Conor Casey is a Lecturer in Law, University of Liverpool School of Law & Social Justice. Adrian Vermeule is the Ralph Tyler Jr. Professor of Constitutional Law, Harvard Law School.
A standard set of claims, common to both classical lawyers and Dworkinians, is that the interpretation of positive legal rules inevitably rests on express or implied arguments from political morality (in particular, for the classical lawyer, the subset of political morality bearing on legal justice), and that positive rules are inevitably under-determinate, due to the inherent limits of language and of lawmakers’ foresight. Originalists have evolved an attempt at a solution: “closure rules.”
At the question and answer session following a recent panel discussion, organized by the American Enterprise Institute, on “Natural Law and Original Meaning,” an audience member asked the panelists (Sherif Girgis, Kevin Walsh, and Lee Strang) if judges should be able to apply “maxims of natural law as canons of construction” to find the meaning of posited constitutional text, where texts remain under-determinate after attempts to ascertain its original meaning from socio-historical sources. Strang’s headline response was that there was no conventional originalist position to address the question (a notable response in itself). But Strang proceeded to note that originalists have worked hard to identify additional conceptual resources to create so-called “closure rules” in cases where constitutional text remains under-determinate, after the search for things like original semantic and conventional public meaning has been exhausted. Strang cited the possibility of seeking out originally understood legal rules of interpretation to try and generate thicker, more robust, original meanings and ameliorate the under-determinacy problem. Where this kind of enterprise still yielded under-determinacy, Strang suggested that originalists had argued powerfully that constitutional actors should proceed in a Thayerian fashion: the legislative branch should be permitted to “construct” meaning and judges should defer to their construction unless they have acted in clear error of other constitutional commitments.
As we have previously noted, however, such “closure rules” just replicate the problem they are intended to address. Closure rules are themselves subject to the same problems as the first-order rules for which they attempt to provide a tourniquet of determinacy. The very fact that Professor Strang mentioned multiple possibilities for “closure rules” underscores the point: the choice between possible closure rules itself depends upon normative arguments. It is not just written in the nature of things, inherent in the nature of language or communication or inherent in the nature of interpretation, that the interpreter should simply “put the statute down” (as Frank Easterbrook once suggested), should rely on party presentation of socio-historical facts rather than attempt their own search (as some originalists suggest), defer to legislators or agencies (the approach Strang finds powerful), or adopt some other approach. To adopt one or other approach itself requires a choice, necessarily based on arguments about which approach serves the public interest or common good — whether or not the interpreter makes those arguments explicit.
Likewise with the idea that originalists can just use the closure rules “used by the founding generation itself,” such as the above-mentioned use of conventional legal rules of interpretation present at the different times constitutional provisions were ratified. Apart from other grievous problems (notably, that the founding generation thought about law in an entirely different way, within a classical framework and based upon a classical legal ontology), such rules themselves come in multiple competing versions, are themselves inevitably partially indeterminate and require further, controversial specification, and themselves require interpretation, resting on normative arguments. Just as first-order rules can be read at multiple levels of generality, cashed out in different ways, and otherwise extrapolated in different directions, so too with the closure rules. Rules of historical evidence, for example, “rest on express or implied normative assumptions and arguments about the costs of decisionmaking and the costs of error, about the collateral and systemic effects of admitting or not admitting certain categories of evidence, and a myriad other topics” (as one of us recently noted). As H. Jefferson Powell put it, “Rather than avoiding the responsibility of choice, history requires of the originalist a whole new range of contestable…decisions.”
As another example, take Strang’s preferred example of “deference.” As administrative lawyers arguing over the best specification of Chevron deference can attest, merely saying “defer to reasonable interpretations of ambiguous text” is only the beginning of a very long story, involving threshold “step zero“ questions (when exactly does deference apply? Are there categories of decisions to which it doesn’t apply at all?), “step one” questions (how ambiguous is ambiguous?), and “step two” questions (how reasonable is reasonable?). At every such step, normative arguments have been and must be called into play. Consider the idea in the leading decision on the step zero problem, United States v. Mead, that a certain complex approach to step zero best serves “fairness and deliberation.” Consider also an argument for the “major questions” clear statement rule, recently advanced by Justice Neil Gorsuch, that it helps to prevent “government by the people” from being supplanted by “government by bureaucracy.” Gorsuch’s argument here can only be described as Dworkinian in method, although needless to say not Dworkinian in substance.
Sometimes, “closure rules” are said to lie within the “construction zone,” where “construction” is taken to be a different activity than “interpretation.” As we have also said before, this is in essence an arbitrary semantic categorization, one that presupposes a conception of “interpretation” that the classical lawyer contests. Actual interpreters do not approach or experience their task as though it involves a switch between an “interpretation” phase and a “construction” phase where “the law runs out.” It is all one process, a process that occurs within law and within legal interpretation, involving — from the classical perspective — the harmonization of two types of law: the harmonization of positive texts (lex) with background principles of legal justice (ius), not as a way of overriding lex, but as a way of understanding and interpreting lex in light of the larger commitments of the legal order.
As John Finnis puts it, basic principles of natural law are not extra-legal considerations or policy arguments brought to bear by lawyers and judge where the law runs out, but themselves function as a “direct source of law (or justification for judicial decision) and, in a certain sense, as already law.” Basic precepts of the natural law, says Finnis, are best regarded as “judicially applicable moral rules and principles” and “ipso iure (i.e., precisely as morally and judicially applicable) rules of law” belonging to the “ius gentium portion of our law.”
The best account of “closure rules,” then, is that they are just another set of arguments about the content of ius, as opposed to lex. Closure rules are just ius for originalists. Here as elsewhere, whatever their abstract theories, interpreters inevitably practice the classical legal approach, even without knowing it.
In the end, Strang’s response nicely highlights a serious fault line between the interpretive approach of the classical legal tradition, on the one hand, and the hybrid approach of so-called natural law originalism on the other. In the resolution of hard cases, both types of jurists will inevitably appeal to considerations of ius – whether they call it that or not. The former approach builds into the structure of the theory regard for the basic precepts of natural law and legal justice as the interpretive lodestar. But the latter appears (from what we have observed) reluctant to invoke them or, at most, will only have recourse to them when they are incorporated by posited law sources, all out of concern for undermining law’s stability, or co-ordination function, or other institutional goods — goods that, in our view, have already been left by the wayside in the class of cases that require recourse to “closure rules.” While both theories may share some fundamentals, this is a wide divergence indeed.
Conor Casey & Adrian Vermeule
