Lorin Kleinman   Reading, writing, and occasionally eating.

Human beings have an interesting dilemma.  Sanity, I suspect, hinges to some degree on believing ourselves to be right more often than not, on believing that what we see is actually there, and that what we remember actually happened.  Unfortunately, we are very often wrong, about almost everything. In Why We Make Mistakes: How We Look Without Seeing, Forget Things in Seconds, and Are All Pretty Sure We Are Way Above Average,  Joseph T. Hallinan fascinatingly plumbs the depths of our errors.

In a style familiar to readers of Freakonomics and Outliers, he gives a plethora of examples, from medicine to criminal justice to the military. We are never safe from mistakes, though this is due less to problems of intelligence and organization, and more to the inherent flaws of our approach to the world.

One facet of this approach is efficiency. For instance, if we are looking for something and conclude that it is unlikely to be there, we tend to stop looking. This has unfortunate implications for radiologists and airport screeners: in one 2002 study, TSA officers missed one in four guns.  (Screeners have a literally one-in-a-million chance of seeing a gun.)

Equally, if we believe that we know what will be there, we don’t look very closely: we skim.  Hallinan describes a misprint in a piece of music that had gone unnoticed for decades, and was not detected until an unskilled pianist played the piece. All other pianists had simply automatically substituted the correct note.  This suggests, counterintuitively, that the best proof-readers may not be experts.

And we see what we expect to see. I read of a study in which wine experts were given white wine that had been colored to make it appear red. Many of the experts failed to notice that it was white, and gave detailed evaluations of a red wine.  I explained this to a delighted waiter at Cookshop once; I wonder if he ever tried it out on the sommelier.

The list of things that we get wrong goes on and on, often driven by just the sorts of devices that have advantages in other situations. We often have trouble drawing a penny from memory or reciting the national anthem (I failed the penny test, but recited the anthem without error), we are terrible at judging risk, and we tend heavily toward overconfidence–think, for instance, of the millions of unused gym memberships we pay for every year. We also think that we can multi-task. We can’t.

Fortunately, Hallinan does not regard these problems as unsolvable, and the last few chapters of the book describe ways in which we can attempt to overcome them. Forewarned is forearmed, at least some of the time.

Readers of  Jonah Lehrer, Atul Gawande, and Malcolm Gladwell will recognize both Hallinan’s style and a number of the examples he gives, but for anyone interested in the disjoint between what we believe and what is  true–and how to tell the difference–this book provides a fascinating account.