Tuesday, July 28, 2015

Pattern recognition and murder mysteries

I recently completed The Oxford Murders by Argentinian mathematician Guillermo Martinez. From the blurb
A paperback sensation in Argentina, Spain, and the United Kingdom, "The Oxford Murders" has been hailed as "a remarkable feat" ("Time Out" London) and its author as "one of Argentina's most distinctive voices" ("The Times Literary Supplement"). It begins on a summer day in Oxford, when a young Argentine graduate student finds his landlady-an elderly woman who helped crack the Enigma Code during World War II -murdered in cold blood. Meanwhile, a renowned Oxford logician receives an anonymous note bearing a circle and the words "the first of a series." As the murders begin to pile up and more symbols are revealed, it is up to this unlikely pair to decipher the pattern before the killer strikes again.
The writing and plotting are both good. What I especially enjoyed was the interweaving of the history of mathematics (the Pythagoreans in particular) and the challenges of pattern recognition with a standard murder mystery. Martinez does a wonderful job, I think, of leaving you hanging till the final pages with three distinct possible resolutions to the mystery. Each layer of reveal changes the structure and probability of each of the hypotheses, all of them remaining viable. Sometimes the new information tilts the probability more towards one solution than towards another, but they are all feasible. Even knowing the outcome in the end, the other two hypotheses remain near credible.

This approach yields a larger perspective of the fragility of knowledge, pattern recognition and interpretation. Every working hypothesis, however useful or probable it might seem, remains contingent on yet undiscovered knowledge.

Much more than your run of the mill murder mystery. This one has cognitive legs.

No comments:

Post a Comment