Book review / Lies, damned lies and statistics

Superficially, these books have something in common. Both are written by science journalists who here display rather more skill in the latter activity than the former. Both, misleadingly, have "lies" in the title and both aim to explain to the lay reader complicated semi-technical ideas in the social sciences. But neither succeeds in doing what Hawking and Penrose did for physics or Dawkins for biology.

Vital Lies, Simple Truths is written by the same author as last year's best-seller Emotional Intelligence. Is this book a rushed sequel? Alas, I am churlish enough to believe it is a prequel. There are 15 pages of references, but none goes beyond 1984. Lots of articles and unpublished papers are referenced around 1981-83, but nothing later. Did science stop in Orwell's 1984? Does this matter? Yes - vitally, as it turns out. A question absolutely central to the theme of the book, concerning false memory or "recovered memory syndrome", blew up in the early Nineties - but is not discussed at all. The accusation of sexual abuse in childhood by adult children against parents is the current issue around "vital lies", but this book mentions nothing about it.

Vital Lies puts its thesis like this: "The mind can protect itself against anxiety by dimming awareness. This mechanism creates a blind spot: a zone of blocked attention and self-deception. Such blind spots occur at each major level of behaviour from the psychological to the social". Each of the six parts mixes scientific experiment and human-interest story: an excellent formula for popularity, as the previous book showed. This makes it easy to dip into, but does not help coherence.

Goleman moves confidently from one research area to another; from neuro- psychology through psychoanalysis to micro-sociology. The concept of "trade-off" litters the first half. He argues that there is a trade-off between attention and anxiety: we do not attend to stimuli of all sorts that provoke too much anxiety.

Taking us briskly through the cognitive psychology of the late Seventies, Goleman emphasises that perception is certainly not complete and may not be "conscious". Without knowing it, we scan, filter and select information. We see what we want to see.

The book then gets rather more clinical as it considers selective memory and that old Freudian chestnut, repression. The ego, as Uncle Sigmund pointed out 100 years ago, is a great censor which controls and distorts information. We are all essentially self-deceivers. But now we have more scientific proof of the fact.

The last section becomes sociological. We move from the individual to group (family and organisation), remembering and forgetting. Curiously, three sound-bites in the concluding section tell us that personal blind spots may indeed be vital lies because they have survival value. In the face of unpalatable truths, it is fairly natural to try a little self- deception.

The author was a magazine editor, and it shows. His knowledge base is broad rather than deep and he is better at describing research findings than at critically evaluating them. He writes well, and is ever on the lookout for themes, segues and stories to support his fundamentally correct, if not new, theme.

Sanjida O'Connell's Mind Reading could never be accused of being old- hat of out-of-date. A number of references (there are 15 pages of them) specify "in press" or "in prep". Indeed, the book does deal with one of the most currently "sexy" ideas in the biological and social sciences. The subtitle, however, is misleading: there are four references to lying in the index, and none elsewhere.

O'Connell's key phrase is Theory of Mind (ToM) - the ability to read other people, to understand their emotions, beliefs and expectations, and hence to predict their behaviour. Problems with ToM have been suggested as the underlying cause of the deeply debilitating psychological illnesses of autism and schizophrenia.

The author is a television producer and ex-primatologist who worked on chimpanzees; she has also written a novel called Theory of Mind. (There are in fact more references to chimpanzees than to children in her book.) O'Connell writes well and paces the work smartly. But the book has some of the problems of television science presentation; it jumps around too much. Chapters are incoherent and there are a number of inaccuracies in the reporting of empirical work. (I checked this out with two colleagues cited by O'Connell, and they independently said the same thing.)

Once again, we have a straightforward theme which is threaded through a disparate literature. If you don't understand novels or plays, it's because you can't understand motives or intentions. Without a ToM, you can't empathise. Equally, the better your ToM, the better you can lie, cheat and love!

But the theme is too dissipated, because the issue - empathy - is itself too big. The topic is sold well for the lay reader, but with too much generalisation and emphasis on peripheral, if interesting, studies that seem to confirm the thesis. It's all too inductive trying to infer general laws and universal truths on the basis of a gamut of experimental results, personal observations and good stories.

Do I have a ToM about O'Connell's motives for writing the book? Money, fame, narcissism? The author is clearly committed to her topic, but the trouble with people who get really excited by scientific concepts is that they find the concept can miraculously explain just about everything. And that is the point at which science becomes journalism.