Imagine: How Creativity Works, by Jonah Lehrer. The worst thing I can say about it is that it seems to promise too much. That is, it seems to suggest that neuroscience has made great strides in understanding creativity. More like baby steps, is what I end up taking away.

But I highly recommend the book. He writes very well and uses examples that appealed to me. He cites interesting studies, and I learned a lot from the book.

Tyler Cowen will appreciate all of the advantages Lehrer cites for cognitive diversity. ADHD can contribute to creativity. So can bipolar disorder. Somewhat counterintuitively, so can being on the autism spectrum. Lehrer also cites studies showing that travel–or even thinking about distant places–increases creativity.

My favorite chapter is called “urban friction,” on the role of cities in promoting innovation. The chapter becomes particularly interesting when he introduces us to Geoffrey West, of the Santa Fe Institute.

According to the equations of West and Bettencourt, every socioeconomic variable that can be measured in cities–from the production of patents to per capita income–scales to an exponent of approximately 1.15…This means that a person living in a metropolis of one million should generate, on average, about 15 percent more patents and make 15 percent more money than a person living in a city of five hundred thousand….even when the numbers are adjusted for levels of education, work experience, and IQ scores…

even when the San Jose region was mostly walnut and apricot farms, and decades before it became the center of Silicon Valley, the area still produced an abnormally large number of patents per capita…

The vertical culture of the Boston tech sector existed in stark contrast to the horizontal interactions of Silicon Valley…

Martin Ruef, a sociologist at Princeton…discovered a small subset of businesspeople who cultivated a large number of weak ties and unexpected friendships…businesspeople with entropic networks full of weak ties were three times more innovative than people with small networks of close friends…

researchers at the University of Michigan…brought groups of people together and had them play a difficult cooperation game…The groups meeting in person quickly solved the problem, finding clever ways to cooperate. The electronic groups, in contrast, struggled to interact

Lehrer, taking his cue from West, asks why corporations are relatively ephemeral and cities are relatively long-lived. Also, why do cities scale up elegantly, while corporations eventually become less effective as they get larger?

The West-Lehrer answer is that corporations are too closed while cities are more open. My answer would be that corporations are necessarily goal-oriented, top-down institutions. Cities are emergent orders, more akin to markets. You could not simply “open up” a corporation to completely decentralized management. Middle managers would end up making large bets with other people’s money. A completely decentralized corporation would, ironically, start to look like Steven Chu’s Department of Energy, with its poorly-aligned incentive structure. See what I wrote fourteen years ago.

The theme of the productivity of cities is much discussed these days. McKinsey has a new study, which proclaims

The United States has a broader base of large cities than any other region, and that explains their greater economic clout.