« Back

The Theory That Would Not Die: How Bayes' Rule Cracked the Enigma Code

Bayes' Theorem — A Love Story

Scott Brookhart

I took several statistics courses in both undergraduate and graduate school, so I was acquainted with Bayes' rule, though not its history. The actual theorem is a one-line algorithm that's pretty easy to understand and calculate. Essentially, it states that given two events, there is some probability that the outcome of one will affect the outcome of the other — for instance, that a prior event will provide information that can be used to describe a later event's outcome. Most of us might think this way about events in our lives, but the theory wasn't widely accepted for many years, and many leading thinkers have periodically rejected it during its over 200-year history.

Thomas Bayes, an English clergyman and amateur mathematician, first described the theorem in 1746 in an unpublished paper. After Bayes' death in 1761, a friend published the paper as an "Essay towards Solving a Problem in the Doctrine of Chances." With the essay, the theorem acquired a life of its own.

In The Theory That Wouldn't Die, author Sharon McGrayne clearly lays out the history and development of an important theorem and the precarious path it took to get to where it is today. McGrayne is a journalist, writer, and lover of science, who articulates difficult ideas in a way that the general public can understand and appreciate. Her story here describes how Bayes' theorem would lie dormant for many years, until someone encountered it and rejuvenated interest in it for awhile, after which it would once again go dormant. Those someone's included Laplace, Alan Turing, Andrei Kolmogorov, and Claude Shannon, although the theorem always had detractors, too. Not until the 20th century did it finally take hold, with important military applications such as cracking the enigma code and assisting naval searches for submarines and lost weapons as well as its current prevalence in medical and academic research.

Readers needn't have a particular interest in statistics to enjoy this book. I highly recommend it to anyone interested in science, history, and the evolution of a theorem over time. The book read as if it were a love story — for an algorithm that grew up neglected, periodically taken out for a ride but mostly sitting home alone, until at long last, it finally found its rightful place of respect and appreciation in the world.

Scott Brookhart is a software engineer living in Austin, Texas. Contact him at sbrookhart@alumni.utexas.net.


More Reviews