Burned by Edward Humes

 

Last trip to the library, I grabbed a handful of nonfiction. This is reported narrative nonfiction about an arson case that at least would be in with a chance of not being prosecuted today--the science of arson investigation has clarified some in the decades since, at least around what can't be proven, though different jurisdictions and different investigators will understand the meaning of that differently--and it serves as something of an indictment (heh) of the entire system and process of forensic criminal investigation: Too many of the methods haven't ever really been tested scientifically. Even fingerprint analysis has room for errors (some where between 1 in 306 cases and one in eighteen cases); other pattern-matching analysis, such as fibers or shoeprints or took impressions is worse, and don't even get me started on bite-marks. The fact that prosecutors tend to flat refuse to admit when some past prosecutor fucked up just makes me so angry I can barely see. Well-written and deeply reported, and very worth reading if it's the kind of thing you want to read about (you should probably read at least a little about this sort of thing, it's kinda important).