Document Type


Publication Date



This paper is part of larger symposium convened for the 2010 AALS annual meeting. In it the author adapts some of his earlier constitutional theoretical work to engage the topic of that symposium: the so-called “interpretation/construction distinction.” The author makes two related criticisms of the distinction: (1) it relies on a flawed conception of linguistic meaning, and (2) while these flaws may be harmless in the “easy” cases of interpretation, they are much more problematic in the difficult cases of most concern. Thus, the author doubts the ultimate utility of the distinction as part of a “true and correct” model of constitutional theory.

First, the author disagrees that there are such things as “linguistic facts” that we might discover as part of an initial search for a clause’s bare “semantic meaning.” A great many sentences make assertions about the world whose correspondence cannot be verified as an empirical matter. This is most obviously true when we make vague assertions such as “John is bald.” This assertion may or may not correspond with the world; it depends upon the linguistic rules and convention surrounding our particular use of the word “bald”—it depends on what we think counts as “baldness.” One of the ways that we might make arguments for or against John’s baldness is by reference to paradigm cases of baldness and not-baldness; and in this essay the author argues that this is analogous to the ways that we use canonical texts in our constitutional argument. But the important point for purposes of the interpretation/construction distinction is that “sentence meaning” depends upon the rules of proper linguistic usage, not the discovery of linguistic facts.

Second, in the easy cases—such as the age requirement for the Presidency—it may make little difference whether the phrase “35 years old” denotes a factual state of affairs or refers us to linguistic rules. In other words, there is such universal agreement about the interpretive rules in these cases that they may serve as something like “facts.” In such cases, it is possible to engage in the two-step kind of analysis that the interpretation/construction distinction suggests: We can agree upon a fairly objective interpretive referent, and then build out applicable legal rules. But in those cases where there is not broad agreement on the proper linguistic rules - such as the proper use of “equal protection” - it becomes quite apparent that there is no “linguistic fact" waiting for us to discover. Instead, our “interpretation” in these cases will depend upon what we can agree on as legitimate or proper uses of the phrase. Thus, in the cases of most concern, the author thinks the underlying conceptual flaws impose a great limitation on the ultimate utility of the interpretation/construction distinction.

In constitutional practice part of the argument about meaning will refer to paradigm cases of “equal protection” and “not-equal protection.” In our practice, the author contends that canonical texts often serve as some these paradigm cases. But the symbolic meaning of these texts evolves also over time within our argumentative usage. And as the meaning of these texts change, so will the interpretive conventions surrounding constitutional language. Here the author presents the evolution of Federalist 10 as an illustrative example.

Publication Citation

27 Const. Comment 9 (2010).