Academic

Stein on the Domestic Violence Clause & the Fixation of Original Meaning (Reposted, With Comments)

By Lawrence Solum

Mark S. Stein (Harvard Law School, Petrie-Flom Center) has posted The Domestic Violence Clause in ?New Originalist? Theory on SSRN. Here is the abstract:

In this brief essay, I address the way in which ?new originalists? Jack Balkin and Lawrence Solum use the Domestic Violence Clause in Article IV, Section 4 of the Constitution to support their theories. Balkin uses the Domestic Violence Clause to provide an example of an impermissible departure from original meaning: surely, he claims, it would be wrong to interpret the constitutional term ?domestic violence? as referring to spousal abuse. Solum uses the same example to support his thesis that the semantic meaning of the Constitution is fixed at the time of origin.

In fact, the constitutional term ?domestic violence? could evolve so that an application of the Domestic Violence Clause to spousal abuse no longer seems absurd. Reflection on this possibility yields some insights about originalism and original expected applications.

And from the paper:

But what if Balkin is not right about the original meaning of ?domestic violence?? Some readers will by now have begun to consider a move often made by ?new originalists? such as Balkin and Solum: drawing a line between original meaning and original expected applications. Maybe the original meaning of the constitutional term ?domestic violence? was not, as suggested by Balkin, ?riots or insurrections;?14 maybe those were merely expected applications of the term. Maybe the original meaning of the constitutional term ?domestic violence? was ?violence internal to a state,? so that spousal abuse was, from the very beginning, embraced within the semantic meaning of the Domestic Violence Clause. Spousal abuse was not initially an expected application of the clause, but a constitutional provision can be given a legal meaning different from its original expected application, without doing violence to the hard core of original semantic meaning.

Highly recommended--a wonderful and illuminating essay.

The discussion that follows draws on my essay, Semantic Originalism: click on the title to download the most recent draft.

Before I respond in some depth to the substance of Stein's discussion of the implications of the point he makes in the paragraph quoted above, I want to say that I agree with his analysis in that paragraph and that it corrects a move that I now see I made too quickly.  In my discussions of the domestic violence clause, I have equated the eighteenth century meaning of the phrase with riots and insurrections, but this equation was done hypothetically.  I still haven't done any serious historical research on the clause, but Stein's hypothesis--that it could have meant something like "violence internal to a state"--seems plausible.  And it also seems plausible that riots and insurrections were something like "original expected applications" of the general category.  Because the distinction between original public meaning (in the semantic sense of meaning) and original expected applications is a key distinction in the New Originalism, I should have been very careful about this point.

Of course, the issue here isn't what the clause actually did mean.  The domestic clause is being used to illustrate a theoretical point--by Stein, Balkin, and me.

This brings me to the main point.  My discussion of the Domestic Violence Clause in Semantic Originalism was offered as part of an argument for the fixation thesis, roughly the claim that the linguistic meaning or semantic content of a constitutional provision is fixed at the time of constitutional utterance, i.e., the time at which the provision is framed and ratified.  So the claim I made was that the linguistic meaning of the phrase "domestic violence" was fixed by linguistic practice in 1787-1789.  During this period, the phrase "domestic violence" was not used to mean "violence within families" or "spousal abuse."

Let us suppose, arguendo, that Stein's hypothesis were true, and that linguistic practice at in the late 18th century was such that the phrase "domestic violence" meant "violence within the territory of a political unit" and that the surrounding context entails that the public meaning of the phrase given the public context of constitutional utterance was "violence within the territory of one of the United States."  What would this entail regarding the linguistic meaning of the clause?  Recall the wording of Article IV, Section 4:

The United States . . . shall protect each of [the States] against Invasion; and on Application of the Legislature, or of the Executive (when the Legislature cannot be convened) against domestic Violence.

Stein disagrees.  Here is the relevant passage:

In my view, the change in legal meaning that Solum hypothesizes contradicts his fixation thesis. The Supreme Court, he assumes, gives the constitutional term ?domestic violence? a new legal meaning, and people acquiesce. But the new legal meaning is based on, and embodies, a new semantic meaning. Therefore, the constitutional term ?domestic violence? has taken on a new semantic meaning. The original semantic meaning may still be relevant to various people (historians, originalists), but the term has a new semantic meaning as well. Solum avoids this conclusion only by defining ?semantic meaning? so as to exclude everything but original semantic meaning. He makes his fixation thesis true by definition ? not a persuasive approach, in my view.

Stein's article is admirably fair and even-handed in almost every respect, but on this particular point, I believe that his characterization of my position is inaccurate and misleading.  I did not argue from a stipulated (or covertly stipulated) definition of linguistic meaning.  Instead, my argument was that our intuitions and considered judgments about linguistic practice support the fixation thesis.  In Semantic Originalism, this argument was developed at length, with many supporting examples and arguments.

Stein's remarks are brief, and he may clarify them in a future draft of what is a work in progress.  But in their current form, they suggest that Stein has made a conceptual mistake by confusing the possible meanings of utterance types with the actual meanings of utterance tokens.  (Note that the type-token distinction is a technical one used by philosophers to express the difference between classes of things or universals (types) and particular instances (tokens).)  Of course, if we consider the phrase-type, "domestic violence," one of its possible meanings today is "violence within the family."  (Everyone agrees upon that.)  And if the Supreme Court were to stipulate that the legal meaning of the phrase in Article IV Section 4 was "violence within a family," then the linguistic meaning of many utterance tokens of "domestic violence" would be "violence within a family."  For example, when the Supreme Court itself used the phrase in the opinion that stipulated the new meaning, the actual linguistic meaning of tokens of the phase "domestic violence" in the opinion would be the stipulated meaning.  Tokens of the phrase in lower court opinions, in newspaper reports, and so forth would also have the stipulated meaning.  This is obviously true, and nothing said in Semantic Originalism is to the contrary: rather, Semantic Originalism clearly acknowledges this fact.

All of that was preliminary to my main point.  The fixation thesis is a claim about the linguistic meaning of particular utterance tokens--the Constitution of 1789 and its amendments.  In the case, of the Domestic Violence Clause, the relevant utterance occurred when the Constitution of 1789 was framed and ratified.  My claim is that the linguistic meaning of the Constitution of 1789 was fixed by linguistic usage at the time it was uttered (framed and ratified).  Of course, the very same words (the utterance type) could be uttered on a different occasion and have a different meaning.  For example, someone might embed the text of the Constitution of 1789 in a science fiction novel set in a world with different linguistic practices--the linguistic meaning of that utterance token would not be fixed by the meaning of the actual Constitution.

This point about the type-token distinction was made in Semantic Originalism.  This was one of many arguments offered in support of the general claim that the linguistic meanings of utterance tokens are fixed at the time of utterance.  Stein has not replied to these arguments, and his claim that my position relies on a stipulated definition of linguistic meaning is not an accurate characterization of my position.

Stein makes another point that is important and which I fully embrace.  He notes that many of the examples used to establish the fixation thesis are "silly"--in the sense that they seem unlikely to be taken seriously in the real world.  Moreover, these examples do not involve cases in which assigning a new meaning to Constitution would be normatively attractive.  For this reason, Stein argues that thought experiments may be biased--pumping intuitions that do not generalize.  I agree.  It is precisely for this reason that my argumentative strategy was to begin with nonlegal examples that illustrate the general claim of linguistic fact in a context in which our normative intuitions (e.g., our beliefs about what the Constitution should say) do not cloud our linguistic intuitions (i.e. our beliefs about how language works in general).  So my discussion begins with the example of the word "deer."  In the twelfth centry, "deer" meant "four footed mammal": a letter written in the twelfth century might have used the word "deer" to refer to a pig or a cow.  Does the fact that modern usage restricts the word deer to a mammal belonging to the family Cervidae change the linguistic meaning of twelfth-century letter?  No, obviously not.

Moreover, I believe that Stein is right when he argues that thought experiments that associate a change in meaning with a result that we strongly prefer on normative grounds might lead some of us to deny the fixation thesis.  But the pull of that kind of thought experiment should not tempt us to commit what is an obvious category mistake.  Theories of linguistic meanings are positive theories--they are theories about how language works.  Arguing for a theory of meaning on the grounds that it produces normatively attractive results in Constitutional cases involves an enormous conceptual error--like arguing that greenhouse gases do not cause global warming because the world would be a better place if physics were different.  The Domestic Violence Clause thought experiment (created by Larry Tribe and introduced to me by Jack Balkin) is designed to illustrate the absurdity of the assertion that the meaning of an utterance token can be altered by subsequent changes in linguistic practice.  But one thought experiment should not be taken as decisive--the possibility of confounding intuitions should be carefully considered.  I believe that the argument in Semantic Originalism does precisely that.

One more point in Stein's essay should be mentioned.  New Originalists, like Jack Balkin, Randy Barnett, Keith Whittington, and myself have argued for a distinction between "interpretation" and "construction."  (Here is the Legal Theory Lexicon entry.)  The core idea of the distinction is that legal practice involves two distinct activities.  The first activity involves the discovery of the linguistic meaning of legal text (a contract, deed, will, statute, or constitution).  The second activity involves the determination of the legal effect of the text.  This distinction is a general one, but New Originalists deploy the distinction to clarify the nature of constitutional practice.  We need terminology to distinguish these two activities.  There is a long tradition in the common law of calling the first activity "interpretation" and the second activity "construction."  (The use of this terminology goes back at least as far as the Nineteenth Century, but I am not aware of any work that definitively traces its historical origins.)

Here is what Stein says about the distinction:

Some originalists, prominently including Solum, have embarked on a campaign to obtain sole possession of the word ?interpretation.? The word ?interpretation,? they believe, should only be used to denote the determination of original meaning, or original semantic meaning.33 But the word ?interpretation,? when attached to the word ?constitution,? carries highly normative connotations about the correct determination of a legal result, connotations that do not accompany the non-tendentious phrase ?determination of original meaning.? The common usage of the word ?interpretation? is such that if an originalist says ?X is the correct interpretation of the Constitution,? it has far more normative force than if he says, non-tendentiously, ?X is the original meaning of the Constitution.? The originalist linguistic imperialists would like to commandeer all the normative force of the word ?interpretation? for themselves; they would invite nonoriginalists, such as myself, to say things like ?The Supreme Court should adopt position X, even though it is contrary to the correct interpretation of the Constitution.? While I admire the audacity of this invitation, I must respectfully decline. As the word ?interpretation? has powerful connotations about the bottom-line determination of legal results, I will continue to use that word, in the phrase ?constitutional interpretation,? to denote the bottom-line determination of legal results.

I do not believe that this remark accurately characterizes my writings or the writings of other new originalists.  Consider the following passage from Semantic Originalism:

For the purposes of this Article, the phrases ?constitutional interpretation? and ?constitutional construction? have stipulated meanings. This need not be the case for other purposes. A historian of constitutional ideas might seek to understand the distinction as it has developed in constitutional practice or theory. A philosopher of law might find a deep connection between interpretation and semantics, on the one hand, and construction and pragmatics, on the other. A constitutional practitioner might find doctrinal significance in the distinction. But for our purposes, all that is necessary is that the distinction be made with precision and that the stipulated definitions be employed consistently.

I do not believe that this passage is consistent with the charge that I have engaged in "a campaign to obtain sole possession of the word 'interpretation.'"  In this passage, and in numerous others, I have done my best to make it clear that the terminology itself does not matter.  I believe that the choice of terminology is fair given the long historical pedigree of this usage, and the fact that this terminology has become standard in contemporary theoretical discourse about originalism, but I have no objection to other theorists using different terminology to mark the same distinction.

The terminology is not important, but the distinction itself is required to avoid conceptual confusion.  There is a real difference between the two activities.  Finding linguistic meaning is not the same activity as the construction of legal doctrine.  A theoretical vocabulary that fails to distinguish these two activities blurs a real difference.  There is a historical story to be told about the connection between the rise of legal realism and the blurring of the interpretation-construction distinction.  My best guess is that many realists sought to erase the distinction, but I would guess that they were motivated by their belief that the theoretical basis for the distinction was unsound.  I have no reason to believe they were trying to take unfair advantage of the normative resonance of the word "interpretation."

One final point.  In footnote 22, Stein makes the following claim accompanied by two citations:

While originalists differ among themselves in their respect for precedent, I believe it is fair to say that they are, on average, less respectful of precedent than nonoriginalists. For the views of the ?new originalists? Barnett and Solum, see Randy E. Barnett, Trumping Precedent With Original Meaning: Not As Radical As It Sounds, 22 CONST. COMMENT. 257 (2005); and Lawrence B. Solum, The Supreme Court in Bondage: Constitutional Stare Decisis, Legal Formalism, and the Future of Unenumerated Rights, 9 U. PA. J. CONST. L. 155 (2006).

I am sure that it was unintentional, but some readers might draw the conclusion that my article, "The Supreme Court in Bondage" offers a view that is "less respectful of precedent than nonoriginalists."  Those who have read the article will know that is not the case.  In fact, I believe that my view--that the Supreme Court is bound by its own precedents--affords the greatest weight to precedent of any contemporary legal theorist.

Read Stein's marvelous essay.

Full post as published by Legal Theory Blog on May 29, 2009 (boomark / email).