Black Swans and the Problem of Induction

Nassim Nicholas Taleb is the author of the best-selling book The Black Swan. The book is concerned with randomness and uncertainty, and our chronic inability to accurately fathom and measure these phenomena. According to Taleb, a Black Swan event is one that is unpredictable yet has wide-spread ramifications. Not only are Black Swan events difficult to predict, but Taleb also argues that we human beings have certain psychological limitations and biases that prevent us from foreseeing these events, while also thinking that the events were perfectly predictable after they occur.

The name for the Black Swan Theory comes from history. Prior to 1697, a black swan had never been observed by anyone in Western civilization. Many people therefore believed that all swans were white and there was no such thing as a black swan. However, a black swan was finally observed in western Australia in 1697, and the fallacy that black swans didn’t exist was proven wrong.

This idea illustrates the problem of induction, one of Taleb’s primary arguments. Just because every swan observed in the west until 1697 was white does not allow you to conclude that every swan is white, or that other types of swans (like black swans) don’t exist. However, this is a mistake that many people make.

Taleb also illustrates the problem of induction with the parable of the Thanksgiving turkey. Imagine that a turkey is born on a farm. Every day, it is fed regularly and has a shelter to sleep in. For 1,000 days, the turkey has a very happy and perfectly content existence. The turkey wakes up on the 1,001st day and expects more of the same. However, that day is Thanksgiving and the turkey finds himself on the chopping block. The turkey used inductive reasoning to conclude that because every day so far had been happy, the next day must be happy as well, and his conclusion was obviously false. Thanksgiving day is a Black Swan event for the turkey because he couldn’t have predicted his fate, but it is a white swan event for the farmer. After all, he’s the one doing the chopping.

As these examples illustrate, conclusions drawn from inductive reasoning are susceptible to falsification, where one instance can nullify a wide-ranging conclusion drawn from insufficient data.