Introduction "It takes a theory to beat a theory"--this is surely one of the top ten all-time comments uttered by law professors to one another in those ritual interactions that are called "faculty workshops" or "colloquia." The first instance of the comment that I can find in the legal literature appears in an article by Richard Epstein [92 Yale Law Journal 1435 (1983)], and the some legal academics associate the comment with Professor Epstein, whose facility in workshops is legendary.
What does "It takes a theory to beat a theory" mean? What is the point or purpose of making this comment in a debate about normative or positive legal theory? Of course, the core idea is relatively straightforward. One can't beat a theory just by nitpicking. We go with the best theory we have, warts and all. So if you want to beat a theory, you must show it is not the best theory we have, and the only way to do that is to produce a better theory. Hence, it takes a theory to beat a theory. Or to put it crudely, if we are playing "king of the hill," whoever is on top stays there until pushed off.
Most of the entries in the Legal Theory Lexicon series deal with some concept, theory, or idea that law students are likely to encounter in their first year of law school. But this entry is a little different. "It takes a theory to beat a theory" is not likely to be heard much in the classroom. It's a "workshop comment" not a "class comment." But in a way, that makes this idea all the important for law students, including first year law students, with an interest in legal theory. And understanding this move and its counters is essential for anyone who plans a career in the legal academy.
The Aphorism and its Applications Here is one of the most famous examples of the aphorism at work--although the exact words do not appear. This is Justice Scalia, in his famous article, Originalism, the Lesser Evil:
Apart from the frailty of its theoretical underpinning, nonoriginalism confronts a practical difficulty reminiscent of the truism of elective politics that "You can't beat somebody with nobody."' It is not enough to demonstrate that the other fellow's candidate (originalism) is no good; one must also agree upon another candidate to replace him. Just as it is not very meaningful for a voter to vote "non-Reagan,"' it is not very helpful to tell a judge to be a "non-originalist."' If the law is to make any attempt at consistency and predictability, surely there must be general agreement not only that judges reject one exegetical approach (originalism), but that they adopt another. And it is hard to discern any emerging consensus among the nonoriginalists as to what this might be.
Here is another version of the argument--this time from George Stigler's 1982 Nobel Prize Lecture:
Nevertheless the economic theory of regulation is achieving a substantial scientific prosperity. Its findings with respect to both the operation and the origins of regulatory policies directed to particular industries (such as the securities markets, transportation, and occupational licensing) command a substantial support. To be sure, the explanatory triumphs have not been overwhelming, and indeed the theory itself is still relatively primitive. The main reason for the considerable acceptance of the approach is that fundamental rule of scientific combat: it takes a theory to beat a theory. No amount of scepticism about the fertility of a theory can deter its use unless the sceptic can point to another route by which the scientific problem of regulation can be studied successfully.
Variations If you've been paying close attention, you will have noticed that Scalia and Stigler were making very different arguments--relying on different premises and with only a surface level similarity. "It takes a theory to beat a theory"--is not a single move. It's several different moves, which can nonetheless be expressed via the same aphorism. Let's quickly catalog some of the variations:
--"It takes a practice to replace a practice." This variation is at the root of Scalia's point. You can't vote against a candidate; you must vote for someone. You can't just stop interpreting the constitution if you reject originalism; it takes some other interpretive practice to substitute for originalism.
--"It takes a better explanatory theory to substitute for a theory that has even limited success." This variation is at the root of Stigler's point. The economic theory has some success; so it will continue as the "dominant paradigm" until something better comes along.
--"It takes a better normative theory to substitute for a normative theory that has plausible support." Thus, one might say that it takes a better theory of justice to displace Rawls's theory.
And I'm sure there are many other variations.
Counter Moves What are the countermoves to "It takes a theory to beat a theory." They are legion, and many are contextual, but here are a few typical countermoves:
--Expose the hidden "presumption." It takes a theory to beat a theory may assume that the theory in question must be "beaten" or it stands. But that assumption requires a justification. It must be argued there is some "presumption" in favor of the theory's validity. What is the source of that presumption? Why does this theory enjoy the presumption rather than a rival theory? Sometimes these questions can't be answered. Other times there are answers, but once they are articulated, they can be contested.
--It doesn't take a theory to beat an argument. Even if it does take a theory to beat a theory, it doesn't take a theory to beat an argument. Arguments fail if they are invalid or unsound. Arguments are invalid if the conclusions don't flow from the premises. Arguments are unsound if their premises are false. So when someone says, "It takes a theory to beat a theory," you can reply, "Oh, you thought I was trying to beat a theory, but what I actually did was to beat an argument."
--Recharacterize the Status Quo. Frequently, "it takes a theory to beat a theory," assumes that the theory to be beaten is, in some sense, the status quo--the current king of the hill, the way we are doing or thinking now. But this is not necessarily the case. Take the "originalism" example. One might argue that "originalism" isn't the "king of the hill." Rather, the status quo is eclecticism--a little precedent, a little originalism, a little instrumentalism, etc. So if the case for originalism is inadequate, the presumption is for the status quo--not originalism. In other words, poking holes may be enough. Once the theory in question is no longer the status quo, the question becomes is the new theory demonstrably better than what we've got now.
--Deny the Theory Assumption. "It takes a theory to beat a theory" assumes that the issue at hand is one which is properly addressed by a theory. Let's call that, "the theory assumption." But the theory assumption can itself be contested. Here's a marvelous example of "denying the theory assumption," from Dan Farber, himself famous for this move:
The last ditch defense of the rational choice theory is to insist that it takes a theory to beat a theory, and that the behavioralists have only assembled a collection of empirical regularities without any unifying theory. The behavioralists indignantly respond that they do have a theory, although an incomplete one. The assumption on both sides is apparently that the sine qua non of social science is having a unified predictive theory. But perhaps this is merely another symptom of economics' famous case of "physics envy." Physics presents a breathtaking example of mathematical elegance combined with fantastically accurate predictions. But taking physics as the paradigm of science may be a mistake. Today's great success story among the sciences may well be biology. Biology does have a central paradigm (evolution) and an understanding of its molecular basis. But organisms, because they are the products of evolution rather than design, are extremely complex, and no one seems to think that their features can be predicted in any detail on the basis of a deductive theory.
(Daniel A. Farber, Toward a New Legal Realism, 68 U. Chi. L. Rev. 279, 295 (2001))
There are lot's of ways of doing this. Here's one way of making the move: "Hmm. You seem to be assuming that we need a theory here. I'm puzzled. Why would you assume that?" Stressing the words in italics. This move puts the burden on the maker of the "It takes a theory to beat a theory" move to justify the theory assumption--and as a practical matter, it also buys you time.
--Go Pragmatic. This is a variation of denying the theory assumption. When someone asks for a theory, you can reply that the subject at hand requires a "contextualist" and "pragmatic" analysis. As the scare quotes indicate, I'm not very fond of this move.
--Go Metatheoretical. This move embraces the notion that it takes a theory to replace a theory, but denies that the way that happens is through direct confrontation. The idea is that the way we replace theories is by accumulating the data points the theory can't handle. Here's a nice example of this move from Tom Ulen:
One frequently hears it said that "it takes a theory to beat a theory." Those who invoke that view are typically doing so in order to justify maintaining the rational choice paradigm until some well-articulated theory comes along to replace it. That is not, however, how scientific advances typically occur. Rather, anomalies arise and are either explained within an amended paradigm or serve as observations that make a case for another paradigm (as yet undefined).
(Thomas S. Ulen, A NOBEL PRIZE IN LEGAL SCIENCE: THEORY, EMPIRICAL WORK, AND THE SCIENTIFIC METHOD IN THE STUDY OF LAW2002 U. Ill. L. Rev. 875, 887, n. 47)
--I've Got a Theory. Of course, this move is especially impressive in the context of a workshop. Your paper makes a critical or destructive point. Someone in the audience goes "It takes a theory to beat a theory." And you go, "Great point. I've got a theory and here it is." You then get 5 minutes of uninterrupted theory exposition, and, of course, no one at the workshop has heard your theory before, and hence, no one has had an opportunity to prepare objections.
--It Takes a Theory to Beat a Theory, but you have no Theory. This is a variation of Lloyd Bentsen's famous, "Senator, you are no Jack Kennedy." The idea is this: "It takes a theory to beat a theory," assumes that we have a theory of X on the table. So one effective counter is an argument that there is no "theory of x" to replace. One way to do this is to set out a list of criteria, C1, C2, C3 . . . Cn, that must be fulfilled if a given cluster of ideas, I, is to count as a theory, T, of a subject, X. You then argue that the view on the table is not a true theory, because it fails one or more of the criteria.
A variation on this move is "It doesn't take a theory to beat a really bad theory." If your objection to a theory makes it clear that the theory cannot be correct, then it doesn't take theory to beat a theory. Or to continue the metaphor, "You don't have to become the king of the hill to knock someone down."
Undoubtedly, there are many other standard countermoves and numerous others that are relevant to particular contexts. I hope this post gets you started.
Conclusion "It takes a theory to beat a theory." Or does it? I hope this Lexicon installment has given you a taste of this famous move and some of the replies. And I also hope that it has provided some of the flavor of the intellectual atmosphere of faculty workshops and colloquia. As always, the Lexicon provides only a simplified (and all too often simplistic) treatment of a subject that deserves an extended discussion.
Adam Kolber, Theory v. Theory (March 16, 2006)
Lawrence Solum, It Takes a Theory to Beat a Theory (September 12, 2002)