by Demosophist

Although the legal decisions about Terri Schiavo's fate are behind us, the ethical and moral issues probably are not. Like Eugene Volokh I'm not sure it's necessary to have an opinion on the Terri Schiavo controversy. But I have noticed something of an anomaly in the coverage, both in MSM and in the blogosphere. The issue is presented as boiling down to two basic questions:

1. What were Terri Schiavo's wishes regarding the "right to die" question?

2. Is she currently in a "persistent and permanent vegetative state?"
Now, I'm not quite sure why both of these issues are important at the same time. From an ethical perspective if she is not in a persistent vegetative state, then she can presumably be brought back at some point to a semblance of human consciousness. If that's the case then only if you believe in the right of assisted suicide would you consider it ethical to allow her to die, which is in effect leaving that decision to her rather than the "experts." So, only if she is not "brain dead" (in the vernacular) is Question 1 relevant.

If, however, she is in a vegetative state (which doesn't mean she's "a vegetable" by the way, since that state is higher functioning than a vegetable) then there really is no person identifiable as "Terri" who is alive, and more importantly there is no real possibility of ever regenerating that person any more than if she were "moldering in the grave." I realize that's harsh, but it's true. Now, I'm not presuming that this is the case. I'm just saying that if it is then what Terri Schiavo wishes or doesn't wish, or for that matter what anyone else wishes for Terri Schiavo, is simply not relevant. She's gone. Get over it. Move on.

(If her husband did something to bring about that situation then seek to find enough evidence to bring charges. But that's a different issue, entirely. And by the way, only in the case where she is "allowed to die" would he be in jeapardy of a conviction for homicide or murder. If her parents have their way the most he could be convicted of is attempted homicide or murder.)

So, I'm actually on both sides of this issue at the same time because there's a "vacuum of information" about the precise circumstances of the issue. And if that's the case then it becomes important to ask what are the parameters of that uncertainty. (Not that I know.) My pal Rusty has an excellent tutorial on decision-making under conditions of uncertainty that's relevant not only to this situation, but to the WMD question in Iraq as well as lots of other personal and public issues. I've written about it in slightly different terms myself. But the long and short of it is that if you're a decision-maker you have a choice between two basic methods, and you're constrained to choosing only one of those two.

Rusty refers to these two methods as Type I and Type II, and I call them "alpha" and "beta" but the terminology is discipline-specific, and doesn't really matter. To draw an analogy, you have the choice between "innocent until proved guilty" (alpha or Type I) and "guilty until proved innocent" (beta or Type II).

Although it's not appropriate to use the term "guilty" for someone who has clearly done nothing wrong, I'll use it for Case 1, where Terri is either "gone permanently from the body" or where she has chosen to have her life ended should she be incapacitated. Again, I'm not implying that she ever did anything wrongn, I'm just trying to keep the categories straight without relying on the slightly confusing nomenclature of hypothesis testing. Likewise, she's "innocent" (Case 2) if she is not in a vegetative state and she has not chosen to have her life ended should she be incapacitated. Note that the first situation involves an "or" condition, while the second involves an "and." This is appropriate because I'm assuming a person has a right to end their own life. If you don't believe that, then whether or not she chooses to end her life if she's incapacitated is irrelevant, and there's only one condition: whether or not she's "brain dead."

Now, in order to better illustrate these points I'm going to make a slight digression. My argument with regard to Saddam Hussein's possession of WMD (and Rusty's) was pretty simple: It was a Type II (beta) decision. This was simply because the consequences of a "false acquittal" were so dire. (They still are, with regard to Iran by the way.) We were therefore compelled by prudence and wisdom to adopt a "guilty until proved innocent" approach. The fact that we may have determined subsequently that he was actually innocent is irrelevant to the propriety of the choice of original method. That wasn't wrong then, and it's not wrong now. Only our assessment of the actual conditions about which we were uncertain was wrong. If you are assailed by an attacker in an alley who may have a lethal weapon, and who behaves as if he has a lethal weapon, it's appropriate to assume he has one until you know for certain that he doesn't. If, after shooting the fellow, you find that he had no weapon that doesn't mean your method was wrong. You still made the right decision under uncertain conditions. And you'd be justified in doing it again, under very similar circumstances (although you'd be required to take new information into account, of course). End of story.

By the way, the reason why the decision-maker has to stick to one and only one approach is that the consequences of mixing approaches can be disastrous, and because you can only optimize for one kind of false outcome at a time: either false conviction or false acquittal. Sorry, but there's no free lunch. Only where the odds of either are exactly 50% is it possible to optimize for both at the same time, and in that case it doesn't matter which method you choose. Note that although Hans Blix originally stated (on Charlie Rose) that he was adopting a "guilty until proved innocent" approach to Saddam, be gradually shaded over into the other method without realizing it. That was dangerously sloppy. He was inconsistent, which is frankly what one ought to expect of a career diplomat.

Now, back to the Schiavo case. This is essentially the opposite of the Saddam/WMD situation. The consequences of a "false acquittal" are not dire at all, at least not for any of us. Someone would have to pay her hospital bills, but apparently there are people willing to shoulder that burden. So, if I were the decision-maker (which I am not) I could afford to adopt a Type I or alpha method, and simply assume until proved otherwise that Case 2 is a fact: she is not in a vegetative state and she has not chosen to have her life ended should she be incapacitated.

So, we're done right? Well, not quite. We still would prefer not to commit an error, even though we're only optimizing to avoid one sort of false outcome: false conviction. So the facts of the case are still important. We can probably dismiss the issue of what her actual wishes are, because we can't really know that beyond the degree of certainty that we currently possess. We have the word of her husband and a couple of sisters-in-law, apparently. Various judges have considered that testimony credible, so from a legal standpoint that issue has been settled as much as it will ever be settled. There's no new evidence coming down the pike. The only part of Case 2 that's potentially still up for grabs is whether or not she's in a persistent vegetative state, and the clearest and most unambiguous evidence of that has to be physical. The CT scan suggested that she has no cerebral cortex, and if that's the case then our assumption has been proved false. That is, even though we've assumed that Terri is "innocent" our actual analysis has proved otherwise, and we therefore are compelled to reject our assumption. But some people, including a Nobel Prize nominee named Hammesfahr, have suggested that the CT scan evidence is not conclusive. Are his objections, or the objections of anyone else on that side of the argument, to be believed?

Well, I feel confident that I can dismiss Hammesfahr as a credible witness. Anyone who solicits a nomination for a Nobel Prize, but who lacks any peer-reviewed articles that might qualify him to actually win the prize, is a con man, charlatan, vulture. The same was true of another man of infamy, Armand Hammer, who solicited a nomination for the Nobel Peace Prize (and nearly won, by the way) but who was actually a KGB bag man. So Hammesfahr not only does not call the CT scan into question, he taints everyone on that side of the argument who relies on or associates with him.

But it's still possible that the CT scan isn't definitive, and I'm simply not sufficiently expert to know whether a PET or an MRI might settle the matter more definitively. If they can I'd be all for using them. That's not something we want to determine conclusively only in an autopsy, after all.

So I'm afraid that I'm still stuck on both sides of the issue. And it has occurred to me that the real meaning of this controversy is less about the fate of Terri Schiavo than about our own deficient perceptions of life and death, and what those mean. Terri Schiavo is a kind of ink blot for all of us, so that we can observe what it is that we really believe about the single greatest uncertainty in all our lives.

Update: I can think of only one rather unconvincing reason to allow Terri Schiavo to die of dehydration (which will get her long before starvation). It's simply our queasiness about making a decision with consequences that prevents us from giving a lethal injection on the chance that even those in a vegetative state can feel pain. (However, for those distressed by this possibility, odds are people in this state don't experience anything, including pain. If that's not true, then perhaps death really is the lesser evil.) The one sorry reason I can think of for not doing that is that if Michael Schiavo ever faces trial he could argue that it wasn't he who killed her, but the state. Thus, he could only be tried for attempted murder. But it's not really much of a reason.

(Cross-posted by Demosophist to Demosophia and Anticipatory Retaliation)

Posted by: Demosophist at 02:20 PM | Comments (10) | Add Comment
Post contains 1801 words, total size 11 kb.