杏吧原创

(Image: Patrick George)

IF YOU want to make petunias a deeper purple, you could just add an extra pigment gene, right? Wrong: the extra gene turns the flowers white. This surprising finding was made independently in the early 1990s by two plant biologists, Richard Jorgensen in the US and Joseph Mol in the Netherlands. Neither dismissed the finding as an error. They suspected they鈥檇 found something big, and they had: an entirely new way in which cells regulate gene expression, now called RNA interference. RNAi has since been the subject of a Nobel prize, has saved lives and promises to save many more.

This is by no means the only example of good luck in science. Percy Spencer, an engineer at the US company Raytheon, was working on a radar set in 1945 when he noticed that a candy bar in his pocket was melting. That observation led, two years later, to Raytheon introducing the first commercial microwave oven. In 1976, chemist Shashikant Phadnis鈥檚 boss asked him to test a chlorinated sugar compound being studied as a potential insecticide. Phadnis misheard it as a request to 鈥渢aste鈥� the stuff 鈥� a scary mistake to make in his line of work 鈥� and found it was extremely sweet. . Viagra was a drug proving not so effective for heart conditions before someone noticed an interesting and highly marketable side effect.

(Image: Patrick George)

Examples like these show that chance plays a role, sometimes a dramatic one, in the progress of science. Yet how much do we really know about its contribution? Its influence would be easier to gauge if we could pin down how it shows up: is it like buying a winning lottery ticket 鈥� something that can happen to anyone 鈥� or was Louis Pasteur right to say that 鈥渃hance favours only the prepared mind鈥�? At least one academic thinks not only that Pasteur was correct, but that it is possible to train minds to be receptive to the subtle signs of chance. This September he will launch a course to do just that.

(Image: Patrick George)

Opinions differ widely as to how frequent a part chance plays in science. 鈥淭here are not so many stories about serendipity. Basically, you have a couple of dozen, but in the scientific literature over the last 200 years there are so many discoveries from just plain hard work,鈥� says , an innovation researcher at the Arison School of Business in Herzliya, Israel. 鈥淚f you tried to assess the ratio between serendipity-based discovery and not, I would say less than half a per cent were the result of serendipity. But we like these stories.鈥�

Others think chance鈥檚 role is more significant. 鈥淓very decent idea I鈥檝e ever had, I had no idea about until I started doing the research and it didn鈥檛 turn out the way I expected,鈥� says , a sociologist of science at Cardiff University, UK. If we underestimate the good-luck factor, it could be to do with scale. 鈥淚 would think little surprises are there often, and big surprises are rarer,鈥� says , a social psychologist at the University of Virginia in Charlottesville.

One reason for the divergent views is the difficulty in defining chance. All of life, after all, is a walk down branching paths, and the direction at each fork often hangs on chance events: having an inspiring science teacher in school, an office mate who happens to know a useful tidbit of information, an experiment that improbably works out well. All of this involves chance; but it doesn鈥檛 necessarily mean discoveries happen by chance.

One of the hottest areas of neurobiology, for example, is optogenetics, which allows researchers to control the behaviour of groups of neurons with great precision. While at Stanford University in California, discovered a key technique in the field, the use of light-sensitive proteins from algae to trigger electrical activity in neurons. He and his co-workers (already like-minded 鈥� the first stroke of luck) had been thinking for years about using light to control neurons. Then they stumbled across the algal studies (more good luck) and decided to try inserting the genes responsible into mouse cells.

鈥淚t kinda worked on the first try,鈥� recalls Boyden, now at the Massachusetts Institute of Technology Media Lab. 鈥淲ho would have known that these molecules from algae, which are very different organisms, would work in neurons? That was also serendipitous.鈥� As they later learned, they were even luckier than they knew: the algal protein requires another molecule to work properly, and mammalian brains just happen to produce it for an unrelated reason.

Even so, serendipity was only half the story. Controlling neurons is an idea Boyden and his colleagues were keen on; in Pasteur鈥檚 parlance, their minds were 鈥減repared鈥�.

Perhaps the most iconic example of chance in science is Alexander Fleming鈥檚 discovery of penicillin. In 1928, a stray fungal spore landed in a discarded bacterial culture in his lab at St Mary鈥檚 Hospital, London. When Fleming looked at it weeks later, he saw a ring around the growing fungal colony where something had killed the bacteria nearby. That something was eventually identified as penicillin.

Yet Fleming鈥檚 finding did not pop out of a vacuum. Other scientists over the preceding century, including Pasteur, had noticed that moulds inhibit bacterial growth. Fleming himself had spent years looking for bacteria-killing compounds and had already found one 鈥� lysozyme, an enzyme he isolated . Fleming鈥檚 prepared mind connected the dots, but even so, it was another decade before other researchers, Howard Florey and Ernst Chain, figured out how to turn the mould into a drug.

Discoveries like these are often called 鈥減seudo-serendipity鈥� 鈥� the scientists knew what they were looking for but found the answer in an unexpected place. The writer Arthur Koestler vividly described such finds as 鈥渁rrivals at the right destination by the wrong boat鈥�. Taken to extremes, this approach can pretty much remove the element of chance from discovery. The inventor Thomas Edison, for example, tested hundreds of materials before he found the right filament for his electric light bulb, and pharmaceutical companies now systematically screen hundreds of thousands of substances looking for new drugs. When such an 鈥淓disonian materials dragnet鈥�, as Gorman puts it, turns up something useful, that鈥檚 a testament to hard work more than luck, he says.

In contrast, true serendipity happens when researchers stumble across something entirely unexpected, as in the discovery of microwave heating or Sucralose. Here luck plays a much more obvious role 鈥� although every case still needs an alert observer to notice the anomaly, not discount it as an error and turn it into a useful result.

Some examples, though, fall in between. Take the case of the scientist at the chemicals giant 3M who was trying to create a super-strong adhesive but ended up with a super-weak one. Years later, a colleague decided it was just the thing to stop place markers falling out of his hymn book in church. That inspiration spawned Post-it notes.

(Image: Patrick George)

This sort of accident turns out to be fairly common in the annals of innovation. When Goldenberg studied the origin of 200 important inventions, he found that in about half of the cases, the old saying had it backwards: invention was the mother of necessity. 鈥淔irst they found the invention, then they discovered the need,鈥� he says. That makes the final product not exactly an accidental discovery. It鈥檚 more a matter of finding the best way to play the hand you鈥檝e been dealt.

鈥淚t鈥檚 much easier to find a function for an existing form than the other way around,鈥� says Goldenberg. 鈥淧eople are very creative when you have a form.鈥� He points to the example of Vaseline, which has its roots in a dark sludge left over from oil processing. Only when chemists began looking for an application did they discover they could use the purified jelly to help burns heal.

Luck clearly helps some technologies bloom, but its impact on the broader world of scientific discovery is unclear. No one seems to have made a systematic survey of scientific breakthroughs to measure how often chance plays a large part.

Indeed, such a survey may be almost impossible to do properly, says , a psychologist at the University of California, Davis, who studies creativity. Scientific papers may not mention what inspired their findings. Besides, chance may be inextricably intertwined with hard work, making it difficult to weigh the relative contribution of each. 鈥淓ven if we accept Newton鈥檚 falling apple experience as valid, how much of his Principia should be credited to serendipity?鈥� he asks.

Perhaps the most direct attempt to quantify scientific serendipity came two decades ago, when Juan Miguel Campanario of the University of Alcala, Spain, surveyed 205 of the most highly cited scientific papers of the 20th century and found that 17 of them, or 8.3 per cent, contributing to the findings. This probably underestimates the true frequency, however, seeing as not every author is likely to mention their good fortune in print.

Even if there鈥檚 little certainty about how common serendipity is in science, there is broad agreement that more of it is a good thing 鈥� if only because it leads to more original discoveries. 鈥淚f you鈥檙e working on something where all you have to do is be smart and work hard, chances are somebody鈥檚 already found it,鈥� says Boyden. 鈥淪o we鈥檙e often trying to do things to deliberately encourage serendipity.鈥�

Boyden has made something of a cottage industry out of wooing Lady Luck. This autumn he plans to teach a course at MIT on nurturing serendipity, in which he will ask each group of students to systematically set out to revolutionise one area of science. 鈥淚 think we鈥檝e learned enough now about how to orchestrate serendipity that maybe we should teach it,鈥� he says.

A starting point for having a good idea is to list all possible ones (Image: Image Source/REX Shutterstock)

Boyden鈥檚 first rule for making your own luck in research: list all possible ideas to pursue. That鈥檚 not as silly as it sounds, he argues. The trick is to subdivide the universe of possibilities into either/or options, and do it over and over again. If you鈥檙e looking for a novel way to image the brain optically, for example, you could either detect photons within the brain or wait for them to leave the brain and detect them outside. If you鈥檙e doing it within the brain, you could use either active electronics or a passive detector. And so on. He calls this approach a 鈥渢iling tree鈥� because it branches like a tree and covers the entire 鈥渋dea space鈥� like tiles on a floor.

Blue sky鈥檚 the limit

In effect, it鈥檚 an Edisonian dragnet for ideas. 鈥淵ou can subdivide into smaller and smaller categories, but you never lose any possible ideas. At the very ends of these branches are things you could try out.鈥� That step is where serendipity might appear.

Boyden鈥檚 second tip is to range widely. His own research group includes engineers, physicists, neuroscientists, chemists, mathematicians and more. This diversity increases the odds that someone will make an unexpected conceptual connection. In the same vein, it鈥檚 good to work on more than one thing at once, as this also boosts the likelihood of cross-pollination. This was a key source of Thomas Edison鈥檚 creativity, for example. In a study of the , Simonton found that the more subjects he was working on, the higher his output of patents.

A more controversial way to encourage serendipity, especially discoveries that open whole new fields of science, is simply to find the smartest, most creative thinkers and give them unrestricted funding to get on with it.

That鈥檚 what used to happen at legendary research centres like Bell Labs, and still happens to some extent at Google, for example, which allows its engineers to spend 20 per cent of their time on side projects. Back in the 1980s, oil giant BP also funded a blue-skies research initiative with the goal of seeking out the very best scientists and funding them with no strings attached. 鈥淚 had 13 years of freedom at BP,鈥� recalls , who ran the programme and is now at University College London. 鈥淲e had 10,000 applicants and I picked just 37,鈥� he says. 鈥淔ourteen of those won major breakthroughs.鈥�

That鈥檚 a lesson funding agencies still need to heed, says Collins. 鈥淚t鈥檚 difficult to have a policy to encourage serendipity,鈥� he says. 鈥淏ut it鈥檚 not difficult to have a policy to discourage it.鈥� Winning research grants is now so competitive 鈥� with just 10 per cent of applicants getting funded in many cases 鈥� that researchers have to play it safe and go after results they know they can achieve, he says. More adventurous proposals, those that might stumble across something altogether new, tend to be too risky to gain funding.

In essence, today鈥檚 system is a self-fulfilling prophecy: it doesn鈥檛 believe in chance and so chance discoveries seldom happen. Yet, with some enlightened thought 鈥� and a little bit of luck 鈥� that could be reversed.

New 杏吧原创鈥榮 latest book Chance is out on 5 November and available to pre-order on .

This article appeared in print under the headline 鈥淕et lucky鈥�