The Black Swan

A black swan was used as far back as Roman times as an example of something so rare that it didn't exist, and it was widely believed in Europe that all swans were white. Then a Dutch explorer in Australia sighted a black swan. All of the sightings and observations of white swans in Europe seemed like proof that all swans were white, but one contrary observation overturned years of shared understanding.

This is called the problem of induction, and it's the primary topic of this highly entertaining book by Nassim Nicholas Taleb. Taleb's thesis is that not only are black swans — new information or events entirely outside our expectations or predictions — more common than we think they are, they're far more influential as well. He argues that we naturally expect tomorrow to be like yesterday, build narrative explanations that establish a sense of continuity and predictability, and have great difficulty wrapping our minds around the possibility of the unexpected. But such aspects of modern life as highly interconnected financial structures, hugely increased access to information, and increases in world population are making black swans more and more likely and more and more influential. Even though most events obey the prosaic rules that Taleb calls Mediocristan, black swans have such an enormous impact when they happen that their effects can easily swamp everything else.

Taleb has a Wall Street background, so many of his examples (and much of his scorn) are drawn from financial forecasting, the stock market, complex derivatives (and this book was written prior to the impact of the mortgage crisis), and particularly the Nobel Prize in Economics. But he draws examples from all over human experience, including, unsurprisingly, 9/11, and, more interestingly, his childhood in Lebanon and the black swan of the Lebanese civil war. His comments on security and risk analysis line up very well with essays by Bruce Schneier and others: we drastically overestimate risks with spectacular narratives attached, drastically underestimate other risks (particularly ones embedded in day-to-day life), and do a horrible job at analyzing risk tradeoffs around improbable events.

This is not a book that offers solutions; rather, it's an extended attempt to open people's eyes to just how little we know and how badly we handle that. Taleb argues that we naturally and through education expect most aspects of the world to follow something akin to a bell curve, where the outliers drop off so quickly that they can be safely ignored. One example he uses is height: if you fill a stadium full of people and then want to measure their average height, missing a person or two won't matter. No single person can be so tall or so short as to throw off your measurement very far. Height is a property that we're good at reasoning about, as are many natural facts that we've lived with since we evolved from apes.

Wealth, on the other hand, is one of Taleb's examples of a more modern, information-driven, and black-swan-susceptible measurement. If you are attempting to measure the average wealth of that same stadium full of people, it suddenly matters a great deal whether you account for everyone. If Bill Gates is in the stadium, he could easily have more money than everyone else in the stadium put together; by himself, he has a significant effect on the outcome. Wealth is, in Taleb's terms, a property from Extremistan, describable with a power rule (20% of people may have 80% of the wealth, and that may be true of each subset of the population including the richest 20%), self-perpetuating (wealthy people tend to become more wealthy), and heavily influenced by outliers. Most of The Black Swan is devoted to an inventory of all the ways in which our reasoning breaks on such quantities and how much we're in denial about that.

This is not an academic paper by any stretch. It's a cross between a popularization and a diatribe, and Taleb doesn't hold back his scorn and contempt for innumerable people, techniques, and institutions that he sees as hopelessly ignorant. The Nobel committee comes under heavy fire, as does most of the field of economics, anyone who does long-range economic forecasts, anyone who teaches the bell curve, and anyone who thinks that black swans are about the sort of randomness you can study in game theory and statistics. What the book loses in convincing somberness it makes up for in sheer entertainment. Taleb comes across as arrogant, self-assured, and brilliant, and while I'm not sure I'd want to spend time with him in person, he's a highly diverting writer. If you like reading sometimes-obnoxious rants by clearly intelligent people, this book is a blast.

The Black Swan is fascinating entertainment, but a bit hard to wrap one's mind around when it comes to practical use. It's a book about the limits of knowledge, and nearly all of its conclusions are negative. It's a book about what we don't know, can't know, don't understand, can't analyze, haven't developed tools for, and are deluding ourselves about. Taleb is highly persuasive and left me wanting to act on his conclusions, but making the leap from the theoretical uselessness of planning to dealing with the practical requirements of day-to-day project timelines and scheduling is hard. Taleb's own approach of seeking out high-risk opportunities and increasing his exposure to positive black swans makes sense both in his analysis and for his personality, but doesn't seem like a reasonable approach for the whole world. Still, I'd rather be thinking about these limits than not. Just be prepared, as a reader, for a bit of frustration at trying to dig actions out of the ideas.

The first half of this book is the strongest, after which Taleb tends to repeat himself a bit, but anyone with statistics background should stick through the final chapters for a thorough defenestration of the usefulness of the bell curve. Contrary to Taleb's loudly stated assertions I think people with a bit of statistical background will get the most out of this book (and will know where to take it with a grain of salt), but Taleb is a wonderfully entertaining writer even if the data models aren't your thing. I laughed out-loud while reading The Black Swan and then thought about this book for days after reading it; that's a rare combination in a non-fiction book about epistemology. Highly recommended.

Rating: 9 out of 10

Reviewed: 2008-04-09