Go to Chapter One Section Go to Book World's Review

Imagined Worlds
By Freeman Dyson

Chapter One: Stories

Successful technologies often begin as hobbies. Jacques Cousteau invented scuba diving because he enjoyed exploring caves. The Wright brothers invented flying as a relief from the monotony of their normal business of selling and repairing bicycles. A little earlier, the bicycle and the automobile began as recreational vehicles, as means for people of leisure to explore the countryside, before smooth roads existed to make riding and driving efficient. In all these technologies, the pioneers were spending their money and risking their lives for nothing more substantial than fun. Scuba diving is fun, flying is fun, riding bicycles and driving cars are fun, especially in the early days when nobody else is doing it. Even today, when each of these four hobbies has grown into a huge industry, when legal regulations are enforced to reduce the risks as far as possible, sport and recreation are still supplying much of the motivation for pushing the technologies ahead.

The history of flying is a good example to look at in detail for insight into the interaction of technology with human affairs, because two radically different technologies were competing for survival--in the beginning they were called heavier-than-air and lighter-than-air. The airplane and the airship were not only physically different in shape and size but also sociologically different. The airplane grew out of dreams of personal adventure. The airship grew out of dreams of empire. The image in the minds of airplane-builders was a bird. The image in the minds of airship-builders was an oceanliner.

We are lucky to have a vivid picture of the creative phases of these technologies, written by a man who was deeply involved in both and was also a gifted writer, Nevil Shute Norway. Before he became the famous novelist Nevil Shute--author of Pied Piper, A Town like Alice, On the Beach, and other wonderful stories--he was an aeronautical engineer working professionally on the design of airplanes and airships. He wrote an autobiography with the title Slide Rule, describing his life as an engineer.

Norway did not start out with any bias for airplanes and against airships. He worked on both with equal dedication, and he was particularly proud of his part in the design of the airship R100. He worked on it for six years, from the moment of conception in 1924 to the delivery in 1930, and flew on its triumphant maiden voyage in 1930, from London to Montreal and back. From a technical point of view, airships then had many advantages over airplanes, and the R100 was a technical success. But Norway saw clearly that the fate of airships and airplanes did not depend on technical factors alone. Even before he became a professional writer, he was more interested in people than in nuts and bolts. He saw and recorded the human factors that made the building of airplanes fun and made the building of airships a nightmare.

After finishing the R100, Norway started a company of his own, Airspeed Limited. It was one of the hundreds of small companies that were inventing and building and selling airplanes in the 1920s and 30s. Norway estimated that 100,000 different varieties of airplane were flown during those years. All over the world, enthusiastic inventors were selling airplanes to intrepid pilots and to fledgling airlines. Many of the pilots crashed and many of the airlines became bankrupt. Out of 100,000 types of airplane, about 100 survived to form the basis of modern aviation. The evolution of the airplane was a strictly Darwinian process in which almost all the varieties of airplane failed, just as almost all species of animal become extinct. Because of the rigorous selection, the few surviving airplanes are astonishingly reliable, economical, and safe.

The Darwinian process is ruthless, because it depends upon failure. It worked well in the evolution of airplanes because the airplanes were small, the companies that built them were small, and the costs of failure in money and lives were tolerable. Planes crashed, pilots were killed, and investors were ruined, but the scale of the losses was not large enough to halt the process of evolution. After the crash, new pilots and new investors would always appear with new dreams of glory. And so the selection process continued, weeding out the unfit, until airplanes and companies had grown so large that further weeding was officially discouraged. Norway's company was one of the few that survived the weeding and became commercially profitable. As a result, it was bought out and became a division of De Havilland, losing the freedom to make its own decisions and take its own risks. Even before De Havilland took over the company, Norway decided that the business was no longer fun. He stopped building airplanes and started his new career as a novelist.

The evolution of airships was a different story, dominated by politicians rather than by inventors. British politicians in the 1920s were acutely aware that the century of world-wide British hegemony based upon sea power had come to an end. The British Empire was still the biggest in the world but could no longer rely on the Royal Navy to hold it together. Most of the leading politicians, both Conservative and Labor, still had dreams of empire. They were told by their military and political advisers that in the modern world air power was replacing sea power as the emblem of greatness. So they looked to air power as the wave of the future that would keep Britain on top of the world. And in this context it was natural for them to think of airships rather than airplanes as the vehicles of imperial authority. Airships were superficially like oceanliners, big and visually impressive. Airships could fly nonstop from one end of the empire to the other. Important politicians could fly in airships from remote dominions to meetings in London without being forced to neglect their domestic constituencies for a month. In contrast, airplanes were small, noisy, and ugly, altogether unworthy of such a lofty purpose. Airplanes at that time could not routinely fly over oceans. They could not stay aloft for long and were everywhere dependent on local bases. Airplanes were useful for fighting local battles, but not for administering a worldwide empire.

One of the politicians most obsessed with airships was the Labor Peer Lord Thompson, Secretary of State for Air in the Labor governments of 1924 and 1929. Lord Thompson was the driving force behind the project to build the R101 airship at the government-owned Royal Airship Works at Cardington. Being a socialist as well as an imperialist, he insisted that the government factory get the job. But as a compromise to keep the Conservative opposition happy, he arranged for a sister ship, the R100, to be built at the same time by the private firm Vickers Limited. The R101 and R100 were to be the flagships of the British Empire in the new era. The R101, being larger, would fly nonstop from England to India and perhaps later to Australia. The R100, a more modest enterprise, would provide regular service over the Atlantic between England and Canada. Norway, from his position in the team of engineers designing the R100, had a front-seat view of the fate of both airships.

The R101 project was from the beginning driven by ideology rather than by common sense. At all costs, the R101 had to be the largest airship in the world, and at all costs it had to be ready to fly to India by a fixed date in October 1930, when Lord Thompson himself would embark on its maiden voyage to Karachi and back, returning just in time to attend an Imperial Conference in London. His dramatic arrival at the conference by airship, bearing fresh flowers from India, would demonstrate to an admiring world the greatness of Britain and the Empire, and incidentally demonstrate the superiority of socialist industry and of Lord Thompson himself. The huge size and the fixed date were a fatal combination. The technical problems of sealing enormous gasbags so that they should not leak were never solved. There was no time to give the ship adequate shake-down trials before the voyage to India. It finally took off on its maiden voyage, soaking wet in foul weather, with Lord Thompson and his several thousand pounds of lordly baggage on board. The ship had barely enough lift to rise above its mooring-mast. Eight hours later it crashed and burned on a field in northern France. Of the fifty-four people on board, six survived. Lord Thompson was not among them.

Meanwhile, the R100, with Norway's help, had been built in a more reasonable manner. Its gasbags did not leak, and it had an adequate margin of lift to carry its designed pay-load. The R100 completed its maiden voyage to Montreal and back without disaster, seven weeks before the R101 left England. But Norway found the voyage far from reassuring. He reports that the R100 was violently tossed around in a local thunderstorm over Canada and was lucky to have avoided being torn apart. He did not judge it safe enough for regular passenger service. The question whether it was safe enough became moot after the R101 disaster. After one such disaster, no passengers would be likely to volunteer for another. The R100 was quietly dismantled and the pieces sold for scrap. The era of imperial airships had come to an end.

The announced purpose of the R100 was to provide a reliable passenger service between England and Canada, arriving and leaving once a week. After the airship failed, Lord Cunard, the owner of the Cunard shipping company, asked his engineers what it would take to provide a weekly service across the Atlantic using only two oceanliners. At that time it took seven or eight days for a ship to cross the Atlantic, so that a weekly service needed at least three ships. To do it with two ships would require crossing in five days, with two days margin for bad weather, loading, and unloading. The Cunard engineers designed the Queen Mary and the Queen Elizabeth to cross in five days. To do this economically, because of the way wave-drag scales with speed and size, the two ships had to be substantially larger than other oceanliners. Lord Cunard felt confident that the business of transporting passengers by ship could remain profitable for a few more decades, and he ordered the ships to be built.

In due course, after the interruption caused by the second world war, they were carrying passengers profitably across the ocean and incidentally breaking speed records. The British public was proud of these ships, which regularly won the famous Blue Ribbon for the fastest Atlantic crossing. The public imagined that the ships were designed to win the Blue Ribbon, but Lord Cunard said the public misunderstood the purpose of the ships completely. He said his purpose was always to build the smallest and slowest ships that could do a regular weekly service. It was just an unfortunate accident that to do this job you had to break records. The ships continued their weekly sailings profitably for many years, until the Boeing 707 put them out of business.

While oceanliners were still enjoying their heyday, before the triumph of the Boeing 707, another tragedy of ideologically driven technology occurred. This was the tragedy of the Comet jetliner. During World War II the De Havilland company had built bombers and jet fighters and acquired an appetite for bigger things. After the war, the company went ahead with the design of the Comet, a commercial jet that could fly twice as fast as the propeller-driven transport planes of that era. At the same time, the British government established the British Overseas Airways Corporation, a state-owned monopoly with responsibility for long-distance air routes. The Empire was disintegrating rapidly, but enough of it remained to inspire the planners at BOAC with new dreams of glory. Their dream was to deploy a fleet of Comets on the Empire routes that BOAC controlled, from London south to Africa and east to India and Australia.

The dream was seductive because it meant that Britain would move into the jet age five years ahead of the slow-moving Americans. While the Boeing Company hesitated, the Comets would be flying. The Comets would display to the world the superiority of British technology, and incidentally demonstrate that the Empire, now renamed the Commonwealth, was still alive. After the BOAC Comets had shown the way, other airlines all over the world would be placing orders with De Havilland. The dreams that inspired the Comet were the same dreams that inspired the R101 twenty years earlier. The heirs of Lord Thompson had learned little from his fate.

The Comet enterprise made the same mistake as the R101, pushing ahead into a difficult and demanding technology with a politically dictated time-table. The decision to rush the Comet into service in 1952 was driven by the political imperative of staying five years ahead of the Americans. One man foresaw the disaster that was coming. Nevil Shute, no longer an aeronautical engineer but a well-informed bystander, published in 1948 a novel with the title No Highway, which described how political pressures could push an unsafe airplane into service. The novel tells the story of a disaster that is remarkably similar to the Comet disasters that happened four years later.

The fatal flaw of the Comet was a concentration of stress at the corners of the windows. The stress caused the metal skin of the plane to crack and tear open. The cracking occurred only at high altitudes when the plane was fully pressurized. The result was a disintegration of the plane and strewing of wreckage over wide areas, leaving no clear evidence of the cause. Two planes were destroyed in this way, one over India and one over Africa, killing everybody on board. After the second crash, the Comets stopped flying. For five years no jetliners flew, until the Americans were ready with their reliable and thoroughly tested Boeing 707. It took a hundred deaths to stop the Comets from flying, twice as many as it took to stop the airships. If the Secretary of State for Air had been on board the first Comet when it crashed, the second crash might not have been necessary.

Nevil Shute explains how it happened that the R101 and the Comets were allowed to carry passengers without adequate flight-testing. It happened because of a clash of two cultures, the culture of politics and the culture of engineering. Politicians were making crucial decisions about technical matters which they did not understand. The job of a senior politician is to make decisions. Political decisions are often made on the basis of inadequate knowledge, and usually without doing much harm. In the culture of politics, a leader gains respect by saying: "The buck stops here." To take a chance of making a bad decision is better than to be indecisive. The culture of engineering is different. An engineer gains respect by saying: "Better safe than sorry." Engineers are trained to look for weak points in a design--to warn of potential disaster. When politicians are in charge of an engineering venture, the two cultures clash. When the venture involves machines that fly in the air, a clash tends to result in a crash.

Aviation is the branch of engineering that is least forgiving of mistakes. But from a wider point of view, unforgivingness may be a virtue. In the long view of history, the victims of the R101 and the Comets did not die in vain. They left as the legacy of their tragedy the extraordinarily safe and reliable airplanes that now fly every day across continents and oceans all over the world. Without the harsh lessons of disaster and death, the modern jetliner would not have evolved.

My friend Albert Hirschman has found other places where unforgivingness is a virtue. He is an economist who has spent much of his life studying Latin American societies and giving advice to their governments. He has also given advice to newly independent countries in Africa. He is often asked by the leaders of poor countries, "Should we put our limited resources into roads or into airlines?" When this question is asked, the natural impulse of an economist is to say "roads," because the money spent on roads provides jobs for local people, and the roads benefit all classes of society. In contrast, the building of a national airline requires the import of foreign technology, and the airline benefits only the minority of citizens who can afford to use it. Nevertheless, long experience in Africa and Latin America has taught Hirschman that "roads" is usually the wrong answer. In the real world, roads have several disadvantages. The money assigned to road-building tends to fall into the hands of corrupt local officials. Roads are easier to build than to maintain. And when, as usually happens, the new roads decay after a few years, the decay is gradual and does not create a major scandal. The end-result of road-building is that life continues as before. The economist who said "roads" has achieved little except a small increase in the wealth and power of local officials.

Contrast this with the real-world effect of building a national airline. After the money is spent, the country is left with some expensive airplanes, some expensive airports, and some expensive modern equipment. The foreign technicians have left and local people must be trained to operate the system. Unlike roads, airplanes do not decay gracefully. A crash of an airliner is a highly visible event and brings unacceptable loss of prestige to the rulers of the country. The victims tend to be people of wealth and influence, and their deaths do not pass unnoticed. The rulers have no choice. Once they own an airline, they are compelled to see to it that the airline is competently run. They are forced to create a cadre of highly motivated people who maintain the machines, come to work on time, and take pride in their technical skills. As a result, the airline brings to the country indirect benefits that are larger than its direct economic value. It creates a substantial body of citizens accustomed to strict industrial discipline and imbued with a modern work ethic. And these citizens will in time find other useful things to do with their skills besides taking care of airplanes. In this paradoxical way, the unforgivingness of aviation makes it the best school for teaching a traditional society how to modernize.

This is not the first time that an unforgiving technology has transformed the world and forced traditional societies to change. The role of aviation today is similar to the role of sailing ships in the preindustrial world. King Henry VIII of England, the most brutal and most intelligent of English monarchs, destroyer of monasteries and founder of colleges, murderer of wives and composer of madrigals, for whose soul regular prayers are still said at Trinity College Cambridge in gratitude for his largesse, understood that the most effective tool for modernizing England was the creation of a Royal Navy. It was not by accident that the industrial revolution of the eighteenth century began in England, in the island where daily life and economics had been dominated for 300 years by the culture of sailing ships. When the young Tsar Peter the Great of Russia, a kindred spirit to Henry, decided that the time had come to modernize the Russian empire, he prepared himself for the job by going to work as an apprentice in a shipyard.

The R101 and Comet tragedies are examples of the baleful effects of ideology, the ideology in those cases being old-fashioned British imperialism. Today, the British Empire is ancient history, and its ideology is dead. But technologies driven by ideology are likely to run into trouble, even when the ideology is not so outmoded. Another powerful ideology that ran into trouble is nuclear energy. All over the world, after the end of World War II, the ideology of nuclear energy flourished, driven by an intense desire to create something peaceful and useful out of the ruins of Hiroshima and Nagasaki. Scientists and politicians and industrial leaders were equally bewitched by this vision, that the great new force of nature that killed and maimed in war would now make deserts bloom in peace. Nuclear energy was so strange and powerful that it looked like magic. It was easy to believe that this magic could bring wealth and prosperity to poor people all over the earth. So it happened that in all large countries and in many small ones, in democracies and dictatorships, in communist and capitalist societies alike, Atomic Energy Authorities were created to oversee the miracles that nuclear energy was expected to perform. Huge funds were poured into nuclear laboratories in the confident belief that these were sound investments for the future.

I visited Harwell, the main British nuclear research establishment, during the early days of nuclear enthusiasm. The first director of Harwell was Sir John Cockcroft, a first-rate scientist and an honest public servant. I walked around the site with Cockcroft, and we looked up at the massive electric power lines running out of the plant, over our heads and away into the distance. Cockcroft remarked, "The public imagines that the electricity is flowing out of this place into the national grid. When I tell them that it is all flowing the other way, they don't believe me."

There was nothing wrong, and there is still nothing wrong, with using nuclear energy to make electricity. But the rules of the game must be fair, so that nuclear energy competes with other sources of energy and is allowed to fail if it does badly. So long as it is allowed to fail, nuclear energy can do no great harm. But the characteristic feature of an ideologically driven technology is that it is not allowed to fail. And that is why nuclear energy got into trouble. The ideology said that nuclear energy must win. The promoters of nuclear energy believed as a matter of faith that it would be safe and clean and cheap and a blessing to humanity. When evidence to the contrary emerged, the promoters found ways to ignore the evidence. They wrote the rules of the game so that nuclear energy could not lose. The rules for cost-accounting were written so that the cost of nuclear electricity did not include the huge public investments that had been made to develop the technology and to manufacture the fuel. The rules for reactor safety were written so that the type of light-water reactor originally developed by the United States Navy for propelling submarines was by definition safe. The rules for environmental cleanliness were written so that the ultimate disposal of spent fuel and worn-out machinery was left out of consideration. With the rules so written, nuclear energy confirmed the beliefs of its promoters. According to these rules, nuclear energy was indeed cheap and clean and safe.

The people who wrote the rules did not intend to deceive the public. They deceived themselves, and then fell into a habit of suppressing evidence that contradicted their firmly held beliefs. In the end, the ideology of nuclear energy collapsed because the technology that was not allowed to fail was obviously failing. In spite of the government subsidies, nuclear electricity did not become significantly cheaper than electricity made by burning coal and oil. In spite of the declared safety of light-water reactors, accidents occasionally happened. In spite of the environmental advantages of nuclear power plants, disposal of waste fuel remained an unsolved problem. The public, in the end, reacted harshly against nuclear power because obvious facts contradicted the claims of the promoters.

When a technology is allowed to fail in competition with other technologies, the failure is a part of the normal Darwinian process of evolution, leading to improvements and possible later success. When a technology is not allowed to fail, and still it fails, the failure is far more damaging. If nuclear power had been allowed to fail at the beginning, it might well have evolved by now into a better technology which the public would trust and support. There is nothing in the laws of nature that stops us from building better nuclear power plants. We are stopped by deep and justified public distrust. The public distrusts the experts because they claimed to be infallible. The public knows that human beings are fallible. Only people blinded by ideology fall into the trap of believing in their own infallibility.

The tragedy of nuclear fission energy is now almost at an end, so far as the United States is concerned. Nobody wants to build any new fission power plants. But another tragedy is still being played out, the tragedy of nuclear fusion. The promoters of fusion are making the same mistakes that the promoters of fission made thirty years earlier. The promoters are no longer experimenting with a variety of fusion schemes in order to evolve a machine that might win in the marketplace. They long ago decided to concentrate their main effort upon a single device, the Tokamak, which is declared by ideological fiat to be the energy producer for the twenty-first century. The Tokamak was invented in Russia, and its inventors gave it a name that transliterates euphoniously into other languages. All the countries with serious programs of fusion research have built Tokamaks. One of the largest and most expensive is in Princeton. To me it looks like a plumber's nightmare, a dense conglomeration of pipes and coils with no space for anybody to go in and fix it when it needs repairs. But the people who built it believe sincerely that it is an answer to human needs. The various national fusion programs are supposed to converge upon a huge international Tokamak, costing many billions of dollars, which will be the prototype for the fusion power producers of the future. The usual claims are made, that fusion power will be safe and clean, although even the promoters are no longer saying that it will be cheap. The existing fusion programs have stopped the evolution of a new technology that might actually fulfill the hopes of the promoters. What the world needs is a small, compact, flexible fusion technology that could make electricity where and when it is needed. The existing fusion program is leading to a huge source of centralized power, at a price that nobody except a government can afford. It is likely that the existing fusion program will sooner or later collapse as the fission program collapsed, and we can only hope that some more useful form of fusion technology will rise from the wreckage.

© 1996 President and Fellows of Harvard College

Harvard University Press

Back to the top