Reflections. II. How do new scientific ideas arise?

The death this past week of James D. Watson, the co-discoverer of the structure of DNA, signals the end of an era. Watson was nearly the last of the great pioneers of molecular biology, the men and women (overwhelmingly men given the social structure of science in the 20^th century) who created the subject in the 1950s and ‘60s. He was 97.

Yet, it also provides an occasion for thinking about creativity in science. After all, his discovery of DNA structure, along with Francis Crick, was one of the crowning intellectual achievements in biological science in the 20^th century. How do great new ideas in science come into being? Of course, this question applies to many other fields as well – for example mathematics, art, music, literature, architecture – but for this piece, we will stick with science and, in particular, biological science.

There are undoubtedly multiple ways in which new ideas are born. Many new ideas are generated following a discovery of some new fact that does not fit with the old ideas (see Reflections. I. How do ideas change in science?). Often, the conceptual path from the discovery of the new fact – either a direct observation or an experiment – to the new idea is fairly short, even obvious. Yet, even such deductions involve some deductive logic and creativity. However, other discoveries involve a longer chain of reasoning and have an aura of specialness about them as a result.

I remember something about scientific creativity I heard in 1991 that really struck me. The occasion was a scientific meeting on the island of Crete, in Greece. The meeting was meant to mark the arrival of a new sub-field of evolutionary biology in the previous decade, the 1980s, evolutionary developmental biology or “evo-devo” as it was tagged.

Evolutionarv biology as a whole had a reigning theory, from the 1940s onwards (see What is odd about the “theory of evolution”?). For decades, however, the field more or less ignored how developmental processes evolve. This lack of interest in development had seemed baked into evolutionary biology even though it was obvious the changes in shape and form (“morphology”) of mature animals and plants had to come from changes in the developmental processes that give rise to them. To the extent one can explain this omission, I think it reflected a belief that the broad outlines of evolution could be explored without getting bogged down in the details. Yet, it came to be realized that one cannot test one’s ideas about the general features of evolutionary change without getting into the messy details. In the 1980s, through advances in molecular biology, it became increasingly possible to do so and more people wanted to.

The Crete conference brought together at least 100 people to discuss where the field was then and to help chart a likely path going forward. It was mostly young scientists but there were a number of major figures too.

One of these luminaries was the French scientist François Jacob. He had pioneered the study of gene regulation but in 1991 would have been about 71. However, he was still vigorous, mentally acute and in good health (despite carrying in his body some German shrapnel from wounds acquired in WWII). I had had the privilege of interviewing him about eight years before but was, of course, eager to hear what he had to say now. His talk was about his work in mouse developmental biology but the bit that has stayed with me all these years concerned how the big insights in molecular biology, in the 1950s and ‘60s, had come about. Jacob remarked that many were generated through talk between two people at a time. In his view, creative insight often bloomed in the interactions of two people. Of course, other people were soon involved in all these pieces of work, certainly in the experimental tests. However it seems that there is something about two people talking to one another, bouncing ideas off each other, a kind of intellectual ping-pong, that can be a tremendously important element in scientific creativity. It does not happen in committees but in pairs of people, talking, looking at each other’s faces.

Indeed, for molecular biology, this was strikingly true. For the initial working out of the structure of DNA, the crucial pair had been Watson and Crick, working in Cambridge, England. A few years later, in testing the Watson-Crick idea of how DNA replicates, it had been two young scientists, Matthew Meselson and Franklin Stahl, at CalTech in California. For solving the nature of the genetic code, it had been Francis Crick and Sydney Brenner, in Cambridge again. And, for understanding the basics of how genes get turned on or off, it had been Jacob himself and his colleague Jacques Monod, working in Paris, France.

Let us look at one of these collaborations in detail. I choose that of Watson and Crick even though that might seem so well known as not to be worth another look. Indeed, it has been the subject of at least two major books, one film, and one play, as well as being routinely discussed in any history of biology in the 20^th century.¹ Yet, there is one aspect that is little known and that deserves attention by anyone interested in how the structure of DNA came to be solved.

However, let us first back up a bit. Before Watson and Crick started thinking about DNA, a lot was known about the chemistry of DNA but very little about its structure. The one certainty about the structure was that it was a polymer, consisting of a string of units called nucleotides. It was the chemistry of the nucleotides that was well understood. Each nucleotide consisted of a particular kind of sugar (deoxyribose), a phosphate and a ring compound called a “nitrogenous base” (because it had nitrogen and had the character of attracting hydrogen ions). However, there were four kinds of bases. Two were double-ring compounds, termed purines, specifically in this case, adenine (A) and guanine (G), and two single ring compounds termed pyrimidines, specifically thymine (T) and cytosine (C) in DNA. Each base was linked to the sugar molecule and each nucleotide, consisting of a sugar molecule, a base, and a phosphate group, was linked to two other nucleotides, through its phosphate and sugar groups, leading to an alternating sequence of sugars and phosphates, the so called sugar-phosphate backbone of the molecule, with the bases linked to the sugars but projecting off them somehow.²

To anyone interested in biochemistry, these facts would make DNA an interesting molecule but why should anyone else care about it? The reason is that two experiments had shown that DNA somehow carried the genetic information; it was the molecular basis of heredity.³ Watson and Crick’s first critical insight was that one could not understand biological heredity without understanding the structure of DNA. Today, this view seems obvious; in 1951-52, when Watson and Crick began their work, it was revolutionary.

In the end, there were four crucial insights that Watson and Crick had into DNA structure. The first two were quite early. First was the idea that each DNA molecule consists of two helical strands wound around each other, a double helix. (In their chemical polarity, they run in opposite directions but for this story, that is a detail.) Secondly, the bases must project inward, into the open space between the two helices, with the phosphate molecules on the outside. Third, and crucially, the bases pair in a highly specific fashion, A with T, and C with G. Fourth, given that the specific base composition of DNA was characteristic of each species, there must be no inherent restrictions on base pair composition – or sequence. Indeed, given that, it seemed quite possible that any order of base pairs might exist. In turn, that might suggest that the base pair sequence might “encode” the genetic information.

The first two conclusions came early and fairly easily. A lab in London had been doing X-ray crystallographic work on DNA and had produced several pictures of the diffraction pattern of the molecule. One of them, the now famous picture 51, was seen by Crick and he instantly deduced – he was expert in this kind of analysis – that the molecule was a double helix. The picture had been made by a woman chemist named Rosalind Franklin. She was a fine scientist but for reasons I won’t go into here, to save space, she did not leap to this conclusion. However, if DNA was a double helix, the next point, that the bases must project inward and the phosphates were on the outside followed almost automatically. The reason is that phosphates are negatively charged and to put them all inside the molecule would concentrate negative charge there. The phosphates, therefore, had to be on the outside.

It was deduction number three that was the most brilliant and provided the key to understanding how DNA might carry heredity. If you have read Watson’s memoir, “The Double Helix” (see end-note 1), you will probably remember the description of how Watson figured out base pairing. He was playing with realistic models of the bases and figured out that only A-T and C-G pairs would bond easily and when they did so, the two base pairs were exactly the same size, making possible a molecule of smooth dimensions. Wonderful! But was it really that simple?

I became aware that it might not be in a conversation I had with two older scientists, sometime in the late 1990s, I think it was 1996. Their names were Geoffrey Grigg, an accomplished Australian microbiologist, and Robin Holliday, an eminent British molecular biologist. We were in Sydney, Australia, where Geoff and Robin worked, and having dinner one evening. At one point, the conversation turned to the discovery of DNA and Geoff, who knew Jim Watson well, started reminiscing about a conversation he had had with him some years previously.

Watson mentioned that a mathematics undergraduate, John Griffiths, doing part time work at the Cambridge lab, had been looking at possible base pair interactions from a quantum physics standpoint, on a suggestion from Crick, and had deduced that A could only pair with T, and G only with C. This was before Watson started working with his models. Apparently, Crick and Watson mentally filed this fact away but, at the time, it did not bowl them over. Geoff, when told this much later by Watson was surprised to learn this important fact and asked Jim why Griffiths hadn’t been a coauthor on their famous DNA paper. According to Geoff, Watson responded with: “We couldn’t do that, he was only an undergraduate!”

I cannot confirm that last detail since all the individuals in this story are deceased. (Crick died in 2004, Geoff in 2008, Griffiths I have read died tragically young in his late 30s from cancer (like Rosalind Franklin), and now Watson is gone.) However, the basic fact of Griffith’s priority in the base-pairing idea is almost certainly true. It is given in Horace Judson’s classic book, “The Eighth Day of Creation”.⁴

What, if anything, can one deduce from this story about the nature of scientific creativity? We tend to think of major insights as involving great moments of inspiration – think of Archimedes in his bathtub or Kekulé figuring out the benzene ring --and this undoubtedly is often part of the process. After all, a new thought arises in an individual mind at one moment where it had not been seconds before.

Yet, I would suggest that there is probably often a lot of prior mental preparation, a lot of it often not fully conscious but which is real and can be significant. Crick’s realization from Rosalind Franklin’s photo that DNA must be a double helix would not have come about had he not had years of experience doing Fourier analysis of pictures from X-ray crystallography. He immediately drew on that knowledge without having to go over it. And while there is no reason to doubt the suddenness of Jim Watson’s inspired idea about base pairing while fiddling with his models, we know from Geoff Grigg’s anecdote that Watson had not forgotten John Griffith’s finding and that it was almost certainly there “at the back of his mind” as he manipulated his molecular models.

Still, these are just a few examples. It is fair to ask: does the conclusion that there is a lot of hidden, that is unconscious, mental work often underlie great discoveries, really hold up? It is impossible to know but I think so. For instance, years ago, I wrote a paper on Darwin’s creativity, titled “Charles Darwin: genius or plodder?”.⁵ The paradox of Darwin is that, by his own account, he was a relatively slow thinker yet he produced profound ideas, many of which still have great value. The conventional Idea of scientific brilliance envisages speed of thinking as its essence. Indeed, many of the men mentioned in this article fit this picture but Darwin doesn’t. He was a “plodder”, to use Crick’s word for another Cambridge scientist of note, Max Perutz. Perutz, like Darwin, was not flashily brilliant but produced great insights.

Perhaps, for some individuals, slower thinking facilitates depth of insight. If so, that capacity may well be tied up with mental processes that are unconscious. Of course, such processes would be very hard to reach and explore, probably harder than conscious ones, which themselves are incredibly difficult to understand (see Consciousness. I. Basic questions). Perhaps, however, we can look forward to new creative thinking about such beneath-the-radar mental processes. New creative thinking about the nature of creative thinking is always welcome.

1. There are two key books about the discovery of DNA. The first is James Watson’s “The Double Helix” (1968. Atheneum Press: New York.) Its subtitle is “A personal account of the discovery of structure of DNA” and that gives a hint of its problems. In a phrase borrowed from literary criticism, Watson in this account is an “unreliable narrator” in the way he distributes praise and criticism. Still, it is a gripping, enjoyable read. The other book is much more reliable and really excellent. It is “The Eighth Day of Creation” by Horace Judson (1979. Jonathan Cape: London). It is a much longer treatment and covers not just the Watson-Crick discovery but the full history of the early days of molecular biology in the 1950s and ‘60s.

2. For readers not familiar with the details of DNA, I recommend the Wikipedia article and Watson’s “Molecular Biology of the Gene” (1970. Cold Spring Harbor Laboratory Press.)

3. The first experiment to show that DNA was the molecular carrier of genetic information was carried out at Rockefeller University by Oswald Avery and colleagues and published in 1944. The second key experiment was done at Cold Spring Harbor Laboratory by Alfred Hershey and Martha Chase and was published in 1952.

4. The description of John Griffith’s contribution is on p. 139 of Judson’s book (see note 1, above.)

5. See Wilkins, A.S. (2009). Charles Darwin: genius or plodder?. GENETICS 183: 773-777. Doi.org/10.1534/genetics 109.110452.