The majority of people reading this column live in the same gigantic, interconnected, unpredictable, dangerous machine. The electrical grids, with their nuclear plants, the industrial infrastucture that supports and transports billions of people, global communications networks, chemical plants and supertankers, are all part of it. It seems that the only we ever discover the limits of these technologies when something goes very wrong and large numbers of people die in a kind of accident that wasn't supposed to happen. I met a fellow recently who makes an impressive argument that big parts of the megamachine we live in are doomed to fail in fatal ways, despite all attempts to prevent such catastrophes.
""The technological disasters at Three Mile Island, Bhopal, Chernobyl triggered political reactions,'' a young man named Wade Roush told me, about a year ago. He and I had struck up an e-mail correspondence about his research into technological disasters, from the Challenger explosion to the great power blackouts in the Northeast, and the social changes that followed the disasters. He came by my office in California to discuss his doctoral research for MIT. He argued that all large technological systems are bound to fail eventually in an unforeseen way, because of the way complex systems give rise to ""emergent phenomena.'' Roush creditedYale sociologist Charles Perrow, author of ""Normal Accidents: Living With High-Risk Technologies'' (Basic Books, 1984) for this idea.
The tendency of complex, high-energy systems to fail because of phenomena that could not have been predicted neutralizes any strategy of building fail-safe systems to ward off such disasters. Additional safety systems increase complexity, and hence increase the possibility of unpredicted breakdown.
Roush sent me his finished thesis. In it he outlines how modern societies react to high-tech disasters. Following disasters, societies can become much more democratic about how such technologies are employed. But Roush also shows how technological failures can help destroy empires:
""The people of the former Soviet Union asserted their freedom bit by bit as glasnost gradually exposed the weakness, corruption, and bloody history of the Communist Party. But if this proces of awakening and recognition had a discrete beginning, it came at 1:23 AM on April 26, 1986, when reactor No. 4 of the state-run V.I. Lenin Atomic Energy Station at Chernobyl disintegrated in a blast of steam, flaming graphite, and deadly radionuclides.''
The Soviets failed to design a safe nuclear plant. On top of that, they failed to operate it safely. When it exploded, they failed to contain the damage, failed to evacuate and decontaminate adequately. Knowing the facts, they failed to tell the truth. These failures, together with the testimony and suicide of Legasov, a top official of the Soviets nuclear program, were among the factors that led to the toppling of the Soviet system. An anti-nuclear movement mobilized at a moment in history when an organized resistance to the Soviet state of any kind was unheard-of: ""The dissolution of the USSR shortly after the abortive coup of August, 1991, was a kind of consensual state suicide in which the various republics agreed to sever the ties, especially those of Soviet Communism, which had bound them for so long. This moment might not have come with such swiftness and finality if not for the Chernobyl disaster's political repercussions in Ukraine, Belarus, Lithuania, Russia, and other republics,'' Roush wrote.
Two important themes emerge from Roush's social-technological scholarship: All complex, high-energy systems fail eventually; the failures are embedded in social structures, which also fail in the wake of disaster. It was the arrogance of the operators of Chernobyl on the night of the explosion that led to the failure of four levels of safety systems. It was the centralization of the political power to make decisions about dangerous technologies that Soviet citizens attacked after Chernobyl.
Can we learn how to avoid mixing arrogant bureacracies with dangeroustechnologies? If we know that powerful, dangerous, complex technologies are bound to fail, what should designers of these technologies do in the future that they failed to do in the past? We need a thousand intellectual detectives like Wade Roush, digging up biggest untold story of the century -- nobody really knows how to operate this big machine that controls our lives.