Into the Wild Blue Yonder : THE FINAL CALL: Why Airline Disasters Will Continue to Happen, <i> By Stephen Barlay (Pantheon: $24.95; 457 pp.)</i>

Every airline crash is preventable in retrospect. The recent USAir-Sky West collision at LAX, which killed 34 people, might have been avoided in a half-dozen ways:

If only the controller hadn’t become mixed up about which commuter plane she had put on which runway; if only a third plane that looked exactly like the one she thought she was talking to had not been where she thought the other should be; if only the USAir pilot had sensed something amiss when his initial requests for landing clearance went unanswered; if only the Sky West pilot had called the tower to ask why he had been left sitting on an active runway for an unusual length of time.

On the other hand, the vast majority of airline flights are long strings of accidents that were prevented. Consider how many different things had to go wrong in order for the USAir-Sky West collision to occur. Major airline crashes caused by single, gross pilot blunders or catastrophic mechanical failures have become sufficiently rare that in the last decade one occurred only every two or three years in the United States. Over that span, about 175 people per year have died aboard U.S. airliners, compared to about 40,000 per year in U.S. automobiles, while about 2 million people per day have arrived safely on flights where nothing went wrong, except that passengers ate the meals.

So is air travel a safety-first industry, or a wild gamble in which you place your life in the hands of absurd twists of fate such as an air controller getting confused? According to “The Final Call,” it’s the latter. Stephen Barlay, an English writer, has assembled numerous examples of what seem like pilots and airline industry officials hell-bent on making idiotic errors, refusing to double-check the obvious, and failing to learn from experience. “Statistics . . . will not guarantee our right to survive airline flights,” Barlay writes. “Death by foreseeable and avoidable accidents could be seen as murders.”

But wait a minute. When crazed pilots and shortsighted air-industry officials take actions that result in accidents, don’t pilots and air-industry officials die too? And if the lunatics are running the asylum, why do overall safety measures show an almost uninterrupted curve of improvement during the jet age?

These are but two of several objections Barlay never addresses in “The Final Call.” Basically, the book boils down to endless repetitions of the banality that accidents are avoidable in retrospect. The author never presents a comprehensive thesis on the nature and significance of air crashes, other than that they are invariably a damn shame.

The book is so confusingly written as to border on unreadable. “The Final Call” jumps all over the map, from American to British to Third World air incidents, in a manner impossible to follow. Chilling narratives of crashes or near-crashes are presented in some detail, then the examples suddenly vanish from the text without any explanation of what went wrong with the plane or what happened to the passengers.

The author has a number of annoying stylistic ticks, among them setting off sentences as single paragraphs, frequent use of italics for no apparent reason (“Just like nonfatal incidents , even minor complaints against manufacturers must be taken seriously”), and worst, sentences that simply don’t make sense. Here is what Barlay has to say of the many accidents involving the DC10: “The accumulation of DC10 cases in this chapter is not meant to imply that the frequently suspected, much-maligned but rightly criticized aircraft is any less safe these days than any other big jets.” A prize to the reader who can decode that cryptography!

Barlay has written an intriguing first draft on airline safety, but instead of sending him back for revisions, his editors have shipped the manuscript directly to press. Not a heartening volume for those who fear that major publishers have stopped editing what they publish.

Nevertheless, in its scattershot way, “The Final Call” has virtues. Barlay provides many reasons to believe that airplane manufacturers and government agencies such as the Federal Aviation Administration (FAA) drag their feet on reforms that everyone knows full well to be inevitable, causing needless deaths in the interim. Lavatory smoke detectors, better tires (blow-outs during the takeoff roll are surprisingly dangerous), ground-proximity warning systems, floor-guide lighting and similar safety features now universally agreed to be essential were, Barlay shows, at early points opposed by people who ought to have known better.

Are there other safety improvements, possible today, that the airline industry will some day be blamed for moving on too slowly? Absolutely. Airborne collision-avoidance systems, airport wind-shear warning networks and improved ground-traffic radars, all of which might have prevented one or more recent accidents, are either being phased in at a leisurely pace or are hung up by bureaucratic details.

Anyone who has read the bone-chilling survivor accounts of what it was like inside USAir 1493 in the minutes after the runway collision knows that FAA procedures for certifying how quickly an airliner can be evacuated are ridiculously unrealistic. The drills usually are carried out by aircraft-manufacturer employees who have choreographed their movements in advance, and are staged on brand-new planes where all the emergency exits are working perfectly, which is almost never the case in the aftermath of an accident.

More, larger emergency exits with easier-opening mechanisms seem like a must for future airliners. For existing planes, the seat immediately adjacent to each exit ought to be removed. With most aircraft now flying less than full, it is, as Barlay would say, close to a crime to have people dying because they must climb over seats to reach the escape point. (Manufacturers don’t like emergency exits because they increase the weight of the plane.)

Next, how about smoke hoods under every seat? They were badly needed aboard USAir 1493; accident post-mortems invariably show that more people die as a result of smoke inhalation than by fire directly. I have been in a serious fire--in a building, not an airliner--and can attest that just a few gulps of thick, toxic smoke saps your strength with a speed that is difficult to believe.

And how about cabin sprinkler systems? Just a few hundred pounds of water, sprayed into the cabin in the key moments as the fire starts, might prevent burning aircraft from becoming instant crematoria. How about shoulder harnesses instead of lap belts? The pilots have them. For that matter, how about air bags?

The fact that many of the above safety improvements would have been vitally important on USAir 1493 leads to an important insight: Most recent crashes have been survivable. Of the 805 people on major airliners involved in accidents in U.S. airspace since 1987 (this excludes the Pan Am bombing), about 528, or 65%, lived. This compares to just 11% of passengers surviving comparable accidents in the previous eight years.

As the quality of airplane mechanical parts increases and pilot knowledge about flying continues to accumulate, it is quite possible that this will be seen not just as some statistical quirk but rather as proof that air accidents do not automatically mean death. In most recent air emergencies, pilots have had at least partial response from the aircraft right until the end, leading to controlled-crash impacts rather than catastrophic nose-down obliterations.

The traditional engineering attitude, however, has been that if there is a crash, your number is up. This outmoded line of reasoning should be abandoned for a new commitment to aircraft designs that protect passengers even when things are going badly. Ten years from now, smoke hoods, shoulder harnesses, better exits, cabin sprinklers and other new features will be required not only on all new airliners, but we will find it difficult to believe there ever was any argument about them.

Here is a place where Stephen Barlay is right as rain: Once any safety improvement becomes possible, it becomes inevitable, and delays inexcusable. If only “The Final Call” made this case in prose that readers could follow.