If you’re an American, you’ll at least have heard of Chris Kyle, the famous “American Sniper” who is the subject of a big-grossing Hollywood movie of that name. Kyle, a Navy SEAL stationed in Iraq, was the most “successful” of military snipers, with 160 confirmed kills and nearly a hundred more suspected kills—one from a distance of over a mile. He also wrote a bestselling book with the same title as the movie.

And as you may also know, Kyle devoted his post-war life to helping veterans recover from post-traumatic stress disorder (PTSD). Ironically, it was one of the people he was trying to help, Ray Routh, who, in February of 2013, shot Kyle and his friend Chad Littlefield in the back when they took him to a shooting range as part of Routh’s treatment. Kyle and Littlefield were both killed, and Routh was tried for the murders.

Routh pleaded “not guilty by reason of insanity,” and there were good reasons to think that he was pretty deranged. He had been diagnosed with schizophrenia before he entered the military, as well as PTSD. He was a heavy drug user and had been in and out of mental hospitals. But the Texas jury didn’t buy the insanity plea and, on Tuesday, found Routh guilty of capital murder. Mandatory sentencing laws mean that he’ll spend the rest of his life in prison without the possibility of parole. The Washington Post analyzes why the insanity defense failed:

To surmount the odds, an attorney must demonstrate not mental illness but legal insanity. Texas makes clear the distinction between the two in its legal standard: “It is an affirmative defense to prosecution that, at the time of the conduct charged, the actor, as a result of severe mental disease or defect, did not know that his conduct was wrong.”

You can be insane, but if you knew at the time of the crime what you were doing was wrong, you’re cooked.

And that’s what cooked Routh, for in his interrogation after the shootings, he was asked by a Texas ranger, “You know what you did today is wrong, right?” Routh answered: “Yes, sir.” That’s all the jury had to hear to convict him. Also, several psychiatrists testified that Routh was faking his illness. (Others took the opposite view.) As an expert witness, I know that you can always buy expertise to testify in your favor, whether you’re the defense or the prosecution, and that’s why I never took money for my services—to buttress the credibility of what I said about DNA evidence. These things tend to become a “battle of experts,” with the jury left baffled and unable to judge. Such bafflement should, in fact, raise some “reasonable doubt”, which is the criteria for exculpation—from capital murder in this case.

As a determinist who thinks (along with most of us) that Routh had no choice about what he did, was his sentence appropriate? Should “the ability to tell right from wrong, and know what you did was wrong” be the criterion for sanity and imprisonment rather than treatment (which is what would have happened had Routh been found not guilty by reason of insanity)? Also, as a determinist I think that since we have no choice about what we do, punishment should be based on three things: 1) sequestering the criminal from society so that no more crimes will be committed until he/she can be trusted to return to society, 2) setting an example for others to deter them from similar acts, and 3) isolation of the prisoner while he/she is being cured or treated for whatever caused the crime.

This is a hard problem, but I object to the criterion of an all-or-none insanity defense, particularly one based on “knowing right from wrong.” Suppose a criminal knows that what he did could be seen as wrong, but is a confirmed psychopath and doesn’t feel that what he did was wrong—or doesn’t care? Should that person spend the rest of his life in jail without treatment? That doesn’t seem right.

In fact, everyone is “insane” in some way when they commit a crime, for something about their genes and environment has led them to a situation where they do a bad thing, and they cannot help themselves, regardless if they know (or internalize) right from wrong. The “knowing right from wrong” criterion is irrelevant for conviction because it didn’t suffice to stop the person from committing a crime based on deterministic factors. Where it may be relevant is how to punish a criminal; but it’s not clear that the “knowing” criterion should be the one factor mandating jail versus psychiatric treatment, or how that treatment is given.

Ideally, in my view, there shouldn’t be a simple all-or-none decision that dooms a criminal to jail versus a more humane form of treatment or incarceration. If that criterion were found to be perfectly correlated with the certainty of doing more crimes regardless of how one is treated, then yes, maybe you can use it. But there are plenty of people who “know right from wrong,” but are so damaged, abused, or mentally ill, that they commit crimes anyway. I can’t see them being treated all the same.

And ideally as well, the jury’s decision should be only “whether the person committed the crime or not,” not “whether there were mitigating circumstances.” Those, and the sentence and treatment, should be left to experts, and that’s where science comes in.  In principle, science could determine what kind of sentences are most valuable for deterrence (I believe that execution hasn’t been shown as a deterrent), what treatment is most efficacious in helping people return to society without endangering it, and how long that treatment should last before the return occurs.  None of that is being done, of course—no society has that kind of money. Instead, we impose simplistic standards on juries, standards that, while they’re better than nothing, are surely far from optimal.

Nearly everyone has mitigating factors for their crimes, I think. Look at the overrepresentation of poor blacks in our prison system, which much surely reflect environments that are improverished, where education and conventional ambition are hard to sustain, as well as discrimination, which leads to both mistreatment and sequestration in substandard environments. One thing we must do is ensure that we try to mitigate those environments. But given that they’re here, and the undoubted fact that they breed crime, we must hold environment as a mitigating factor and see if there are ways we can cure its influence on the convicted.

This is not an easy problem. But one thing I know: we don’t invest nearly enough money in the scientific study of crime (which involves a mixture of psychology, sociology, and neurobiology), and yet it’s one of the biggest problems America has.

As for the fate of Routh, I wasn’t on the jury and don’t have their knowledge. But I know from personal experience how easily juries can be swayed by experts, as they were in the O. J. Simpson case. In that instance the jurors knew nothing about DNA, and merely the presence of some experts arguing about statistics and contamination threw sufficient sand in their faces that they simply thought, “oh, reasonable doubt.” In Routh’s case it went the other way.  We need experts to judge such things, not untutored laypeople, no matter how well motivated they are.