Legal Theory Blog

Introduction

As frequent readers of Legal Theory Blog already know, I believe that the legal and theoretical distinction between interpretation and construction is an important one.  Although these words can be used as synonyms, "interpretation" is also used to refer to the activity of discerning the linguistic meaning of a text, whereas "construction" can refer to the activity that determines the text's legal effect.  Recently, there has been considerable discussion of the interpretation-construction distinction in the blogosphere lately, prompted by a series of scholarly events, including the main session of the Constitutional Law Section at the AALS, a debate at the Federal Society's law faculty meeting, and a conference at the University of San Diego.  Several interesting issues arose at the conference, and in a recent exchange between Randy Barnett and Michael Rappaport.

The latest post from Mike Rappaport responds to Randy Barnett (here and here) responding to an earlier post by Mike. I've already made a brief comment here.  I think it might be helpful to put the disagreement in a larger perspective by saying a few words about the origins of the distinction, and then to engage more specifically with some of Rappaport's comments.  I have the greatest respect for both Randy and Mike, so I want to try to make my position on these issues clear.  For a brief introduction to the issues, check out Legal Theory Lexicon 063: Interpretation and Construction.

The Origins of the Interpretation-Construction Distinction

The interpretation-construction distinction is actually an old one, it appears in common-law cases on contracts, trusts, and wills and in cases on patent law and statutory construction.  The classical formulation of the distinction identifies "interpretation" as the activity that discerns the linguistic meaning of a legal text and "construction" as the activity that determines their legal effects.  For example, if a court is deciding a contracts case, it might first determine the meaning of the words and phrases (the linguistic meaning) and then determine what legal effect those words will have.  The two might be identical, but this is not necessarily the case, because contract law contains mandatory rules that can nullify or modify the effect that would otherwise be given to the linguistic meaning of the contract.  Another familiar example of this phenomenon is the notion that a will that violates the rule against perpetuities may be given a "saving construction."  The saving construction is the legal effect given to a linguistic meaning that violates the rule against perpetuities.

If you are like me, you probably did not learn of the interpretation-construction distinction in law school–although the distinction is discussed in contemporary cases.  There is an interesting story to be told about why that is the case.  I suspect that the distinction came into disfavor in the academy with the rise of American Legal Realism, which emphasized the role of policy in construction and minimized the binding effect of the linguistic meaning of legal texts of all kinds (including contracts, statutes, and constitutional provisions).  Farnsworth suggested that that the realists played a role in downplaying the distinction between linguistic meaning in his discussions of the interpretation-construction distinction in contract law.  (His Yale Law Journal article, "Meaning in the Law of Contracts," can be found here.)

I first learned of the interpretation-construction distinction via the work of Randy Barnett and later Keith Whittington.  (Whittington's work was first in time and influenced Barnett).  At first I was highly skeptical of the distinction, because it just seemed so unlikely that I could have completely missed a fundamental distinction, and because the terms "interpretation" and "construction" were used interchangeably in the legal literature with which I was familiar.  But I gradually came to appreciate the idea that there was a real difference between the linguistic meaning of a text and its legal effect.  At first, I thought that this difference could be cashed out in terms of the distinction between vagueness and ambiguity, but as time went on, I began to realize that the distinction between interpretation and construction was a deep one, and that its relationship to vagueness and ambiguity was merely a surface manifestation.  That realization led to the paper that I gave at the San Diego conference–which is a very early draft of an attempt to provide a theoretical explication of the distinction.

Evolving Views of the Nature of the Distinction

Mike Rappaport suggests in his most recent post that my view of the distinction has changed.  He is correct, although I believe that my current position is a refinement of the old one, and not a significant shift in position.  Let me explain this in three steps: (1) the old view, (2) the new view, and (3) the relationship between the two.

(1) The Old View.  My old view related the interpretation-construction distinction to the distinction between ambiguity and vagueness.  For an introduction to that distinction, see Legal Theory Lexicon 051: Vagueness and Ambiguity.  A word, phrase, or legal provision is ambiguous when it has two or more senses (e.g., cool can refer to style or temperature), but vagueness occurs when a word, phrase, or legal provision admits of borderline cases (i.e., H.L.A. Hart's notion of the core and the penumbra).  When we resolve ambiguity, we attempt to determine the linguistic meaning of a text, and usually this can be done by reference to the context in which the text was written.  (If "cool" appears in the sentence, "It's very cool outside, so I'll bring my coat," the relevant sense of "cool" is likely to be the temperature sense.)  So "interpretation" resolves ambiguity.  Vagueness is different, when a word is vague, the linguistic meaning itself is vague.  Something else is required to create a bright line–in the legal context, that might be a supplementary doctrine that permits application of the vague text to a particular case.  So "construction" is required to give legal effect to a vague legal text.  This suggested that "interpretation" and "construction" were alternative methods or modalities of legal practice–courts either engage in "interpretation" or "construction" but not both.  Simplistically, the view was: ambiguity requires interpretation, but vagueness requires construction.

(2) The New View.  But I came to realize that the old view wasn't quite right.  The key insight is a simple one: we always give "legal effect" to the linguistic meaning of legal texts–whether that meaning is clear, ambiguous, or vague.  Linguistic meaning is one thing, but legal effect is another.  Of course, I felt somewhat silly once I realized this–because, after all, the definition of "construction" in terms of the determination of legal effect had been there all along in the common-law cases–staring me in the face, as it were.  This suggested that the relationship between "interpretation" and "construction" was not that of two alternatives, but instead was better captured as two steps or moments in legal practice: first, determine the linguistic meaning of the legal text (interpretation), and then, second, determine what legal effect that linguistic meaning produces (construction).

(3) The Relationship Between the Old View and the New View.  I also came to realize that the old view could actually be explained in terms of the new view.  It is true that when a legal text is ambiguous, the ambiguity requires interpretation.  That was right.  When a text is ambiguous, the activity of interpretation come to the fore–we focus on linguistic meaning.  Likewise, it is also true that once we have determined that the linguistic meaning is vague, then construction is required.  Drawing a line to resolve the vagueness puts the activity of construction at the center of our attention.  But this is consistent with the idea that both interpretation and construction are always at work.  When we decide cases or engage in practice that is governed by a legal text, we always need to know what the text means and we always engage in the practice of giving the text legal effect.  The seeming plausibility of the alternative methods view was a function of the fact that in some contexts "interpretation" comes to fore and "construction" goes unnoticed–and vice versa.

In other words, the new view provides a fuller and deeper account of the distinction which accounts for and subsumes the old view.  The old view described a surface manifestation of the interpretation-construction distinction; the new view describes its deep structure.

The Relationship Between "Construction" and "Vagueness"

This brings me to the criticisms that Michael Rappaport and John McGinnis have offered in connection with the interpretation-construction distinction.  I highly recommend their article in the Northwestern University Law Review, which is also available on SSRN: Original Methods Originalism.  One of the points that they make in that article is reflected in the following comment made by Mike Rappaport in his latest post:

Larry (and Randy) drew a distinction between interpretation as determining semantic meaning and construction as giving nonsemantic effect to ambiguity and vagueness. (Call this the resolving vagueness sense of construction). I largely rejected this distinction, because I believe that ambiguity and vagueness can be resolved by interpretation to yield a semantic meaning (using the original methods of the constitutional enactors). Thus, I reject this sense of construction. My rejection is not based merely on a rejection of using normative matters to resolve vagueness. Rather, it is because I believe that interpretation can resolve these matters semantically. There is no need for the resolving vagueness sense of construction.

In other words, Rappaport and McGinnis believe that the original methods of constitutional interpretation permit the complete resolution of each and every instance of either ambiguity or vagueness in the constitutional text, and therefore, they believe that "interpretation" (in these sense of a determination of linguistic meaning or semantic content) can "do all the work."

I have four separate and distinct points to make about this position and its relationship to the interpretation-construction distinction:

(1) I agree with Rappaport that it is an error to think that there is a distinct "vagueness sense of construction."  A caveat first: the word "construction" has different senses.  Some of these senses have nothing to do with theories of legal interpretation and construction, and the word "construction" has a familiar sense in which it is synonymous with "interpretation."  But the technical sense of "construction" that is deployed in theorizing the interpretation-construction distinction does not make its linguistic meaning or semantic sense equivalent to "resolving vagueness."  "Construction" (in the relevant sense) means something like "the determination of legal effect."  To the extent that I have given the impression that the sense or linguistic meaning of the term "construction" is conceptually equivalent to "resolving vagueness," that was in error, and I hope that I have now made my position explicit and clear.

(2) From the fact that there is no distinct "vagueness sense of construction," it does not follow (conceptually or logically) that there is no connection between the "legal effect sense of construction" and vagueness.  If there are legal texts with vague linguistic meanings (i.e. legal texts with semantic content that creates borderline cases or penumbras), then it follows that "interpretation" cannot do all the work necessary to apply the text to particular case.  Something more than "strict construction" would be required.  I believe that, in fact, there are legal texts with vague linguistic meaning, but this is an empirical claim–which must be supported by evidence.

(3) Based on Rappaport's most recent posts and remarks that both he and John McGinnis have made, I believe that they fail to fully appreciate the implications of the distinction between vagueness and ambiguity and its relationship to construction.  In the passage quote above, Rappaport writes, "Larry (and Randy) drew a distinction between interpretation as determining semantic meaning and construction as giving nonsemantic effect to ambiguity and vagueness"  Of course, this is only a blog post and Rappaport's formulation may not reflect the point he was trying to make, but a similar point occurs in their article.  Assuming that Rappaport meant what he said, I have several reactions to this point:

• If Rappaport means to refer to the paper that I gave at the San Diego conference and to Barnett's oral remarks at the conference, then I believe his assertion is simply in error.  My remarks (both oral and written) explicitly rejected the notion that the interpretation-construction distinction is based on the ambiguity-vagueness distinction.  Indeed, debunking that notion was a major purpose of my remarks.  Barnett's oral remarks at the conference explicitly tied "interpretation" to linguist meaning and construction to "application" and "legal effect."  If Rappaport means to refer to the position that I took in the November 2008 draft of "Semantic Originalism," then his point is correct, but it would have been more helpful had he said that explicitly.

• It has never been my position that ambiguity always or usually requires construction.  Quite the opposite, my position has always been that ambiguity is usually or almost always resolved by interpretation.  When a legal provision is ambiguous as between two possible linguistic meanings, we can usually resolve the ambiguity by reference to the context of utterance.  I do think that it is possible that there are cases of what I call "irreducible ambiguity"–for example, in cases in which deliberately ambiguous language was chosen and the public meaning of the provision would have been understood as ambiguous.  If there are such cases (I am agnostic about their actual existence), then construction would be required in order to resolve the irreducible ambiguity.  I believe that Rappaport mischaracterized my position in the passage quoted above and elsewhere–although I am sure that this is accidental and not deliberate.

• The distinction between ambiguity and vagueness is important in the context of the interpretation-construction distinction, because it goes to the limits of interpretation.  If linguistic meaning is vague, then borderline cases are built into the semantic content of the legal provision: there is no more work that interpretation can do and construction is required.  Of course, Rappaport and McGuiness believe that there are no vague provisions in the Constitution–I will say more about that in a subsequent post in this series.  But I want to emphasize the point that I just made: if the linguistic meaning of a legal provision is vague, then underdeterminacy in borderline cases is built into the meaning of the provision.  This means that there is quite a lot at stake in the question, "Are there any provisions of the Constitution that are vague?"

• In sum, "vagueness" and "ambiguity" need to be distinguished.  Ordinarily ambiguity can be resolved by interpretation, but vagueness cannot be so resolved, because it is the linguistic meaning of some words, phrases, and legal provisions that can be vague.  Hence, construction would be required if there are constitutional provisions that produce "borderline cases" or have a "penumbra" in the sense of H.L.A. Hart's distinction between the core and penumbra.

(4) I believe that Rappaport can and should accept the interpretation-construction distinction as it is now formulated and has long been formulated in the common law as a distinction between discerning linguistic meaning and determining legal effects.  Assuming arguendo that Rappaport and McGinnis are correct and the Constitution does not contain any contain any vague provisions, their position could then be reformulated as follows:

The proper method of constitutional construction is "strict construction," which requires that each and every rule of constitutional law be equivalent to the linguistic meaning of the corresponding portion of the constitutional text.  The legal content of constitutional doctrine can and should be equivalent to the semantic content of the constitutional text, because the correct theory of constitutional interpretation (original methods originalism) eliminates all vagueness or ambiguity.

Of course, this principle of "strict construction" has the consequence that all of the hard work of constitutional practice is done at the interpretation stage–McGinnis and Rappaport argue that interpretation can resolve all ambiguities and eliminate all vagueness.

Mike Rappaort's original post partially agrees on this point:

[Solum] says interpretation is determining meaning and construction is given legal effect to a provision.  He even says that when one gives effect to an unambigous provision, one is engaged in (a kind of) construction.  Under this view, my approach may be consistent with Larry’s as well.  I believe virtually all cases can be decided based on interpretation. And giving effect to these interpretations can be deemed construction.  If that is construction, then I have little reason to reject it.  Of course, most people think of construction differently.

I agree with Rappaport that he has "little reason to reject" construction–if the relevant sense of "construction" is giving legal effect to the linguistic meaning of a legal text.  I am not quite sure what to make of his last point–"most people think of construction differently."  The theory of the interpretation-construction distinction that I offered in the paper delivered at the San Diego conference is, in fact, the traditional common-law conception of the distinction.  In that paper, I offered a series of arguments that are designed to show that this conception of the distinction provides the deep structure that can explain the various ways in which the distinction is used.  Rappaport has not disputed any of those claims.  I do not know to whom "most people" refers.  If Rappaport is referring to the group of contemporary constitutional theorists who work with the interpretation-construction distinction, it is a small group, and "most people" could be identified by name.

These four reactions to Rappaport's comment do not directly contest his central claim.  That is, nothing that I have said in this post directly engages (or denies) the claim that interpretation can do all the work, leaving nothing in what I have called "the construction zone."  That is the topic of the next post in this series.

Conclusion

Let me say one more thing, before I close this post.  The point of my remarks at the San Diego conference was not to argue against McGinnis and Rappaport's argument that interpretation can eliminate vagueness and ambiguity.  Quite the contrary, one of my central aims was to show that their position is actually consistent with the interpretation-construction distinction–once that distinction is properly understood.