What Technology Wants (20 page)

Inevitable! There is that word again. Common instances of equivalent inventions independently discovered at the same moment suggest that the evolution of technology converges in the same manner as biological evolution. If so, then if we could rewind and replay the tape of history, the very same sequence of inventions should roll out in a very similar sequence every time we reran it. Technologies would be inevitable. The appearance of morphological archetypes would further suggest that this technological invention has a direction, a tilt. A tilt that is independent to a certain extent of its human inventors.

Indeed, in all fields of technology we commonly find independent, equivalent, and simultaneous invention. If this convergence indicated that discoveries were inevitable, the inventors would appear as conduits filled by an invention that just had to happen. We would expect the people making them to be interchangeable, if not almost random.

That is exactly what psychologist Dean Simonton found. He took Ogburn and Thomas's catalog of simultaneous invention before 1900 and aggregated it with several other similar lists to map out the pattern of parallel discovery for 1,546 cases of invention. Simonton plotted the number of discoveries made by 2 individuals against the number of discoveries made by 3 people, or 4 people, or 5, or 6. The number of 6-person discoveries was of course lower, but the exact ratio between these multiples produced a pattern known in statistics as a Poisson distribution. This is the pattern you see in mutations on a DNA chromosome and in other rare chance events in a large pool of possible agents. The Poisson curve suggested that the system of “who found what” was essentially random.

Certainly talent is unequally distributed. Some innovators (like Edison, or Isaac Newton, or William Thomson Kelvin) are simply better than others. But if geniuses aren't able to jump far ahead of the inevitable, how do the better inventors become great? Simonton discovered that the higher the prominence of a scientist (as determined by the number of pages his biography occupies in encyclopedias), the greater the number of simultaneous discoveries he participated in. Kelvin was involved in 30 sets of simultaneous discoveries. Great discoverers not only contribute more than the average number of “next” steps, but they also take part in those steps that have the greatest impact, which are naturally the areas of investigation that attract many other players and so produce multiples. If discovery is a lottery, the greatest discoverers buy lots of tickets.

Simonton's set of historical cases reveals that the number of duplicated innovations has been increasing with time—simultaneous discovery is happening more often. Over the centuries the velocity of ideas has accelerated, speeding up codiscovery as well. The degree of synchronicity is also gaining. The gap between the first and last discovery in a concurrent multiple has been shrinking over the centuries. Long gone is the era when 10 years could elapse between the public announcement of an invention or discovery and the date the last researcher would hear about it.

Synchronicity is not just a phenomenon of the past, when communication was poor, but very much part of the present. Scientists at AT&T Bell Labs won a Nobel Prize for inventing the transistor in 1948, but two German physicists independently invented a transistor two months later at a Westinghouse laboratory in Paris. Popular accounts credit John von Neumann with the invention of a programmable binary computer during the last years of World War II, but the idea and a working punched-tape prototype were developed quite separately in Germany a few years earlier, in 1941, by Konrad Zuse. In a verifiable case of modern parallelism, Zuse's pioneering binary computer went completely unnoticed in the United States and the UK until many decades later. The ink-jet printer was invented twice: once in Japan in the labs of Canon and once in the United States at Hewlett-Packard, and the key patents were filed by the two companies within months of each other in 1977. “The whole history of inventions is one endless chain of parallel instances,” writes anthropologist Alfred Kroeber. “There may be those who see in these pulsing events only a meaningless play of capricious fortuitousness; but there will be others to whom they reveal a glimpse of a great and inspiring inevitability which rises as far above the accidents of personality.”

The strict wartime secrecy surrounding nuclear reactors during World War II created a model laboratory for retrospectively illuminating technological inevitability. Independent teams of nuclear scientists around the world raced against one another to harness atomic energy. Because of the obvious strategic military advantage of this power, the teams were isolated as enemies or kept ignorant as wary allies or separated by “need to know” secrecy within the same country. In other words, the history of discovery ran in parallel among seven teams. Each discrete team's highly collaborative work was well documented and progressed through multiple stages of technological development. Looking back, researchers can trace parallel paths as the same discoveries were made. In particular, physicist Spencer Weart examined how six of these teams each independently discovered an essential formula for making a nuclear bomb. This equation, called the four-factor formula, allows engineers to calculate the critical mass necessary for a chain reaction. Working in parallel but in isolation, teams in France, Germany, and the Soviet Union and three teams in the United States simultaneously discovered the formula. Japan came close but never quite reached it. This high degree of simultaneity—six simultaneous inventions—strongly suggests the formula was inevitable at this time.

However, when Weart examined each team's final formulation, he saw that the equations varied. Different countries used different mathematical notation to express it, emphasized different factors, varied in their assumptions and interpretation of results, and awarded the overall insight different status. In fact, the equation was chiefly ignored as merely theoretical by four teams. In only two teams was the equation integrated into experimental work—and one of those was the team that succeeded in making a bomb.

The formula in its abstract form was inevitable. Indisputably, if it had not been found by one, five others would have found it. But the specific expression of the formula was not at all inevitable, and that volitional expression can make a significant difference. (The political destiny of the country that put the formula to work, the United States, is vastly different from those that failed to exploit the discovery.)

Both Newton and Gottfried Leibniz are credited with inventing (or discovering) calculus, but in fact their figuring methods differed, and the two approaches were only harmonized over time. Joseph Priestley's method of generating oxygen differed from Carl Scheele's; using different logic they uncovered the same inevitable next stage. The two astronomers who both correctly predicted the existence of Neptune (John Couch Adams and Urbain Le Verrier) actually calculated different orbits for the planet. The two orbits just happened to coincide in 1846, so they found the same body by different means.

But aren't these kinds of anecdotes mere statistical coincidences? Given the millions of inventions in the annals of discovery, shouldn't we expect a few to happen simultaneously? The problem is that most multiples are unreported. Sociologist Robert Merton says, “All singleton discoveries are imminent multiples.” By that he means that many potential multiples are abandoned when news of the firstborn is announced. A typical notebook entry goes like this one found in the records of mathematician Jacques Hadamard in 1949: “After having started a certain set of questions and seeing that several authors had begun to follow the same line, I happen to drop it and to investigate something else.” Or a scientist will record their discoveries and inventions but never publish the work due to busyness, or their own dissatisfaction with the results. Only the notebooks of the great get a careful examination, so unless you are either Cavendish or Gauss (the notebooks of both reveal several unpublished multiples), your unreported ideas will never be counted. Further concurrent research is hidden by classified, corporate, or state-secret work. Much is not disseminated because of fear of competitors, and until very recently, many examples of duplicate discoveries and inventions remained obscure because they were published in obscure languages. A few coexistent inventions went unrecognized because they were described in impenetrable technical language. And occasionally a discovery is so contrarian or politically incorrect that it is ignored.

Furthermore, once a discovery has been revealed and entered into the repository of what is commonly known, all later investigations that arrive at the same results are reckoned as mere corroborations of the original—no matter how they are actually arrived at. A century ago the failure of communication was in its slow speed; a researcher in Moscow or Japan might not hear about an English invention for decades. Today the failure is due to volume. There is so much published, so fast, in so many areas, that it is very easy to miss what has already been done. Re-inventions arise independently all the time, sometimes in full innocence centuries later. But because their independence can't be proven, these Johnny-come-latelies are counted as confirmations and not as evidence of inevitability.

By far the strongest bits of evidence for ubiquitous simultaneity of invention are scientists' own impressions. Most scientists consider getting scooped by another person working on the same ideas the unfortunate and painful norm. In 1974 sociologist Warren Hagstrom surveyed 1,718 U.S. academic research scientists and asked them if their research had ever been anticipated, or scooped, by others. He found that 46 percent believed that their work had been anticipated “once or twice” and 16 percent claimed they had been preempted three or more times. Jerry Gaston, another sociologist, surveyed 203 high-energy physicists in the UK and got similar results: 38 percent claimed to have been anticipated once and another 26 percent more than once.

Unlike scientific scholarship, which places a huge emphasis on previous work and proper credit, inventors tend to plunge ahead without methodically researching the past. This means reinvention is the norm from the patent office's viewpoint. When inventors file patents, they need to cite previous related inventions. One-third of inventors surveyed claimed they were unaware there were prior claims to their idea while developing their own invention. They did not learn about the competing patents until preparing their application with the required “prior art.” More surprising, one-third claimed to be unaware of the prior inventions cited in their own patent until notified by the survey takers. (This is entirely possible, since patent citations can be added by the inventor's patent attorney or even the patent office examiner.) Patent law scholar Mark Lemley states that in patent law “a large percent of priority disputes involve near-simultaneous invention.” One study of these near-simultaneous priority disputes, by Adam Jaffe of Brandeis University, showed that in 45 percent of cases both parties could prove they had a “working model” of the invention within six months of each other, and in 70 percent of cases within a year of each other. Jaffe writes, “These results provide some support for the idea that simultaneous or near-simultaneous invention is a regular feature of innovation.”

There is the air of inevitability about these simultaneous discoveries. When the necessary web of supporting technology is established, then the next adjacent technological step seems to emerge as if on cue. If inventor X does not produce it, inventor Y will. But the step will come in the proper sequence.

This does not mean the iPod, with its perfect, milky case, was inevitable. We can say the invention of the microphone, the laser, the transistor, the steam turbine, and the waterwheel and the discovery of oxygen, DNA, and Boolean logic were all inevitable in roughly the era they appeared. However, the particular form of the microphone, its exact circuit, or the specific engineering of the laser, or the particular materials of the transistor, or the dimensions of the steam turbine, or the peculiar notation of the chemical formula, or the specifics of any invention are not inevitable. Rather, they will vary quite widely due to the personality of their finder, the resources at hand, the culture or society they are born into, the economics funding the discovery, and the influence of luck and chance. A light based on a coil of tungsten strung within an oval vacuum bulb is not inevitable, but the electric incandescent lightbulb is.

The general concept of the electric incandescent lightbulb can be abstracted from all the specific details allowed to vary (voltage, height, kind of bulb) while still producing the result—in this case, luminance from electricity. This general concept is similar to the archetype in biology, while the specific materialization of the concept is more like a species. The archetype is ordained by the technium's trajectory, while the species is contingent.

The electric incandescent lightbulb was invented, reinvented, coin-vented, or “first invented” dozens of times. In their book
Edison's Electric Light: Biography of an Invention,
Robert Friedel, Paul Israel, and Bernard Finn list 23 inventors of incandescent bulbs prior to Edison. It might be fairer to say that Edison was the very
last
“first” inventor of the electric light. These 23 bulbs (each an original in its inventor's eyes) varied tremendously in how they fleshed out the abstraction of “electric lightbulb.” Different inventors employed various shapes for the filament, different materials for the wires, different strengths of electricity, different plans for the bases. Yet they all seemed to be independently aiming for the same archetypal design. We can think of the prototypes as 23 different attempts to describe the inevitable generic lightbulb.

Quite a few scientists and inventors, and many outside science, are repulsed by the idea that the progress of technology is inevitable. It rubs them the wrong way because it contradicts a deeply and widely held belief that human choice is central to our humanity and essential to a sustainable civilization. Admitting that anything is “inevitable” feels like a cop-out, a surrender to invisible, nonhuman forces beyond our reach. Such a false notion, the thinking goes, may lull us into abdicating our responsibility for shaping our own destiny.