In his book Brilliant Blunders, Mario Livio offers a detailed and fascinating examination of major errors made by five great scientists – Charles Darwin, Linus Pauling, Lord Kelvin, Fred Hoyle and Albert Einstein – as they sought to understand the evolution of life on Earth, the evolution of the Earth itself and the evolution of the universe as a whole. The stories of how these blunders came about, and what happened next, are extremely well researched, and they shed a welcome, informative, entertaining and sometimes new light on science as a deeply human activity. They also pass my private test – that non-scientific friends and relatives should be able to read and enjoy them – with flying colours.

Livio’s stories do not, however, always support his contention that blunders “acted as catalysts for impressive breakthroughs”. In four of the five cases (Darwin is the exception, and Einstein is a partial exception), the scientist concerned simply dug his heels in, refused to accept that he had made an error and was overtaken by other scientists working along different lines. Far from catalysing progress, the errors were dead ends, and the views of their proponents became marginalized.

Pauling, for example, hung on to his abortive three-chain model for the structure of DNA and kept trying to rescue it with minor tweaks. Its obvious failings did not stimulate him to investigate substantive alternatives. Nor did Pauling’s model really catalyse Watson and Crick’s development of a two-chain model, except in the sense that it stimulated Watson’s competitive instinct. The catalysts that really mattered were Rosalind Franklin’s discovery of the “B” structure of DNA in the presence of a large amount of water, along with Crick’s understanding of what its X-ray diffraction pattern meant. As Livio points out, both were necessary. By 1951, he writes, a rival team of researchers at Leeds, Elwyn Beighton and William Astbury, “had excellent X-ray photographs of the B form [but] unfortunately, neither of them was familiar with how a helix would appear in X-ray photographs. Just like that, the Leeds lab missed a chance to play a significant role in the DNA story.”

The book is full of such informative sidelights on well-known stories. Fred Hoyle, for example, introduced the term “big bang” in a BBC radio broadcast, but according to Livio’s research, he used it as a term of description, not of disparagement as is commonly supposed. Hoyle hung on stubbornly, however, to his belief in a “steady-state” model of the universe, and a consequent need for the continuous creation of matter at a rate that precisely compensates for the dilution caused by observed cosmic expansion. As evidence in favour of a “big bang” theory continued to accumulate, Hoyle used increasingly complex logical convolutions to try to fit the new data to his theory. As a result, he was sidelined by the mainstream cosmological community, and his opinions were no longer taken seriously.

The story of Lord Kelvin’s calculation of the age of the Earth at around 100 million years, using Fourier’s new theory of heat transfer, is also well known, as is Kelvin’s ongoing argument with the biologist T H Huxley on the subject. Kelvin got it wrong mainly because he did not allow for convection in the Earth’s mantle. When his former pupil, John Perry, introduced this factor into the calculations, he found that the calculated age could be as much as three billion years, which was much more in line with the geological evidence being adduced by Huxley.

The discovery of radioactive elements such as radium suggested another possible means of heat production in the Earth’s core, but although their disintegration did not turn out to be a significant factor, radioactivity contributed to the debate in another way. Livio tells a nice story about an encounter between the Canadian geologist Frank Dawson Adams and Ernest Rutherford, who was, at the time, carrying a piece of black rock. “How old is the Earth supposed to be?” asked Rutherford. Adams answered that several methods had given an estimate of 100 million years. Rutherford commented quietly, “I know that this piece of pitchblende is 700 million years old,” and walked on.

Kelvin stuck to his guns and, as with Hoyle, his opinions on the subject became marginalized. Nevertheless, Livio has a case when he argues that Kelvin’s error (unlike Hoyle’s or Pauling’s) catalysed the advance of science, since it “completely transformed geochronology from vague speculation into an actual science, based on the laws of physics”.

Livio is an astrophysicist at the Space Telescope Science Institute in Maryland, US, and his insights and knowledge are particularly apparent in the chapters on his own field. His meticulous research reveals, for example, that Einstein is unlikely ever to have said that the assignment of a non-zero value to the cosmological constant was his “biggest blunder”. Most probably this widely quoted phrase was an invention of George Gamow, whose interaction with Einstein the author examines in some detail. Livio argues that, in any case, Einstein’s real blunder was in returning to his initial assignment of a value of zero, whereas the discovery that the expansion of the universe is accelerating now suggests that a non-zero value is necessary.

The biggest blunder of all was made by Charles Darwin, who failed to see that his mechanism of natural selection was incompatible with the contemporary belief that inheritance occurs by blending the characteristics derived from both parents. As the Scottish engineer Fleeming Jenkin quickly pointed out, such a mechanism would swamp any advantageous variation within a few generations. Only when Mendelian heredity, and its genetic basis, were discovered and accepted many years later was this problem resolved. Even so, Darwin’s proposed mechanism of natural selection prospered and survived.

The real message of this vastly entertaining book is that scientists often make major blunders when they try to reconcile new ideas with established paradigms. Scientific insight can be a fickle jade but, as Darwin’s example shows, it can be best (unwittingly or otherwise) sometimes to have the insight to ignore such paradigms. But then, you probably need to be a Darwin to get away with it.