Instrumental commensurability.

• Author: Schauer, Frederick
• Position: Symposium: Law and Incommensurability

What is commensurability, and what is incommensurability? Although the topic of commensurability has been discussed extensively by philosophers, lawyers, and others for more than two decades,(1) the typical discussion of commensurability is an inquiry in moral ontology. That is, commensurability is taken to be a property inhering in values, reasons, options, or norms, and much of the debate is about whether that property does or does not inhere in all values, reasons, options, or norms, or, more accurately, in all pairs or sets of values, reasons, options, or norms. Those who argue for commensurability maintain that all values, reasons, options, or norms(2) are reducible to some common and thus comparable metric (perhaps dollars,(3) perhaps some other medium of exchange, perhaps utils, or perhaps something else). According to this position, the value or reason having or producing more of the property measured in the common metric is better than the value or reason having or producing less of that property. Relatedly, those who subscribe to the somewhat different position known as comparability(4) maintain that all values, reasons, options, or norms can be compared to each other, even if they cannot be reduced to a common metric.(5) Naturally, those who argue for incommensurability or incomparability deny the phenomenon of commensurability or comparability, and thus maintain that the members of some pairs or sets of reasons, values, options, or norms are irreducibly different.

My emphasis on the word "is" in my opening question is intended to signal a species of inquiry different from the one that provides the focus of most of the existing commmensurability literature. I Will argue here that commensurability and comparability(6) are not (or are not only) properties of sets of values or reasons that do or do not obtain. Rather, commensurability and comparability often have, or can be constructed to have, the character of attitudes, dispositions, presumptions, or conceptual frameworks, and, as such, they are best thought of as being chosen rather than as simply existing and, furthermore, as being chosen for instrumental and not intrinsic reasons. Why this is so, and why we might want to choose one or the other of commensurability and incommensurability, or of comparability and incomparability, are the central questions I will address in this Article.

I

I own a white 1990 Subaru Legacy station wagon. Suppose I happen to mention this fact to a friend, who responds that she has the same car. "No you don't," I reply. "You could not possibly have the same car. I have seen my car almost every day for the last seven years. And if you do not believe me when I say that the car you have is not the same car that I have, I will show you the serial number of mine, and that will prove conclusively that my car is not the same car as your car.

If I were to reply in this way, perhaps my friend would assume that I was being odd, or that I was making a weak and unsuccessful attempt at humor. And she would assume one of these things because the word "same," which in some contexts denotes identity (the person who came to my door today was the same person who came to your door yesterday(7)), in many other contexts, including this one, denotes something else. The use of the word "same" often suggests not only that a number of relevant similarities exist even in the face of dissimilarities, but also that in the instant context these dissimilarities are immaterial. Although your car is not my car, and although we are talking about two cars and not one car, we commonly use words of identity--"same," "equal," "alike," "identical," "exact"(8)--in a particular kind of context-dependent way. Even though we recognize that there are, in a literal sense, differences of one kind or another, we suppress these differences or inexactitudes, because the points of convergence are far more important in some contexts than the points of divergence. My example shows that there are contexts in which you would say that your car is the same as my car even if the two were different cars, and even if you would not trade your car for mine. Numerous common expressions--"close enough for government work"; "six of one and half a dozen of the other"; "it's all the same to me"; "Tweedledum and Tweedledee"--all reflect the frequency with which, for various strategic or instrumental reasons, we commonly treat things that are in some respects different as if they were the same.(9)

We see the same phenomenon in the context of the ordinary discourse surrounding the concept of political, moral, and social equality. When the authors of the Declaration of Independence announced that "all men are created equal," they did not suppose that even the people of whom they spoke (a subset of all people, and a subset of all men) were in fact equal in all respects, or even in all relevant respects. Their claim of equality was not a descriptive one, but one that was in part aspirational, and in even larger part normative. As with claims of identity, likeness, exactness, and sameness, claims of equality are not ordinarily claims of literal equality in all respects, or even in all potentially relevant respects, but rather they are claims that people should be treated the same in some number of respects because they are the same in some, but clearly not in all, respects.(10)

The same phenomenon of attributing sameness on the basis of partial similarity could, and often does, work in reverse as well, such that we attribute difference even in the face of substantial similarity, Thus, we frequently exaggerate differences in some contexts that in other contexts would be inconsequential. The (perhaps apocryphal)(11) claim that the Inuit have a very large number of different words for snow is a prominent example, but the same phenomenon occurs in ordinary talk. When we say "it's not the same," we often mean that in the face of numerous similarities, and in the face of superficial identity, there is a difference that is particularly relevant in some context. Just as assertions of sameness are typically context-dependent, instrumental assertions of relevant sameness in the face of some number of differences, so too are assertions of difference typically context-dependent, instrumental assertions of relevant difference in the face of some number of similarities between the compared particulars.

II

What emerges from this brief look at the ordinary language of sameness and difference is the suspicion that claims of equality (and inequality), sameness (and difference), and equivalence (or nonequivalence) are more purposive and instrumental and less descriptive than they may seem at first sight. Moreover, the purposes and goals behind the use of such words and phrases may involve normative and prescriptive claims as well as descriptive statements. And when claims that look descriptive on their face turn out to be prescriptive and normative, a cluster of related philosophical contributions becomes relevant.

The first member of this cluster of ideas is the concept of emotive meaning. The notion of emotive meaning, often associated with the work of Charles Stevenson,(12) suggests that the purpose of a great deal of superficially descriptive language is to express the speaker's pro or con attitude about what it is she is literally only describing. For instance, in modern discourse, whether a speaker refers to certain employment and admissions policies as "affirmative action" or as "reverse discrimination" tells us a great deal about her views on such policies. Similarly, we can surmise a person's opinions on some issues of social policy based on whether she identifies an individual as a "homeless person" or as a "tramp."

Stevenson and his followers believe that this kind of "Boo!--Hurrah!" connotation is a large part of ordinary talk, and, in particular, the ordinary talk that claims to be morally descriptive, as in "it is wrong to have sex outside of marriage." For Stevenson and many others,(13) the best translation of "it is wrong (or right)" and other seemingly descriptive statements is "I like it" or "I do not like it."

Related to this account is the more recent literature on "thick descriptions,(14) as most carefully developed by Philippa Foot(15) and Bernard Williams.(16) When we appear to describe some behavior as "rude," to use one of Foot's examples,(17) we are not just describing it. Instead, we are either personally condemning the behavior or, at least, describing the fact of extant societal condemnation. Whether the condemnation comes from the speaker or from the community at large, the important point is that the condemnation is part of the meaning of words like "rude." To describe behavior as both rude and praiseworthy would be to engage in an act of linguistic self-contradiction. Many other descriptions--for example, "adorable," "pushy," "imposing," "bum," "hero"--are in part descriptive because someone in agreement with the normative standards implicit in one of these words could still disagree with the word's application to a particular case. The words, however, are not only descriptive, because, in the process of describing, the describer is also making or subscribing to a normative or evaluative claim, or at the very least describing a normative or evaluative claim made by someone else or by society at large.

The idea of thick descriptions suggests, in turn, the concept of ascriptive language. In his earliest work, H.L.A. Hart suggested that all of legal language is ascriptive, in that legal language characteristically announces a legal conclusion rather than describes a hard fact about the world.(18) To say that a transaction is a contract, or that a defendant is guilty, or that some form of behavior is a tort, Hart argued, is not to identify some raw and pretheoretical fact. Instead, it is to announce a legal outcome, and thus to ascribe to a set of facts a certain...