Book review: Brain Bugs

Authored by Dean Buonomano in 2011 (ISBN 978-0-393-07602-8).

My wife likely picked this book for me from the library shelf because of my recent interest in psychology. I have been listening to psychology lectures via CD while commuting, enjoying Robert Cialdini's columns, and relating my recollections of The Invisible Gorilla.

The thesis running through these sources is that our brains strongly resemble the brains of humans from 10,000 years ago because little natural selection has occurred since then, both because of the relatively short time elapsed (merely 500 generations) and because few modern-day humans are unable to reproduce due to the inability to survive until reproductive age. The human brains of 10,000 years ago were well adapted to life at that time. The human brains of today, being largely the same, are perhaps not so well adapted, as the world has changed substantially during that time: agriculture, the industrial revolution, the computer revolution, and now the information revolution.

Buonomano introduces us to the brain as neural network, rather than computer. The brain primarily works with associations, and he gives the following example. Answer these questions: 1) What continent is Kenya in? What are the two opposing colors in the game of chess? Name any animal. 20% of us answer "zebra" to the third question when it is asked after the first two. If the first two are not asked, less than 1% of us answer "zebra". Plainly, the relationship between zebras, chess colors and Africa influences our answer. As another example, try to memorize these words: rest, bed, dark, snore, wake, slumber midnight, dream, snooze.

This quirk of answering "zebra" is the rule, not the exception. Our ability to recognize patterns is at the heart of our humanness. Language is all about patterns, and computers, for all their wonder, are not so good at them, as anyone who has suffered through a call-center's voice recognition software can confirm. A five-year-old is much better than the strongest computer at understanding speech (at least what words were said, if not their meaning). Likewise, a two-year-old can recognize Grandma even after not seeing her for some time, and even if Grandma changes her hair and glasses. Computers struggle, and that is the basis for CAPTCHA tests.

Recalling the list of words from two paragraphs ago, do you remember the word "sleep"? If yes, check again, it was not on the list.

This facility with patterns was probably very helpful to early humans attempting to find prey and avoid predators. Precision, as is sometimes required in our modern world, was less valuable. Even today, some hunter-gatherer cultures do not employ numbers beyond 1; everything is labeled "zero", "one" or "many". As Buonomano explains, when faced with a den of snakes or surveying a herd of deer, a human able to form the concept of "many" is just as well prepared as the one who can distinguish between 42 and 43.

As a former attorney I am particularly interested in human perception and memory, as it forms the basis for most evidence introduced in courtrooms. The legal field has long recognized that testimony is potentially inaccurate owing to defects in one or more of the following: perception, memory, communication and bias. Cross-examination is designed to allow the opposing party to highlight these sources of error. Most people recognize that memory degrades after time, but I was fascinated by the experiments conducted by Elizabeth Loftus.

After learning about her work I strongly suspect my memory and the memories of others. She would show a movie and split the observers into two groups, say 20 people in each group. The movie consisted of scenes from a suburban school. There was a school bus, children milling about, but no child entered or exited the bus. She would question those in the first group, including "do you remember a child entering the bus?" Perhaps a few would answer yes. (This by itself is an interesting result.) She would not ask that question of the second group. A week later she would invite both groups back; this time, both groups were asked that question. With the second group, again a few would answer yes. With the first group, those that were "primed" with the question a week earlier, many would answer yes. The conclusion: one can influence the memory of another person about a past event just by asking about that event. I have always been ambivalent about the death penalty because so many convictions rely on eyewitness testimony.

Some people suffer injuries to their brains that cause them to think abnormally. Oliver Sacks' The Man Who Mistook His Wife for a Hat covers many such cases. Buonomano describes Capgras syndrome and phantom limbs. People who lose limbs often continue to "feel" them; the feelings can be painful and no intervention has proven effective in all cases. There is a converse condition called somatoparaphrenia: patients deny that a portion of their body belongs to them, stating calmly and bizarrely that they have never seen that limb before.

Scenario 1: a plane crashes because of a never-before-observed, unpreventable mechanical failure. Scenario 2: a plane crashes because of a terrorist attack. Which scenario generates more anger? Scenario 2, of course, but why? Terrorism has killed a few thousand people over the last decade, as against 450,000 motor vehicle accidents. Humans instinctively fear other humans and animals, and that made sense 10,000 years ago when fights for territory and survival were critical to perpetuating your genes; death from mechanical failure was unknown and death from poor health habits was at the least poorly understood. Darwin commented on this instinctive fear when describing the Galapagos Islands, which contained some species with no natural predators. Guns were unnecessary to shoot birds. He could walk up to a bird and with his gun push if off a branch.

For me, the most humorous, yet dark, examples of maladaptive brain functioning is in marketing. Whereas we have the power to reason carefully and analytically, our brains have a strong preference for shortcuts: fast, contextual analysis. Baby ducks will imprint on the first thing that moves. 99.9% of the time thing will be Mom, but experiments have shown that a man in a white lab coat will suffice, too. 99.9% of the time the brain activity from 10,000 years ago is sufficient for us, too, but marketers are keen to exploit the remaining fraction. Much has been written on this subject, so I will comment on one Buonomano includes. Monkeys tend to spend more time looking at dominant members. They will do so even at the expense of missing out on a treat. "Mirror" neurons are activated by these visual impressions. And so it is with us, explaining why Accenture showcased Tiger Woods, even though we are aware Mr. Woods probably knows nothing about business process re-engineering.