Photograph by Ken Light.
The State of Texas, acting on behalf of the people of Texas, and, more broadly, on behalf of the people of The United States, executed Cameron Todd Willingham. The evidence does not, at least, justify the conviction--and subsequent execution--of this man, and the evidence does, at most, demonstrate Willingham's innocence.
To put it another way the State of Texas, acting on our behalf, executed a man who is either innocent in the eyes of the law, or, probably, innocent in matter of fact.
David Grann's article in a recent New Yorker lays out the case against Willingham and its many, many flaws. Since then, and since a number of articles in papers across the country, this story has--albeit slowly--made it's way to cable news. Though I recommend reading the Grann's article, here's CNN's segment:
As both this segment--and Grann's article--points out, Gov. Perry was probably negligent, and his inaction was inexcusable. What's more, Perry, since this has become national news, has engaged in Nixon-esque firings of public officials to ensure that no state agency reviews the case further.
There are, I think, at least two sorts of people opposed to the death penalty. One sort are like me: People who are simply opposed to our government killing people--even if they are guilty of heinous crimes. The second sort are more pragmatic: People who are not opposed in principle to killing those guilty of heinous crimes, but who recognize the myriad structural problems in our justice system, and are opposed because there is a chance--a big enough chance--that we might execute someone innocent.
For the second sort of people, Willingham's execution is sort of tragic, unwanted, and unneeded vindication. There are too many flaws. There are pseudoscientific investigations (the arson investigator and expert in the Willingham case); there are overworked defense attorneys and over-eagar DAs (as in the Willingham case); there are "tough-on-crime" politicians who can't bring themselves to review evidence (as in the Willingham case). Let's not even start on racial discrimination in death penalty cases, which, by the by, wasn't even applicable here.
But even putting aside these problems, there are basic epistemological issues to consider. When determining if a person has committed a crime, we use sort of reasoning called "inference to the best explanation" (IBE) or "abductive reasoning." The idea is this: we have a body of evidence, x, for which there are many competing hypotheses, h_n, that purport to explain x. IBE bestows epistemic merit on a given hypothesis by evaluating that hypothesis's explanatory strength vis-a-vis it's competitors. Simpler explanations are better than more complicated ones. Deep explanations (i.e. explanations that include motives) are better than facile ones. Broad explanations (i.e. explanations that explain all--or nearly all--of the evidence) are better than narrow ones, etc.
So, the best explanation gets epistemic merit. But how much? In criminal cases, we want enough to overcome the burden of "reasonable doubt." I take this to mean that a hypothesis's competitors must be so ludicrously impoverished in comparison. And, in this way, that hypothesis can be awarded so much merit that we are rationally constrained to accept it.
The problem comes when we consider the finality of execution. That is, imprisonment is reversible (more or less). Execution is not.
So let's take away all of the structural problems with the justice system. Let's assume perfect rationality on the part of all the actors. Let's imagine that our justice system was perfect. Even with all of this, there is always a chance with any sort of non-deductive inference that we are wrong. We can even be rationally constrained to accept the truth of hypothesis, but in matter of fact be wrong. And if we accept the maxim on which our justice system is founded that it is "better that ten guilty men go free than one innocent man be wrongly convicted," we must not take irreversible action on the basis of non-deductive evidence.
Really, though, you should read the article. Go, read.
|