"Math On Trial" by Leila Schneps and Coralie Colmez, is a recent volume out from a mother/daughter mathematics team. It piggybacks on the current trendy interest in statistical-and-probability analyses of human affairs. Anyone with a penchant both for the 'true crime' genre of reporting, and for math, will likely enjoy this volume, which combines a look at various famous true crimes with the mathematical evidentiary material brought to bear in such cases...
Unfortunately, I'm not one of those people who takes much interest in crime reporting, so for me the volume did not hold undue interest. It covers ten notorious diverse cases (from murder to gender discrimination), but most of them are decades old, or even from over a century ago. A volume of more modern cases, say from the last dozen years, would have made for a more interesting and powerful read to me (...the Amanda Knox murder and Bernie Madoff swindle cases are among the more recent ones addressed in the volume). My own interests run more toward current events than history, and so the reaching back in time wasn't particularly effective, but other readers may find these historical accounts more keenly interesting. Each chapter focuses on a different kind of mathematical error (rather than the same sort of probability error occurring over and over) which is probably one reason for the somewhat odd selection of cases employed here.
At times the focus on math seems to oversimplify the multitude of legal issues and details involved in a given case, while at other times the mathematical material seems a bit buried under the legalities and details of a case. Still, the basic underlying point that math is often misused/misinterpreted in the courtroom can't be denied, and certainly anyone who doesn't already know that should read a book like this, because it is clearly a message worth spreading.
The book ends on the question of whether mathematical evidence should even be allowed in the courtroom, given how easily it is manipulated before a naive jury or judge. The final "Conclusion" chapter discusses Laurence Tribe's 40-year-old "beautifully argued, passionate" article on the subject, before noting that a lot has changed in the four decades hence, especially with the routine use of DNA forensics. Of course DNA evidence can bear problems as well, if not carefully administered, but is also a boon to those falsely-accused who may be exonerated through its use. Lastly, the authors acknowledge the more recent emphasis on Bayesian analysis and the attempt to set standards for its inclusion in the courtroom as well.
This volume didn't happen to grab me, but those of you more enamored of true-crime fare, may find it an interesting and worthwhile read, and its ultimate message that the public needs to be better educated as to the uses and abuses of statistics in legal matters is an important one.
No comments:
Post a Comment