I have been reading Lawrence Solum's Legal Theory Blog on some problems of intellectual property with great profit. Solum is engaged in a discussion with Eugene Volokh [1,2] on certain problems in economic analysis of intellectual property rights. My concern here is with two methodological problems. The first problem comes from the ex ante/ex post distinction, the second from the use of hypothetical cases.
Since I'm not a legal theory scholar, I don't know if Solum's usages are standard practice. It's also clear that legal theorists are concerned (of course) with right and wrong, while comparative analysis in the sciences is concerned with true and false, and it's an open question (at least in my mind) if the two kinds of analysis necessarily come to the same conclusions.
The ex post/ex ante distinction.
Solum specifies the ex post.ex ante distinction as follows:
The ex post perspective is backward looking. From the ex post point of view, we ask questions like: Who acted badly and who acted well? Whose rights were violated? …
The ex ante perspective is forward looking. From the ex ante point of view, we ask questions like: What affect will this rule have on the future? Will decision of a case in this way produce good or bad consequences? …
In focusing on looking "backward" and "forward", the ex post/ex ante distinction is similar to the distinction that researchers in clinical medicine (and many social scientists) make between "prospective" and "retrospective" studies. In the retrospective (ex post) approach, we have the facts in hand, and our job is to explain or interpret them-- to tell what happened. In the prospective (ex ante) approach, we have a (possibly tentative) model of the world, and our job is to use that to specify what will (or should) happen next.
In his discussion with Volokh, Solum is arguing that the conclusions one draws from a hypothetical case (here, the construction of property rights in an inexhaustible well) may differ if the case is viewed ex ante or ex post. This is where we run into the first problem. The distinction between retrospective and prospective seems conflated with another distinction: that between reasoning from particular cases to general principles on the one hand, and reasoning from principles to particulars on the other. That the two contrasts can be seen by crossing them and providing examples of the relevant kinds of work (I use examples from research work here):
| |
Retrospective |
Prospective |
| General to particular |
Identification |
Prediction |
| Particular to general |
Abduction |
Classification |
Identification means assigning an individual case to a place in a scheme of taken-for-granted categories (what physicians call diagnosis); it answers the question "What did we bring home from our field trip?" Classification means constructing a categorical scheme; it answers the question "What sorts of thing might we find next time?" Of course, classifications don't have to be constructed particular-to-general, but that way has been good scientific practice in recent centuries.
Prediction means drawing conclusions about particular cases from a model. Abduction occurs when we specify a hypothesis which would, if true, account for the cases we have already observed.
Are hypothetical cases particular or general?
The cross-classification of distinctions leads us to the second methodological problem in the Solum-Volokh discussion. Volokh's argument uses a hypothetical case, in which a farmer owns an inexhaustible well. Hypothetical cases, scenarios, have an interesting property: in an argument, they can be treated as either particular cases, or as abstractions (and hence, generals). It's possible to shift back and forth between the two seamlessly; here an example from Volokh:
My original hypothetical explicitly assumed that "the water table is huge, and the farmers aren't going to exhaust it," because there were only a few hundred farmers living nearby. Naturally, that's not always so; in many places, water tables are exhaustible. But it's not implausible that in some lightly populated places, the water table would indeed be practically inexhaustible -- not theoretically, since if transportation costs were nil, farmers from all over the world would come by, but practically.
Here, Volokh has postulated an individual case-- but the case is edited in order to eliminate objections which are irrelevant to the point that Volokh wants to make. He retains transportation costs as a way of getting rid of too many users. In his original post, he retained the costs of digging a well, in order to motivate the farmer's neighbors' willingness to come for someone else's water. At the same time, all sorts of other characteristics are tacitly abstracted from the well. Solum for example, restores location, and puts the well on the edge of the farmer's property, so as to avoid trespass issues. But this is after he makes a much more general point:
Because [Volokh] can easily fix the capacity problem by stipulating that the well itself has sufficient capacity so that assuming the water were free and all of the farmers in the neighborhoods used the well simultaneously, there would still be excess capacity. If I tried to argue with Volokh about this, I would simply be fighting the hypothetical. When you fight the hypo, even if you win the battle, you lose the war. Volokh could simply switch to another particular case in which a tangible resource is a toll good, not a pure private good or a club good.
This is a critical point. We can change the characteristics of a hypothetical case in order to suit the requirements of the argument we wish to make; this usage is prospective. It is also general-- we are setting the premises of an argument. This way of working then, is analogous to making predictions from a theory.
But of course, the same scenario, once tailored, serves as a particular case, and we reason from it to some conclusion. This usage is retrospective as well as particular-to-general. This way of working is thus analogous to abduction.
Needless to say, it's very difficult to avoid serious problems using this double-ended approach, which is why scientists are often very suspicious of it. There are good uses for scenarios; for example, a scenario which ends in absurdity is a good way of ruling out a hypothesis. Galileo's famous "experiment" is an excellent example: if lighter objects fall more slowly than heavier ones, then attaching a lighter object to a heavier one should slow it down. But this is absurd. Hence, all objects fall at the same rate, other things equal.
Since debates in law and policy are very often focused on "ought" rather than "is", it often isn't clear if using a scenario in a double-ended way is fallacious-- the outcomes of such debates are measured in fairness and justice, not in truth and wisdom. In general however, I think it's much better to have real particulars somewhere in the argument.
The comments to this entry are closed.
Comments