Why Is the Law So Either/Or?

One of the main things a law teacher does in class is to explore the boundary regions of a legal doctrine—the hard-to-categorize cases that are most likely to give rise to litigation: Is it an assault for a man to kiss a sleeping woman? Has the hunter acquired ownership of a wild animal when one of his bullets has been lodged in it, whereupon the animal stumbles into another hunter’s trap? Is the alumnus’ promise of a donation a binding contract? Posing these kinds of questions may be the activity law professors most often indulge in, often to excess.

One of the most fascinating problems raised by such boundary cases is why the law is so disturbingly rigid when dealing with them, why it paints in black and white, when reality seems to come in shades of gray. The law insists that something either is an assault or is not, either gives rise to ownership or does not, either is a binding contract or is not. There are no in-betweens.

The trouble is that many of the actual cases seem like in-betweens. In fact a case would never reach the courts if it did not seem to be in-between. But why not let the court say that while this is not a clear-cut case of an assault, it is sort of an assault and that therefore we will award the victim a fraction of the damages, or impose on the perpetrator a fraction of the punishment, that would be appropriate if this were a clear-cut case? Wouldn’t that avoid a lot of unnecessary hair-splitting hairsplitting? That’s what many legal scholars have come to believe. Are they right? Would it be better if the law were less either/or?

My own conclusion in thinking about this question at some length is that it quite simply cannot be done—not for practical, but for logical reasons. To give you the flavor of the kind of difficulty one runs into if one tried to make the law less either/or, let’s consider a very specific kind of boundary problem: the transition from life to death.

As we look more closely at the stages through which everyone passes as he moves from being fully alive to being fully dead, it starts to feel increasingly artificial to designate any one point in this progression as demarcating the boundary between life and death. A person is a bundle of organs, every single one of which can function well, adequately, poorly, barely, or not at all. The brain is just one of these. Not only is brain death not an obvious, indisputable divider between life and death, there is the fact that brain death itself is a matter of degree, of stages, at any one of which someone might linger and languish for shorter or longer periods.

Would it not make sense to treat these intermediate stages as intermediate legal categories, during which someone enjoys only a portion of the rights that accrue to him while fully alive and a good deal more than would accrue to him after he is deemed dead? Let’s see what would happen if we tried to make good on this.

Imagine, then, a human being at a thousand (or even a hundred thousand, or a million if that seems more plausible) separate stages, ranging from full aliveness at one end to utter cellular decomposition at the other. This is the continuum along which, if we departed from the usual either/or approach, we would gradually reduce someone’s rights: the less alive he is, the less like a human being we will be required to treat him.

Let’s call the creature just one notch below a fully alive, regular human being H-minus-1, the creature two notches below a regular human being H-minus-2, and so on. Now suppose we had a choice to make between an H on one side and an H-minus-1 on the other side. What I have in mind here is a choice—any choice—about whether H or H-minus-1 should suffer a certain adverse effect, a choice, for instance, between giving aid to H or to H-minus-1 (or between imposing an unavoidable risk on either H or H-minus-1). To sharpen the issue, let us assume that the decision is made by someone with a duty to render aid, an emergency room physician, say, or a lifeguard.

If the choice is between an H and an H-minus-1 and we subscribe to the gradualist approach, we should presumably give the edge, by a narrow margin, to H. Being more fully alive, he has a greater claim to the kind of aid we accord the living. What if the choice is between an H-minus-1 and an H-minus-2? Presumably, by the same token, H-minus-1 should win out over H-minus-2 and so on down the line to H-minus-1,000.

Next, however, let us complicate the choice slightly. Suppose the choice is not between H and H-minus-1, but between one H and two H-minus-1’s. Now it would seem that the edge belongs to the latter. Two H-minus-1s seem worth more, as it were, than one H. The doubling in quantity seems to more than make up for the slight drop in quality.

Let us repeat this process by comparing three H-minus-2s with two H-minus-1’s and one H. Presumably the three H-minus-2s beat out the two H-minus-1s, which in turn beat out the one H. Now keep reiterating this process, so that we have a menu of alternatives that contains at one end 1,001 H-minus-1,000s, 1,000 H-minus-999s, and 999 H-minus-998s and at the other end 3 H-minus-2s, 2 H-minus-1s and 1 H.

To be painfully explicit, here is the menu:
1 H
2 H-minus-1s
3 H-minus-2s
4 H-minus-3s
. . .
999 H-minus-998s
1,000 H-minus-999s
1,001 H-minus-1000s
Apparently the package in the second row (2 H-minus-1s) should be chosen over the package in the first row (a single H); the package in the third row (3 H-minus-2s) should be chosen over the package in the second row and therefore also over the package in the first row; and continuing all the way down the line, the package consisting of 1,001 H-minus-1,000s should be chosen over the single H as well.

More schematically put:
H < 2 H-minus-1s < 3 H-minus-2s < 4 H-minus-3s . . . < 1,001 H-minus-1,000s where < of course means “less deserving of help than.” But this is absurd! A single H-minus-1,000 is a collection of completely decomposed cells. How can a set of 1,001 such collections possibly trump a single living human being, which is what an H represents? Something has gone wrong. But what? One or more steps in our chain of inference must be in error. Somewhere in our chain of comparisons, we must have made a mistake. At least one of these inequality signs must not be correct. But what would it mean to say that one of these inequality signs is wrong? It can mean only one thing: that at some point it must be the case that subtracting an increment in quality from H-minus-n as we go down to H-minus-(n + 1) cannot be made up for by increasing the quantity of such creatures by one. A small decrease in quality cannot be made up for by a small increase in quantity. In other words, there is no gentle transition into that good night, as it were. Death is a cliff not a gentle slope. Conceptually it is not possible to be just a little bit dead or a little bit alive. Death is never partial, it is quintessentially an either/or phenomenon. The same turns out to be true of many legal concepts. If you are interested in reading more about this, you might either look at my book Why the Law Is So Perverse, or at the work of Larry Temkin, on whom my argument here draws (though the problem of either/or is of no particular concern to him), specifically A Continuum Argument for Intransitity, Philosophy and Public Affairs, 2, no. 3 (1996).