Brilliant Blunders

BALTIMORE – Thomas Edison is reputed to have said, “I have not failed. I’ve just found 10,000 ways that won’t work.” This statement sums up a fundamental – but often misunderstood – truth about scientific inquiry. Progress in science – as in any creative discipline – is not a direct march to the answer, but a complex, zigzag path, involving many false starts and blind alleys. Blunders are not only inevitable; they are essential to innovative thinking, because they point the way for other explorers.

One may wonder whether today’s highly competitive, funding-starved scientific atmosphere, in which publications and citations have become a primary criterion for success, can accommodate such mistakes. The simple answer is yes. Indeed, they are as important as ever – and not only in academia.

French leaders

Confronting France’s Jihadis

Brigitte Granville examines what Noëlle Lenoir, Bernard-Henri Lévy, Fawaz Gerges, and other Project Syndicate commentators think is driving repeated terrorist attacks, and what should be done to stop them.

In fact, the entire scientific method is based on the notion that discovering what does not work is vital to learning what does. Any scientific theory must be falsifiable – that is, based on existing observations or experimental results. For a theory to be considered scientific, it must yield specific predictions of future observations or experimental results. If those observations or results contradict the predictions, the theory is discarded, or at least must be modified.

The mistakes that are integral to scientific progress are not those that result from haste, sloppiness, or inexperience. Rather, they are the mistakes that arise from thoughtful, meticulous experimentation based on bold ideas – the kind of ideas that can lead to major breakthroughs.

Fred Hoyle, one of the twentieth century’s greatest astrophysicists, provided a perfect example of such a “brilliant blunder.” Hoyle and two of his colleagues proposed what became known as the Steady State model of the universe, according to which the universe did not evolve following the so-called “big bang” (a term that Hoyle coined); instead, it was constant, remaining the same throughout eternity.

The idea was brilliantly elegant: just as our universe is homogeneous (the same at every point in space) and isotropic (looking the same in all directions), it remains the same at every point in time. While the Steady State theory was eventually falsified – our universe is expanding, and it most likely started from a big bang – it energized the entire field of cosmology, because it brought into sharp focus the questions that had to be addressed. In fact, currently fashionable models of the multiverse – the concept that our universe is but one of a huge ensemble of universes – are consistent with the idea that they are collectively in a kind of steady state.

The nineteenth-century physicist William Thomson, later known as Lord Kelvin, made his own brilliant blunder when he calculated that the Earth was less than 100 million years old – about fifty times younger than the age deduced from modern radiometric measurements. Though Kelvin’s estimate was seriously flawed, the effort remains central to the history of knowledge, because it applied real science – the laws of physics – to what had long been a subject of vague speculation.

Kelvin’s insights helped to launch a fruitful dialogue between geologists and physicists – a dialogue that eventually resolved even problems related to the length of time needed for Darwin’s theory of evolution to operate. And the oversight that warped Kelvin’s estimate – the possibility that fluid motion could efficiently transport heat within the Earth’s interior – turned out to be critical to understanding plate tectonics and continental drift.

Startup companies exemplify the potential benefits of risk-taking. While only about 49% of manufacturing startups and 37% of information startups survive for four or more years, those that do have managed to produce breakthrough innovations.

Tom Watson, Jr., who led IBM through decades of strong growth, is known for having supported brilliant blunders. As he put it, “We should have the courage to take risks when they are thoughtful risks….We must forgive mistakes which have been made because someone was trying to act aggressively in the company’s interest.”

Funding agencies for academic research should adopt a similar philosophy, awarding a certain share of financing to thoughtful, unconventional proposals – those deemed risky, owing to a relatively low probability of success, but that could lead to important discoveries. Such a scheme would create opportunities to take advantage of serendipity – a major component of scientific discovery.

Until about a decade ago, the Space Telescope Science Institute adopted a similar policy for allocating observation time for the Hubble Space Telescope. In addition, each year, the institute’s director was allotted a certain amount of discretionary time to grant to special projects that he or she deemed worthwhile. In 1995, Robert Williams used that time to take a major risk: he aimed the telescope at a seemingly uninteresting area for nearly ten days. The result was an image of more than 3,000 galaxies some 12 billion light-years away – the so-called Hubble Deep Field.

Support Project Syndicate’s mission

Project Syndicate needs your help to provide readers everywhere equal access to the ideas and debates shaping their lives.

Learn more

Likewise, closer to home, as many as half of our discoveries of new medicines have originated from accidents. For example, isoniazid was initially tested as a tuberculosis drug; iproniazid, one of its derivatives, later proved to be effective in the treatment of depression.

Space for brilliant blunders is vital to achieving the kind of creative breakthroughs that drive scientific progress. It is time for funding institutions to recognize that.