Between the lines

The dark side of statistics

The more pairs of edible underwear that a nation’s people eat, the longer citizens of that nation will, on average, live. So does that mean that edible underwear makes you live longer? Not at all. Rather, this is a classic case of correlation not causation: longer lifespans and high edible-underwear consumption are both linked to a country’s wealth, but not to each other. However, some examples of what science journalist Charles Seife calls “proofiness” are harder to spot. In his book of the same title, Seife describes how proofiness, or “the art of using bogus mathematical arguments to prove something…is true, even when it’s not”, has infected science, business, journalism and politics. According to Seife, scientists who come up with absurd formulae for the perfect bum are guilty of a form of proofiness called “regression to the Moon”, meaning that they are disguising nonsense with mathematics to make it seem respectable. He also pokes fun at pseudoscientific statements made by manufacturers, including L’Oreal’s claim that Extra Volume Collagen Mascara gives lashes “12 times more impact”. (“Perhaps they had someone blink and listened to how much noise her eyelashes made when they clunked together,” he adds drily.) UK readers may detect some parallels with Ben Goldacre’s “Bad Science” columns in the Guardian newspaper; certainly, Goldacre’s fans will find plenty to enjoy in the book’s opening chapters. Later, though, Seife sets his sights on rather different targets. After devoting one chapter to risk mismanagement and another to misleading opinion polls, he arrives at the book’s polemical heart: proofiness in voting. Whether it pops up in inaccurate census data, gerrymandered election districts or the surprising impossibility of counting votes accurately, Seife is convinced that vote-related proofiness is a serious threat to democracy. His rhetoric can get overheated and he offers few solutions, but his underlying message is both fascinating and disturbing.

• 2010 Viking Adult £16.70/$25.95 hb 304pp

A more-connected mess

If you think meetings are a waste of time, perhaps you should think again. After all, an observational study of molecular biologists found that most of the group’s important ideas emerged not as individual “eureka moments” at the microscope, but during their regular lab meetings. Such anecdotes feature prominently in Steven Johnson’s book Where Good Ideas Come From: The Natural History of Innovation, in which he describes how several factors – from hunches to error and, especially, networks – influence the innovation process. Full of unlikely links and a dizzying array of anecdotes about inventions, scientists and history, Johnson’s book is an excellent example of the sort of networked, connected thinking he advocates. It is unsurprising to learn that he wrote it with the help of a computer program called DEVONthink, a searchable repository of quotes, fragments and hunches that functions as Johnson’s 21st-century version of a Victorian “commonplace book”. But the downside of this type of thinking is also on display in the book: once you get past the tangled web of surface connections, what lies beneath is surprisingly insubstantial, and the results are often less than satisfying. There was a lot more to the controversy surrounding the discovery of DNA, for example, than the fact that Watson and Crick were interdisciplinarians who took long, rambling coffee breaks, while Rosalind Franklin was a more conventional biophysicist – but you will not find it in Johnson’s book. Moreover, the dark side of innovation is barely mentioned here, save in a footnote in which Johnson admits that the innovation-fostering atmospheres of big cities and the Internet also seem to make criminals and spammers work more efficiently. Ideas may want to be free, as Johnson claims, but the question remains: free for whom, and for what purpose?

• 2010 Riverside $26.95 hb 336pp

Lawless but not flawless

It must be a temptation, after retiring as a physicist, to go beyond one’s research specialism and write a book outlining your “philosophy” of science and the scientific method. The latest offering in that mould is Lawless Universe by Joe Rosen who was, until retirement, a theorist at the universities of Tel Aviv and Central Arkansas with a particular interest in symmetry. After ploughing through the nature of science, theory and the difference between objectivity and subjectivity, Rosen then comes to the meat of the matter – his view that science, despite its successes, can only explain part of what the universe is about. So cosmology, for example, is metaphysics, not science, because we cannot run reproducible experiments on new universes; cosmology lets us describe the universe, but not explain it. Moreover, as quantum theory cannot be a literal description of objective reality, then, in Rosen’s view, objective reality must be mostly hidden from us. Reality, in other words, transcends nature and surpasses human understanding. Quite how scientific laws can then exist in an intrinsically orderless universe is a bit unclear, but Rosen is a genial enough guide through some mind-bending stuff.

• 2010 Johns Hopkins University Press £39.00/$75.00 hb £15.50/$30.00 pb 184pp