The world as it is, we never lack for horrific images: A car-bomb in the high street; a pile of adolescent bodies stacked by the locked fire escape of a burnt-out disco; the aftermath of hurricanes, earthquakes and floods.
But, watching the "live" pictures broadcast from Gonesse last Tuesday evening, it seemed to me - not for the first time - that there is something especially compelling about the scene of an aircraft's destruction.
An aircraft rising in flight - and none more so than Concorde - is a potent symbol of the control modern man has achieved over the forces of the physical universe. So, a wrecked aircraft scattered in some field is an equally potent symbol of the fragility of that control and the drastic consequences of its failure.
Almost 30 years ago, on a crystal-bright Sunday afternoon, I flew to Paris to attend the scene of what was then the world's worst airline crash: a Turkish Airlines DC-10 that took off from Orly Airport in perfect conditions and plunged uncontrollably into a forest.
It came down just north of Paris, transforming what had been a pleasant shallow grove into how I imagine the Somme looked after the worst battle. The impact had been extremely violent and the 346 people on board had been reduced to body parts as unrecognisable as the fuselage that had contained them.
Except, that is, for a pair of clasped hands that had withstood disintegration - an image that still haunts me and, I imagine, all those who saw it. One was a man's hand, the other a woman's: a couple, we assumed, ambushed by death, reaching out to each other for a last, brief assurance of existence.
But all of this took place in a self-contained world of horror. Even as the salvage crews and the police and we professional voyeurs picked our way warily around the edges of the site, afraid of what we might step on, a steady procession of other aircraft took off from Orly, glinting in the sky above us. Like the "delicate ship" in Brueghel's Icarus that saw something amazing, they had "somewhere to get to and sailed calmly on".
So it was this Tuesday evening and through the night as thousands of passengers and crew departing Charles de Gaulle Airport were given - could hardly avoid - a bird's-eye view of the still-burning debris of Concorde and the small hotel it crashed into. Those departing on Wednesday morning, when the fire was more or less extinguished, could have seen the traffic cones that marked the position of the bodies.
The philosophy that life must go on with barely a pause is, of course, instinctive - as innate, perhaps, as the need to fix the blame with all possible speed.
And in the case of Concorde, as in the case of the DC-10 - as in the case of the dozen other airline disasters I have covered in the interim - there is a pattern to the blaming process that seems as inevitable as it is unfair.
Within hours of the DC-10 disaster, after the plane's rear cargo door and six bodies had been found in a field several miles short of the crash site, the finger of suspicion was pointed at the 39-year-old Algerian cargo handler who had closed the door at Orly. In all-too-short a time that suspicion grew to near certainty, and Mohammed Mahmoudi found himself publicly accused - by the president of the DC-10's manufacturers, no less - of, in effect, causing the deaths of 346 people.
Similarly, on Wednesday, the Air France mechanics who, it emerged, had worked on the Concorde's number 2 engine shortly before it took off, found themselves suspected of botching the job; of, perhaps, leaving a metaphorical spanner in the works that caused the engine to explode.
But like the unfortunate Mr Mahmoudi - and like those pilots accused of the most basic flying errors and even of committing suicide; like the naval gunners accused of accidentally shooting down a Boeing 747 off the shore of Long Island - they serve a crucial need: reassuring us, the flying public, that the occasional catastrophes are conjunctions of the rarest chance and human error; that the chances of it happening again, and to us, are minuscule.
For, if we did not believe that, if we believed that a crash had occurred because of a fundamental flaw in the design or operation of a particular type of aircraft, it is not likely we would be willing to climb aboard its sister ships and sail calmly on.
Yet, all too often, it emerges - months and sometimes years later; long after we've put the scenes of destruction from our minds - that the aircraft's crash was due to "human error" and came from those who designed or built the planes, not those who flew or maintained them.
It was not Mr Mahmoudi's fault that the DC-10 crashed into the Forest of Ermenonville. Rather, it was a design flaw that allowed him to close the cargo door improperly; a flaw that was identified and should have been fixed but wasn't, even though the manufacturer's paperwork said it had been.
It was not an errant missile or a bomb that brought down TWA's Flight 800 in 1996, but the arrangement of the fuel tanks in early models of the Boeing 747 that, combined with refuelling procedures, can allow hot explosive gasses to build up in the tank that sits beneath the passenger cabin.
It was not "pilot error" or extreme weather that caused two early models of the Boeing 737 to crash - and sent a string of others plunging to near-disaster - but a flaw in the design of the rudder.
OR at least, those are the opinions of the American National Transportation Safety Board, which, by mandate of the US Congress, investigates every aviation accident or serious incident that occurs in America, or involves American-built aircraft, or aircraft that fly to America (which means virtually all commercial jets), and it routinely issues warnings of aircraft systems or components or operating procedures it believes to be "unsafe".
In light of what has now emerged about the catalogue of disasters that overtook Concorde as it accelerated down the runway - at least one tyre burst, one engine out, one more failing, flames shooting from the wing - it is sobering to remember that in 1981 the NTSB recognised and warned of the "potentially catastrophic" problems concerning the tyres on planes that have much faster take-off and landing speeds than its subsonic cousins.
Between 1996 and 1998 there were repeated warnings of problems that can arise with engines that operate at maximum speed and temperature for 60 per cent of their flying time, compared to 10 per cent of the time for more conventional engines.
But the NTSB can only issue "findings of probable cause". These are frequently disputed by the manufacturers, and "recommendations" are frequently resisted by the airlines. It has no authority to impose its will and the agency that does - the Federal Aviation Administration - must draw the balance between safety and the practicalities of commercial aviation. (Safety is the FAA's priority, but the only truly safe plane would never leave the ground.)
It is true that commercial aviation is, by any measure, extraordinarily safe - but it is also true that it is nowhere near as safe as it could and should be. The FAA now has the declared "number one" aim of cutting the number of aviation accidents in the US by 80 per cent by the year 2007.
It can only do that by forcing manufacturers and airlines to acknowledge that the line they walk is exceedingly narrow and that the "absolute confidence" they express in the safety of their aircraft - particularly immediately after an accident, as British Airways quickly and loudly declared its faith in Concorde on Wednesday - is hubris.
The machines they build and fly, however ingenious, are imperfect - as flawed as the human agencies that created them. And as Rudyard Kipling reminded us in one of his poems, machines are not built to "comprehend a lie".
They live by an immutable law:
We can neither love, nor pity, nor forgive.
If you make a slip in handling us, you die.
Paul Eddy is a freelance journalist who was co-author (with Elaine Potter and Bruce Page) of Destination Disaster - The Risk of Flying, published in 1976. His first novel, Flint, has been sold around the world and will be published in Ireland by Headline Features early next year.