The saddest instruction I ever heard came from a news editor who was planning how to cover an air disaster. He ordered his reporters to go, not to the place where the plane had crashed, but to its destination. "That," he said, "is where the real victims are - the ones who are waiting. They are the story."

The drawn and agonised faces we saw yesterday of those who had lost friends and relatives in the Concorde tragedy suggest he may have been right. In the immediate aftermath of a plane crash the first question - why it happened - is almost always unanswerable. The suffering of those left behind is palpable.

Whatever secrets may be yielded by Concorde's black box flight recorders, establishing the cause of the engine explosion is likely to be more than usually complex. It will take the accident investigators back, not just to the last check-up or even the last risk assessment report, but to the very origins of the plane's design.

Of course, as the aircraft's defenders were insisting yesterday, one accident is not enough to write off a whole fleet. Concorde will continue to fly, and at its present rate, will survive at least until 2015. It remains a widely admired machine, the "flagship" of two major airlines, an object of curiosity whenever it takes to the air. But its first, terrible crash will raise again the nagging question of why, 25 years after it went into service, it remains the world's only supersonic airliner; why no carrier other than British Airways and Air France has felt it worthwhile to invest in it; why there is still no high-speed successor in sight.

While Boeing's 747 has spawned a generation of successful planes, Concorde has become the Neanderthal of aircraft design - a line leading nowhere; it may be a glorious creature, but it is ultimately an anachronism.

Similar questions about the viability of a plane were asked 26 years ago, when another airliner crashed outside Paris, amid scenes of far greater horror. On March 4, 1974, a Turkish Airlines DC10 came down in the forest of Ermenonville, killing 354 people, including the crew of 11. It was the first jumbo jet disaster, a catastrophe of such unimaginable proportions that it seemed to challenge the whole principle of mass transit by air.

Only gradually, as the cause of the accident was traced, did the truth finally emerge. A cargo door had fallen off, damaging the rudder-controls that ran underneath the floor of the cabin. A lengthy inquiry by the Insight team from The Sunday Times established that McDonnell Douglas, who had developed the DC10 as a scaled up version of the robust and well-tested DC9, had failed to build sufficient "redundancy" - or resilience - into its control systems.

In a determined effort to get its plane on to the market in advance of Boeing and Lockheed, the company had not developed systems that could cope with an unexpected emergency on this scale. Its frailties had nothing to do with the size of the plane, everything to do with the manufacturing philosophy.

Boeing, by contrast, had used the very size of its 747 to make it a safer plane. It had built in four separate control systems, made certain that its flaps were near the back rather than the front of the wings, placed its engines well apart and away from the fuel tanks. As a result, its safety record, based on the millions of miles it has flown, is among the best in the world, despite the fact that it requires far fewer check-ups than Concorde.

Those who point to Concorde's 25 years of incident-free flying should recall that, compared to the Boeing 747, it has scarcely taken off - at any time there are 1,200 Boeings in the skies, criss-crossing the world, compared to about three Concordes flying to New York and back. This is because the 747 and its successors were designed for the rapidly expanding mass transit market of the late 20th century, rather than to prove a point about Europe's technological virility in the 1960s.

Concorde was a triumph of design over commercial potential. Its origins lay with the brilliant aerodynamic theories of Dietrich Küchemann, who came to England after the last war. He and his colleagues, many of them from the little town of Göttingen in Germany, developed the swept-back wings and raked outline of the modern jet, which finally resolved the problem of the shock-waves that formed over the wings when a plane approached the speed of sound.

Küchemann is regarded as the man who made Concorde possible, but he was never happy about the way his design was adapted to produce a passenger jet. It was not just the massive investment - one estimate is that at current prices each Concorde cost the taxpayer £1 billion, the price of ten Jumbos - it was the structural compromises that had to be made.

Unlike most modern airliners, its four engines are set into the wings, next to the fuel tanks, rather than slung beneath them; it needs to achieve a land speed of 220mph to take off, compared to 165mph for most subsonic aircraft; it requires reheat to give it the extra thrust to take off. None of these are exactly safety features. Concorde is not so much a passenger jet as a military fighter posing as an airliner.

It is a grim irony that news of the disaster should break during the air show at Farnborough, where Küchemann did most of his work. The talk there this week was not about the next generation of supersonic planes. Instead it concerned the lumbering Airbus A3XX, a double-deck jumbo jet, which carries more passengers, more cheaply, and far more slowly than Concorde. On the other hand, it might just find a market.

Times Recommends

• Resist urge to throw voters red meat
• Slay sacred cows and be candid
• We cannot let these 100 deaths deflect us