Theory has it that unlike religion, politics or journalism, science has built-in machinery for getting things right. Peer review ought to sort out anything that is obviously wrong, while the actual publication of anything untrue is swiftly followed by rebuttal supported by experimental evidence.
So far so good. But this book, by a mathematics professor at the University of Western Ontario, shows that in plenty of cases things do not work out so smoothly. The cases he documents do not involve deliberate fraud. He cites instances in which discoveries were announced that never were, experiments were started that would have been better left alone, or linguistic misunderstandings mistaken for scientific insight.
One common theme is that anything that subverts the normal process of peer review and publication tends to end in tears. An example is the case that opens the book, that of the N-rays "discovered" by Rene Blondot in the early years of the 20th century.
Blondot's discovery was not about truth but about la gloire. Herr Dr Rontgen had found X-rays, and French physics needed something to allow it to look the Boche in the eye. The N-ray tale reveals that self-deception is possible even in a laboratory full of equipment meant to guarantee objectivity. Blondot was able to detect the rays, view their spectrum and measure their properties. In 1904, however, a number of non-French scientists. perturbed by their inability to find N-rays, despatched Robert Wood to Blondot's laboratory in Nancy, where it soon became obvious that Blondot was deceiving himself about the existence of the rays, although he still got an award from the Academie des Sciences.
Apart from the amour propre of French science, the N-ray affair left no wounded. The same cannot be said of other examples cited by A. K. Dewdney. When rotten science interacts with people instead of with atoms, somebody is likely to be hurt.
In the same year Blondot was discredited, the French education ministry asked Alfred Binet to devise a test that would allow students with learning problems to be identified. So far so altruistic. He warned often that the results would not yield any sort of general measure of intelligence, but on his structure an edifice has been built which allows people who get 140 in a test to think they are cleverer than people who get 135. The tests, as Americanised by H. H. Goddard, were let loose on would-be immigrants who spoke next to no English, and showed that (for example) 87 per cent of Russian applicants for immigration were "feeble-minded", an excuse for them to be deported.
A present-day successor of Goddard, says Dewdney, is Phillipe Rushton, an American scientist who has updated the theory to allow for the present unease in the United States over the economic success of the Pacific Rim nations. Now, it seems, white folks are still comfortingly ahead of black ones, brain-wise, but fall behind the mongoloids. By contrast, the breeding capacity league table runs the other way.
Here Dewdney mocks attempts to apply the r and K nomenclature - used to describe the reproductive tactics of different species - to different "races" of humans. The idea is that K species like elephants or whales have a few offspring and look after them well, while r species have many - thousands, in the case of some insects - and do not worry when most of them die. We are definitively a K species and there is no point trying to prove that different climate zones cause different patterns of human reproduction. Indeed, the original idea, as applied to animal populations, implies an outcome opposite to that Rushton suggests. Researchers including E. O. Wilson proposed that r animals would occur far from the equator, where there is a risk of very bad weather causing catastrophic casualties, while in Rushton's theory reproduction enthusiasm is at fever pitch among the black races of the equatorial lands.
Rushton is also an enthusiast for the Bell Curve notion of human intelligence, whereby measures of cranial size show interracial differences in mental capacity. There is already one huge assumption here, that cranial capacity has a direct link to intellectual power. But even if it does, Dewdney shows, the overlap between different human groups is massive compared to the differences between them, which do not allow any conclusions to be drawn about innate ability.
After this tragedy, it is almost light relief to turn from Ellis Island to the open spaces of Arizona and the fiasco of Biosphere 2, out of which, at last, some good is now coming. You recall the tale: eight brave and photogenic scientists locked away for two years in a huge sealed greenhouse to see whether a viable ecoystem could be created. It could not: for example, oxygen had to be imported and it turned out that a clandestine carbon dioxide scrubber had been installed to keep the gas under control.
The problem with Biosphere 2 was in its origins. It sprang not from the dull old world of peer-reviewed science, but from an eccentric sect looking for technology to permit the colonisation of Mars. When their enthusiasm met up with the chequebook of Ed Bass, a Texan oil billionaire, the project was on, whatever real ecologists thought. Now it is turning into a centre for valid science under the guidance of Columbia University, although it will never produce as much good research as the $150 million it cost would have bought if spent on academic ecology.
This fiasco attracted much public attention - but nothing in the history of science compares with the egg-on-face count of the events that give Yes, We Have No Neutrons its title, the theatrical tale of cold fusion. Here, at least, the progenitors of the idea were genuine scientists, in this case physical chemists, and the idea they were pursuing, while speculative, was by no means cranky.
Their experiments showed that heat was being produced but did not find enough neutrons, a sign that fusion is occurring. Ordinarily, a further experiment would have shown that there was some other explanation for the heat, and a bit of scientific progress would have been made. But this time, Martin Fleischmann and Stanley Pons published too early, because of perceived competition from another researcher, and were left to look foolish when nobody could replicate their results.
Some of the bad science Dewdney cites has run on for decades, such as the work of Sigmund Freud or the search for extraterrestrial intelligence. But in the latter case he may have underestimated the changes in the project since Project Ozma in the 1960s and the crazed anthropomorphism of the Drake Equation, which purported to reveal the number of habitable planets per galaxy. Seti is a speculative and much-mocked venture, but it just might produce a discovery of unimaginable importance.
These cases show that bad science can get funded (although often through odd channels) and can have real effects and influence. Most of the time, the processes that are meant to guarantee scientific truth actually work. But do not be surprised if new outbreaks as bogus as cold fusion recur in future. And the book is worth a read, although you might not buy it at Pounds 17.99 for a pretty slender volume.
Martin Ince is deputy editor, The THES.
Yes, We Have No Neutrons: An Eyeopening Tour Through the Twists and Turns of Bad Science
Author - A. K. Dewdney
ISBN - 0 471 10806 5
Publisher - Wiley
Price - £17.99
Pages - 180