Modesty About the System's Ability to Find Truth

The big criminal justice related cases, at the moment, are the killing in Ferguson and Rolling Stone's UVA gang rape story.  It is remarkable that in both cases, a key lesson is that witnesses cannot be trusted.  Of course, it has long been known that eyewitness identifications are unreliable.  But it is becoming clearer that statements about what people said and did, or where they were, are not necessarily so. 

And yet, prosecutions for perjury are apparently rare--other than in cases involving financial benefits or demonstrable harm.  The Ferguson prosecutor, for example, is giving a pass to witnesses who lied on both sides, perhaps because prosecutions would have a chilling effect on future witnesses.  For the most part, then, there is no real deterrent to lying. 

Another disturbing piece of the puzzle comes from J. Guillermo Villalobos and Deborah Davis, psychologists at Nevada-Reno, and USF Law Professor Richard Leo.  in Honest False Testimony in Allegations of Sexual Offences, they propose, essentially, that the general social context, coupled with use of alcohol and the plasticity of memory make it quite possible for two people to testify honestly about very different versions of a single event.  The paper is about sexual assault cases, but the moral certainly extends beyond that.

My reaction is that we should recognize that criminalization comes with significant costs, which likely include a signficant error rate.  For rape, robbery, murder, and the other common-law felonies, there is no choice but to criminalize, even recognizing that mistakes will be made; the alternative is too terrible.  But for lesser forms of undesirability, perhaps greater caution  is warranted.