Stephen Griffin has posted some very perceptive and illuminating comments about Semantic Originalism at Balkinization. In this post, I reply to the first of Griffin's comments. (See also Part One, Part Two Part Three Part Four Part Five Part Six Part Seven Part Eight Part Nine) Quotes from Griffin are in ordinary roman type, indented, and black. My discussion is in blue. Griffin raises a number of points, so I am going to break my reply into several parts, posted over the next few days.
Before discussing Griffin's comments in detail, let me say a few words about the the argument of the article, beginning with two points of clarification:
First, the central aim of semantic originalism is to disentangle the different kinds of claims that play a role in contemporary theoretical debates over originalism. Some claims are nonnormative claims about "meaning." But once we say "meaning," there is potential for misunderstanding, becuase "meaning" is an ambiguous term: I am using "meaning" in the sense of "linguistic meaning" or "semantic content." Other claims in the originalism debate are about the relationship between the semantic content of the constitutional text and the content of consittutional law--these claims are legal claims. In theoretical terms, they are claims about the rule of recognition and the role that it assigns to the constitutional text. And a final set of claims are normative (or claims of political morality): these are claims about what how whether we should respect constitutional law, and whether we shoudl affirm or reject our current legal practices.
Second, Semantic Originalism deploys a technical distinction between "construction" and "interpretation." The basic idea of the distinction is that there are two distinction activities involved in the movement from constitutional text to application. The first activity, "constitutional interpretation," is the determination or discovery of the semantic content or linguistic meaning of the constitutional text. The second activity, "constitutional construction," involves the translation of the semantic content into rules of constitutional law. When the meaning is plain (neither vague nor ambiguous), then the interpretation and construction happen without our even noticing: the constitutional text gives each state two senators, that means two senators, and the legal rule is two senators. But sometimes the text is ambiguous: in the strict sense, "ambiguity" occurs when a word, phrase, or expression has two or more possible meanings. Usually, we resolve ambiguity by interpretation: the context of utterance makes it possible to determine which sense provides the correct understanding of the meaning. Other times, the text is vague: in the strict sense, "vagueness" occurs when there are borderline cases. Usually, the resolution of vagueness requires a construction and cannot be resolved by interpretation. That's because it is usually the case that the semantic content is vague.
With those two distinctions in place, here is a summary of the four central claims of Semantic Originalism:
The fixation thesis claims that the semantic content of each constitutional provision is fixed at the time the provision is framed and ratified: subsequent changes in linguistic practice cannot change the semantic content of an utterance. Here is an example. The constitution refers to "domestic violence." The meaning of this phrase did not change when the contemporary sense of "domestic violence" as "spouse abuse" entered linguistic practice. Instead, the original sense of "domestic violence" as (roughly) "riot, rebellion, or insurrection within a state" is the fixed.
The clause meaning thesis claims that the semantic content of each clause of the Constitution is given by the conventional semantic meaning (or original public meaning) of the text with four modifications:
The first modification is provided by the publicly available context of constitutional utterance: words and phrases that might be ambiguous in isolation can become clear in light of those circumstances of framing and ratification that could be expected to known to interpreters of the Constitution across time.
The second modification is provided by the idea of the division of linguistic labor: some constitutional provisions, such as the natural born citizen clause may be terms of art, the meaning of which are fixed by the usages of experts.
The third modification is provided by the idea of constitutional implicature: the constitution may mean things it does not explicitly say.
The fourth modification is provided by the idea of constitutional stipulations: the constitution brings into being new terms such as House of Representatives and the meaning of these terms is stipulated by the Constitution itself.
The contribution thesis asserts that the semantic content of the Constitution contributes to the law: the most plausible version of the contribution thesis is modest, claiming that the semantic content of the Constitution provides rules of constitutional law, subject to various qualifications. Our constitutional practice provides strong evidence for the modest version of the contribution thesis.
The fidelity thesis asserts that we have good reasons to affirm fidelity to constitutional law: virtuous citizens and officials are disposed to act in accord with the Constitution; right acting citizens and officials obey the constitution in normal circumstances; constitutional conformity produces good consequences. Our public political culture affirms the great value of the rule of law.
Let's get started. Here is Griffin's first comment:
1. Solum provides a theory of how the various clauses (provisions) in the Constitution acquire meaning. But why is meaning limited to clauses? Articles might also have meaning, especially in relationship to one another. So some might view the message of Articles I, II, III as saying there should be three co-equal branches of government. And some view the entire Constitution as communicating a meaning best summarized in the Preamble. Is there a reason to limit meaning to clauses?
Griffin's question is a good one, and it addresses a general issue in the philosophy of language: what is the unit of meaning? One possibility is that the proper unit of meaning is the individual "word," but counter-examples show that this cannot be correct. Most obviously, when words are combined into phrases, the meaning of the individual words may be transformed by their inclusion in the phrase. For example, the words "natural," "born," and "citizen" take on a new meaning in the Constitution when they are combined into the phrase "natural born citizen." This phrase does not mean, citizens whose birth was natural. Instead, it has a technical meaning that is not the same as its concatenated word meaning: roughly, a "natural born citizen" is a citizen whose status as citizen was determined at birth. The word "natural" plays no role in the technical meaning of the phrase.
When we try to determine the meaning of the constitution, we need to identify some unit as the starting point--the object of interpretation. There is an analagous question that philosophers need to answer when they theorize about the meaning of language generally. They ask, what unit of language can be true or false? And the answer they give is, the sentence. The Constitution rarely uses declarative sentences: it is composed of legal rules, and the smallest meaningful unit of a rule is the clause. The idea that we should begin with clauses is very familiar to lawyers and judges. They want to know the meaning of the commerce clause, the appointments clause, the necessary and proper clause, and so forth. Clauses are the appropriate place to start, because they are the unit that can be translated into legal rules.
So far, I haven't really answered Griffin's queston. Why not Articles? Why not the whole constitution? Of course, the whole constitution and each article provides context which can be used to disambiguate the semantic content of the individual clauses--Griffin understands this. His question is about a deeper point--whether the meaning of the Constiution is in some sense "holistic."
Let us use the name constitutional holism for the view that the meaning of the Constitution is the meaning of the whole document or the holistic meaning.
Does holism provide an alternative to clause meaning? The clause meaning thesis asserts that the semantic content of the constitution is a function of the meanings of individual clauses given the publicly available context of constitutional utterance. But what if individual words and phrases cannot be understood in isolation because the Constitution is an organic whole? For example, the phrase “rights . . . retained by the People” in the Ninth Amendment might not be comprehensible without reference to “We the People” in the Preamble: are “the People” individuals or are they a polity? Likewise, the Ninth uses the phrase “the enumeration of certain rights in this Constitution” Gleaning the meaning of this phrase seems to require reference to what is now called “the Bill of Rights,” and once that has been accomplished, the meaning of the phrase “rights . . . retained by the People” may be clarified: the “retained rights” that are not to be denied or disparaged may be of the same type or kind as the “enumerated rights” such the freedom of speech and press, the right to bear arms, the right to due process, and so forth.
Does holistic meaning provide a better account of the semantic content of the constitution than does clause meaning? To get at this question, we first need to identify and then deflate a misleading picture of the relationship between the meaning of individual clauses and the whole Constitution. It might be thought that there are only two alternative positions on the relationship of the whole to the parts when it comes to constitutional meaning. The first alternative might be called clause bound interpretivism, the view that the meaning of each clause must be determined from within the four corners of the clause. The second alternative might be called organic unity holism, the view that meaning only attaches to the whole constitution as an organic unity and that as a consequence individual clauses are not meaningful units of constitutional communication. This picture, which suggests we must choose between clause bound interpretivism and organic unity holism, might be called the all-or-nothing picture: either the constitution’s meaning is all (the whole constitution all at once) or it is nothing (no meaning attaches to the individual clauses by virtue of their relationship to the whole document).
The all-or-nothing picture creates a false dilemma. There is an alternative picture of the relationship between the meaning of individual clauses and the whole Constitution: that picture can be expressed via two theoretical ideas: (1) familiar device of the hermeneutic circle, and (2) the related notion of intratextualism. The idea of hermeneutic circle figured prominently in protestant theological hermeneutics to understand the relationship of the meaning of individual biblical passages to the whole text: the meaning of each individual passage of scripture is gleaned in light of the meaning of the Bible as a whole. As Gadamer puts it, “For it is the whole of scripture that guides the understanding of the individual passage: and again this whole can be reached only through the cumulative understanding of individual passages.” Justice Story's first recommendation for constitutional construction is based on the same notion: “In construing the constitution of the United States, we are, in the first instance, to consider, what are its nature and objects, its scope and design, as apparent from the structure of the instrument, viewed as a whole, and also viewed in its component parts.”
Intratextualism as articulated by Akhil Amar expresses a closely related idea with a different metaphor:
Textual argument as typically practiced today is blinkered (“clause-bound” in Ely's terminology), focusing intently on the words of a given constitutional provision in splendid isolation. By contrast, intratextualism always focuses on at least two clauses and highlights the link between them. Clause-bound textualism paradigmatically stresses what is explicit in the Constitution's text: “See here, it says X!” By contrast, intratextualism paradigmatically stresses what is only implicit in the Constitution's text: “See here, these clauses fit together!” But there is no clause in the Constitution that says, explicitly and in so many words, that the three Vesting Clauses should be construed together, or that the Article III grant of federal question jurisdiction should be read alongside the Article VI Supremacy Clause. Intratextualism simply reads the Constitution as if these implicit linking clauses existed. Clause-bound textualism reads the words of the Constitution in order, tracking the sequence of clauses as they appear in the document itself. By contrast, intratextualism often reads the words of the Constitution in a dramatically different order, placing textually nonadjoining clauses side by side for careful analysis. In effect, intratextualists read a two-dimensional parchment in a three-dimensional way, carefully folding the parchment to bring scattered clauses alongside each other.
Both the idea of the hermeneutic circle and the idea of intratextualism undermine the all-or-nothing picture. Our choices are not limited to clause-bound interpretivism and organic unity holism. The excluded middle is to read individual clauses in the context of the whole Constitution. The clause meaning thesis squarely occupies the excluded middle: it insists that clause meaning is bound by the publicly available context, and the whole of the constitutional text is indisputably part of that! Once the all-or-nothing picture is out of the way, it becomes apparent that contextual clause meaning can be reconciled with a plausible version of holistic meaning.
One final point: organic unity holism is utterly implausible as a theory of semantic content. The whole constitution is not the relevant unit for determining semantic content. It is no accident that when we apply the Constitution our focus is on clauses and the interaction between clauses. The Constitution as an organic unity says both too much and too little. Too much, because the whole Constitution from top to bottom considered a single unit of meaning doesn’t translate into rules of constitutional law: organic unity holism makes the Constitution one long primal scream. Too little, because organic unity prevents us from assigning meaning at the level of particularity required to do the work of constitutional practice: organic unity holism transforms individual clauses into meaningless concatenations of phonemes. In sum, if holistic meaning is construed plausibly (as incorporating the ideas of the hermeneutic circle and intratextualism) then it is absorbed into the theory of clause meaning via the notion of the publicly available context of utterance. But if construed in accord with organic unity holism, holistic meaning is no meaning at all.
In my next post, I will deal with Griffin's second question, which concerns the interpretation of the amendments to the Constitution.