computer for the tower (a system separate from the center, designed for a different function) also fails. In addition, even the basic radio frequencies fail. When that happens, the pilot and the tower cannot talk to each other, which can make the approach to landing a thrilling experience.

March 25, 1979, from the Miami Tower log. The regular frequency for directing outbound traffic from Fort Lauderdale (controlled by Miami) failed. The stand-by was used. It failed less than an hour later. The log is quite clear about the consequences of such failures: "When [it] failed, the controller had three departing [aircraft] on radar vectors with whom he could not communicate." The statement of the controller says that Eastern Airlines flight 886 "passed one and a half [nautical miles] in front of" a private plane. "Both [aircraft] should have been level at 5000."

Three minutes later, the Eastern 727 came within a mile of another aircraft. To give you an idea of how close that is, two jets, converging at normal cruising speed, cover a mile in about three seconds.

Numerous log entries show the same pattern of problems: The computer or frequencies fail, come back on, fail again. Pilots are sometimes reluctant to report near mid-air collisions because it might reflect badly on them, and the controllers are not allowed to ask a pilot if he wants to report an incident. So when those failures occur, when the controller is left staring at a blank scope and hollering into a dead microphone, all he can do is file what is called a U.C.R. (Unsatisfactory Condition Report).

From a ream of such reports, here is a typical selection. Summer of 1979: "I was working FLL [Lauderdale] arrival position.... I was working about seven to eight aircraft when this frequency became intermittent and unusable. This happened at a time when I had four aircraft about to turn final. [Eastern flight] 200 turned by [himself; a private (continued on page 284)Airline Safety(continued from page 278) plane] was already on the final about ten miles out when I first attempted to turn him off the final because of [Delta flight] 148 heading south...and not hearing my transmissions.... This is a constant and dangerous problem at Miami approach and should be taken care of immediately before we have a disaster around the Fort Lauderdale area."

Yet another controller report filed with the FAA: "Repeated failure of radios.... This is a level-five approach-control facility and we need radios to talk to airplane drivers before someone gets killed."

And: "We have been lucky so far."

Or: "Real danger to the flying public."

The very day I visited the Miami Center, after I had left, the computer went down again. There is considerable controversy about the magnitude of this problem. The FAA says it's getting better. The controllers say, next time you have to fly to south Florida, fly to Grand Rapids instead.

Lonnie D. Parrish, chief of the Air Traffic Division of the Southern Region of the FAA, wrote in a memo to the director of Air Traffic Service on November 28, 1978: "The mix of a high volume of both V.F.R. and I.F.R. traffic, compounded by increased speeds and sophistication of general-aviation aircraft, will raise the mid-air collision potential to a level that demands affirmative action by those of us responsible for air-traffic management. The automation capacity of Miami A.R.T.C.C. will be reached in the immediate future, with today's level of service, and will be totally inadequate for services we should be providing in the highly congested south Florida area."

In August 1979, the Aviation Safety Institute warned, "We see more reports each week of radar data processing (R.D.P.) failures...and many are catastrophic.... The FAA headquarters more readily admits that the 9020 computer systems are reaching their performance limits. The computer manufacturer, IBM, warned the FAA back in the late Sixties that the 9020 would not do the job of controlling the 1980 traffic volume."

I interviewed John Galipault, president of A.S.I., the day after he returned from testifying before Congress on R.D.P. problems. I asked what they were going to do. "Nothing," he said. "I should learn not to go to these things. They're going to wait for a big mid-air before they do anything." Wasn't San Diego a big mid-air? I asked. At the time, it was the biggest crash in U.S. history. "Not big enough," Galipault said. "Maybe when two 747s collide...." It may only be a matter of time, too: In November 1979, in San Diego alone, there were at least two near mid-air collisions involving large jets. One of them took place over a packed football stadium. And at Los Angeles International, the FAA is casually allowing operations that can only be characterized as suicidal. Planes take oft and land in opposite directions on parallel runways. It's something like an airborne, 200-mile-an-hour game of chicken. For that reason, the International Federation of Airline Pilots Associations has given LAX its Black Star rating, reserved for only the worst airports.

In a December 17, 1979, editorial, Aviation Week asked: "Why did the FAA fail to specify higher reliability for an air-traffic-control system in which thousands of passengers daily are placed at risk? Why does it take such an inordinately long time for the FAA to procure and introduce new technology equipment into the traffic-control system? The answer to the first question is that when the present computer system was bought by the FAA, absolute reliability... was costly.... The answer to the second question is controversial...lack of funding is not standing in the way of better air-traffic-control technology--at least for high-priority items."

The FAA may have legitimate complaints about controllers, who are not making life any easier for the flying public by maintaining a constant state of red alert. But, as the controllers would say, being dead right is being nowhere at all.

•

I am walking around on top of a Boeing 747, high above the concrete floor of the TWA overhaul base in Kansas City, Missouri. The metal-mesh scaffolding has a spring to it and a lurching, drunken sway that would give a hardened high-steel worker a taste of copper in his mouth. It is the twilight shift and I am getting my ear filled by TWA mechanics, who speak a dialect that is known in bars from Nova Scotia to Arizona.

"We had one ol' boy here, he's dummer 'n a box o' rocks."

"He was one dumb shit."

"An" that ain't no shit."

My host looks like a younger, heavier Slim Pickins and greets all the night-shift workers with malicious glee and an elevated middle finger. Everywhere we look there are 707s, 727s, L-lO1ls and 747s in various stages of disassembly. One room is set aside for paint stripping. They just spray the stripper on and scrape off the old paint job, just as you'd do with your woodwork. There are areas for hydraulics, for wire harnesses, oxygen bottles, explosively deployed emergency slides (a locked room with a lot of black-powder canisters in it), wheels, landing gear, brakes, tires; and there are places where they take the plane down to the bone, where a cockpit can look like an eye socket that has taken a direct hit from some monolithic mousehawk, where the cabin floor is just a few skinny beams ("You wanna watch where you're steppin', buddy").

An entire building is devoted to engine reworking and repairing, row on row of jet fans lined up like cocoons. The creep through the TWA labyrinths takes all night and the place is as busy as a bus terminal on Christmas Eve.

For several days, I spent time with these mechanics who repair and maintain airliners in noisy rooms of such dimensions that a glance at the ceiling can make your stomach pitch. I learned that they are the salt of the earth, these men. We hung out in the taverns of Kansas City, with Kenny Rogers singing out of jukeboxes, "Who picks up the pieces, every time two fools collide?" and I came to realize that I was up against another group of ethanol Blue Angels--a Precision Drinking Team.

We discussed economic considerations versus safety.

"It's gettin' butts in seats," one said.

"An' that ain't no shit."

We talked quality control.

"Them tars," one said of the Goodyear products we had seen, "gotta be tight."

"Tighter 'n a gnat's ass stretched acrost a rain barrel."

"That's tight."

"An' that ain't no shit."

We talked superior mechanical skills.

"We got one ol' boy here's smarter 'n shit."

"An' that ain't no shit."

"You don't argue nothin' with 'im, from crotch crickets to the Bible."

"An' that ain't no shit, neither."

I asked if the regulations were ever bent to keep the airplanes moving. "We never do anything by the book," my host said. "If we did wheels by the book, we'd build four a day. Right now we're buildin' 25 a day and still have plenty of fuck-off time."

He said that with the present system of wheel rework, they are perfectly safe. "Some of them bolts might not be torqued up just right, but there's 16 or so anyway. Ain't nothin' gonna happen."

"It's all redundancy, anyway," another said.

"We cut corners like a mother, but we do it safe. It isn't ever worth it to stick your neck out that far."

I asked if there were ever pressure from above to do work that was not right. They admitted that sometimes a part would come down that was not the part needed and they would be told to put it on the aircraft anyway. The mechanic, in such cases, would tell the superior to sign it off. That would shift the responsibility if anything went wrong. It was loud and clear: "I ain't signin' off nothin' if I don't know it's right."

It is commonly accepted that the men who repair and maintain airliners undergo intense training and are qualified by the Government to do that work. "I've got a hot flash for you," an FAA safety counselor told me. "Ninety percent of the airline mechanics are not A & Ps." A & P is the FAA designation for a qualified mechanic. It means airframe and power plant. It takes a lot of training--years, in fact. You have to pass rigorous testing, written, oral and practical. It costs a lot in both time and money.

But the law doesn't specify that repair of an airplane must be done by an A & P; it only specifies that the work must be signed for by an A & P, no matter who has actually done it. That signature indicates that the A & P has inspected and approved of the work. I asked the TWA mechanics how many of their men were A & Ps.

"A large percentage," one man told me. "Maybe as much as twenty-five percent."

They talked about how they had gotten their A & P tickets. First, you need a certificate from a certified aviation-maintenance school or 18 months' experience and must have a qualified person recommend you for the written test. Alter passing the written, you have up to two years before you must take your oral and your practical. My host said he went for his test at Johnson County Airport in Kansas City, where he crawled around on an airplane for two days before the examiner asked, "You realize if the FAA gave you this test you'd flunk?"

"Yes, sir, but so would half the FAA guys."

"That's right," the examiner said, and gave him his A & P license.

"All you need," said my host, "is common sense and to be a mechanic." Another TWA mechanic told of getting his license in Oklahoma City at a diploma mill. His oral consisted of six questions. The entire examination took about four hours.

I asked the mechanics to choose--if they could go on any airplane--how they would get from point A to point B. "If I had to go," one of them said, "I'd fly Delta." They all nodded solemn agreement.

They further agreed that they would fly any plane, as long as it were not a DC-10. "DC-10's an original piece of garbage," one of them said.

"Ail1 that ain't no shit."

Finally, they made it crystal-clear that under no circumstances would they fly Braniff International--they would sooner jump off the Statue of Liberty in pink tights. "We pull Braniff maintenance out in L.A.," they said, "and we know."

These mechanics weren't just blowing smoke. The FAA recently ordered Braniff to pay an unprecedented $1,500,000--in what the FAA calls a "civil penalty"--for improper maintenance and illegal practices. A comprehensive article in Aviation Week describes a situation that makes Braniff look like a Third World bus company. The FAA apparently warned Braniff time after time and the airline did nothing. After 39 extended over-water flights, Braniff was found to be carrying illegal life rafts, one with a hole in it. According to the FAA charges, the airline ignored cracked wings, engines showing strong vibrations, seats that could recline to block emergency exits, the installation of parts on one type of engine that were made for another type, and main cabin doors that didn't operate normally. One airplane alone was taken on 447 flights during almost a year in an unairworthy condition, says the Government. Braniff, according to the charges, wasn't even keeping proper records of maintenance.

A Braniff spokesman told me, when asked to respond to the charges, "We have not been fined: let's get the facts straight. We filed a detailed, thorough response--highly technical. The FAA is studying this response, which is several hundred pages long. The FAA has asked that neither party make any further comment on the matter while it is studying the matter." At this writing, the penalty has not been paid and the Government's case against Braniff is still pending. Langhorne Bond told me, "When we make charges like this, we give the carrier a chance to respond and if we're wrong, we're wrong. But we haven't been wrong yet."

•

The maintenance problems put me in mind of the NTSB hearings into the Chicago crash--because maintenance was the big question. The NTSB was trying to figure how a contemporary jetliner with only a handful of hours since a major overhaul could fall out of a clear blue sky. How a seven-ton engine-pylon assembly could just rip off. How, even with the lost engine, the plane could not be driven around to a safe landing, as other planes had been in the past. No one at the hearings seemed to have any idea of what was going on.

On a raised platform in the middle of the convention hall were cameras and producers and sound men from every network and local station. A table in front of the platform was jammed with reporters, as was another off to the side of the Safety Board, which was elevated on its own bench like the Supreme Court. You could tell that the truly heavy engineering types were in town. In the rest room outside the hearing room, among the crude obscenities scratched on the walls, one lone graffito offered, "Time is just nature's way of keeping everything from happening all at once."

The hearings were long and arduous. McDonnell Douglas pointed the finger at American Airlines, American pointed the finger at Douglas, the FAA pointed the finger at both and both pointed back. Chairman Driver tried ineffectually to keep the fingers away from the throats. Everyone except American and the Air Line Pilots Association (A.L.P.A.) pointed the finger at the pilot, Captain Walter Lux, though that was done in the most subtle ways imaginable. There was grumbling among the participants because the hearings were being held too soon after the crash. One A.L.P.A. member told me they had been given boxes of material and had only three days to master it. No one, it seemed, was up to speed.

At each recess, the pilots in the audience, who had come from nearby O'Hare out of curiosity, would stand in the hall and say things like, "Shit, that's not how it is." The reporters would put their heads together and try to decipher the technical language (TOGO mode, roll departure, sonic eddy current, gust loading, hard time limits, induced load, failure mode and effect analysis). If you hadn't done your homework, you could be lost inside half an hour and the spectacle went on for almost two weeks--it's no wonder the reports on it were mostly incomprehensible. Yet the basics were relatively simple, once you stripped away the jargon.

Douglas had designed a pylon that attached the engine to the wing of the DC-10, then sold a lot of DC-10s. The pylon had at least one fault. When it showed up, Douglas told its customers to fix it. The fault itself did not promise to be fatal. But the fact that it had to be fixed did. Engines are removed routinely from aircraft, but pylons are not. American Airlines devised a short-cut method for fixing the faulty Douglas pylon. Instead of removing the engine first and supporting it from above with a crane (as Douglas had suggested), American removed the engine and pylon together by shoving a forklift truck underneath and just dropping the whole assembly down.

Continental was doing the same thing. Continental dinged some pylon mounts, too, but happened to notice the mistake and fix it. Continental didn't tell American, though. And American dinged one of its pylons in the same way. "They were just bangin" on it with a big ol' hammer," Langhorne Bond told me. Since they had just serviced and reassembled the thing, they didn't check to see if it was destroyed, which it was.

And it flew 166 landings and 430 hours of service with passengers on board before the cracked part let loose, triggering a complex sequence of systems failures and resulting in the loss of the aircraft and all on board. That sequence of events dramatically pointed out numerous design deficiencies that had only been suspected before the crash of flight 191. But on that day, many things that had been waiting to go wrong with the DC-10 went wrong.

That left everyone with the sticky problem of why the plane couldn't continue to fly after losing an engine. There were only two choices: Either the DC-10 airplane could fly in that condition (and therefore other DC-10s could be allowed to continue flying their 400-odd daily missions, carrying 150,000 passengers) or it could not fly in that condition (which would mean admitting that the plane was a dog that had to be taken out of service). The choice was made: The plane was safe. It had to be safe. The industry couldn't afford for it not to be safe--because it would cost too much to take the DC-10 out of service.

That led to a confusing series of discussions about why (if the plane was, indeed, flyable) the crew of flight 191 hadn't flown it. And therefrom came a subtle line of argument that pinned the blame on the crew, even though no one was crude enough to come out with "pilot error" as a cause. In fact, in its final report, the NTSB went out of its way to say that the pilot was not to blame.

Simulator tests run in the "accident configuration" helped determine that the plane was flyable after the engine came off. The NTSB said the manual instructed the pilot to reduce his speed during that critical phase of flight after the loss of the engine. Therefore, he did reduce his speed and his left wing stopped flying. And the plane rolled over. A significant number of pilots who flew the accident configuration, however, were able to recover and fly away from the problem. The language was very roundabout, but the message continued to echo through it all: "Our pilots were able to fly the thing; what's the matter with your pilots?"

The way they got around actually blaming the pilot was to say that, since the electrical system was disabled, the crew would not have had the normal warnings another crew would have had--stall warning, for example. And the captain's instrument panel was out of service. The NTSB reasoning here is a bit fuzzy. A properly powered instrument panel displays a "command bar" indicating to the pilot the pitch he should fly (i.e., how far up or down he should put the plane's nose). The NTSB concluded: "The consistent 14-degree pitch attitude indicated that the flight-director command bars were being used for pitch attitude guidance and, since the captain's flight director was inoperative, confirmed the fact that the first officer was flying the aircraft." This, of course, is sheer supposition on the NTSB's part. But it sounds good and makes the pieces seem to fit neatly.

A former accident investigator with 30 years' experience said, "This is really a strange report."

Indeed, the report is odd, making sweeping assumptions and skipping over critical questions. Take, for example, the stall-warning device, which became inoperative on flight 191 when the electrical system was lost. The simulator tests were run with the device operating in some cases and the conclusion was that if the captain had had this warning, he would have lived. At the most fundamental level, the report fails to address the question of why a pilot with 22,000 hours would sit there and let his copilot do something that even a primary student of flying would hesitate to do--pull the nose up during a critical phase of flight when there existed the opportunity to put the nose down and keep the show on the road. There were no obstacles out beyond runway 32 Right at O'Hare. The day was clear: The crew could see for 15 miles. In addition, the captain would not have gotten a stall warning even if the equipment had been operating, because it was the tip portion of the wing that stalled and the warning device takes its information from the inboard portion, which was still flying at the point where the roll began. On the other hand, there were so many things going wrong at once that the crew was probably overwhelmed by the problems. "Their panel would have lit up like a Christmas tree," one airline captain told me at the hearings.

During the hearings, a witness was asked, "Would you please discuss the rationale of why the aircraft--why it would be considered safe to operate the aircraft with what appears to be a Russian roulette type of system?" And a few minutes later: "Could you discuss the rationale behind the certification of a system in which an engine loss causes you to lose the system which tells you that you lost the engine in the first place?" But participants who asked such questions were accused of browbeating witnesses and were silenced. And so the question was left hanging: How come this crew couldn't fly this nice flyable plane? And while it is certainly possible that they were merely incompetent, there are just too many unanswered questions left by the NTSB report.

An American Airlines captain told me, "If they think they've got a bunch of heroes up in the cockpit who are going to pull back their speed because the manual says so, they're crazy. They say Lux was going back to V2--bullshit! I can't believe anybody would be dumb enough to pull it back to 158 knots sitting out in the clear blue sky. That is simply not a satisfactory explanation." In fact, during the simulator tests, it was noted that one of the test pilots was asked to fly the way flight 191 flew and he couldn't do it. His pilot instincts would not let him pull the nose up. Because, as any pilot knows, that is an invitation to the harp farm.

•

All of which leaves open the distasteful possibility that the airplane was simply out of control during most of its 31-second flight, that the hydraulic system folded up on them and that there was no way to put the nose down. The subject is hotly debated as we go to press, but no one will ever know what happened on that flight. Not even the cockpit voice recording survived--well, the actual machine survived the crash, but the critical portions of the tape were never recorded, because the electrical system was so fouled up that the recorder lost power.

To get a more complete picture of the DC-10, I visited the Douglas plant where it is manufactured--in Long Beach, California. A further examination of the DC-10 will be included in part two of this article, along with discussions of airline economics and the philosophy of crashworthiness.

But while I was at the Douglas plant, I saw a curious thing. I was in a room so large that the new Douglas Super 80s and a number of DC-9s in various stages of construction seemed like toys tossed into corners and forgotten by children. Except that I could see the tiny men swarming over them and hear the ricochet of rivets popping like machine-gun fire throughout the room.

I stood at the electrical assembly line for the DC-10 and the DC-9. The man in charge was explaining how they create and then move the wire harnesses, and I couldn't shake the feeling that there was something awfully spooky about what I was seeing. There were computers to check the accuracy of the connections, all of which were made by hand, mostly by women, from the look of the line. I was staring at a virtual wall of white wire that was but one quarter of one section of an embryonic DC-10's nose ("One hundred miles of wire in the DC-10," I was told); all this white wire, coiling endlessly.

And then it hit me. If you look closely, you can see that thousands upon thousands of pieces of white wire are embedded in the rich, black mud that sticks to your shoes as you walk along. ...And I suddenly realized where all that wire had come from that I'd seen in that field in Chicago, that scarred patch of ground where American flight 191 had gone down.

I toured the rest of the wire-assembly area, but I didn't hear another word the man said.

"The crew isn't necessarily a trio of full-bird idiots; but they aren't steel-eyed superheroses, either."

"It puts one in mind of a man who jumps from a plane wearing a parachute--and neglects to pull the cord"

" 'Ninety percent of the airline mechanics are not A & Ps'--the FAA designation for qualified mechanics."

This is the first of two parts of PLAYBOY's investigation of airline safety. The conclusion will appear next month.