But now reporters came to the story with the lead fixed in their minds; they saw their job as proving what they already knew. They didn't want information so much as evidence of villainy. In this mode, they were openly skeptical of your point of view, since they assumed you were just being evasive. They proceeded from a presumption of universal guilt, in an atmosphere of muted hostility and suspicion. This new mode was intensely personal: they wanted to trip you up, to catch you in a small error, or in a foolish statement-or just a phrase that could be taken out of context and made to look silly or insensitive.

Because the focus was so personal, the reporters asked continuously for personal speculations. Do you think an event will be damaging? Do you think the company will suffer? Such speculation had been irrelevant to the earlier generation of reporters, who focused on the underlying events. Modem journalism was intensely subjective-"interpretive"- and speculation was its lifeblood. But she found it exhausting.

And Jack Rogers, she thought, was one of the better ones. The print reporters were all better. It was the television reporters you really had to watch out for. They were the really dangerous ones.

Crossing the plant, she fished her cell phone out of her purse, and called Marder. His assistant, Eileen, said he was in a meeting. "I just left Jack Rogers," Casey said. "I think he's planning a story that says we're shipping the wing to China, and there's trouble in the executive suite." "Uh-oh," Eileen said. "That's not good." "Edgarton better talk to him, and put it to rest." "Edgarton isn't doing any press," Eileen said. "John will be back at six o'clock. You want to talk to him then?" "I better, yes." "I'll put you down," Eileen said.

It looked like an aviation junkyard: old fuselages, tails, and wing sections littered the landscape, raised up on rusty scaffolding. But the air was filled with the steady hum of compressors, and heavy tubing ran to the airplane parts, like intravenous lines to a patient This was Proof Test, also known as Twist-and-Shout, the domain of the infamous Amos Peters.

Casey saw him off to the right, a hunched figure in shirtsleeves and baggy pants, bent over a readout stand, beneath an aft fuselage section of the Norton widebody.

"Amos," she called, waving as she walked over to him.

He turned, glanced at her. "Go away."

Amos was a legend at Norton. Reclusive and obstinate, he was nearly seventy, long past mandatory retirement age, yet he continued to work because he was vital to the company. His specialty was the arcane field of damage tolerance, or fatigue testing. And fatigue testing was of vastly greater importance than it had been ten years before.

Since deregulation, the carriers were flying aircraft longer than anybody ever expected. Three thousand aircraft in the domestic fleet were now more than twenty years old. That number would double in five years. Nobody really knew what would happen to all those aircraft as they continued to age.

Except Amos.

It was Amos who had been brought in by the NTSB as a consultant on the famous Aloha 737 accident, back in 1988. Aloha was an inter-island carrier in Hawaii. One of their airplanes was cruising at 24,000 feet when suddenly eighteen feet of the airplane's outer skin peeled off the fuselage, from the cabin door to the wing; the cabin decompressed, and a stewardess was sucked out and killed. Despite the explosive pressure loss, the plane managed to land safely at Maui, where it was scrapped on the spot.

The rest of Aloha's fleet was examined for corrosion and fatigue damage. Two more high-time 737s were scrapped, and a third underwent months of repairs. All three had extensive skin cracks and other corrosion damage. When the FAA issued an Airworthiness Directive mandating inspections of the rest of the 737 fleet, forty-nine more planes, operated by eighteen different carriers, were found to have extensive cracking.

Industry observers were perplexed by the accident, because Boeing, Aloha, and the FAA were supposedly all watching the carrier's 737 fleet. Corrosion cracking was a known problem on some early-production 737s; Boeing had already warned Aloha that the salty, humid Hawaiian climate was a "severe" corrosion environment.

Afterward, the investigation found multiple causes for the accident. It turned out that Aloha, making short hops between islands, was accumulating flight cycles of takeoff and landing at a faster rate than maintenance was scheduled to handle. This stress, combined with corrosion from ocean air, produced a series of small cracks in the aircraft skin. These were unnoticed by Aloha, because they were short of trained personnel. The FAA didn't catch them because they were overworked and understaffed. The FAA's principal maintenance inspector in Honolulu supervised nine carriers and seven repair stations around the Pacific, from China to Singapore to the Philippines. Eventually, a flight occurred in which the cracks extended and the structure failed.

Following the incident, Aloha, Boeing, and the FAA formed a circular firing squad. The undetected structural damage in Aloha's fleet was variously attributed to poor management, poor maintenance, poor FAA inspection, poor engineering. Accusations ricocheted back and forth for years afterward.

But the Aloha flight had also focused industry attention on the problem of aging aircraft, and it had made Amos famous within Norton. He'd convinced management to begin buying more old aircraft, turning wings and fuselages into proof test articles. Day after day, his test fixtures applied repetitive pressures to aging aircraft, stressing them to simulate takeoffs and landings, wind shear and turbulence, so Amos could study how and where they cracked.

"Amos," she said, coming up to him, "it's me. Casey Singleton."

He blinked myopically. "Oh. Casey. Didn't recognize you." He squinted at her. "Doctor gave me a new prescription… Oh. Huh. How are you?" He gestured for her to walk with him, and he headed toward a small building a few yards away.

No one at Norton could understand how Casey was able to get along with Amos, but they were neighbors; he lived alone with his pug dog, and she had taken to cooking him a meal every month or so. In return, Amos regaled her with stories of aircraft accidents he'd worked on, going back to the first BO AC Comet crashes in the 1950s. Amos had an encyclopedic knowledge of airplanes. She had learned a tremendous amount from him, and he had become a sort of adviser to her.

"Didn't I see you the other morning?" he said.

"Yes. With my daughter."

"Thought so. Want coffee?" He opened the door to a shed, and she smelted the sharp odor of burned grounds. His coffee was always terrible.

"Sounds great, Amos," she said.

He poured her a cup. "Hope black is okay. Ran out of that creamer stuff."

"Black is fine, Amos." He hadn't had creamer for a year.

Amos poured a cup for himself in a stained mug, and waved her to a battered chair, facing his desk. The desk was piled high with thick reports. FAA/NASA International Symposium on Advanced Structural Integrity. Airframe Durability and Damage Tolerance. Thermographic Inspection Techniques. Corrosion Control and Structures Technology.

He put his feet up on the desk, cleared a path through the journals, so he could see her. "I tell you, Casey. It's tedious working with these old hulks. I long for the day when we have another T2 article in here."

"T2?" she said.

"Of course you wouldn't know," Amos said. "You've been here five years, and we haven't made a new model aircraft in all that time. But when there's a new aircraft, the first one off the line is called Tl. Test Article 1. It goes to Static Test-we put it on the test bed and shake it to pieces. Find out where the weaknesses are. The second plane off the line is T2. It's used for fatigue testing-a more difficult problem. Over time, metal loses tensile strength, gets brittle. So we take T2, put it in a jig, and accelerate fatigue testing. Day after day, year after year, we simulate takeoffs and landings. Norton's policy is we fatigue test to more than twice the design life of the aircraft. If the engineers design an aircraft for a twenty-year life span- say, fifty thousand hours and twenty thousand cycles-we'll do more than twice that in the pit, before we ever deliver to a customer. We know the planes will stand up. How's your coffee?"