Eight Preposterous Propositions: From the Genetics of Homosexuality to the Benefits of Global Warming

  • Robert Ehrlich
Princeton University Press: 2003. 360 pp. $27.95, £18.95

A little way into Robert Ehrlich's assault on obfuscation and unreason, your appalled eye will light on a table which reveals that more than a quarter of the population of the United States believes in witches, 41% in possession by the devil, fully a half in extrasensory perception (ESP), and no less than 45% are in no doubt that extraterrestrial beings have been stalking the Earth. (The physicist Leo Szilard said so too, but added that they are called Hungarians.) Worse still, even among the beneficiaries of a college education, only 16.5% are prepared to concede that Homo sapiens is a product of evolution, unaided by the hand of God.

Such dense fog between the ears is invariably linked to an inability to grasp that improbable events are merely manifestations of the rules of chance, and not of divine intervention. Oscar Wilde understood this ingrained disorder of the human intellect: “Man can always believe the impossible, but man can never believe the improbable,” he observed. Ehrlich has set himself the heroic task, concealed beneath his flippant title, of confronting the tide of irrationality in what is in effect a manual of scientific reasoning.

His method, originating in his earlier book Nine Crazy Ideas in Science, is to test eight quite diverse propositions, extending from the unquestionably absurd (telekinesis, or moving matter around by thought alone) to the probably valid, such as a part for genetic factors in determining sexual inclination. He grades these on a scale of 'flakiness': zero flakes implies that the proposition may well be true, and four flakes that it is unarguably nonsense. My dictionary defines 'flaky' as “adj. eccentric, crazy”, but this is not altogether what Ehrlich means by it; he conceives it as “lacking in empirical evidence or internal consistency”, thereby distinguishing it from his 'crazy ideas' in science, some of which (like practically all of Kuhn's 'paradigm shifts') could be true.

Ehrlich's longest chapter is devoted to the weighty question: “Should you worry about global warming?” After an impeccably neutral analysis of the passionate opinions on either side, Ehrlich awards it a (judiciously qualified) 'one-flake' rating — in other words, a negative answer to the question could just about be entertained. A strength of Ehrlich's treatment is that he approaches the truly preposterous theses — telekinesis and the eclipse of evolution by “intelligent design” — with a straight face. His reasoned demolition of the evidence for these aberrations is vastly more effective than the red-eyed apoplexy that seizes the average scientist at their mere mention.

But it is the final two chapters that I found the most compelling. Ehrlich is at his most incisive on the placebo effect, and on the recent assertion in a widely publicized paper that its extent has been grossly exaggerated. The arguments hinge, for the most part, on the interpretation of statistics, which Ehrlich manages to make accessible to all who will make the effort. Only at one point, isolated in a box from which innumerate readers can avert their eyes, does he set out the mathematics in full.

He makes a powerful case that many, and especially psychotropic, drugs which make extravagant profits for the pharmaceutical industry are ineffective or worse. He uncovers the weaknesses in conventionally designed double-blind trials and, both here and in his final chapter (“Should you worry about your cholesterol?”), he expatiates on the lax standards by which the industry is now regulated, and the way in which the once-proud US Food and Drug Administration has been emasculated. What is especially striking is the low confidence limit (P < 0.05) considered adequate to establish the efficacy of a new drug in clinical trials — a level that is perhaps acceptable in sociological research, but is generally considered risible in the exact sciences.

Ehrlich records the concern of one psychiatrist that if antidepressant drugs were shown to be ineffective, patients would suffer by being deprived of the benefits of the placebo effect. A striking study, published last year, revealed that in PET scans the same region in the brains of patients became active whether they were given placebos or opioid-receptor drugs. What we clearly need, then, are better placebos.

There is a short story by the US humorist Josh Billings in which a farmer discovers that his black horses are eating more hay than his white horses. Eventually he gets to the root of the matter and finds the explanation: he has more black horses than white. The confusion of mind that Ehrlich exposes to view is often close to that of Billings's farmer.

Here and there I found myself wondering whether, in striving for objectivity, Ehrlich was not putting too much weight on publications of uncertain authority in dubious journals. All the same, he has dug consistently deep and marshalled the evidence in masterly style. He is unfailingly lucid, and if his colloquial, rather jokey manner brings his book more readers, then so much the better, for the lessons that it teaches are important to us all. It upholds, moreover, the principle enunciated by an academic politician of legendary dexterity, a famous British vice-chancellor: one must always keep scientists away from committees — they are apt to change their minds in response to the evidence.